Combination therapy of mesenchymal stem cells and serelaxin effectively attenuates renal fibrosis in obstructive nephropathy

The FASEB Journal • Research Communication

Combination therapy of mesenchymal stem cellsand serelaxin effectively attenuates renal fibrosisin obstructive nephropathy

Brooke M. Huuskes,* Andrea F. Wise,* Alison J. Cox,† Ee X. Lim,* Natalie L. Payne,‡

Darren J. Kelly,† Chrishan S. Samuel,§,1 and Sharon D. Ricardo*,1

*Department of Anatomy and Developmental Biology and §Department of Pharmacology, MonashUniversity, Clayton, Victoria, Australia; †Department of Medicine, University of Melbourne, St. Vincent’sHospital, Melbourne, Australia; and ‡Australia Regenerative Medicine Institute (ARMI) and MonashUniversity, Clayton, Victoria, Australia

ABSTRACT Chronic kidneydisease (CKD) results fromthe development of fibrosis, ultimately leading to end-stage renal disease (ESRD). Although human bonemarrow–derived mesenchymal stem cells (MSCs) can ac-celerate renal repair following acute injury, the establish-ment of fibrosis during CKD may affect their potential toinfluence regeneration capacity. Here we tested the novelcombination of MSCs with the antifibrotic serelaxin to re-pair and protect the kidney 7 d post-unilateral ureteralobstruction (UUO), when fibrosis is established. MaleC57BL6 mice were sham-operated or UUO-inured (n =4–6) and received vehicle, MSCs (1 3 106), serelaxin(0.5 mg/kg per d), or the combination of both. In vivotracing studies with luciferin/enhanced green fluorescentprotein (eGFP)–taggedMSCs showed specific localizationin the obstructed kidney where they remained for 36 h.Combination therapy conferred significant protectionfrom UUO-induced fibrosis, as indicated by hydroxypro-line analysis (P < 0.001 vs. vehicle, P < 0.05 vs. MSC orserelaxin alone). This was accompanied by preservedstructural architecture, decreased tubular epithelial injury(P < 0.01 vs. MSCs alone), macrophage infiltration, andmyofibroblast localization in the kidney (both P < 0.01 vs.vehicle). Combination therapy also stimulated matrix met-alloproteinase (MMP)-2 activity over either treatment alone(P < 0.05 vs. either treatment alone). These results suggestthat the presence of an antifibrotic in conjunction withMSCsameliorates establishedkidneyfibrosis andaugmentstissue repair to a greater extent than either treatmentalone.—Huuskes, B. M., Wise, A. F., Cox, A. J., Lim, E. X.,Payne, N. L., Kelly, D. J., Samuel, C. S., Ricardo, S. D.Combination therapy of mesenchymal stem cells andserelaxin effectively attenuates renal fibrosis in obstructivenephropathy. FASEB J. 29, 000–000 (2015). www.fasebj.org

Key Words: interstitial kidney injury • macrophage in-filtration • transforming growth factor-b • myofibroblast dif-ferentiation • collagen deposition • matrix metalloproteinase

REGARDLESS OF ETIOLOGY, fibrosis results from aberrantwound healing and represents a hallmark of all forms ofCKD, leading to renal dysfunction and ultimately ESRD,requiring renal replacement therapy (1). Histologically,ESRD manifests as glomerulosclerosis, vascular sclerosis,and tubulointerstitial fibrosis which results from in-terstitial expansion due to collagen accumulation (2).This accumulation of extracellular matrix (ECM) in thetubulointerstitium ultimately leads to tubular atrophy,capillary loss, and podocyte depletion, which directlycorresponds to proteinuria and the loss of renal func-tion (3). Although controlling diet and exercise, alongwith pharmaceutical interventions, such as antihyper-tensive drugs, can somewhat slow the progression ofrenal failure (4), there are currently no effective curesfor CKD, leaving a high demand for renal replacementtherapy (5).

Human MSCs have been proposed as a plausible ther-apeutic tool as they are easily isolated from the bone mar-row and can be cultured and expanded into adequatenumbers for therapeutic administrationdue to their abilityto extensively proliferate in vitro (6). In numerous modelsof kidney injury, the administration of MSCs has beenreported to protect from acute injury and enhance re-nal regeneration (7). Although the exact mechanismsof repair are yet to be elucidated, there is substantialevidence thatMSCs elicit repair through endocrine andparacrine mechanisms (8–10). The ability of MSCs torelease a plethora of growth factors and chemokines,including IGF-1, hepatocyte growth factor, and VEGF,may be responsible for the reduction in endogenousapoptosis and promotion of mitogenesis, vasculo-genesis, and angiogenesis (11–13) and hence offers

Abbreviations: a-MEM, a-minimal essential media; a-SMA,a-smooth muscle actin; CKD, chronic kidney disease; CUK,contralateral unobstructed kidney; ECM, extracellular matrix;eGFP, enhanced green fluorescent protein; ESRD, end-stagerenal disease; FBS, fetal bovine serum; fluc, firefly luciferase;Kim-1, kidney injury molecule-1; MMP, matrix metalloproteinase;MSC, mesenchymal stem cell; RXFP-1, relaxin family peptidereceptor-1; sr, steradian; UUO, unilateral ureteral obstruction

1 Correspondence: S.D.R., Department of Anatomy and De-velopmental Biology, Monash University, Clayton, Victoria,3800, Australia. E-mail: [email protected]; C.S.S.,Department of Pharmacology, Monash University, Clayton,Victoria, 3800, Australia. E-mail: [email protected]: 10.1096/fj.14-254789

0892-6638/15/0029-0001 © FASEB 1

The FASEB Journal article fj.14-254789. Published online November 13, 2014.

http://www.fasebj.org

mailto:[email protected]

mailto:[email protected]

advantages over using single-factor therapy to treat kidneydisease (14). Furthermore, preclinical studies have dem-onstrated that MSCs posses immunomodulatory proper-ties, resulting in the up-regulation of anti-inflammatorycytokines (15) and inhibition of proinflammatory cyto-kines (11), including TGF-b (16), specifically in a modelof fibrotic kidney disease (17). Recent studies have fur-ther demonstrated that the presence of MSCs, but nottheir conditioned media, can attenuate macrophage in-filtration into the injured kidney (18), with the solublefactors released by MSCs responsible for altering macro-phage polarization from an inflammatory M1 phenotypeto an anti-inflammatory M2 state (19). MSCs can alsomodulate MMPs, which correlate to a decrease in acutekidney injury–induced collagen deposition (19). De-spite their regenerative capacity in settings of acutekidney disease, the development of fibrosis in chronicdisease models has hampered MSC survival and limitsintegration of transplanted cells into the host tissue,thus significantly impacting on MSC-dependent repairand restoration of function (20, 21). Therefore, weproposed that combining MSC treatment with an agentthat has antifibrotic properties may enhance MSC-mediated tissue repair.

The endogenous hormone relaxin is primarily pro-duced by the corpus luteum during pregnancy and is re-sponsible for softening the pelvic ligaments and wideningthe birth canal for parturition (22); it also plays a criticalrole in maintaining collagen homeostasis in non-reproductive organs (23).Of the 3 relaxin geneproductsidentified in humans, human gene-2 relaxin (relaxin-2)and its functional relaxin ortholog in rodents are thepredominant forms of stored and circulating relaxin(23). The administration of exogenous serelaxin(recombinant human relaxin-2) in animals models hasbeen shown to abrogate aging (24, 25) and disease-induced kidney fibrosis (26–28), including that inducedby UUO (29), while fundamentally improving renalfunction in clinical trials (23, 30). The antifibroticeffects of serelaxin, as shown in preclinical studies, areprimarilymediated though its binding to Relaxin FamilyPeptide Receptor (RXFP)-1 and ability to down-regulateSmad2 phosphorylation, resulting in the inhibition ofTGF-b1-induced myofibroblast differentiation (31, 32).This in turn suppresses myofibroblast-induced collagensynthesis while also promoting collagen degradationthrough increasing iNOS availability, thereby up-regulating MMPs (33). Serelaxin additionally pro-motes angiogenesis and vasodilation while decreasingapoptosis (23).

These findings suggest that serelaxin may attenuatescarring, which ideally would aid cell-based therapies in thesetting of chronic disease. This was demonstrated ina porcine myocardial infarct model whereby mouse skel-etal myoblasts, engineered to produce serelaxin, signifi-cantly decreased pathologic collagen deposition whilepromoting myoblast viability, MMP activity, angiogenesis,and cardiac function compared with the effects of myo-blasts alone (34).

On the basis of the preceding findings, the currentstudy investigated the combined potential of MSCs andserelaxin therapy in a fibrotic model of UUO-inducedkidney injury. The results obtained demonstrated that

MSCs home to the kidney and in conjunction with ser-elaxin, ameliorated pathologic fibrosis.

MATERIALS AND METHODS

Experimental animals

Male C57BL/6J mice weighing 20–25 g were obtained fromMonash University Animal Services (Clayton, VIC, Australia).Mice were housed at Monash Animal Research Laboratories ina controlled environment on a 14h light, 10 h dark schedulewithfree access to water and lab chow.Micewere separated into sham-operated (n = 4) or UUO-induced injury groups (n = 24); wherethe latter received treatments of vehicle alone (administered atthe time of injury onset; n = 6), MSC alone (administered at thetime of injury onset; n = 6), serelaxin alone (kindly provided byCorthera Incorporated, San Carlos, CA, USA; a subsidiary ofNovartis AG, Basel, Switzerland; administered 2 d prior to injuryto ensure that it was actively circulating at the time of injury onset;n = 6), or the combination of MSC and serelaxin (as describedpreviously; n = 6). All experiments were approved by theMonashUniversity Animal Ethics Committee, which adheres to theAustralian Code of Practice for the Care andUse of Animals forScientific Purposes.

