The American Journal of Pathology, Vol. 183, No. 4, October 2013

ajp.amjpathol.org

CARDIOVASCULAR, PULMONARY, AND RENAL PATHOLOGY

miR-21eContaining Microvesicles from Injured TubularEpithelial Cells Promote Tubular Phenotype Transition byTargeting PTEN ProteinYang Zhou,* Mingxia Xiong,* Li Fang,* Lei Jiang,* Ping Wen,* Chunsun Dai,* Chen-yu Zhang,y and Junwei Yang*

From the Center of Kidney Disease,* 2nd Affiliated Hospital, Nanjing Medical University, Nanjing; and the Jiangsu Diabetes Center,y State Key Laboratory ofPharmaceutical Biotechnology, School of Life Sciences, Nanjing University, Nanjing, China

Accepted for publication

C

P

h

June 24, 2013.

Address correspondence toJunwei Yang, M.D., Ph.D., 2ndAffiliated Hospital, NanjingMedical University, 262 NZhongshan Rd, Nanjing,Jiangsu 210003, China. E-mail:jwyang@njmu.edu.cn.

opyright ª 2013 American Society for Inve

ublished by Elsevier Inc. All rights reserved

ttp://dx.doi.org/10.1016/j.ajpath.2013.06.032

Renal fibrosis is inevitably progressive no matter what the initial insult is or whether the insult persists.In an experimental fibrosis model induced by unilateral ureteral obstruction, the accelerated patho-logical changes could hardly be explained by aggravated pressure caused by hydronephrosis afterligation. Moreover, at the initial stage, tubular phenotype transition and matrix deposition inobstructive kidneys are always local and scattered; however, these renal lesions expand and progresswith time. In this study, cultured recipient tubular cells underwent phenotype transition after incu-bation with conditioned media derived from transforming growth factor-b1etreated donor tubularcells. Thus, it is reasonable to speculate that some secretable molecules from injured tubules contributeto the progression of renal fibrosis. Herein, we report that secreted miRNA-21 (miR-21) can serve as themolecule mediating intercellular communication. miR-21 was packaged into microvesicles, which enterand deliver miR-21 into recipient tubular cells, and exogenous miR-21 enhances Akt signaling by targetdepression of phosphatase and tensin homolog (PTEN) protein, and promotes tubular phenotypetransition. These results demonstrate that tubular cells can secrete miR-21 and deliver it into recipienttubules by microvesicles, where the exogenous miR-21 can target PTEN protein and enhance Aktsignaling in recipient cells. Microvesicle-mediated delivery of miR-21 among tubular epithelialcells might shed new light on the mechanism of progressive renal fibrosis. (Am J Pathol 2013, 183:1183e1196; http://dx.doi.org/10.1016/j.ajpath.2013.06.032)

Supported by the 973 Science Program of the Ministry of Science andTechnology of China grant 2012CB517603 and National Science Foun-dation of China grant 31200870.

Y.Z. and M.X. contributed equally to this work.

Renal fibrosis is an inevitable outcome of nearly all kindsof chronic kidney disease.1,2 Despite the removal of initialinsult, renal lesions are always hardly relieved, but progresstoward end-stage renal disease. The mechanism thatpromotes the progression of renal injury remains unclear.Tubular phenotype transition3e11 is one of the contributors torenal fibrogenesis; however, similarly as observed in clinicalsettings, injured tubules are always local and scattered atthe initial stages and then expand and progress in murineobstructive kidneys. It indicates that injured tubules mightinfluence the normal ones, which results in the progression ofrenal lesions.

microRNA (miRNA), discovered as a family of non-coding RNAs, is responsible for regulating the expressionsof up to 30% of mammalian protein-encoding genes. It hasbeen proved in recent years that most proteins involved in

stigative Pathology.

.

renal fibrosis were regulated by miRNA,12e14 whichprovides a new mechanism on miRNA-regulated renal fi-brogenesis. In our previous investigations, examination ofthe expression profile of miRNAs identified miR-21, whichwas temporally up-regulated in obstructive kidneys andtransforming growth factor (TGF)-b1estimulated tubularcells. However, whether miR-21 also mediates cell-to-cell communication between injured and normal tubulesremains unclear.

Microvesicles (MVs) are circular fragments of membranethat are released from the endosomal compartment as

Zhou et al

exosome or shed from the surface membranes of almost allkinds of cell types under both normal and pathological con-ditions.15e17 Increased evidence indicates that MV-mediateddelivery of molecules plays pivotal roles in cell-to-cellcommunication. It has been reported that MVs are idealmiRNA carriers for their delivery in cells.18e24 We determinewhether MV-mediated delivery of miR-21 regulates progres-sion of renal fibrosis. In this study, we report that secretedmiR-21 can serve as the molecule mediating intercellularcommunication. miR-21 was packaged into MVs, which enterand deliver miR-21 into recipient tubular cells, and exogenousmiR-21 enhances Akt signaling by target depression of PTENprotein, and promotes tubular phenotype transition. Theseresults demonstrate that tubular cells can secrete miR-21 anddeliver it into recipient tubules by MVs, where the exogenousmiR-21 can target PTEN protein and enhance Akt signaling inrecipient cells. MV-mediated delivery of miR-21 amongtubular epithelial cells might shed a new light on the mecha-nism of progressive renal fibrosis.

Materials and Methods

Cell Culture and Treatment

A rat renal proximal tubular epithelial cell line (NRK-52E)was purchased from the Cell Resource Center of the ShanghaiInstitutes for Biological Sciences Chinese Academy ofSciences (Shanghai, China), which was originally obtainedfrom ATCC (Manassas, VA; CRL-1571TM). Cells werecultured in Dulbecco’s modified Eagle’s mediumeF12medium supplemented with 10% fetal bovine serum (Invi-trogen, Carlsbad, CA). For TGF-b1 treatment, NRK-52E cellswere seeded at 80% confluence in complete medium con-taining 10% fetal bovine serum. Twenty-four hours later, thecells were changed to serum-freemediumand incubated for 16hours. Then, cells were treated with recombined human TGF-b1 (rhTGF-b1; R&D Systems, Minneapolis, MN) at a dose of5 ng/mL. Conditioned medium was generated as follows:donor NRK-52E cells were incubated with rhTGF-b1 for 48hours. The culture media were then changed into normalDulbecco’s modified Eagle’s mediumeF12medium (withoutserum) to exclude the influence of rhTGF-b1 and incubated foranother 48 hours. The later rhTGF-b1efree media were thencollected and centrifuged at 300� g, 1200� g, and10,000� gfor 5, 20, and 30 minutes separately, respectively, to harvestconditioned media. Meanwhile, the control medium wasgenerated in the same procedure from donor cells that were notincubated with rhTGF-b1 in the first 48 hours. The recipientnormal tubular cells were then incubated with control orconditioned media for various time periods or doses, as indi-cated, and then collected for further characterization. TGF-b neutralization antibody was obtained from R&D Systems. Amatrix metalloprotease (MMP) inhibitor, C20H28N4O4, wasobtained from Sigma-Aldrich (St. Louis, MO; M5939), whichinhibits a variety of MMPs. Wortmannin and LY294002 wereobtained from Sigma-Aldrich.

1184

Western Immunoblot Analysis

NRK-52E cells were lysed with SDS sample buffer. Thesupernatants were collected after centrifugation at 13,000� g at4�C for 20minutes. Protein concentrationwas determined usinga bicinchoninic acid protein assay kit (Sigma-Aldrich), andwhole cell lysates weremixedwith an equal amount of 2� SDSloadingbuffer. Sampleswereheated at 100�Cfor approximately5 to 10 minutes before loading and were separated on precasted10%or 5%SDS-polyacrylamide gels (Bio-Rad, Hercules, CA).Detection of protein expression by using Western blot analysiswas performed according to the established protocols describedpreviously.6 The primary antibodies were as follows: antieE-cadherin (BD Transduction Laboratories, Franklin Lakes,NJ); antiea-smooth muscle actin (a-SMA), anti-actin, andanti-fibronectin (Sigma-Aldrich); antiephospho-Smad3, anti-Smad3, anti-PTEN, anti-phospho Akt (Ser473), and anti-Akt(Cell Signaling Technology, Danvers, MA). Quantificationwas performed by measurement of the intensity of the signalswith the use of ImageJ (NIH, Bethesda, MD).

