Bioreactor culture of embryonic stem cell aggregates: Oxygen transfer dynamics and population balance modeling of aggregation
Cadavid Olaya, Diana P.
MetadataShow full item record
This work studies in detail the oxygen mass transfer in mouse embryonic stem cell (mESC) and human embryonic stem cell (hESC) aggregates. Dissolved oxygen (DO) profiles for both mESC and hESC aggregates (known as embryoid bodies (EBs)) were obtained and it was found that mESC aggregates formed and cultured at 60 rpm (agitation rate) have a higher oxygen consumption rate followed by mESC aggregates at 80 rpm, static cultured (dishes), 100 rpm and hESCs at 60 rpm. Furthermore, size distribution data obtained from these experiments showed that an agitation rate increases the formation of smaller and more compact aggregates is promoted. With the results obtained from the DO experiments, parameters such as the maximum oxygen consumption rate and the oxygen permeability in the aggregates were calculated for mESC aggregates at 60, 80, 100 rpm of suspension culture and in static culture by solving the unsteady state diffusion-reaction equation for spherical coordinates with Michaelis-Menten kinetics to describe the oxygen consumption by the cells in the aggregates. The kinetic and diffusive parameters were consequently used to perform a simulation to calculate the effectiveness factor, Thiele modulus and oxygen profiles inside the aggregates for each condition. Porosity and tortuosity values from mESC and hESC aggregates were also determined and not significant variation was found on the porosity and tortuosity of mESC aggregates with time or agitation rate. On the other hand, hESC aggregates cultured at 60 rpm showed a significantly higher porosity than mESC aggregates under the same culture conditions which is predictable since mESCs have a higher proliferation rate (doubling time: 20 h) than hESCs (doubling time: 36 h) making the EBs more compact and less porous. In addition no significant difference was found in the tortuosity value of hESC and mESC aggregates at 60 rpm. Hypoxia Inducible factor (HIF-1–α) is known as a regulator of stem cell gene transcription when cells are exposed to low O 2 environments. In this study the expression of this protein was measured using qPCR assay to establish if mESC cell aggregates formed at 60, 80, 100 rpm and static conditions were experiencing hypoxia after 3 days of culture. The results showed that all cell studied expressed (HIF-1–α); however, it was markedly up-regulated and fairly similar in cells cultured as aggregates compared to cells cultured as monolayer. Application or population balance model to the bioreactor cultivation of mESC. Population balance models allow the prediction of changes in cell aggregate properties such as size, mass, growth, aggregation and disaggregation with time. Furthermore, these models give insights into how the formation of cell aggregate takes place as a result of their interactions with themselves and the surrounding medium . In this study mESCs were cultured as aggregates at different concentrations in static cultures. Size data were collected every day under no-shear conditions and several growth models were studied. The Gompertz equation was found to adequately describe the growth of stem cells within aggregates. Aggregates were also cultured in spinner flasks at different agitation rates and seeding cell concentrations. Size data were collected every day and aggregate size distributions under shear conditions were generated. A population balance equation model was set up to predict the temporal evolution of aggregate size depending on the rate of cell proliferation and collisions among cells and aggregates. Various expressions for the collection kernel were investigated and a coalescence frequency kernel was selected to model the interaggregate adhesion. The resulting aggregate size distributions were compared with experimental data from stem cell cultures. Such models are expected to be integral in the development of robust and scalable bioprocesses for stem cells. (Abstract shortened by UMI.)