Renal Physiology

Human marrow-derived mesenchymal stromal cells decrease cisplatin renotoxicity in vitro and in vivo and enhance survival of mice post-intraperitoneal injection

Nicoletta Eliopoulos, Jing Zhao, Manaf Bouchentouf, Kathy Forner, Elena Birman, Shala Yuan, Marie-Noelle Boivin, Daniel Martineau


Acute kidney injury (AKI) can occur from the toxic side-effects of chemotherapeutic agents such as cisplatin. Bone marrow-derived mesenchymal stromal cells (MSCs) have demonstrated wide therapeutic potential often due to beneficial factors they secrete. The goal of this investigation was to evaluate in vitro the effect of human MSCs (hMSCs) secretome on cisplatin-treated human kidney cells, and in vivo the consequence of hMSCs intraperitoneal (ip) implantation in mice with AKI. Our results revealed that hMSCs-conditioned media improved survival of HK-2 human proximal tubular cells exposed to cisplatin in vitro. This enhanced survival was linked to increased expression of phosphorylated Akt (Ser473) and was reduced by a VEGF-neutralizing antibody. In vivo testing of these hMSCs established that ip administration in NOD-SCID mice decreased cisplatin-induced kidney function impairment, as demonstrated by lower blood urea nitrogen levels and higher survival. In addition, blood phosphorous and amylase levels were also significantly decreased. Moreover, hMSCs reduced the plasma levels of several inflammatory cytokines/chemokines. Immunohistochemical examination of kidneys showed less apoptotic and more proliferating cells. Furthermore, PCR indicated the presence of hMSCs in mouse kidneys, which also showed enhanced expression of phosphorylated Akt. In conclusion, our study reveals that hMSCs can exert prosurvival effects on renal cells in vitro and in vivo, suggests a paracrine contribution for kidney protective abilities of hMSCs delivered ip, and supports their clinical potential in AKI.

  • acute kidney injury
  • cisplatin
  • cell therapy
  • kidney protection/repair

acute kidney injury (AKI), a common and grave illness with a high mortality rate, is caused by toxic and/or ischemic insult from chemotherapy, antibiotics, or shock occurring from infection or major surgery (2, 26, 50). AKI can lead to dysfunction and apoptosis/necrosis of renal tubular epithelial cells, in addition to a loss of renal endothelial cells (50). Cell therapy may improve the treatment of AKI, as various types of stem cells have shown potential for renal protection and repair (5, 26, 50, 56). Among these are bone marrow-derived mesenchymal stem cells, also commonly referred to as mesenchymal stromal cells (MSCs), cells with promise for many tissue repair and regenerative medicine applications, including treatment of AKI (4, 5, 43, 46, 53, 55, 56, 65).

Many studies using MSCs have shown that damages to renal tubular as well as to renal endothelial cells can be prevented and/or repaired by MSCs. In rodent models of AKI caused by ischemia-reperfusion (I/R), or chemically induced such as through the use of cisplatin, rodent MSCs have been found to be beneficial through their ability to differentiate into renal tubular cells, to differentiate into renal endothelial cells, but most significantly through their ability to secrete beneficial factors in response to tissue injury (3, 5, 20, 21, 28, 37, 44, 50, 56, 62, 63). These factors secreted by MSCs can engender antiapoptotic, anti-inflammatory, mitogenic, and angiogenic actions on injured tissue (3, 7, 17, 28, 62, 63).

For instance, among these investigations, murine MSCs delivered intravenously (iv) or intraperitoneally (ip) in a mouse model of cisplatin-induced AKI were noted to lower the apoptosis and increase the proliferation of renal tubular cells of recipient mice (3). A similarly beneficial outcome was also seen following the ip injections of conditioned media from these MSCs and led to the proposal of a mainly endocrine effect exerted by these cells due to their secreted factors (3). Among these factors, the implication of IGF-1 was demonstrated in a study where small interfering-RNA (siRNA) was utilized to knock down the expression of this secreted protein in murine MSCs, thereby decreasing the renoprotective action of these cells in mice with cisplatin-induced AKI (28). Furthermore, the involvement of VEGF was shown in another investigation where murine MSCs with siRNA gene-silenced expression of VEGF demonstrated decreased effectiveness in mice with I/R-induced AKI (I/R AKI) (61), an observation also made in rats with I/R AKI (64).

Although there are many reported studies utilizing mouse- or rat-derived MSCs, only one investigation has been published to date testing human MSCs (hMSCs) in a model of AKI. In the recent study by Morigi et al. (45), hMSCs were delivered by iv injection in NOD-SCID mice with cisplatin-induced AKI and found to reduce mortality, and be protective functionally by lowering blood urea nitrogen (BUN) levels, as well as structurally by preventing damage to tubular epithelium and to peritubular microvessels. A phase I clinical trial is evaluating the infusion of hMSCs for prevention and treatment of patients at increased risk for AKI following cardiac surgery (17).

These promising findings with hMSCs, and their clinical interest, indicate that additional experiments will be important to investigate these cells as well as their secretome to further understand and support their potential for use in patients. The objective therefore of our present study was to test the conditioned media of hMSCs on human kidney cells in vitro and the ip delivery of hMSCs in NOD-SCID mice with cisplatin-induced AKI to assess the in vivo outcomes. In brief, we noted that the conditioned media from hMSCs enhanced the survival of human kidney cells exposed to cisplatin in vitro and that the ip administration of hMSCs provoked renoprotective and prosurvival effects in vivo in NOD-SCID mice with AKI.