Surgical procedures

All surgery was performed under 2% isoflurane anesthesia(Abbott Australasia Pty Limited, Kurnell, NSW, Australia), ad-ministered via inhalation.

Osmotic pump implantation

Micro-osmotic pumps (ALZET Models 1007D Alzet, Cupertino,CA, USA) were loaded with 0.5 mg/kg per d of serelaxin,allowing a constant infusion rate of 0.5 ml/h for up to 9 d (giventhe drug reservoir of each pump). Serelaxin is bioactive inmice,and this rate of infusion produces circulating serelaxin levels ofapproximately 20–40 ng/ml, a concentration that mimics thoseobserved during pregnancy (35). In relevant groups, 2 d prior toUUO surgery, pumps were subcutaneously implanted as pre-viously described (29) and maintained until 7 d post injury.

Unilateral ureteral obstruction

UUO surgery was performed whereby the left ureter was visual-ized via a flank incision and ligated using double tracks with 5.0surgical silk. The right contralateral unobstructed kidney (CUK)served as a biologic control. In sham-operated animals, the leftureter was manipulated only. A subgroup of UUO-injured micereceived vehicle (PBS), and further subgroups of UUO-injuredmice received MSCs (as detailed below) either with or withoutserelaxinpretreatment.Afterflank incision and/or ligation,micewere sutured and monitored until they were fully recovered(approximately 10–15 min in each case) and maintained until7 d postinjury.

MSC culture

Human bone marrow-derived MSCs purchased from the TulaneCentre for Stem Cell Research and Regenerative Medicine(Tulane University, New Orleans, LA, USA) were cultured aspreviously described (36). Briefly, cells were cultured in media

2 Vol. 29 February 2015 HUUSKES ET AL.The FASEB Journal x www.fasebj.org


comprising of a-minimal essential media (MEM), supplementedwith 20% fetal bovine serum (FBS) and antibiotics (10,000 U/mlpenicillinG,10mg/ml streptomycin sulfate, 200mML-glutamine,all from Invitrogen /GIBCO, Life Technologies, VIC, Australia).Cells were seeded at a density of 60 cells/cm in tissue cultureflasks (BD Falcon, North Ryde, NSW, Australia) with mediachangedevery3–4d.At70%confluence(after7–10d inculture),cells were trypsinized, pelleted, and resuspended in 1 ml ofmedia.Of these, 13106 cells were taken and suspended in120mlof PBS (Invitrogen) for delivery into the renal vein of micewith/without UUO injury.

Bioluminescence imaging

To trace MSCs in vivo, a subgroup of animals (n = 4–6) wereadministered 1 3 106 MSCs expressing enhanced greenfluorescent protein (eGFP) and firefly luciferase (fluc) (36) i.v.at the timeof shamorUUOsurgery. To image these cells in vivo,anesthetized animals were injected intraperitoneally with 200 mlof D-luciferin (15 mg/ml in PBS; VivoGlo Luciferin; Promega,San Luis Obispo, CA, USA) at 1, 24, and 36 h post-cell injection.Mice were imaged with the IVIS 200 system (Xenogen, Alameda,CA, USA), and the fluc luminescent signal was captured andanalyzed as photons/ s per cm2 (Living Image 3.2; Xenogen).After 36 h, animals were culled and both left and right kidneyswere harvested and imaged ex vivo.

Confirmation of MSCs in the kidney

Kidneys from sham-operated and UUO-injured mice that hadreceived eGFP expressing MSCs were removed at 36 h, and ge-nomic DNA was extracted and then analyzed for the eGFP geneand b-actin housekeeping gene by PCR as previously described(36). Primer pairs were eGFP: 59-TACGGCAAGCTGACCCT-GAAGTTC-39, 59-CGTCGTCCTTGAAGAAGAATGGT; GCG-39;b-actin: 59-GGCACCACACCTTCTACAA-39, 59-CTGCTGCTGAA-GCTCTAG-39. The PCR products were electrophoresed in 2%agarose gels and bands visualized under UV light.

Histopathology

After 7 d of UUO-induced injury, transverse, paraffin-embeddedkidneys sectioned at 4 mm were stained with hematoxylin andeosin to assess changes in kidney structure.

Immunohistochemistry

Immunohistochemical staining was performed on 4 mmparaffinsections to assess tubular epithelial injury via the detection ofkidney injury molecule (Kim)-1 expression, and the localizationof mature macrophages and myofibroblasts using F4/80 anda-smoothmuscle actin (a-SMA) antibodies, respectively. Sectionsunderwent microwave antigen retrieval technique in a sodiumcitrate buffer as previously described (37). Sections were in-cubated with primary antibodies against Kim-1 (1:200 dilution;R&D Systems, Minneapolis, MN, USA), F4/80 (1:200 dilution;AbD Serotec, Oxford, United Kingdom), and a-SMA (1:250 di-lution; Dako, North Sydney, NSW, Australia) for 90 min. Un-conjugated anti-sera was detected by incubating sections witha biotinylated anti-IgG (for Kim-1 and F4/80; Vector Laboratories,Burlingame, CA,USA) or horseradish peroxidase-conjugated anti-IgG (for a-SMA; Dako) for 30 min. Binding was visualized with anavidin-biotincomplex(Vector), followedby3,39-diaminobenzidine(Dako). Sections were counterstained with hematoxylin. Toquantify these factors, 5 consecutivenonoverlappingfields at3400

were viewed (Provis AX70; Olympus, Tokyo, Japan) and acquired(DP70 color camera; Olympus) from 3 nonserial mouse kidneysections.All imageswere semiquantifiedbasedon theproportionalarea and intensity of brown staining, determined using imageanalysis (ImageJ 1.46r; National Institute ofHealth, Bethesda,MD,USA). Results were expressed as the percentage of stained arearelative to the total area. Appropriate isotype controls were con-sistently negative.

Immunofluorescence

Immunofluorescence was conducted on frozen kidney sectionsembedded in optimal cutting temperature (TissueTek, Tokyo,Japan) to confirm the presence of myofibroblasts and visualizethe colocalization of macrophage with type IV collagen as pre-viously described (38). Sections were incubated with primaryantibodies against F4/80 (1:200; Serotec) and goat anti-humancollagen type IV (1:400; Southern Biotech, Birmingham, AL,USA) for 1 h, respectively. Secondary antibodies for F4/80 (AlexaFluor 555; Invitrogen) and type IV collagen (Alexa Fluor 647;both at dilutions of 1:1000; Invitrogen) were added to sectionsfor 30 min. Additional sections were incubated with antia-SMAprimary antibody (1:100; Sigma-Aldrich, St. Louis, MO,USA) for 1 h followed by Alex Fluor 555 (1:1000; Invitrogen)for 30 min. All sections were counterstained with DAPI(1:10,000; Invitrogen) for 5 min and mounted with Fluores-cent Mounting Medium (Dako). Sections were analyzed witha fluorescent microscope (Provis AX70) and photographed(F-view II digital camera).

Analysis of kidney collagen content

Hydroxyproline analysis

Equivalent frozen portions of kidney tissue (containing equalparts of cortex and medulla) were assessed for total collagencontent, determined by hydroxyproline content as previouslydescribed (24, 29).Hydroxyproline valueswere thenconverted tocollagen content by multiplying by a factor of 6.94 (as hydroxy-proline represents ;14.4% of the amino acid composition ofcollagen) and further expressed as a percentage of the dry tissueweight to yield collagen concentration.

Picrosirius red staining

Paraffin-embedded kidney sections were stained with picrosiriusred to visualize thedistributionof interstitial collagenasdescribedpreviously (39). Five consecutive nonoverlapping fields at 3400were viewed (Provis AX70) and acquired (DP70 color camera)from each mouse kidney. All images were semiquantified basedon the proportional area and intensity of red staining (collagen),determinedusing imageanalysis (ImageJ 1.46r;National Instituteof Health). Results were expressed as the percentage of stainedarea relative to the total area.

Gelatin zymography

Gelatin zymography was used to detect changes in latent andactive forms ofMMP-2 in kidney tissues from treatment groups aspreviously described (29). MMPs were extracted as detailed be-fore (40) and assessed on 7.5% acrylamide gels containing1 mg/ml gelatin. Gelatinolytic activity was visualized as clearbands. Densitometry of active MMP-2 bands was performed(GS710 Densitometer; Bio-Rad Laboratories, Gladesville, NSW,Australia) and the relative mean6 SEM density was graphed.

COMBINATION THERAPY AMELIORATES KIDNEY FIBROSIS 3

33P in situ hybridization for TGF-b

Changes of mRNA expression for TGF-b were detected by in situhybridization andquantified asdescribedpreviously (41). Briefly,33P antisense RNA probes for mouse TGF-b were generated thenhybridized onto 4 mm paraffin sections and incubated in a hu-midified chamber at 60°C for 14–16 h. Sections were then rinsedinanRNasebuffer and treatedwith 150mg/mlofRNaseA for 1hat 37°C. Sections were dehydrated in ethanol and, once air-dried,exposed to Kodak Biomax MR Autoradiography film for 4 d atroom temperature. A sense probe was used as a negative control(data not shown). Films were quantified by densitometry usingthe Micro Computing Imaging Device (Imaging Research, St.Catherines, ON, Canada) as previously described (41). A stan-dard curve of optical vs. radioactive density was created usingAmersham 14C, which were coexposed with the hybridized sec-tions. A threshold was set to discriminate against areas of highareas of probe labeling intensity in the area of the cortex only.Probe binding in the area of the medulla was not used forquantification.

