Highly efficient magnetic labelling allows MRI tracking of the homing of stem cell‐derived extracellular vesicles following systemic delivery

Abstract Human stem‐cell‐derived extracellular vesicles (EVs) are currently being investigated for cell‐free therapy in regenerative medicine applications, but the lack of noninvasive imaging methods to track EV homing and uptake in injured tissues has limited the refinement and optimization of the approach. Here, we developed a new labelling strategy to prepare magnetic EVs (magneto‐EVs) allowing sensitive yet specific MRI tracking of systemically injected therapeutic EVs. This new labelling strategy relies on the use of ‘sticky’ magnetic particles, namely superparamagnetic iron oxide (SPIO) nanoparticles coated with polyhistidine tags, to efficiently separate magneto‐EVs from unencapsulated SPIO particles. Using this method, we prepared pluripotent stem cell (iPSC)‐derived magneto‐EVs and subsequently used MRI to track their homing in different animal models of kidney injury and myocardial ischemia. Our results showed that iPSC‐derived EVs preferentially accumulated in the injury sites and conferred substantial protection. Our study paves a new pathway for preparing highly purified magnetic EVs and tracking them using MRI towards optimized, systemically administered EV‐based cell‐free therapies.

S2. Labeling liposomes with DiR dye S3. Western blot analysis of EV markers Supplementary Data Figure S1. Fourier-transform infrared spectra of SPIO-His and SPIO-COOH. Figure S2. Photos of SPIO-His (left) and magneto EVs (right) before (eluent) and after (eluate) passing through a Ni-NTA column.        Figure S11. LFQ intensity-based heatmap of the top differentially expressed proteins between iPSC-EVs and plasma derived EVs using |LFQ log2 fold-change intensity| ≥5 as a threshold. Figure S12. GO pathway enrichment analysis for top differentially expressed proteins between iPSC-EVs and plasma derived EVs.
Movie S1. Dynamic T2*w images of a normal mouse after injection of magneto-EVs.
Movie S2. Dynamic T2*w images of an LPS-AKI mouse after injection of magneto-EVs Movie S3. 3D visualization of the distribution of magneto-EVs in an LPS-AKI kidney.

S1. Preparation of liposomes
Liposomes (DPPC: cholesterol: DSPE-PEG-2000 =57:40:3) were formed by the lipid film hydration method as described previously 1 . In brief, lipid mixture was dissolved in chloroform at the concentration of 25 mg/mL, followed by being air-dried for 90 min and vacuum-dried for 30 min. The lipid film was hydrated with 2 mL PBS (pH 7.4) at 50 o C for 1.5 h with intermittent shaking. Then the solution was extruded sequentially with 400 nm and 200 nm polycarbonate membranes. The size and distribution (polydispersity index, PDI) were measured at room temperature by dynamic light scattering using a Nanosizer ZS90 (Malvern Instruments, Southborough, MA) and the concentration of liposomes (in terms of number per milliliter) was measured using nanoparticle tracking analysis (Zetaview, Particle Metrix, Germany) after dilution. The size of liposomes was 226 nm in diameter (PDI=0.152).

S2. Labeling liposomes with DiR dye
DiR-labeled liposomes were prepared using the same protocol as that for EVs. In brief, liposomes of the concentration of 1.1 × 10 11 per milliliter were added with DiR dye (1,1'-Dioctadecyl-3,3,3',3'-Tetramethylindotricarbocyanine Iodide, ThermoFisher Scientific) to a final concentration of 1 µM. After incubation for 15 min with gentle shaking, the free DiR dye was removed by eluting through a Sephadex G-50 column.
C. Survival of LPS-AKI mice co-injected with FBS-EV or PBS (vehicle control), respectively.
D. Serum creatinine (SCr) levels at 24 hours in the LPS-AKI mice receiving FBS-EV or PBS (vehicle control), and normal mice without any treatment (negative control).

Figure S11. LFQ intensity-based heatmap of the top differentially expressed proteins
between iPSC-EVs and plasma derived-EVs using |LFQ log2 fold-change intensity| ≥5 as a threshold. The first two columns of the heatmap are the normalized log2 LFQ intensity values from iPSC and plasma, respectively. The third column of the heatmap is the log2 fold-change comparison between iPSC-EVs and plasma derived-EVs, with red color representing higher expression in iPSC-EVs and blue color representing higher expression in plasma derived-EVs .
The row name of the heatmap is formatted as "protein ID/corresponding gene symbol". values, using Benjamini-Hochberg procedure) was used as the threshold to select GO terms. Ontology.