Bernstein members involved: Marion Silies

When light conditions change quickly, the eye has to react in a fraction of a second in order to continue to see well. This is helpful or even necessary when, for example, we drive through a forest and steer out of the shade of trees into the sunlight and then back into the shade. “In such situations, it is not enough for the photoreceptors in the eye to adapt; an additional correction mechanism is required,” explained Professor Marion Silies of Johannes Gutenberg University Mainz (JGU). In previous studies, her research group has already shown that the fruit fly Drosophila melanogaster has a correction mechanism that is located directly behind the photoreceptors. Silies’ team has now decoded the algorithms, mechanisms and neuronal circuits that make it possible to maintain stable vision in rapidly changing light conditions. The work has been published in Nature Communications.

Abrupt changes in light pose a challenge

Whether we are moving around ourselves in our environment or following an object with our eyes as it moves from light to shadow, our vision has to function in many different situations. This is true for us humans and also for thousands of animal species that rely heavily on visual orientation. Even in the inanimate world, rapid changes in light pose a challenge for information processing, for example for camera-based navigation systems. This is why many self-driving cars also use radar or laser technology to correctly detect the contrast between an object and its background. “Animals can do this without the corresponding technology. So can we learn from animals how visual information is processed stably in rapidly changing light conditions?” says Silies, formulating the research question.

Theoretical and experimental work were combined

The compound eye of Drosophila consists of 800 individual eyes. Behind the photoreceptors, the contrast between an object and the background is determined. However, differences in the contrast calculation occur here if light conditions, i.e. the background, suddenly change, for example if an object moves into the shadow of a tree. This would have consequences for all subsequent steps of visual processing and would make the object look different. Using two-photon microscopy, the study with first author Dr. Burak Gür shows at which point in the visual system stable contrasts are encoded for the first time. In this way, neuronal cell types were identified that are located two synapses behind the photoreceptors.

However, these cell types react very locally to visual information. In order for the brightness of the background to be correctly included in the contrast calculation, this information must be bundled across the space, so to speak, as calculations by co-author Dr. Luisa Ramirez show using a computer model. “So on the one hand, we started from a theoretical approach that shows which radius would be optimal in images of natural environments in order to capture the background brightness over a certain space, and in parallel we looked for a cell that takes over this function,” explains Prof. Dr. Marion Silies, who heads the Neural Circuits research group at the Institute of Developmental Biology and Neurobiology (IDN) at JGU.

Brightness information is bundled before transmission

The neurobiologists have found a cell type that meets all the necessary criteria. These cells, called Dm12, bundle brightness signals over a certain radius, which in turn corrects the contrast calculation between object and background in rapidly changing light conditions. “We have thus uncovered the algorithms, circuits and molecular mechanisms that make vision stable, even when brightness changes abruptly,” summarizes Silies. The scientist began researching the visual system of fruit flies 15 years ago and suspects that brightness correction in vertebrates and even humans works in a similar way, especially as the neuronal prerequisites are in place.