Mesenchymal Stromal Cells Epithelial Transition Induced by Renal Tubular Cells-Derived Extracellular Vesicles

Mesenchymal-epithelial interactions play an important role in renal tubular morphogenesis and in maintaining the structure of the kidney. The aim of this study was to investigate whether extracellular vesicles (EVs) produced by human renal proximal tubular epithelial cells (RPTECs) may induce mesenchymal-epithelial transition of bone marrow-derived mesenchymal stromal cells (MSCs). To test this hypothesis, we characterized the phenotype and the RNA content of EVs and we evaluated the in vitro uptake and activity of EVs on MSCs. MicroRNA (miRNA) analysis suggested the possible implication of the miR-200 family carried by EVs in the epithelial commitment of MSCs. Bone marrow-derived MSCs were incubated with EVs, or RPTEC-derived total conditioned medium, or conditioned medium depleted of EVs. As a positive control, MSCs were co-cultured in a transwell system with RPTECs. Epithelial commitment of MSCs was assessed by real time PCR and by immunofluorescence analysis of cellular expression of specific mesenchymal and epithelial markers. After one week of incubation with EVs and total conditioned medium, we observed mesenchymal-epithelial transition in MSCs. Stimulation with conditioned medium depleted of EVs did not induce any change in mesenchymal and epithelial gene expression. Since EVs were found to contain the miR-200 family, we transfected MSCs using synthetic miR-200 mimics. After one week of transfection, mesenchymal-epithelial transition was induced in MSCs. In conclusion, miR-200 carrying EVs released from RPTECs induce the epithelial commitment of MSCs that may contribute to their regenerative potential. Based on experiments of MSC transfection with miR-200 mimics, we suggested that the miR-200 family may be involved in mesenchymal-epithelial transition of MSCs.


Introduction
Mutual interactions between epithelial cells and mesenchymal cells coordinate kidney development, play a pivotal role in maintaining organ integrity in the adult, and contribute to renal regeneration after injury. Bone marrow-derived mesenchymal stromal cells (MSCs) have

Immunofluorescence analyses
Indirect immunofluorescence was performed on cells cultured on chamber slides (Nalgen Nunc International, Rochester, NY, USA). The cells were fixed in 2,5% paraformaldehyde and permeabilized with Hepes-Triton X100 buffer (Sigma-Aldrich, St. Louis, MO). The following antibodies were used: mouse monoclonal anti-vimentin (Sigma-Aldrich), mouse monoclonal anti-pan-cytokeratin (Bio-Rad, Hercules, CA), rabbit polyclonal anti-cytokeratin 18 and goat polyclonal antiaminopeptidase A (Santa Cruz Biotechnology, Dallas, TX). As a control, primary antibodies were omitted and substituted with nonimmune mouse IgG. Alexa Fluor 488 anti-mouse (Invitrogen, Carlsbad, CA) was used as secondary antibody. For nuclear staining, Hoechst 33258 dye (Sigma-Aldrich) was added to fixed cells. A Zeiss LSM 5 Pascal Model Confocal Microscope (Carl Zeiss International, Germany) was used to perform confocal microscopy analysis.
MSCs expressed high levels of vimentin (S1A Fig

Extracellular vesicles isolation and conditioned medium preparation
EVs were obtained from supernatant of 80% confluent RPTECs cultured overnight in RPMI (without phenol red, Invitrogen) supplemented with 2% FBS deprived of EVs after 15 hours of centrifugation at 100,000 g (Optima L-90K ultracentrifuge, Beckman Coulter, Indianapolis, IN). Cell supernatant was collected, centrifuged at 2,000 g for 20 minutes to remove debris, and processed to obtain: (1) extracellular vesicles (EVs), (2) total conditioned medium with EVs (TOT-CM) and (3) conditioned medium after EVs deprivation (CM) after ultracentrifugation at 100,000 g for 15 hours.
EVs were isolated from cell-free supernatant by differential ultracentrifugation as previously described by Théry and colleagues [28]. The pellet collected after ultracentrifugation at 100,000 g was resuspended in RPMI containing 10% dimethyl sulfoxide (DMSO, Sigma-Aldrich) and stored at -80°C.
TOT-CM and CM were concentrated~20-fold by centrifugation at 2,700 g for 40 minutes at 4°C, using ultrafiltration units with a 3 kDa molecular weight cut-off (Amicon Ultra -15, centrifugal filter devices, Millipore, Billerica, MA). To preserve EVs integrity, TOT-CM was supplemented with 10% DMSO and stored at -80°C, while CM was stored directly at -20°C.

