Exosomes containing miR-148a-3p derived from mesenchymal stem cells suppress epithelial-mesenchymal transition in lens epithelial cells

Abstract

Epithelial-mesenchymal transition (EMT) of lens epithelial cells (LECs) is responsible for the development of fibrotic cataracts, which contribute to severe visual impairment. Recent evidence has shown that mesenchymal stem cell-derived exosomes (MSC-Exo) can attenuate EMT in several tissues. However, the effect of MSC-Exo on EMT in LECs (LECs-EMT) has not been determined. In this study, we isolated exosomes from human umbilical cord MSCs (hucMSC-Exo) and evaluated their effect on LECs-EMT both in vitro and in vivo. HucMSC-Exo application significantly suppressed the expression of mesenchymal cell-associated genes while increasing the expression of epithelial cell-associated genes. Cell proliferation and migration of LECs undergoing EMT were inhibited after hucMSC-Exo treatment. The volume of EMT plaques in mice with injury-induced anterior subcapsular cataract (ASC) was significantly reduced in the hucMSC-Exo-treated group. Furthermore, miR-148a-3p was abundant in hucMSC-Exo. After transfection with miR-148a-3p inhibitor, the anti-fibrotic effect of hucMSC-Exo was attenuated in LECs-EMT. A dual-luciferase reporter assay identified PRNP as a direct target gene of miR-148a-3p. Furthermore, we verified that hucMSC-Exo inhibited LECs-EMT through the miR-148a-3p/PRNP axis and the potential downstream ERK signaling pathway. Taken together, our work reveals the inhibitory effect of hucMSC-Exo on LECs-EMT and the underlying mechanism involved, which may provide potential therapeutic options for fibrotic cataracts.

Keywords: EMT; LECs; PRNP; hucMSC-derived exosomes; miR-148a-3p.

Graphical Abstract

Figure 1.

Identification and characterization of hucMSC-derived exosomes. A. Flow cytometric analysis of the surface markers of hucMSCs. The cells were positive for CD90, CD105, and CD44, and negative for HLA-DR, CD34, and CD45. B. TEM analysis of typical hucMSC-Exo. C. The relationship between the size and concentration of exosomes was measured by NTA (n = 3) D. The specific surface markers (CD63, CD9, and CD81) of the exosomes were evaluated by Western blot.

Figure 2.

HucMSC-Exo were internalized into LECs and attenuated TGFβ2-induced EMT. A. Confocal images of FHL124 cells co-cultured with Dil-labeled hucMSC-Exo for 48 hours. Scale bar = 50 μm. B. Confocal images showing the time course of Dil-labeled hucMSC-Exo uptake by the LECs. Dashed lines indicate nuclei. Asterisks and triangles indicate soma regions where the fluorescence intensity of exosomes changed significantly with time. Scale bar = 10 μm. C. Time course of the fluorescence intensity over time in the asterisked region of Figure 2B. D. Immunofluorescence analysis of EMT markers in LECs after treatment with hucMSC-Exo and TGFβ2 for 48 hours. Scale bar = 50 μm. E. Western blot analysis of EMT-associated proteins in LECs after treatment with hucMSC-Exo and TGFβ2 for 48 hours. F. Real-time PCR analysis of EMT-associated gene expression in LECs after treatment with hucMSC-Exo and TGFβ2. The ACTA-2 gene encodes the α-SMA protein. *P < .05, **P < .01, and ***P < .001.

Figure 3.

HucMSC-Exo inhibited extracellular matrix deposition, LECs proliferation, and migration. A. Immunofluorescence analysis of the extracellular matrix protein Col Ⅳ, the mesenchymal cell marker vimentin, and the epithelial cell marker ZO-1 in LECs after treatment with TGFβ2 and hucMSC-Exo. Scale bar = 50 μm. B. An EdU assay was used to analyze the proliferation of LECs after treatment with TGFβ2 and hucMSC-Exo. Scale bar = 100 μm. C. Quantification of the EdU⁺/DAPI ratio (%). *P < .05 and ***P < .001. D. Wound healing analysis of LEC migration after treatment with TGFβ2 and hucMSC-Exo for 48 hours. Scale bar = 100 μm. E. Quantification of the remaining wound area per field. n = 5; *P < 0.05; ns, not significant.

