Exosomes Produced from 3D Cultures of MSCs by Tangential Flow Filtration Show Higher Yield and Improved Activity

Abstract

Exosomes can deliver therapeutic RNAs to neurons. The composition and the safety profile of exosomes depend on the type of the exosome-producing cell. Mesenchymal stem cells are considered to be an attractive cell type for therapeutic exosome production. However, scalable methods to isolate and manufacture exosomes from mesenchymal stem cells are lacking, a limitation to the clinical translation of exosome technology. We evaluate mesenchymal stem cells from different sources and find that umbilical cord-derived mesenchymal stem cells produce the highest exosome yield. To optimize exosome production, we cultivate umbilical cord-derived mesenchymal stem cells in scalable microcarrier-based three-dimensional (3D) cultures. In combination with the conventional differential ultracentrifugation, 3D culture yields 20-fold more exosomes (3D-UC-exosomes) than two-dimensional cultures (2D-UC-exosomes). Tangential flow filtration (TFF) in combination with 3D mesenchymal stem cell cultures further improves the yield of exosomes (3D-TFF-exosomes) 7-fold over 3D-UC-exosomes. 3D-TFF-exosomes are seven times more potent in small interfering RNA (siRNA) transfer to neurons compared with 2D-UC-exosomes. Microcarrier-based 3D culture and TFF allow scalable production of biologically active exosomes from mesenchymal stem cells. These findings lift a major roadblock for the clinical utility of mesenchymal stem cell exosomes.

Keywords: exosomes; mesenchymal stem cell; tangential flow filtration.

Copyright © 2018. Published by Elsevier Inc.

Figures

Figure 1

Umbilical Cord Mesenchymal Stem Cells Yield the Most Exosomes (A) Yield of exosomes isolated by differential ultracentrifugation from mesenchymal stem cells derived from umbilical cord (U-MSC), bone marrow (BM-MSC), or adipose (A-MSC). Yield calculated as the number of exosomes in the isolated sample measured by Nanoparticle Tracking Analysis divided by the number of cells in the source cultures. Results of seven experiments are shown, with mean ± SD, one-way ANOVA. (B) Average sizes of U-MSC, BM-MSC, or A-MSC exosomes purified in (A).

Figure 2

Scheme of Mesenchymal Stem Cell Culturing Methods and Exosome Isolation Methods (A) Schematic of flask-based (two-dimensional) mesenchymal stem cell culture. Cells are cultured in triple-layer flasks in mesenchymal basal medium to a density of 20,000 cells/cm2. (B) Schematic of microcarrier-based (three-dimensional) mesenchymal stem cell cultures. Cells are cultured on microcarriers in serum-free/GMP-compatible medium in 250-mL spinner flasks to ∼40,000 cells/cm2. (C) Isolation of exosomes by differential ultracentrifugation. Exosomes are enriched from culture supernatants by sequential ultracentrifugation, with filtration and wash steps, as indicated. (D) Isolation of exosomes by tangential flow filtration. Exosomes are enriched from culture supernatants by tangential flow filtration using a 500-kDa cutoff cartridge, as indicated.

Figure 3

Characterization of Exosomes (A) Yield of exosomes isolated by 2D-UC, 2D-TFF, 3D-UC, or 3D-TFF (n = 12 measurements each). Yield calculated as the number of exosomes in an isolated sample measured by Nanoparticle Tracking Analysis divided by the number of cells in the source cultures. Plots show yield for each experiment, and the mean ± SD of all measurements, one-way ANOVA with Tukey’s multiple comparison test. (B) Western blot analyses of proteins present in exosomes (CD81, CD9, and CD63). Calnexin is a negative control, present in cells, but not in exosomes. (C) Protein levels detected via LC-MS/MS proteomics in exosome variants and cells. Protein content was determined by intensity-based absolute quantification (iBAQ) analysis. (D) Size distribution of exosomes isolated from two-dimensional (2D) or three-dimensional (3D) cultures by differential ultracentrifugation (UC) or tangential flow filtration (TFF). Concentration and size of exosomes were measured by nanoparticle tracking analysis. (E) Average size of 2D-UC- (n = 23), 2D-TFF- (n = 12), 3D-UC- (n = 18), and 3D-TFF-exosomes (n = 14) plotted, showing the mean ± SD of all measurements, one-way ANOVA with Tukey’s multiple comparison test. (F) Number of particles per microgram protein in exosome preparations isolated from 2D or 3D cultures by UC or TFF. Number of particles measured by Nanoparticle Tracking Analysis, and protein concentration measured by Bradford protein assay. Plots show result for each measurement, and the mean ± SD of all measurements, one-way ANOVA with Tukey’s multiple comparison test.

Figure 4

Proteomic Content of Exosomes (A) Venn diagram of proteins detected in 2D-UC-exosomes (light gray), 2D-TFF-exosomes (dark gray), 3D-UC exosomes (light blue), and 3D-TFF-exosomes (dark blue). Numbers represent the number of proteins detected in each group. Percentages represent the fraction of unique proteins of 2D-UC-only, 2D-TFF-only, 3D-UC-only, and 3D-TFF-only in the total protein amount of 2D-UC, 2D-TFF, 3D-UC, and 3D-TFF, respectively. Protein amount was determined by intensity-based absolute quantification (iBAQ) analysis. (B) Levels of proteins specific to two-dimensional culture (2D-UC and 2D-TFF), three-dimensional culture (3D-UC and 3D-TFF), differential ultracentrifugation (2D-UC and 3D-UC), and tangential flow filtration (2D-TFF and 3D-TFF), from (A), one-way ANOVA. Protein level was determined by intensity-based absolute quantification (iBAQ) analysis. (C) Size distribution of proteins specific to two-dimensional culture (2D-UC and 2D-TFF), three-dimensional culture (3D-UC and 3D-TFF), differential ultracentrifugation (2D-UC and 3D-UC), and tangential flow filtration (2D-TFF and 3D-TFF), from (A), one-way ANOVA. (D) Gene ontology analysis of proteins shared or unique to all exosome variants (lavender), 2D-only-exosomes (black), 3D-only-exosomes (dark blue), UC-only-exosomes (light blue), and TFF-only-exosomes (gray), from (A).

Figure 5

TFF-Exosomes Are More Efficient at Delivering siRNAs to Neurons (A) Dose-response analysis showing Huntingtin (Htt) mRNA levels in mouse primary neurons treated with 2D-UC-, 2D-TFF-, 3D-UC-, and 3D-TFF-exosomes containing the indicated doses of siRNA. Each data point represents the mean ± SEM of n = 3 experiments, two-way ANOVA. (B) Time course of fluorescence in primary neurons treated with 2D-UC-, 2D-TFF-, 3D-UC-, and 3D-TFF-exosomes containing Cy3-labeled siRNA. Each data point represents the mean ± SEM of 62–143 cells per time point, two-way ANOVA. UNT, untreated.