Stromal Cells Protect against Acute Tubular Injury via an Endocrine Effect
Baoyuan Bi,
Roland Schmitt,
Malika Israilova,
Hitoshi Nishio and
Lloyd G. Cantley
Section of Nephrology, Department of Medicine, Yale University, New Haven, Connecticut
Correspondence: Dr. Lloyd G. Cantley, Yale University School of Medicine, 333 Cedar Street, PO Box 208029, New Haven, CT 06510. Phone: 203-785-4186 Fax: 203-785-3756; E-mail: lloyd.cantley{at}yale.edu
Received for publication February 1, 2007.
Accepted for publication May 22, 2007.
Emerging evidence suggests that the intravenous injection ofbone marrow–derived stromal cells (BMSC) improves renalfunction after acute tubular injury, but the mechanism of thiseffect is controversial. In this article, we confirm that intravenousinfusion of male BMSC reduced the severity of cisplatin-inducedacute renal failure in adult female mice. This effect was alsoseen when BMSC (or adipocyte-derived stromal cells (AdSC)),were given by intraperitoneal injection. Infusion of BMSC enhancedtubular cell proliferation after injury and decreased tubularcell apoptosis. Using the Y chromosome as a marker of donorstromal cells, examination of multiple kidney sections at oneor four days after cell infusion failed to reveal any examplesof stromal cells within the tubules, and only rare examplesof stromal cells within the renal interstitium. Furthermore,conditioned media from cultured stromal cells induced migrationand proliferation of kidney-derived epithelial cells and significantlydiminished cisplatin-induced proximal tubule cell death in vitro.Intraperitoneal administration of this conditioned medium tomice injected with cisplatin diminished tubular cell apoptosis,increased survival, and limited renal injury. Thus, marrow stromalcells protect the kidney from toxic injury by secreting factorsthat limit apoptosis and enhance proliferation of the endogenoustubular cells, suggesting that transplantation of the cellsthemselves is not necessary. Identification of the stromal cell–derivedprotective factors may provide new therapeutic options to explorein humans with acute kidney injury.
Recent studies of injury in organs such as the heart, liver,brain, and pancreas have introduced the controversial conceptthat adult bone marrow cells can provide a source of organ protection,organ repair, or both.1–4 Multiple mechanisms for theseeffects have been proposed, including differentiation of bonemarrow–derived cells into organ-specific phenotypes, fusionof bone marrow–derived cells with existing differentiatedcells, protection of existing cells by either paracrine or endocrineactions of the bone marrow cells, and/or inhibition of inflammatoryresponses associated with the organ injury. The bone marrowcontains multiple cell types, with the majority of cells belongingto the hematopoietic lineages (hematopoietic stem cells [HSC]and their lineage-positive derivatives) and the supporting stromalcells (marrow stromal cells [MSC], which are believed to providethe niche for HSC to survive for long periods under adverseconditions). Although still controversial, differentiation intoorgan-specific cell types has been proposed for both HSC andMSC, whereas fusion has been predominately associated with HSCand organ protection and suppression of inflammation have beenproposed as features of MSC action.
HSC are well characterized and can be purified on the basisof the surface expression of c-Kit and Sca-1 and the absenceof the lineage markers that are found on their downstream derivatives(thus HSC are lin-cKit+Sca1+). In contrast, the MSC populationis not well characterized and historically has been definedas bone marrow–derived cells that adhere to tissue culturedishes and proliferate extensively in vitro. These cells lacklineage marker expression but may express other surface proteinssuch as CD133 and Stro1.5 As noted, MSC are believed to playa supportive role in maintaining the viability of HSC,6 butthese cells have also have been found to be capable of differentiatinginto adipocytes, chondrocytes, and osteocytes in vitro and thereforeare alternatively called "mesenchymal stem cells."7 Furthermore,a rare population of MSC termed multipotent adult progenitorcells have been cultured in vitro and found to differentiateinto other cell types, including neurons and myocytes.8
A possible role of bone marrow–derived cells in the kidney’sresponse to injury was first proposed when studies of humanfemale kidneys that were transplanted into male recipients demonstratedY chromosome–positive cells within the tubules.9,10 Becausebone marrow cells can be mobilized into the circulation afterhypoxia or ischemia,11 these studies supported the concept thatbone marrow–derived cells might directly participate inkidney tubule repair. In conceptual agreement with this possibility,studies in which whole bone marrow from a donor mouse was infusedinto a recipient mouse that had been subjected to renal injuryhave confirmed that bone marrow–derived cells enter theinjured kidney. However, the distinct majority of these cellsare present in the renal interstitium and express markers ofinflammatory cells such as leukocytes and macrophages,12–15thus suggesting that these cells might worsen the acute renalinjury as a result of their proinflammatory effects rather thanimprove repair.16 Previous work by our laboratory demonstratedthat ablation of bone marrow by irradiation worsens the courseof ischemia/reperfusion–induced acute renal failure andthat infusion of lineage-depleted bone marrow cells (containingboth bone marrow–derived stromal cells [BMSC] and HSC)can restore the repair process to normal,17 again supportingthe idea that bone marrow cells normally play a role in eitherpreventing renal injury or improving renal repair.
Further insight into the importance of bone marrow–derivedcells in modulating renal injury was provided by the work ofMorigi et al18 When they separated whole bone marrow into HSCand MSC fractions and injected them separately, they found thatthe MSC fraction, not the HSC fraction, provided protectionagainst acute tubular injury. This general observation has beenreproduced in several models of acute kidney injury, includingglycerol injection and ischemia/reperfusion.13,14,19,20 However,the mechanism of the MSC effect remains controversial becausesome groups reported that the injected BMSC infiltrate the kidneyand directly populate the injured renal tubule,18,19 whereasothers have found only transient evidence for injected MSC inthe renal vasculature and no evidence for direct BMSC incorporationinto tubules during the repair process.13,14 These transientlypresent MSC have been shown to provide paracrine support forvascular endothelial cells in the injured kidney.21
Understanding the mechanism by which bone marrow cells protectagainst acute tubular damage is a critical aspect of developingtherapeutic interventions for patients on the basis of thesecells. If the cells act by engrafting the tubules long term,then either they will need to be host-derived (to prevent rejection)or the patient will require immunosuppressive therapy. In contrast,if the cells merely transit through the kidney and act in aparacrine manner to protect or stimulate the endogenous renalcells, then they might only need to survive for a few days andthus could be expanded from a single donor for use by multiplerecipients. Finally, if the protective effect is mediated inan endocrine manner, then injection of the cells themselveswould not be required but rather the factors that those cellssecrete could be provided immediately at the time of kidneyinjury.