Incorporation of EVs in MSCs
To study EVs incorporation by MSCs, we incubated these cells with 30,000 labelled-EVs/cell for 6-15-24 hours at 37°C. The EVs were previously stained with a red-fluorescent dye which binds to lipid membranes (Vybrant™ DiI cell-labeling solution, Invitrogen), in accordance with the manufacturer's instructions. Hoechst 33258 (Sigma-Aldrich) staining was used to visualize cellular nuclei. EVs uptake was evaluated by confocal microscopy.

Stimulation experiments using RPTEC-derived EVs, or TOT-CM or CM and co-culture experiments
To test the role of RPTEC-derived paracrine factors in MSC differentiation, MSCs were seeded in MSCBM, at a density of 50,000 cells/well in 6-well plates (Becton Dickinson) and stimulated with 150,000 EVs/cell, 30% CM, or 30% TOT-CM for 7-14 days before RNA extraction. When using TOT-CM to stimulate MSCs, we maintained the same ratio EV:cells that we used to stimulate MSCs with EVs. For co-culture experiments, a transwell system with a 1 μm pore size permeable membrane (Becton Dickinson) was used to separate RPTECs physically from bone marrow MSCs in 6-well plates (Becton Dickinson). RPTECs were seeded into the upper insert of the transwell system in DMEM with 5% FBS, at a density of 100,000 cells/well. MSCs were seeded into the lower chamber of this co-culture system, at a density of 50,000 cells/well. MSCs cultured alone in MSCBM were used as control. Cells were maintained in co-culture conditions for 7 days before RNA extraction. Six different experiments with similar results were done.

Trans-epithelial electric resistance
Trans-epithelial electrical resistance (TEER) was used as an indicator of epithelial differentiation and integrity (30). Control MSCs, or MSCs cultured in the presence of EVs or TOT-CM, or MSCs co-cultured with RPTECs were plated on polycarbonate membrane transwell (Falcon Corning Corp., Cambridge, MA) and allowed to reach confluency. An epithelial volt-ohm meter (EVOM, World Precision Instruments, Inc., Sarasota, FL) was used to determine the TEER. Measurement of cell-free membrane inserts were performed and subtracted from all subsequent measurement and expressed as ohm/cm 2 . All values were normalized for the area of the membrane. All experiments were done in triplicate.
was refreshed and MSCs were incubated with 50 μg/ml of FITC-labeled BSA (Sigma-Aldrich) for two hours at 37°C. Fluorescence emission was measured by cytofluorimetric analysis.

Western Blot Analysis
For protein analysis, EVs and cells were lysed at 4°C for 30 minutes in RIPA buffer (20 nM Tris-HCl, 150 nM NaCl, 1% deoxycholate, 0.1% SDS, 1% Triton X-100, pH 7.8) supplemented with PMSF, protease and phosphatase inhibitors cocktail (Sigma-Aldrich). EV samples were quantified using a proteic Bradford assay. Since the EVs protein concentration was lower than 2 mg/ml, a mixture of methanol-chloroform-water was used for the quantitative precipitation of proteins, as previously described [33,34]. Aliquots of the cell lysates containing 30 μg proteins were run on 10% acrylamide gel SDS-PAGE under reducing conditions. Proteins were transferred onto a PVDF membrane using the 7 minutes transfer program of the iBlot™ Dry Blotting System (Life Technology, Carlsbad, CA). The blots were blocked with 5% non fat milk in PBS supplemented with 0.1% Tween-20 (PBS-T). The following primary antibodies were used at 1∶200 dilution: anti-human aminopeptidase A and anti-human pan-cytokeratin (Santa Cruz Biotechnology). After incubation for 15 hours at 4°C with primary antibodies, membranes were washed with PBS-T and then incubated with for one hour at room temperature with peroxydase conjugated secondary antibodies (Santa Cruz Biotechnology). Finally, membranes were washed with PBS-T, developed with ECL detection reagents (GE healthcare, Amersham, Buckinghamshire, UK) and detected by Chemidoc XRS system (Bio-Rad).

