DISCUSSION

Over the last years, numerous basic and clinical studies have focused on MSC due to their powerful regenerative potential and immunomodulatory features [reviewed in (Zhou et al., 2021)]. More recently, most of the potential beneficial effects of MSC therapies have been primarily attributed to EVs, with several studies confirming that EVs recapitulate many of the features of their parental cell line (MSC). Overall, MSC-EVs have been being explored for the potential treatment of spinal cord injury (Wang et al., 2021), acute kidney injury (Bruno et al., 2012), atherosclerotic cardiovascular disease (Badimon et al., 2022), myocardial ischemia (Charles et al., 2020; Ma et al., 2017; Zhu et al., 2018), among other diseases. To support all these clinical applications, large doses of EVs will be required and the limited manufacture capacity of traditional static culture systems will be a major hurdle towards the implementation of EVs-based therapies. In this context, the large-scale production of therapeutic MSC and their EVs will require the development of a cost-effective manufacturing platform based on fully controlled bioreactor systems for the expansion of well-characterized MSC populations and the production of their MSC-CM, as well as a robust downstream process to isolate and purify MSC and MSC-EVs [reviewed in (Mawji et al., 2022; Syromiatnikova et al., 2022)]. In this work, a S/X-free microcarrier-based culture system was successfully established for the expansion of MSC(M) and the production of MSC-EV using a 2 L-scale controlled STR as it represents a scalable, robust, cost-effective, and well-characterized platform widely used to produce biotherapeutics. In our previous work, MSC(M) and umbilical cord-derived Wharton’s jelly MSC [MSC(WJ)] were expanded on CellStart-coated plastic and Cultispher S microcarriers, respectively, using S/X-free StemPro culture medium and a bench-scale STR (1L) operated under repeated medium exchange mode (Dos Santos et al., 2014; Fernandes-Platzgummer et al., 2016). Based on those works, the operational process parameters were set to pH 7.2, dissolved oxygen (DO) 20%, temperature 37ºC and agitation rate 60 rpm (40 rpm during cell adhesion to the microcarriers). The entire culture was carried out in the STR, representing an advance compared to our previous protocol, where cell adhesion and initial culture (4 days) were performed in a spinner flask and subsequently transferred to the reactor (i.e. cell-containing microcarriers plus culture medium) (Dos Santos et al., 2014). After 24h, MSC(M) successfully adhered to the microcarriers, with adhesion efficiencies between 30-50%, and with a very homogeneous cell distribution, a factor identified to be crucial for a successful microcarrier-based culture (Carmelo et al., 2015). Despite losing more than 50% of the inoculated cells, carrying out the entire process inside the reactor has the advantages of having the whole protocol controlled from the first day of culture, while minimizing the risk of contamination associated to transferring the cell suspension between culture systems. Additionally, the adhesion efficiency can be maximized by optimizing the culture conditions in the first 24h, including the use of other agitation regimes or alternative microcarriers [reviewed in (Tsai & Pacak, 2021)]. For example, in previous studies from our group under S/X-free culture conditions, adhesion efficiencies of MSC(WJ) on gelatin-based Cultispher S microcarriers around 75% were attained in a 2L bioreactor culture (Mizukami et al., 2016), whereas initial adhesion efficiencies of 71±7.4 and 74±0.3% were obtained for MSC(M) using Low Concentration SynthemaxTM II and CellBIND® microcarriers, respectively, in spinner flask cultures (Carmelo et al., 2015).
Since culture medium represents one of the major costs in the manufacturing of human MSC, the delineation of feeding strategies able to maximize cell densities in a cost-effective way is of utmost importance. The feeding scheme adopted in most dynamic microcarrier-based culture systems used to expand MSC, is the repeated medium exchange strategy, where the cell-containing beads are allowed to settle before changing the culture medium once or twice a day (Cunha et al., 2015; de Almeida Fuzeta et al., 2020; Dos Santos et al., 2014; Lembong et al., 2020). This procedure requires stopping the reactor operation (i.e. discontinuous operation), which has the disadvantage of increasing the probability of aggregation between the cell-containing microcarriers during stagnant periods, potentially impairing cell growth and impacting cell identity. In this work, we hypothesized that an automated feeding protocol would potentially result in fewer fluctuations in cell proliferation/metabolism patterns and that a continuous operation would minimize cell-containing microcarrier aggregation. Therefore, two feeding operation modes were compared: FB, where the fresh culture medium was added discretely to the STR and FB/CP, where there was an automated continuous removal/replenishment of the medium with retention of the cell-containing microcarriers through a spin-filter. Growth curves of MSC(M) expanded under the two feeding regime strategies were quite similar during the first 6-7 days of cultivation, diverging upon this time point to reach considerable different maximal cell densities of (2.0±0.51)×105 and (4.1±0.90)×105 cells/mL at days 8 and 12 of cultivation for FB and FB/CP cultures, respectively. This difference could be explained by the accumulation of toxic by-products such as lactate, although it never reached values described as inhibitory to cell growth (35 mM (Schop et al., 2009)), and/or by the lack of replenishment of other important nutrients, as glucose concentration in the culture medium was always ranging from 2-6 mM, above the non-limiting value (over 1 mM (Schop et al., 2009)). Indeed, the concentrated glucose pulses added throughout culture, allowed us to maintain the glucose concentration within the desire range without adding volumetric capacity to the bioreactor.