Serelaxin effects on MSC proliferation

MSCs were seeded at 53 103 per well in 96-well plates containingMSCmedia as described previously or media supplemented with1ng/mlof serelaxin,a concentration typicallyobserved inhumanpregnancies (42)andatwhich it effectively inhibits renalfibroblastproliferation (31). After 72 h, cell proliferation was determinedusing the CellTitre 96 AQueous One Solution Proliferation Assay(Promega, Fitchburg, WI, USA) according to manufacturer’sinstructions. Absorbance was read at 490 nm using the Bench-mark Plus microplate spectrophotometer (Bio-Rad). Resultswere expressed as a percentage of proliferation relative to thepositive control (MSCs in media).

Serelaxin effects on MSC migration

MSC migration was performed as previously described (43).Briefly, migration was performed in 6.5 mm transwell plates con-taining8mmpore inserts (CorningCostar,Cambridge,MA,USA).Inserts were coated with 0.1%wt/vol gelatin for 1 h at 37°C before53 105 of stimulated (1 ng/mlTNF-a, 24 h)MSCswere seeded inthe insert. Migration media [RPMI 1640 (Invitrogen) substitutedwith 0.25% bovine serum albumin (Sigma-Aldrich)] alone orsupplemented with 30% FBS were added to the bottom chamberto serve as negative and positive controls, respectively. Varyingconcentrationsof serelaxin(0.1, 1, 10, and100ng/ml)wereaddedto the migration media and plated in the bottom chamber todetermine if serelaxin was a chemoattractant forMSCs. After 24 hof incubation at 37°C and 5% CO2, cells remaining on the upperside of the inserts weremechanically removed. Inserts were washedwith PBS and fixed in 2.5% glutaraldehyde before being stainedwith hematoxylin (Monash Histology Platform [Monash Univer-sity]) and mounted on glass slides. Ten random fields from eachfilter were captured and the total number of cells counted in eachfield. Each experiment was performed in duplicate. Data wereexpressed as a percentage of cells relative to the negative control.

Statistical analysis

Statistical analysis was performed usingGraphPad Prism softwareversion 6.0d (GraphPad Software Incorporated, San Diego, CA,USA). Data were analyzed via 1-way ANOVA with a Tukey’s posthoc test for intergroup comparisons. Cell culture data were ana-lyzed using a Student’s t test. All data were expressed as mean6SEM, and P, 0.05 was considered statistically significant.

RESULTS

MSCs home to the UUO-injured kidney

Whole-bodybioluminescence imagingwasused to confirmthat MSCs home to the site of UUO injury. MSCs trans-duced with a lentiviral vector encoding for eGFP and flucwere transplanted into sham-operated or UUO-injuredmice. Bioluminescent measurements were made on thedorsal surface of mice 1, 24, and 36 h post-MSC trans-plantationafter intraperitoneal administrationof luciferin.UUO-injured mice showed accumulation over the areacorresponding to the kidney as early as 1 h post-MSCtransplantation (Fig. 1A), and sham-operated animalsshowed a distinct accumulation of MSCs in the lungs (Fig.1B). By 24–36 h post-MSC transplantation, the majority ofthe fluc signal was observed in the left injured kidney (Fig.1A). After 36 h,mice were culled and kidneys were excisedand imagedalone(Fig. 1A), confirming thatMSCswereonlypresent in the left, UUO-injured kidney and not the CUK.Sham-operated control mice displayed a declining fluc sig-nalover36h in theareaof the lungsonly(Fig.1B), indicatingthat MSCs were likely trapped in the pulmonary capillaries.Additionally, theflucsignalover theareaof the lungs insham-operated animals rapidly diminished and was hardly detect-able at 36 h (0.0462250 3 107 photons/s/cm2/steradian[sr]) compared with the fluc signal in injured mice at thesame time point (2.2419345 3 107 photons/s/cm2/sr),demonstrating thatMSCs diminishedmore rapidly in theabsence of an inflammatory signal (Fig. 1C).

As bioluminescence may not be sensitive enough todetect smallnumbersofMSCs,PCRwasused to identify theeGFP gene in both sham-operated and MSC-treated kid-neys 36 h post-MSC injection (Fig. 1D). PCR results dem-onstrated that the eGFP gene was present in 5 of the 6UUO-injured kidneys. No GFP, however, was detected inthe kidneys of sham-operated controls, confirming thatMSCs only home to the kidney in the presence of in-flammation and injury (Fig. 1D).

Combination therapy preserves structural integrity ofkidney tubules

The histology of all UUO-injured mouse kidneys showedprogressive hydronephrosis characterized by dilation andcollapse of tubules after 7 d of sustainedUUO(Fig. 2D–O).Sham-operated kidneys displayed regular alignment ofproximal and distal tubules with minimal interstitial space(Fig. 2A–C). Ablation of the renal parenchyma was mostobvious in UUO-injured kidneys treated with vehicle,which was accompanied by an increase in inflammatorycell infiltrate and the loss of the brush border on the lu-minal surface of the proximal tubules, as well as tubulardilation and atrophy (Fig. 2D–F). The loss of tubule in-tegrity is a direct cause of increased ECM accumulation,which directly correlates to a decline in renal function(44). Histologic differences between vehicle-treated,MSC, and serelaxin-treated groups were indistinguish-able (Fig. 2D–L). Conversely, treatment with the combi-nation of MSCs and serelaxin led to the amelioration ofinterstitial matrix accumulation and resulted in the atten-uation of inflammatory cell infiltrate. Renal architecture



was maintained as evident by the presence of numerousproximal tubules in the corticomedullary region of thekidney (Fig. 2I–J). The right CUK served as a biologiccontrol in the UUO-injured mice and was histologicallycomparable to sham-operated kidneys.

Combination therapy attenuates established fibrosis inobstructive nephropathy

The extent to which kidney fibrosis had developed after7 dof sham-operationorUUO-induced injurywith vehicle,MSCs, serelaxin, andcombination therapywasdeterminedby hydroxyproline analysis of total collagen concentrationand morphometric analysis of picrosirius red-stained in-terstitial collagens (Fig. 3). Mice treated with vehicleresulted in a 2.5-fold increase in collagen accumulation(percent collagen content per dry weight tissue) after7 d post-UUO injury compared with sham-operated ani-mals (P, 0.001; Fig. 3C). A trend toward reduced collagenaccumulation was measured in kidneys from MSC- orserelaxin-treated animals, and the combination of MSCsand serelaxin significantly ameliorated UUO-inducedcollagen deposition by approximately one-third (P ,0.001 vs. vehicle; Fig. 3C). Furthermore, combinationtherapy conferred a significant reduction in fibrosis com-pared with either treatment alone (P, 0.05; Fig. 3C). Theaccumulation of interstitial collagen, consisting of types I,III, and V, in sham and UUO-injured kidneys were

visualized using picrosirius red staining (Fig. 3A) andsemiquantified using morphometric analysis (Fig. 3B).Interstitial collagen in UUO-injured kidneys treated withvehicle significantly increased (by.2-fold) comparedwithsham-operated controls (P , 0.01; Fig. 3B), which wasagain marginally affected by either therapy alone. How-ever, the combined effect of MSCs and serelaxin resultedin approximately a two-thirds decrease in interstitial colla-gen compared with UUO-injured animals treated withvehicle injection (P , 0.01). Together these results dem-onstrate that in this model of renal fibrosis, the combina-tion of MSCs and serelaxin attenuated establishedtubulointerstitial fibrosis to a greater extent than eithertherapy alone.

Gelatinase activity is increased withcombination therapy

Gelatin zymography of protein extracts from each of thegroups studied was used to detect changes in the 72 kDagelatinase MMP-2 of all animals 7 d post-UUO (Fig. 4).Although both latent and active MMP-2 bands were ob-served (Fig. 4A), densitometric analysis was performed ononly the active form of MMP-2 (Fig. 4B) to gauge endog-enous activity of kidney MMP-2 between the groups stud-ied. Sham-operated animals displayed minimal MMP-2activity comparedwith that detected fromall UUO-injuredmouse kidneys in which activity was significantly increased

Figure 1. MSCs home to and localize in the kidney after UUO injury. Luciferin/GFP + MSCs were injected i.v. into mice aftersham operation or UUO induction, and the bioluminescence fluc signal was examined at 1, 24, and 36 h postinjection on thedorsal surface of all animals. A) MSCs home to the left injured kidney as early as 1 h postinjection and remained for up to 36 hafter ureteral ligation. B) MSCs home to the lung in sham animals after surgery but are hardly detectable at 36 h post-MSCinjection. C) Quantification of the bioluminescence signals at 1, 24, and 36 h from the lung area of sham-operated mice and overthe area of the kidney in UUO-injured mice. D) PCR of genomic DNA extracted from whole kidneys show that the GFP gene waspresent in MSC-treated groups but not sham-operated controls.


(P , 0.001 vs. sham-operated). No differences in renalMMP-2 activity were observed between vehicle, MSC, orserelaxin therapy. However,MMP-2 activity was elevated toa greater extent in animals that were treated with thecombined therapy of MSCs and serelaxin compared withUUO-injured kidneys treated with either therapy alone(P , 0.01 combination vs. vehicle, P , 0.05 combinationvs.MSC and serelaxin).

