Evolutionary innovation using EDGE, a system for localized elevated mutagenesis

Abstract

Mutations arising across the whole genome can hinder the emergence of evolutionary innovation required for adaptation because many mutations are deleterious. This trade-off is overcome by elevated mutagenesis to localized loci. Examples include phase variation and diversity-generating retroelements. However, these mechanisms are rare in nature; and all have narrow mutational spectra limiting evolutionary innovation. Here, we engineer a platform of Experimental Designed Genic Evolution (EDGE) to study the potential for evolutionary novelty at a single locus. Experimental evolution with EDGE shows that bacterial resistance to a novel antibiotic readily evolves, provided that elevated mutagenesis is focused on a relevant gene. A model is proposed to account for the cost and benefit of such single loci to adaptation in a changing environment and explains their high mutation rates, limited innovation, and the rarity of localized mutagenesis in nature. Overall, our results suggest that localized mutation systems can facilitate continuing adaptive evolution without necessarily restricting the spectrum of mutations. EDGE has utility in dissecting the complex process of adaptation with its localized, efficient evolution.

Fig 1. The EDGE genetic circuit to increase the mutation rate of a target gene, here a TetA-gfp fusion.

A Hin specific endonuclease cleaves both DNA strands at the Hin sites (yellow). Next, resection removes the 5' strands on both sides up to the chi sites (blue). Error-prone DNA repair generates mutations in the target gene and flanking regions.

Fig 2. EDGE induction increases mutagenesis, especially in the TetA-gfp target gene.

a. Growth on tetracycline plates selects for the gain of tetA efflux pump activity. EDGE induction increases the appearance of gain-of-function tetracycline resistant mutants on plates supplemented by 4 μg/mL tetracycline (squares, solid black line, p = 0.005). Growth on streptomycin also selects for streptomycin resistance, but EDGE induction only marginally increases the appearance of streptomycin resistance mutants (diamonds, dashed line, p = 0.074). The strep resistance gene is not the target locus. b. Growth on apramycin plates selects for the loss of tetA efflux pump activity [6]. EDGE induction increases the appearance of apramycin resistant loss of function mutants on plates supplemented by 4 μg/mL apramycin (circles, black line, p = 3.17 x 10⁻¹¹). Means and confidence intervals (CI) are determined by two-way ANOVA. CI are standard errors.

Fig 3. Growth rates of the ancestral genotype (WT) and two tetA antibiotic-resistant mutants.

The growth rate of the EDGE derived mutant (I235V) is larger than both the “wild type” and that previously obtained by in vitro methods (I235F). * indicates p< 0.05.

Fig 4. Homology model of tetA.

The arrow points to the mutated residue (isoleucine 235 with side chain shown) in both traditional directed evolution and EDGE to evolve resistance to tigecycline. This mutation lies at the exit of the pump on the periplasmic side. Colors distinguish the different alpha helices for clarity. The model was created using Phyre2 with intensive mode [10].

Fig 5. The fitness of localized mutators is strongly affected by the number of allelic states and the rate of mutagenesis to alleles.

The model predicts that optimal mutators have the minimum number of necessary allelic states and high rates of localized mutagenesis, as is seen in natural systems. Fitness is determined by the difference in malthusian parameters [31, 32].