Immunofluorescent Staining

Indirect immunofluorescent staining was performed aspreviously described.6 Briefly, cells cultured on coverslipswere washed twice with cold PBS and fixed with coldmethanol/acetone (1:1) for 10 minutes at �20�C. After threeextensive washings with PBS, the cells were blocked with0.1% Triton X-100 and 2% normal donkey serum in PBSbuffer for 40 minutes at room temperature and then incubatedwith the specific primary antibodies previously described,followed by staining with fluorescein isothiocyanateeconjugated secondary antibody. Cells were double stainedwith DAPI to visualize the nuclei. Slides were viewed witha Nikon Eclipse 80i Epi-fluorescence microscope equippedwith a digital camera (DS-Ri1; Nikon, Tokyo, Japan). In eachexperimental setting, immunofluorescent images werecaptured with identical exposure settings.

MV Isolation

MVs were isolated from conditioned media by differentialcentrifugation, according to previous publications.18e20

Briefly, after removing cells and other debris by centrifu-gation at 300 � g, 1200 � g, and 10,000 � g for 5, 20, and30 minutes, respectively, the supernatant was centrifuged at110,000 � g for 1 hour (all steps were performed at 4�C).MVs were collected from the pellet and resuspended in fetalbovine serumefree media. Total RNA of MVs derived fromcellswas then extracted using TRIzol LS reagent (Invitrogen).

Cryoelectron Microscopy and Transmission ElectronMicroscopy

Cryoelectron microscopy and transmission electron micros-copy were performed as previously described.20 Briefly, for

ajp.amjpathol.org - The American Journal of Pathology

miRNA-Containing MV in Tubular Transition

cryoelectron microscopy, a droplet of resuspended MV at theappropriate concentration was applied to carbon-coatedholey film supported by a copper grid. After removingexcess fluid by blotting, the grid was frozen by plunging itinto liquid ethane. The vitrified specimens were stored inliquid nitrogen and transferred into a transmission electronmicroscope using a Gatan 626 cryoholder (Thomson Scien-tific Instruments Pty Ltd, Carlton, Australia). The specimenswere imaged on an FEI (Burlington, VT) Tecnai 20 electronmicroscope with source LaB6, operated at 120 kV. Theimages were recorded by a Gatan UltraScan 894 charge-coupled device (Gatan, Warrandale, PA) in low-dose mode(30e/A2). For conventional transmission electron micros-copy, the MV pellet was placed in a droplet of 2.5% glutar-aldehyde in PBS buffer and fixed. Samples were rinsed andpostfixed in 1% osmium tetroxide. The samples were thenembedded, fixed, and cut into several blocks (<1 mm3). Thesamples were dehydrated in increasing concentrations ofalcohol and infiltrated with increasing concentrations ofQuetol-812 epoxy resin (Nisshin EM, Tokyo, Japan) mixedwith propylene oxide. Samples were embedded in pure,fresh, Quetol-812 epoxy resin and polymerized. Sections(100 nm thick) were cut using a Leica (Solms, Germany)UC6 ultramicrotome and poststained with uranyl acetate for10 minutes and with lead citrate for 5 minutes at roomtemperature before observation in an FEI Tecnai T20 trans-mission electron microscope, operated at 120 kV.

Fluorescence Labeling of MV and FluorescentMicroscopy

NRK-52E cells were labeled with Dil-C18 (a gift from Prof.Chen-yu Zhang, Nanjing University, Nanjing, Jiangsu,China) for 1 hour and then washed three times with PBS. Theconditioned media, prepared as previously described, werecollected and centrifuged to harvest MVs. MVs were resus-pended in Dulbecco’s modified Eagle’s mediumeF12medium and incubated with cultured recipient NRK-52Ecells. After incubation for various time periods, as indicated,cells were washed, fixed, and observed under a Nikon Eclipse80i Epi-fluorescence microscope equipped with a digitalcamera. In each experimental setting, immunofluorescentimages were captured with identical exposure settings.

Microarray Analysis

The procedure was conducted as previously described.25 Thecomplementary probes (in triplicate) against miRNAs weredesigned based on miRBase release 12.0 (http://www.mirbase.org). RNA labeling, microarray hybridization, andarray scanning were performed as follows: briefly, 25 mg oftotal RNA was used to isolate lowemolecular-weight RNAusing polyethylene glycol solution precipitation. Subse-quently, lowemolecular-weight RNAs were labeled withCy3 and hybridized with miRNA microarray (CapitalBioCorp, Beijing, China). Finally, hybridization signals were

The American Journal of Pathology - ajp.amjpathol.org

detected and quantified. Four independent adult amphioxusRNA samples were hybridized with miRNA microarrayseparately. Hybridization intensity values from individualamphioxus samples were filtered and global mediannormalized. We considered candidate miRNAs with a signal>3000 and P < 0.001 from a Student’s test (compared withblank spotting solution) to be positive.

Quantitative PCR of miRNAs

The miRNAs were quantified as previously described, withminor modification.26 Total RNAwas prepared using a TRIzolisolation system, according to the instructions by the manu-facturer (Invitrogen). To generate anmiRNA cDNA library, thefirst strand of cDNA was synthesized using 1 mg of RNA in 20mL of reaction buffer using miScript RT II buffer (Qiagen,Düsseldorf, Germany). The mix was incubated at 37�C for60 minutes, followed by 95�C for 5 minutes. Subsequently,real-time quantification was performed using an Applied Bio-systems 7300 Sequence Detection system (Applied Bio-systems, Inc., Carlsbad, CA). The 20-mL PCR reaction systemconsisted of 1 mL RT product, 2 mL 10� miScript UniversalPrimer, 2mL10�miScript PrimerAssay, 10mL2�QuantiTectSYBRGreenPCRMasterMix, andRNase-freewater (Qiagen).The mixtures were incubated at 95�C for 10 minutes, followedby 40 cycles of 95�C for 15 seconds and 60�C for 1minute. Allreactions were run in triplicate. The CT data were determinedusing default threshold settings, and themeanCTwas calculatedfrom the triplicate PCRs. The ratio of miRNAs in kidney withunilateral ureteral obstruction (UUO)/those in sham-controlwascalculated by using the equation 2-DCT, in which

DCT Z CT Treatment � CT Sham-Control:

All primers were purchased from Qiagen. U6 was usedfor normalization in miRNA quantitative PCR (qPCR) whentotal RNA was extracted from cell or tissue samples.

miRNA ISH

The miRNA in situ hybridization (ISH) was performed usinga mercury LNATM microRNA ISH optimization kit (Exi-qon, Vedbaek, Denmark) for formalin-fixed, paraffin-embedded kidney samples, according to the protocol by themanufacturer. Briefly, sections (10 mm thick) were prepared,followed by deparaffinization in xylene and ethanol. Theslides were incubated with 15 mg/mL of proteinase-K (Exi-qon) for 20 minutes at 37�C. After washing and dehydrating,the slides were hybridized with double digoxigenin-labeled,LNATM miR-21 probe, LNATM-scrambled miRNA probe,LNATM U6 snRNA probe, and LNATM miR-126 probe(positive control) (Exiqon) for 1 hour at 55�C. The slideswere washed with standard saline citrate buffer, and thenincubated with blocking solution for 15 minutes, followed byincubation with anti-digoxigenin reagent for 60 minutes,alkaline phosphatase substrate for 2 hours at 30�C, andKTBT buffer twice for 5 minutes. The slides were mounted

1185

Zhou et al

with mounting medium, and the results were analyzed bylight microscopy (Nikon Eclipse 80i).

