Identification of calcium current reversal in bipolar cells

Electrical stimulation of neurons is employed in the field of neuroprosthetics to artificially generate or block neural activity. Retinal implants aim to restore vision to the blind suffering from degeneration of their photoreceptors, cells which transform light input into neuronal output. The activation of the remaining, healthy neurons in a controlled way is one possibility to restore a rudimentary kind of vision to these patients. A general approach for coordinated electrical stimulation of retinal neurons, however, has not been found so far. Bipolar cells are the main targets when stimulation is applied from the side of the degenerated photoreceptors, so called subretinal stimulation. These cells connect photoreceptors with ganglion cells, the output neurons of the retina, and thereby perform several pre-processing steps of incoming neuronal signals. Bipolar cells have to be depolarized, that is, increasing the membrane potential, at their terminals in order to generate synaptic release, which consequently activates ganglion cells. Whereas bipolar cells mainly generate so called graded potentials which control their synaptic output, ganglion cells generate action potentials to code visual information. As reported recently, stimulus amplitude and consequently strength of depolarization has to be in a specific range in order to evoke synaptic activity at bipolar terminals. By over-stimulating bipolar cells these are assumed to shut down their activity and thus are not likely to generate physiologic responses as during normal vision. The reason for decreased synaptic activity are inverted calcium currents, which do not elevate the intracellular calcium concentration necessary to initiate the synaptic signaling cascade. This calcium current reversal and all further implications during stimulation of the retina, however, were obtained in a modeling study and only little physiologic data is currently available to confirm these results. Therefore, in the course of this project multiple electrophysiological experiments (e.g. patch clamp technique, electrical micro-stimulation) in mammalian retina in-vitro will be performed in order to prove the existence and the impact of the current reversal. Understanding the implications of this biophysical principle is of importance for both, neuroprosthetics as well as neurophysiology.

Electrical stimulation of nerve cells is employed in a variety of neuroprosthetic devices that aim to restore sensory and tactile body functions. The artificial activation of nerve signals works in a range of stimulus amplitudes, between the so-called lower and upper threshold. Especially in retinal implants, that restore vision to the blind, this phenomenon is not well understood. Therefore, this project initially aimed to exploit the varying sensitivity of lower and upper thresholds in different retinal cell types to develop more sophisticated stimulation strategies; this was expected to result in more natural artificial vision. Within the course of the project we found a second biophysical element in retinal neurons which can be exploited to achieve varying sensitivity to extracellular electric stimulation. The so-called axon initial segment in retinal ganglion cells is a highly attractive target for electrical stimulation as it is located on the output neurons of the retina. We could show that the axon initial segment is highly important for both, physiological function as well as the response to electrical stimulation. Our results suggest that the axon initial segment is tailored to the pre-synaptic inputs that a given cell type receives from the retinal network. Therefore, micro-adjustments in the length and location of the axon initial segment are necessary to generate meaningful output in single retinal ganglion cells. A second important finding originated from a computational study that investigated the response of retinal bipolar cells to repetitive electrical stimulation as used in the PRIMA System, a retinal prosthetic currently tested in clinical trial. Our results provide a biophysical explanation why stimulation frequencies larger than 20 Hz can improve the quality of elicited phosphenes by showing that higher frequencies can mimic natural vision (i.e. elicited by light input) more closely than lower frequencies. This is particularly important for the development of new stimulating strategies that can generate visual perceptions which resemble natural vison as close as possible. Furthermore, our results show the variability of different cell classes and types of the retina to electrical stimulation. We probed the response to single pulses as well as to high frequency stimulation and found distinct differences between cell types. Importantly, we could also show that these responses are similar across species; this raises hope that results obtained in lower order species such as mouse will also translate into humans. In sum, results from this project revealed several intriguing response characteristics of retinal neurons during electrical stimulation which could potentially be useful for the development of the next generation of retinal prostheses.