Adipose stromal cells bioproducts as cell-free therapies: manufacturing and therapeutic dose determine in vitro functionality

Abstract

Background: Extracellular vesicles (EV) are considered a cell-free alternative to mesenchymal stromal cell (MSC) therapy. Numerous reports describe the efficacy of EV in conferring immunomodulation and promoting angiogenesis, yet others report these activities to be conveyed in EV-free bioproducts. We hypothesized that this discrepancy may depend either on the method of isolation or rather the relative impact of the individual bioactive components within the MSC secretome.

Methods: To answer this question, we performed an inter-laboratory study evaluating EV generated from adipose stromal cells (ASC) by either sequential ultracentrifugation (UC) or size-exclusion chromatography (SEC). The effect of both EV preparations on immunomodulation and angiogenesis in vitro was compared to that of the whole secretome and of the EV-free protein fraction after SEC isolation.

Results: In the current study, neither the EV preparations, the secretome or the protein fraction were efficacious in inhibiting mitogen-driven T cell proliferation. However, EV generated by SEC stimulated macrophage phagocytic activity to a similar extent as the secretome. In turn, tube formation and wound healing were strongly promoted by the ASC secretome and protein fraction, but not by EV. Within the secretome/protein fraction, VEGF was identified as a potential driver of angiogenesis, and was absent in both EV preparations.

Conclusions: Our data indicate that the effects of ASC on immunomodulation and angiogenesis are EV-independent. Specific ASC-EV effects need to be dissected for their use as cell-free therapeutics.

Keywords: Angiogenesis; Immune modulation; Mesenchymal stromal cells; Secretome.

Fig. 1

Scheme of isolation of ASC bioproducts. The ASC were cultured in traditional 2D culture flasks in serum-free medium for the production of EV-UC (A). The supernatant was collected and then ultracentrifuged for 2 h at 100,000 × g, the EV-UC were collected, cryopreserved and shared between the centers. For the production of EV-SEC (B), the ASC were cultured in a 3D hollow fiber bioreactor (HFBR), from which the supernatant was collected and concentrated with 100 kDa filter. The samples were then processed with size exclusion chromatography and the EVs and protein fraction were concentrated with a 100 kDa filter and collected. The EV-SEC and the protein-rich fraction were shared between centers. Finally, the ASC were cultured in traditional 2D culture flasks, the supernatant was collected and concentrated with 3 kDa and the conditioned medium and wash-off were collected during the process (C), cryopreserved and shared between centers

Fig. 2

Characterization of the ASC bioproducts. A FACS analysis using the MACSPlex exosomal kit. EV-UC, EV-SEC, Protein-rich fraction, CM and CM-WO were tested for 39 surface markers divided into groups based on their origin and function (embryonic cells, pro-coagulation activity, neurite growth, cell adhesion, hematopoietic cells, immune system regulation, immune cells, mesenchymal cells, tetraspanins and two controls). Data represented as a heatmap with the mean fluorescence intensity of N = 3 ASC donors per marker. B Representative super-resolution microscopy pictures of EV-UC, EV-SEC, CM. The samples are stained with CD81 red, CD44 green and CD9 blue. The scale bar is 100 nm. C Protein content analysis showed higher concentration of protein in CM and Protein-Rich Fraction preparations and lower in EV-UC. No protein content was detected in the EV-SEC or the CM-WO groups. D Western blot analysis of Calnexin of EV-UC, EV-SEC, Protein-rich fraction, CM-WO, CM. Cell lysates of Hela cells were used as a positive control

Fig. 3

Immunomodulatory properties of ASC bioproducts. A PBMC proliferation after five days of co-culture with ASC under PHA stimulation. All values were normalized to PHA-stimulated monoculture PBMCs. B LPS-stimulated THP-1 cells showed an increase in their phagocytic activity when compared to non-stimulated cells (Neg Ctrl). Treatment with EV-SEC, Protein-Rich Fraction and CM enhanced the number of positive phagocytic cells with respect to the Neg Ctrl and the EV-UC group. C Increasing the ratio of EV:THP-1 cells to 20:1 significantly enhanced the phagocytic stimulation of both EV preparations. Data are represented as the mean ± SD of N = 3 ASC donors and n = 6 technical replicates. Statistical analysis was performed using Two-Way ANOVA with Tukey’s multiple comparison test; * = p < 0.05, ** = p < 0.001, *** = p < 0.0001, **** = p < 0.00001. # Significance versus the negative control

Fig. 4

Enhancement of cell migration by ASC bioproducts. A Protein-Rich Fraction, CM and EV-UC preparations significantly enhanced endothelial cell migration with respect to the negative control (EndoGro-LS medium without FBS and VEGF, positive control with FBS and VEGF) in an in vitro wound healing model at 8 h after injury. CM and Protein-Rich Fraction displayed higher abilities to promote cell migration than any other treatment. B Migratory capacity of ASC 24 h after injury showed a tendency to increase after the addition of CM respect to the negative control (MEM-⍺ media without FBS) and positive control (MEM-⍺ media with FBS) (C) Increasing the ratio of EV:MSC leads to a significant increase in cell migratory capacity of EV-UC. Data are represented as the mean ± SD of N = 3 ASC donors and n = 6 technical replicates. Statistical analysis was performed using Two-Way ANOVA with Tukey’s multiple comparison test; * = p < 0.05, ** = p < 0.001, *** = p < 0.0001, **** = p < 0.00001. # Significance versus the negative control

Fig. 5

Angiogenic properties of ASC-derived bioproducts. Tube formation assay was chosen as inter-center comparison between centers. A Ability of ASC bioproducts to stimulate tube formation in vitro is represented as the relative tube formation of the negative control (EndoGRO-LS medium without VEGF). A wide range of tube formation was achieved in each center, with similar trends observed among groups, whereby CM and Protein-Rich Fraction showed superior angiogenic support abilities. B Representative phase contrast images of tubule-like networks in culture. C Increasing the EV ratio to 20:1 did not impact tube formation. D VEGF concentration was highly found in CM and protein-rich fraction but absent in both EV preparations. E Addition of 1 µM of ZM32381 in the CM and the protein-rich fraction preparations resulted in reduced tube formation. Data are represented as the mean ± SD of N = 3 ASC donors and n = 3 technical replicates per center. Statistical analysis was performed using One-Way ANOVA (D, E). * p < 0.05, ** p < 0.01, *** p < 0.001