Modeling the Electro-chemical Properties of Microbial Opsin ChrimsonR for Application to Optogenetics-based Vision Restoration

Abstract

Optogenetic activation of neurons [1] have greatly contributed to our understanding of how neural circuits operate, and holds huge promise in the field of neural prosthetics, particularly in sensory restoration. The discovery of new channelrhodopsins, Chrimson -which is 45 nm more red-shifted than any previously discovered or engineered channelrhodopsin -and its mutant ChrimsonR with faster kinetics [2] made this technology available for medical applications. However, a detailed model that would be able to accurately reproduce the membrane potential dynamics in cells transfected with ChrimsonR under light stimulation is missing. We address this issue by developing the first model for the electrochemical behavior of ChrimsonR that predicts its conductance in response to arbitrary light stimulation. Our model captures ON and OFF dynamics of the protein for stimuli with frequencies up to 100 Hz and their relationship with the brightness, as well as its activation curve, the steady-state amplitude of the response as a function of light intensity. Additionally, we capture a slow adaptation mechanism at a Introduction 1 This paper studies the activation dynamics of ChrimsonR, a channelrhodopsin which is 2 currently the best candidate for optogenetic-based sensory restoration in human 3 patients, due to its peak spectral sensitivity in the red at 590 nm, which is 45 nm more 4 red-shifted than all previously known channelrhodopsins [3]. However, design of 5 light-stimulation protocols for optogenetics-based vision restoration would greatly 6 benefit from a model of the channel's behavior, as its dynamics will influence the 7 PLOS 1/25ChrimsonR mediated temporal dynamics of membrane potential in transfected neurons. 8To address this issue, here we introduce for the first time the relevant kinetics and the 9 expected behavior of the channel. In particular, we propose a predictive model of the 10 conductance of a population of channels expressed in a given cell in response to 11 arbitrary illumination stimuli spanning the whole dynamic range of the protein. 12Upon photostimulation, channelrhodopsins modify their conformation and form a 13 water-filled pore allowing some ions to flow through the channel. The illumination 14 signal therefore drives the conductance of the cell's membrane, which then induces a 15 current through the membrane, the subsequent modification of the membrane potential 16 and the possible triggering of action potentials. A predictive model of the 17 channelrhodopsin's conductance provides the mapping between the controlled 18 illumination and the physiologically relevant variable. 19The main features of this transformation that we capture are the ON and OFF 20 dynamics (and their relationship to the brightness value of the stimulus); the activation 21 curve, the steady-state amplitude of the response as a function of light intensity as well 22 as a slow adaptation mechanism leading to a progressive decrease in the amplitude of 23 the response. These features can be captured using a five-state Markov k...

Optogenetic therapy: high spatiotemporal resolution and pattern discrimination compatible with vision restoration in non-human primates

“…Consistent with this observation, recordings of activity in response to achromatic binary white noise with a 50% pseudorandom selection rate revealed a gradual decline of evoked activity. The underlying mechanism of this modulation may involve an inactivated state of the microbial opsin 30 or the inactivation of the voltage-gated channels in the ganglion cells. A simple monochrome transformation of natural images would result in a large number of pixels with high values (i.e., light gray) potentially leading to rapid deactivation of retinal ganglion cells.…”

Cortical visual prosthesis: a detailed large-scale simulation study

“…The model has been implemented using the Mozaik neural simulation workflow framework [8] and the Arkheia tool [7]. Here, we have extended the Mozaik framework with three additional components: the model of MLEE, the model of light propagation in cortical tissue, and a model of channelrhodopsin dynamics [65]. The NEST simulator [36] was used as the back-end for all simulations described in this paper.…”

“…The parameters of this model have been fitted by Sabatier et al [65] to light (590nm wavelength) stimulation experiments in ChrimsonR-expressing HEK293 cells, and here we use the parameter values reported in that study. 4…”