Isolation, culture, and phenotypic analysis of human bone marrow-derived MSCs.

Bone marrow was obtained from a 64-year-old man at the Jewish General Hospital (Montreal, QC) under guidelines approved by the hospital's ethics committee. The mononuclear cells were harvested following Ficoll-Paque PLUS (Amersham Biosciences, Fairfield, CT) density gradient and plated in tissue culture flasks in complete media, i.e., α-MEM media (Invitrogen-GIBCO, Carlsbad, CA) supplemented with 10% heat-inactivated FBS (Wisent, St.-Bruno, QC), 100 U/ml penicillin, 100 μg/ml streptomycin (Pen/Strep; Wisent), and 2 mM l-glutamine (GIBCO) and placed in a humidified 37°C incubator with 5% CO2. Media was changed twice per week, and the adherent cells were passaged by 1:3 dilution once a week until passage 6, when a portion of the resulting hMSCs were prepared for phenotypic analysis by flow cytometry. Briefly, cells were recovered and stained with anti-human CD31 PE clone WM59, CD44 APC clone G44–26, CD45 biot clone HI30, CD73 PE clone AD2, CD90 biot clone 5E10 (BD Pharmascience, Mississauga, ON), and CD105 FITC clone 8E11 (Millipore-Chemicon, Billerica, MA). All antibodies were used at a 1:50 dilution and incubated with the hMSCs for 1 h at 4°C in the dark. Cells were then washed with 3% FBS in PBS. The hMSCs incubated with the biotin-conjugated antibodies were stained with the secondary antibodies (streptavidin-PE, dilution 1:400) for 30 min. Isotypic control analyses were conducted in parallel. All cells were washed with 3% FBS in PBS, fixed with 1% paraformaldehyde, and analyzed using a fluorescence-activated cell sorter (FACS) Calibur flow cytometer (BD Immunocytometry Systems, San Jose, CA), and data analysis was conducted with FCSExpress (De Novo Software, Los Angeles, CA).

Analysis of hMSCs-conditioned media.

For conditioned media collection from hMSCs for analysis, the media from hMSCs in a 75-cm2 culture flask was changed for 13 ml fresh complete media and collected 72 h later, at which time cell number was 2 × 106. Any floating cells were removed by centrifugation, and the supernatant (i.e., conditioned media) was frozen at −80°C until later use. A human cytokine antibody array membrane-based kit (RayBiotech, Norcross, GA) was used to analyze the conditioned media from hMSCs. The black spots revealed were analyzed using Scion Image software, version 4.0.2. The intensity of the spots of the resulting membrane was compared with that of the corresponding spots of the control membrane, and levels of expression relative to control determined for VEGF and HGF.

Testing of hMSCs-conditioned media on human kidney cell survival and phosphorylated Akt levels.

For conditioned media collection from hMSCs for in vitro testing, media from passage 6 hMSCs in a 150-mm tissue culture plate was changed for 25 ml fresh complete media and collected when at 60–70% confluency 96 h later. Any floating cells and debris were removed by centrifugation, and the supernatant (i.e., conditioned media) was frozen at −80°C until use in the following in vitro assay. HK-2 human kidney cortex proximal tubular cells (ATCC, Manassas, VA) were plated at 2.8 × 105 cells/well in six-well plates in the ATCC-recommended media [i.e., Keratinocyte Serum-Free Media (K-SFM) media (Invitrogen) supplemented with bovine pituitary extract (0.05 mg/ml) and human recombinant epidermal growth factor (5 ng/ml, Invitrogen)]. The next day, when cell confluency was ∼60%, media was replaced by serum-free media for 7 h to synchronize the cells and then replaced by hMSCs-conditioned media or complete media (i.e., α-MEM media, FBS, Pen/Strep, l-glutamine) in the presence or absence of cisplatin (Mayne Pharma Canada, Montreal, QC) at a final concentration of 12.5 μg/ml, and placed in tissue culture incubator for 42 h. In addition, in some wells, a VEGF-neutralizing antibody (MAB293; R&D Systems, Minneapolis, MN) and/or a HGF-neutralizing antibody (MAB294; R&D Systems) was added, for a final concentration of 550 and 20 ng/ml, respectively. The concentrations of VEGF and HGF in the hMSCs-conditioned media were quantified using ELISAs specific for human VEGF or human HGF (R&D Systems).

Subsequently, the cells, both adherent and floating, were recovered and analyzed following annexin V and propidium iodide (PI) staining by flow cytometry. Briefly for this, cells were pelleted, resuspended in 100 μl 1× binding buffer (BD Biosciences Pharmingen), and then 5 μl annexin V FITC and 10 μl PI (BD Biosciences Pharmingen) were added. After 15-min incubation at room temperature in the dark, cells were pelleted and resuspended in 400 μl 1× binding buffer and analyzed within 1 h with the FACSCalibur flow cytometer FCSExpress software. This experiment was conducted four separate times.