RNA extraction and quantitative Real Time Reverse Transcriptase Polymerase Chain Reaction
Total RNA was isolated from both EVs and cells (used in co-culture and stimulation experiments) using miRNeasy mini kit (Qiagen, Valencia, CA), according to manufacturer's instructions, and quantified spectrophotometrically (Nanodrop ND-1000, Wilmington DE). The quality of EVs-derived RNA was evaluated by capillary electrophoresis on an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc, Santa Clara, CA) using the eukaryotic total RNA 6000 Pico Kit. The presence of small RNAs was verified in both EV and cell samples using the Small RNA Kit (Agilent Tech).
The mRNA expression in MSCs co-cultured or stimulated with EVs was assessed by quantitative real-time PCR. High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) and the Power SYBR1 Green PCR Master Mix (Applied Biosystems) were used as previously described [26]. Negative cDNA controls (no cDNA) were cycled in parallel with each run. Quantitative real-time polymerase chain reaction (qRT-PCR) was performed using a 96-well StepOne Real-Time System (Applied Biosystems). Sequence-specific oligonucleotide primers were purchased from MWG-Biotech AG (Ebersberg, Germany, www.mwg-biotech.com) and are shown in Table 1. Fold change in mRNA expression compared to control was determined as 2 -ΔΔCt for all samples, using TBP as normalizer. The endogenous control in EVs was significantly different from that of their cells of origin. For this reason, miRNA comparisons between cells and EVs was not performed.

MicroRNA profiling by quantitative real-time PCR
Mature miRNA expression levels in EVs were analyzed as previously described [29,30]. The Applied Biosystems TaqMan1 MicroRNA Assay Human Panel Early Access Kit (Life Technology) was used to profile 754 mature miRNAs by sequential steps of reverse transcription (Megaplex RT Pools; Life Technology) using an Applied Biosystems 7900H qRT-PCR instrument. The SDS software (version 2.3) was used to calculate raw C T values, automatic baseline, and threshold. Since a C T value of 35 represents single molecule template detection, C T values higher than 35 were considered to be below the detection level of the assay and excluded from the analyses.
To confirm the expression of MET-related miRNAs, screened by microarray analysis, miScript Reverse Transcription Kit and miScript SYBR Green PCR Kit (both from Qiagen) were used as previously described [22]. All samples were run in triplicate and 3 ng of cDNA were used for each reaction, as described by the manufacturer's protocol (Qiagen). MiRNAs specific primers to hsa-miR-200a, 200b, 200c, 141 and 429 are listed in Table 2 and were all used in the same reaction. The small nucleolar RNA RNU-6B was used as reference control to normalize data. A 96-well StepOne TM Real Time System (Applied Biosystems) was used to perform qRT-PCR.

MicroRNA target prediction and pathway analysis
The web-based application DIANA-mirPath (version 2.0) was used to perform the enrichment analysis of predicted target genes by miRNAs in biological pathways. The algorithm microT-CDS was chosen to predict EVs-derived miRNA targets, using the default microT threshold of 0.8.
DIANA-mirPath performed an enrichment analysis of multiple miRNA target genes to all known KEGG pathways, as previously described [36,37]. The statistical significance value Table 1. Primers used in qRT-PCR to evaluate mRNAs expression.

