Mesenchymal Stem Cells Are Renotropic, Helping to Repair the Kidney and Improve Function in Acute Renal Failure

Murine Model of ARF

C57BL6/J female or male mice (Charles River Italia s.p.a., Calco, Italy), 2 mo of age at the start of the experiments, were used. Animal care and treatment were conducted in conformity with the institutional guidelines that are in compliance with national (D.L. n.116, G.U., suppl 40, 18 Febbraio 1992, Circolare No. 8, G.U., 14 Luglio 1994) and international laws and policies (EEC Council Directive 86/609, OJL 358, Dec 1987; NIH Guide for the Care and Use of Laboratory Animals, U.S. National Research Council, 1996). Animals were housed in a constant temperature room with a 12-h dark 12-h light cycle and fed a standard diet. ARF was induced in female mice by subcutaneous injection of cis-diaminedichloroplatinum (cisplatin; 12.7 mg/kg; Sigma, St. Louis, MO) dissolved in sterile 0.9% saline solution. The dose of cisplatin was chosen on the basis of preliminary experiments.

Female C57BL6/J mice were divided in two groups and 1 d after cisplatin received an intravenous (i.v.) injection as follows: group 1, saline (n = 27); group 2, mesenchymal stem cells (MSC; 2 × 10⁵ cells) derived from bone marrow of male C57BL6/J mice (n = 23). Mice were killed at different time intervals (4, 7, 11, and 29 d after cisplatin), and blood samples for blood urea nitrogen (BUN) determination were collected at all times until the animals were killed. Two additional groups of female mice on day 1 after cisplatin were treated as follows: group 3 (n = 12) received an i.v. injection of Lin⁻ c-kit^POS hematopoietic stem cells (HSC; 2 × 10⁵ cells) derived from the bone marrow of male mice; group 4 (n = 12) received an injection of saline. Mice were killed at 4, 11, and 29 d after cisplatin, and blood samples for BUN determination were collected. Renal function was assessed as BUN in heparinized blood by the Reflotron test (Roche Diagnostics, Indianapolis, IN). BUN levels that exceeded 30 mg/dl were considered abnormal (normal range in our laboratory: 14 to 29 mg/dl). Kidneys were taken for histologic and ultrastructural analysis and for Y-chromosome examination by in situ hybridization (groups 1, 2, and 3). The expression of Ki-67, a proliferation marker, was also evaluated. Normal mice served as controls. A group of mice that received cisplatin were killed after 1 d for light and electron microscopy studies. At this time, BUN levels were within the normal range. An additional group of mice (n = 12) 1 d after cisplatin received an i.v. injection of MSC immunodepleted of hematopoietic precursor CD45-positive cells (2 × 10⁵ cells). Renal function was evaluated before (day 0) and at day 4 after cisplatin and compared with BUN of mice that received saline (n = 10) or MSC (n = 8). Renal histology was assessed at day 4.

Isolation and Purification of MSC and HSC

Bone marrow (BM) was obtained from 2-mo-old male C57BL6/J mice. Briefly, mice were killed, and femurs and tibias were aseptically removed. BM was flushed from the shaft of the bone with DMEM medium (Sigma) containing 5% FCS (Invitrogen, Paisley, Scotland) plus penicillin/streptomycin (100 U/ml to 0.1 mg/ml; Invitrogen) and then filtered through a 100-μm sterile filter (Falcon) to produce a single-cell suspension. MSC were recovered from BM by their tendency to adhere tightly to plastic culture dishes, as described previously (29). Filtered BM cells were plated in DMEM plus 10% FCS and penicillin-streptomycin (100 U/ml to 0.1 mg/ml) and allowed to adhere for 6 h. Nonadherent cells were then removed. Medium was changed regularly every 3 d; after 2 to 3 wk, adherent cells were detached by trypsin-EDTA (0.5 to 0.2 g/L; Invitrogen), washed with PBS, and used for the in vivo experiments.