Moreover, for Western blot analysis, HK-2 human kidney cells were plated at 1.63 × 106 cells/dish in 10-cm tissue culture dishes and exposed the next day, after 7-h serum-starvation as indicated above, to hMSCs-conditioned media or to complete media, with or without cisplatin at 12.5 μg/ml, and with or without neutralizing antibodies against VEGF and/or HGF at 550 and 20 ng/ml, respectively. After 42 h, all cells (adherent and floating) were recovered and lysed using a lysis buffer [i.e., 20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM b-glycerophosphate, 1 mM Na3VO4, 1 μg/ml leupeptin, and 1 mM PMSF]. Protein concentrations were determined using a Bio-Rad Protein Assay kit (Bio-Rad, Hercules, CA). Whole cell lysates containing 80 μg total protein were separated by 4–20% Precise Protein Gels (Thermo Scientific, Rockford, IL) and transferred to a nitrocellulose membrane (Bio-Rad Laboratories). The membranes were then incubated for 1 h at room temperature in the solution of TBST [i.e., 10 mM Tris·HCl (pH 8.0), 150 mM NaCl, and 0.05% Tween 20] supplemented with 10% nonfat dry milk. Subsequently, the membranes were probed overnight at 4°C with the appropriate primary antibody, first with phosphorylated Akt (P-Akt; Ser473, no. 4051, 1:1,000)-neutralizing antibody (Cell Signaling Technology, Beverly, MA). The membranes were washed three times for 10 min each with TBS containing 0.1% Tween 20. The bound primary antibodies were visualized by incubating for 1 h with a suitable secondary antibody conjugated with horseradish peroxidase (Bethyl Laboratories, Montgomery, TX). The membranes were washed and developed using Chemiluminescent HRP Substrate (Millipore). Blots were later stripped with Re-Blot Plus Strong Solution (Millipore, Temecula, CA) and reprobed with an anti-β-tubulin (no. 2146, 1:2,000, Cell Signaling Technology) antibody as a loading control. This experiment was conducted three separate times. Band intensities were determined by densitometry, and results from the three distinct experiments were averaged.

Implantation of hMSCs in NOD-SCID mice with cisplatin-induced AKI and analysis of blood.

To generate a mouse model of AKI for testing the hMSCs in vivo, NOD-SCID female mice, 9–11 wk old (The Jackson Laboratory, Bar Harbor, ME), were injected with cisplatin ip at a concentration of 11 mg/kg after a prior 6-h removal of food and water. The following day, hMSCs that had been expanded in 150-mm plates and at passages 6–7 were recovered, pelleted, washed, and resuspended in RPMI media (Invitrogen). These hMSCs were then implanted by ip injection at a final concentration of 5 × 106 cells/NOD-SCID mouse with AKI, in 370 μl RPMI/mouse. A group of AKI mice was similarly implanted with the RPMI, i.e., the vehicle only, at 370 μl RPMI/mouse. Animals were handled under the guidelines promulgated by the Canadian Council on Animal Care.

For long-term experiments, blood was collected from the saphenous vein at several time points following cisplatin administration, and urea nitrogen concentration in plasma was determined by a colorimetric method (Urea Nitrogen Reagent, Colorimetric Method; Teco Diagnostics, Anaheim, CA). Survival of these mice was also determined over time. The experiment was repeated with similar findings, and the results were pooled.

For short-term experiments, mice were killed, blood was collected by cardiac puncture, and kidneys were harvested and placed in formalin for later immunohistochemical analysis. Blood chemistry analysis to assess BUN, phosphorous, alanine aminotransferase, amylase, and creatinine levels was conducted using 100 μl of blood in the Abaxis VetScan VS2 chemistry analyzer (VetNovations, Barrie, ON). The experiment was repeated with similar findings, and the results were pooled.

In addition, plasma was isolated from a second sample of blood collected by cardiac puncture and utilized for the determination of mouse IL-6, IFN-γ, KC, macrophage inflammatory protein (MIP)-2, monocyte chemoattractant protein (MCP)-1, granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), IL-1α, and PDGF plasma levels, using a custom-made Bio-Plex multiplex protein expression assay kit for detecting mouse cytokines/chemokines with the Bio-Plex 200 apparatus according to the user's guide (Bio-Rad Laboratories Canada, Mississauga, ON).

Immunohistochemical analysis of kidney sections.

For assessment of apoptotic and thus caspase-3-expressing cells, as well as of proliferating and thus Ki-67-expressing cells in kidneys of mice, kidneys were retrieved, placed in formalin, paraffin-embedded, and 3-μm sections were prepared and stained. Immunohistochemical detection was performed on a Discovery XT automatic immunostainer (Ventana Medical Systems, Tucson, AZ) for deparaffinization and antigen retrieval with proprietary reagents (Cell Conditioning 1, Ventana Medical Systems). The anti-caspase-3 antibody (1:100 with PBS, Biocare Medical) was incubated for 2 h at room temperature. The anti-Ki-67 antibody (1:200 with PBS, Biocare Medical) was incubated for 60 min at room temperature. This was followed by an incubation with anti-rabbit biotin-conjugated antibody (Jackson ImmunoResearch). Streptavidin-horseradish peroxidase, a 3,3′-diaminobenzidine detection kit, and an Avidin-Biotin Blocking kit were used according to the manufacturer's instructions (Ventana Medical Systems). The sections were counterstained with hematoxylin, and a bluing reagent was applied for postcounterstaining. The resulting stained kidney sections were scanned using the NanoZoomer Virtual Microscopy System (Hamamatsu). Images of three different areas per kidney section were processed using the Paint Shop Pro software (version 12.0) using a preprogrammed script and then transferred to Scion Image software (version 4.0.2). Objects of interest (caspase-3- or Ki-67-stained areas) were discriminated from surrounding background using the thresholding mode. In this mode, pixels equal to or greater than threshold level were displayed in black, and all other pixels were displayed in white (background). Evaluation of stained cells was performed by counting the number of black spots.

