Stochastic Spatial Simulation of Genetic Information Processes in the Minimal Cell

Abstract

JCVI-syn3A is a genetically minimal cell whose genome consists of 493 genes, of which 452 are protein-coding. The genetic information processing reactions in JCVI-syn3A were recently solved stochastically using a well-stirred assumption which predicts the times of DNA replication initiation and elongation to be 8 and 50 minutes on average respectively after averaging over 1,000 replicate cells. This model approximately doubles the number of proteins according to genome-wide proteomics data as well as the number of ribosomal components over the course of a cell cycle. We introduce spatial heterogeneity and diffusion by constructing a 3D spatial model of JCVI-syn3A using cryo-electron tomograms to infer placement of large cellular components such as membrane boundaries and ribosomes. The ensemble of DNA configurations for the circular chromosome were generated using a random polymer where self-intersections and intersections with ribosomes were forbidden. Membrane-bound degradosome complexes were randomly distributed across the interior of the cell membrane. We fixed positions of the DNA, degradosomes, and ribosomes while simulating diffusion of RNA and proteins with more than 3,000 reactions including transcription of all 452 protein-coding genes and translation and degradation of all 452 mRNA under a reaction-diffusion master equation formulation using our GPU-based software, Lattice Microbes. Introducing spatial heterogeneity and requiring mRNA to diffuse to ribosomes and degradosomes to be translated or decayed results in a distribution of mRNA half-lives in agreement with measurements in other gram-positive bacteria Mycoplasma gallisepticum and Bacillus subtilis.