In additional experiments, MSC were immunodepleted of CD45-positive cells (CD45⁻ MSC). Briefly, after blocking with PBS that contained 0.5% BSA (Sigma), cells were incubated for 20 min with rat anti-mouse CD45 antibody, 0.2 μg/10⁶ cells (Caltag Laboratories, Burlingame, CA). After washing, cells were incubated with magnetic microbeads coated with goat anti-rat IgG (Miltenyi Biotec, Caldara di Reno, Bologna, Italy), and CD45⁻ MSC were isolated by magnetic cell-sorting separation system.

For isolating HSC (30,31⇓), total BM cells were suspended in PBS that contained 1% FCS and incubated on ice with rat anti-mouse mAb specific for the following lineage markers: CD4 (0.125 μg/10⁶ cells) and CD8 (0.5 μg/10⁶ cells; T lymphocytes), CD45R (0.05 μg/10⁶ cells; B-lymphocytes), CD11b (0.5 μg/10⁶ cells; macrophages), Gr-1 (0.125 μg/10⁶ cells; granulocytes), and Ter-119 (0.5 μg/10⁶ cells; erythrocytes; Caltag Laboratories). After washings, cells were incubated with magnetic microbeads and magnetically cell sorted as described above. The obtained lineage negative cells (Lin⁻), which represented 10% of the total BM cells, were incubated with rat mAb against mouse CD117 (c-kit) PE conjugate (Caltag Laboratories). A wavelength of 488 nm was used to excite c-kit–PE–labeled cells. The c-kit^POS cells were sorted on a FACS Vantage cytofluorimeter (Becton Dickinson, BD Bioscience, Milan, Italy) to obtain purified HSC preparations. For the in vivo experiments, 2 × 10⁵ MSC, CD45⁻ MSC, or Lin⁻ c-kit^POS HSC, isolated from BM of male mice, were washed with PBS, suspended in 500 μl of saline and then injected intravenously into syngeneic female mice.

Fibroblast-Like Colony-Forming Unit Assay

Fibroblast colony growth was evaluated on primary cells grown on tissue culture six-well dishes (32). Total BM-derived cells were plated at the density of 25 × 10⁶ cells/well. After 7 d, the capability of MSC to form fibroblast-like colonies was assessed. Images that showed MSC morphology were acquired by contrast-phase microscope.

In Vitro Differentiation Assays

BM-derived cells obtained by plastic adhesion, as described above, were studied to verify their mesenchymal potential to differentiate toward osteoblasts, adipocytes, and chondroblasts. MSC were grown until confluence, and the growth medium was replaced with the inductive medium consisting of Iscove’s modified Dulbecco’s medium (Invitrogen), 20% FCS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.05 mM β-mercaptoethanol supplemented with specific differentiation reagents as follows.

Immunohistochemistry and In Situ Hybridization

To verify the presence of male mesenchymal or hematopoietic stem cells and to characterize their phenotype in the kidney of female recipients, we stained sections for lectin Lens culinaris binding sites in the brush border of proximal tubuli and further processed for Y-chromosome detection using in situ hybridization. Renal biopsies were fixed in Dubosq-Brazil, routinely processed, and paraffin wax embedded. Slides were dewaxed by xylene, taken through graded alcohols (100, 90, 70, and 50%) to PBS, and incubated with trypsin (0.1% in CaCl₂ 0.1%) at 37°C for 15 min. After washing with PBS, slides were incubated with biotinylated lectin (Vector Laboratories, Burlingame, CA) for 30 min (1:200), followed by treatment with streptavidin-alkaline phosphatase (1:50). Vector Red substrate (Vector Laboratories) was then applied for 15 min at room temperature. After washing in PBS, the slides were processed using the in situ hybridization protocol as described previously (21). Sections were washed again in PBS and digested with 0.4% wt/vol pepsin (Sigma) in 0.1 M HCl for 10 min at 37°C. For quenching the reaction, 0.2% glycine was used and then sections were postfixed with 4% paraformaldehyde and dehydrated through graded alcohols. Samples were air dried and incubated with FITC-labeled mouse Y-chromosome paint (Star-FISH; Cambio, Cambridge, UK). The probe was added to the sections, and a 10-min incubation at 60°C was performed followed by overnight incubation at 37°C. Slides were washed with 50% formamide/2× SSC and then 0.1× SSC followed by PBS. The sections were incubated with 1/50 peroxidase-conjugated antifluorescein antibody (Boehringer Mannheim), developed in 3,3′-diaminobenzidine, counterstained with hematoxylin, and mounted with either an aqueous mounting medium (Bio Optica) or 60% glycerol. Slides were analyzed under a light microscope. In additional experiments, sections were further stained with PAS reagent after initial examination and removal of the coverslip to stain tubular basement membrane structures for more precise localization of Y chromosome—positive cells.