PCR and Western blot analysis of kidney sections.

Genomic DNA was isolated from pieces of retrieved mouse kidneys using an AllPrep DNA/RNA/Protein Mini Kit (Qiagen, Mississauga, ON). DNA samples of 125 ng were used to amplify a 1,171-base pair (bp) fragment of a human chromosome 17-specific α-satellite using the following primers: forward 5′-ACACTCTTTTTGCAGGATCTA-3′ and reverse 5′-AGCAATGTGAAACTCTGGGA-3′ as published (60). The HotStar Taq DNA polymerase system was used under the following conditions: 95°C for 15 min (1 cycle), 94°C for 1 min, 61.5°C for 1 min, 72°C for 2 min (45 cycles), and 72°C for 7 min.

Total protein was also extracted from pieces of harvested mouse kidneys. Briefly, kidney tissue samples of 30 mg were homogenized in 300 μl of ice-cold lysis buffer [i.e., 1% SDS, 20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM b-glycerophosphate, 1 mM Na3VO4, 1 μg/ml leupeptin, and 1 mM PMSF]. Protein concentrations were determined using a Pierce BCA Protein Assay Kit (Thermo Scientific, Rockford, IL). Lysates of 100 μg of protein were fractionated using 4–20% Precise Protein Gels (Thermo Scientific) and transferred to a nitrocellulose membrane (Bio-Rad Laboratories). The membrane was blocked for 1 h at room temperature with Tris-buffered saline (TBS) containing 10% nonfat dry milk and 0.05% Tween 20 and then probed overnight at 4°C with the primary antibody against P-Akt (Ser473; 1:1,000, Cell Signaling Technology). The bound antibody was visualized with a suitable secondary antibody conjugated with horseradish peroxidase using Chemiluminescent HRP Substrate (Millipore). The blot was stripped with Re-Blot Plus Strong Solution (Millipore) and reprobed with an antibody against β-tubulin (1:2,000, Cell Signaling Technology) as a loading control.

Statistical analysis.

A t-test was used to compare the various groups, and P ≤ 0.05 was considered significant. For survival experiments, a log-rank test was conducted, and P ≤ 0.05 was deemed significant.


Phenotypic analysis of hMSCs by flow cytometry.

To demonstrate that our preparation of hMSCs express the cell surface antigens characteristic of hMSCs, flow cytometry analysis was performed on these cells at passage 6 and revealed the expected positive expression of CD44, CD73, CD90, and CD105 and absence of CD31 and CD45 expression, as shown in Fig. 1A.

Fig. 1.

Flow cytometry characterization of human mesenchymal stromal cells (hMSCs) and analysis of conditioned media. Flow cytometry analysis was performed on culture-expanded primary hMSCs to assess the expression of cell surface antigens CD31, CD44, CD45, CD73, CD90, and CD105, as described in materials and methods. The solid line represents the specific antibody, whereas the dashed line indicates the isotype control (A). Conditioned media from hMSCs was analyzed using a human cytokine antibody array membrane-based kit. The intensity of the spots on the resulting membrane was compared with that of the corresponding spots on the control membrane where control media was used. Levels of expression of VEGF and HGF were calculated relative to the control membrane (B).

Analysis of hMSCs-conditioned media.

To investigate potential reasons for any beneficial in vitro and/or in vivo effects of hMSCs, conditioned media from these cells was analyzed by antibody array. Several proteins were detected with high expression levels relative to control i.e., to complete media. These include VEGF and HGF, with levels that were 686 and 276 U2, respectively, relative to control (Fig. 1B). Specific ELISAs revealed that human VEGF and human HGF were present at concentrations of 5.5 and 1 ng/ml, respectively, in the hMSCs-conditioned media to be tested on kidney cells with/without anti-VEGF and/or anti-HGF neutralizing antibodies.

Effect of hMSCs-conditioned media on human kidney cell survival in vitro.

To determine any effect of hMSCs-conditioned media on cisplatin-induced toxicity to human kidney cells, HK-2 human proximal tubular cells were exposed to cisplatin in the presence or absence of hMSCs-conditioned media for 42 h in vitro. As seen in Fig. 2A, hMSCs-conditioned media, compared with complete media, reduced cisplatin-induced kidney cell death and thereby led to significantly improved survival, as was ascertained by flow cytometry analysis following annexin V and PI staining. This figure shows 74.6 ± 5.03% (means ± SE) survival of cisplatin-treated HK-2 cells when these cells were also exposed to hMSCs- conditioned media, in contrast to 48.5 ± 2.42% survival when complete media only (i.e., not exposed to hMSCs) was used instead (P = 0.018, t-test). Survival of cisplatin-treated HK-2 cells when exposed to hMSCs-conditioned media in the presence of an anti-VEGF antibody was 67.6 ± 3.93%, a value significantly lower than that noted in the absence of this antibody (P = 0.020) (Fig. 2A). When cisplatin-treated HK-2 cells were exposed to hMSCs-conditioned media and an anti-HGF antibody, their survival was 72.7 ± 3.11%, not significantly different than that noted without the antibody (Fig. 2A). In the presence of both the anti-VEGF and the anti-HGF antibodies, survival of cisplatin-treated HK-2 cells exposed to hMSCs-conditioned media was 67.5 ± 5.18%, a value similar to that obtained with the anti-VEGF antibody and also significantly different from survival without antibody use (P = 0.014) (Fig. 2A).