To address these disparate possibilities, we compared intravenousand intraperitoneal injection of bone marrow–and adipocyte-derivedMSC22 and found that they provided indistinguishable levelsof protection in a model of cisplatin-induced acute tubularinjury. Furthermore, we found only rare examples of these transplantedcells in the renal parenchyma but did find that animals thatreceived these cells displayed a decrease in endogenous tubularcell apoptosis. Conditioned medium produced by cultured MSCwas found to induce epithelial cell growth and survival in vitroand to protect mice from acute kidney injury when injected intraperitoneally.
In Vivo Effects of BMSC
For determination of the mechanism by which BMSC improve theoutcome of acute kidney injury, mice were administered injectionsof two doses of cisplatin to induce acute tubular injury followed24 h later by intravenous infusion of either 2 x 105 BMSC orvehicle control (200 µl volume). Both groups of mice becamesystemically ill with evidence of dehydration and anorexia (weightloss), although mice that received BMSC demonstrated greatersurvival as compared with control mice (Figure 1A). Evaluationof renal function in mice that were treated with cisplatin alonedemonstrated a marked rise in blood urea nitrogen (BUN) andmoderate rise in creatinine by day 3, whereas mice that receivedBMSC exhibited less of an initial decline in renal function(Figure 1, B and C). All surviving mice were killed after BUNand creatinine determination on day 6. Of note, the apparentimprovement in renal function parameters on day 6 in the cisplatincontrol group was primarily due to the previous death of micethat exhibited the highest BUN and creatinine values on day3. Examination of renal histology on day 6 demonstrated diffusetubular injury in the cisplatin control group, primarily inthe cortical proximal tubules (Figure 1D). Kidneys from micethat received intravenous BMSC infusion revealed fewer numbersof necrotic tubules and fewer tubular casts.
Figure 1. Effects of intravenous bone marrow–derived stromal cells (BMSC). Mice were given intraperitoneal injections of cisplatin (10 mg/kg) on day 0 and day 1, followed by intravenous injection of BMSC or vehicle control on day 2. (A) Survival curve of cisplatin-treated mice with or without BMSC (the numbers in parentheses are surviving mice on day 6 per total mice in that group). (B and C) Blood urea nitrogen (BUN) and creatinine values at the beginning of the experiment and on days 3 and 6. *P < 0.05 versus cisplatin alone; **P < 0.01 versus cisplatin alone. (D) Renal histology on day 6 (images shown are representative areas of the cortex). Magnification, x40 (hematoxylin and eosin [H&E] stain).
In the kidney as well as other organs, MSC have been proposedin some cases to contribute directly to tissue repair, whereasother reports suggested a paracrine or endocrine action. Todetermine whether injected MSC must transit through the kidneyto provide their protective effect, we examined the abilityof stromal cells injected intraperitoneally to protect againstacute kidney injury. For these experiments, we compared BMSCwith a similar population of cells derived from adipose tissue(AdSC). Mice were again subjected to two doses of cisplatinfollowed 24 h later by intraperitoneal injection of either 2x 105 BMSC or 1 x 105 AdSC or vehicle control (500 µlof PBS). Similar to the results with intravenous BMSC, micethat received cisplatin followed by either intraperitoneal BMSCor AdSC demonstrated improved survival as compared with controlmice (Figure 2A). BUN values were less elevated in both groupsof treated mice, with no statistical difference between theBMSC and AdSC groups (Figure 2B). Renal pathology on day 6 againshowed more severe tubular damage in the cisplatin control groupas compared with the mice that received intraperitoneal BMSCor AdSC (Figure 2C).
Figure 2. Effects of intraperitoneal BMSC. Mice were given intraperitoneal injections of cisplatin (10 mg/kg) on day 0 and day 1, followed by intraperitoneal injection of BMSC, adipocyte-derived stromal cells (AdSC), or vehicle control on day 2. (A) Survival curve of cisplatin-treated mice with or without BMSC or AdSC (the numbers in parentheses are surviving mice on day 6 per total mice in that group). (B) BUN values at the beginning of the experiment and on days 3 and 6. *P < 0.05 versus cisplatin alone. (C) Renal histology on day 6 (images shown are representative areas of the cortex). Magnification, x40 (H&E stain).
Because cells that were injected into the peritoneum could slowlyenter the circulation and thus transit through the kidney, weused Y chromosome tracing to determine whether injected stromalcells could ever be detected in the renal parenchyma of cisplatin-treatedmice. For these experiments, we used male BMSC or AdSC injectedeither intravenously or intraperitoneally into cisplatin-treatedfemale mice. Fluorescence in situ hybridization for the Y chromosomewas performed on multiple sections from two to three mice ineach group. This approach generally results in the detectionof the Y chromosome in approximately 60 to 80% of cells froma male kidney, depending on the thickness of the section. Inour experiments, 65% of tubular and nontubular cells were foundto be Y chromosome positive in our male control kidneys, whereas0 cells were positive in female control kidneys (Figure 3A,quantified in 3D). We found no examples of Y chromosome–positivetubular cells in any kidney section examined at either 24 or96 h after cell injection (Figure 3, B and C). A single Y chromosome–positivenontubular cell was detected in the interstitium of one of themice 24 h after receiving BMSC intravenously (data not shown).To determine where the injected cells localized, we killed cisplatin-treatedfemale mice 1 h after injection of male BMSC and examined theliver, spleen, lungs, and kidneys for the presence of Y chromosome–positivecells. Numerous Y chromosome–positive cells were foundin the vasculature of the lung at this early time point, withno Y chromosome–positive cells detected in the liver,spleen, or kidneys (data not shown).