Gene
Symbol/gene ID: Primer sequence doi:10.1371/journal.pone.0159163.t002 associated with the identified biological functions was calculated by the mirPath software (http://microrna.gr/mirpath). Biological pathways showing p-value less than 0.001 were considered as significantly enriched.

MSC transfection with miRNA mimics
MSCs were transiently transfected with miR-200a, miR-200b, miR-200c, miR-141 and miR-429 mimics (20 μM, MiScript miRNA mimics, Qiagen) using the HiPerFect Transfection Reagent (Qiagen), according to the manufacturer's protocol. Briefly, MSCs were seeded into 6-well plates at a density of 10,000 cells/cm 2 . MiRNA mimics at a concentration of 20 nM were mixed with an appropriated volume of HiPerFect Transfection Reagent (Qiagen) and incubated at room temperature for 10 minutes. Allstars negative control siRNA (20 μM, Qiagen) was used as scrambled control. The transfection mix was added dropwise to the cells in FBSdepleted culture medium, in agreement with the manufacturer's instructions. After 24-hours of incubation, the transfection medium was removed, and cells were cultured in MSCBM for 7 days.

Statistical analysis
Statistical analysis was performed by using the t test. Statistical significance was set at P < 0.05.

RPTECs induce epithelial commitment of MSCs
To investigate the effects of RPTECs on MSCs, human bone marrow-derived MSCs and RPTECs were co-cultured in not-contact conditions. After 7 days, mRNA levels of specific mesenchymal markers Twist (TWIST1), vimentin (VIM) and FOXC2 were significant reduced in MSCs (Fig 1A and 1E); moreover, MSCs expressed higher levels of the epithelial-specific marker cytokeratin 18 (KRT18), suggesting that MET occurred in MSCs (Fig 1B and 1H). Same results were observed when MSCs were incubated with TOT-CM (Fig 1A, 1B, 1D and 1G). To investigate the contribution of paracrine factors in epithelial differentiation of MSCs induced by RPTECs, MSCs were also incubated with CM or with purified EVs. Interestingly, after one week of incubation with CM, we did not observe any change in the expression of mesenchymal and epithelial markers in MSCs, suggesting that the contribution of soluble factors is not sufficient to induce the epithelial commitment of MSCs (Fig 1A and 1B). In MSCs stimulated with EVs, we observed a significant reduction in the expression of mesenchymal markers ( Fig 1A and 1C) and an increase in the expression of the epithelial-specific marker KRT18, in a manner comparable with co-culture system (Fig 1B and 1F). These results indicate that EVs play a relevant role in inducing MET in MSCs. Notably, the down-regulation of the expression of mesenchymal markers and the increase in cytokeratin 18 expression were also maintained after two weeks of stimulation with EVs (Fig  2A, 2B and 2C), TOT-CM (Fig 2A, 2B and 2D) and of co-culture with RPTECs (Fig 2A, 2B and 2E). A slight but significant increase in expression of the renal tubular-specific marker aminopeptidase A (ENPEP) was also observed in MSCs stimulated with EVs (Fig 2B and 2F), TOT-CM (Fig 2B and 2G). The early expression of ENPEP in MSCs co-cultured with RPTECs ( Fig 1B) was also maintained after 14 days (Fig 2B and 2H), suggesting that the renal tubular commitment of MSCs was facilitated by a bidirectional exchange of paracrine factors between the two cell populations.
Human proximal tubular epithelial cells can effectively reabsorb albumin by receptor-mediated endocytosis [39]. To verify whether this tubular epithelial-specific function was induced by EVs in MSCs, we quantified fluorescently-labeled albumin uptake using cytofluorimetric analysis. MSCs stimulated with EVs, TOT-CM and co-cultured with RPTECs successfully incorporated FITC-labeled albumin (Fig 2I). Moreover, evaluation of TEER as an indicator of epithelial differentiation and integrity showed that EVs and TOT-CM induced a significant increase of TEER in MSCs (Fig 2J). Taken together, these results confirm that EVs can induce the epithelial commitment of MSCs.

