Exosomes mediate an epithelial‐mesenchymal transition cascade in retinal pigment epithelial cells: Implications for proliferative vitreoretinopathy

Abstract Exosomes have recently emerged as a pivotal mediator of many physiological and pathological processes. However, the role of exosomes in proliferative vitreoretinopathy (PVR) has not been reported. In this study, we aimed to investigate the role of exosomes in PVR. Transforming growth factor beta 2 (TGFß‐2) was used to induce epithelial‐mesenchymal transition (EMT) of retinal pigment epithelial (RPE) cells, as an in vitro model of PVR. Exosomes from normal and EMTed RPE cells were extracted and identified. We incubated extracted exosomes with recipient RPE cells, and co‐cultured EMTed RPE cells and recipient RPE cells in the presence of the exosome inhibitor GW4869. Both experiments suggested that there are further EMT‐promoting effects of exosomes from EMTed RPE cells. MicroRNA sequencing was also performed to identify the miRNA profiles in exosomes from both groups. We identified 34 differentially expressed exosomal miRNAs (P <. 05). Importantly, miR‐543 was found in exosomes from EMTed RPE cells, and miR‐543‐enriched exosomes significantly induced the EMT of recipient RPE cells. Our study demonstrates that exosomal miRNA is differentially expressed in RPE cells during EMT and that these exosomal miRNAs may play pivotal roles in EMT induction. Our results highlight the importance of exosomes as cellular communicators within the microenvironment of PVR.

occur in PVR, the transdifferentiation of retinal pigment epithelial (RPE) cells into mesenchymal cells via epithelial-mesenchymal transition (EMT) has been regarded as the trigger in PVR pathogenesis. [3][4][5] The retinal pigment epithelium, located between the neural retina and Bruch's membrane, is a monolayer of highly polarized epithelial cells. EMT enables RPE cells to lose epithelial properties, transdifferentiate into mesenchymal cells, migrate into vitreous cavity and proliferate into membranes. Ultimately, contraction of these membranes leads to structural and functional damage to the retina. 5 Exosomes are nano-sized membrane vesicles  nm in diameter) secreted by a range of cells. 6,7 Exosomes contain microRNA (miR-NAs), mRNA, DNA, cytoplasmic proteins and lipids. These components can deliver specific molecular messages and cause various responses in recipient cells. 8 It has been shown that exosomes play a fundamental role in the regulation of physiological situations, as well as in pathological processes including central nervous system diseases, myocardial ischaemia/circulation damage, liver and kidney injuries, and the modulation of tumour angiogenesis and metastasis. 9 Recent studies have MicroRNAs are noncoding RNAs that average 22 nucleotides in length. They repress target mRNAs, regulate gene expression [13][14][15] and participate in many biological processes. 16 In addition to their endogenous actions, miRNAs can be secreted into the extracellular space within exosomes. [17][18][19] Cell-derived exosomes contain many miRNAs, and these exosomes can be taken up into neighbouring or distant cells to modulate the function of the recipient cell. [20][21][22] In recent years, exosomal miRNAs have received increased attention, particularly in tumour microenvironment research. Li et al found that CAFs contribute to cancer cell proliferation and metastasis via exosomal miR-34a-5p. 23 Another study conducted by Wang et al suggest that exosomal delivery of miR-155-5p could promote EMT and chemoresistance in gastric cancer cells. 24 Exosomal miR-32-5p has also been reported to play a role in multidrug resistance in hepatocellular carcinoma. 25 With the delivery of their contents, exosomes can exert precise effects on cellular interactions within the local microenvironment, as well as on signal spreading for distant cellular communications. In carcinogenesis, the roles of exosomes in both tumour microenvironments and metastasis have been well studied. However, the effect of exosomes on EMT of RPE cells in the vitreoretinal microenvironment and PVR has not yet been investigated. In this study, we examined the effects of transitioning from the epithelial state to the mesenchymal state on the release and contents of exosomes in RPE cells.
Furthermore, we investigated the role of exosomes from EMTed RPE cells on recipient RPE cells.

| Exosome isolation
Exosomes were collected using the ExoQuick (System Biosciences) precipitation method. In brief, cell supernatants were collected and were centrifuged at 3000 g for 15 minutes to remove cell debris.
ExoQuick Solution was then added to the supernatants (1:5), mixed well and incubated overnight at 4°C. After incubation, the mixture was centrifuged at 1500 g for 30 minutes, and the supernatant was removed. Exosome pellets were suspended in PBS and passed through a 0.22-μm filter. The characterization of exosomes was confirmed by nanoparticle tracking analysis (NTA), transmission electron microscopy (TEM) analysis and Western blotting.

