Rapid evolution of diminished transformability in Acinetobacter baylyi

Abstract

The reason for genetic exchange remains a crucial question in evolutionary biology. Acinetobacter baylyi strain ADP1 is a highly competent and recombinogenic bacterium. We compared the parallel evolution of wild-type and engineered noncompetent lineages of A. baylyi in the laboratory. If transformability were to result in an evolutionary benefit, it was expected that competent lineages would adapt more rapidly than noncompetent lineages. Instead, regardless of competency, lineages adapted to the same extent under several laboratory conditions. Furthermore, competent lineages repeatedly evolved a much lower level of transformability. The loss of competency may be due to a selective advantage or the irreversible transfer of loss-of-function alleles of genes required for transformation within the competent population.

FIG. 1.

Free DNA is detrimental to the growth of A. baylyi. Competent and noncompetent lineages were grown in minimal medium in the presence of genomic DNA from A. baylyi ΔilvC::Spec^r-sacB. Dashed lines indicate the range of concentrations of genomic DNA found in the medium after 16 h of growth of ancestral, competent A. baylyi. Curve fits have r² values of 0.85 and 0.67 for Δcom and wild-type lineages, respectively; shown are means and standard deviations for five lineages.

FIG. 2.

Competent and noncompetent Acinetobacter baylyi strain ADP1 lineages adapt equivalently to laboratory conditions. (A) Growth rates of competent and noncompetent lineages are improved to comparable extents. Shown are means and standard deviations of three measurements each for five lineages of each genotype. (B) The fitness of each evolved lineage was determined by competition with the ancestor of the opposite type at high and low densities. The high-density fitness of the evolved competent lineages had increased by ∼0.06 h⁻¹ greater than the increase of the noncompetent lineages. Shown are the means and standard deviations for five lineages, with two replicates each.

FIG. 3.

A. baylyi evolves a diminished transformability in response to adaptation to laboratory conditions. (A) Transformability in the evolved and ancestral competent lineages was assayed. Transformation was tested with two markers in order to ensure that the transforming marker was not having an effect. Regardless of marker, evolved lineages have transformability that is ∼15 to 20% of that of the ancestral lineages. Shown are means and standard deviations for five lineages. (B) Transformability of lineages over the course of the evolution experiment. Transformability was normalized to the initial fraction transformed within each lineage; means and standard deviations are shown for five lineages. (C) Six clones from each of five lineages, evolved and ancestral, were assayed for transformability. Rank-ordered clones are shown, with similar fill patterns of bars indicating clones from the same lineage.

FIG. 4.

A. baylyi evolves a diminished transformability regardless of mutation frequency. (A) Over the course of adaptation to LB, wild-type and ΔmutS lineages adapted to similar extents. (B) After ∼730 generations of parallel adaptation, wild-type and ΔmutS lineages achieved similar reductions in transformability. Shown are means and standard deviations for five lineages.

FIG. 5.

Adaptation to harsh conditions results in the evolution of diminished transformability. (A) Competent and noncompetent lineages were adapted to LB plus 300 mM NaCl at 40°C for ∼400 generations. Growth rates changed at a greater rate in the Δcom lineages than in the wild-type lineages. (B) Growth rates of competent and noncompetent lineages were measured in benign conditions over the course of adaptation to harsh conditions. Adaptation to harsh conditions elicited neither correlated benefits nor trade-offs with regard to growth rates when measured in benign conditions. (C) Similar to the case for experiments in benign conditions, adaptation to harsh conditions resulted in a diminished transformability. Shown are means and standard deviations for five lineages.

FIG. 6.

Noncompetent lineages of A. baylyi overtake competent lineages more rapidly than predicted in direct competition. The rate at which the noncompetent lineages overtake the population is higher than predicted by two equations: (i) using equation 2, which was used to determine selection coefficients in the initial competitions, assuming that s = 0.012 generation⁻¹ (see Materials and Methods) (29), and (ii) using equation 4 (see Materials and Methods) (13). Shown are the means and standard deviations for five competitions, as well as predicted values based on the two equations discussed.