Effects of combination therapy on markers of renalinjury, macrophage localization, and myofibroblastaccumulation in the obstructed kidney

Immunohistologic staining of paraffin-embedded kidneysectionswithan anti–Kim-1 allowed visualizationof injuredproximal tubular epithelial cells. Sham-operated kidneyshadminimal Kim-1 staining compared to all UUO-injured

Figure 2. Combination therapy maintains structural integrity of kidney tubules. Representative photomicrographs illustratinghematoxylin and eosin–stained kidney tissue 7 d post-UUO. A) Sham-operated kidneys display normal histoarchitecture withdefined cortex, medullar, and papilla regions and (B, C) regular alignment of proximal (P) and distal (D) tubules with minimalinterstitial space. D) After 7 d of UUO, there was noticeable shrinkage of the renal parenchyma with (E) obvious dilation and (F)flattening of tubular epithelial cells (arrow), inflammatory cell infiltrate (arrowhead), and interstitial space widening as a resultfrom collagen deposition (asterisk). G) MSC and (J) serelaxin therapy had little protection of renal architecture, and (H, I, K, L)inflammatory infiltrate and tubular dilation was similar to that seen in vehicle-treated kidneys with minimal proximal tubulepreservation. M) Combination therapy protects from parenchyma ablation and (N, O) maintains structural integrity of thekidney, evident by the presence of the proximal tubule brush border (arrow). Scale bars, 500 mm (A, D, G, J, M) and 100 mm(B, C, E, F, H, I, K, L, N, O).



kidneys (Fig. 5A). Semiquantification of Kim-1 stainingindicated that MSC treatment alone had no effect onUUO-induced tubule injury, whereas serelaxin and com-bination therapy resulted in a significant decrease inKim-1expression (both P, 0.01 vs.MSC; Fig. 5F).

The localization of F4/80-positive macrophages wassignificantly increased in animals treated with vehicle,MSC, and serelaxin compared with sham-operated coun-terparts (P, 0.01 vehicle and serelaxin vs. sham-operated,P , 0.05 MSCs vs. sham-operated; Fig. 5B). This injury-induced accumulation of macrophages was attenuatedwith the administration ofMSCs (P, 0.05 vs. vehicle) andwith combination therapy (P , 0.01 vs. vehicle; Fig. 5G).Immunofluorescence staining (Fig. 5C) established thecolocalization of macrophages (green) with type IV colla-gen (red), which were both increased in obstructed kid-neys post-UUO. A reduction in collagen IV staining wasobserved in combination-treated animals compared withvehicle-treated animals (Fig. 5C) consistent with the hy-droxyproline (Fig. 3B) and picrosirius red morphometry(Fig. 3C) data obtained.

a-SMA, a marker of myofibroblast differentiation, wasonly located within the smooth muscle arterioles of sham-operated kidneys (Fig. 5D). Paralleling hydroxyprolineanalysis, UUO-induced injury significantly increased myo-fibroblast localization in kidneys treated with vehicle (P,0.001 vs. sham-operated; Fig. 5H), with a trend towarda decreased myofibroblast accumulation observed in

kidneys treated with MSC alone. Although treatment withserelaxin alone resulted in a significant decrease in myo-fibroblast accumulation (P, 0.05 vs. vehicle), the combi-nation of MSCs and serelaxin reduced myofibroblastlocalization to a greater extent (P , 0.01 vs. vehicle). Im-munofluorescent detection of a-SMA protein expressionconfirmed the UUO-induced up-regulation of myofibro-blast accumulation, which was lowered by combinationtherapy. Appropriate isotype controls were tested andconsistently negative.

TGF-b mRNA expression is not significantlydecreased with combination therapy

The mRNA expression of TGF-b was investigated using insitu hybridization, as TGF-b is a therapeutic target of bothserelaxin (31, 32, 39) and MSCs (16). Autoradiographicimages (Fig. 6A) showed the dispersion of TGF-b mRNAthroughout the kidney. Bright and dark field images (Fig.6B) revealed diffuse levels of TGF-b in sham-operatedanimals throughout the kidney, and localization was spe-cific to the proximal tubules of UUO-injured mice.Quantification of the cortex in autoradiography images(Fig. 6C) revealed that TGF-b levels were significantlyhigher in all treatment groups compared with sham-operated animals (P , 0.001). However, treatment withMSCs significantly reduced kidney TGF-b mRNA levels

Figure 3. Combination therapy ameliorates collagen deposition in obstructed kidneys. A) Representative images of kidneysections stained with picrosirius red to visualize interstitial collagen types I, III, and V in sham-, vehicle-, MSC-, serelaxin-, andcombination-treated kidneys after 7 d of UUO. Interstitial collagen was quantified (B) using morphometric analysis. Vehicletreatment resulted in significantly higher collagen staining compared to sham-operated animals (P , 0.01) and combination-treated kidneys (P, 0.01). C) Hydroxyproline analysis of equivalent kidney tissue portions (n = 4–6) was used to quantify changesin collagen concentration, a measure of fibrosis. Vehicle treatment resulted in a significant increase in collagen accumulationcompared with sham-operated animals (P , 0.0001). This increase in aberrant collagen was significantly prevented with thecombined effects of MSCs and serelaxin (P , 0.001); however, only a modest reduction in collagen was detected when MSC orserelaxin therapy was used alone. Results are expressed as the mean 6 SEM relative collagen staining per field (%) (B) or totalcollagen concentration (percent collagen content per dry weight tissue) (C) and both are expressed as a ratio to that in the shamoperated group (expressed as 1). Comb, combination therapy; Rln, serelaxin; Veh, vehicle. *P , 0.05, **P , 0.01, ***P , 0.001.Scale bar, 100 mm.


compared with that in vehicle- and serelaxin-treated ani-mals (P, 0.05 vs. vehicle and serelaxin alone).

Serelaxin increases MSC proliferation and migrationin vitro

To determine if the significant reduction in collagenobservedwith combination treatmentwasdue toa serelaxin-dependent increase in MSC proliferation and/or mi-gration, MSCs were cultured with varying dose of serelaxinfor 72 h. The presence of serelaxin (1 ng/ml) increasedMSC proliferation (Fig. 7) by ;14% (P , 0.05 vs. MSCproliferation alone). Additionally, TNF-a–stimulatedMSCshad an increased migration capacity in the presence ofserelaxin at 10 ng/ml (P , 0.01 vs. negative control) and100 ng/ml (P , 0.001 vs. negative control) after 24 h.

DISCUSSION

This study provides the first report demonstrating the im-proved reparative and protective effects of MSCs whencoadministered with serelaxin in the setting of interstitial

kidneyfibrosis. Combination therapy significantly reducedECM expansion and improved renal architecture after7 d of UUO, which was associated with a decrease in in-terstitial and total collagen levels. This enhanced ability ofcombination therapy to prevent the progression offibrosisappeared to result fromanup-regulationofMMP-2activity,which can stimulate the breakdown of the basementmembrane and interstitial collagens. Of further signifi-cance, we demonstrated that i.v. administered MSCs werecapable of homing to the UUO-injured kidney as early as1 h post-cell transplantation. Taken together, our findingsindicated thatMSCs and serelaxin synergistically protectedthe kidney from fibrotic injury to a greater extent thaneither therapy alone.

In numerous experimental models of disease, the routeof MSC administration has remained of some concern asvenous systemic transplantation has been associated withaccumulation of MSCs in the small capillary beds of thelungs and not in the target organ (45). To overcome this,the use of a vasodilator at the time of MSC administration(46), or administration via the arterial circulation haspreviously been employed (47). However, consistent withprevious work in an ischemia-reperfusion model (19), ourresults demonstrate that after venous transplantation,MSCeither home to the UUO-injured kidney directly or locateto both the lung and the kidney as early as 1 h post-transplantation.Regardless ofMSC locationwithin thefirsthour, by 24 h post-transplantation, the majority of cellsmigrated to the injured kidney, where they remained forup to 36 h. This indicates that MSCs have the ability tomigrate out of the lungs and home to the injured kidneywhen they are present for a longer time, and in largernumbers (47) than their sham-operated counterparts,likely due to the inflammatory signals that are producedfrom the constant insult of UUO-injury. It was noted,however, that MSC therapy alone had no significant effecton collagen accumulation in treated kidneys. Poor survivalof MSCs in the hypoxic microenvironment of the injuredkidney could offer an explanation. Given the well-documented angiogenic and vasorelaxant properties ofserelaxin (48), we hypothesize that when used in combi-nation with MSCs, serelaxin may have aided MSC migra-tion out of the lungs and into the obstructed kidney.Consistent with serelaxin’s ability to improve the numberof viable stem cells following administration into the is-chemic porcine heart (34), we also demonstrated thatserelaxin increased MSC proliferation and migration ca-pacity in vitro. Hence, serelaxin may be important foraugmenting MSC proliferation and migration in vivo,which would allow for increased numbers of MSCs to me-diate their therapeutic effects, in combination-treatedmice in the model evaluated, which is consistent with ser-elaxin’s previously documented ability to promote endo-thelial cell proliferation and migration (49).