Figure 3. Y chromosome staining of BMSC-infused mouse kidneys. Kidneys were fixed and fluorescence in situ hybridization performed for detection of Y chromosome. (A) Male control kidney showing Y chromosome staining (red dots) in nuclei of tubular and nontubular cells (arrows). The 5-µm sections exclude the Y chromosome from some cells, resulting in 65% of cells being Y chromosome positive in the male kidney. No cells were Y chromosome positive in the female kidney control (data not shown). (B and C) Fluorescence in situ hybridization analysis of kidneys from female mice that were treated with cisplatin followed by intravenous injection of 2 x 105 male BMSC. Kidneys were examined 24 h (B) and 96 h (C) after cell injection. (D) Numbers of Y chromosome–positive cells found in the kidneys of control mice or female mice that were treated with cisplatin followed by intravenous or intraperitoneal injection of male BMSC. Multiple sections of kidneys from two to four mice were examined for each time point. Magnification, x100.
The failure to detect injected stromal cells in the kidney ledus to investigate the hypothesis that the primary effect ofinfused BMSC or AdSC is on endogenous tubular cell function.For examination of this question, mice that were treated withcisplatin with or without BMSC were killed on day 4 to determinethe degree of tubular cell apoptosis. Examination of the renalcortex revealed approximately 30 terminal deoxynucleotidyl transferase–mediateddigoxigenin-deoxyuridine nick-end labeling (TUNEL)-positivecells per high-power field in cisplatin control mice, whereasmice that received BMSC had fewer than five TUNEL-positive cellsper high-power field (Figure 4, A and B, quantified in C). Bothtubular and nontubular cells were found to be TUNEL positive,although the majority of the cells were in the tubular compartment.
Figure 4. Renal cell proliferation and apoptosis after cisplatin injury. Mice were given intraperitoneal injections of cisplatin (10 mg/kg) on day 0 and day 1, followed by intravenous injection of BMSC or vehicle control on day 2. (A and B) Terminal deoxynucleotidyl transferase–mediated digoxigenin-deoxyuridine nick-end labeling (TUNEL) staining was performed on kidneys 48 h after cisplatin+vehicle (A) or cisplatin+BMSC (cis+BMSC; B) injection. (C) Quantification of TUNEL-positive cells/high-power field (hpf). (D and E) Proliferating cell nuclear antigen (PCNA) staining was performed on sections of kidneys 96 h after cisplatin+vehicle (D) or cis+BMSC (E) injection (day 6 of the experiment). (F) Quantification of the PCNA-positive cells/hpf. (G and H) Bromodeoxyuridine (BrdU) was injected on days 3, 4, and 5 of the experiment, followed by kidney harvest and detection of BrdU-positive cells in mice that were treated with cis+vehicle (G) or cis+BMSC (H). (I) Quantification of BrdU-positive cells/hpf. All quantitative data were obtained from three to five mice per group with each mouse coming from a separate experiment. *P < 0.01 versus the respective cisplatin control. Magnifications: x40 in A, B, G, and H; x20 in D and E.
Renal tubular repair after acute injury is believed to requireproliferation of the surviving endogenous tubular cells. Immunostainingfor proliferating cell nuclear antigen (PCNA) demonstrated thatkidneys from mice that had received BMSC after the cisplatininjection had a greater number of PCNA-positive cells as comparedwith cisplatin control mice (Figure 4, D and E, quantified inF). Because PCNA expression is also upregulated during the processof DNA repair and therefore might be increased as a result ofcisplatin-mediated DNA damage independent of cell proliferation,a separate group of mice were administered and injection ofbromodeoxyuridine (BrdU) for 3 d before being killed, and proliferatingcells were detected by BrdU uptake. Again, a significantly largernumber of BrdU-positive cells were detected in the BMSC-treatedmice, with the majority of these cells lying within the tubularbasement membrane (Figure 4, G and H, quantified in I).
In Vitro Effects of BMSC
In a separate series of experiments, isolated BMSC were culturedunder various conditions in an attempt to induce differentiationtoward an epithelial lineage. Culture of BMSC in the presenceof growth factor combinations including hepatocyte growth factor(HGF), EGF, retinoic acid, and erythropoietin, as well as culturein serum from mice that had been subjected to renal ischemicor toxic injury, failed to result in expression of any epithelialmarkers by the cultured cells (data not shown). However, co-cultureof fragments of kidney with the BMSC, even in medium that lackedFCS, resulted in the appearance of small islands of cells withan epithelial appearance. These cells expressed the epithelialmarkers cytokeratin, zonula occludens-1, and E-cadherin (Figure 5,A and B, and Supplemental Figure 2) and proliferated in culturefor more than 10 d. Co-culturing male BMSC with female kidneyfragments or GFP-positive BMSC with wild-type kidney fragmentsdemonstrated that these islands of epithelial cells were derivedfrom the kidney fragment and not from the BMSC (Figure 5C).Even in cases in which the BMSC were entirely surrounded byan epithelial island, there was no detectable expression ofepithelial markers by the BMSC (Supplemental Figure 2).
Figure 5. In vitro effects of BMSC on renal tubular cells. (A and B) Direct co-culture of BMSC (arrows) with fragments of injured kidney cortex resulted in the outgrowth of colonies of cobblestone-appearing cells that express cytokeratin (A) and zonula occludens-1 (B). (C) Direct co-culture of green fluorescence protein (GFP)-positive BMSC (green cells noted with arrows) with GFP-negative kidneys demonstrates that the epithelial cells are derived from the kidney. Microdissection and isolation of tubule segments (shown in D through G) followed by culture in BMSC conditioned medium (CM) resulted in epithelial colony growth from all tubule segments tested (H) Colonies derived from S3. (I) Colonies grown from distal convoluted tubule (DCT) with the DCT fragment visible (arrow). S1, S2, and S3, segments of the proximal tubule; TDL, thin descending limb; CS, connecting segment; CCD, cortical collecting duct.
