Human mesenchymal stromal cells release functional mitochondria in extracellular vesicles

Abstract

Cartilage and other skeletal soft tissues heal poorly after injury, in part due to their lack of vascularity and low metabolic rate. No pharmacologic approaches have proven effective in preventing chronic degenerative disease after joint injury. Mesenchymal stromal cells (MSCs) have been investigated for their ability to treat pain associated with osteoarthritis (OA) and preserve articular cartilage. Limitations of MSCs include variability in cell phenotype, low engraftment and retention rates, and inconsistent clinical outcomes. Therefore, acellular biologic therapies such as extracellular vesicles (EVs) are currently being investigated. MSC-derived EVs have been found to replicate many of the therapeutic effects of their cells of origin, but the mechanisms driving this remain unclear. Recent evidence in non-orthopedic tissues suggests MSCs can rescue injured cells by donating mitochondria, restoring mitochondrial function in recipient cells, preserving cell viability, and promoting tissue repair. Our group hypothesized that MSCs package mitochondria for export into EVs, and that these so-called "mitoEVs" could provide a delivery strategy for cell-free mitochondria-targeted therapy. Therefore, the goals of this study were to: 1) characterize the vesicle fractions of the MSCs secretome with respect to mitochondrial cargoes, 2) determine if MSC-EVs contain functional mitochondria, and 3) determine if chondrocytes can take up MSC-derived mitoEVs. We isolated exosome, microvesicle, and vesicle-free fractions from MSC-conditioned media. Using a combination of dynamic light scattering and nanoparticle tracking, we determined that MSC-EV populations fall within the three size categories typically used to classify EVs (exosomes, microvesicles, apoptotic bodies). Fluorescent nanoparticle tracking, immunoblotting, and flow cytometry revealed that mitochondrial cargoes are abundant across all EV size populations, and mitoEVs are nearly ubiquitous among the largest EVs. Polarization staining indicated a subset of mitoEVs contain functional mitochondria. Finally, flow cytometry and fluorescent imaging confirmed uptake of mitoEVs by chondrocytes undergoing rotenone/antimycin-induced mitochondrial dysfunction. These data indicate that MSCs package intact, functional mitochondria into EVs, which can be transferred to chondrocytes in the absence of direct cell-cell interactions. This work suggests intercellular transfer of healthy MT to chondrocytes could represent a new, acellular approach to augment mitochondrial content and function in poorly-healing avascular skeletal soft tissues.

Keywords: MSCs; mitoEVs; mitochondria; mitochondrial transfer; regenerative orthobiologic; secretome.

FIGURE 1

Basic characterization of human MSC-derived extracellular vesicles. (A) Dynamic Light Scattering (DLS) of cell-conditioned media (CCM) from murine MSCs revealed three sub-populations of EVs based on size: small, (∼5–10 nm) medium, (∼100–1000 nm) and large (5,000–10,000 nm), which likely represent exosomes, microvesicles, and apoptotic bodies, respectively. Thin lines represent individual trials (n = 3), thick lines represent averaged curves. (B) Analysis by one way ANOVA of average percent area under the curve on DLS data revealed a significantly smaller 1^st (exosome) peak and a significantly larger 2^nd (microvesicle) peak for both stimulated groups (* = p≤0.05). (C) Immunoblotting of whole cell lysate (WCL) and microvesicle (MV), exosome (EX), and vesicle free media (VFM) fractions of CCM. Ubiquitous marker HSP90 is present in WCL, EX, and VFM fractions. The EV marker flotillin is present in all fractions except VFM, and the cellular marker IκBα is absent from exosome (EX) and microvesicle (MV) fractions, as expected. MT protein ATP5A1 was found only in WCL, but COXIV (MT) was found in WCL and MVs. (D) Nanoparticle tracking analysis (NTA) of PBS (negative control), EX, and MV fractions indicates that EXs outnumber MVs by several orders of magnitude amongst particles less than 1μm, the maximum detection size of NTA.

FIGURE 2

MSC-derived EVs contain mitochondria (MT). (A) Events fluorescing above control thresholds for Calcein AM (Green) and Mitotracker (Red) on flow cytometry were classified as EVs containing intact MT (i.e., mitoEVs; dark red). (B) Quantification of mitoEVs (red-green double positive events) and three control groups; Mitotracker only (▴), calcein only (■), and unstained control (▾), compared to EVs stained with calcein + Mitotracker (●). (C) Representative flow plot depicting backgating strategy for assessing relative sizes of mitoEVs (red) and MT-negative EVs (gray). (D) Backgating revealed that EVs of all sizes contain MT, but EVs lacking MT are generally small, and the majority of the largest EVs contain MT. (E) Fluorescent NTA revealed mitoEVs (solid red line) trend larger than the general EV population (dotted lines) and cluster into distinct sub-groups. Groups that do not share letters (Panel B) are significantly different (p ≤0.05) by one way ANOVA.

FIGURE 3

A subset of MitoEVs contain functional mitochondria. (A) Representative flow plot identifying polarized mitoEVs; events fluorescing above control thresholds for calcein AM (green) and TMRM (orange) were classified as mitoEVs containing functional MT (red). (B). Quantification of functional mitoEVs (red-green double positive events) and three control groups; TMRM only (▴), calcein only (■), and unstained control (▾), compared to EVs stained with calcein + TMRM (●). Data is expressed as the fraction of double-positive events in each group. of percent red/green double positive showed statistically significant differences between the double stained experimental group and single color/unstained controls Groups that do not share a letter are significantly different (n = 3, p ≤0.05) by one-way ANOVA.

FIGURE 4

Murine Chondrocytes take up MT from MSC-derived Microvesicles (A) Schematic depicting experimental method for imaging non-contact intercellular MT transfer (B) Confocal images of MSC-MitoEVs and mitoEV mediated MT transfer. MT Transfer demonstrated with mitoEVs (green) incorporated into chondrocyte MT networks (red). Chondrocytes were co-cultured with MSC-microvesicles for 12 h, then fixed and imaged.

FIGURE 5

Chondrocytes take up MSC-derived MT following incubation with MitoEVs (A) Representative flow cytometry plots of chondrocytes incubated with (+MV) and without (-MV) microvesicles isolated from MSCs expressing endogenous, MT-specific dendra2 fluorescence. Dendra2 (green) threshold was set using control chondrocytes. Events fluorescing above threshold were considered chondrocytes having taken up mitoEVs. (B) Quantification of mitoEV uptake by chondrocytes in culture (n = 2) reveals the +MV group had a higher rate of transfer events than -MV. Data are expressed as the fraction of chondrocytes above threshold for green fluorescence. * = p≤0.05, by one-way ANOVA.