Fibrosis is the final common pathway of ESRD, regard-less of its etiology, andUUO is used as amodel offibrosis toidentify novel therapeutic targets as it mimics human pa-thology (50). An end point of 7 d was selected to examinethe effect of combination therapy as published data hadpreviously shown that constant delivery of exogenous ser-elaxin could ameliorate establishedfibrosis by up to;30%after 9 d of UUO-induced injury (29). Serelaxin treatmentalone induced a trend toward decreased collagen in this

Figure 4. Combination therapy increases MMP-2 activity in theobstructed kidney. Gelatin zymography of equal amounts ofkidney protein extract (10 mg; n = 4–6) was used to determineMMP-2 expression and activity. Shown is a representativegelatin zymograph (A) of latent and active MMP-2 levels ineach treatment group studied. A separate zymograph of 2–3samples per group produced similar results. Densitometry ofthe active MMP-2 bands was performed. The relative mean 6SEM OD of MMP-2 levels (B), expressed as a ratio of that in thesham-operated group (which was expressed as 1) is alsoshown. There was a significant increase in active MMP-2 levelsin UUO-injured kidneys compared to sham-operated animals(P , 0.001) (B). Although MSC and serelaxin therapy alonehad no significant effect on active MMP-2 levels, combinationtherapy resulted in an up-regulation of MMP-2 to a greaterextent than vehicle injections (P , 0.01) in addition to eithertherapy alone (P , 0.05). Comb, combination therapy; OD,optical density; Rln, serelaxin; Veh, vehicle. *P , 0.05, **P ,0.01, ***P , 0.001.



Figure 5. Changes in Kim-1, macrophage, and myofibroblast accumulation in UUO-injured kidneys. Representativephotomicrographs show cortical sections from sham-operated and UUO-injured animals (n = 4–6) 7 d post injury. A, F) Kim-1expression was elevated in the kidneys of vehicle- and MSC-treated groups (P , 0.001 vs. sham-operated) yet was significantlydecreased in serelaxin- and combination-treated animals (P , 0.01 vs. MSC). B, G) Mature macrophages were detected with F4/80 primary antibody and showed that macrophage localization was increased in UUO-injured kidneys treated with vehicle (P ,0.01 vs. sham), MSCs (P , 0.05 vs. sham), and serelaxin (P , 0.01 vs. sham). Both MSC (P , 0.05 vs. vehicle) and combinationtherapy (P , 0.01 vs. vehicle) significantly reduced macrophage accumulation within the injured kidney. C) Immunofluorescentstaining shows the relationship between macrophage localization (green) and type IV collagen (red) accumulation. D, E)Myofibroblast accumulation was detected with an anti–a-SMA antibody. H) UUO-induced myofibroblast accumulation within thekidney was significantly decreased with serelaxin treatment (P, 0.05 vs. vehicle) and combination therapy (P, 0.01 vs. vehicle).Results are expressed as the mean 6 SEM relative percentage area and expressed as a ratio to that in the sham operated group(expressed as 1). Comb, combination therapy; Rln, serelaxin; Veh, vehicle. *P , 0.05, **P , 0.01, ***P , 0.001. Scale bars,100 mm.


study, likely due to the severity of fibrosis being strain-dependent (50, 51) with previous studies utilizingC57B6Jx129SV mice to demonstrate that serelaxin couldsignificantly prevent renal fibrosis post-UUO (29). How-ever, given its well-known antifibrotic and antiapoptoticactions (23–25, 29, 31–34, 39, 52), we propose that ser-elaxin was able to create a more favorable environment,with less tissue damage and fibrosis, in which MSCs couldnot only survive but have an increased migration andproliferative capacity to mediate their antifibrotic and re-parative effects, resulting in a greater reduction of fibrosiswith combination treatment.

The combined treatment of MSCs and serelaxin wasassociated with less obvious parenchymal ablation andre-epithelialization of injured proximal tubules. This im-proved renal architecture was closely associated witha significant reduction of total collagen accumulation.Consistent with this, Kim-1 is a marker of proximal tubularepithelial cell damage, and expression levels have beencorrelated to the degree of interstitial injury, not only inanimalmodelsoffibrosis (53, 54)but also inhumankidneybiopsy samples from numerous pathologies (55). Kim-1expression was found to be elevated in the superficialkidney cortex (56) in response to UUO-injury. ComparedwithMSCtreatment,Kim-1was significantly reduced in thekidneys of serelaxin and combination-treated mice, con-firming that serelaxin itself was able to reduce the extent ofepithelial injury.

In our study, macrophages were present in significantlyhigher numbers after UUO-induced injury treated withvehicle. Consistent with previous findings (16, 57), MSCadministration resulted in decreased macrophage accu-mulation, which correlated with a decrease in TGF-b1mRNA.Conversely, administrationof serelaxinhadnoeffecton the localizationofmacrophages in theobstructedkidney,as reported previously (58). The involvement of macro-phages in the development of kidney fibrosis is well docu-mented; however, studies now suggest that macrophages

have a more complex role in organ repair and re-generation (59, 60). Furthermore, MSC-mediated repairwas shown to be dependent on the presence of macro-phages in a sepsis model of organ injury (61). Our find-ing that serelaxin did not affect TGF-b1mRNAor proteinexpression (29) in the UUOmodel studied suggests thatit is more likely to affect TGF-b1 signal transduction atthe level of Smad2 phosphorylation (pSmad2) andtranslocation to the nucleus (31, 32), rather thanmacrophage-mediated TGF-b1 expression itself. Hence,our findings suggest that MSCs and serelaxin workthrough different but synergistic mechanisms to reduceTGF-b-mediated renal fibrosis.

Serelaxin’s renoprotective effects are exerted throughbinding to its receptor, RXFP1, and altering the down-stream signaling pathways of the TGF-b/Smad2 axis toinhibitmyofibroblast transformation in the kidney (31, 32)and other organs (62, 63), rather than altering the pro-duction of TGF-b itself (29). Consistent with our findings,MSCs have also been described to alter the balance of fi-brosis by limiting macrophage infiltration and down-regulating TGF-b production via paracrine mechanisms(16) and more specifically altering macrophage polariza-tion in an acutemodel of kidney injury (19). Nevertheless,we postulate that the most likely explanation for the com-bined ability of MSC and serelaxin to prevent UUO-induced renal collagen deposition/fibrosis was via theircombined ability to up-regulate MMP-2 activity over eithertreatment alone.

Fibrosis occurs due to an imbalance in the rate of col-lagen synthesis (of which at least half arises through fi-broblast proliferation and differentiation from within thekidney) (64), and collagen degradation controlled byMMPs (65). The interdependent effect of MSCs and ser-elaxin in ameliorating collagen accumulation is likely dueto their combined effect on MMP activity. Gelatin zymo-graphic analysis revealed that MMP-2 activity was signifi-cantly up-regulated in all UUO-injured groups compared

Figure 6. TGF-b in situ hybridization. Representative in situ hybridization (A) autoradiographs and (B) bright field images forTGF-b localization in sham-operated and UUO-treated kidneys. There was little mRNA for sham-operated animals; however, at7 d post-UUO an increase in mRNA in all treated groups was observed (all P , 0.001 vs. sham-operated). C) Quantification of insitu hybridization shows that mRNA expression of TGF-b was significantly decreased with MSC treatment (P, 0.05 vs. vehicle andserelaxin treatment). Data are expressed as 6 SEM of the nCi/g of TGF-b mRNA expression. Comb, combination therapy; Rln,serelaxin; Veh, vehicle. *P , 0.05, **P , 0.01. Scale bars, 100 mm.



with sham-operated counterparts, consistent with previousdata investigating diabetes-induced fibrosis (66). Inter-estingly, MMP-2 activity was further increased in kidneyextractsderived fromcombination-treatedmicecomparedwith that measured from vehicle, MSC, or serelaxin treat-ment alone. These data are consistent with a study con-ducted in a swinemyocardial infarctmodel, demonstratingthat in conjunction with serelaxin, stem cells were able toup-regulate MMP activity and significantly reduce cardiacfibrosis (34). Conversely, individual treatment of MSCsalone or serelaxin alone resulted in kidney MMP-2 activitycomparable to that of UUO-injured mice treated with

vehicle.Of note, changes inMMP-2 activity appeared to beinversely proportional to changes in total collagen accu-mulation from the groups analyzed in this study. OnlywhenMSCsand serelaxin therapywere combinedwere themost striking reductions in collagen observed, correlatingwith a significant increase in MMP-2 activity.

Although there is limited literature on the ability ofMSCs to produce or influence MMP activity in vivo, MSCscan stimulate MMP production from cultured cardiacfibroblasts (67), and increased MMP-2 activity may be themain mechanism for reducing lung fibrosis when admin-istering umbilical cordMSCs (68).Of further significance,the migration of MSCs into injured tissues is reliant onMMP-2–dependent degradation of the basement mem-brane (69). Increases in serelaxin-induced MMP activityhas been demonstrated both in vitro and in vivo in a num-ber of fibrotic pathologies including those of the kidney(23, 29, 52, 70). The pathway that influences the up-regulation of MMPs by serelaxin has only recently beenelucidated and is associated with the increased availabilityof NO (33). This may provide a possible mechanismexplaining the synergistic effectsof combination therapy asthe delivery of an NO donor enhanced MSC homing ina model of liver fibrosis (71).