Urine Collection and RNA Isolation

Urine samples of obstructive kidney and sham-control werecollected at different time points, as indicated. An equalvolume of 200 mL of urine from each group was appliedfor total RNA isolation, and acidic phenol was used, followedby chloroform/isopropanol purification. The quantity of totalRNA was then determined and normalized. A total of 1 mg ofRNA was applied for reverse transcription to generate a urinemiRNA cDNA library, using anmiScript II RT kit (Qiagen) in20mLof reaction buffer. Themixwas incubated at 37�C for 60minutes, followed by 95�C for 5 minutes. Subsequently, real-time quantification was performed as previously described.The products were then size fractionated on agarose gel.

RNA Transfection

miRNA-21 mimic, inhibitor, and their negative control (NC)RNA (purchased from Qiagen) were transfected into NRK-52E cells using Lipofectamine 2000 reagent (Invitrogen)following the protocols provided by the manufacturer. Aftertransfection, cells were incubated at 37�C in a CO2 incu-bator for 24 hours until they were ready to assay for geneexpression or further treatment.

Animal Models

Male CD-1 mice, weighing 18 to 20 g, were purchased fromthe Shanghai Experimental Animal Center. Theywere housed

1186

in the animal facilities of the Experimental Animal Center ofNanjing Medical University with free access to food andwater. Animals were treated humanely in accordance withNational Medical Advisory Committee guidelines and by useof approved procedures of the Institutional Animal Use andCare Committee at the Nanjing Medical University. CD-1mice were randomly assigned into five groups (with five miceper group): control group and UUO for 1, 3, 7, and 14 daysgroups.UUOwas performed using an established procedure.6

Briefly, under general anesthesia, complete left ureteralobstructionwas performed by double ligation of the left ureterusing 4-0 silk after a midline abdominal incision. Animals inthe sham-control group had their ureter exposed and manip-ulated, but not ligated. Mice were euthanized at different timepoints, as indicated after surgery, and the obstructive kidneyswere removed for further investigation.

Morphological Assessment

Kidney tissues were immersed in 4% neutral-buffered form-aldehyde at 4�C for 48 hours. The tissues were paraffinembedded, processed for light microscopy, and divided intosections (3 mm thick). Sections were then stained with H&Efor general histological analysis and Masson for extracellularmatrix deposition. Pictures were taken with a Nikon Eclipse80i microscope equipped with a digital camera.

Immunohistochemical Staining for Collagen I

Kidney sections (3 mm thick) were deparaffinized andrehydrated by xylene, a graded alcohol series, and double-deionized water. Briefly, after blocking for 30 minutes at

Figure 1 Conditioned media derived from TGF-b1etreated donor tubular cells promotes recipienttubular cells to undergo phenotype transition.Western blot analysis demonstrates loss of E-cad-herin, de novo induction of a-SMA, and expressionof fibronectin (FN) in recipient NRK-52E cellsincubated with 100 mL of conditioned media forvarious time periods, as indicated. The whole celllysates were immunoblotted with specific anti-bodies. The same blot was reprobed with actin toensure equal loading of each lane (A). Westernblot shows loss of E-cadherin, de novo induction ofa-SMA, and expression of fibronectin in NRK-52Ecells incubated with conditioned media atvarious doses, as indicated for 48 hours (B). Theepithelial or myofibroblast markers (green) weredetected by indirect immunofluorescent staining.NRK-52E cells were treated with control (CeE) orconditioned (FeH) media for 48 hours. Thetransformed cells lost E-cadherin (C and F),acquired a-SMA (D and G), and formed fibronectin(E and H).

ajp.amjpathol.org - The American Journal of Pathology

Figure 2 Conditioned mediaeinduced phenotype transition is independent of TGF-b1 or MMPs. Western blot analysis demonstrates that TGF-b neutrali-zation antibody (ab) was incapable of blocking the conditioned mediaeinduced (48-hour) loss of E-cadherin, and expression of a-SMA and fibronectin (FN)(A), unlike 2 ng/mL TGF-b1, conditioned media failed to induce Smad3 phosphorylation (B), and MMP inhibitor could not retrieve conditioned mediumeinduced (48 hour) phenotype transition of NRK-52E, as evaluated by E-cadherin suppression, and a-SMA and fibronectin induction (C).

miRNA-Containing MV in Tubular Transition

room temperature with blocking buffer, sections wereincubated with anti-collagen I antibody (Santa CruzBiotechnology, Santa Cruz, CA), then with horseradishperoxidaseeconjugated secondary antibodies.

Statistical Analysis

Animals were randomly assigned to control and treatmentgroups. Statistical analysis was performed using SigmaStatsoftware version 3.5 (Jandel Scientific Software, San Rafael,CA). Comparisons between groups were made using one-way analysis of variance, followed by the t-test. P < 0.05was considered significant.

Results

Conditioned Media Derived from TGF-b1eTreatedDonor Tubular Cells Promote Recipient Tubular Cells toUndergo Phenotype Transition

In mice with obstructive kidney induced by UUO, theexpansion rate of matrix deposition area was markedlyaccelerated in the later 4-day time compared with the initial3-day time (Supplemental Figure S1), suggesting anunknown internal mechanism, rather than increased hydro-static pressure promoting renal fibrosis progression. Tubularepithelial cells and tubulointerstitial spaces are the major

The American Journal of Pathology - ajp.amjpathol.org

locations of renal lesions. We, therefore, examined whetherinjured tubular cells were capable of promoting normalcongener cells to undergo phenotype transition. Thecontrol and conditioned media were generated as describedin Materials and Methods. Despite the removal of rhTGF-b1 for 48 hours, the donor tubular cells underwentphenotype transition when the conditioned media wereharvested (data not shown). As shown in Figure 1A, afterincubation with the conditioned media for various timeperiods, as indicated, normal NRK-52E cells were inducedto undergo phenotype transition in a time-dependentmanner, as demonstrated by the loss of E-cadherin, denovo expression of a-SMA, and fibronectin. The condi-tioned media were then used to incubate normal NRK-52Ecells in different doses, as indicated, for 48 hours. Westernblot analysis also shows that the recipient tubular cellsunderwent phenotype transition in a dose-dependentmanner (Figure 1B). Immunofluorescent staining showsthe phenotype transition of tubular epithelial cells aftertreatment with the conditioned media for 48 hours, whichexhibited as total loss of the E-cadherin staining (Figure 1,C and F). Meanwhile, the a-SMAepositive microfilaments(Figure 1, D and G) appeared in the cytoplasm, and fibro-nectin was deposited in the intercellular area (Figure 1, Eand H). All together, these data suggest that some secretedmolecules from injured tubular cells probably mediatedtubular epithelial cell-to-cell communication.

Figure 3 MVs derived from conditioned mediainduce phenotype transition of recipient tubularcells. Cryoelectron microscopy micrograph of MVsisolated from conditioned media. Transmissionelectron microscopy micrograph of MVs isolatedfrom conditioned media. Scale bar Z 100 nm (Aand B). Western blot analysis shows that MVs, butnot supernatant from the conditioned media,induced tubular phenotype transition (C). Fluo-rescent staining illustrates the intercellulardelivery of MVs. Microscopy image shows theinternalization of fluorescently labeled MVs intoNRK-52E cells. Donor NRK-52E cells were labeledwith Dil-C18 (red). Control (D), NRK-52E cells withMVs for 12 hours (E), and NRK-52E cells with MVsfor 24 hours (F). FN, fibronectin.

