Artificial laboratory evolution procedure

Cells were evolved in either 384 or 1,536 colony arrays arranged on top of solid medium. Deletion mutant strains, and wild-type cell populations to be used as a baseline for calling genetic effects on adaptation rates, were stored in 384 colony arrays as −80°C glycerol stocks. Stored cell populations were thawed and recovered on YPD + G418 media, leaving every fourth colony position empty. After 3 days of cultivation at 30°C on the recovery plate, recovered cell populations were subsampled and replicated 3 times onto different SC medium preculture plates (Supplementary Fig. 1). In parallel, wild-type cell populations to be used as nonevolving spatial controls on experimental plates were recovered on YPD, as 384 colony arrays, and transferred to a separate preculture plate containing SC media. The evolution of gene deletion strains and interleaved evolving wild-type controls was initiated and continued by pinning first the precultures, and then each successive evolutionary batch culture, onto stressor-containing evolution plates. After each batch cycle of evolution, evolving gene deletion strain and wild-type cell populations were transferred to stressor-containing preculture plates and preculture for growth phenotyping experiments. At the precultivation stage, we introduced nonevolving wild-type spatial controls into every fourth, previously empty, colony position and later used these to account for environmental variations within and between plates. Precultures for growth phenotyping were transferred to stressor-containing experimental plates after 72 h at 30°C. These experimental plates were used for the growth phenotyping experiments described below. We generated cycle 0 estimates of preadaptation growth by pinning cell populations from recovery plates onto SC medium precultivation plates without stressor. The wild-type spatial controls were transferred using 384 pin-pads, while all other cell transfers used 1,536 pin-pads.