An evolvant ofEscherichia coli that employs thel-fucose pathway also for growth onl-galactose andd-arabinose

Summary l-Galactose,d-arabinose, andl-fucose form six-membered rings with identical stereoconfigurations. However, onlyl-fucose can serve as the sole carbon and energy source of wild-typeEscherichia coli K-12. A mutant that can grow onl-galactose andd-arabinose was isolated by alternate selection on the two sugars. Thel-fucose pathway became inducible by all three sugars. Transduction into the mutant of the wild-type fuc⁺ region containing both the regulatory and structural genes abolished the novel growth abilities onl-galactose andd-arabinose, whereas transduction into the mutant of a fuc deletion abolished the growth abilities on all three sugars. Introduction of the wild-type fucR⁺ (which encodes the activator protein for the fuc regulon) on a multicopy plasmid depressed the growth abilities of the mutant onl-galactose andd-arabinose, but not onl-fucose. The results suggest that the effector specificity of the activator protein in the mutant was broadened. It is proposed that an adaptive response of an activator-controlled system is more likely than that of a repressor-controlled system to achieve fixation in a population, because the first variant to emerge in response to a novel metabolic demand has a good chance of having an altered specificity of regulation. Such a change entails little or no metabolic liability during the absence of the novel substrate. In contrast, the first variant of a negatively controlled system to emerge has an overwhelming chance of being the result of a random mutation that destroys repressor function. Although negatively controlled systems can be more opportunistic in exploiting new conditions than positively controlled systems, an adaptive change is less likely to become fixed because of the cost associated with gratuitous constitutive gene expression in the absence of the substrate.