Characterization of EVs and incorporation into MSCs
To better clarify the mechanisms underlying the epithelial commitment induced by RPTECs on MSCs, we evaluated the capacity of EVs to be incorporated into MSCs and characterized EVs for the presence of specific proteins and mRNAs. Starting from 6 hours of incubation, labeled EVs were internalized by MSCs, with a progressive compartmentalization of EVs inside the cell at 15 and 24 hours of incubation (Fig 3).
Nanosight analysis revealed that the average size of EVs was 185 nm with a mode of 163 nm and a standard deviation of 84 nm (Fig 5A). EVs presence was also confirmed in TOT-CM, with an average size of 162 nm, a mode of 141 nm and a standard deviation of 84 nm (Fig 5B). After RPTEC supernatant ultracentrifugation, EVs concentration was highly reduced in CM ( Fig 5B).

MicroRNA profiling in EVs and comparative pathway analyses
We verified the quality of EVs-derived RNA and the presence of small RNAs, with a bioanalyzer profile on total RNA from EVs and their cell of origin. EVs contained a broad range of RNA sizes, with a RNA Integrity Number (RIN) of 7.30 out of 10 ( Fig 5C). A more comprehensive analysis using a bioanalyzer kit specific for small RNAs showed a relevant peak characteristic of small RNA classes inside EVs, with an enrichment of miRNAs around 36% (Fig 5D). RNA extracted from EVs was profiled for 754 known human mature miRNAs. We found 237 miRNAs expressed (S1 Table). Among them, we found a subset of five miRNAs (miR-200a, miR-200b, miR-200c, miR-141 and miR-429) that belong to miR-200 family, known to be involved in EMT [40,41].
In order to examine which biological pathways may be involved in the regulation of the epithelial commitment of target cells, we applied the software DIANA MirPath on the 65 most expressed miRNAs derived from EVs. We performed an enrichment analysis of predictive miRNA target genes included in KEGG database and we found that 63 KEGG biological processes were significantly enriched (p<0.001, FDR corrected) (S2 Table). Twenty miRNAs found in EVs revealed a strong association to several biological pathways that are known to modulate EMT (Table 4).
Our analysis showed that the predicted target genes of these miRNAs are principally involved into the following EMT and MET-associated pathways: ECM-receptor interaction, Wnt pathway, gap junction, transcriptional misregulation in cancer, TGF-beta pathway, pathways in cancer, MAPK signaling pathway, p53 signaling pathway, focal adhesion, PI3K-Akt signaling pathway, regulation of actin cytoskeleton (Table 5) [11,[42][43][44][45]. Fig 6 provides a graphical representation of the most significantly enriched pathways modulated by the 20 miR-NAs that we selected from our analysis.
We selected two miRNA targets from our predictive analysis, CCND1 and IGF-1R, and examined their expression in MSCs after one week of incubation with EVs or TOT-CM. In both conditions, we observed a reduction in the expression of CCND1 and IGF-1R in MSCs, that could be attributed to miRNA-mediated inhibitory effects (Fig 7). These data suggest that EVs may prompt an epithelial commitment in target cells, by the delivery of specific miRNAs involved in MET-related pathways.

Role of the miR-200 family in the epithelial commitment of MSCs
We employed qRT-PCR to validate the expression levels of the five miRNAs belonging to miR-200 family, selected from the miRNA array. We showed that all these five miRNAs were strongly expressed in EVs and we selected them to study their possible role in MET induction in MSCs. A mixture of mimics was used to increase the expression of miR-200a, miR-200b, miR-200c, miR-141 and miR-429 in MSCs. After 1 week of transfection, MSCs showed lower levels of MET markers VIM and fibronectin 1 (FN1) (Fig 8A), and higher levels of epithelial marker KRT18 and tubular marker ENPEP (Fig 8B). Moreover, MSCs transfection with  Extracellular Vesicle-Induced MSC Differentiation miR-200 mimics reduced the expression of two miRNA targets from our predictive analysis, CCND1 and IGF-1R (Fig 8C).
To assess whether miR-200 family can effectively repress translation through binding to CCND1 and IGF1R 3'UTR, we performed luciferase reporter assay. MiR-200 family mimics   were co-transfected in HEK 293T cells with the luciferase reporter construct pEZX-MT05 containing the human CCND1 or IGF1R 3'UTR immediately following the luciferase coding sequence. As a result, miR-200 family induced a 90% reduction of luciferase activity in transfected cells. Taken together, these results suggest that the miR-200 family carried by EVs can induce MET in MSCs by binding to the 3'UTR sequences of predicted target genes, like CCND1 and IGF1R (Fig 8D).