Overall, continuous medium perfusion operation allowed MSC expanded under S/X-free conditions to reach densities in the STR that had previously only been achieved on small scale stirred culture systems (≤400 mL) operated under S/X-free conditions (Cunha et al., 2015), or often employing culture media containing human serum components (de Almeida Fuzeta et al., 2020; Lembong et al., 2020). The specific growth rates estimated were 0.6±0.1 and 0.5±0.1 for FB and FB/CP cultures, respectively, which are in agreement with the literature for MSC cultures cultured under S/X-free conditions (Carmelo et al., 2015; Cunha et al., 2015; Heathman et al., 2018). Other studies have compared different feeding modes of operation. In a previous work by our group, two different feeding regimes were compared for MSC(M) cultivation using a combined system employing a spinner flask transferred thereafter to a 1L-scale controlled STR: daily medium renewal every day or every 2 days versus fed-batch addition of concentrated nutrients and growth factors every 2 days. No significant differences were observed in terms of MSC(M) proliferation, although the fed-batch approach led to a faster accumulation of metabolites, namely lactate, as expected, as no culture medium was withdrawn from the STR. Moreover, a continuous perfusion process was tested in a smaller 400 mL STR with a dilution rate of 0.25 day-1 (starting from day 3) throughout the whole process. A maximal cell density of 5.0x105 cells/mL was reached at day 11, demonstrating the advantage of working under perfusion operation mode (Dos Santos et al., 2014). Cunha and colleagues also compared two different culture operation modes for expanding MSC(M) in a STR, 50% medium renewal every 2.5 days versus continuous perfusion cultures at a dilution rate of 0.2 day-1. The results attained showed that MSC achieved higher cell concentrations (3.7x105 cell/mL) and maximum growth rate (0.38 day-1) in continuous perfusion cultures when compared to the repeated medium exchange strategy (2.9 x 105cell/mL and 0.26 day-1), respectively. Although continuous perfusion processes led to higher cell densities when compared to fed-batch or repeated medium exchange operation modes, it utilizes larger quantities of culture media, which increases the operation costs. Moreover, the greater logistic of implementing it (especially in what concerns the cell retention device) and higher probability of technical failures have been hindering the use of this feeding operation mode for the production of MSC. In the present work, a spin-filter was used to retain the cell-containing microcarriers inside the STR. This system consists in a cylinder cage with a porous mesh wall, normally mounted on the impeller shaft. Perfusate (bleed) is pumped out from inside the spin-filter at the same rate at which fresh culture medium is pumped into the bulk of the STR (i.e. outside the spin-filter). Minimum fouling and an optimum cell retention at the necessary medium perfusion rate are crucial parameters for a successful operation of a spin-filter-based STR (Castilho & Medronho, 2002). No fouling phenomenon was observed in the mesh of the spin-filter in the present work and the cell-containing microcarriers were efficiently retained inside the STR, as no microcarriers were detected in the STR bleed. Another important parameter in continuous perfusion cultures is the dilution rate, as low perfusion rates result in growth inhibition due to nutrient exhaustion and/or accumulation of metabolites, and high perfusion rates results in wasting valuable medium components and over dilution of autocrine factors promoters of cell growth. In this context, as it has been previously demonstrated for human hematopoietic stem/progenitor cells (Madlambayan et al., 2005), the expansion of stem cell populations can be boosted by removing inhibitory factors produced by their more differentiated progeny. For those reasons, it would be interesting in the future to study different medium residences times in the STR platform operating under a continuous perfusion mode and their impact on MSC attributes.
