Cortical Actin Depolymerisation in 3D Cell Culture Enhances Extracellular Vesicle Secretion and Therapeutic Effects

Abstract

Three-dimensional (3D) culture systems have been shown to enhance cellular secretion of small extracellular vesicles (sEVs) compared to two-dimensional (2D) culture. However, the molecular mechanisms driving sEV secretion and influencing their potential for disease treatment have not been elucidated. In this study, we discovered the depolymerisation of cortical actin as a new mechanism that leads to increased sEV release, and that in 3D cultured mesenchymal stem cells (MSCs), this process was modulated by the downregulation of integrin-α1 (ITGA1) and subsequent inhibition of the RhoA/cofilin signalling pathway. Interestingly, the knockdown of Rab27A and Rab27B significantly reduced sEV secretion by MSCs to 0.5- and 0.1-fold, respectively. However, there was no difference in expression levels of Rab27A/B between MSCs cultured in 2D and 3D environments. In addition, sEVs derived from 3D cultured MSCs demonstrated enhanced therapeutic function both in vitro and in rat models of osteoarthritis (OA) and wound healing. Collectively, this study illustrates a new mechanism for enhanced secretion of sEVs, involving RhoA/cofilin pathway-dependent cortical actin depolymerisation, which is independent of Rab27A/B. These findings provide novel insights for optimising the yield of stem cell-derived sEVs, as well as their therapeutic efficacy for treating chronic diseases.

Keywords: 3D culture; Rab27A/B; actin depolymerisation; cortical actin; small extracellular vesicles.

FIGURE 1

Schematic diagram of study design. MSCs, mesenchymal stem cells; EVs, extracellular vesicles; MVBs, multivesicular bodies; M1, represents type I macrophage; M2, represents type II macrophage; MSC^2D‐EVs, represent EVs derived from MSCs grown on a 2D plate; MSC^3D‐EVs, represent EVs derived from MSCs grown in a 3D culture system.

FIGURE 2

3D culture of MSCs increases sEV secretion. (a) Schematic flow chart of 3D dynamic culture of MSCs and extraction of sEVs. (b) Live/dead staining and scanning electron microscopy (SEM) images of MSCs grown in 3D on microcryogels. (c) Western blotting (WB) results for sEV‐specific markers CD9, Alix, TSG101 and CD63, along with the negative marker Calnexin. (d) Nanoparticle tracking analysis (NTA) showing the particle size distribution of sEVs derived from MSCs cultured in 2D (MSC^2D‐sEVs) and 3D (MSC^3D‐sEVs). (e) Representative transmission electron microscopy (TEM) images and semi‐quantitative analysis of MSC^2D‐sEVs and MSC^3D‐sEVs. (f) Gene Ontology (GO) analysis of MSCs cultured on 2D plates and in 3D culture system. (g) KEGG analysis of MSCs cultured on 2D plates and in 3D culture system. (h) Gene Set Enrichment Analysis (GSEA) of RHOA in GO terms of macrophage migration, muscle system process, ncRNA processing, somatic stem cell population maintenance, external encapsulating structure, vesicle lumen, exoribonuclease activity and extracellular matrix structure constituent. (i) GSEA of LIMK2 in GO terms of acute inflammatory response, ncRNA metabolic process, ncRNA processing, somatic stem cell population maintenance, collagen containing extracellular matrix, gamma tubulin complex, interleukin‐1 receptor binding and tumour necrosis factor receptor binding.

FIGURE 3

Cortical actin depolymerisation increases sEV secretion independent of Rab27A/B. (a) Representative 2D‐SIM immunofluorescence staining images of Rab27A, Rab27B and F‐actin after siRNA knockdown of Rab27A and Rab27B. (b) Semi‐quantitative analysis of Rab27A and Rab27B in immunofluorescence staining images following transfection (n = 5). (c) Representative results of western blot analysis for the expression of Rab27A and Rab27B in MSCs after siRNA knockdown. Vinculin was used as a loading control. (d) Number of secreted sEVs normalised to the cell number. (e) Representative 2D‐SIM immunofluorescence staining images of Rab27A, Rab27B and F‐actin in 3D and 2D cultures. (f) Semi‐quantitative analysis of Rab27A and Rab27B immunofluorescence staining images in 3D and 2D cultures (n = 5). (g) Representative results of western blot analysis for the expression of Rab27A and Rab27B in 3D and 2D cultures. Vinculin was used as a loading control. (h) Representative immunofluorescence staining images of G‐actin and F‐actin of MSCs cultured in different conditions. (i) Representative immunofluorescence staining images of CD9 (specific marker of sEVs) and F‐actin of MSCs cultured in different conditions. (j) Cell membrane of MSCs cultured under different conditions were pre‐labelled with a live cell dye, and time‐lapse sequence photographs of live cells were taken at 5‐s intervals to obtain time‐lapse sequence maps. (k) Representative TEM images of different sEVs. (l) Nanoparticle tracking analysis (NTA) showing the particle size distribution of different sEVs. (m) Number of secreted sEVs in different groups normalised to the cell number.