Proliferating tubular cells were immunohistochemically identified by labeling with mAb against the proliferating cell nuclear antigen Ki-67 (Novocastra Laboratories, Newcastle, UK) as follows (35). Sections fixed in Dubosq-Brazil were deparaffinized and dehydrated. For obtaining an adequate signal with Ki-67 antibody, the sections were microwaved twice for 5 min in citrate buffer, and for reducing background, they were blocked for 30 min in 1% BSA (Sigma). The sections were incubated overnight with anti–Ki-67 Ab (1:100) at 4°C. After rinsing with PBS, sections were incubated with biotinylated sheep anti-mouse IgG (Chemicon International, Temecula, CA), 1:100 for 30 min at room temperature. Then streptavidin-alkaline phosphatase conjugate (Roche Diagnostic) was applied at a dilution of 1:100 for 30 min. The reaction product was visualized by incubation with Fast Red substrate (Vector Laboratories). Counterstaining was performed by using Harris hematoxylin (Bio-Optica, Milan, Italy). For establishing whether MSC may be proliferating, serial sections of kidneys of selected female mice at 4 d were stained as above for Y chromosome and Ki-67, respectively. In control experiments, the first antibody was omitted or replaced by normal serum. Signals in negative control sections were negative. Ki-67–positive cells within tubuli were counted in 10 randomly selected HPF (×400) for each animal (n = 5 mice for each group).

Statistical Analyses

The results are expressed as mean ± SEM. ANOVA with Tukey-Cicchetti test for multiple comparisons was used to analyze BUN data. Histology and immunohistochemistry data were analyzed using the nonparametric Kruskal-Wallis test or Mann-Whitney test, as appropriate. Statistical significance level was defined as P < 0.05.

MSC Characterization

MSC after isolation from BM by their adherence to plastic consisted of an heterogeneous cell population with a predominant spindle-shaped morphology and were able to form fibroblast-like colonies (Figure 1a). Under appropriate culture conditions, primary MSC were able to produce extracellular mineral deposition, indicating that MSC possess osteogenic potential (Figure 1b). In addition, MSC differentiated to adipocytes, as indicated by the accumulation of neutral lipid vacuoles, visualized by phase-contrast microscopy or stained with Oil Red O (Figure 1c). Finally, the accumulation of glycosaminoglycans within the extracellular matrix suggested that chondrogenesis had occurred (Figure 1d).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Photomicrographs of mesenchymal stem cells (MSC) showing differentiative potential. (a) MSC after 1 wk of culture grew in colonies that contained heterogeneous small spindle-shaped fibroblastoid cells and more rounded cells. (b through d) In vitro differentiation of MSC maintained in culture with specific inductive media. The differentiation toward osteoblasts is indicated by the formation of calcium-rich hydroxyapatite detected with Alizarin red (b). MSC-adipocyte differentiation is visualized by highly refractive intracellular lipid vacuoles (c, left) and by Oil Red O staining (c, right). Chondrogenesis was induced by incubating MSC as a micromass pellet in chondrogenic medium. Glycosaminoglycans stained by safranin O indicate chondroblast differentiation (d). Magnifications: ×100 in a; ×40 in b; ×400 in c and d.