Discussion
The mechanism of cell communication mediated by EVs allows exchange of proteins and genetic information between cells, as EVs may transfer mRNAs, long non coding RNAs (lncRNAs) and miRNAs to target cells [18,20,[46][47][48]. The epithelial commitment of bone marrow-derived cells induced by EVs derived from differentiated tissues has been demonstrated by Quesenberry's group [49]. They showed that EV-mediated transfer of proteins and genetic information from lung injured cells to bone marrow-derived cells may reprogram marrow cell phenotype to acquire lung epithelial-specific markers [19,20]. Moreover, EVs derived from prostate or lung cancer cells and EVs derived from lung injured cells were also shown to trigger stable epigenetic changes in marrow cells, activating the expression of prostate-or lung-specific genes [50,51], following internalization of EVs into target cells [52]. Taken together, these results indicate that EVs collected from differentiated injured tissues can effectively induce epigenetic modifications of marrow cells, that acquire phenotypic characteristics of the tissue of origin.
Previous studies have shown that tubular epithelial cells and their conditioned medium may induce an epithelial transition of MSCs [1][2][3][4]. In the present study, we have demonstrated that EVs released by RPTECs are the main mediators of the epithelial differentiation of bone marrow-derived MSCs. In fact, we noticed that RPTEC-derived conditioned medium depleted of EVs was no longer able to induce MET in MSCs. In particular, after one week of incubation of MSCs with EVs, we observed the decreased expression of TWIST1, FOXC2, known to repress epithelial genes expression [53,54], and of vimentin, a protein involved in cellular adhesion and migration [55,56]. Moreover, after two weeks of MSC stimulation with EVs, we observed the increased expression of epithelial marker cytokeratin 18, which contributes to maintenance of epithelial cell polarity and plays an important role in cellular adhesion and cell-to-cell interactions [57][58][59], and a slight increase in the expression of aminopeptidase A. This zinc-dependent metalloproteinase is typically expressed in the brush border membrane of renal proximal related pathways (highlighted by the circle line) are targeted by the selected miRNAs with a very low p-value (p<10 −7 , FDR corrected). The attached dendrograms illustrate hierarchical clustering results for miRNAs and pathways. MiRNAs that show similar pathway targeting patterns were clustered together. An analogous clustering was also performed for the biological pathways.
doi:10.1371/journal.pone.0159163.g006 tubule cells and it is known to be involved into the removal of N-terminal acid aminoacidic remainings from different peptides, for example angiotensin II [60,61]. Taken together, these results indicate an epithelial differentiation of MSCs induced by EVs.
To understand the mechanism of epithelial commitment of MSCs induced by EVs, we characterized EV protein and RNA content. We found that EVs contained epithelial-specific proteins and transcripts that after delivering to target cells may be responsible for the early epithelial commitment of MSCs. The persistence of KRT18 expression in MSCs after 14 days of stimulation with EVs suggests that epigenetic changes contributed to maintain MSCs epithelial commitment. The long-term modification in MSC transcriptome was probably due to epigenetic changes that follow the transfer of non-coding RNAs, such as miRNAs [62]. Recently, the EVs-mediated transfer of miRNAs has been observed between differentiated normal cells [21,63], between MSCs and differentiated cells [64][65][66], between MSCs and cancer cells [67][68][69] and inside the tumor microenvironment [70][71][72].