Tubulointerstitial fibrosis not only is a common pheno-typeobserved inESRDbutalso is a commoncause for renalallograft rejection (72) and as such, a number of ongoingclinical trials are investigating the use of MSCs in a trans-plant setting (73). Although long-term studies are yet to beconcluded, early reportsdemonstrated thatMSC therapy iswell tolerated as a form of induction therapy (74) and alsoreduced fibrosis when subclinical rejection had com-menced (75). It is, however, important to note that thetimingofMSCadministration is akey factor indeterminingthe implications of this therapy (76). Recent clinical trialswith serelaxin have mainly been conducted in patientssuffering from acute heart failure. Results from phase IIband phase III trials consistently demonstrated that the ad-ministration of serelaxin is well tolerated by patients, re-lieved dyspnea, prevented the in-hospital worsening ofheart failure, and decreased mortality (30, 77). A trial isunderway tounderstand thepharmacokineticsof serelaxinwhen administered to patients with severe renal disease(clinicaltrials.gov: NTC01875523). Given our findings, it ishopeful that future clinical trials may incorporate thecombined therapy of MSCs and serelaxin to treat kidneydisease that results from pathologic fibrosis.

In conclusion, our study provides evidence for a novelcombined cellular-based therapy that aided kidney repairand limited fibrosis progression associated with UUO-induced injury. The synergistic benefit observed by amal-gamating MSCs and serelaxin therapy is likely mediatedby increased MMP activity, allowing significant reversalof aberrant collagen accumulation and enhanced pro-liferation and homing ofMSCs. The results from this studydemonstrate that MSC therapy, combined with an agentthat has antifibrotic properties, may provide improvedtherapeutic options for patients with ESRD.

The authors acknowledge technical support from Dr.Stephen Firth from Monash Micro Imaging, and the staff atMonash Histology Platform, Monash University. B.M.H. isa recipient of a Kidney Health Australia Medical and Science

Figure 7. Serelaxin increases MSC proliferation and migrationcapacity in vitro. A) MSCs were cultured with 1 ng/ml ofserelaxin for 72 h and then proliferation was measured usinga colorimetric assay. Absorbance was measured at 490 nm anddemonstrated that proliferation of MSCs was significantlyincreased in the presence of 1 ng/ml of serelaxin whencompared with MSC proliferation in media alone. Results areexpressed as the mean 6 SEM relative percentage proliferationfrom 4 independent experiments and expressed as a ratioto that of MSC proliferation alone (expressed as 100%).Migration studies were performed in the presence of 0.1, 1,10, and 100 ng/ml of serelaxin to determine if serelaxin wasa chemoattractant for TNF-a-stimulated MSCs. B) There wasno difference in the migration capacity of MSCs between thenegative control (serum-free media) and 0.1 and 1 ng/ml ofserelaxin. Only when 10 or 100 ng/ml of serelaxin was presentdid MSC migration increase compared with the negativecontrol (P , 0.001) and lower concentrations of serelaxin(P , 0.001 for 100 ng/ml vs. 0.1 and 1 ng/ml; P , 0.01 for10 ng/ml vs. 0.1 and 1 ng/ml). Results are expressed as themean 6 SEM relative percentage migration from 4 indepen-dent experiments and expressed as a ratio to that of MSCmigration in the presence of the negative control (expressedas 100%). Rln, serelaxin. *P , 0.05, **P , 0.01, ***P , 0.001.


Research Scholarship. C.S.S. is a recipient of a NationalHealth and Medical Research Council (NHMRC) of AustraliaSenior Research Fellowship (APP1041766).

REFERENCES

1. Zeisberg, M., and Neilson, E. G. (2010) Mechanisms of tubu-lointerstitial fibrosis. J. Am. Soc. Nephrol. 21, 1819–1834

2. Anders, H.-J., Vielhauer, V., and Schlondorff, D. (2003) Che-mokines and chemokine receptors are involved in the resolutionor progression of renal disease. Kidney Int. 63, 401–415

3. Bohle, A., Mackensen-Haen, S., and von Gise, H. (1987) Signif-icance of tubulointerstitial changes in the renal cortex for theexcretory function and concentration ability of the kidney:a morphometric contribution. Am. J. Nephrol. 7, 421–433

4. Roy, L., White-Guay, B., Dorais, M., Dragomir, A., Lessard, M.,and Perreault, S. (2013) Adherence to antihypertensive agentsimproves risk reduction of end-stage renal disease. Kidney Int. 84,570–577

5. Perico, N., Codreanu, I., Schieppati, A., and Remuzzi, G. (2005)Prevention of progression and remission/regression strategiesfor chronic renal diseases: can we do better now than five yearsago? Kidney Int. Suppl. 68, S21–S24

6. Humphreys, B. D., and Bonventre, J. V. (2008) Mesenchymalstem cells in acute kidney injury. Annu. Rev. Med. 59, 311–325

7. Wise, A. F., and Ricardo, S. D. (2012) Mesenchymal stem cells inkidney inflammation and repair. Nephrology (Carlton) 17, 1–10

8. Morigi, M., Imberti, B., Zoja, C., Corna, D., Tomasoni, S., Abbate,M., Rottoli, D., Angioletti, S., Benigni, A., Perico, N., Alison, M.,and Remuzzi, G. (2004) Mesenchymal stem cells are renotropic,helping to repair the kidney and improve function in acute renalfailure. J. Am. Soc. Nephrol. 15, 1794–1804

9. Lange, C., Togel, F., Ittrich, H., Clayton, F., Nolte-Ernsting, C.,Zander, A. R., and Westenfelder, C. (2005) Administered mes-enchymal stem cells enhance recovery from ischemia/reperfusion-induced acute renal failure in rats. Kidney Int. 68,1613–1617

10. Bi, B., Schmitt, R., Israilova, M., Nishio, H., and Cantley, L. G.(2007) Stromal cells protect against acute tubular injury via anendocrine effect. J. Am. Soc. Nephrol. 18, 2486–2496

11. Togel, F., Weiss, K., Yang, Y., Hu, Z., Zhang, P., andWestenfelder, C. (2007) Vasculotropic, paracrine actions ofinfused mesenchymal stem cells are important to the recoveryfrom acute kidney injury. Am. J. Physiol. Renal Physiol. 292,F1626–F1635

12. Imberti, B., Morigi, M., Tomasoni, S., Rota, C., Corna, D.,Longaretti, L., Rottoli, D., Valsecchi, F., Benigni, A., Wang, J.,Abbate, M., Zoja, C., and Remuzzi, G. (2007) Insulin-like growthfactor-1 sustains stem cell mediated renal repair. J. Am. Soc.Nephrol. 18, 2921–2928

13. Morigi, M., Introna, M., Imberti, B., Corna, D., Abbate, M., Rota,C., Rottoli, D., Benigni, A., Perico, N., Zoja, C., Rambaldi, A.,Remuzzi, A., and Remuzzi, G. (2008) Human bone marrowmesenchymal stem cells accelerate recovery of acute renal injuryand prolong survival in mice. Stem Cells 26, 2075–2082

14. Medina, A., Fernandez-Fuente, M., Carbajo-Perez, E., Santos, F.,Amil, B., Rodrıguez, J., and Crespo, M. (2002) Insulin-like growthfactor I administration in young rats with acute renal failure.Pediatr. Nephrol. 17, 1005–1012

15. Semedo, P., Palasio, C. G., Oliveira, C. D., Feitoza, C. Q.,Gonçalves, G. M., Cenedeze, M. A., Wang, P. M. H., Teixeira,V. P. A., Reis, M. A., Pacheco-Silva, A., and Camara, N. O. S.(2009) Early modulation of inflammation by mesenchymalstem cell after acute kidney injury. Int. Immunopharmacol. 9,677–682

16. Ueno, T., Nakashima, A., Doi, S., Kawamoto, T., Honda, K.,Yokoyama, Y., Doi, T., Higashi, Y., Yorioka, N., Kato, Y., Kohno,N., and Masaki, T. (2013) Mesenchymal stem cells ameliorateexperimental peritoneal fibrosis by suppressing inflammationand inhibiting TGF-b1 signaling. Kidney Int. 84, 297–307

17. Liu, Y.-L., Wang, Y.-D., Zhuang, F., Xian, S.-L., Fang, J.-Y., Su, W.,and Zhang, W. (2012) Immunosuppression effects of bonemarrow mesenchymal stem cells on renal interstitial injury in ratswith unilateral ureteral obstruction. Cell. Immunol. 276, 144–152

18. Xing, L., Cui, R., Peng, L., Ma, J., Chen, X., Xie, R.-J., and Li, B.(2014) Mesenchymal stem cells, not conditioned medium, con-tribute to kidney repair after ischemia-reperfusion injury. StemCell Res Ther 5, 101–113

19. Wise, A. F., Williams, T. M., Kiewiet, M. B. G., Payne, N. L.,Siatskas, C., Samuel, C. S., and Ricardo, S. D. (2014) Humanmesenchymal stem cells alter macrophage phenotype andpromote regeneration via homing to the kidney followingischemia-reperfusion injury. Am. J. Physiol. Renal Physiol. 306,F1222–F1235

20. Eun, L. Y., Song, H., Choi, E., Lee, T. G., Moon, D. W., Hwang,D., Byun, K. H., Sul, J. H., and Hwang, K. C. (2011) Implantedbone marrow-derived mesenchymal stem cells fail to metaboli-cally stabilize or recover electromechanical function in infarctedhearts. Tissue Cell 43, 238–245

21. van Dijk, A., Naaijkens, B. A., Jurgens, W. J. F. M., Nalliah,K., Sairras, S., van der Pijl, R. J., Vo, K., Vonk, A. B. A., vanRossum, A. C., Paulus, W. J., van Milligen, F. J., and NiessenH. W. (2011) Reduction of infarct size by intravenous injection ofuncultured adipose derived stromal cells in a rat model is de-pendent on the time point of application. Stem Cell Res. (Amst.)7, 219–229