To determine whether direct contact between the BMSC and thekidney fragment was required for migration and proliferationof the renal epithelial cells, we performed separation of theBMSC from the kidney fragment using a Transwell filter apparatus(Becton Dickinson Labware, Franklin Lakes, NJ), l1;6qagain inthe absence of added FCS. In these experiments, epithelial islandswere easily detected and were exclusively found in the chambercontaining the kidney fragment. Removal of the Transwell insertcontaining the BMSC, in combination with changing the culturemedium, resulted in death of the epithelial cells after 24 to48 h. These results strongly argue that BMSC exert an endocrineeffect on renal epithelial cell migration, growth, and survival.To investigate this further, we cultured BMSC in Iscove’sModified Dulbecco’s Medium (IMDM) in the absence of serumfor 48 h, and this BMSC-conditioned medium (BMSC-CM) was centrifugedto remove cell debris. Culture of kidney fragments in the BMSC-CMagain resulted in the appearance of numerous epithelial colonies(data not shown).
For determination of whether the islands of cells that weredetected after co-culture of kidney fragments with BMSC werespecific to a single kidney compartment, renal tubule segmentswere microdissected and cultured in the presence of BMSC-CM.In these experiments, all tubular segments were found to containcells that were capable of migrating onto the culture dish andproliferating as epithelial colonies, although the greatestnumber of colonies per millimeter of tubule length were obtainedfrom the collecting duct (Figure 5, D through I). These resultssuggest that endogenous tubular cells from all tubular regionsretain the capacity to proliferate and migrate in response toBMSC-derived stimuli.
BMSC Protect against Cisplatin-Mediated Tubular Cell Injury via an Endocrine Effect
The ability of BMSC to secrete factors that stimulate renalepithelial cell survival and proliferation led us to investigatethe possibility that these same endocrine effects could inhibittubular cell death. Immortalized mouse proximal tubule (MPT)cells were cultured in the presence of cisplatin with or withoutBMSC-CM. Sustained cisplatin exposure resulted in the detachmentand death of many MPT cells at 24 h and most cells at 48 h (Figure 6,A and B). However, the addition of BMSC-CM significantly inhibitedcell detachment at 24 h (Figure 6C, quantified in D) and decreasedcisplatin-induced cell death after 48 h (Figure 6E, quantifiedin F). In contrast, conditioned medium from cultured MPT cellsfailed to provide protection against cisplatin-induced celldeath (data not shown).
Figure 6.In vitro effects of BMSC on cisplatin injury. mouse proximal tubule (MPT) cells were cultured in the absence (A) or presence (B) of cisplatin (5 µg/ml) for 24 h. The presence of 50% BMSC-CM in the cisplatin-treated cells (C) resulted in fewer detached cells after 24 h.)D) MPT cells were cultured with or without cisplatin and with or without BMSC-CM for 48 h, followed by trypsinization and counting of viable (trypan blue negative) cells. (E) MPT cells cultured as in D were incubated with propidium iodide (PI) and anti–Annexin V followed by FACS analysis to detect dead (PI+) or apoptotic (Annexin V+) cells. Sorting from a single representative experiment is shown for each condition. (F) Quantification of FACS sorting as in E. n = 3 separate experiments; *P < 0.01 versus cisplatin control.
Cumulatively, these results fail to demonstrate that BMSC orAdSC directly contribute to renal epithelial cell protectionor repair but rather support the model that these cells primarilyact in an endocrine manner to secrete factors that can preventtubular epithelial cell death and stimulate proliferation. Thisfinding suggests that in vivo injection of the cells may notbe necessary to obtain the protective effect. For addressingthis possibility, mice were subjected to cisplatin injectionwith or without administration of BMSC-CM or AdSC-CM intraperitoneally.On the basis of the prediction that the half-life of many ofthe factors present in CM would be short, mice were given 1ml of conditioned medium twice daily beginning at the time ofthe first cisplatin injection. Control mice received the cisplatinwith an equal volume of IMDM that had not been used for BMSCor AdSC culture. Even though the large volume of fluid preventeddehydration in these mice, six of 15 of those in the cisplatincontrol group had died by day 6, whereas no mice died in theCM-injected groups (Figure 7A). Furthermore, BUN values weremarkedly improved in both BMSC-CM–and AdSC-CM–treatedmice at 3 and 5 d after cisplatin injection (Figure 7B). Consistentwith the results obtained with injection of the BMSC themselves,the rate of apoptosis was diminished in the BMSC-CM group (Figure 7C),and renal pathology was significantly better in the two groupsof mice that received CM as compared with the cisplatin controls(Figure 7D).
Figure 7.In vivo effects of BMSC-CM on cisplatin injury. Mice were given intraperitoneal injections of cisplatin (10 mg/kg) on day 0 and day 1. All mice received twice-daily intraperitoneal injections of either 1 ml of IMDM (control) or 1 ml of BMSC-CM or AdSC-CM. (A) Survival curve of cisplatin-treated mice with or without BMSC-CM or AdSC-CM (the numbers in parentheses are surviving mice on day 6 per total mice in that group). (B) BUN values at the beginning of the experiment and on days 3 and 6. *P < 0.05 versus cisplatin alone; **P < 0.01 versus cisplatin alone. (C) TUNEL staining was performed on day 3 in mice that received cisplatin with or without BMSC-CM. n = 3 mice from separate experiments; *P < 0.01. (D) Renal histology on day 6 from a mouse in each group (images shown are representative areas of the cortex). Magnification, x40 (H&E stain).
In agreement with the studies of Duffield et al.13 and Togelet al.,14 our experiments provide no evidence that injectedstromal cells entered the renal parenchyma in significant numbersor directly contributed to tubular cell repopulation. In fact,the complete absence of injected MSC in the renal parenchymaargues that the bone marrow–derived cells that are foundin the renal parenchyma after either whole bone marrow infusionor lineage-negative bone marrow infusion must be derived fromHSC and their derivatives.12,13,17,23 Instead, our data suggestthat BMSC as well as AdSC secrete a factor or factors that caninhibit cisplatin-induced renal epithelial cell apoptosis invitro and in vivo and thereby limit the renal injury inducedby this toxin as well as improve the survival of mice that aresubjected to cisplatin injection.