| Transmission electron microscopy
Exosomes were analysed using TEM. We fixed 20 μL of exosome suspension (5 µg/µL) on a continuous grid, negatively stained with 2% uranyl acetate solution for 1 minute and air-dried. The samples were observed using a FEI Tecnai G2 spirit transmission electron microscope (FEITM, Hillsboro, OR, USA) at an acceleration voltage of 120 kV.

| Nanoparticle tracking analysis
Nanoparticle tracking analysis (NTA) measurements were per-

| miRNA library construction and sequencing
Total RNA from exosomes was used for miRNA library preparation and sequencing. Library preparation and sequencing were performed at RiboBio. Briefly, total RNA samples were fractionated on a 15% Trisborate-EDTA (TBE) polyacrylamide gel and small RNAs ranging between 18 and 30 nucleotides in size were used for library preparation.
Small RNAs were reverse transcribed and amplified by PCR. The PCR products were sequenced using an Illumina HiSeq 2500 platform.

Exosome transfection was performed using an Exo-Fect Transfection
Kit. Briefly, miR543 mimics/MOCK miR was incubated with Exo-Fect transfection reagent for 15 minutes at RT and exosomes from normal ARPE-19 cells were added to the mixture and incubated at 37°C for 1 hour. The reaction was transferred to the pre-washed spin-column and incubated with gentle rotation for 10 minutes at RT. Exosomes loaded with miR543/MOCK miR were collected by centrifuging the spin-column for 30 seconds at 1000 g.

| Real-time quantitative PCR
Total RNA was extracted at the indicated timepoints using TRIzol SnRNA U6 and Cel-miR-39 were used to normalize for technical variation between cell samples and exosome samples, respectively, as previously described. 26

China) supplemented with phenylmethylsulphonyl fluoride and
PhoSTOP EASY pack phosphatase inhibitor (Roche, Mannheim, Germany) on ice for 30 m, 48 hours after treatment. The lysates were clarified by centrifugation at 8317 g for 5 m at 4°C. Total protein concentration was quantified by a bicinchoninic acid assay kit (Thermo Scientifics), and 40 μg protein was loaded and separated on SDS-PAGE gels and transferred onto nitrocellulose membrane (Bio-rad). The membranes were blocked using 5% BSA (Sigma-Aldrich) in PBS for 45 m at room temperature to prevent non-specific binding. The membranes were then incubated with primary antibodies diluted in 2% BSA in PBS with 0.1% Tween-20 (PBS-T) at 4°C overnight. After rinsing with PBS-T for three times, the membranes were incubated with IRDye ® 680LT Goat anti-rabbit or IRDye ® 800CW Goat anti-mouse secondary antibodies (Li-Cor Biosciences) at room temperature for 1 hour. After three washes with PBS-T, the bound antibody was detected using an Odyssey infrared imaging system (Li-Cor Biosciences). The band intensities were analysed using Odyssey software and normalized to β-actin or GAPDH.

| Immunofluorescence analysis
Cells were seeded and cultured in a 24-well plate inlaid with glass coverslips. After treatment, cells were washed and fixed in cold acetone for 5 m. After three washes with PBS, cells were blocked with 2% BSA for 1 hour at room temperature and incubated with the primary antibodies overnight at 4°C. After three rinses with PBS, the coverslips were then incubated with FITC-conjugated secondary antibodies for 1 hour at room temperature. After counterstaining with 4,6-diamidino-2-phenylindole (DAPI), the stained coverslips were mounted and visualized under a confocal microscope (Carl Zeiss, LSM710, Jena, Germany).

| Statistical analysis
All experiments were performed at least three times. The mean and SEM were calculated on all parameters determined in this study. Statistical significance was analysed by one-way ANOVA or two-tailed Student's t test. A value of P < .05 was accepted as statistically significant.

| EMT induction in ARPE-19 cells
We first used TGFβ2 to induce EMT in ARPE-19 cells. ARPE-19 cells were seeded and cultivated for 24 hours. Then, cells were starved using DMEM/F12 medium supplemented with 1% penicillin-streptomycin without FBS for 24 hours before treatment with TGFβ2. Treatment with 10 ng/mL TGFβ2 for 48 hours significantly reduced the expression of E-cadherin and increased the expression of α-SMA and fibronectin at both the mRNA and protein level ( Figure 1A,B). These results were also validated by immunofluorescence analysis ( Figure 1C). Our results suggested that ARPE-19 cells went through EMT after 48 hours treatment with 10 ng/mL TGFβ2.
Morphological feature of extracted vesicles was observed by TEM, which was consistent with the characteristics of exosomes. The concentration and size distribution of extracted vesicles were analysed by NTA. Peaks vesicle size for both groups was within the expected size of exosomes. The concentration of exosomes from EMTed ARPE-19 cells was 6.5-fold higher than that of normal ARPE-19 cells

| Verification of exosome internalization
To verify that ARPE-19 cells can internalize exosomes, CM-Dillabelled exosomes were incubated with PKH67-labelled ARPE-19 cells for 24 hours, and the cellular uptake of exosomes was imaged via confocal microscopy. The internalization of exosomes was confirmed visually by the presence of intracellular punctate fluorescence in the cytoplasm of the cells. There were no differences in the uptake of normal ARPE-19-derived exosomes and EMTed ARPE-19derived exosomes (Figure 4).