MSC Can Cure ARF

To determine whether MSC and HSC obtained from BM of male C57BL6/J mice ameliorated acute renal dysfunction, we injured the kidney of female C57BL6/J mice with cisplatin, an antitumor drug whose clinical use is accompanied by a high incidence of nephrotoxicity, mainly in the form of renal tubular damage. In the murine model of cisplatin nephrotoxicity, renal function, as measured by serum BUN and creatinine, is impaired between days 4 and 7 after drug administration (10,36⇓). In our setting, subcutaneous injection of 12.7 mg/kg cisplatin induced significant increases in serum BUN, which peaked at days 4 and 5, declined at day 7, and stabilized at days 11 and 29 to values slightly higher than baseline (Figure 2). Intravenous injection of 2 × 10⁵ MSC obtained from the BM of male mice into syngeneic female mice on day 1 after receiving cisplatin strongly protected renal function at days 4 and 5 as reflected by significantly lower BUN values with respect to cisplatin-treated mice that were given saline (Figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Mesenchymal stem cells (MSC) but not hematopoietic stem cells (HSC) protected cisplatin-treated mice from renal function deterioration. MSC or Lin⁻c-kit^POS HSC were purified from bone marrow (BM) of male C57BL6/J mice and injected intravenously (2 × 10⁵ cells) into female C57BL6/J mice 1 d after cisplatin (12.7 mg/kg subcutaneously). Blood urea nitrogen (BUN) levels were measured at different time intervals in cisplatin-treated mice that received saline (○), MSC (▪), or HSC (▴). Data are mean ± SE. °P < 0.01 versus day 0 (basal) and cisplatin+saline at 7, 11, and 29 d; *P < 0.01 versus cisplatin+saline at the same time.

Next, we investigated whether the renal function improvement by MSC treatment was associated with preservation of tubular structure. By light microscopy, no significant changes were detectable in the kidneys taken 1 d after cisplatin injection. Focal and severe tubular changes were observed at day 4 (Table 1, Figure 3b). Proximal tubuli revealed loss of brush border, cytoplasmic vacuolization, flattening and loss of the epithelial cells, PAS-positive droplets, nuclear fragmentation, and luminal cell debris. Hyaline casts were prominent. Distal nephron segments revealed cytoplasmic swelling, fragmented nuclei, and luminal debris. The most damaged tubuli displayed areas of apparent denudation of the tubular basement membranes. Scores of tubular damage reached the highest values at day 4 (Table 1). At day 7, tubular changes were still present. They recovered within 29 d, except for sparse casts (Figure 3c). Like the effect on renal dysfunction, MSC had an impressive effect on renal pathologic changes. Injection of MSC almost completely abrogated tubular damage. In particular, the kidneys of MSC-treated mice that were killed on day 4 after cisplatin exhibited low degrees of proximal tubular cell damage, cast formation, and focal cell loss (Table 1, Figure 3d).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Renal histology of C57BL6/J control or cisplatin-treated mice that received saline, MSC, or Lin⁻ c-kit^POS HSC. (a) Normal mouse kidney (control). (b) Cisplatin-treated mouse at 4 d. Severe tubular changes (epithelial cell flattening and loss) are visible in addition to loss of brush border, luminal casts, and cell debris. (c) Tissue recovery at 29 d. (d) MSC-injected cisplatin-treated mouse at 4 d shows less severe tubular damage consisting of occasional luminal casts and, as shown here, very mild cell swelling. Cisplatin-treated mice that received MSC (e) show overall comparable tissue recovery as other groups at 29 d. HSC-injected cisplatin-treated mice at 4 d were similar to the cisplatin-treated mice that were given saline at this time (f). Most tubular changes in HSC-treated mice recovered at 29 d (g). Magnification, ×400.

Electron microscopy analysis of the kidneys of all cisplatin-treated mice revealed changes on day 1 at the time of MSC injection, both in proximal tubuli and in segments of the distal nephron (Figure 4, c and d). Tubular cells showed loss of brush border, cytoplasmic vacuolization, apical blebbing, phagolysosomes, and enlarged mitochondria with irregular cristae. Cell lysis and detachment from the tubular basement membrane were prominent in kidneys taken at day 4 after cisplatin injection (Figure 4, e and f). In contrast, the kidneys of mice that received MSC and were killed on day 4 showed much less severe ultrastructural changes. Remarkably, areas of cell detachment were either absent or detectable in just a few scattered tubuli (Figure 4, g and h).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4. Ultrastructural changes after cisplatin injection in C57BL6/J mice and effect of treatment with MSC. Normal proximal (a) and distal (b) tubular cells in a control mouse kidney. Cisplatin-treated mouse at 1 d, time of MSC injection, shows increase in apical vacuoles in the epithelial cells of a proximal tubule (c) and mitochondrial swelling in a distal tubule (d). At 4 d after cisplatin, severe cell blebbing (*), cell lysis (arrows), and tubular basement membrane denudation (arrowheads) are evident in a proximal tubule (e) and in a distal segment (f). In cisplatin-treated mouse that received an injection of MSC, at 4 d, tubular cell injury is markedly attenuated (g and h). Magnification, ×2800.