MSC have been shown to secrete a variety of factors, includinggrowth factors such as HGF, vascular endothelial growth factor,IGF-1, and EGF; prostaglandins such as PGE2; and cytokines includingG-CSF, stem cell factor, and M-CSF.14,24–26 Of these,HGF and IGF-1 have been shown to reduce tubular injury whengiven to mice subjected to either toxic or ischemic acute kidneyinjury27,28 whereas vascular endothelial growth factor can mediateendothelial as well as epithelial cell proliferation and survivalafter injury.29,30 Finally, MSC-secreted factors such as TGF-and PGE2 can inhibit lymphocyte activation and thereby suppressthe inflammatory responses that might otherwise augment thetubular injury and increase rates of apoptosis.25,26,31 Therefore,it is likely that multiple factors that are secreted by theBMSC are acting in concert to limit the acute injury associatedwith cisplatin exposure. In agreement with this, we found thatBMSC-CM fractions of <10, 10 to 30, and >30 kD all arecapable of stimulating renal tubular cell colony growth in vitro(data not shown).
To date, no successful therapies that alter the outcome of acutekidney injury in patients have been developed. Several individualgrowth factors have shown promise in mouse models but have notproved effective in the limited human trials attempted. Thisstudy demonstrates excellent protection against cisplatin-inducedinjury in the mouse using conditioned medium from cultured stromalcells. Identification of the protective factors present in thismedium should provide us with new therapeutic avenues to considerin humans with acute kidney injury.
Reagents
Antibodies for FACS analysis and immunocytochemistry were obtainedfrom BD Biosciences (San Jose, CA) unless otherwise indicated.Anti-mouse CD105/endoglin-FITC was purchased from R&D Systems(Minneapolis, MN). IMDM and FCS were purchased from Invitrogen(Carlsbad, CA). Cisplatin was obtained from Sigma Chemical Co.(St. Louis, MO). MPT cells are a line derived from the ImmortoMouseand were a gift of Dr. Jon Schwartz (Boston University, Boston,MA).30,32
Stromal Cell Isolation and Expansion
BMSC were isolated and cultured from bone marrow of 6- to 8-wk-oldmale C57/Bl6 mice by the method of Peister et al.33 Whole bonemarrow cells were plated at 1 to 2 x 107/10-cm dish in IMDMthat was supplemented with 10% FCS, and the nonadherent cellswere removed by a medium change at 48 to 72 h and every 4 dthereafter. When the cells reached near confluence, they weretrypsinized and passaged at low density for further expansion.At the end of the second passage or at 5 to 8 wk of expansion,BMSC were used for transplantation or generation of conditionedmedium (Supplementary Figure 1).
AdSC were purified from abdominal adipose samples by the methodof Meyerrose et al.34 Briefly, abdominal adipose tissue washarvested at the same time as the bone marrow, and samples werewashed thoroughly with PBS followed by digestion in 0.075% collagenase(Sigma) for 30 min at 37°C. Digestion was stopped with IMDMmedia (Invitrogen) containing 10% fetal bovine serum (FBS).The digested adipose tissue was centrifuged at 1200 x g for10 min to obtain a cell pellet. The pellet was then resuspendedin PBS and filtered through a 70-µm nylon screen. Cellswere plated at a density of 1 x 107 cells/10-cm dish, and nonadherentcells were removed by washing. Adherent cells were maintainedat subconfluence in IMDM with 10% FBS, 50 U/ml penicillin, 50µg/ml streptomycin sulfate, and 2 mM l-glutamine (Invitrogen).Twenty-four hours after initial plating, nonadherent cells wereremoved by washing and the cells were refed with fresh medium.Cells were maintained at subconfluence by dissociation with0.25% trypsin-EDTA (Sigma-Aldrich) and replating under the sameculture conditions at a 1:4 dilution.
Characterization of Adult Stromal Cells
Surface marker expression of BMSC and AdSC was determined byFACS analysis. Briefly, cells were detached using trypsin/EDTAfor 5 min, immediately washed with PBS to remove trypsin, andresuspended at 106/ml. Of note, no difference in marker expression(including MSC-related and hematopoietic molecules) was observedwhen cells were detached using a cell scraper rather than trypsinization.The cell suspension (100 µl) was incubated at 4°Cfor 15 to 30 min with 10% FCS, followed by incubation with thespecific antibody at 4°C for 30 min. The cells were washedwith PBS and interrogated by flow cytometry (FACSCalibur; BectonDickinson). Negative controls were performed using isotype controlantibodies (BD Bioscience, San Jose, CA) and the number of positivecells per 10,000 events was determined using the Cell Questsoftware (BD Bioscience).
Cisplatin-Induced Acute Tubular Injury
Experiments were performed using 10- to 12-wk-old male C57BL/6mice weighing approximately 18 to 20 g. For induction of acuterenal failure, mice were given an intraperitoneal injectionof cisplatin (10 mg/kg body wt) on 2 successive days (day 0and day 1, total dosage 20 mg/kg). Twenty-four hours after thesecond cisplatin dose (day 2), BMSC (2 x 105 cells/mouse), AdSC(1 x 105 cells/mouse), or vehicle (control) was injected eithervia tail vein or intraperitoneally. Blood for BUN and creatininewas obtained before the first cisplatin dose (baseline) andin surviving mice on days 3 and 6 of the experiment. For TUNELstaining, mice were killed 48 h after the second cisplatin doseand kidneys were harvested. For detection of proliferation,mice were administered an injection of BrdU (100 mg/kg intraperitoneally;BD Bioscience) for 3 successive days before being killed. Kidneytissues were processed for histology (hematoxylin and eosin),TUNEL, or immunostaining as described. All surviving mice werekilled 6 d after the first cisplatin injection, and kidneyswere collected for histology. Each experiment was repeated onat least three separate occasions with five mice per experimentalgroup each time. All animal protocols were approved by the YaleUniversity Institutional Animal Care and Use Committee.
