Scientists have engineered a protein-based neural network within living cells, promising breakthroughs in decision-making biology and AI-inspired therapies.
Yes, scientists are recreating the human brain.
At a glance:
- Researchers developed a protein-based system, “perceptein,” capable of processing multiple signals and making biological decisions.
- The system mimics artificial neural networks to classify inputs and trigger cellular responses like apoptosis.
- Unlike slower DNA-based systems, perceptein uses proteins for faster and more direct signal processing.
- Potential applications include programmable therapies and biology-based artificial intelligence.
In a groundbreaking collaboration, scientists from Westlake University in China and the California Institute of Technology have designed a protein-based artificial neural network inside living cells. Dubbed “perceptein,” this innovation combines the principles of protein engineering and neural network theory to mimic the binary classification capabilities of perceptrons, a fundamental concept in AI.
The perceptein network processes complex signals at the protein level, enabling cells to make decisions such as initiating apoptosis (programmed cell death) or responding to stress. Unlike traditional DNA- or RNA-based systems, perceptein utilizes synthetic proteins for faster and more efficient biological computations.
Revolutionizing Cellular Decision-Making
Cells naturally integrate various signals to perform critical functions. For example, immune cells detect threats, and the p53 signaling pathway determines whether to repair damaged DNA or self-destruct to prevent cancer. However, replicating such intricate decision-making processes artificially has posed significant challenges.
The perceptein system overcomes these hurdles by employing engineered proteins that bind in specific ways to form networks. These proteins activate or inhibit one another, ensuring that only the strongest signals trigger a response while weaker ones are ignored. This protein-level approach offers a more streamlined alternative to slower, transcription-dependent methods.
How Perceptein Works
The researchers created a perceptein circuit using six synthetic protein components and two input proteins. By integrating well-known proteases, including split tobacco etch virus protease, they constructed a system capable of distinguishing inputs with remarkable precision.
To test the circuit, the team engineered human embryonic kidney cells that expressed fluorescent proteins as visual markers. When the perceptein circuit was activated, specific proteases cleaved degradation signals attached to the fluorescent proteins, reducing their brightness. This setup allowed scientists to monitor the circuit’s activity in real-time.
By adjusting the levels of perceptein components, the researchers fine-tuned the network’s decision-making boundaries, demonstrating robust performance even under noisy or varying input conditions.
The study, published in Science, demonstrates the feasibility of artificial neural network-inspired circuits in mammalian cells, marking a significant step forward in synthetic biology and AI-driven therapeutic design.
This is honestly terrifying…don’t you think?
