(306c) Modeling Liver Metabolism for Drug Development

Abstract

Successful evaluation of efficiency and safety of new medicines is a time consuming and expensive process.  One way to improve the initial drug screening is to develop high-throughput 3D cultures that mimic liver metabolism.  Liver tissue is targeted due to its pivotal role in metabolism of medicinal drugs, and hepatocytes are grown outside the body to test hepatotoxicity.  Recent advances have used porous structures onto which cells are seeded and cultured in static cultures or in bioreactors.  Media is introduced to the cells providing nutrients for cell survival and growth. Cells are subjected to initial drug concentrations to determine toxicity levels.  However, lack of understanding of the culture conditions and limitations has hampered the process of developing high-throughput technologies.  The objective of this study was to develop a liver metabolism model using computational fluid dynamics (CFD) for the purpose of creating an artificial environment capable of supporting three-dimensional liver tissue cultures.

We used a commercially available CFD software (Comsol Multiphysics 4.3a) to assess the effect of scaffold thickness, location within the culture, and porosity of the scaffold.  A three dimensional static culture with a scaffold 14mm in diameter and of various thicknesses to be placed in a six-well plate was simulated.  In static cultures, changes in the concentration of three components oxygen, estrogen and urea were evaluated for 48 h (typical media culture duration) using Michaelis–Menten rate constants from literature.  Initial concentrations were either based on solubility or physiological level.  Fluid properties were that of water, and effective diffusivities were based on Mackie-Meares relationship. Overall concentrations in the static culture were determined through multiple point analysis of Fick’s Law and rate laws at nodes throughout the three dimensional structure.  Model concentration distributions were tracked as scaffold thickness, elevation, and porosity were varied.  A constant cell density was used and increased thickness lead to increased cell number.

Time-dependent concentration profiles were analyzed for 1mm, 0.75mm, 0.5mm, and 0.25mm thick scaffolds of 0.85 and 0.25 porosities in order to determine limitations due to scaffold thickness.  These results showed that scaffolds thicker than 0.5mm scaffolds were not capable of supporting three-dimensional liver tissue cultures due to oxygen diffusion limitations.  Scaffold elevations of 0mm, 0.5mm, 1mm, and 1.5mm from the bottom of the bioreactor, for each of the scaffold thicknesses, were examined to investigate limitations due to scaffold placement.  Concentrations of all three molecules were affected by placement of the scaffold at 0mm as molecules could not diffuse into the scaffold through the bottom.  Only oxygen concentration profiles were affected by a further increase in elevation as the static bioreactor is open to the atmosphere and oxygen can diffuse into the media from air.

Due to scaffold thickness limitations in the static culture, a flow-through bioreactor of identical dimensions was simulated.  Scaffolds of 1mm and 2mm thicknesses were investigated at 0.25 and 0.85 porosities.  The shape of the bioreactor was initially optimized for uniform nutrient distribution.  Then concentration profiles were determined at steady-state for each molecule, using similar rate kinetic parameters in convective-diffusion equation.  Scaffold elevation was set at 1mm from the bottom of the bioreactor.  Fluid velocity was varied to obtain sufficient nutrient distribution for cell survival.  Flow-through bioreactors were constructed in-house according to the simulated dimensions.  Chitosan-gelatin scaffolds with 0.85 porosity were prepared by freeze drying. For static cultures, 6-well plates were used and experiments were performed for each media change using HepG-2 cells (from ATCC) and protocols recommended by the vendor.  Simulation results were validated by measuring oxygen consumption and urea production.  Using these conditions, one could develop multiple wells and bioreactors that could be used in rapid drug screening.