Learning without a brain – how bacteria store memories and remember the past like artificial neural networks

E. coli bacteria possess the ability to learn from past experiences and use that knowledge to prepare for future situations, a process that does not require a brain or neurons. This discovery, published in the journal PRX Life, challenges the traditional notion that learning is confined to organisms with complex nervous systems. Researchers from the author's laboratory investigated how single-celled organisms like E. coli can retain memories and apply them to future decisions.

Bacteria inhabit environments that fluctuate constantly, facing challenges such as varying nutrient levels, temperature changes, and antibiotic threats. In order to survive, a bacterium must respond swiftly to present conditions while retaining valuable information about past experiences. Striking the right balance between adaptability and stability is crucial for bacterial survival.

To study whether bacteria can learn from their past, the researchers employed a microfluidic device to monitor the behavior of thousands of individual E. coli cells as they altered their nutrient supply at different rates. They discovered that bacteria not only react to current nutrient levels but also keep track of their nutrient history to cope with changing conditions.

While bacteria react to their immediate environment, the specific behavior observed in response to a sudden pulse of food depends on their prior experiences. When exposed to the same influx of nutrients, bacteria that recently experienced a feast-and-famine environment adapted faster than those from a stable environment. This suggests that the bacteria's prior experiences influence their present behavior, demonstrating a form of learning.

To understand how bacteria store this information, the researchers developed a mathematical model of the internal molecular network that controls bacterial growth. The model revealed that a component called ribosomes, responsible for building proteins and determining how fast a cell grows, likely serves as the site of memory storage in bacteria.

Ribosomes change their response to nutrient changes, with some responding quickly to present conditions while others retain traces of past experiences. This combination of fast and slow responders allows bacteria to maintain a memory spanning different timescales, from minutes to hours.

The molecular system in bacteria follows a similar computational logic to gated recurrent neural networks, a type of artificial intelligence used for processing sequences like speech and sensor data. Within the bacterial cell, molecular gates control how much existing memory is kept and overwritten when new information is introduced. These gates enable bacteria to retain memory at the cost of growth, as prioritizing adaptation comes at the expense of energy resources dedicated to growth.

This discovery highlights the evolutionary advantage of bacteria's unique strategy for learning and memory, balancing history-dependent computation without requiring neurons. The findings bridge the gap between biology and artificial intelligence, offering insights into how cells process information and make decisions while operating under energy constraints.

Moreover, the study has implications for medicine, as many pathogenic bacteria adapt to changing conditions to survive, providing a potential blueprint for developing targeted therapies.