How many neurons are required for learning and memory? None, according to Saigusa et al., who report basic learning behavior in unicellular amoebae in a recent article in Physical Review Letters.

The amoeba Physarum polycephalum is sensitive to environmental conditions. At room temperature, Physarum move at a constant rate. However, dry air slows the rate of Physarum movement.

The authors puffed dry air on Physarum once an hour for three hours. On the fourth hour, Physarum slowed down, even when no puff of air was delivered. Subsequent hours without air puffs slowly extinguished the periodic slowing of Physarum movement. However, one dry air puff six hours later reactivated the hourly behavior pattern.

These behaviors are consistent with rudimentary learning in higher organisms. Do these data indicate that unicellular organisms can learn? Physarum, like other organisms, have precise biological rhythms set by cellular oscillators. So, Physarum may be particularly sensitive to events occuring at regular intervals, and their periodic slow-down may represent the setting of a biological rhythm. However, rhythms alone do not explain extinction of the behavior in the absence of additional dry air puffs.

Do these data indicate a potential origin for learning, or do they indicate that our definition of learning in complex organisms is too simplistic? I’m a bit torn.