Immunocytochemistry
Kidney sections were subjected to antigen retrieval, and slideswere blocked and labeled with a 1:50 dilution of monoclonalanti-BrdU antibody (Sigma) and Alexa488-conjugated goat anti-mousesecondary antibody (Invitrogen). For PCNA detection, a monoclonalmouse anti-rat PC12 antibody was used at dilution 1:40 (Calbiochem,San Diego, CA) and developed using the Vectastain ABC kit (VectorLaboratories, Burlingame, CA). Apoptotic cells were identifiedusing a TUNEL assay using the In Situ Cell Death Detection Kit(Roche, Mannheim, Germany). Accordingly, kidney sections weredeparaffinized and rehydrated, and antigen retrieval was performedwith BD Pharmingen Retrievagen A kit (BD Biosciences) and labeledwith the TUNEL reaction mixture for 60 min at 37°C. Scoringfor BrdU-, PCNA-, or TUNEL-positive cells was carried out bycounting the number of positive nuclei per field in 10 randomlychosen sections of kidney cortex and outer medulla using x40magnification. Approximately 1000 to 1500 nuclei were screenedper each field. For each condition, all data from three to fiveseparate mice obtained from separate experiments were pooledto obtain final cell numbers.
Fluorescence In Situ Hybridization
Fluorescence in situ hybridization was performed as describedpreviously.35 Briefly, sections were deparaffinized and rehydratedas described previously, antigen retrieval was performed inBD Retrievagen A solution (BD Pharmingen) followed by washingin 2x SSC, and HCl and postfixation were performed with 4% paraformaldehyde.Air-dried sections were incubated with the mouse Y chromosome–specificprobe overnight, washed in 2x SSC, and then incubated with anti-rhodamine(1:20; Roche Diagnostics, Indianapolis, IN) in 4x SSC/1% BSA/0.1%Tween20 at 37°C for 45 min. Slides were finally mountedwith VECTASHIELD/DAPI (Vector Laboratories) and sealed beforeviewing.
Measurement of Serum Creatinine and BUN
Serum BUN values were performed by the Mouse Metabolic PhenotypingCenter at Yale University using the COBAS Integra system (Roche).Creatinine measurements were performed using HPLC as per themethod of Yuen et al.36,37
Experiments Using BMSC-CM and AdSC-CM
The CM was generated as follows: 1 x 106 BMSC or AdSC at 5 to8 wk of culture were washed and refed with serum-free IMDM for4 d. The medium was harvested, cell debris was removed usingcentrifugation, and the supernatant (CM) was used immediatelyfor in vivo injection or in vitro co-culture. A total volumeof 1 ml of CM or control IMDM was injected in the peritonealspace two times each day for the 6 d of the experiment.
BMSC Co-Culture Experiments
BMSC were isolated and maintained as described previously, followedby co-culture in the absence of FCS with either kidney fragmentsor microdissected tubule segments. Individual tubule segmentswere identified using the criteria of Chabardes et al.38 andisolated by the method of Velazquez et al.39 In some experiments,BMSC were harvested from enhanced green fluorescent protein–expressingmice, and kidney fragments were obtained from enhanced greenfluorescent protein–negative mice to allow determinationof the source of the epithelial colonies. Immunocytochemistryof the cultured cells was performed by fixation with 4% paraformaldehyde/PBSfollowed by primary antibody addition (anti–pan-cytokeratin;1:200; Sigma) or anti-E-cadherin (1:100; BD transduction). Insome experiments, BMSC were cultured in the presence of HGF(40 ng/ml), erythropoietin (10 IU/ml), EGF (20 ng/ml), retinoicacid (0.1 to 1 µM), serum from mice that had undergoneprevious ischemia/reperfusion (10% vol/vol), or combinationsof these in an attempt to induce mesenchymal-epithelial transformationof the cultured BMSC.
In Vitro Assay for Cell Death
Immortalized MPT cells at 60 to 70% confluence were exposedto IMDM with or without cisplatin (5 µg/ml) and with orwithout BMSC-CM for 24 to 48 h. BMSC-CM was obtained as describedpreviously and then diluted 1:1 with fresh medium containing10% FCS (final concentration 50% CM, 5% FCS). Control cellsreceived fresh IMDM with 5% FCS. After 24 h, adherent cellswere harvested by trypsinization and viable cell numbers weredetermined by manual counting of cells that excluded trypanblue. Cell death was determined in both adherent and floatingnonpermeabilized cells by using flow cytometry to detect cellswith surface phosphatidylserine expression (annexin V binding)and propidium iodide (BD Bioscience) uptake using a commercialkit (BD Pharmingen).
This work was supported by an American Heart Association EstablishedInvestigator award (0440005N) and National Institutes of Healthaward (DK66216) to L.G.C. and a Emmy-Noether stipend from theDeutsche Forschungsgemeinschaft to R.S.
Parts of this work were presented at the November, 2006 annualmeeting of the American Society of Nephrology; November 14 through19, 2006; San Diego, CA; and published in abstract form (J AmSoc Nephrol 17: 622A, 2006).
We thank Heino Velazquez for help in the tubule microdissectionexperiments.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
Supplemental information for this article is available onlineat http://www.jasn.org/.
See the related editorial, "The Continuing Story of Renal Repairwith Stem Cells," on pages 2423–2424
Orlic D, Kajstura J, Chimenti S, Jakoniuk I, Anderson SM, Li B, Pickel J, McKay R, Nadal-Ginard B, Bodine DM, Leri A, Anversa P: Bone marrow cells regenerate infarcted myocardium.
Nature 410
: 701
–705, 2001[CrossRef][Medline]
Prockop DJ, Gregory CA, Spees JL: One strategy for cell and gene therapy: Harnessing the power of adult stem cells to repair tissues.
Proc Natl Acad Sci U S A 100
: 11917
–11923
Cogle CR, Yachnis AT, Laywell ED, Zander DS, Wingard JR, Steindler DA, Scott EW: Bone marrow transdifferentiation in brain after transplantation: A retrospective study.
