From Protein Structure to Function with Bioinformatics.pdf

6 Function Diversity Within Folds and Superfamilies 157families are grouped together into sub-groups that constitute a convenient intermediatelevel whose definition varies between superfamilies.Currently, the SFLD only covers six superfamilies. But the conservation of partsof the reaction chemistry within superfamilies appears very common, beingobserved in 22 out of the 31 enzyme superfamilies that were studied by Todd et al.(2001). In contrast, substrate specificity was not conserved in 20 of these superfamilies(see below).The occurrence of a common mechanistic step in mechanistically diverse superfamiliessuggests that enzymes in these superfamilies have maintained aspects oftheir catalytic mechanism in the course of their evolutionary diversification. Suchsituations hint at an evolutionary scenario in which enzymes evolve new functions,via duplication and recruitment, by maintaining partial reaction mechanisms (ratherthan partial substrate specificity, see below), thus resulting in the mechanisticallydiverse superfamilies observed nowadays (Gerlt and Babbitt 2001).Specificity Diverse SuperfamiliesAn alternative scenario for the divergent evolution of enzymatic functions withinsuperfamilies is one in which an ancestral enzyme with broad specificity duplicatesand the descendant copies specialise in binding more specific substrates. In such ascenario, substrate specificity is the dominant factor for function evolution in thesuperfamily. In their extensive analysis of enzyme superfamilies, Todd et al. (2001)showed that in most cases, reaction mechanisms were more conserved than substratespecificities between homologous enzymes. Out of 28 superfamilies thatwere involved in substrate binding, ten displayed no conservation of the substratewhatsoever, and another ten had very varied substrates with only a small commonchemical moiety such as a peptide bond (Todd et al. 2001).The expectation that substrate specificity might be conserved between homologousenzymes in a superfamily derives from Horowitz’s proposal on the backwardevolution of metabolic pathways (Horowitz 1945). This hypothesis suggests thatwhen the substrate of an enzyme becomes depleted, an organism possessing a newenzyme that is able to produce that substrate from a precursor compound which isavailable will have a selective advantage over others, and the new enzyme will befixed by evolution thus giving rise to an initial 2-step metabolic pathway. A similarevolutionary process can then take place for the other steps of the extant pathway.According to this scenario, pathway evolution goes backward as compared with thedirection of the metabolic flow (Rison and Thornton 2002). Because the originalenzyme has the ability to bind a substrate molecule that is the same as the product ofthe new enzyme, it has been suggested that this common property may be used as abasis for the evolution of the latter enzyme. Following this idea, all enzymes within ametabolic pathway would be homologous, and the enzyme catalysing the final stepof the pathway would be the most ancient. In addition, the evolution of these enzymeswould have been driven by their substrate selectivity, and this would result in a tendencyof extant superfamilies to display commonalities in substrate specificity.