Longitudinal Study of Neurosphere Metabolism in Bioreactor Culture for Scalable Production of Human Embryonic Stem Cell Derived Neurons

Human embryonic stem cells  (hESCs) offer great potential for multiple tissue engineering applications including drug screening and future clinical applications such as treatment of neurological disease.  Today, state-of-the-art technology lack the capacity to support systematic larger scale applications. The challenge resides not only in enlarging the cultivation size but also in controlling the quality Herein, long-term expansion of hESC derived free-floating neurospheres in stirred bioreactor systems enhanced proliferation and viability and also improved the differentiation potential into various neuronal lineages. We examined in detail variations in cellular metabolism during expansion under hypoxic and normoxic conditions in static versus stirred cultivation systems, orbitally or impeller-mixed over an extended time by analyzing nutrients and metabolites including amino acid content in growth medium.  We found that an orbitally stirred bioreactor system promotes larger sphere size and higher cell yield. Furthermore, the bioreactor system promoted a higher ammonium release, mainly a by-product of oxidative phosphorylation. This was seen also during hypoxia and was accompanied by changes in glutamine and glutamate consumption, indicating a switch towards increased efficiency of metabolic pathways generally activated at atmospheric oxygen. This metabolic pattern observed in the stirred bioreactor was accompanied with enhanced consumption and/or production of many non-essential amino acids mimicking a high oxygen condition. Furthermore, glutamic acid, aspartic acid and serine could be partially or fully depleted thus indicating that these amino acids may be a limiting factor during large-scale production of neurospheres. Together, these data give new insight in the specific needs when designing platforms for large-scale propagation of neuronal progenitors and differentiated neuronal subtypes.