On the natural and laboratory evolution of an antibiotic resistance gene

TEM-1 ß-lactamase is one of the most notorious antibiotic resistance enzymes around. It exists at high frequencies in antibiotic-resistant bacteria around the world and confers resistance to ß-lactam antibiotics, including penicillins (e.g. ampicillin) and cephalosporins. The enzyme displays a remarkable phenotypic plasticity in response to the introduction of new drugs; within a few years after the clinical debut of most new ß-lactam antibiotics resistance conferring variants of TEM-1 are isolated. Such a shift in resistance phenotype is typically caused by just a few amino acid substitutions. Until today, more than 150 variants of TEM-1 with a unique amino acid sequence have been identified. Because of the clear link between genotype and phenotype (i.e. level of resistance or fitness) and because of the ease of selecting for increased antibiotic resistance, TEM-1 has been used as a model in studies that seek new methods to optimize proteins. These studies combine the power of in vitro mutagenesis and in vivo selection and have resulted in a wealth of information about which mutations can increase resistance when the enzyme is exposed to an antibiotic that it initially hydrolyzes inefficiently. At a later stage, these techniques were adopted and used to repeat and predict the natural evolution of TEM-1 under various selective conditions. Recently, TEM-1 is increasingly being used as an experimental model for the study of fundamental evolutionary questions, particularly those that benefit from the direct relationship between genotype and phenotype. In this thesis, both the natural and laboratory evolution of TEM-1 are studied. The aim of the laboratory work is to increase our understanding of the way in which adaptive mutations interact. For this purpose, TEM-1 is mutagenized using error-prone PCR, which creates variation in the resulting copies of the TEM-1 gene. Mutated gene-copies are placed in bacteria which are subsequently selected for increased resistance to cefotaxime (an antibiotic that TEM-1 hydrolyzes poorly). By repeating this process multiple times in independent experiments, the mutations and mutational trajectories involved in the increase of cefotaxime resistance are studied. At a fundamental level, this has lead to a better understanding of the nature of mutation interaction and its consequences for evolutionary contingency and constraint. Evidence indicating that certain ‘silent’ mutations (i.e. mutations that alter the codon sequence but not the amino acid that the respective codon encodes) can also play a role in increased resistance was found in these data as well. A phylogenetic study of the sequences of the ~150 different TEM-alleles that have been isolated in hospitals and clinics so far indicates that recombination has played a significant role in the evolution of TEM-alleles, contrary to what is often assumed. Furthermore, amino acid substitutions present in these clinical isolates are compared to those found in laboratory evolution studies of TEM-1, in order to investigate to what extent laboratory evolution can be used as a predictive tool for the natural evolution of antibiotic resistance genes. This overview indicates that laboratory evolution very accurately repeats the natural evolution of TEM-1. Based on these findings, predictions are made about substitutions that may appear in future clinical TEM-isolates, and directions are given how laboratory evolution can be exploited as a predictive tool most efficiently.