MSC Engraft and Differentiate to Tubular Epithelial Cells

To assess whether the protective effects of MSC in cisplatin-induced ARF could be associated with recruitment of the cells into the renal parenchyma, we analyzed kidneys of cisplatin-injected mice for the presence of Y chromosome by in situ hybridization. The Y chromosome was detected in the majority of cell nuclei in sections of male control mouse kidneys (Figure 5a). No signal was present in kidneys of female controls (not shown) or of the female mice that received an injection of cisplatin and were left untreated (Figure 5b). Sections of kidney cortex from mice that received an injection of cisplatin and were treated with MSC revealed Y chromosome–positive cells both in proximal tubuli and in distal tubuli in focal areas at day 4. The distribution of Y chromosome–positive cells closely reproduced the focal distribution of injury otherwise found in mice that were given cisplatin. The morphology of the tubuli that contained Y-positive cells was well preserved or even comparable to normal (Figure 5, c through e). Y chromosome–positive cells were aligned within epithelial areas showing well-recognizable brush borders positive for the tubular differentiation marker binding sites for Lens culinaris lectin. In addition, Y chromosome–positive cells were focally found in peritubular areas in the proximity of tubular profiles. The localization of the cells in close contact to the outer aspect of the tubular basement membrane (Figure 5e) was confirmed on PAS-stained sections (Figure 5f), possibly consistent with migration into the tubule from the surrounding peritubular structures. Y chromosome–positive cells were also detected within the context of the tubular epithelium on day 29 in cisplatin-treated mice that were given MSC (Figure 5, g and h). These findings suggest local recruitment of MSC through renotropism at sites of injury and provide evidence that mesenchymal cells actively participate to the reconstitution of the differentiated epithelial lining.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5. Detection of Y chromosome–positive cells by in situ hybridization in mouse kidney sections simultaneously stained with the tubular cell marker Lens culinaris lectin, and comparison with periodic acid-Schiff (PAS) staining in sequential sections to identify relationships with the tubular basement membrane. (a) Male control shows staining for the Y chromosome (brown color). (b) Cisplatin-treated female mouse that received an injection of saline and was killed at 4 d. Tubular damage is severe in one area with extensive disruption of Lens culinaris lectin sites (red color). (c through f) Cisplatin-treated mice that received injections of MSC and were killed at 4 d. Signal for Y chromosome is seen in tubular epithelial linings (arrows). Y chromosome–positive nuclei are recognizable in the context of the epithelium of proximal tubuli showing differentiated brush borders (arrows in d and e). One Y chromosome–positive cell is located at the periphery of a proximal tubular profile (arrowhead in e). (f) PAS staining of the same section confirms the localization of the Y- chromosome–positive cells at either the tubular (double arrows) or the interstitial aspect (double arrowheads) of the basement membranes. (g and h) Cisplatin-treated mouse that received an injection of MSC and was killed at 29 d. Magnifications: ×400 in a, b, c, and g; ×1000 in d, e, f, and h.