1187

Zhou et al

Conditioned Media-Induced Phenotype Transition IsIndependent of TGF-b1 or MMPs

The conditioned media of rhTGF-b1etreated NRK-52E cellsshould be a mixture of various ingredients, among whichendogenously produced TGF-b1 and MMPs are probablytwo major profibrogenesis factors. We first investigatedwhether de novoeproduced TGF-b1 by NRK-52E cells afterrhTGF-b1 treatment accounted for the phenotype transitionof epithelial cells. The rhTGF-b1efree conditioned mediawere collected as previously described. The concentration ofTGF-b1 in the conditioned media was undetectable (data notshown). Neutralizing TGF-b1, using neutralization antibody,could hardly abolish conditioned media-induced tubularphenotype transition (Figure 2A). Because the Smad signaling

Figure 4 Tubular expression and urine excretion of miR-21 are up-regulatedUUO and sham-control kidneys. Dendrogram generated by cluster analysis showedmiRNA profiling (A). Up-regulated miRNAs in fibrotic kidneys. The average fold cIncluded miRNAs show more than two-fold up-regulation (value, >1 after log2-tra(B). qPCR analysis of relative miR-21 expression in UUO versus sham-control with tshows ISH analysis of miR-21 expression in control (D) and 3 days after UUO (E) kurine (F). *P < 0.01 versus control (n Z 5) (C and F).

1188

pathway is crucial in TGF-b1einduced tubular phenotypetransition, we examined whether the conditioned media couldactivate Smad signaling. Figure 2B shows that, after 1 hour ofincubation, unlike 2 ng/mL rhTGF-b1, conditioned mediafailed to induce phosphorylation of Smad3.It has been reported that TGF-b1 can induce the

production of MMPs that may directly promote tubularphenotype transition through disruption of E-cadherin.27 Totest if MMPs could affect tubular phenotype transition, weapplied a broad-spectrum inhibitor (inhibits MMP2 andMMP9). Figure 2C shows that MMP inhibitor was inca-pable of relieving conditioned media-induced phenotypetransition of NRK-52E cells. These data indicated thatconditioned media-induced phenotype transition wasattributed to neither TGF-b1 nor MMPs.

after obstruction. Cluster analysis of the miRNAs differentially expressed inthe separation of UUO from sham-control kidney samples on the basis of

hanges obtained were log2-transformed and normalized with sham-control.nsformed and then normalized with sham-control) at least at one time pointhe same RNA preparations used in microarray assay. Representative stainingidneys. qPCR analysis of relative miR-21 level in obstructive or sham-control

ajp.amjpathol.org - The American Journal of Pathology

Figure 5 Up-regulated miR-21 from injuredtubule is packaged into MVs. qPCR analysis ofrelative miR-21 expression in NRK-52E cells,without or with 5 ng/mL of TGF-b1 treatment forvarious periods, as indicated. *P < 0.01 versuscontrol (n Z 3) (A), without or with differentconcentrations of TGF-b1 treatment for 12 hours.*P < 0.01 versus control (n Z 3) (B), without orwith 5 ng/mL of TGF-b1 treatment. *P < 0.01versus control (nZ 3) (D), and relative preemiR-21expression in donor NRK-52E cells, without or with5 ng/mL of TGF-b1 treatment. *P < 0.01 versuscontrol (n Z 3) (C). Comparison of miR-21 level inMVs isolated from conditioned media and controlmedia (E). Results are presented as means � SEof three independent experiments. *P < 0.01,conditioned MVs versus control MVs. qPCR ampli-fication curves of miR-21 from MVs or MV-freemedia (F). Comparison of miR-21 in MVs isolatedfrom murine urine (G). MVs were obtained fromurine of sham or UUO 7 days by sequentialcentrifugation. miR-21 expression levels in MVswere determined by qPCR. Results are presented asmeans � SE of three independent experiments.*P < 0.01, UUO 7 day urine MVs versus sham urineMVs.

miRNA-Containing MV in Tubular Transition

MVs Derived From Conditioned Media InducePhenotype Transition of Recipient Tubular Cells

Recent studies identified cell-derived microparticles asideal carriers for transport among cells to induce renalfibrosis.28 To determine whether MVs are carriers of thetubular-secreted molecules, MVs were obtained from con-ditioned media by sequential centrifugation. Under electronmicroscopy, the MVs appeared as clusters of vesicles(approximately 50 to 200 nm in diameter) surrounded bya double-layer membrane (Figure 3, A and B). We nextexamined whether MVs could enter into the recipient cells.As shown in Figure 3, DeF, MVs were labeled with Dil-C18, as previously reported,20 and then applied to treatrecipient NRK-52E cells. Dil-C18elabeled MVs rapidlyentered into the recipient NRK-52E cells at 37�C in a time-dependent manner. These results also suggest that recipientNRK-52E cells labeled with Dil-C18 are not due to Dil-C18 carryover but result from Dil-C18 internalization intotargeted cells. To further exclude the possible effects offree molecules in conditioned media, the effects of condi-tioned media, supernatant, and MVs on recipient cells were

The American Journal of Pathology - ajp.amjpathol.org

compared. As previously mentioned, supernatant camefrom the same conditioned media, whose MVs were iso-lated. Figure 3C shows that MVs, but not supernatant,induced tubular phenotype transition. These results showedthat MVs derived from conditioned media contained themolecule that induced tubular phenotype transition and thatMVs were probably the major carriers between tubularcells.

Tubular Expression and Urine Excretion of miR-21 AreUp-Regulated after Obstruction

As newly discovered regulators of renal fibrosis, miRNA isa promising mediator of tubular cell-to-cell communication.A microarray assay was adopted to screen a mature miRNAexpression profile in control and obstructive (UUO) kidneysamples. The expression abundance of almost 1000 miR-NAs in kidney samples was compared. A list of 45 miRNAsdifferentially expressed in UUO compared with controlsamples was generated (Figure 4A). Because the molecule isprobably secreted by tubular cells, we focused on those up-regulated miRNAs. Among the 19 up-regulated miRNAs

1189

Figure 6 Exogenous miR-21 is delivered by MVs into recipient tubular cells and promotes tubular phenotype transition. qPCR analysis of relative preemiR-21expression in recipient NRK-52E cells incubated with MVs isolated from conditioned or control media. Results are presented as means � SE of three inde-pendent experiments. A: qPCR analysis of relative miR-21 expression in recipient NRK-52E cells. Results are presented as means � SE of three independentexperiments. B: Western blot analysis shows that inhibition of miR-21 by transient transfection of miR-21 inhibitor into recipient cells blocked conditionedMV-induced E-cadherin suppression, a-SMA induction, and fibronectin (FN) expression (C). Representative indirect immunofluorescence staining of E-cadherin(DeF), a-SMA (GeI), and FN (JeL) in NRK-52E cells. Cells were treated with MVs from control media (D, G, and J). NC-transfected cells treated with MVs fromconditioned media (E, H, and K). miR-21 inhibitor (50 nmol/L) transfected cells treated with MVs from conditioned media (F, I, and L). *P < 0.01 forconditioned versus control MVs (n Z 3) (A and B).

Zhou et al

(Figure 4B), only the expression of miR-21 was increased asearly as day 1 and maintained at high levels until day 14after UUO. qPCR analysis confirmed the result of miR-21expression after UUO (Figure 4C). ISH was applied tolocalize miR-21 in kidney. As shown in Figure 4D, miR-21was hardly detectable in normal kidney. However, theexpression of miR-21 was markedly increased in the prox-imal tubule of obstructive kidney after UUO (Figure 4E).An ISH assay using a probe with a scrambled sequencesuggested that the staining with anmiR-21 probe was specificfor miR-21 (data not shown). The up-regulated expression ofmiR-21 made its secretion by tubular cells possible.

To further verify whether the up-regulated miR-21 intubular cells could be secreted, the urine miR-21 level inobstructive kidney was examined. After obstruction for 3 or7 days, total RNA was isolated using acidic phenol froma volume of 200 mL of obstructive or sham-control urine.qPCR analysis using a specific miR-21 primer was appliedto quantify the miR-21 level in urine total RNA. As weexpected, miR-21 in sham-control urine was hardly detect-able (CT Z 36.2 � 0.2); however, in obstructive urine, themiR-21 level was markedly increased in a time-dependentmanner (Figure 4F), which is in accordance with the miR-21 expression level in obstructive kidney. Moreover, theup-regulated urine excretion of miR-21 was not caused byurine concentration because total RNA concentration wasmarkedly decreased in obstructive urine compared withsham-control urine (Supplemental Figure 2A). In addition,

1190

Supplemental Figure 2B illustrates the results of semi-quantitative PCR analysis of miR-21 directly isolated frommurine urine samples. These results indicated that increasedexpression of miR-21 in obstructive tubular cells could besecreted, which made the delivery of miR-21 betweentubular cells possible.

