Chance and memory

Memory and chance are two properties or, more evocatively, two strategies that cells have evolved. Cellular memory is beneficial for broadly the same reasons that memory is beneficial to humans. How well could we function if we forgot the contents of the refrigerator every time the door was closed, or forgot where the children were, or who our spouse was? The inability to form new memories, called anterograde amnesia in humans, is tremendously debilitating, as seen in the famous psychiatric patient H.M. [1] and the popular movie “Memento” (Christopher Nolan, 2000). Cells have limited views of their surroundings. They are able to sense various chemical and physical properties of their immediate environment, but distal information can come only through stochastic processes, such as diffusion of chemicals. Cells lack the ability to directly sense distal parts of their environment, as humans can through vision and hearing. Thus, we might expect that cells have adopted memory of the past as a way of accumulating information about their environment and so gain an advantage in their behaviour. In a simple application, memory can be used to track how environmental signals change over time. For example, in the chemotaxis network of E. coli, cells use memory of previous measurements of chemical concentrations to estimate the time derivative of the chemical concentration and so to infer whether the organism is swimming up or down a chemical gradient [2]. Conversely, memory can be useful in enabling persistent behaviour despite changing environments. For example, differentiated cells in a multicellular organism must remain differentiated and not redifferentiate despite existing under a variety of conditions. If such cells become damaged or infected, they may undergo apoptosis, but we do not observe neurons becoming liver cells, or liver cells becoming muscle cells, and so on. Different mechanisms for achieving cellular memory operate at different time scales. For example, the chemotaxis machinery of E. coli can change response in seconds. Gene regulatory systems controlling sugar catabolism, such as the lac operon in E. coli [3] or the galactose network in Saccharomyces cerevisiae [4], take minutes or hours to switch state. Terminally differentiated cells can retain their state for years – essentially permanently, although scientists are learning how to reset their memories and return cells to a pluripotent status [5]. Of course, that cells