FIGURE 4

ITGA1/RhoA/cofilin pathway downregulation promotes cortical actin depolymerisation and sEV secretion. (a) Heatmap representation of Differentially Expressed Genes (DEGs) in MSC‐2D and MSC‐3D. (b) Volcano diagram highlighting DEGs. (c) Circular plot of DEGs mainly associated with integrin binding, integrin‐mediated signalling pathway, cell adhesion mediated by integrin, positive regulation of kinase activity and integrin complex. (d) Representative immunofluorescence staining images of G‐actin and F‐actin in MSCs with blocked ITGA1. (e) Representative immunofluorescence staining images of CD9 and F‐actin in MSCs with blocked ITGA1. (f) Representative immunofluorescence staining images of ITGA1 and F‐actin in MSCs cultured in different conditions. (g) Representative western blots and (h) Corresponding quantification for the expression of RhoA, RhoA‐GTP, cofilin, p‐cofilin, T‐actin, F‐actin and G‐actin in MSCs cultured in different conditions. The insoluble F‐actin pool was separated from the soluble G‐actin pool using Triton‐X100 extraction and subsequent centrifugation. GAPDH was used as a loading control. (i) Representative TEM images of sEV secretion from MSCs in different conditions. (j) Schematic illustrating the potential mechanism regulating sEV secretion from MSCs cultured in a 3D environment. Inhibition of the integrin α1 (ITGA1) receptor reduces activation of the RhoA/cofilin pathway, thereby preventing actin polymerisation from G‐actin to F‐actin, and promoting the transportation and secretion of multivesicular bodies (MVBs) from the cell into the extracellular space. CytoD, cytochalasin D; integrin α1: ITGA1; MVBs: multivesicular bodies.

FIGURE 5

MSC^3D‐sEVs enhance cell activity and macrophage polarisation compared to MSC^2D‐sEVs. (a) Schematic illustrating the experiments using different groups of sEVs to evaluate their effects on the proliferation, migration and anti‐senescence activity of chondrocytes and human dermal fibroblasts (HDFs). (b) EdU staining and (c) semi‐quantitative analysis of cells treated with different sEV groups. (d) Scratch wound healing assay, and (e) Semi‐quantitative analysis of scratch gaps in different groups (n = 5). (f) Representative images of SA‐β‐Gal staining in chondrocytes and HDFs, and (g) Percentage of SA‐β‐Gal⁺ cells (n = 5). (h) Relative gene expression of M1‐related genes (CD86, TNF‐α, iNOS and IL‐1β) and M2‐related genes (CD206, IL‐4, IL‐10 and Arg‐1). (i) Immunofluorescence staining and (j) Semi‐quantitative analysis of CD86 and mannose expression.

FIGURE 6

In vivo effects of MSC^3D‐sEVs treatment in OA rats. (a) Schematic illustration of establishing the rat OA model, and experimental design for evaluating the protective effects of different sEVs. (b) Representative footprint images and 3D footprint intensity (RH/LH) (n = 5). (c) Evaluation of swing speed, mean intensity, print area, and maximum contact area during walking. (d) X‐ray and 3D reconstructed μ‐CT images. (e) Quantitative analysis of articular space width, trabecular separation (Tb.Sp), bone volume fraction (BV/TV), and trabecular number (Tb.N) from μ‐CT (n = 5). (f) H&E and safranin‐O staining of joint sections at 6 weeks post‐surgery. (g) Heatmap representation of Mankin scoring for joint histopathology. (h) Quantitative Mankin score analysis of safranin‐O staining. (i) Immunohistochemical staining of aggrecan, collagen II (COL2), ADAMTS5 and MMP13. (j) Quantitative Mankin score analysis of aggrecan staining. (k) Quantitative Mankin score analysis of COL2 staining. (l) Quantitative Mankin score analysis of ADAMTS5 staining. (m) Quantitative Mankin score analysis of MMP13 staining (n = 6).

FIGURE 7

In vivo effects of MSC^3D‐sEVs treatment on wound healing in rats. (a) Schematic illustration of the rat wound healing model. (b) Representative images depicting the dynamic progress of wound healing in different groups on days 0, 1, 3, 5, 7, 9, 11 and 14. (c) Wound healing ratio on days 0, 1, 3, 5, 7, 9, 11 and 14 (n = 5). (d) H&E and Masson staining with semi‐quantitative analysis of the wound tissue on day 14.