Scalable Production of Human Mesenchymal Stromal Cell-Derived Extracellular Vesicles Under Serum-/Xeno-Free Conditions in a Microcarrier-Based Bioreactor Culture System

Abstract

Mesenchymal stromal cells (MSC) hold great promise for tissue engineering and cell-based therapies due to their multilineage differentiation potential and intrinsic immunomodulatory and trophic activities. Over the past years, increasing evidence has proposed extracellular vesicles (EVs) as mediators of many of the MSC-associated therapeutic features. EVs have emerged as mediators of intercellular communication, being associated with multiple physiological processes, but also in the pathogenesis of several diseases. EVs are derived from cell membranes, allowing high biocompatibility to target cells, while their small size makes them ideal candidates to cross biological barriers. Despite the promising potential of EVs for therapeutic applications, robust manufacturing processes that would increase the consistency and scalability of EV production are still lacking. In this work, EVs were produced by MSC isolated from different human tissue sources [bone marrow (BM), adipose tissue (AT), and umbilical cord matrix (UCM)]. A serum-/xeno-free microcarrier-based culture system was implemented in a Vertical-Wheel^TM bioreactor (VWBR), employing a human platelet lysate culture supplement (UltraGRO^TM-PURE), toward the scalable production of MSC-derived EVs (MSC-EVs). The morphology and structure of the manufactured EVs were assessed by atomic force microscopy, while EV protein markers were successfully identified in EVs by Western blot, and EV surface charge was maintained relatively constant (between -15.5 ± 1.6 mV and -19.4 ± 1.4 mV), as determined by zeta potential measurements. When compared to traditional culture systems under static conditions (T-flasks), the VWBR system allowed the production of EVs at higher concentration (i.e., EV concentration in the conditioned medium) (5.7-fold increase overall) and productivity (i.e., amount of EVs generated per cell) (3-fold increase overall). BM, AT and UCM MSC cultured in the VWBR system yielded an average of 2.8 ± 0.1 × 10¹¹, 3.1 ± 1.3 × 10¹¹, and 4.1 ± 1.7 × 10¹¹ EV particles (n = 3), respectively, in a 60 mL final volume. This bioreactor system also allowed to obtain a more robust MSC-EV production, regarding their purity, compared to static culture. Overall, we demonstrate that this scalable culture system can robustly manufacture EVs from MSC derived from different tissue sources, toward the development of novel therapeutic products.

Keywords: bioreactors; extracellular vesicles; mesenchymal stromal cells (MSC); scalable production; serum-/xenogeneic-free.

FIGURE 1

Workflow of the production and characterization of MSC-EVs in bioreactors and static systems. MSC were isolated from three different human tissue sources: BM, AT, and UCM. Firstly, MSC were expanded in static conditions (i.e., T-flasks) in hPL supplemented DMEM. These cells were subsequently used to inoculate a VWBR (5M cells; 100 mL final working volume), as well as to maintain a static culture in T-175 flasks. For each cell source, MSC from three independent donors (n = 3; BM1, 2, 3; AT1, 2, 3; UCM1, 2, 3) were used to inoculate either the final T-flasks for EV production or the VWBR, in passages from P4 to P5 [specifically, BM1 (P4); BM2 (P5); BM3 (P4); AT1 (P4); AT2 (P4); AT3 (P5); UCM1 (P4); UCM2 (P4); UCM3 (P5)]. Upon reaching stationary growth phase in VWBR or maximum confluency in static, the culture medium was changed for supplement-free culture medium and culture was maintained for 48 h. Over this period, culture medium was enriched in EVs secreted by cultured MSC. This conditioned culture medium was recovered and EVs were isolated by precipitation using a commercially available kit. Finally, EV production was quantified in both static and dynamic systems and samples were characterized using multiple techniques. MSC, mesenchymal stromal cells; EV, extracellular vesicles; VWBR, Vertical-Wheel^TM bioreactor; hPL, human platelet lysate; BM, bone marrow; AT, adipose tissue; UCM, umbilical cord matrix; NTA, nanoparticle tracking analysis; AFM, atomic force microscopy. The cells, T-flask and Eppendorf cartoons were obtained from Smart Servier Medical Art (https://smart.servier.com).

