A COMSOL Multiphysics® Finite Element Model of the Diffusion Profile of Brain Derived Neurotrophic Factor: Biological Validation through a Microfluidics Device
Background: Sensorineural hearing loss is the most common type of hearing impairment. One novel method to treat this form of hearing loss is by regenerating the synaptic connections between the extant Spiral Ganglion Neurons (SGNs) and transplanted human stem cell derived SGNs. Brain-Derived Neurotrophic Factor (BDNF) plays an essential role in directing the growth of SGN neurites towards one another. The diffusion profile of BDNF as it is released from Polyhedrin Delivery System (PODS®, Cell Guidance Systems, Cambridge, UK) can help estimate parameters for desired experimental conditions in vivo. Using COMSOL Multiphysics® simulation software, we have developed a finite element model to analyze the diffusion profile of varying concentrations of BDNF through a 3D surface of a Xona Microfluidics device. We will then compare it to empirical experiments using immunohistochemistry as measurement tools to check for neurite growth.
Methods: Xona Microfluidics device measurements were taken to recreate a 3D surface in COMSOL Multiphysics®. This mesh surface was then used to create mathematical models of the diffusion profile of BDNF. The Coefficient Form PDE interface of COMSOL Multiphysics® can be used to create a mathematical model of the problem. The diffusivity coefficient of BDNF was approximated through the diffusivity of b-Lactoglobulin, a protein of similar size as BDNF. The reaction kinetics of BDNF from the PODS and its subsequent degradation was estimated from data of PODS containing Leukemia Inducing Factor (LIF). A MATLAB® curve-fitting algorithm was used to fit the concentration-time data from the LIF experiment and obtain kinetic constants for the model. The release of BDNF can be represented as a source term, and the degradation constant of BDNF will be used as the absorption coefficient. A Time Dependent study will be run for 14 days on the set-up described above.
Mathematical results were recorded and used to conduct empirical experiments on Xona Microfluidics device, using the same initial parameters. Immunohistochemistry was used to check for neurite growth. Standard estimates show that 10 ng/ml of BDNF is needed for optimal neurite growth in vitro. Thus, empirical results could be compared to mathematical model results to verify accuracy of the finite element model.
Results: The initial results from the Xona Microfluidics 3D surface model revealed the concentration diffusion of 50 million PODS crystals. Our simulation showed that this initial concentration could sustain optimal neurite growth over a period of time. Results of empirical experiments will need to be obtained and analyzed using immunohistochemistry. This will be compared to simulation results, and these results will be analyzed and presented at the conference.
Conclusions: Our computational models can help predict the optimal initial parameters needed to achieve desired experimental conditions in vitro. From the simulation results, it can be determined whether the desired BDNF concentration across the Xona Microfluidics device can be achieved with 50 million crystals. Further simulations on COMSOL Multiphysics®, which will be iteratively improved with empirical experiments, will be used to test other combinations of number of crystals and number of days that the simulation is run.
Herunterladen
- nella_poster.pdf - 1.32MB