cells and EMTed ARPE-cells were examined via Illumina
HiSeq 2500 high-throughput sequencing. We compared the ex-

| Exosomal miR-543 induces EMT in ARPE-19 cells
We selected miR-543 for further study, as it was significantly increased in exosomes from EMTed ARPE-19 cells, and was the most highly expressed miRNA ( Figure 5B). In addition, several studies have suggested that miR-543 plays a role in EMT in a variety of cell types. [27][28][29][30][31] To test whether exosomal miR-543 could induce

| D ISCUSS I ON
Proliferative vitreoretinopathy, which is a severe blinding complication that usually occurs after RRD pre-or post-operatively, is char- In this study, we first investigated the effect of EMT on the secretion of exosomes from ARPE-19 cells. We use TGFβ to obtain EMTed ARPE-19 cells and collect exosomes for subsequent experiments. TGFβ, a classic EMT inducer that is used in many cell types, was also found in the vitreous fluid of PVR patients. 35 Our group has utilized TGFβ to induce EMT in RPE cells since 2011. 36 Consistent with previous studies, 48 hours treatment with TGFβ resulted in obvious changes to both epithelial and mesenchymal marker proteins in ARPE-19 cells. [36][37][38][39][40] We next extracted exosomes from both normal ARPE-19 cells and EMTed ARPE-19 cells. For exosome extraction, many methods have been previously used, including differential ultracentrifugation, 41 PEG precipitation, 42 sucrose and iodixanol ultracentrifugation, 43 immunoaffinity capture 44 and size-exclusion chromatography. 45 Of these, differential centrifugation and PEG precipitation are the two methods that are most commonly used. The deposit from both methods often contains lipoprotein particles and large protein aggregates. As the classic exosome extraction protocol, differential centrifugation shows better purity, higher efficiency and lower cost compared with other methods. In addition, PEG precipitation has been previously used to separate viruses and has been shown to be a viable method for exosome extraction in recent years. 42,46,47 PEG precipitation method-based kits such as ExoQuick and Total Exosome Isolation reagent have also become increasingly popular due to their ease of use and improved protection of exosome contents. 48,49 In our experiments, PEG precipitation was adopted because of its efficiency and convenience in dealing with large volumes of supernatant.
The concentration and content of secreted exosomes vary by cell type and cell status. Previous studies have suggested that massive exosome release from cells underlies specific biologic processes.

Research conducted by Wang et al suggested that breast cancer
cells secrete much more exosomes when cultured under hypoxic conditions. 50 Another study suggested that LPS, an EMT inducer, can significantly increase the release of exosome from macrophages. 51 Wang et al also observed massive exosome release from astrocytes upon treatment with TNFα. 52 Consistent with previous reports, our results suggested that ARPE-19 cells secreted massive amounts of exosomes when EMT was induced by TGFβ-2. As for