Fig. 2.

Effect of hMSCs-conditioned media on HK-2 human kidney cells. HK-2 cells were exposed to cisplatin with/without hMSCs-conditioned media (hMSCs CM), with/without a neutralizing anti-VEGF antibody (hVEGF Ab), and/or a neutralizing anti-HGF antibody (hHGF Ab). Cell survival was determined 42 h later by flow cytometry following annexin V and propidium iodide (PI) staining (n = 4, means ± SE; A). Phosphorylated Akt (P-Akt; Ser473) expression as well as β-tubulin levels were measured in cell lysates by Western blot analysis (n = 3, means ± SE; C), with a representative Western blot shown (B).

Furthermore, to help elucidate a contributing mechanism for this increased survival of cisplatin-exposed human kidney cells seen with the use of hMSCs-conditioned media, lysates from the HK-2 human kidney cells were analyzed by Western blotting for expression levels of P-Akt (Ser473). A representative example is shown in Fig. 2B. Results as shown in Fig. 2C indicate that in cisplatin-treated HK-2 cells, the hMSCs-conditioned media led to a 2.27 ± 0.46-fold increase in P-Akt expression compared with control, i.e., cisplatin only (with culture media only, second column vs. first). Addition of an anti-VEGF antibody to the hMSCs-conditioned media reduced this increased P-Akt expression to only 1.04 ± 0.09-fold. Instead, when an anti-HGF antibody was added to the hMSCs-conditioned media, the P-Akt level in cisplatin-treated HK-2 cells was not reduced as much, as it was 1.66 ± 0.29-fold higher than control. The concomitant use of both the anti-VEGF and anti-HGF antibodies led to a 1.16 ± 0.15-fold increase over control, a level more similar to use of anti-VEGF antibody (Fig. 2C).

Effect of hMSCs on plasma urea nitrogen and survival of cisplatin-treated mice.

To demonstrate any renoprotective effect of hMSCs in vivo, these cells were injected ip in NOD-SCID mice that had been administered cisplatin the previous day to induce AKI. Blood was collected to assess renal function, and mice were followed over time to also determine survival. We had previously determined that BUN levels were highest at 5 days postcisplatin in NOD-SCID mice. Urea nitrogen levels measured in the plasma of blood collected 5 days postcisplatin, corresponding to 4 days post-hMSCs, were significantly lower, i.e., 82.5 ± 18.6 compared with 138 ± 16.0 mg/dl in mice that received the vehicle only (P = 0.03, t-test, means ± SE, n = 15–16), as shown in Fig. 3A. By day 8 following cisplatin administration, the BUN levels in the surviving mice had returned to baseline (Fig. 3B).

Fig. 3.

Effect of hMSCs on blood urea nitrogen (BUN) and survival of cisplatin-treated mice. NOD-SCID mice were injected intraperitoneally (ip) with hMSCs (■, n = 15) or vehicle (▴, n = 25) at day 1 postcisplatin, and blood was collected for BUN determination at day 5 (individual mice; bars indicate means ± SE; A), and at later time points (means ± SE; B). Survival of mice over time was also determined (C).

Moreover, hMSCs significantly enhanced survival of mice with cisplatin-induced AKI. More specifically, as revealed in Fig. 3C, 47% of cisplatin-treated mice implanted with hMSCs were alive for over 15 days, as significantly opposed to only 10% of those that received the vehicle only (P = 0.04, log-rank test). Mice that died usually did so in the first week following cisplatin.

Chemistry and cytokine/chemokine analysis of blood from euthanized mice.

To be able to obtain higher blood volumes for additional blood chemistry analysis, in addition to kidneys for immunohistochemical analysis, the implantations were repeated another two times, and mice were killed at day 5 for harvesting of blood and kidneys. As shown in Fig. 4, blood chemistry analysis confirmed significantly reduced levels of BUN, i.e., 70.8 ± 10.4 vs. 136 ± 29.4 mg/dl (P = 0.03, t-test) in hMSCs- vs. vehicle only-implanted mice, respectively, and in addition, significantly lower phosphorous (2.96 ± 0.19 vs. 4.16 ± 0.57 mmol/l, P = 0.04) and amylase (1,808 ± 106 vs. 2,288 ± 199 Ul/l, P = 0.03) concentrations, as well as, notably but not significantly, decreased alanine aminotransferase (94.1 ± 15.0 vs. 133 ± 22.2 U/l, P = 0.15) and creatinine (68.5 ± 10.3 vs. 93.1 ± 17.6 U/l, P = 0.22) levels (Fig. 4).

Fig. 4.

Chemistry analysis of blood from mice. NOD-SCID mice received ip hMSCs (filled bars) or vehicle (open bars) 1 day postcisplatin, then killed 4 days later, and blood was collected and analyzed for BUN, phosphorous, amylase, alanine aminotransferase, and creatinine levels (AE). Values are means ± SE; n = 14–17.