Up-Regulated miR-21 from Injured Tubule Is Packagedinto MVs

As shown in Figure 5A, in cultured tubular cells, miR-21was induced as early as 3 hours after incubation with 5ng/mL TGF-b1; 2 ng/mL of TGF-b1 could markedlyinduce miR-21 expression as early as 12 hours (Figure 5B),suggesting that the induction is also an early event intubular phenotype transition in vitro. We next determinedwhether the up-regulated miR-21 can be packaged intoMVs, as demonstrated in previous studies.17,20e24 In thedonor NRK-52E cells, expression levels of preemiR-21and miR-21 were both markedly up-regulated (Figure 5, Cand D). The persistent up-regulation of miR-21 contributesto the progression of tubule injury after the removal of theinitial insult (rhTGF-b1). MVs were collected from controlor conditioned media and accumulated by sequentialcentrifugation. Thereafter, total RNA was extracted fromthe isolated MVs. qPCR analysis revealed that the miR-21level was markedly increased in MVs isolated fromconditioned media compared with those from control media

ajp.amjpathol.org - The American Journal of Pathology

Figure 7 miR-21 induces tubular phenotype transition via targeting PTEN protein and enhancing Akt signaling. NRK-52E cells were transfected with NC ordifferent amounts (nmol/L) of miR-21 mimic for 72 hours. Western blot analysis shows that miR-21 mimic induces loss of E-cadherin, de novo induction ofa-SMA, and expression of fibronectin (FN) in a dose-dependent manner (A). NRK-52E cells were treated with NC (BeD) or 50 nmol/L of miR-21 mimic (EeG)for 72 hours. Indirect immunostaining demonstrates that the transformed cells lost E-cadherin (B and E), acquired a-SMA (C and F), and formed FN (D and G).Western blot analysis shows that miR-21 mimic dose dependently inhibited PTEN protein expression (H), transfection of 50 nmol/L of miR-21inducedphosphorylation of Akt signaling, however diminished by wortmannin and LY294002 (I), and miR-21einduced tubular phenotype transition was abolishedby wortmannin and LY294002 (J).

miRNA-Containing MV in Tubular Transition

(Figure 5E). In addition, after removal of MVs, miR-21became undetectable in the MV-free conditioned media(Figure 5F). Similarly, the miR-21 level was markedlyincreased in MVs isolated from urine of UUO micecompared with those of sham mice (Figure 5G). It was alsodifficult to detect miR-21 in MV-free urine of UUO mice(data not shown). These results demonstrated that miR-21can be produced, secreted, and further packaged into MVsin response to injury.

Exogenous miR-21 Is Delivered by MVs Into RecipientTubular Cells and Promotes Tubular PhenotypeTransition

qPCR analysis demonstrates that miR-21 expression wasmarkedly increased in recipient NRK-52E cells (Figure 6B)after incubationwith isolatedMVs from conditionedmedia ofdonorNRK-52E cells for 24 hours; however, the preemiR-21level remained unchanged (Figure 6A). These findings

The American Journal of Pathology - ajp.amjpathol.org

suggested that the up-regulated miR-21 in recipient cells wasexogenous. To further confirm the role of miR-21 inpromoting recipient tubular phenotype transition, miR-21inhibitor, single-strand antisense oligonucleotides, whichinhibits the binding between miR-21 and its target genes bycomplementary pairing with miR-21, was pretransfected intorecipient NRK-52E cells; then, the conditioned MVs werecollected and used to incubate the pretransfected recipient cellsfor 48 hours. As shown in Figure 6C, compared with cellstransfected with NC, tubular cells transfected with miR-21inhibitor were resistant to conditioned MV-induced pheno-type transition. Furthermore, the conditioned MV-inducedphenotype transition of recipient cells was almost completelyeliminated by miR-21 inhibitor at a 50 nmol/L concentration.Immunofluorescent staining also demonstrated that condi-tioned MV-induced tubular phenotype transition (Figure 6, E,H, and K) was blocked by miR-21 inhibitor (Figure 6, F, I,and L). These results demonstrated that MV-induced tubularphenotype transition was attributed to the included miR-21.

1191

Figure 9 MVs induce tubular phenotype transition through Akt signaling.Western blot analysis shows that tubular phenotype transition induced byMVswas abolished by Akt signaling inhibitor, wortmannin and LY294002.

Figure 8 Exogenous miR-21 delivered by MVs reduces PTEN proteinlevel and enhances Akt phosphorylation in recipient tubular cells. Westernblot analysis shows that PTEN protein expression in recipient NRK-52E cellswas markedly reduced by MVs. Depression of PTEN expression was abolishedin recipient cells transfected with miR-21 inhibitor (A). Western blotanalysis shows that MVs induced the phosphorylation of Akt signaling,which was abolished by transfection of miR-21 inhibitor into recipient cells(B) and Akt phosphorylation induced by MVs was abolished by Akt signalinginhibitors, wortmannin and LY294002 (C).

Zhou et al

miR-21 Induces Tubular Phenotype Transition viaTargeting PTEN Protein and Enhancing Akt Signaling

The effects of miR-21 on tubular phenotype transition weredetermined directly by transient transfection of miR-21mimic. As shown in Figure 7A, after transfection of miR-21mimic, NRK-52E cells displayed loss of E-cadherin,induction of a-SMA, and expression of fibronectin, whichare similar to those after TGF-b1 stimulation. A dose-dependent study revealed that miR-21 was able to inducetubular phenotype transition as low as 10 nmol/L. Immu-nofluorescent staining also demonstrated that NRK-52Ecells lost E-cadherin (Figure 7, B and E), expresseda-SMA (Figure 7, C and F), and expressed fibronectin(Figure 7, D and G) after miR-21 mimic transfection. Thesedata confirmed the effect of miR-21 on enhancing tubularphenotype transition.

PTEN was demonstrated to be the target of miR-21.29e42

The dephosphorylation of phosphatidylinositol (3,4,5)-tris-phosphate (PIP3) by PTEN protein results in inhibition ofthe Akt signaling pathway.43 The effect of miR-21 on PTENexpression was determined in cultured tubular cells. Asshown in Figure 7H, PTEN protein was expressed in normalNRK-52E cells; however, miR-21 mimic dose dependentlyinhibited PTEN protein expression. The depression ofPTEN expression by 10 nmol/L of miR-21 was similar tothat expressed by 5 ng/mL of TGF-b1. Thus, transfection of

1192

50 nmol/L of miR-21 resulted in a marked phosphorylationof Akt signaling (Figure 7I). However, the activation of Aktby miR-21 was diminished by phosphatidylinositol 3-kinase(PI3K)eAkt inhibitors (wortmannin and LY294002).Furthermore, the effects of miR-21 on tubular phenotypetransition were also abolished by PI3K-Akt inhibitors(Figure 7J). These results suggested that the effect of miR-21on tubular phenotype transition was specifically dependent onthe enhanced Akt signaling pathway, which was attributed tothe target depression of PTEN protein by miR-21.

