Table 1.
A list of strains used in this study.
Fig 1.
Graphic view of the experimental evolution setup.
Three flasks labeled A, B, and C in red (striated caps) were inoculated with both yeast and bacteria whereas the other three flasks labeled D, E, and F in black (caps with no fill) were inoculated with yeast only (controls). The cultures were successively passaged for at most passages depending on the strain (see Table 2). After every 10 passages, cultures were withdrawn and karyotyped using pulse field gel electrophoresis (PFGE). Those that underwent genomic rearrangements were then phenotyped using a high-throughput micro-cultivation instrument, Bioscreen C (Oy Growth Curves, Finland).
Fig 2.
Experimental evolution scheme used to evolve yeasts.
A) We used a modified serial dilution transfer method in which yeast and bacteria were co-cultured to compete only during the exponential phase to allow selection at maximum growth rates when resources were unlimited [46] by decreasing with time. Numbers 1 and 2 show successive passaging of cultures, whereas the red line indicates yeast biomass. Each passage entailed a 48-hour incubation period consisting of initial 4 hours of yeast adaptation, followed by 40 hours of bacteria-yeast competition and then 4 hours of antibiotic treatment to “kill off” bacterial survivors before a 500-fold dilution into fresh media. B) Bacterial “hurdles” scheme. The scheme shows the order of sequential exposure of yeasts to six bacteria species. P. agglomerans was used to added to growing yeasts for 20 passages and then another 20 passages with S. plymuthica and e.t.c. Pseudomonas fluorescens was the only species that was used for longer than 20 passages. Longitudinal samples were stored for analyses every 10 passages. C) Effects of streptomycin on growth of yeasts. The plot shows control yeasts grown in YPD (filled diamond) and yeasts grown in YPD supplemented with streptomycin (filled square). D) Effects of dead bacteria on growth of yeasts. The plot shows control yeasts grown in YPD (filled diamond) and yeasts grown in YPD together with dead bacteria (filled triangle). The figure shows that both Streptomycin and dead bacteria (heat-killed by incubating for 5 hours at 60°C) did not affect yeasts growth.
Fig 3.
Long-term yeast-bacteria experimental evolution.
A total of 18 yeasts covering over 250 million years of evolutionary history (S. cerevisiae A1 and A2 (2 strains), S. eubayanus, C. glabrata, T. pretoriensis, L. thermotolerans, L. kluyveri, K. non-fermentans, K. wickerhamii, K. lactis A1 and A2 (2 strains), K. marxianus, E. gossypii, D. hansenii, D. bruxellensis, D. anomala, B. naardenensis, and B. custersianus) [36] evolved in this study are shown. Yeasts are phylogenetically ordered such that species at the bottom are the least related to S. cerevisiae [60]. Red bars represent yeasts that underwent genomic rearrangements whereas blue bars represent those that did not. *Poor, intermediate and good ethanol producers (i.e. ethanol yield of 0.1 ± 0.1g, 0.25 ± 0.05g and 0.37 ± 0.06g of ethanol per gram of glucose consumed in aerobic batch fermentation respectively as reported elsewhere) [61] are all represented in the Saccharomyces and non-Saccharomyces yeasts. Species with a lower number of passages had difficulties in growing in the presence bacteria as described in the main text.
Fig 4.
Evolutionary trajectories based on electrophoretic karyotypes.
The figure shows chromosomal bands separated by PFGE using a CHEF Mapper XA PFGE apparatus (Bio-Rad). To determine karyotypes, overnight cultures from each of the frozen samples stored after every 10 passages were used. Preparation of chromosomal plugs for PFGE was done by standard methods as reported [40]. Plugs were run using a multistate program for 110 h (Block 1, 25 h at 1.5 V/cm with a 2, 700 s pulse time and an angle of 53°; Block 2, 25 h at 1.5V/cm with a 2, 200s pulse time and an angle of 60°; Block 3, 30 h at 2 V/cm with a 1.500 s pulse time and an angle of 60°; and Block 4, 30 h at 2.5 V/cm with a 500-s pulse time and an angle of 60°) at a constant temperature (14°C) [52]. The gels were stained with ethidium bromide before photographing. Most ancestral strains used in this study are completely sequenced to be used as standards to estimate the sizes of novel bands based electrophoretic migration distance regression curve. In some cases, S. cerevisiae (S288c) and Schizosaccharomyces pombe (SJA148) were used as standards. Strains that underwent a rearrangemnt event are labelled in red. Bands that were not found on the respective ancestral lane are shown by a red arrow whereas those that were lost are shown by a cancelled red arrow. D. anomala (Y863), B. custercianus (Y893), D. bruxellensis (Y879), L. kluyveri (Y057) [16], C. glabrata (Y475), L. thermotolerans (Y688), K. nonfermentans (Y1057) and T. pretoriensis (Y1055).
Table 2.
Long-term evolution experiment.
The table lists culture collection identities and the number of passages they were evolved (in the presence of a bacterial selection pressure). All controls were passaged for 120 times except C. glabrata, which was passaged for 130 passages). Strains that underwent large-scale genomic rearrangements were derived from species in bold.
Fig 5.
Phenotyping of the evolved strains.
Strain phenotyping was performed using a high-throughput micro-cultivation instrument, Bioscreen C (Oy Growth Curves Ab Ltd, Helsinki, Finland). Turbidimetric readings were recorded every 20 minutes for 168 hours. Strains were grown at 25°C in rich medium, YPD (0.5% yeast extract, 1% peptone, 2% glucose, pH 6.2). The plots show average growth data from duplicates experiments. The experiments were carried out for 168 hours although we only show the lag and exponential phases of growth on these plots. A) D. anomala (Y863), B) B. custercianus (Y893), C) D. bruxellensis (Y879), D) K. nonfermentans (Y1057), and E) T. pretoriensis (Y1055). C. glabrata could not be tested using the same facilities as the species is an opportunistic pathogen. Phenotypes of L. thermotolerans (Y688) are not reported in this work, whereas L. kluyveri (Y057) has been comprehensively described elsewhere [16].
Fig 6.
Growth rates variation among the evolved strains.
Growth rates were calculated from the data in Fig 3. The growth rates were calculated from turbidimetric readings acquired at OD600nm during the exponential phase (S1 File). A linear regression of log of readings was done to calculate the growth rates using Microsoft Excel. The data shown is an average of duplicate assays. A) D. anomala (Y863), B) B. custercianus (Y893), C) D. bruxellensis (Y879), D) T. pretoriensis (Y1055) and E) K. nonfermentans (Y1057). C. glabrata could not be tested using the same facilities as the species is an opportunistic pathogen. Phenotypes of L. thermotolerans (Y688) are not reported in this work whereas L. kluyveri (Y057) has been comprehensively described elsewhere [16].