In Memoriam Francis H.C. Crick

Quickly I narrowed down my interests to two main areas: the borderline between the living and the nonliving, and the workings of the brain. Further introspection showed me that what these two subjects had in common was that they touched on problems which, in many circles, seemed beyond the power of science to explain. He was fortunate to get a studentship in Max Perutz's group in the Cavendish Laboratory at Cambridge Univer-He once said that the ultimate goal of a theory was not sity, where he learned about X-ray crystallography. The to explain the brain directly but indirectly by asking the discovery of the structure of DNA by James Watson and right question and designing a key experiment that Francis Crick and, later, the structures of proteins by would " give the game away. " An early interest in the Perutz's group set the stage for the extraordinary ad-molecular basis of memory led him to ask: Do spines vances in molecular biology that have occurred in the twitch? (Crick, 1982). His prediction that synaptic spines last 50 years. He once mentioned to me that it had might be motile and that actin might be found in them never occurred to him during the years he worked on was recently confirmed (Matus, 2000). He also explored the genetic code that the human genome would be se-the issue of what was responsible for the stability of quenced in his lifetime. memories despite the rapid turnover of protein in neu-Crick's interest turned to neuroscience after moving rons. He suggested that proteins might be found at from Cambridge to the Salk Institute for Biological Stud-synapses with stable binary states (Crick, 1984b). ies in 1976, where he began to think about the workings Autophosphorylation of an important molecule in the of the brain. His approach to the brain focused on per-postsynaptic spine, Ca 2ϩ /calmodulin-dependent pro-ception, particularly visual perception (Crick, 1979). His tein kinase, was found to have the predicted properties preferred way to attack a problem was to read the litera-(Miller et al., 1988). In a study on the function of dream ture, which he knew better than most experts, and to sleep (Crick and Mitchison, 1983), he used an associa-discuss the details directly with the authors. In preparing tive network model to explore the possibility of " un-a paper on the distribution of long-range horizontal con-learning " during sleep. Evidence that memory consoli-nections …