Simulation Of The Electrical Field Generated By Bipolar And Ganglion Cells In The Electrically Stimulated Retina

Abstract

Event Abstract Back to Event Simulation Of The Electrical Field Generated By Bipolar And Ganglion Cells In The Electrically Stimulated Retina Paul Werginz1* and Frank Rattay1 1 Vienna University of Technology, Institute for Analysis and Scientific Computing, Austria Motivation Recording of the (summed) response of bipolar (BCs) and ganglion cells (GCs) is essential for examination of cell activation in in-vitro experiments in order to gain a better understanding of the electrically stimulated retina. Here, we describe a proof of concept to simulate these electrical fields during extracellular stimulation. Material and Methods The response of BCs and GCs to extracellular electrical stimulation is computed with multi-compartment models. Our modeling discretizes a simplified 2-D cell geometry in space, using cylindrical and spherical compartments. Current flow between compartments and across the cell membrane leads to a system of ordinary differential equations (Eq. 1 [1]). Ionic currents are computed with a Hodgkin-Huxley-like model approach for 35 degree Celsius [2]. Here the extracellular field generated by a stimulating electrode is computed by a simple analytical solution for point sources in homogeneous media (Eq. 2) Extracellular recordings were computed with the same formula; see [3] for details. The used model of the ribbon synapse at BC terminals is presented in [4]. Eq. 1 Eq. 2 Results A single extracellular action potential is bi- or triphasic, depending on the position of the recording electrode and the site of spike initiation. For direct GC stimulation recordings at or close to the soma resulted in biphasic waveforms whereas recordings at the distal axon resulted in triphasic waveforms (Fig. 1). Additionally, the amplitude of the recorded waveforms varies upon the location of the recording electrode, the geometric properties of the model neuron (e.g. fiber diameter in Fig. 1B) and ion channel equipment along the neural membrane. When BC input was fed to the model GC bursts of action potentials were generated and in the extracellular recording the synaptic input could be displayed (Fig. 2). Calcium currents that drive the synaptic signaling cascade in BC terminal compartments, however, could not be revealed because of the stimulus artifact. Discussion Direct stimulation of GCs from the epiretinal space generated responses that are qualitatively similar to data obtained in in-vitro experiments [5]. Periodic synaptic input during repetitive stimulation led to bursting spiking activity in GCs. Applied stimulus amplitudes resulted in synaptic (indirect) activation whereas no direct GC spikes were generated. The spiking rate is decreased for later pulses because of vesicle depletion in the ribbon synapses. Recorded intra- and extracellular responses revealed the synaptic inputs, however, the signal was rather small in the extracellular recording. Conclusion The presented initial modeling study shows that it is generally feasible to simulate the extracellular field generated by retinal neurons during electrical stimulation.