Information Coding Strategy Of Bullfrog Retinal Ganglion Cells

Vision is crutial for humans and other primates, and it is one of the most important ways to get information from the environment. How visual system effectively encodes the rich and colorful visual scenes is one of important issues of neuroscience researches. Retina is the first stage in visual information processing. Retinal ganglion cell(RGC) is the only output of the retina, and its firing activities will change with different visual stimuli. RGCs can effectively encode visual information via many manners, such as firing rate, temporal structure of spike trains, and concerted activity, etc. Then, what is the relationship among RGCs’ coding strategy, visual stimulus patterns and retinal neural network activity? Using micro-electrode recording system, we performed electrophysiological/pharmacological experiments, combining with various methods such as information theory and computational model, to study the coding strategy of bullfrog RGCs, and also analyzed possible physiological mechanisms and significances of the coding strategy.Stimulus duration is one of the important features of visual stimulation. In the first part, we studied the coding strategy of ON and OFF responses of bullfrog changing contrast detector(transient ON-OFF RGC) to different stimulus durations. The results showed that both single cells’ firing patterns(response latency and spike count) and population activity pattern(correlation strength between RGCs) were changed with the stimulus durations, but more than 80% information about the stimulus durations was carried by the single cells’ activities. On the other hand, further studies showed that the coding strategy of ON and OFF responses was different, i.e. ON response mainly adopted response latency to encode stimulus durations, whereas OFF response mainly adopted spike count to encode. These results suggest that retinal ON and OFF pathways may adopt different coding strategies to transmit visual information.Dopamine is an important neuromodulator in the retina. In the second part, we analyzed the effects of dopamine on information coding of single cell. The results showed that though dopamine could shorten neuronal response latency and increase spike count, it only significantly attenuated the stimulus-duration-dependent response latency change, but had little effect on the stimulus-duration-dependent spike count change. Information analysis further showed that dopamine mainly decreased the information encoded by response latency, and did not obviously change the information encoded by spike count. These results suggest that retinal dopaminergic pathway is involved in modulating the stimulus-duration-dependent response latency change.Adaptation is one of the basic characteristics of the nervous system. In the third part, we studied coding strategy of bullfrog dimming detector(sustained OFF RGC) during luminance adaptation. The results showed that at the early stage of the adaptation, the stimulus information was mainly encoded in firing rates, whereas at the late stage of the adaptation, it was more encoded in synchronized correlations. Based on the results of computational models, we speculated that such a transition in coding properties was not a simple consequence of the attenuation of neuronal firing rates, but involved an active change in the neural correlation strengths. Our results revealed that in encoding a prolonged stimulation, the neural system may utilize synchronized, but less active, firings of neurons to encode information.In summary, our results showed that the information coding strategy of RGCs is related with retinal neural network state, and in exposure to different stimuli, RGCs can also dynamically and effectively change their way of information encoding. This dynamical information coding strategy can reflect the highly dynamic and efficient coding manner of RGCs.