Exogenous miR-21 Delivered by MVs Reduces PTENProtein Level and Enhances Akt Phosphorylation inRecipient Tubular Cells

The protein level of PTEN was determined to assess theeffects of exogenous miR-21 delivered by MVs in recipienttubular cells. As shown in Figure 8A, the expression level ofPTEN in NRK-52E cells was markedly reduced afterincubation with conditioned media-derived MVs. Toexclude the possibility that the decreased PTEN protein wascaused by factors other than miR-21, miR-21 inhibitor wastransfected into recipient cells. As shown in Figure 8A,miR-21 inhibitor abolished the depression of PTENexpression by MVs. Furthermore, MVs induced the phos-phorylation of Akt signaling (Figure 8B). Because theactivation of Akt was abolished by transfection of miR-21inhibitor into recipient cells, the effect of MVs on Aktsignaling was specifically caused by miR-21. Moreover, theeffect of MVs on Akt phosphorylation was also abolishedby PI3K-Akt inhibitors (Figure 8C).

MVs Induce Tubular Phenotype Transition through AktSignaling

The specificity of Akt signaling in mediating MV-inducedtubular phenotype transition was finally determined usingPI3K-Akt signaling inhibitors. Figure 9 shows that tubularphenotype transition induced by MVs was abolished bywortmannin and LY294002, which suggested that Aktsignaling was indispensable for the effects of MVs. Takentogether, these results demonstrate that secreted miR-21

ajp.amjpathol.org - The American Journal of Pathology

Figure 10 Schematic illustration depictingthat miR-21 containing MVs from injured tubularepithelial cells promote tubular phenotype transi-tion. The diagram illustrates that tubular expres-sion of miR-21 is up-regulated in response toinjury (TGF-b1). The up-regulated miR-21 from theinjured tubule is then packaged into MVs. Exoge-nous miR-21 is delivered by MVs into recipienttubular cells, where they function to promotetubular phenotype transition. ECM, extracellularmembrane.

miRNA-Containing MV in Tubular Transition

presented in MVs can be delivered into recipient cells,where they function to promote tubular phenotype transitionvia enhanced Akt signaling after target depression of PTEN.

Discussion

In clinical settings, renal fibrosis is inevitably progressivedespite what the initial insult is and whether the insultpersists. However, because of the limitation of biopsyindications for patients with chronic kidney disease, it isdifficult to get full insight into the pathological character-ization of progressive renal fibrosis in patients. In this study,by using a murine renal fibrosis model induced by UUO,up-regulated miR-21 was identified as an inner mechanism,whose secretion and delivery by MVs promote phenotypetransition of tubule and progression of renal fibrosis.

In obstructive kidney, tubular cells and tubulointerstitialspaces are inevitably the major locations of renal lesion.6

Because the increased speed of pressure gradually sloweddown because of a negative feedback caused by increasedvolume of obstructive urine, the pathological feature ofacceleration of tubular damage and matrix deposition couldhardly be explained by aggravated hydronephrosis afterligation (Supplemental Figure S1). Moreover, phenotypetransition of recipient tubular cells after incubation withconditioned media derived from TGF-b1etreated donortubular cells (Figure 1) suggests that tubular epithelial cell-to-cell communication was probably mediated by somesecreted molecule from injured tubular cells. Therefore, it isreasonable to speculate that tubular epithelial cells are theinner mechanism, which form a self-aggravation circle,contributing to renal fibrosis progression. However, ingre-dients of the conditioned media from TGF-b1etreatedNRK-52E cells were complex, among which TGF-b1 andMMPs are well-known profibrogenesis factors. AlthoughTGF-b1 is ubiquitously expressed in all cell types and all

The American Journal of Pathology - ajp.amjpathol.org

cells appear to respond to TGF-b1, it was undetectable inthe conditioned media. The conditioned media also failed toinduce the Smad signaling pathway (Figure 2B). The effectsof conditioned media on recipient cells were not blockedby TGF-b neutralization antibody or MMP inhibitor(Figure 2, A and C), suggesting that conditioned media-induced phenotype transition was not attributed to TGF-b1 or MMPs. The possible effect of free molecules wasfurther excluded because the supernatant from MV-freeconditioned media failed to induce tubular phenotype tran-sition (Figure 3C).

Proteins are the possible molecules that mediate cellcommunication. However, the quantity of protein inconditioned media is limited. Moreover, the effect of proteinafter delivery is uncertain. We hypothesized that the mole-cule capable of mediating cell-to-cell communication shouldbe stable and have the capability of gene regulation, whichcould activate a magnified gene expression cascade inrecipient cells, possibly an miRNA.23

The discovery of miRNA has greatly expanded ourknowledge of gene regulation in various kinds of diseases.miRNA, a family of endogenously produced, short, non-coding RNAs, functions as an inhibitor of target geneexpression by inducing mRNA degradation or translationalrepression, thereby playing important roles in gene regula-tion to modulate physiological and pathological proc-esses.44e46 In recent years, the imperative need to determinetheir target genes and disease relevance has sparked anunprecedented explosion of research in the miRNA field. Inthe recent perfect storm of tiny RNAs,47 studies haverevealed critical functions for specific miRNAs in cellularevents, such as proliferation, differentiation, development,and immune responses, and in the regulation of genesrelevant to human diseases. Of particular interest to renalresearchers are recent reports that key miRNAs are highlyexpressed in the kidney and act as effectors of TGF-b1actions in chronic kidney disease.48e50 Hence, studying

1193

Zhou et al

miRNA expression profiles in an obstructive nephropathymodel might be an important and promising project forrevealing the mechanism involving fibrogenesis.

Applying a microarray profiling approach, we illustratedherein the miRNA expression profiles in murine interstitialfibrosis kidney disease induced by UUO. Among a list of 45miRNAs differentially expressed in obstructive kidneyscompared with control (Figure 4A), 19 were up-regulated(Figure 4B), among which the expression pattern of miR-21 was unique as up-regulated in a time-dependentmanner after UUO. qPCR and ISH analyses confirmed therapid and marked induction of miR-21 in degenerated renaltubules (Figure 4, C and D). Herein, we present a hypothesisthat persistent signaling in injured tubular epithelium causesprogression of renal fibrosis and deterioration of renalfunction. miRNAs, as previously mentioned, are present inbodily fluids, such as serum and plasma, and resistant toRNase A digestion. On these grounds, we presumed that up-regulated miR-21 may be released from impaired tubuleepithelial cells and then exert a profibrogenesis effect onnormal tubular cells. However, the mechanisms concerningthe delivery of miRNA under fibrotic conditions are largelyunknown.

It was recently reported that MVs from cultured cellscontain miRNAs18,19,21e24 and serve as carriers for deliveryof selectively secreted miRNA.16 MVs are circular mem-brane fragments, which were considered to be inert cellulardebris that came from damaged cells or dynamic plasmamembrane. However, recent studies have assigned a newdefinition to MVs released in the microenvironment asa cell-to-cell communicator.19,28 MVs, when released fromoriginal cells, may remain in the extracellular spaces inproximity or enter into the biological fluid to distant sites.By far, mechanisms involved in MV-mediated cell-to-cellcommunication included acting as a signaling complex thatdirectly stimulates target cells, transferring receptorsbetween cells, delivering functional proteins or infectiousparticles to target cells, and transferring genetic informationvia mRNA, miRNA, or transcription factors from one cell toanother.20 Our result suggested that up-regulated miR-21 ininjured tubular cells was secreted by MVs but not freenucleic acid (Figure 5F). Moreover, in donor tubular cells,the up-regulation of preemiR-21 and miR-21 was stimu-lated by the initial insult of TGF-b1 treatment; however, theup-regulation of miR-21 in recipient cells was induced byexogenous transfection via MVs (Figure 6B). Therefore, thepreemiR-21 level in recipient cells remained the same afterincubation with conditioned media (Figure 6A). MVsmediated the delivery of miR-21 into recipient tubules,thereafter inducing tubular phenotype transition andcontributing to the progression of renal fibrosis.

