A mechanistic model of bacteriophage T7 infection, replication, and evolution

点击翻译按钮获取中文摘要

Summary Genetically modified viruses have important applications in the prevention and treatment of disease, such as viruses used as live vaccines and for phage therapy. In these applications, we need to modify viruses to have specific phenotypes (e.g., attenuated fitness or the ability to kill multidrug-resistant pathogens) while also pre- venting rapid adaptation that may compromise these functions. In practice, viruses for these applications are often created haphazardly, via trial-and-error. A critical barrier to further progress in this field is the ability to engineer viral genomes rationally, while being able to predict the phenotypic consequences of the engineering as well as the likelihood of further adaptation or evolutionary reversion of the engineered viruses. This project will develop a detailed mechanistic model of a viral study system that can be used to study genome engineering, targeted viral attenuation, and evolutionary recovery. The virus is a dsDNA bacteriophage (T7) that is safe and can be easily manipulated and engineered. With its extensive background of genetic, biochemical and evolutionary studies, T7 offers the best empirical and theoretical foundation of all viruses for addressing this problem. Our approach consists of three Aims that collectively combine computational modeling of the viral life cycle with genome engineering, molecular studies of viral infections, fitness measurements, and evolution of modified genomes. In Aim 1, we will assess the principles of gene regulation in T7. We hypothesize that complex, dynamic expres- sion patterns do not require explicit gene regulatory networks, and that instead gene regulation in T7 is the result of a finely tuned balance between transcript synthesis and degradation. We will test this hypothesis both in three- gene model systems and in simulations of the entire T7 life cycle, validated against high-throughput measure- ments of T7 transcript and protein abundances. In Aim 2, we will extend our simulator into a predictive fitness model. We hypothesize that bacteriophage fitness can be predicted from the rate of production and cellular abundance of bacteriophage genomes, transcripts, and proteins. We will extend the simulator with modules for genome replication, capsid assembly, and lysis. All sim- ulations will be calibrated using experimental measurements of phage fitness for a panel of different engineered and evolved T7 genomes. Aim 3 will apply the insights generated from Aims 1 and 2 to larger-scale genome disruptions and phage evolu- tion. We hypothesize that the T7 genome architecture imposes quantifiable constraints on the ways in which the phage can evolve and/or respond to genetic manipulation. We will engineer T7 variants with inserted transgenes, rearranged gene order, or more fragmented gene expression modules, and we will assess to what extent we can predict the phenotypic and evolutionary consequences of these modifications in silico.