To investigate the miRNA content of EVs, we profiled their RNA for 754 known human mature miRNAs and we identified 20 miRNAs with a strong association to EMT pathways. In particular, we detected the expression of some miRNAs belonging to miR-200 family. Recent  [73][74][75][76] have demonstrated that miR-200 family miRNAs (miR-200a, miR-200b, miR-200c, miR-141, miR-429) are involved in EMT. In fact, mature miR-200 family miRNAs promote E-cadherin expression with the acquisition of an epithelial cell phenotype via post-transcriptional repression of zinc finger E-box binding transcription factor 1 and 2 (ZEB1 and ZEB2) [77]. During EMT, ZEB1 and ZEB2 over-expression leads to transcriptional repression of E-cadherin, promoting a mesenchymal cell phenotype. This process may be reversible, since an increase in miR-200 family miRNAs expression initiates MET, restoring Ecadherin expression and promoting an epithelial phenotype [77].
Since miR-200 family miRNAs are shuttled by EVs, we investigated the effects of these miR-NAs in MSCs and their possible contribution to MET. After one week of transfection of bone marrow derived-MSCs with 5 synthetic miR-200 mimics, we observed a reduction in the expression of VIM and FN1, which is another mesenchymal marker involved in cell adhesion, cell migration, cell growth and cell differentiation [78]. Moreover, we observed an increased expression of ENPEP, indicating that the miR-200 miRNAs can induce the epithelial commitment of MSCs. These data suggest that EVs may contribute to the epithelial commitment of MSCs by transferring a small subset of miRNAs that belong to the miR-200 family.
Furthermore, in order to validate the results of our predictive analysis on EV-derived miR-NAs, we randomly selected cyclin D1 (CCND1) and insulin growth factor-1 receptor (IGF1R), from all predictive miRNA target genes.
CCND1 encodes a regulatory subunit of the cyclin-dependent kinase 4 (CDK4) and of the cyclin-dependent kinase 6 (CDK6), whose activity is required for cell cycle G1/S transition [79]. CCND1 is involved in cell proliferation, survival, invasion, metastasis formation and its over-expression is linked to tumorigenesis [80,81]. Recent studies have demonstrated that CCND1 activation can induce EMT in tumor cells, such as ovarian cancer [82], breast cancer [83], esophageal cancer [84] and epidermoid carcinoma cells [85]. In particular, Su et al. [84] have shown that the inhibition of CCND1 down-regulates the expression of other mesenchymal markers and reverses EMT in esophageal cancer cells.
IGF1R is a trans-membrane receptor tyrosine kinase which binds to its ligand IGF1, leading to activation of the PI3K-Akt pathway and the Erk-MAPK pathway. Previous studies have shown that IGF1R activation can induce EMT in prostate cancer [86], breast cancer [87,88], mammary epithelial cells [89] and lung cancer cells [34,90]. Nurwidya et al. [35] found that inhibition of IGF1R reversed hypoxia-induced EMT. Zhou et al. [90] reported that IGF1R activation induces EMT in lung cancer cells by up-regulating the expression of Snail and promoting beta-catenin translocation from the cell membrane into the nucleus, which, in turn, downregulates E-cadherin expression.
We evaluated the expression of CCND1 and IGF1R in MSCs after incubation with EVs or TOT-CM and after cell transfection with miR-200 mimics. We showed that both EVs and miR-200 family miRNAs can effectively reduce CCND1 and IGF1R expression in MSCs. These results confirmed that CCND1 and IGF1R down-regulation is indicative of MET. Luciferase reporter assay has proved that both CCND1 and IGF1R are target genes of EV-delivered miR-NAs, thus validating the results of our miRNA predictive analysis.
In conclusion, EVs released from tubular epithelial cells may modify the phenotype of MSCs by inducing an epithelial commitment that may contribute to the regenerative potential of MSCs. Whether this phenomenon occurs in vivo remains to be defined.