Red cell manufacturing using parallel stirred‐tank bioreactors at the final stages of differentiation enhances reticulocyte maturation

Abstract

The aim of this study was to develop a robust, quality controlled, and reproducible erythroid culture system to obtain high numbers of mature erythroblasts and red blood cells (RBCs). This was achieved using a fully controlled stirred‐tank bioreactor by the design of experiments (DOE) methods in the serum‐free medium by defining the appropriate culture parameters. Human cord blood CD34+ cells were first cultured in static flasks and then inoculated to stirred‐tank bioreactors. Cell diameter was gradually decreased and final RBC yields were significantly higher when cells were inoculated at sizes smaller than 12 μm. The larger immature cells in the basophilic stage did not survive, while smaller mature erythroid cells were successfully expanded at high agitation speeds, demonstrating that appropriate seeding timing is critical. A high inoculation cell density of 5 × 106 cells/ml was achieved reaching 1.5 × 107 cells/ml. By using DOE analysis fitted to precise stages of erythropoiesis, we were able to acquire the optimal culture parameters for pH (7.5), temperature (37°C), dissolved oxygen, agitation speed (500 rpm), inoculation timing (cell diameter 12–13 μm), media feeding regimen, and cell seeding density (5 × 106 cells/ml). The final pure RBCs showed appropriate functions compared with fresh donor RBCs, confirming that manufacturing mature RBCs with reproducibility is possible.

Expansion and differentiation ofex vivocultured erythroblasts in scalable stirred bioreactors

“…Bayley et al also observed a decrease in proliferation in stirred bioreactors compared to static cultures, although no significance difference on growth was observed between stirring speeds of 300 and 450 rpm (tip speeds = 180 and 270 mm/s, respectively) (Bayley et al, 2017). Han et al observed a dependence on the inoculum age to agitation tolerance, with proerythroblast and basophilic erythroblast cultures showing a quick decrease in growth and viability in microbioreactors agitated at 300 rpm (tip speed = 180 mm/s), while more mature cells (day 12 after CD34 + isolation and start of culture), could tolerate the same conditions (Han et al, 2021). However, it is difficult to directly compare the hydrodynamic conditions between these reports, partly due to differences in the culture set-up but also due to a lack of consensus on which mixing parameter (rpm, tip speed, volumetric power input) is more appropriate to do this comparison, which together co-define the dynamic shear stress experienced by cells.…”

“…Bayley et al also observed a decrease in proliferation in stirred bioreactors compared to static cultures, although no significance difference on growth was observed between stirring speeds of 300 and 450 rpm (tip speeds = 180 and 270 mm/s, respectively) (Bayley et al, 2017). Han et al observed a dependence on the inoculum age to agitation tolerance, with proerythroblast and basophilic erythroblast cultures showing a quick decrease in growth and viability in microbioreactors agitated at 300 rpm (tip speed = 180 mm/s), while more mature cells (day 12 after CD34 + isolation and start of culture), could tolerate the same conditions (Han et al, 2021).…”

“…For cRBCs, STRs were used in small scale cultures (Han et al, 2021;Lee et al, 2018;Ratcliffe et al, 2012Ratcliffe et al, , 2012, or in an attempt to scale-up production (Law & Gilbert, 2021). Some studies report a large increase in cell numbers, but the conditions that were used varied widely with respect to medium composition, cell density, oxygenation, agitation (impeller type, speed), and nutrient feeding strategy, making a direct comparison to other cultivation systems challenging (Bayley et al, 2017;Han et al, 2021).…”

“…The different stages of erythroid cultures show distinct sensitivity to shear stress. Immature erythroid cells responded to shear stress with low viability and increased cell fragility, whereas cultures of more mature cells were robust and showed a linear increase in enucleation through the range of tested stirring speeds (Han et al, 2021). The effect of shear stress on erythroblasts may be mediated, at least partially, by the mechanosensitive channel PIEZO1, which activates multiple calcium-dependent signaling transduction cascades including the Calcineurin-NFAT pathway, the modulation of STAT5 and ERK signaling, and inside-out integrin activation (Aglialoro et al, 2020(Aglialoro et al, , 2021Caulier et al, 2020).…”

PIEZO1, sensing the touch during erythropoiesis

“…Indeed, although it is now possible to produce reticulocytes and RBCs in vitro without requirement of feeder cells [62], efficiency of the process is improved in three-dimensional culture systems mimicking the erythroid niche [63]. Shaking incubators or bioreactors are often required to upgrade RBC production to clinical scale [64][65][66]. The use of turbulent flow suggests that mechanical activation of differentiating erythroid cells might partly overcome the absence of cell-cell contacts with the micro-environment.…”