The Multiplicity of Domains in Proteins

The domainal nature of proteins is well established. What is less certain is how many domains are evolutionarily mobile in that they occur in otherwise nonhomologous proteins or in different sequential locations in homologous proteins. The combinatorial advantage of shuffling domains around into diverse settings is obvious. Those domains that have been shuffled about in recent evolutionary times, within the last half billion years or so, can usually be identified on the basis of sequence resemblances alone. Contrarily, domains that were rearranged in ancient times may only be apparent after three-dimensional analysis, their sequence resemblances having been eroded over time. The shuffling of domains in recently evolved proteins has been greatly promoted by introns, but this does not imply that all domainal rearrangements involve introns. Only a small fraction of known exons show evidence of having been shuffled. Taken in aggregate, the available data best fit a scenario whereby a relative small number of genes encoding domain-sized polypeptides has been expanded by duplication and modification with a burst of exceptional genomic rearrangement.