Immunophenotype analysis before and after STR cultures revealed that MSC(M) cultured in S/X-free culture medium under stirred conditions maintained the high expression of CD73, CD90, and CD105, whereas the expression levels of hematopoietic cell markers (CD34, CD45, CD14 and CD19), and HLA-DR molecules were very low in all conditions, satisfying the minimal phenotypic criteria for describing human MSC (Viswanathan et al., 2019). The expression of CD105 decreased after the STR cultures, which was expected, as this event has been reported previously by our group and others (de Soure et al., 2016; Dos Santos et al., 2014), and may be attributed to longer exposure times to the enzymatic agent for cell detachment, which is known to affect surface receptors (Brown et al., 2007; Tsuji et al., 2017).
Overall, the results obtained in this work are in line with previous results from our group and others showing that MSC main features are well maintained upon cultivation under S/X-free stirred conditions (Carmelo et al., 2015; Cunha et al., 2015; Dos Santos et al., 2014).
At the end of STR cultures, the MSC-CM was collected and the EVs were successfully isolated and characterized by different techniques proposed by the International Society of Extracellular Vesicles (Witwer et al., 2017). TEM and Western blot techniques confirm the “cup-shaped” morphology of EVs and the presence of their characteristic markers, respectively and no significant different were found between EVs mean and mode sizes [(163±5.27) nm vs (162±4.44) nm and (134±4.23) nm vs (137±6.92) nm] for FB and FB/CP cultures, respectively (P>0.05). In what concerns the concentration of EVs and their purity in the MSC-CM retrieved from the STR cultures, average concentrations of (2.4±0.35)x1011 and (3.0±0.48)x1011 EVs/ml and similar PPRs of (1.7±0.21)x108 and (2.0±0.22)x108particle/µg protein were estimated by NTA and protein quantification for FB and FB/CP cultures, respectively. The MSC-EV densities obtained herein and in agreement with the results reported on the production of MSC-EVs using microcarriers in a Vertical-Wheel bioreactor (VWBR) by our group (de Almeida Fuzeta et al., 2020) in spinner flasks by others (Haraszti et al., 2018). To the best of our knowledge, this is the first work describing a fully controlled process that maximizes the proliferation of MSC(M) under S/X-free conditions in a 2L STR and the subsequent isolation of EVs from the enriched MSC-derived conditioned medium.

Conclusions

The bioreactor-based platform developed herein will allow to transform laboratory-based protocols into robust MSC and MSC-EVs manufacturing processes, with a tight control over the culture process and significant reduction of the production times. By addressing the manufacturing challenges of cell-based products, this technology is expected to facilitate translation of MSC therapies and likely to impact the development of therapeutic strategies employing MSC-EVs, which could rapidly progress towards clinical studies exploiting their potential as intrinsic therapeutics or as drug delivery systems (de Almeida Fuzeta et al., 2022; Syromiatnikova et al., 2022). In addition, this platform could be applied to the production of EVs from other parental cells lines (i.e. dendritic cells, natural killer cells) in therapeutic settings as cancer.
Author Contributions : Ana Fernandes-Platzgummer: Conceptualization, Investigation, Methodology, Writing - Original Draft, Funding Acquisition. Raquel Cunha: Investigation, Methodology; Marta Carvalho: Investigation; Sara Morini: Investigation; Juan Moreno-Cid: Methodology; Joaquim M.S. Cabral: Funding acquisition and Writing - Review & Editing; Cláudia Lobato da Silva: Conceptualization, Writing - Review & Editing and Funding Acquisition. All authors have read and agreed to the published version of the manuscript.
Acknowledgments : We would like to acknowledge Instituto Português de Oncologia Francisco Gentil and Clínica de Todos-os-Santos for their kind donations of BM samples for this study.
Funding : This research was financed by national funds from FCT—Fundação para a Ciência e Tecnologia (FCT), I.P., within the scope of the projects UIDB/04565/2020 and UIDP/04565/2020 of the Research Unit Institute for Bioengineering and Biosciences‐iBB, LA/P/0140/2020 of the Associate Laboratory Institute for Health and Bioeconomy—i4HB, and PTDC/EQU‐EQU/31651/2017 of the EXOpro project. This research was also funded by the project Bioprocessing Strategies for high productivity in the Culture of Human Stem Cells in Stirred Tank Bioreactors and their downstream. CELLS4ALL H2020-INNOVOUCHER-2018-0039.
Conflicts of Interest : J.M.-C. is employee of Bionet company.