Abstract
Unlike most modern enzymes, which perform a single reaction, primordial enzymes are believed to have each performed a broad range of reactions. Consequently, most modern enzymes are poor comparative models for inferring the properties of primordial enzymes. The aim of this thesis was to identify and characterise modern enzymes that have evolved to catalyse multiple reactions on multiple substrates, in order to better understand the properties of primordial enzymes.
Using bioinformatics and phylogenetics, I have discovered that, in three distinct clades, one enzyme (MetC) has taken over the role of the absent alanine racemase (Alr). Two of the three MetC enzymes have also taken over the role of the absent glutamate racemase (MurI). Even though MetC, MurI and Alr are not homologous. In other organisms, such as Escherichia coli, MetC catalyses the β-elimination of cystathionine in methionine biosynthesis. The E. coli MetC has a small promiscuous alanine racemisation activity, but no detectable glutamate racemisation activity.
Two of the clades —the genus Pelagibacter and the family Anaplasmataceae, which includes Wolbachia— were in the same class, the Alphaproteobacteria, but the precise location was under debate in the literature. Therefore, I used phylogenetic methods to determine that they did belong to sister orders and that the AT-richness of their genomes is ancestral, rather than a source of bias.
The three enzymes investigated are located in different groups along the MetC tree. Using in vitro activity assays, I found that the enzyme most similar to E. coli MetC, Pelagibacter ubique MetC, could racemise alanine, but could not racemise glutamate. Further away on the tree, Wolbachia MetC has glutamate and alanine racemisation activities that are both stronger than its cystathionine β-elimination activity in terms of turnover numbers (kcat). While, in a basal group, Thermotoga maritima MetC possess the strongest glutamate racemising activity of the enzymes studied. These three enzymes were also able to promiscuously cleave cysteine.
The three enzymes had Michaelis constants (KM) for the various substrates that are comparable to those of each dedicated enzyme (MetC, Alr and MurI) in other organisms, while having turnover numbers (kcat) that are much lower than those of each dedicated enzyme. It was conjectured that this balance of kinetic parameters in these enzymes is due to the physiological necessity of operating with low concentrations of each substrate, yet producing a limited amount of product due to the low cellular demand for each metabolite.
The crystal structure of T. maritima MetC was solved, showing that it may possess glutamate racemisation activity because an active site tryptophan forms a hydrogen bond with the terminal carboxyl group of the bound glutamate. The structure also revealed a latch-like loop close to the active site entrance.
I used directed evolution in an attempt to improve the cystathionine elimination activity of T. maritima MetC. A mutant (S86T/S305C) was identified that possessed a more permissive active site with decreased Michaelis constants for all activities, including those that had not been under selection. This may indicate that wild-type T. maritima MetC has a tight balance of kinetic parameters that needs to be relaxed before it can be evolved into an enzyme with a single activity.
Overall, the properties of these enzymes reveal how multiple activities are balanced in a single active site and offer new insights into the likely nature of primordial enzymes.