MSC Accelerate Tubular Epithelial Cell Regeneration

The effect of engrafted MSC on tubular cell regeneration was explored by analyzing the proliferation marker Ki-67 in cisplatin-treated mice at different time intervals after injection with saline or MSC. In normal control kidneys and in cisplatin-treated mice at day 4, low numbers of Ki-67–positive cells were detected, reflecting low frequency of tubular cell turnover (control, 2.5 ± 0.5; cisplatin day 4, 4.6 ± 1.9 cells/HPF). The numbers of Ki-67–positive cells in kidneys of cisplatin-treated mice were significantly higher since day 11 as compared with normal controls (day 11, 6.2 ± 1; day 29, 24.5 ± 5.1 cells/HPF; P < 0.05 and P < 0.01 versus control, respectively). The effects of MSC on cell proliferation over time in cisplatin-treated mice are shown in Figure 6. Injection of MSC into cisplatin-treated mice resulted in a fourfold increase in tubular cell proliferation at day 4 over cisplatin-treated mice that were given saline (P < 0.05; Figure 6a). With respect to cisplatin-treated mice showing sparse Ki-67–positive cells (Figure 6b), high numbers of Ki-67–positive cells were detected at this time both in proximal tubuli and, to a lesser extent, in distal nephron profiles after MSC injection (Figure 6c). Higher numbers of Ki-67–positive cells within tubuli in MSC transplant recipients were also detected on day 11 (P < 0.05 versus cisplatin+saline). Similar degrees of proliferation were recorded thereafter in both groups (Figure 6, a, d, and e). Staining for Y chromosome and Ki-67 on serial sections revealed co-localization of both markers to the same nuclei (Figure 7), indicating actual proliferation of MSC within tubule.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6. Effects of MSC on tubular cell regeneration after cisplatin. (a) Counts of Ki-67–positive nuclei within tubuli of cisplatin-treated mice that were treated with saline or MSC at different time intervals (4, 11, and 29 d). (b through e) Ki-67 staining in kidneys of cisplatin-treated mice at 4 (b) and 29 d (d) and effects of MSC at 4 (c) and 29 d (e). Data are mean ± SEM. HPF, high-power field. °P < 0.05 versus cisplatin+saline 4 and 11 d; *P < 0.05 versus cisplatin+saline at the corresponding time.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7. Photomicrographs of proximal tubule in sequential serial sections stained for detection of Ki-67 (a) or Y chromosome (b) in the kidney of a cisplatin-treated mouse that received an injection of MSC and was killed at 4 d. Co-localization is seen in a nucleus in the context of the epithelial lining (arrow). The arrowhead points to Y chromosome staining. Tubuli were encircled for easier recognition. Magnification, ×1000.

MSC Immunodepleted of CD45⁺ Cells Protect Cisplatin-Treated Mice against Renal Damage

On the basis of the evidence (32) that murine MSC cultures established from various inbred mice contain heterogeneous populations, including hematopoietic precursors, we performed additional experiments by injecting an enriched population of MSC immunodepleted of CD45⁺ cells (CD45⁻ MSC), 1 d after cisplatin. Injection of CD45⁻ MSC significantly reduced serum BUN in cisplatin-treated mice at day 4 to levels that were comparable to MSC (CD45⁻ MSC, 44.2 ± 5.8 and MSC, 32 ± 4.4 versus saline, 103.6 ± 8.9 mg/dl; P < 0.01). The light microscopy analysis confirmed the protective effect on renal structure (score values: cisplatin+CD45⁻ MSC, casts 0.75 ± 0.16; tubular degeneration 0.25 ± 0.16; cell loss 0 versus cisplatin +saline, casts 1.8 ± 0.24; tubular degeneration 1.8 ± 0.24; cell loss 2 ± 0.21; P < 0.01 for all lesions).

HSC Fail to Protect Cisplatin-Treated Mice against ARF

To examine whether HSC may have the same protective potential as MSC against renal injury, we assessed the effect of Lin⁻ c-kit^POS HSC purified from BM of male mice. Injection of HSC into female cisplatin-treated mice (2 × 10⁵ cells given intravenously on day 1) did not ameliorate renal function, as indicated by serum BUN levels that were similar to those of cisplatin-treated mice that were given saline, at any time points considered (Figure 2). By histology, at day 4 after cisplatin, mice that received HSC had severe tubular changes comparable to those of mice that received saline (Table 1, Figures 3f and 8a⇓). As shown in Figure 8b, occasional Y chromosome–positive cells were localized, mostly in the least severely damaged tubular structures, at day 4 in kidneys of cisplatin-treated mice that received HSC.

• Download figure
• Open in new tab
• Download powerpoint

Figure 8. Simultaneous detection of Y chromosome–positive cells by in situ hybridization and of Lens culinaris lectin sites by double staining in a kidney section of a cisplatin-treated mouse that received an injection of HSC and was killed at 4 d (a,b). No or very few Y chromosome–positive cells (one is marked by arrow in b) could be identified. Magnification, ×400.