(698a) A Detailed Model of Electroenzymatic Glutamate Biosensors Aids Applications in vivo

Electroenzymatic glutamate microbiosensors have proven useful for enabling local measurements of concentration transients in the brain, but due to the diffusional and kinetic resistances of the biosensor’s permselective and immobilized enzyme coatings, the sensor response cannot generally be assumed to be a true representation of actual glutamate signals. To model the biosensor’s ability to track changing concentrations and to optimize its ability to do so, a mathematical model describing transport through sensor coatings, ping-pong enzyme kinetics, and H₂O₂ electrooxidation kinetics was constructed and solved using a finite element solver (COMSOL). Optimization studies showed that reducing permselective Nafion thickness to 1 µm from ~10 µm, and the enzyme layer thickness to 3 µm from ~20 µm, could increase sensitivity several-fold to ~400 nA/µM/cm² and bring response times down from ~0.7 s to ~30 ms. Simulated biosensors with ms response times were also shown to accurately detect glutamate concentration peaks with standard deviations as small as <0.25 s, while slower simulated sensors reflective of the state-of-the-art underestimated the glutamate in such peaks substantially. The modeling results also showed that changing the relative proportions of glutamate oxidase and bovine serum albumin in the crosslinked enzyme layer had little effect on biosensor performance unless the enzyme layer was <5 µm thick. Finally, oxygen depletion within the enzyme layer at high (>5 mM) glutamate concentrations explains experimental observations in vitro of apparent K_m values for the immobilized enzyme that deviate significantly from the intrinsic. These modeling results suggest that the performance of electroenzymatic glutamate biosensors can be improved dramatically, which will contribute to better resolution of glutamate transients in vivo and better correlation of glutamate signaling with observed behavior.