rmf , a conserved ribosome hibernation marker of convergent dormancy in experimentally evolved Escherichia coli

ABSTRACT Bacteria adapt to remarkably diverse chemical stresses, yet how far independent adaptive trajectories converge, and on what, remains poorly defined. We reanalyzed transcriptomes from 188 Escherichia coli lineages evolved under 47 different chemical stresses, 4 M9 controls, and 4 ancestral replicates. We used multiple public data sets to test whether divergent lineages converge on a shared physiological endpoint. Lineages evolved under chemically diverse stresses converged on a shared transcriptional axis, with their expression change positively correlated across chemicals ( r = 0.157; P = 0.001). This axis corresponds to a slow-growth, RpoS-governed stationary-phase and ribosome hibernation program, aligning strongly with an RpoS/stationary-phase index ( r = 0.97) and inversely with a ribosomal index ( r = −0.65). The ribosome modulation factor rmf , which forms translationally silent 100S ribosome dimers, rose along the PC1 axis ( r = 0.65) and co-expressed specifically with the 100S hibernation factors hpf , yfi A, and sra . In contrast, its associations with persister toxin-antitoxin, multidrug efflux, and oxidative stress genes were weak or inconsistent. Cross-species analysis supported purifying selection on ribosome modulation factor, and a genome-wide analysis placed rmf within the lowest-diversity 4% of core genes, consistent with convergence through regulation rather than coding change. Neither the axis nor rmf expression tracked the resistance level of the lineages, identifying the program as a conserved state of adaptation rather than a resistance mechanism. These findings recast a canonical stress response program as the convergent endpoint of adaptive evolution and identify rmf as a conserved readout of that state. IMPORTANCE Bacteria survive antibiotics not only by becoming resistant but also by entering a dormant, slow-growing state that tolerates drugs, and how predictably different populations reach such a state has been unclear. By reanalyzing Escherichia coli populations experimentally evolved under dozens of chemically distinct stresses, we show that each population acquired largely stress-specific mutations, yet their gene expression profiles converged on a shared dormancy program centered on ribosome hibernation. A single gene, rmf , whose product switches ribosomes off, reliably marked how deeply a population had entered this state, was reproducible across many independent data sets, and was almost invariant across hundreds of E. coli genomes. Because this convergence arose through gene regulation rather than mutation and did not correspond to resistance, rmf offers a simple readout of a dormant state associated with tolerance, a distinction useful for understanding and countering how bacteria survive antibiotics.