F I G U R E 3 Co-culture of normal and EMTed ARPE-19 cells. A, Western blot analysis. Recipient ARPE-19 cells were co-cultured with
EMTed ARPE-19 cells with or without GW4869 for 48 h. The protein expression of E-cadherin, α-SMA and fibronectin was detected by Western blot. Relative protein expression (normalized to β-actin) was quantified in Western blots based on grey scale values. The data are presented as the mean ± SEM, n = 3. Statistical significance was analysed by one-way ANOVA. *P < .05; B, Real-time quantitative PCR analysis. Recipient ARPE-19 cells were co-cultured with EMTed ARPE-19 cells with or without GW4869 for 48 h. The mRNA expression levels of EMT-related proteins were detected with real-time quantitative PCR. The data are presented as the mean ± SEM, n = 3. Statistical significance was analysed by one-way ANOVA. *P < .05; (C) Immunofluorescence analysis of EMT-related proteins in recipient ARPE-19 cells. After co-culturing with EMTed ARPE-19 cells for 48 h, EMT-related proteins were detected using appropriate antibodies. Nuclei were stained with DAPI. The slides were examined by confocal microscopy. Original magnification: 630×, oil. Scale bar: 10 μm exosomal contents, previous research compared the protein differences between exosomes from normal MDCK and EMTed MDCK cells and found significantly different levels of EMT-related proteins such as E-cadherin, MMP and integrin between the two groups. 53 Another study conducted by Tang et al suggested that the function and miRNA profile of exosomes change after EMT in human lung cancer cells. 54 In this study, we focused on the miRNA profile of exosomes from TGFβ-2. 57 Jun et al also demonstrated that miR124 was down-regulated after TGFβ-1 treatment, and that overexpression of miR124 repressed the TGFβ-1-induced EMT of RPE by targeting RHOG. 58 The role of exosomal miRNAs has also been investigated. A study conducted by Li et al suggested that CAFs contribute to oral cancer cell proliferation and metastasis via exosome miR-34a-5p. 23 In addition, Ota et al found that exosomal miR-30e suppresses cell invasion and migration via inhibiting the EMT of cholangiocarcinoma cells. 59 However, the effect of exosomal miRNAs in the EMT of RPE has not yet been reported. In the study mentioned above, 53 exosomes from EMTed cells were speculated to further promote EMT; however, no functional analyses were conducted. 60 In our study, we incubated recipient ARPE-19 cells Of these, miR-543 has been reported to promote cell migration and invasion by targeting SPOP in gastric cancer. 61 Another study conducted by Zhao et al suggested that miR-543 promotes migrant, invasion and EMT in oesophageal cancer cells. 27 MiR-543 has also been shown to promote EMT in prostate cancer, via targeting RKIP. 31 In our study, miRNA-543 mimics were transfected into exosomes, and these exosomes were incubated with recipient ARPE-19 cells. Up-regulation of mesenchymal proteins and down-regulation of epithelial proteins were found in recipient ARPE-19 cells. These results suggest that exosomal miR-543 induces EMT in ARPE-19 cells, and that the EMT cascade process can be at least partially completed by exosomal miR-543.
In PVR pathology, the disease progresses from several RPE cells floating within the vitreous at disease onset to the formation of a PVR membrane containing huge amounts transdifferentiated RPE cells, suggesting that more and more RPE cells go through EMT as the disease progress. In the local vitreoretinal environment, EMTed RPE cells may secrete exosomes loaded with pro-EMT cargo. These exosomes F I G U R E 5 Exosomal microRNA sequencing and exosomal miR-543 promote EMT of ARPE-19 cells. A, Heat map of differentially expressed exosomal miRNA from normal and EMTed ARPE-19 cells. B, Real-time quantitative PCR verification of miRNA-seq. Exosomal miR-10a-5P, miR-543 and miR-323a-3p expression in normal and EMTed ARPE-19 cells was detected by RT-qPCR. n = 3, Statistical significance was analysed by two-tailed Student's t test. *P < .05; (C) Real-time quantitative PCR analysis of miR-543 in recipient ARPE-19 cells. ARPE-19 cells were incubated with 100 μg/mL exosome loaded with miR-543 mimics or miRNA mimics control for 3, 12, 24 and 48 h, and the cellular miR-543 levels were detected with RT-qPCR. Statistical significance was analysed by one-way ANOVA. D, Western blot analysis. Recipient ARPE-19 cells were incubated with 100 μg/mL exosomes loading with miR-543 mimics or miRNA mimics control for 48 h. The protein expression of E-cadherin, α-SMA and fibronectin was detected with Western Blot. Relative protein expression (normalized to β actin) was quantified in the Western blots based on their grey scale values. The data are presented as the mean ± SEM. n = 3. Statistical significance was analysed by one-way ANOVA. *P < .05; E, Real-time quantitative PCR analysis. Recipient ARPE-19 cells were incubated with 100 μg/ mL exosomes loaded with miR-543 mimics or miRNA mimics control for 48 h. The mRNA expression levels of EMT-related proteins were detected with real-time quantitative PCR. The data are presented as the mean ± SEM, n = 3. Statistical significance was analysed by oneway ANOVA. *P < .05; F, Immunofluorescence analysis of EMT-related proteins in ARPE-19 cells. After incubation with 100 μg/mL exosomes loaded with miR-543 mimics or miRNA mimics control for 48 h, EMT-related proteins were detected using appropriate antibodies. Nuclei were stained with DAPI.

| CON CLUS ION
Our preliminary study suggests that ARPE-19 cells secret massive amounts of exosomes after EMT, and that the exosomes from EMTed ARPE-19 cells induce further EMT of recipient ARPE-19 cells.
This EMT cascade mediated by exosomes could be at least partially accounted for by exosomal miR-543. Primary RPE and animal models of PVR may also be adopted to confirm the results derived from ARPE-19 in the future. Moreover, the comprehensive and precise role of exosomes in the pathogenesis of PVR and the detailed underlying mechanisms still need further investigation.

ACK N OWLED G EM ENTS
We thank Prof. Kai Wang from Shanghai First maternity and Infant hospital for providing methods and consultation for exosome collection. This work was supported by the National Natural Science

Foundation of China (No. 81770939) and Shanghai Municipal
Commission of Health and Family Planning, 20174Y0052.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the finding of this study are available from the corresponding author upon reasonable request.