22. Zhao, L., Roche, P. J., Gunnersen, J. M., Hammond, V. E.,Tregear, G. W., Wintour, E. M., and Beck, F. (1999) Mice withouta functional relaxin gene are unable to deliver milk to their pups.Endocrinology 140, 445–453

23. Samuel, C. S., Hewitson, T. D., Unemori, E. N., and Tang, M. L.K. (2007) Drugs of the future: the hormone relaxin. Cell. Mol. LifeSci. 64, 1539–1557

24. Samuel, C. S., Zhao, C., Bond, C. P., Hewitson, T. D., Amento,E. P., and Summers, R. J. (2004) Relaxin-1-deficient mice developan age-related progression of renal fibrosis. Kidney Int. 65,2054–2064

25. Danielson, L. A., Welford, A., and Harris, A. (2006) Relaxinimproves renal function and histology in aging Munich Wistarrats. J. Am. Soc. Nephrol. 17, 1325–1333

26. Garber, S. L., Mirochnik, Y., Brecklin, C. S., Unemori, E. N.,Singh, A. K., Slobodskoy, L., Grove, B. H., Arruda, J. A., andDunea, G. (2001) Relaxin decreases renal interstitial fibrosisand slows progression of renal disease. Kidney Int. 59,876–882

27. Garber, S. L., Mirochnik, Y., Brecklin, C., Slobodskoy, L., Arruda,J. A. L., and Dunea, G. (2003) Effect of relaxin in two models ofrenal mass reduction. Am. J. Nephrol. 23, 8–12

28. Yoshida, T., Kumagai, H., Suzuki, A., Kobayashi, N., Ohkawa, S.,Odamaki, M., Kohsaka, T., Yamamoto, T., and Ikegaya, N. (2012)Relaxin ameliorates salt-sensitive hypertension and renal fibrosis.Nephrol. Dial. Transplant. 27, 2190–2197

29. Hewitson, T. D., Ho, W. Y., and Samuel, C. S. (2010) Antifibroticproperties of relaxin: in vivo mechanism of action in experi-mental renal tubulointerstitial fibrosis. Endocrinology 151,4938–4948

30. Teerlink, J. R., Cotter, G., Davison, B. A., Felker, G. M.,Filippatos, G., Greenberg, B. H., Ponikowski, P., Unemori, E.,Voors, A. A., Adams, K. F., Jr., Dorobantu, M. I., Grinfeld, L. R.,Jondeau, G., Marmor, A., Masip, J., Pang, P. S., Werdan, K.,Teichman, S. L., Trapani, A., Bush, C. A., Saini, R.,Schumacher, C., Severin, T. M., and Metra, M., RELAXin inAcute Heart Failure (RELAX-AHF) Investigators. (2013) Serelaxin,recombinant human relaxin-2, for treatment of acute heartfailure (RELAX-AHF): a randomised, placebo-controlled trial.Lancet 381, 29–39

31. Heeg, M. H. J., Koziolek, M. J., Vasko, R., Schaefer, L., Sharma,K., Muller, G. A., and Strutz, F. (2005) The antifibrotic effects ofrelaxin in human renal fibroblasts are mediated in part by in-hibition of the Smad2 pathway. Kidney Int. 68, 96–109

32. Mookerjee, I., Hewitson, T. D., Halls, M. L., Summers, R. J.,Mathai, M. L., Bathgate, R. A. D., Tregear, G. W., and Samuel,C. S. (2009) Relaxin inhibits renal myofibroblast differentiationvia RXFP1, the nitric oxide pathway, and Smad2. FASEB J. 23,1219–1229

33. Chow, B. S. M., Chew, E. G. Y., Zhao, C., Bathgate, R. A. D.,Hewitson, T. D., and Samuel, C. S. (2012) Relaxin signalsthrough a RXFP1-pERK-nNOS-NO-cGMP-dependent pathway toup-regulate matrix metalloproteinases: the additional involve-ment of iNOS. PLoS ONE 7, e42714



34. Formigli, L., Perna, A.-M., Meacci, E., Cinci, L., Margheri, M.,Nistri, S., Tani, A., Silvertown, J., Orlandini, G., Porciani, C.,Zecchi-Orlandini, S., Medin, J., and Bani, D. (2007) Paracrineeffects of transplanted myoblasts and relaxin on post-infarctionheart remodelling. J. Cell. Mol. Med. 11, 1087–1100

35. Samuel, C. S., Zhao, C., Bathgate, R. A. D., Bond, C. P., Burton,M. D., Parry, L. J., Summers, R. J., Tang, M. L. K., Amento, E. P.,and Tregear, G. W. (2003) Relaxin deficiency in mice is associ-ated with an age-related progression of pulmonary fibrosis.FASEB J. 17, 121–123

36. Payne, N. L., Sun, G., McDonald, C., Layton, D., Moussa, L.,Emerson-Webber, A., Veron, N., Siatskas, C., Herszfeld, D., Price,J., and Bernard, C. C. A. (2013) Distinct immunomodulatory andmigratory mechanisms underpin the therapeutic potential ofhuman mesenchymal stem cells in autoimmune demyelination.Cell Transplant. 22, 1409–1425

37. Cochrane, A. L., Kett, M. M., Samuel, C. S., Campanale, N. V.,Anderson, W. P., Hume, D. A., Little, M. H., Bertram, J. F., andRicardo, S. D. (2005) Renal structural and functional repair ina mouse model of reversal of ureteral obstruction. J. Am. Soc.Nephrol. 16, 3623–3630

38. Alikhan, M. A., Jones, C. V., Williams, T. M., Beckhouse, A. G.,Fletcher, A. L., Kett, M. M., Sakkal, S., Samuel, C. S., Ramsay,R. G., Deane, J. A., Wells, C. A., Little, M. H., Hume, D. A., andRicardo, S. D. (2011) Colony-stimulating factor-1 promotes kid-ney growth and repair via alteration of macrophage responses.Am. J. Pathol. 179, 1243–1256

39. Samuel, C. S., Cendrawan, S., Gao, X.-M., Ming, Z., Zhao, C.,Kiriazis, H., Xu, Q., Tregear, G. W., Bathgate, R. A. D., and Du,X.-J. (2011) Relaxin remodels fibrotic healing following myo-cardial infarction. Lab. Invest. 91, 675–690

40. Woessner, J. F., Jr. (1995) Quantification of matrix metal-loproteinases in tissue samples. Methods Enzymol. 248, 510–528

41. Zhang, Y., Edgley, A. J., Cox, A. J., Powell, A. K., Wang, B.,Kompa, A. R., Stapleton, D. I., Zammit, S. C., Williams, S. J.,Krum, H., Gilbert, R. E., and Kelly, D. J. (2012) FT011, a newanti-fibrotic drug, attenuates fibrosis and chronic heart failure inexperimental diabetic cardiomyopathy. Eur. J. Heart Fail. 14,549–562

42. Goldsmith, L. T., and Weiss, G. (2009) Relaxin in human preg-nancy. Ann. N. Y. Acad. Sci. 1160, 130–135

43. Ponte, A. L., Marais, E., Gallay, N., Langonne, A., Delorme, B.,Herault, O., Charbord, P., and Domenech, J. (2007) The in vitromigration capacity of human bone marrow mesenchymal stemcells: comparison of chemokine and growth factor chemotacticactivities. Stem Cells 25, 1737–1745

44. Liu, Y. (2006) Renal fibrosis: new insights into the pathogenesisand therapeutics. Kidney Int. 69, 213–217

45. Burst, V. R., Gillis, M., Putsch, F., Herzog, R., Fischer, J. H., Heid,P., Muller-Ehmsen, J., Schenk, K., Fries, J. W. U., Baldamus, C. A.,and Benzing, T. (2010) Poor cell survival limits the beneficialimpact of mesenchymal stem cell transplantation on acute kid-ney injury. Nephron, Exp. Nephrol. 114, e107–e116

46. Schrepfer, S., Deuse, T., Reichenspurner, H., Fischbein, M. P.,Robbins, R. C., and Pelletier, M. P. (2007) Stem cell trans-plantation: the lung barrier. Transplant. Proc. 39, 573–576

47. Togel, F., Yang, Y., Zhang, P., Hu, Z., and Westenfelder, C.(2008) Bioluminescence imaging to monitor the in vivo distri-bution of administered mesenchymal stem cells in acute kidneyinjury. Am. J. Physiol. Renal Physiol. 295, F315–F321

48. Jeyabalan, A., Shroff, S. G., Novak, J., and Conrad, K. P.(2007) The vascular actions of relaxin. Adv. Exp. Med. Biol.612, 65–87

49. Bitto, A., Irrera, N., Minutoli, L., Calo, M., Lo Cascio, P., Caccia,P., Pizzino, G., Pallio, G., Micali, A., Vaccaro, M., Saitta, A.,Squadrito, F., and Altavilla, D. (2013) Relaxin improves multiplemarkers of wound healing and ameliorates the disturbed healingpattern of genetically diabetic mice. Clin. Sci. 125, 575–585

50. Chevalier, R. L., Forbes, M. S., and Thornhill, B. A. (2009)Ureteral obstruction as a model of renal interstitial fibrosis andobstructive nephropathy. Kidney Int. 75, 1145–1152