PTEN is a generally accepted target of miR-21. The cor-responding PTEN protein is found in almost all tissues in thebody. PTEN protein acts as a phosphatase to dephosphorylatephosphatidylinositol PIP3. PTEN specifically catalyzes thedephosphorylation of the 30 phosphate of the inositol ring

1194

in PIP3, resulting in the biphosphate product, phosphatidyl-inositol 4,5-bisphosphate. This dephosphorylation is impor-tant because it results in inhibition of the Akt signalingpathway.43 Two inhibitors of Akt signaling, wortmannin andLY294002, which block phospho-activation of Akt by inhi-bition of the production of PIP3, were used to confirm thespecificity of the involvement of Akt signaling in miR-21einduced tubular phenotype transition. We demonstratethat the effect of miR-21 on tubular phenotype transition isspecifically dependent on the enhanced Akt signalingpathway, resulting from target depression of PTEN protein bymiR-21 (Figure 7). The effects of MVs on promoting tubularphenotype transition via PTEN-Akt signaling are mostlyattributed to delivery of functional miR-21 into recipient cells(Figures 8 and 9). Recent investigation demonstrated thatTGF-b1 mRNA-containing exosomes from injured epithelialcells activated fibroblasts.28 An exosome is a type of micro-vesicle.51 Further analysis of the ingredients in MVs isolatedfrom conditioned media might provide more valuable infor-mation and inspiration.Since the discovery of lin-4 in Caenorhabditis elegans

nearly a decade ago,52 miRNA has been implicated ina wide variety of physiological and pathological processes.To date, it is reported that nearly 30% of protein-encodinggenes are predicted to be regulated by miRNAs. Thepresent study extends our knowledge of the pathogenesisand progression of renal fibrogenesis by illustrating thatmiR-21, which can be secreted by injured tubular cells andthen packaged and delivered by MVs into recipient tubularcells (Figure 10), in which the exogenous miR-21 functionsas targeting PTEN, enhances Akt signaling and alters thephenotype of recipient cells; finally, miR-21 plays animportant role in mediating cell-to-cell communicationsbetween the primary lesion and adjacent cells. Investigatingthe miRNA-mediated intercellular communication wouldhelp further in the understanding of the pathogenesis ofprogressive renal fibrosis and designing novel therapeuticstrategies.

Supplemental Data

Supplemental material for this article can be found athttp://dx.doi.org/10.1016/j.ajpath.2013.06.032.

References

1. Liu Y: New insights into epithelial-mesenchymal transition in kidneyfibrosis. J Am Soc Nephrol 2010, 21:212e222

2. Zeisberg M, Neilson EG: Mechanisms of tubulointerstitial fibrosis.J Am Soc Nephrol 2010, 21:1819e1834

3. Docherty NG, Calvo IF, Quinlan MR, Perez-Barriocanal F,McGuire BB, Fitzpatrick JM, Watson RW: Increased E-cadherinexpression in the ligated kidney following unilateral uretericobstruction. Kidney Int 2009, 75:205e213

4. Zeisberg EM, Potenta SE, Sugimoto H, Zeisberg M, Kalluri R:Fibroblasts in kidney fibrosis emerge via endothelial-to-mesenchymaltransition. J Am Soc Nephrol 2008, 19:2282e2287

ajp.amjpathol.org - The American Journal of Pathology

miRNA-Containing MV in Tubular Transition

5. Yang J, Zhang X, Li Y, Liu Y: Downregulation of Smad transcrip-tional corepressors SnoN and Ski in the fibrotic kidney: an amplifi-cation mechanism for TGF-beta1 signaling. J Am Soc Nephrol 2003,14:3167e3177

6. Yang J, Liu Y: Dissection of key events in tubular epithelial tomyofibroblast transition and its implications in renal interstitialfibrosis. Am J Pathol 2001, 159:1465e1475

7. Iwano M, Plieth D, Danoff TM, Xue C, Okada H, Neilson EG:Evidence that fibroblasts derive from epithelium during tissuefibrosis. J Clin Invest 2002, 110:341e350

8. Humphreys BD, Lin SL, Kobayashi A, Hudson TE, Nowlin BT,Bonventre JV, Valerius MT, McMahon AP, Duffield JS: Fate tracingreveals the pericyte and not epithelial origin of myofibroblasts inkidney fibrosis. Am J Pathol 2010, 176:85e97

9. Inoue T, Takenaka T, Hayashi M, Monkawa T, Yoshino J,Shimoda K, Neilson EG, Suzuki H, Okada H: Fibroblast expressionof an IkB dominant-negative transgene attenuates renal fibrosis. J AmSoc Nephrol 2010, 21:2047e2052

10. Li L, Zepeda-Orozco D, Black R, Lin F: Autophagy is a componentof epithelial cell fate in obstructive uropathy. Am J Pathol 2010, 176:1767e1778

11. Higgins DF, Kimura K, Bernhardt WM, Shrimanker N, Akai Y,Hohenstein B, Saito Y, Johnson RS, Kretzler M, Cohen CD,Eckardt KU, Iwano M, Haase VH: Hypoxia promotes fibrogenesisin vivo via HIF-1 stimulation of epithelial-to-mesenchymal transition.J Clin Invest 2007, 117:3810e3820

12. Chandrasekaran K, Karolina DS, Sepramaniam S, Armugam A,Wintour EM, Bertram JF, Jeyaseelan K: Role of microRNAs inkidney homeostasis and disease. Kidney Int 2012, 81:617e627

13. Li JY, Yong TY, Michael MZ, Gleadle JM: Review: the role ofmicroRNAs in kidney disease. Nephrology (Carlton) 2010, 15:599e608

14. Akkina S, Becker BN: MicroRNAs in kidney function and disease.Transl Res 2011, 157:236e240

15. Cocucci E, Racchetti G, Rupnik M, Meldolesi J: The regulatedexocytosis of enlargeosomes is mediated by a SNARE machinery thatincludes VAMP4. J Cell Sci 2008, 121:2983e2991

16. Thery C, Zitvogel L, Amigorena S: Exosomes: composition,biogenesis and function. Nat Rev Immunol 2002, 2:569e579

17. Peinado H, Ale�ckovi�c M, Lavotshkin S, Matei I, Costa-Silva B,Moreno-Bueno G, Hergueta-Redondo M, Williams C, García-Santos G, Ghajar C, Nitadori-Hoshino A, Hoffman C, Badal K,Garcia BA, Callahan MK, Yuan J, Martins VR, Skog J, Kaplan RN,Brady MS, Wolchok JD, Chapman PB, Kang Y, Bromberg J,Lyden D: Melanoma exosomes educate bone marrow progenitor cellstoward a pro-metastatic phenotype through MET. Nat Med 2012, 18:883e891

18. Skog J, Wurdinger T, van Rijn S, Meijer DH, Gainche L, Sena-Esteves M, Curry WT Jr., Carter BS, Krichevsky AM,Breakefield XO: Glioblastoma microvesicles transport RNA andproteins that promote tumour growth and provide diagnosticbiomarkers. Nat Cell Biol 2008, 10:1470e1476

19. Valadi H, Ekstrom K, Bossios A, Sjostrand M, Lee JJ, Lotvall JO:Exosome-mediated transfer of mRNAs and microRNAs is a novelmechanism of genetic exchange between cells. Nat Cell Biol 2007, 9:654e659

20. Zhang Y, Liu D, Chen X, Li J, Li L, Bian Z, Sun F, Lu J, Yin Y,Cai X, Sun Q, Wang K, Ba Y, Wang Q, Wang D, Yang J, Liu P,Xu T, Yan Q, Zhang J, Zen K, Zhang CY: Secreted monocytic miR-150 enhances targeted endothelial cell migration. Mol Cell 2010, 39:133e144

21. Zhuang G, Wu X, Jiang Z, Kasman I, Yao J, Guan Y, Oeh J,Modrusan Z, Bais C, Sampath D, Ferrara N: Tumour-secreted miR-9promotes endothelial cell migration and angiogenesis by activatingthe JAK-STAT pathway. EMBO J 2012, 31:3513e3523

22. Palma J, Yaddanapudi SC, Pigati L, Havens MA, Jeong S,Weiner GA, Weimer KM, Stern B, Hastings ML, Duelli DM:

The American Journal of Pathology - ajp.amjpathol.org

MicroRNAs are exported from malignant cells in customized parti-cles. Nucleic Acids Res 2012, 40:9125e9138