FIGURE 2

MSC culture in the microcarrier-based bioreactor system. (A) Evolution of cell number (upper panel) and cell viability (lower panel) over culture period time, for MSC from three different human tissue sources (bone marrow, adipose tissue, and umbilical cord matrix). MSC from three different donors (i.e., three biological replicates) were used per tissue source, which are represented in three different shades of gray. Two data points are presented for the same day when the medium conditioning stage (i.e., EV production) started. Results are presented as mean ± SD of cell count for each time point. (B) Representative images of microcarrier occupation by MSC throughout culture. Cell nuclei were stained with DAPI and images were acquired using a fluorescence microscope. In this case, EV production started on day 9 of culture and finished on day 11. Scale bar = 100 μm. (C) LDH activity profile during the medium conditioning (i.e., EV production) stage in the VWBR system. Culture medium samples were taken at 0, 24, and 48 h after medium conditioning started. Results from one experiment for each MSC source (BM, AT, and UCM). Results are presented as mean ± SD (n = 3). LDH, lactate dehydrogenase; VWBR, Vertical-Wheel^TM bioreactor; BM, bone marrow; AT, adipose tissue; UCM, umbilical cord matrix.

FIGURE 3

Characterization of MSC-EVs. (A) Representative AFM images of MSC-EVs obtained in the VWBR system, using MSC from three different human tissue sources (bone marrow, adipose tissue, and umbilical cord matrix). AFM height images (top) and respective 3D projections (bottom), capturing a total area of 10 × 10 μm. A close-up image focusing on a single EV is presented for each AFM height image. (B) Western blots of MSC lysates and MSC-EV samples. (i) Representative Western blot images of synthenin, CD63, CD81 and calnexin detection in MSC-EVs and corresponding WCL (i.e., cells) obtained from VWBR cultures. (ii) Western blot detection of synthenin, CD63 and CD81 in MSC-EV samples and corresponding WCL (i.e., cells), obtained from BM, AT and UCM MSC after EV production in static and VWBR systems. Detection of the housekeeping protein GAPDH in the same WCL preparations. (C) Zeta potential measurements of the surface charge of MSC-EVs (mV), obtained in either static or VWBR systems, using MSC from three different human sources (BM, AT, and UCM). Results correspond to one representative experiment for each condition. Results are presented as mean ± SD. AFM, atomic force microscopy; WCL, whole cell lysates; BM, bone marrow; AT, adipose tissue; UCM, umbilical cord matrix; VWBR, Vertical-Wheel^TM bioreactor.

FIGURE 4

Size distribution of MSC-EVs. (A) Representative size distribution curves of EV samples obtained from BM, AT, and UCM MSC, cultured in static or Vertical-Wheel^TM bioreactor systems. (B) Box plots representing the size distribution profiles of EV samples obtained from BM, AT, and UCM MSC, cultured in static or Vertical-Wheel^TM bioreactor systems. The minimum, 1st quartile, median, 3rd quartile and maximum values are represented for each condition. MSC from three different donors were used for each tissue source (i.e., n = 3 biological replicates). BM, bone marrow; AT, adipose tissue; UCM, umbilical cord matrix.

FIGURE 5

Comparing MSC-EV production in bioreactor and static culture systems, using MSC from different sources. (A) EV concentration (particles/mL) in the cell culture conditioned medium from BM, AT, and UCM MSC cultures in static and Vertical-Wheel^TM bioreactor systems. MSC from three different donors were used for each tissue source (i.e., n = 3 biological replicates). Results are presented as mean ± SEM (n = 3). Upper-right panel: Summarized paired analysis comparing EV concentration in static and Vertical-Wheel^TM bioreactor systems, for each MSC donor. Paired statistical analysis (paired t-test ^∗∗P = 0.0027) (n = 9). (B) Specific EV concentration (particles/cell) in the cell culture conditioned medium from BM, AT, and UCM MSC cultures in static and Vertical-Wheel^TM bioreactor systems. MSC from three different donors were used for each tissue source. In static cultures, each T-175 yielded 1.2 – 6.6 × 10⁶ cells upon 4 – 9 days of expansion, regardless of the cell tissue source. Results are presented as mean ± SEM (n = 3; n = 2 for UCM-static). Upper-right panel: Summarized paired analysis comparing specific EV concentration in static and Vertical-Wheel^TM bioreactor systems, for each MSC donor. (C) Particle to protein ratios (PPR) (particle/μg protein) of EV samples obtained from BM, AT and UCM MSC, cultured in static and Vertical-Wheel^TM bioreactor systems. MSC from three different donors were used for each tissue source. Results are presented as mean ± SEM (n = 3). Upper-right panel: Violin plot of PPR of MSC-EV samples obtained in static and Vertical-Wheel^TM bioreactor systems.