Evolution of copper tolerance
The artificial evolution experiment was performed to test if selection on a population’s standing variation of strains allows rapid evolution of copper tolerance (Fig. S1). A detailed description of the experimental approach and the outcome (changes in density, growth rate, pH, and Fv/Fm) is included in the Supplement (Fig. S3). Briefly, chronic toxic effects developed over time, and the copper concentration had to be modified to allow stable growth and maintain inhibition levels at around 50%. Strain RO5AC served as a reference point between the two experiments and was used to control for these treatment differences. During the experiment, the mining population’s growth rate increased to about 0.2 day-1 higher than RO5AC, while the reference population grew slower than RO5AC throughout the experiment (Fig. S3). The absolute growth rate increases in the mining population, from the start to the end of the selection phase, was 3-fold higher than the growth rate of the reference population (0.50 versus 0.15 day-1). In both populations, and the RO5AC mono-clonal control, the EC50 values increased from around eight to ten µM copper (Fig. 2), which was similar to the mining inlet strain VG1-2_81’s initial EC50, but significantly higher than any strain in the reference population (Fig. 1A). The response ranges to copper also expanded relative to the control (Table 1), resulting in adapted populations being tolerant to a wider range of copper concentrations beyond what was lethal to all strains during acute 72-hrs exposure.
In the reference population and in the RO5AC strain, a large part of the increase in copper tolerance (50-100%) was plastic and rapidly lost once the toxic stress was relaxed (Fig. 2A and B). The PAM assay showed that this tolerance was induced in RO5AC after only three days of exposure to copper, whereas it took up to 21 days before it was attained in the reference population (Fig. 3). In the reference population, four of the copper-exposed experimental replicates grew poorly in ambient media without excessive copper (Fig. S4), indicating that copper tolerance carries a cost in terms of limited ability to reach high cell densities in low copper environments. Collectively, these observations show that at least some strains of the reference population acclimated to toxic conditions via a plastic response to copper stress, and that this effect was not accurately captured by the acute 72-h dose-response experiment.