Abstract
Organisms often face multiple selective pressures simultaneously (e.g., mine tailings with multiple heavy metal contaminants), yet we know little about when adaptation to one stressor provides cross-tolerance or cross-intolerance to other stressors. To explore the potential for cross-tolerance, we adapted Saccharomyces cerevisiae to high concentrations of six single metals in a short-term evolutionary rescue experiment. We then measured cross-tolerance of each metal-adapted line to the other five metals. We tested three predictors for the degree of cross-tolerance, based on similarity in 1) the physiochemical properties of each metal pair, 2) the overlap in genes known to impact tolerance to both metals, and 3) their co-occurrence in the environment. None of these predictors explained significant variation in cross-tolerance. Instead, the strongest predictor was the metal in which adaptation occurred: Cobalt-adapted lines performed well in most metals (generalists) while manganese-adapted lines typically performed poorly (specialists). To determine the genetic basis, we sequenced the genomes of 109 metal-adapted lines. Broader cross-tolerance characterized lines bearing mutations affecting phosphorus metabolism, with three genes related to phosphate metabolism bearing several independent mutations (PHO84, SIW14, VTC4). Thus, while a genome-wide analysis failed to predict cross-tolerance, a subset of genes facilitated growth in multiple metals. We also observed two "mutator" lines (both in manganese) and report evidence that cadmium, cobalt, and manganese altered the mutation spectrum. While it is challenging to predict how evolutionary adaptation to one stressor will impact tolerance to other stresses, our work helps reveal the environments and pathways that contribute to cross-tolerance among metals.