Figure 4.

HucMSC-Exo ameliorated EMT in the injury-induced mouse model of ASC. A. Schematic diagram of the injury-induced ASC mouse model. B. The morphology of LECs 2 days after injury. Polygonal LECs were observed in the blank control group, and the shape of the injured area was observed in the negative control group. Scale bar = 50 μm. Exosomes were observed in the abnormal LECs of the group that received anterior chamber injection of Dil-labeled hucMSC exosomes. The dashed lines represent the outline of the cells. Scale bar = 10 μm. C. Images of whole mounts of the lens capsule showing areas of subcapsular plaque, along with α-SMA and FN staining at 7 days post injury. Scale bar = 50 μm. D. Volume quantitation of the subcapsular plaques. n = 8, ***P < 0.001. Partial illustrations were created using templates from www.motifolio.com.

Figure 5.

The miRNA expression profiles of hucMSC-Exo and the expression of miR-148a-3p in exosomes. A. Identification of conserved miRNAs in hucMSC-Exo using 2 different GSE datasets; the overlap of the miRNAs is shown in a Venn diagram. B. Heatmap of the 30 miRNAs with the highest expression levels. C. Ten candidate miRNAs associated with lens epithelial repair and negative regulation of EMT. D. Real-time PCR analysis of the 10 candidate miRNAs in hucMSC-Exo. *P < .05, **P < .01, and ***P < .001. E. The levels of miR-148a-3p in hucMSC-Exo were assessed using real-time qPCR after treatment with RNase and Triton X-100 for 30 minutes. ***P < .001; ns, not significant.

Figure 6.

The inhibitory effect of hucMSC-Exo on EMT was attenuated after application of the miR-148a-3p inhibitor. A. The expression of miR-148a-3p was downregulated after TGFβ2 treatment. ***P < .001. B. Analysis of the transfection efficiency of the miR-148a-3p mimic by qPCR. **P < .01; ns, not significant. C. Immunofluorescence staining analysis of FN and α-SMA in LECs that were transfected with the miRNA mimic negative control (mimic-NC) or the miR-148a-3p mimic, and then treated with TGFβ2 for 48 hours. Scale bar = 50 μm. D. Immunofluorescence staining analysis of FN and α-SMA in LECs transfected with inhibitor negative control (inhibitor-NC) or miR-148a-3p inhibitor, followed by treatment with TGFβ2 for 48 hours. Scale bar = 50 μm. E. Immunofluorescence results showing FN and α-SMA expression in LECs after treatment with hucMSC-Exo and transfection with the miR-148a-3p inhibitor. Scale bar = 50 μm. F. Real-time PCR analysis of the expression levels of EMT markers in LECs after the application of hucMSC-Exo and transfection with the miR-148a-3p inhibitor. *P < .05. G. Effect of hucMSC-Exo and the miR-148a-3p inhibitor on EMT-related proteins assessed by Western blot. H. Quantification of EMT-related protein expression levels assessed by Western blot. *P < .05; **P < .01, and ***P < .001.

Figure 7.

HucMSC-derived exosomal miR-148a-3p inhibited the EMT process by suppressing the PRNP and the downstream ERK pathway. A. Potential target genes of miR-148a-3p were predicted using 5 databases (TargetScan, miRanda, miRTarBase, miRSystem, and miRDB). B. The 3ʹUTR of the PRNP gene was complementary to miR-148a-3p C. A dual-luciferase reporter assay was performed to validate PRNP as a target gene of miR-148a-3p. ***P < .001; ns, not significant. D. Real-time PCR analysis of PRNP expression in LECs after transfection with the mimic negative control or the miR-148a-3p mimic. ***P < .001. E. Western blot analysis of PRNP expression in LECs after transfection with the mimic negative control or the miR-148a-3p mimic. F. Real-time PCR analysis of PRNP expression in LECs after the application of hucMSC-Exo and transfection with the miR-148a-3p inhibitor (inh-148a-3p). ***P < .001. G. The effects of hucMSC-Exo and the miR-148a-3p inhibitor on the expression of PRNP were assessed by Western blot analysis. H. Western blot analysis was performed to examine the effect of hucMSC-Exo, the miR-148a-3p inhibitor, and the miR-148a-3p mimic on the expression of p-ERK and t-ERK.