Mechanisms of color processing in the retina
by Mohammad Hossein Khani
Date of Examination:2017-12-14
Date of issue:2018-11-27
Advisor:Prof. Dr. Tim Gollisch
Referee:Prof. Dr. Tobias Moser
Referee:Prof. Dr. Siegrid Loewel
Referee:Prof. Dr. Stefan Treue
Referee:Prof. Dr. André Fiala
Referee:Marion Dr Silies
Files in this item
Name:Mohammad Khani, PhD Thesis.pdf
Size:22.3Mb
Format:PDF
Description:PhD Thesis
Abstract
English
In the sensory system, many neurons are specified to detect different environmental signals and transfer the signals to downstream neurons. These downstream neurons integrate the detected signals and generate an output that is evolutionary relevant for the organism. In the retina, the photoreceptors are specified to detect light from the environment at different wavelength, thus separating the input light into different chromatic channels. The chromatic signals from the photoreceptors are integrated by downstream bipolar and ganglion cells. Although chromatic properties of the photoreceptors and the bipolar cells are relatively known, it is not clear how retinal ganglion cells integrate their chromatic inputs. In this project, we studied the properties of chromatic signal integration in the mouse retina. We used the mouse retina because of its dichromatic light sensitivity to both UV (ultra-violet) and green light. We first built a new projection system that allowed us to stimulate both s-cones (UV sensitive) and m-cones (green sensitive) of the mouse. We designed a new stimulation approach that allowed us to probe the question of chromatic integration. In this stimulation approach, we showed opposing contrasts of the UV and green light simultaneously and checked the responses of the ganglion cells with multi-electrode arrays. We focused on the stimulus combinations where the opposing contrasts of UV and green could effectively cancel each other out (cancellation point). Based on the responses of the ganglion cells at the cancellation point, we asked whether ganglion cells integrate their chromatic inputs linearly or nonlinearly. The chromatically linear cells did not showed any different responses from their spontaneous activity at the cancellation point. The nonlinear cells, on the other hand, showed reliable responses at the cancellation point. Moreover, we found a separation between the responses of the On and Off nonlinear cells. The Off cells showed increased firing rates at the cancellation point. The On cells, however, showed suppressive responses by reducing their activity below their spontaneous firing rates. Next, we classified the ganglion cells based on the chromatic integration properties and their responses to the preferred and non-preferred light stimulus. Further, we used computational models to simulate both linear and nonlinear chromatic integration. In the final step of this thesis, we looked into potential mechanisms of nonlinear chromatic integration. We used a grating variant of our color integration stimulus that allowed us to stimulate the center of the receptive more effectively than the surround. We observed that by reducing the effect of the surround the chromatically linear cells do not change their integration properties while the nonlinear cells become linear. Thus, we concluded that the nonlinearity of the chromatic integration is induced by the receptive field surround of the ganglion cells. To conclude, we found a new type of nonlinear computation in the mouse retina, which can assist the current functional classifications schemes of the retinal ganglion cells that usually ignore the chromatic integration properties. Furthermore, we believe that our experimental approach could potentially be used in other sensory systems to investigate the nonlinear computations of neurons.
Keywords: color; retina; chromatic signal integration; retinal ganglion cells; visual system; neuronal signal integration; linear and nonlinear signal processing in the retina; linear and nonlinear chromatic integration; color processing