To analyze and compare plasma concentrations of several mouse cytokines/chemokines of cisplatin-treated mice injected ip with hMSCs or with vehicle, and quantify their difference with regard to levels in normal mice, a Bio-Plex Multiplex protein expression analysis was performed. This analysis revealed the cisplatin-induced rise of several mouse cytokines/chemokines which were significantly lowered by hMSCs. More specifically, as seen in Fig. 5, hMSCs, compared with vehicle only, significantly lowered the cisplatin-induced fold-increase in plasma MIP-2 (2.75 ± 0.11 vs. 8.13 ± 3.02-fold higher than normal mouse levels, P = 0.05, t-test); G-CSF (1.55 ± 0.18 vs. 7.54 ± 3.35-fold, P = 0.05); KC (4.61 ± 0.27 vs. 7.30 ± 1.36-fold, P = 0.04); IL-α (1.48 ± 0.12 vs. 3.46 ± 0.85-fold, P = 0.01); MCP-1 (1.57 ± 0.05 vs. 2.73 ± 0.54-fold, P = 0.02); PDGF (0.96 ± 0.17 vs. 2.44 ± 0.42 fold, P = 0.002); IFN-γ (1.19 ± 0.12 vs. 1.78 ± 0.21-fold, P = 0.02); GM-CSF (1.08 ± 0.05 vs. 1.45 ± 0.16-fold, P = 0.02); and IL-6 (0.92 ± 0.05 vs. 1.37 ± 0.16-fold, P = 0.01).

Fig. 5.

Cytokine analysis of blood from mice. Blood was collected from NOD-SCID mice 4 days post ip administration of hMSCs (filled bars) or vehicle (open bars), and corresponding to 5 days postcisplatin. Plasma cytokine/chemokine levels were determined, and results are represented as fold-increase over normal mouse levels. Values are means ± SE; n = 6–10.

Immunohistochemical analysis of kidneys.

To assess the amounts of apoptotic and of proliferating cells in the kidneys of cisplatin-treated mice injected ip with hMSCs or with vehicle, kidney sections were stained for detection of caspase-3- and of Ki-67-expressing cells, respectively. As shown in Fig. 6, A and B, hMSCs reduced the amount of apoptotic cells in the kidneys of cisplatin-treated mice. More specifically, there were 2.29 ± 0.13% caspase-3-positive cells/mm2 area in the kidneys of hMSCs-implanted mice, significantly lower than 4.75 ± 0.78% detected for vehicle-recipients (P = 0.01, t-test, means ± SE, n = 5–6). Also, as seen in Fig. 6, A and C, hMSCs, compared with vehicle, led to a significantly higher percentage of proliferating cells in kidneys of cisplatin-treated mice, more precisely 6.42 ± 0.87 vs. 2.90 ± 0.58% (P = 0.01).

Fig. 6.

Immunohistochemical analysis of kidneys. NOD-SCID mice were implanted ip with hMSCs or vehicle 1 day postcisplatin, then killed 4 days later, and kidneys were harvested, sectioned, and stained for detection of caspase-3- or Ki-67-expressing cells, indicated in representative images by black arrows (A). The percentage of caspase-3 (B)- or Ki-67-positive cells (C) per mm2 area was assessed. Values are means ± SE (B and C).

PCR and western blot analysis of kidneys.

To determine whether hMSCs engrafted in the kidneys of hMSCs-implanted cisplatin-treated mice, PCR specific for a human chromosome 17-specific α-satellite fragment of 1,171 bp was used on genomic DNA. As shown in Fig. 7A, PCR amplification of the expected fragment was revealed in DNA samples from hMSCs-implanted mice and not in any of the vehicle-only implanted mice tested. These five positive signals were from a cohort of seven mice, two of which did not lead to fragment detection, the same two of which the BUN was not affected or not as reduced by the hMSCs.

Fig. 7.

PCR and Western blot analysis of kidneys. Genomic DNA isolated from kidneys of cisplatin-treated mice implanted with hMSCs or vehicle only were used to amplify a 1,171-base pair (bp) fragment of human chromosome 17-specific α-satellite, and results from individual mice are shown (A). Total protein was also extracted from the same mouse kidneys and used for Western blot detection of P-Akt (Ser473) and of control β-tubulin. Results in individual mice are shown in B, and C reveals P-Akt expression relative to β-tubulin expression. Values are means ± SE; n = 5–6.

Subsequently, to assess the levels of Akt phosphorylation in these same mice, Western blot analysis was conducted and the resulting blot shown in Fig. 7B. Quantitative analysis of the signal intensity of P-Akt (Ser473) compared with internal control β-tubulin showed a significant difference between the two mouse groups. More specifically, as revealed in Fig. 7C, the kidney tissues from hMSCs-implanted mice showed a significantly higher level of P-Akt expression compared with those from vehicle only-implanted mice (i.e., 0.899 ± 0.138 vs. 0.471 ± 0.089, P = 0.025).


MSCs, known for their important role in the bone marrow microenvironment and in hematopoiesis, are also very promising for tissue repair and regenerative medicine purposes (4, 8, 14, 24, 25, 29, 46, 51, 52, 55), such as for AKI treatment as several preclinical studies have indicated (3, 5, 7, 17, 20, 21, 28, 30, 37, 43, 44, 50, 56, 62, 63). Most in vivo studies, however, on MSCs for AKI have used mouse- or rat-derived MSCs. Only one study has been reported thus far utilizing hMSCs for this application (45). It demonstrated that hMSCs delivered by iv injection in cisplatin-treated NOD-SCID mice lead to less renal structural injury, better kidney function, and increased survival (45). We therefore decided to perform additional testing of hMSCs to further define and extend their potential for AKI treatment.

