Contrast Adaptation in Natural Stimulus Encoding by Retinal Ganglion Cells

• Applicant: Professor Dr. Tim Gollisch
• Subject Area: Cognitive, Systems and Behavioural Neurobiology
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 515774656

Natural visual scenes vary strongly in contrast, that is, in the range of light intensities that they cover. Neurons in the vertebrate visual system adapt to the experienced contrast, and this contrast adaptation begins in the retina, the first stage of visual processing. Retinal ganglion cells, the output neurons of the retina, are known to adjust their sensitivity and temporal filtering characteristics when contrast changes. This contrast adaptation is likely a crucial component in the encoding of natural visual stimuli by retinal ganglion cells. However, most studies so far have investigated contrast adaptation in the context of simple stimuli that typically switch periodically between two levels of contrast, whereas natural scenes contain a wide range of continuously changing contrast levels. This project therefore aims at characterizing contrast adaptation in different types of ganglion cells of the mouse retina under complex spatiotemporal stimuli. A particular goal will be to develop computational models of visual stimulus encoding in ganglion cells with explicit contrast-adaptation modules and test these models on naturalistic stimuli. The model parameters will be fitted to the spiking activity of ganglion cells in response to various artificial stimuli, designed to tease out different aspects of contrast adaptation. The naturalistic stimuli used for probing the models will contain movie sequences as well as images that are shifted according to eye movement traces. Yet, to minimize effects of light adaptation, which are not the topic of this project, the movies will be altered by filtering out slow fluctuations in mean light intensity. To obtain the necessary data, we will perform extracellular multielectrode-array recordings of ganglion-cell spiking activity in isolated mouse retinas while projecting visual stimuli onto the retina’s photoreceptor layer. The recordings allow us to monitor hundreds of ganglion cells simultaneously for many hours and to classify different types of ganglion cells according to their characteristic light responses.

• DFG Programme: Research Grants