(293h) Residence Time Distribution and Evaluation of Exit Concentration in a New Generation Axial Flow Bioreactor

The axial flow bioreactors have recently attracted attention in tissue engineering applications for regenerating a variety of high aspect ratio tissues, such as skin, liver, and bladder.  They offer several advantages over conventional perfusion bioreactors such as flow through and parallel flow.  The nutrient distribution in axial flow bioreactors is mainly convection-driven, and the ability to operate at high flow rate with relatively low pressure.  However, the current designs use randomly selected dimensions with little analyses related to the hydrodynamic characteristics and nutrient distribution.  The objectives of this study were to analyze the RTD in the axial flow bioreactor with two different scaffold preparations, and validate the exit concentration predictions from CFD and segregation model using an experimental setup.  

The design of axial flow bioreactor used in this study was optimized using Computer Aided Design (CAD) and Computational Fluid Dynamics (CFD) to support 100-mm diameter and 2-mm thick scaffolds.  Various factors such as inlet diameter, distributor system, and outlet diameter were investigated to maximize nutrient distribution across the scaffold and minimize hold up volume, dead volume, and pressure drop across the bioreactor.  In addition, RTD experiments were performed on the bioreactor to examine nutrient distribution and evaluate the mean residence time, similar to previous publication related to flow-through bioreactors [1].  The RTD experiments were performed using scaffolds prepared from two different technique, chitosan-gelatin using freeze-drying technique [2] and PCL scaffold using salt leaching technique.  The scaffolds were characterized using scanning electron microscopy (SEM) for pore size and porosity.  The physical properties were estimated using Instron 5542 tension compression test machine (INSTRON, Canton, MA) for Young’s Modulus and a custom-build apparatus for Poisson’s ratio values.  

Each hepatocyte within the scaffold was considered to be a small batch reactor and the RTD was combined with the Michaelis–Menten rate expression using the segregation model to determine exit concentration.  Furthermore, to validate the assumption of cells as small batch reactors in segregation model, the exit concentrations of nutrients were simulated using CFD, similar to our previous publication [3].  CFD has been shown to be effective in analyzing biological systems, studying flow pattern, and understanding consumption kinetics in bioreactors [4].  The model was setup with fluid flow in the porous scaffold described by Brinkman’s equation and other domains were described by Navier-Stokes equation.  The exit concentrations and nutrient distribution were evaluated using the convection-diffusion equation, which links flow field with concentration variances.  The structural mechanics of the scaffold was coupled with fluid flow and consumption kinetic on a moving mesh, using the two way coupling method [5], to include the effect of scaffold deformation on nutrient consumption.  Experiments were also performed using three different cells densities of HepG2 cells on chitosan-gelatin scaffold.  A cell density of 0.6×10¹² cells/m³, 1.2×10¹² cells/m³, and 2.4×10¹² cells/m³ was used for the experiments.  The cells were first cultured to desired number and seeded on the scaffold in a petri dish.  The cells were allowed to attach the scaffold in a static culture for one day before transferring the scaffold into the bioreactor.  The experimental setup was fitted with oxygen probes at the inlet and outlet of the bioreactor to measure oxygen concentrations, a sampling port was used at the outlet to collect sample for exit glucose concentration measurement.  A pressure transducer was connected to the flow line to monitor pressure drop.  The flow system was allowed to reach steady state (4 times the space time) following which oxygen concentration and glucose sample was obtained.  The results showed uniform nutrient distribution across the scaffold.  The dead volume in the bioreactor was less than 3 mL, which corresponds 93% nutrient distribution in the bioreactor.  The chitosan-gelatin scaffold showed better nutrient distribution than PCL scaffold due to its hydrophilic nature.  Our integrated approach provides a useful strategy to designing bioreactors for tissue regeneration.

[1] Lawrence, B.J., M. Devarapalli, and S.V. Madihally, Flow dynamics in bioreactors containing tissue engineering scaffolds. Biotechnology and Bioengineering, 2009. 102(3): p. 935-947.

[2] Podichetty JT, Madihally SV. Dynamics of Diffusivity and Pressure Drop in Flow-Through and Parallel-Flow Bioreactors During Tissue Regeneration. Biotechnology Progress. 

[3] Devarapalli M, Lawrence BJ, Madihally SV. Modeling Nutrient Consumptions in Large Flow-Through Bioreactors for Tissue Engineering. Biotechnology/ Bioengineering. 103(5):1003-1015, 2009.

[4] Patrachari AR, Podichetty JT, Madihally SV. Application of computational fluid dynamics in tissue engineering. Journal of Bioscience and Bioengineering. 2012.

[5] Podichetty JT, Madihally SV. Modeling Growth Medium Perfusion-induced Porous Scaffold Deformation. Acta Biomaterialia,(submitted). 2013.