Boy fly meets girl fly meets AI: Training AI to recognize fly mating identifies mating gene

A research team at the Graduate School of Science, Nagoya University in Japan has used artificial intelligence to discover that Piezo, a method that receives mechanical stimuli, helps control the posture of male flies (Drosophila melanogaster). The inhibition of the Piezo made the flies unable to climb which reduced their reproduction. Their findings were reported in iScience.

Most previous studies of animal mating have focused only on behavior, limiting our understanding of this important process. Since most animals are stable when mating, successful mating is critical to reproductive success. In fruit flies, the male mounts the female and maintains this position until he has transferred enough sperm to fertilize the female, which occurs about eight minutes after mating. A research team at Nagoya University discovered that there is another factor that makes this possible.

One contender is Piezo. Piezo is a family of transmembrane proteins found in bristle cells, sensitive cells in the male genitalia. The piezo opens when a mechanical force is applied to the cell membrane, allowing ions to move through the cell and create an electrical signal. This signal triggers a cellular response, including the release of neurotransmitters in neurons and the contraction of muscle cells. Such a response helps the fly to maintain its perch.

After discovering that the piezo gene is involved in fruit fly mating, Professor Azusa Kamikouchi (him), Assistant Professor Ryoya Tanaka (him), and student Hayato M. Yamanouchi (him) used optogenetics to investigate. the neural mechanism of this phenomenon. The technique combines genetic engineering and eye science to create genetically modified neurons that can be blocked by light of a specific wavelength. When the light was turned on during mating, the neuron was blocked. This allowed the researchers to control the activity of the piezo-expressing neurons.

“This was difficult for us,” Kamikouchi said. “Using optogenetics, specific neurons are silent only when exposed to photostimulation. However, our interest was to prevent neural activity during mating. Therefore, we had to ensure that the light was only turned on during mating. However, if the experimenter manually turned on photostimulation in response to the attraction of the animal , he had to watch the animal the whole time. Waiting for flies to come takes time.”

The tracking problem led the team to develop a deep learning algorithm that can detect interactions. By training the AI ​​to recognize when sex occurs, it can control photostimulation. This allowed them to discover that when piezo-expressing neurons are inhibited, men adopt a wonky posture, which is ineffective. As one would expect, men who showed difficulty in having a proper sexual relationship had fewer children. They concluded that a major part of the piezo gene was helping the man change his side in response to the woman to be successful.

“Piezo proteins have been implicated in various physiological processes, including touch, hearing, blood pressure control, and bladder function,” Kamikouchi said. “Now our findings suggest that reproduction should be added to the list. Since mating is an important reproductive behavior that is widely conserved in animals, understanding how it works can help to better understand reproduction in all animals.”

Kamikouchi is excited to use AI for such research. “With the recent development of informatics, testing systems and analysis methods have advanced significantly,” he concludes. “In this research, we have succeeded in creating a device that automatically detects mating using real-time machine learning and controls the necessary recording of optogenetics. In order to investigate the neural mechanisms that control animal behavior, it is necessary to conduct experiments in which neural activity is controlled only when a person shows a certain behavior. Methods established in this study can be used not only to study the mating behavior of fruit flies but also the different behaviors of other animals. It should greatly contribute to the promotion of neurobiological research.”

Disclaimer: AAAS and EurekAlert! is not responsible for the accuracy of the news posted to EurekAlert! by supporting organizations or using any information through the EurekAlert system.