Revealing the mechanisms underlying drug interactions

Drugs such as antibiotics are often used in combination. When two drugs are combined they may interact synergistically or antagonistically. Synergistic drug pairs have the advantage of enhancing the drugs ef- fects at fixed concentration, making them attractive for medical applications. In contrast, antagonistic drug pairs weaken the drugs effects but have potential for slowing down the evolution of drug resistance. Thus, smartly designed drug combinations are relevant for optimally treating disease and controlling the emer- gence of resistance. In addition, combining drugs is a powerful means for revealing complex relationships in cell physiology. However, while cellular targets of individual drugs are often known and their effects on global cell physiology have been characterized, our understanding of the effects of drug combinations is limited. Here we ask: what are the underlying causes of drug interactions? We propose a combined exper- imental and theoretical approach to reveal, model, and manipulate the genetic and cellular mechanisms of drug interactions that occur between antibiotics. Specifically: (1) We will use an established robotic system to perform high-throughput measurements of the individ- ual and joint effects of eight representative antibiotics on the growth rate of all strains from genome- wide Escherichia coli gene and sRNA deletion libraries. (2) We will systematically analyze these data using techniques from bioinformatics and statistical phys- ics to pinpoint genes that affect drug interactions and identify empirical laws that can capture the growth rate of mutants dependent on drug combinations. Further, we will identify plausible scenari- os and produce theoretical models of bacterial growth, gene regulation, and specific cellular func- tions that quantitatively describe the underlying mechanisms of selected drug interactions. (3) We will elucidate the cellular mechanisms of drug interactions, distinguish between plausible sce- narios, and test theoretical predictions by mimicking specific drug effects genetically and by meas- uring changes in the regulation of key genes, cell composition, and cell physiology using fluores- cent reporters, microscopy, and biochemical assays. This basic research project will involve close interaction between experimentalists and theorists. Its suc- cessful completion will reveal the mechanisms underlying antibiotic synergism and antagonism. It will fur- ther expose a fundamental principle of the phenotypic landscape of the cell, allowing us to predict the growth rate of most mutants in drug combinations from their growth rates in the individual drugs. By resolv- ing mutants that deviate from this prediction, we will identify the genetic factors and cellular functions that control drug interactions. This will provide a set of potential targets for new drugs which, in the long term, could be used to design treatments in which drug interactions are modulated to achieve effective clearance alongside the prevention of emergence of antibiotic resistance. Overall, this work will open the door for a new approach to the rational design of drug combinations.

Antibiotics are important in modern medicine. They are still the only effective treatment option for most infectious diseases. A problem is that the number of resistant pathogenic bacteria is rising. At the same time, the development of new antibiotics has almost stalled. A potential way to circumvent the looming antibiotic resistance crisis is to combine multiple antibiotics: Smartly designed combinations of drugs can improve treatment efficacy and even slow the evolution of drug resistance. However, finding such combinations is hard: Brute-force screening approaches are not feasible since the number of possible combinations of drugs is too large. An improved understanding of drug combination effects has potential to remedy this situation. In this project, we developed a systematic approach to understand what causes the observed effects of drug combinations. We used collections of mutant strains of the common model bacterium Escherichia coli to find genes that change the effects of drug combinations - even if they do not alter the sensitivity to the individual drugs. Combining this new method with mathematical models of bacterial cell physiology and targeted experiments, we were able to elucidate the causes of various antibiotic interactions and make first steps toward a new computational tool that can predict the effects of drug combinations. In the long term, our results can contribute to new strategies for the design of more effective drug cocktails.