23. Cantaluppi V, Gatti S, Medica D, Figliolini F, Bruno S,Deregibus MC, Sordi A, Biancone L, Tetta C, Camussi G: Micro-vesicles derived from endothelial progenitor cells protect thekidney from ischemia-reperfusion injury by microRNA-dependentreprogramming of resident renal cells. Kidney Int 2012, 82:412e427

24. Gallo A, Tandon M, Alevizos I, Illei GG: The majority of micro-RNAs detectable in serum and saliva is concentrated in exosomes.PLoS One 2012, 7:e30679

25. Chen X, Li Q, Wang J, Guo X, Jiang X, Ren Z, Weng C, Sun G,Wang X, Liu Y, Ma L, Chen JY, Wang J, Zen K, Zhang J, Zhang CY:Identification and characterization of novel amphioxus microRNAsby Solexa sequencing. Genome Biol 2009, 10:R78

26. Chen C, Ridzon DA, Broomer AJ, Zhou Z, Lee DH, Nguyen JT,Barbisin M, Xu NL, Mahuvakar VR, Andersen MR, Lao KQ,Livak KJ, Guegler KJ: Real-time quantification of microRNAs bystem-loop RT-PCR. Nucleic Acids Res 2005, 33:e179

27. Zheng G, Lyons JG, Tan TK, Wang Y, Hsu TT, Min D, Succar L,Rangan GK, Hu M, Henderson BR, Alexander SI, Harris DC:Disruption of E-cadherin by matrix metalloproteinase directly medi-ates epithelial-mesenchymal transition downstream of transforminggrowth factor-beta1 in renal tubular epithelial cells. Am J Pathol2009, 175:580e591

28. Borges FT, Melo SA, Özdemir BC, Kato N, Revuelta I, Miller CA,Gattone VH 2nd, LeBleu VS, Kalluri R: TGF-b1-containing exo-somes from injured epithelial cells activate fibroblasts to initiate tissueregenerative responses and fibrosis. J Am Soc Nephrol 2013, 24:385e392

29. Qi L, Bart J, Tan LP, Platteel I, Sluis T, Huitema S, Harms G, Fu L,Hollema H, Berg A: Expression of miR-21 and its targets (PTEN,PDCD4, TM1) in flat epithelial atypia of the breast in relation to ductalcarcinoma in situ and invasive carcinoma. BMC Cancer 2009, 9:163

30. Li LQ, Li XL, Wang L, Du WJ, Guo R, Liang HH, Liu X, Liang DS,Lu YJ, Shan HL, Jiang HC: Matrine inhibits breast cancer growth viamiR-21/PTEN/Akt pathway in MCF-7 cells. Cell Physiol Biochem2012, 30:631e641

31. Buscaglia LE, Li Y: Apoptosis and the target genes of microRNA-21.Chin J Cancer 2011, 30:371e380

32. Zhang JG, Wang JJ, Zhao F, Liu Q, Jiang K, Yang GH: MicroRNA-21 (miR-21) represses tumor suppressor PTEN and promotes growthand invasion in non-small cell lung cancer (NSCLC). Clin Chim Acta2010, 411:846e852

33. Ling HY, Hu B, Hu XB, Zhong J, Feng SD, Qin L, Liu G, Wen GB,Liao DF: MiRNA-21 reverses high glucose and high insulin inducedinsulin resistance in 3T3-L1 adipocytes through targeting phosphataseand tensin homologue. Exp Clin Endocrinol Diabetes 2012, 120:553e559

34. Bai H, Xu R, Cao Z, Wei D, Wang C: Involvement of miR-21 inresistance to daunorubicin by regulating PTEN expression in theleukaemia K562 cell line. FEBS Lett 2011, 585:402e408

35. Lou Y, Yang X, Wang F, Cui Z, Huang Y: MicroRNA-21 promotesthe cell proliferation, invasion and migration abilities in ovarianepithelial carcinomas through inhibiting the expression of PTENprotein. Int J Mol Med 2010, 26:819e827

36. Xiong B, Cheng Y, Ma L, Zhang C: MiR-21 regulates biologicalbehavior through the PTEN/PI-3K/Akt signaling pathway in humancolorectal cancer cells. Int J Oncol 2013, 42:219e228

37. Ma WJ, Lv GD, Tuersun A, Liu Q, Liu H, Zheng ST, Huang CG,Feng JG, Wang X, Lin RY, Sheyhidin I, Lu XM: Role of microRNA-21 and effect on PTEN in Kazakh’s esophageal squamous cellcarcinoma. Mol Biol Rep 2011, 38:3253e3260

38. Liu ZL, Wang H, Liu J, Wang ZX: MicroRNA-21 (miR-21)expression promotes growth, metastasis, and chemo- or radio-resistance in non-small cell lung cancer cells by targeting PTEN. MolCell Biochem 2013, 372:35e45

1195

Zhou et al

39. Qin X, Yan L, Zhao X, Li C, Fu Y: microRNA-21 overexpressioncontributes to cell proliferation by targeting PTEN in endometrioidendometrial cancer. Oncol Lett 2012, 4:1290e1296

40. Dey N, Ghosh-Choudhury N, Kasinath BS, Choudhury GG: TGFb-stimulated microRNA-21 utilizes PTEN to orchestrate AKT/mTORC1 signaling for mesangial cell hypertrophy and matrixexpansion. PLoS One 2012, 7:e42316

41. Han M, Liu M, Wang Y, Chen X, Xu J, Sun Y, Zhao L, Qu H, Fan Y,Wu C: Antagonism of miR-21 reverses epithelial-mesenchymaltransition and cancer stem cell phenotype through AKT/ERK1/2inactivation by targeting PTEN. PLoS One 2012, 7:e39520

42. Yang SM, Huang C, Li XF, Yu MZ, He Y, Li J: miR-21 conferscisplatin resistance in gastric cancer cells by regulating PTEN.Toxicology 2013, 306:162e168

43. Lee JO, Yang H, Georgescu MM, Di Cristofano A, Maehama T,Shi Y, Dixon JE, Pandolfi P, Pavletich NP: Crystal structure of thePTEN tumor suppressor: implications for its phosphoinositidephosphatase activity and membrane association. Cell 1999, 99:323e334

44. Bartel DP: MicroRNAs: target recognition and regulatory functions.Cell 2009, 136:215e233

1196

45. Filipowicz W, Bhattacharyya SN, Sonenberg N: Mechanisms of post-transcriptional regulation by microRNAs: are the answers in sight?Nat Rev Genet 2008, 9:102e114

46. Zamore PD, Haley B: Ribo-gnome: the big world of small RNAs.Science 2005, 309:1519e1524

47. Ruvkun G: The perfect storm of tiny RNAs. Nat Med 2008, 14:1041e1045

48. Tian Z, Greene AS, Pietrusz JL, Matus IR, LiangM:MicroRNA-targetpairs in the rat kidney identified by microRNA microarray, proteomic,and bioinformatic analysis. Genome Res 2008, 18:404e411

49. Choi WY, Giraldez AJ, Schier AF: Target protectors reveal damp-ening and balancing of Nodal agonist and antagonist by miR-430.Science 2007, 318:271e274

50. Davis BN, Hilyard AC, Lagna G, Hata A: SMAD proteins controlDROSHA-mediated microRNA maturation. Nature 2008, 454:56e61

51. Camussi G, Deregibus MC, Bruno S, Cantaluppi V, Biancone L:Exosomes/microvesicles as a mechanism of cell-to-cell communica-tion. Kidney Int 2010, 78:838e848

52. Lee RC, Feinbaum RL, Ambros V: The C. elegans heterochronicgene lin-4 encodes small RNAs with antisense complementarity tolin-14. Cell 1993, 75:843e854

ajp.amjpathol.org - The American Journal of Pathology