Due to the complex paracrine effects reported with rodent MSCs, we chose to evaluate the conditioned media of hMSCs on the survival of human kidney cells in vitro, an effect not yet investigated. First, we demonstrated that hMSCs-conditioned media contained increased levels of VEGF and HGF compared with media only (Fig. 1B). We then observed that hMSCs-conditioned media significantly reduced the death of HK-2 human proximal tubular cells exposed to cisplatin (Fig. 2A). Murine MSCs-conditioned media has been reported to augment survival of murine proximal tubular cells exposed to cisplatin in vitro (3). Therefore, the in vitro protective capability clearly extends to hMSCs-conditioned media, as we here revealed through our present investigation. In addition, we showed the involvement of the hMSCs-secreted VEGF since the neutralizing anti-VEGF antibody significantly reduced the prosurvival effect of the hMSCs-conditioned media (Fig. 2A). This beneficial effect of the conditioned media was also lowered by an anti-HGF antibody, but not to a significant extent, thus indicating the greater implication in vitro of VEGF among these two growth factors (Fig. 2A).

Furthermore, we demonstrated that hMSCs-conditioned media increased the expression level of phosphorylated serine/threonine protein kinase Akt1, a downstream effector of phosphatidylinositol 3-kinase (PI3K) (59), in human proximal tubular cells exposed to cisplatin (Fig. 2C), thereby reducing their apoptosis and promoting their survival. We revealed the implication of VEGF in this effect, since the use of a neutralizing anti-VEGF antibody lowered the hMSCs-conditioned media-induced rise in P-Akt (Ser473) expression (Fig. 2C). In a recent study on cisplatin-induced AKI, the PI3K-Akt pathway was found to be activated and to be important in decreasing renal tubular cell apoptosis and preserving kidney function (36). Also, hMSCs-conditioned media was reported to activate the PI3K-Akt pathway in primary human aortic endothelial cells exposed to hypoxia in vitro and thus decrease apoptosis and increase their survival (27). These previous reports support our observation of a prosurvival effect of hMSCs on human kidney cells, and the potential therapeutic aptitude of hMSCs for renal protection/repair clinical applications.

Since our above findings propose that hMSCs secrete factors, such as VEGF, with in vitro renoprotective abilities, we speculated that ip implantation of these cells would also provoke advantageous consequences in a mouse model of chemically induced AKI, as had been found recently when hMSCs were instead delivered iv (45). Indeed, we noted that ip delivery of hMSCs in NOD-SCID mice significantly lowered the BUN rise caused by cisplatin (Fig. 3A) and also significantly enhanced the survival of recipient mice (Fig. 3C). Moreover, we noted through immunohistochemical analysis, that hMSCs delivered ip significantly reduced the amount of apoptotic cells and significantly enhanced the quantity of proliferating cells in kidneys of cisplatin-treated mice (Fig. 6), results consistent with the antiapoptotic and mitogenic effects, respectively, of these cells. This protection of kidney function, improved survival, as well as antiapoptotic and mitogenic actions are comparable to the reported results following iv delivery of hMSCs (45). An advantage of ip delivery is that it avoids the substantial trapping of cells within the lungs that has been demonstrated to occur following iv injection of MSCs (57) and thereby removes the restriction in the maximum number of cells that can be implanted in vivo. We injected ip 5 million cells/mouse, a cell amount that leads to immediate death if delivered iv in mice. Cell number is important when envisioning future clinical translation since a high number of cells may be required in humans to achieve an optimal therapeutic outcome.

AKI can exert systemic extrarenal repercussions on other organ systems (15, 16, 32). In fact, the prognosis of patients with AKI is worse when kidney injury is accompanied by AKI-related damage to other organs (16, 19). Cisplatin, a widely used chemotherapeutic agent for the treatment of various malignancies (1), has a major dose-limiting and thus effectiveness-restricting side effect of nephrotoxicity (49). Cisplatin also has toxic effects on other normal tissues and can lead for instance to nausea, vomiting, neurotoxicity, and ototoxicity (49), and its distribution in patients was found through autopsy sample analysis to include various organs, mainly the liver, kidney, and testis, but also the pancreas, lung, muscle, and nerve tissue (22). Therefore, since most investigations on MSCs for cisplatin-induced AKI assess kidney function of recipient mice through measurement of BUN or blood creatinine levels, we decided to perform a broader blood chemistry analysis. We first confirmed less cisplatin-induced renal function impairment as our results showed that hMSCs-implanted mice not only had significantly reduced BUN but also significantly decreased phosphorous blood levels (Fig. 4). With further evaluation, we perceived that hMSCs administration also led to significantly reduced amylase, as well as to lower alanine aminotransferase blood concentrations (Fig. 4), findings suggesting that the hMSCs may have also benefited extrarenal sites such as the liver and pancreas. Liver (35) and pancreas repair (38) are among the reported beneficial applications of MSCs. MSCs, being valuable for protection/repair of various tissues, are thus of additional advantage when used for example against chemically induced toxicities that affect different organs. Using MSCs instead of the conditioned media from MSCs has the advantage that MSCs can be directed to the specific sites of injury by homing signals and consequently improve function and the microenvironment mainly via paracrine effects (6, 7, 53). In fact, we were able to detect human MSCs in kidneys of hMSCs-implanted mice (Fig. 7A). We also noted that in these kidneys where hMSCs were detected, P-Akt expression levels were significantly increased compared with control mice (Fig. 7C). These results, as well as our in vitro observations, are in correlation and indicate that hMSCs injected ip can home to injured kidneys where they can protect against or lessen cisplatin-induced damage, through their secretion for instance of beneficial growth factors such as VEGF, leading to increased kidney expression of pro-survival-activated Akt. In some mouse kidneys, however, (data not shown), we could not detect the hMSCs-specific sequences, and interestingly, these were in mice that responded less well to the hMSCs, i.e., had higher BUN levels. In these mice, the hMSCs may still have homed to sites of injury, but may not have persisted for as long, such as until 5 days postcisplatin, when we tested kidneys of recipient mice.