Lancet 363
: 1432
–1437, 2004[CrossRef][Medline]
Yamada M, Kubo H, Kobayashi S, Ishizawa K, Numasaki M, Ueda S, Suzuki T, Sasaki H: Bone marrow-derived progenitor cells are important for lung repair after lipopolysaccharide-induced lung injury.
J Immunol 172
: 1266
–1272, 2004[Abstract/Free Full Text]
Tondreau T, Lagneaux L, Dejeneffe M, Delforge A, Massy M, Mortier C, Bron D: Isolation of BM mesenchymal stem cells by plastic adhesion or negative selection: Phenotype, proliferation kinetics and differentiation potential.
Cytotherapy 6
: 372
–379, 2004[CrossRef][Medline]
Goncalves R, Lobato da Silva C, Cabral JM, Zanjani ED, Almeida-Porada G: A Stro-1(+) human universal stromal feeder layer to expand/maintain human bone marrow hematopoietic stem/progenitor cells in a serum-free culture system.
Exp Hematol 34
: 1353
–1359, 2006[CrossRef][Medline]
Alhadlaq A, Mao JJ: Mesenchymal stem cells: Isolation and therapeutics.
Stem Cells Dev 13
: 436
–448, 2004[CrossRef][Medline]
Jiang Y, Jahagirdar BN, Reinhardt RL, Schwartz RE, Keene CD, Ortiz-Gonzalez XR, Reyes M, Lenvik T, Lund T, Blackstad M, Du J, Aldrich S, Lisberg A, Low WC, Largaespada DA, Verfaillie CM: Pluripotency of mesenchymal stem cells derived from adult marrow.
Nature 418
: 41
–49
Poulsom R, Forbes SJ, Hodivala-Dilke K, Ryan E, Wyles S, Navaratnarasah S, Jeffery R, Hunt T, Alison M, Cook T, Pusey C, Wright NA: Bone marrow contributes to renal parenchymal turnover and regeneration.
J Pathol 195
: 229
–235, 2001[CrossRef][Medline]
Gupta S, Verfaillie C, Chmielewski D, Kim Y, Rosenberg ME: A role for extrarenal cells in the regeneration following acute renal failure.
Kidney Int 62
: 1285
–1290, 2002[CrossRef][Medline]
Rochefort GY, Delorme B, Lopez A, Herault O, Bonnet P, Charbord P, Eder V, Domenech J: Multipotential mesenchymal stem cells are mobilized into peripheral blood by hypoxia.
Stem Cells 24
: 2202
–2208
Lin F, Cordes K, Li L, Hood L, Couser WG, Shankland SJ, Igarashi P: Hematopoietic stem cells contribute to the regeneration of renal tubules after renal ischemia-reperfusion injury in mice.
J Am Soc Nephrol 14
: 1188
–1199, 2003[Abstract/Free Full Text]
Duffield JS, Park KM, Hsiao LL, Kelley VR, Scadden DT, Ichimura T, Bonventre JV: Restoration of tubular epithelial cells during repair of the postischemic kidney occurs independently of bone marrow-derived stem cells.
J Clin Invest 115
: 1743
–1755, 2005[CrossRef][Medline]
Togel F, Hu Z, Weiss K, Isaac J, Lange C, Westenfelder C: Administered mesenchymal stem cells protect against ischemic acute renal failure through differentiation-independent mechanisms.
Am J Physiol Renal Physiol 289
: F31
–F42, 2005[Abstract/Free Full Text]
Cantley L: Adult stem cells in the repair of the injured renal tubule.
Nat Clin Pract Nephrol 1
: 1
–11, 2005[Medline]
Togel F, Isaac J, Westenfelder C: Hematopoietic stem cell mobilization-associated granulocytosis severely worsens acute renal failure.
J Am Soc Nephrol 15
: 1261
–1267, 2004[Abstract/Free Full Text]
Kale S, Karihaloo A, Clark PR, Kashgarian M, Krause DS, Cantley LG: Bone marrow stem cells contribute to repair of the ischemically injured renal tubule.
J Clin Invest 112
: 42
–49
Morigi M, Imberti B, Zoja C, Corna D, Tomasoni S, Abbate M, Rottoli D, Angioletti S, Benigni A, Perico N, Alison M, Remuzzi G: Mesenchymal stem cells are renotropic, helping to repair the kidney and improve function in acute renal failure.
J Am Soc Nephrol 15
: 1794
–1804, 2004[Abstract/Free Full Text]
Herrera MB, Bussolati B, Bruno S, Fonsato V, Romanazzi GM, Camussi G: Mesenchymal stem cells contribute to the renal repair of acute tubular epithelial injury.
Int J Mol Med 14
: 1035
–1041, 2004[Medline]
Lange C, Togel F, Ittrich H, Clayton F, Nolte-Ernsting C, Zander AR, Westenfelder C: Administered mesenchymal stem cells enhance recovery from ischemia/reperfusion-induced acute renal failure in rats.
Kidney Int 68
: 1613
–1617, 2005[CrossRef][Medline]
Togel F, Weiss K, Yang Y, Hu Z, Zhang P, Westenfelder C: Vasculotropic, paracrine actions of infused mesenchymal stem cells are important to the recovery from acute kidney injury.
Am J Physiol Renal Physiol 295
: F1626
–F1635, 2007
Nakagami H, Morishita R, Maeda K, Kikuchi Y, Ogihara T, Kaneda Y: Adipose tissue-derived stromal cells as a novel option for regenerative cell therapy.
J Atheroscler Thromb 13
: 77
–81, 2006[Medline]
Lin F, Moran A, Igarashi P: Intrarenal cells, not bone marrow-derived cells, are the major source for regeneration in postischemic kidney.
J Clin Invest 115
: 1756
–1764, 2005[CrossRef][Medline]
Caplan AI, Dennis JE: Mesenchymal stem cells as trophic mediators.
