Relevant Degree Programs
Graduate Student Supervision
Doctoral Student Supervision (Jan 2008 - April 2022)
New protein functions often evolve through the recruitment and optimization of latentpromiscuous activities. How do mutations alter the molecular architecture to change function? Theoverarching goal of my thesis is to provide answers to this question, utilizing a novel xenobioticorganophosphate hydrolase (OPH) activity as model. Directed evolution performed on an N-acylhomoserine (AHL) lactonase enzyme possessing promiscuous OPH activity demonstrated that thenew function can be quickly optimized via a handful of mutations that rearranged active siteresidues to adapt to the new substrate. Ancestral sequence reconstruction (ASR) conducted on arecently evolved OPH enzyme, methyl-parathion hydrolase (MPH), revealed that the OPH activityemerged from an ancestral lactonase enzyme via five mutations that enlarged the active site toincrease complementarity to the new substrate. Subsequent generation of the adaptive fitnesslandscapes formed by these five mutations uncovered a prevalence of epistatic interactions thatconstrained the number of accessible evolutionary trajectories. Furthermore, the topologies of thelandscapes drastically change in response to subtle differences in substrate substituents. Finally,characterization of several extant lactonase orthologs of MPH revealed that sequence divergencehas resulted in lower levels of promiscuous OPH activities in the orthologs compared to theancestral enzyme that gave rise to MPH. Moreover, the five mutations fail to substantially increaseOPH activity in the genetic backgrounds of the orthologs. Comparative directed evolutionconducted on the MPH ancestor and the orthologs towards OPH activity show that the ancestralenzyme is able to improve the new function more rapidly. Overall, the results of this thesiscontribute to our understanding of enzyme evolution, and will help to better protein engineeringand design in the future.
Metallo-β-lactamases (MBLs) are powerful enzymes capable of conferring pathogenic bacteria with effective resistance against all major classes of β-lactam antibiotics. Their continuing global dissemination, paired with a lack of therapeutic inhibitors, has combined to pose a significant threat to human health. This thesis aims to use an evolutionary perspective to better understand the structure, function, and behaviour of the MBLs. The comprehensive characterization of eight MBLs in three different host organisms, Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa, revealed that host specific constraints can limit the effective periplasmic expression of the enzymes, and as a result, might restrict the dissemination of MBLs to certain bacteria. The directed evolution of NDM-1, VIM-2, and IMP-1 for the provision of greater ampicillin resistance in Escherichia coli exposed the mechanisms by which MBLs may adapt to overcome these expression barriers, while revealing the critical role that the signal peptide plays in host adaptation. The subsequent directed evolution of the same three MBLs with two other β-lactam antibiotics, cefotaxime and meropenem, demonstrated the relative robustness of the family’s broad substrate specificity, as only two of seven complete trajectories featured a narrowing of specificity and changing the selection pressure on one of these trajectories swiftly restored broad specificity. The long-term genetic drift of VIM-2 under purifying selection at different thresholds revealed the plasticity of the MBL’s sequence and structure, but also the robustness of its activity and function. Overall, the results presented in this thesis contribute to our understanding of the MBL family and will help to develop better treatment strategies in the future.
Enzyme superfamilies have expanded over billions of years from the descendants of a potentially single common ancestral function. Understanding the evolution of their functional diversity is central to biochemistry, molecular and evolutionary biology. The overarching question of my thesis is how enzyme promiscuity, the serendipitous ability to catalyze non-native reactions and reactions, connects enzyme functions and facilitates molecular evolution by providing evolutionary starting points towards new functions. In particular, I primarily focus on proteins across the metallo-β-lactamase (MBL) superfamily by comparing evolutionary and functional connectivity based on the functional profiling of 24 enzymes against 10 distinct hydrolytic MBL reactions. This analysis revealed that MBL enzymes are generally promiscuous, as each enzyme catalyzes on average 1.5 reactions in addition to its native one, which leads to high functional connectivity. Furthermore, the ability to promiscuously bind different metal ions, enzymatic co- factors of MBL enzymes, provide additional mechanisms whereby the function profile of some MBL enzymes can be broadened, and thus further extends the connectivity between functions. In addition, I expand and compare the analyses of function connectivity through promiscuity to three previously published superfamily-wide function profiling studies, which revealed common trends that are discussed in the context of enzyme superfamily evolution. Finally, I assess the evolvability of promiscuous enzymes to determine their potential as evolutionary starting points towards a novel function by performing a comparative laboratory evolution experiment of two related β-lactamases, NDM1 and VIM2, towards a shared promiscuous phosphonate monoester hydrolase activity. Both trajectories accumulate 13 mutations over ten rounds of directed evolution, however the mutational solutions and evolvability is strikingly different for the two enzymes. NDM1 improves catalytic efficiency by over 20,000-fold and loses much of its solubility, i.e. the amount of functional enzyme in the cell. Contrarily, VIM2 improves catalytic efficiency only by 60-fold, but improves solubility. Detailed structural analysis, combined with molecular dynamics simulations, reveals a molecular understanding for the observed differences in evolvability between NDM1 and VIM2. Overall, my research contributes to our understanding of enzyme evolution and will help to advance functional annotation and engineering of enzyme.
Master's Student Supervision (2010 - 2021)
Metallo-β-lactamases (MBLs) are powerful enzymes conferring antibiotic resistance to various pathogens. They are actively disseminated in a plethora of pathogenic organisms via horizontal gene transfer, and raise clinical concerns. Interestingly, a predominant number of pathogens that favor MBL expression belong to γ-proteobacteria such as Pseudomonas aeruginosa, implying biases in the host distribution. A previous study in the Tokuriki lab also revealed restricted MBL gene expression in phylogenetically distanced bacterial species. However, it is still obscure how the enzymes adapt to bacterial hosts. This thesis aims to understand mutations in the signal peptides of MBLs contributing to the adaptive phenotype development in Escherichia coli. A series of dose-response curve assays revealed a highly positive correlation between the sequenced individual variants and their respective population where a considerably narrow phenotypic diversity is observed. When the phenotypic diversity is in a limited range at a high ampicillin concentration, the variability in mutational effects associated with phenotypic variations may be confined by fitness costs, resulting in the development of similar phenotypes. This suggests that the sequence profiles of the variants are a reasonable representation of the mutational population. A bioinformatic analysis of the sequence profiles of NDM-1, VIM-2, and IMP-1 variants reveals that mutation-driven changes in translational steps are moderately correlated with minimum inhibitory concentration (MIC), implying that the mutations in the signal peptides are associated with adaptive phenotypes elevating MBL gene expression. In addition, I synthesized a new broad host range vector using three individual plasmid vectors by the Golden Gate assembly technique to analyze mutational effects of MBLs in Acinetobacter baumannii, Escherichia coli, Pseudomonas aeruginosa, and Pseudomonas putida. The newly synthesized plasmid vector is compatible with the listed bacterial species enabling future research. Overall, these results in the thesis will help understand how the mutations in the signal peptide of MBLs, in part, promote adaptive phenotypes in E. coli to elevate MBL gene expression, and may help develop future therapeutic strategies.