51. Puri, T. S., Shakaib, M. I., Chang, A., Mathew, L., Olayinka, O.,Minto, A. W. M., Sarav, M., Hack, B. K., and Quigg, R. J. (2010)Chronic kidney disease induced in mice by reversible unilateralureteral obstruction is dependent on genetic background. Am. J.Physiol. Renal Physiol. 298, F1024–F1032

52. Samuel, C. S., and Hewitson, T. D. (2009) Relaxin and theprogression of kidney disease. Curr. Opin. Nephrol. Hypertens. 18,9–14

53. van Timmeren, M. M., Bakker, S. J. L., Vaidya, V. S., Bailly, V.,Schuurs, T. A., Damman, J., Stegeman, C. A., Bonventre, J. V.,and van Goor, H. (2006) Tubular kidney injury molecule-1 inprotein-overload nephropathy. Am. J. Physiol. Renal Physiol. 291,F456–F464

54. Lekawanvijit, S., Kompa, A. R., Zhang, Y., Wang, B. H., Kelly,D. J., and Krum, H. (2012) Myocardial infarction impairs renalfunction, induces renal interstitial fibrosis, and increases renalKIM-1 expression: implications for cardiorenal syndrome. Am. J.Physiol. Heart Circ. Physiol. 302, H1884–H1893

55. Zhang, P. L., Rothblum, L. I., Han, W. K., Blasick, T. M., Potdar,S., and Bonventre, J. V. (2008) Kidney injury molecule-1 ex-pression in transplant biopsies is a sensitive measure of cell in-jury. Kidney Int. 73, 608–614

56. Humphreys, B. D., Xu, F., Sabbisetti, V., Grgic, I., Naini, S. M.,Wang, N., Chen, G., Xiao, S., Patel, D., Henderson, J. M.,Ichimura, T., Mou, S., Soeung, S., McMahon, A. P., Kuchroo,V. K., and Bonventre, J. V. (2013) Chronic epithelial kidney in-jury molecule-1 expression causes murine kidney fibrosis. J. Clin.Invest. 123, 4023–4035

57. Sung, S. A., Jo, S. K., Cho, W. Y., Won, N. H., and Kim, H. K.(2007) Reduction of renal fibrosis as a result of liposome en-capsulated clodronate induced macrophage depletion afterunilateral ureteral obstruction in rats. Nephron, Exp. Nephrol. 105,e1–e9

58. Hewitson, T. D., Mookerjee, I., Masterson, R., Zhao, C., Tregear,G. W., Becker, G. J., and Samuel, C. S. (2007) Endogenous re-laxin is a naturally occurring modulator of experimental renaltubulointerstitial fibrosis. Endocrinology 148, 660–669

59. Lee, S., Huen, S., Nishio, H., Nishio, S., Lee, H. K., Choi, B.-S.,Ruhrberg, C., and Cantley, L. G. (2011) Distinct macrophagephenotypes contribute to kidney injury and repair. J. Am. Soc.Nephrol. 22, 317–326

60. Braga, T. T., Correa-Costa, M., Guise, Y. F. S., Castoldi, A., deOliveira, C. D., Hyane, M. I., Cenedeze, M. A., Teixeira, S. A.,Muscara, M. N., Perez, K. R., Cuccovia, I. M., Pacheco-Silva, A.,Gonçalves, G. M., and Camara, N. O. (2012) MyD88 signalingpathway is involved in renal fibrosis by favoring a TH2 immuneresponse and activating alternative M2 macrophages. Mol. Med.18, 1231–1239

61. Nemeth, K., Leelahavanichkul, A., Yuen, P. S. T., Mayer, B.,Parmelee, A., Doi, K., Robey, P. G., Leelahavanichkul, K.,Koller, B. H., Brown, J. M., Hu, X., Jelinek, I., Star, R. A., andMezey, E. (2009) Bone marrow stromal cells attenuate sepsisvia prostaglandin E(2)-dependent reprogramming of hostmacrophages to increase their interleukin-10 production. Nat.Med. 15, 42–49

62. Hossain, M. A., Man, B. C. S., Zhao, C., Xu, Q., Du, X.-J., Wade,J. D., and Samuel, C. S. (2011) H3 relaxin demonstrates anti-fibrotic properties via the RXFP1 receptor. Biochemistry 50,1368–1375

63. Bennett, R. G., Heimann, D. G., Singh, S., Simpson, R. L., andTuma, D. J. (2014) Relaxin decreases the severity of establishedhepatic fibrosis in mice. Liver Int. 34, 416–426

64. LeBleu, V. S., Taduri, G., O’Connell, J., Teng, Y., Cooke, V. G.,Woda, C., Sugimoto, H., and Kalluri, R. (2013) Origin andfunction of myofibroblasts in kidney fibrosis. Nat. Med. 19,1047–1053

65. Ronco, P., Lelongt, B., Piedagnel, R., and Chatziantoniou, C.(2007) Matrix metalloproteinases in kidney disease pro-gression and repair: a case of flipping the coin. Semin. Nephrol.27, 352–362

66. Takamiya, Y., Fukami, K., Yamagishi, S., Kaida, Y., Nakayama, Y.,Obara, N., Iwatani, R., Ando, R., Koike, K., Matsui, T., Nishino,Y., Ueda, S., Cooper, M. E., and Okuda, S. (2013) Experimentaldiabetic nephropathy is accelerated in matrix metalloproteinase-2 knockout mice. Nephrol. Dial. Transplant. 28, 55–62

67. Mias, C., Lairez, O., Trouche, E., Roncalli, J., Calise, D.,Seguelas, M.-H., Ordener, C., Piercecchi-Marti, M.-D., Auge,N., Salvayre, A. N., Bourin, P., Parini, A., and Cussac, D. (2009)Mesenchymal stem cells promote matrix metalloproteinasesecretion by cardiac fibroblasts and reduce cardiac ventricularfibrosis after myocardial infarction. Stem Cells 27, 2734–2743


68. Moodley, Y., Atienza, D., Manuelpillai, U., Samuel, C. S.,Tchongue, J., Ilancheran, S., Boyd, R., and Trounson, A.(2009) Human umbilical cord mesenchymal stem cells reducefibrosis of bleomycin-induced lung injury. Am. J. Pathol. 175,303–313

69. Ries, C., Egea, V., Karow, M., Kolb, H., Jochum, M., and Neth, P.(2007) MMP-2, MT1-MMP, and TIMP-2 are essential for the in-vasive capacity of human mesenchymal stem cells: differentialregulation by inflammatory cytokines. Blood 109, 4055–4063

70. Masterson, R., Hewitson, T. D., Kelynack, K., Martic, M., Parry,L., Bathgate, R., Darby, I., and Becker, G. (2004) Relaxin down-regulates renal fibroblast function and promotes matrix remod-elling in vitro. Nephrol. Dial. Transplant. 19, 544–552

71. Ali, G., Mohsin, S., Khan, M., Nasir, G. A., Shams, S., Khan,S. N., and Riazuddin, S. (2012) Nitric oxide augments mesen-chymal stem cell ability to repair liver fibrosis. J. Transl. Med.10, 75–83

72. Nankivell, B. J., Borrows, R. J., Fung, C. L.-S., O’Connell, P. J.,Allen, R. D. M., and Chapman, J. R. (2003) The natural historyof chronic allograft nephropathy. N. Engl. J. Med. 349,2326–2333

73. Reinders, M. E. J., de Fijter, J. W., and Rabelink, T. J. (2014)Mesenchymal stromal cells to prevent fibrosis in kidney trans-plantation. Curr. Opin. Organ Transplant. 19, 54–59

74. Tan, J., Wu, W., Xu, X., Liao, L., Zheng, F., Messinger, S., Sun, X.,Chen, J., Yang, S., Cai, J., Gao, X., Pileggi, A., and Ricordi, C.

(2012) Induction therapy with autologous mesenchymal stemcells in living-related kidney transplants: a randomized con-trolled trial. JAMA 307, 1169–1177

75. Reinders, M. E. J., de Fijter, J. W., Roelofs, H., Bajema, I. M., deVries, D. K., Schaapherder, A. F., Claas, F. H., van Miert, P. P.,Roelen, D. L., van Kooten, C., Fibbe, W. E., and Rabelink, T. J.(2013) Autologous bone marrow-derived mesenchymal stromalcells for the treatment of allograft rejection after renal trans-plantation: results of a phase I study. Stem Cells Transl. Med. 2,107–111

76. Perico, N., Casiraghi, F., Introna, M., Gotti, E., Todeschini, M.,Cavinato, R. A., Capelli, C., Rambaldi, A., Cassis, P., Rizzo, P.,Cortinovis, M., Marasa, M., Golay, J., Noris, M., and Remuzzi, G.(2011) Autologous mesenchymal stromal cells and kidneytransplantation: a pilot study of safety and clinical feasibility. Clin.J. Am. Soc. Nephrol. 6, 412–422

77. Teerlink, J. R., Metra, M., Felker, G. M., Ponikowski, P., Voors,A. A., Weatherley, B. D., Marmor, A., Katz, A., Grzybowski, J.,Unemori, E., Teichman, S. L., and Cotter, G. (2009) Relaxinfor the treatment of patients with acute heart failure (Pre-RELAX-AHF): a multicentre, randomised, placebo-controlled,parallel-group, dose-finding phase IIb study. Lancet 373,1429–1439

Received for publication April 23, 2014.Accepted for publication September 26, 2014.



Documents

Combination therapy of mesenchymal stem cells and serelaxin effectively attenuates renal fibrosis in obstructive nephropathy