To investigate the potential systemic changes engendered by cisplatin that may affect kidney and other organ function and microenvironment, we performed more blood analyses. In sum, we observed that hMSCs significantly reduced plasma levels of several mouse cytokines/chemokines that we noted to be increased by cisplatin (Fig. 5). These include IL-6, IFN-γ, KC, MIP-2, MCP-1, G-CSF, GM-CSF, IL-1α, and PDGF (Fig. 5). Our findings are consistent with studies that have shown that AKI can lead to increased serum/plasma concentrations of multiple cytokines/chemokines. For instance, serum levels of TNF-α and IL-18 have been reported to be increased by cisplatin, and levels of IL-6, KC, GM-CSF, IL-1β, IL-12, MIP-2 and MIP-1α by I/R-induced AKI in rodents (13, 23, 33, 41, 54). Other than AKI-provoked damage to the kidneys, the consequent rise in serum IL-6 has been demonstrated to precede and mediate lung damage in mice (34). In patients with AKI, plasma levels of certain proinflammatory cytokines, such as IL-6 and IL-8 (human homolog of mouse KC), have been revealed to be indicative of mortality (58). Interestingly, the MSCs tested in our AKI mice reduced plasma levels of both these cytokines. Furthermore, our hMSCs also significantly lowered plasma PDGF levels in AKI mice, suggesting their potential role in preventing the occurrence of fibrosis, as earlier rodent studies have reported the implication of PDGF in kidney fibrosis (9, 48). Overall, our findings provide supplemental information to published reports, as they reveal various cytokines/chemokines that are increased in plasma of mice administered cisplatin and furthermore that MSCs administration can significantly reduce their cisplatin-induced rise, thereby lessening their detrimental effects on the kidneys and potentially on other affected tissues as well.

MSCs have been suggested to have the rare capability of reacting according to the damaged tissue microenvironment (53). Cross talk between MSCs and the injured tissue microenvironment with its necessities for repair is proposed to influence the types of factors secreted by these cells (53). We thus wondered about the cause of the renal- and extrarenal-protective effect of our hMSCs in vivo and, more specifically, their secreted factors that may be implicated therapeutically. The in vitro analysis of the conditioned media of our hMSCs revealed the expression of various proteins (Fig. 1B), among which, as discussed above, were VEGF and HGF, two growth factors with reported kidney-protective and -reparative abilities (18, 39, 42, 66). For instance, studies in rodents with AKI indicated that treatment with VEGF-121 preserved the structure of renal microvasculature (39) and that administration of HGF can exert renoprotective and antifibrotic effects (31, 40). Our findings of the favorable outcomes from hMSCs are consistent with other studies where MSCs engendered renotropic effects understood to be caused by their expression of growth factors such as VEGF, HGF, and IGF1, due to their mitogenic, antiapoptotic, angiogenic, and/or anti-inflammatory abilities (3, 7, 28, 47, 56, 6164). Interestingly, the secretion of VEGF, HGF, and IGF by hMSCs has been found to be significantly augmented following TNF-α stimulation (67). Therefore, we speculate that although our hMSCs were found to produce HGF and VEGF in vitro, that in vivo these growth factors and others may be released in therapeutically optimal amounts in response to the TNF-α produced consequent to tissue damage from cisplatin toxicity.

In conclusion, we have shown that the conditioned media of hMSCs can reduce cisplatin-induced human kidney cell death in vitro, that hMSCs when injected ip in NOD-SCID mice with cisplatin-induced AKI can lead to less kidney function impairment and higher survival, that hMSCs can be detected in the kidneys, and that factors secreted by hMSCs, such as VEGF, may be involved in the paracrine protective effect of hMSCs on kidneys and other organs. We therefore deduce that promising cellular therapy applications employing hMSCs include AKI. Future investigations may involve the evaluation of gene-enhanced MSCs for treatment of AKI, as we have previously reported for anemia correction in mice with chronic kidney failure and in previous proof-of-concept studies (10–12).


This research was supported by a Special Grant from the Roche Foundation for Anemia Research (RoFAR) awarded to N. Eliopoulos.


No conflicts of interest, financial or otherwise, are declared by the authors.


We thank Dr. Jacques Galipeau (Emory University) for valuable feedback on the manuscript and Moira Francois, B.Sc. (McGill University) for the hMSCs.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
View Abstract