Periodic Reporting for period 1 - MOVIS (Neural arithmetic of motion vision)
Reporting period: 2021-07-01 to 2023-06-30
The brain is thought to derive much of its computational power from arithmetic operations carried out by individual nerve cells. The aim of the project ‘Neural arithmetic of motion vision’ — MOVIS is to understand how these arithmetic operations are implemented at the cellular and molecular level.
The visual system in the brain of the fruit fly houses nerve cells with remarkable abilities: These neurons compute the direction of visual motion based on information from adjacent light-sensitive cells of the eye. A formal model postulates that this would require the multiplication of two signals, but how a single nerve cell can perform a multiplication has been a puzzle for decades.
The visual system in the brain of the fruit fly houses nerve cells with remarkable abilities: These neurons compute the direction of visual motion based on information from adjacent light-sensitive cells of the eye. A formal model postulates that this would require the multiplication of two signals, but how a single nerve cell can perform a multiplication has been a puzzle for decades.
Work performed over the course of the project describes how direction-selective neurons implement such a multiplicative mathematical operation. Experiments measuring the voltages across the surfaces of nerve cells in live flies revealed that the electrical activities of these cells are subject to constant inhibition. A brief release from this inhibitory signal, occurring only when visual stimuli move along one specific direction, amplifies the coincident excitatory signal from a neighbouring location. This amplification is formally equivalent to a multiplication. The nerve cell’s ability to multiply was linked to a receptor molecule on its surface. Animals lacking this receptor on direction-selective neurons failed to stay on course in a virtual reality. This illustrates the behavioural relevance of a specific computation in a clearly defined type of neuron.
Similar neural computations are thought to underly our abilities to localize sounds, to navigate, or to focus our attention. The comparatively simple brain of the fruit fly offers an unparalleled opportunity to draw back the curtain on some of these seemingly intractable neuroscientific problems.