J Cell Biochem 98
: 1076
–1084, 2006[CrossRef][Medline]
Groh ME, Maitra B, Szekely E, Koc ON: Human mesenchymal stem cells require monocyte-mediated activation to suppress alloreactive T cells.
Exp Hematol 33
: 928
–934, 2005[CrossRef][Medline]
Miller SB, Martin DR, Kissane J, Hammerman MR: Hepatocyte growth factor accelerates recovery from acute ischemic renal injury in rats.
Am J Physiol 266
: F129
–F34, 1994[Medline]
Miller SB, Martin DR, Kissane J, Hammerman MR: Insulin-like growth factor I accelerates recovery from ischemic acute tubular necrosis in the rat.
Proc Natl Acad Sci U S A 89
: 11876
–11880, 1992[Abstract/Free Full Text]
Asahara T, Chen D, Tsurumi Y, Kearney M, Rossow S, Passeri J, Symes JF, Isner JM: Accelerated restitution of endothelial integrity and endothelium-dependent function after phVEGF165 gene transfer.
Circulation 94
: 3291
–3302, 1996[Abstract/Free Full Text]
Karihaloo A, Karumanchi SA, Cantley WL, Venkatesha S, Cantley LG, Kale S: Vascular endothelial growth factor induces branching morphogenesis/tubulogenesis in renal epithelial cells in a neuropilin-dependent fashion.
Mol Cell Biol 25
: 74417448
, 2005
Sotiropoulou PA, Perez SA, Gritzapis AD, Baxevanis CN, Papamichail M: Interactions between human mesenchymal stem cells and natural killer cells.
Stem Cells 24
: 74
–85
Sinha D, Wang Z, Price VR, Schwartz JH, Lieberthal W: Chemical anoxia of tubular cells induces activation of c-Src and its translocation to the zonula adherens.
Am J Physiol Renal Physiol 284
: F488
–F497
Peister A, Mellad JA, Larson BL, Hall BM, Gibson LF, Prockop DJ: Adult stem cells from bone marrow (MSCs) isolated from different strains of inbred mice vary in surface epitopes, rates of proliferation, and differentiation potential.
Blood 103
: 1662
–1668
Meyerrose TE, De Ugarte DA, Hofling AA, Herrbrich PE, Cordonnier TD, Shultz LD, Eagon JC, Wirthlin L, Sands MS, Hedrick MA, Nolta JA: In vivo distribution of human adipose-derived mesenchymal stem cells in novel xenotransplantation models.
Stem Cells 25
: 220
–227, 2007[Abstract/Free Full Text]
Theise ND, Badve S, Saxena R, Henegariu O, Sell S, Crawford JM, Krause DS: Derivation of hepatocytes from bone marrow cells in mice after radiation-induced myeloablation.
Hepatology 31
: 235
–240, 2000[CrossRef][Medline]
Yuen PS, Dunn SR, Miyaji T, Yasuda H, Sharma K, Star RA: A simplified method for HPLC determination of creatinine in mouse serum.
Am J Physiol Renal Physiol 286
: F1116
–F1119
Tsikas D, Wolf A, Frolich JC: Simplified HPLC method for urinary and circulating creatinine.
Clin Chem 50
: 201
–203, 2004[Free Full Text]
Chabardes D, Imbert-Teboul M, Montegut M, Clique A, Morel F: Catecholamine sensitive adenylate cyclase activity in different segments of the rabbit nephron.
Pflugers Arch 361
: 9
–15, 1975[CrossRef][Medline]
Velazquez H, Silva T, Andujar E, Desir GV, Ellison DH, Greger R: The distal convoluted tubule of rabbit kidney does not express a functional sodium channel.
Am J Physiol Renal Physiol 280
: F530
–F539, 2001[Abstract/Free Full Text]
Related Articles
This Month's Highlights
J. Am. Soc. Nephrol. 2007 18: A12.
[Full Text][PDF]
The Continuing Story of Renal Repair with Stem Cells
Enyu Imai and Hirotsugu Iwatani
J. Am. Soc. Nephrol. 2007 18: 2423-2424.
[Full Text][PDF]
This article has been cited by other articles:
B. Bi, J. Guo, A. Marlier, S. R. Lin, and L. G. Cantley Erythropoietin expands a stromal cell population that can mediate renoprotection
Am J Physiol Renal Physiol,
October 1, 2008;
295(4):
F1017 - F1022.
[Abstract][Full Text][PDF]
M. Morigi, M. Introna, B. Imberti, D. Corna, M. Abbate, C. Rota, D. Rottoli, A. Benigni, N. Perico, C. Zoja, et al. Human Bone Marrow Mesenchymal Stem Cells Accelerate Recovery of Acute Renal Injury and Prolong Survival in Mice
Stem Cells,
August 1, 2008;
26(8):
2075 - 2082.
[Abstract][Full Text][PDF]
F. Togel, Y. Yang, P. Zhang, Z. Hu, and C. Westenfelder Bioluminescence imaging to monitor the in vivo distribution of administered mesenchymal stem cells in acute kidney injury
Am J Physiol Renal Physiol,
July 1, 2008;
295(1):
F315 - F321.
[Abstract][Full Text][PDF]
R. Schmitt and L. G. Cantley The impact of aging on kidney repair
Am J Physiol Renal Physiol,
June 1, 2008;
294(6):
F1265 - F1272.
[Abstract][Full Text][PDF]
E. Imai and H. Iwatani The Continuing Story of Renal Repair with Stem Cells
J. Am. Soc. Nephrol.,
September 1, 2007;
18(9):
2423 - 2424.
[Full Text][PDF]
Top
Introduction
RESULTS
and PGE2 can inhibit lymphocyte activation and thereby suppress the inflammatory responses that might otherwise augment the tubular injury and increase rates of apoptosis.25,26,31 Therefore, it is likely that multiple factors that are secreted by the BMSC are acting in concert to limit the acute injury associated with cisplatin exposure. In agreement with this, we found that BMSC-CM fractions of <10, 10 to 30, and >30 kD all are capable of stimulating renal tubular cell colony growth in vitro (data not shown). 
