The rate of translation elongation inEscherichia coliis limited by diffusive transport of matching aminoacyl-tRNAs (aa-tRNAs) to ribosomes. Our previous work revealed that, as cell growth quickens, stoichiometric crowding speeds this diffusive search by optimizing encounters between cognate translation molecules, inclusive of chemical kinetics taken fromin vitroexperiments. However, we predicted absolute elongation rates three-fold slower thanin vivomeasurements. We hypothesized that 'pre-loading' of EF-Tu:GTP:aa-tRNA ternary complexes onto ribosomal L7/L12 subunits -- suggested experimentally but not included in our initial model -- might further speed elongation and close this gap. Here, we develop a first-principles physico-chemical model of theE. colicytoplasm including explicit EF-Tu:L7/L12 interactions and elongation reaction kinetics, which quantitatively predictsin vivobinding and rheology. Our model reveals that transient co-localization of the translation machinery by EF-Tu:L7/L12 interactions shortens wait times at the ribosomal A-site, doubling elongation speed and improving prediction of the absolute elongation rate. We posit pre-loading efficiency as a competition between durable binding and frequent sampling of new ternary complexes, and show that the naturally-observedE. colicopy number of four L7/L12 subunits optimizes this tradeoff. Paired with literature data supporting a correlation between lower L7/L12 copy number and faster bacterial growth rate, we suggest a colloidal-scale evolutionary and functional advantage of having fewer L7/L12 per ribosome: frequent ternary complex sampling in dense, fast-growing cytoplasm.

No abstract