Scientists from the University of Westleyk in China and the California Technological Institute have recently unveiled a groundbreaking protein system that has the ability to process multiple signals within living cells and make decisions based on them. This new innovation, known as “perceptin”, integrates the principles of artificial neural networks and protein engineering, resulting in a biological equivalent of basic artificial intelligence. The study detailing this advancement can be found here.
Perceptin functions as a network of synthetic proteins capable of categorizing signals and triggering corresponding cellular reactions, such as activating survival mechanisms or programmed cell death. By leveraging the unique properties of proteins, perceptin enables rapid and precise responses, surpassing the limitations associated with systems based on DNA and RNA.
During the research, scientists developed pairs of proteins that are interconnected in a specific manner to form a network where certain components activate themselves while others inhibit. This design ensures that the response prioritizes the strongest signal while disregarding weaker ones. To assess the system’s efficacy, a cell line with fluorescent markers was engineered to visually monitor the system’s activity. Notably, the activity of essential proteins resulted in reduced fluorescence levels, confirming the accuracy of perceptin.
The practical application of this technology was demonstrated by linking the system to the apoptosis mechanism, enabling cells to undergo programmed cell death in response to defined conditions. Perceptin presents the opportunity to create customizable cell therapies that can respond to specific diseases, potentially eradicating tumor formations or modulating immune system function with precision.
This breakthrough technology paves the way for the advancement of programmable therapies, where cells can react to disease signals with tailored responses such as selective cell death or other reactions. Furthermore, this work highlights the prospect of developing intricate computational systems from interacting proteins, representing a significant stride towards biological artificial intelligence.