Testing the adaptive value of sporulation in budding yeast using experimental evolution

Saccharomyces yeast can grow through mitotic vegetative cell division while they convert resources in their environment into biomass. When cells encounter specific low nutrient environments, sporulation may be initiated and meiotic division produces 4 haploid cells contained inside a protective ascus. The protected spore state does not acquire new resources but is partially protected from desiccation, heat, and caustic chemicals. Because cells cannot both be protected and acquire resources simultaneously, committing to sporulation represents a trade-off between current and future reproduction. Recent work has suggested that one of the major environmental factors that select for the formation of spores is passaging through insect guts, as this also represents a major way that yeasts are vectored to new food sources. We subjected replicate populations of a panel of 5 yeast strains to repeated, predictable passaging through insects by feeding them to fruit flies (Drosopila melanogaster) and then allowing surviving yeast cell growth in defined media for a fixed period of time. We also evolved control populations using the same predictable growth environments but without being exposed to flies. We assayed populations for their sporulation rate, as measured by the percentage of cells that had sporulated after resource depletion. We found that the strains varied in their ancestral sporulation rate such that domesticated strains had lower sporulation. During evolution, all strains evolved increased sporulation in response to passaging through flies, but domesticated strains evolved to lower final levels of sporulation. We also found that exposure to flies led to an evolved change in the timing of the sporulation response relative to controls, with a more rapid shift to sporulation, and that wild-derived strains showed a more extreme response. We conclude that strains that have lost the ability to access genetic variation for total sporulation rate and the ability to respond to cues in the environment that favor sporulation due to genetic canalization during domestication.

. 55 In nature, populations of the budding yeast Saccharomyces cerevisiae are thought to 56 primarily grow as a vegetative mitotic diploid cells and to disperse to novel habitats 57 through the guts of insect vectors as meiotic haploid quiescent spores encapsulated within 58 a protective structure called the ascus (Stefanini et al., 2012; Gibbs and Stanton, 2001; survive ingestion and establish new colonies in the next habitat when the insect defecates. 78 If ingestion by insects happens only rarely, or only after the local resources are depleted, 79 then we expect genotypes that maximally convert resources into spores will evolve. In  On the other hand, S. cerevisiae dispersal ability depends on the ability to survive insect 86 guts (as spores or vegetative cells) as well as mate recognition and germination once a 87 new suitable growth habitat is reached (Murphy and Zeyl, 2012). 88 We set out to test if fitness-trade offs between growth and dispersal traits in S.

103
Yeast and fruit-fly strains 104 We used a set of five genetically distinct strains of S. cerevisiae that were provided as using a 10% bleach solution for 40 minutes at 22°C. Fly eggs were collected by sterile 126 pipette, washed with sterile water, and transferred using sterile technique to clean media.

127
Clean flies were reared and propagated on this media so that other yeasts and fungi 128 were minimized, but also so that the ingestion of antifungal elements did not reduce the 129 viability of living yeasts traveling through the gut.

130
Selection protocol 131 We grew each of our five diploid yeast strains in 2 mL of YPD (Yeast-Peptone-Dextrose) 132 liquid culture over a 5 day period. Samples of these initial strains were then frozen in 133 15% glycerol solution at -80°C and labelled as the "Ancestral" treatment. The five 134 strains were split into 4 replicate each, and each was then split again into a paired 135 control and treatment population ( Figure 1).

136
At the start of each selection cycle, each population was incubated in YPD at 30 137°C for 5 days without shaking. In order to reduce chances of contamination from other 138 yeasts or bacteria, cultures were always grown in YPD media with G418, tetracycline,   Figure 1: Each lineage was grown in liquid media and the sample was adjusted to an optical density of 0.3 and split into control and treatment tubes. From that point, the control and treatment went through parallel procedures lasting 7 days: 2 days of exposure to treatment (control: 22°C bench-top incubation; treatment: 22°C exposure to flies) and 5 days of growth or sporulation in liquid media at 30°C. This was repeated for 30 cycles.
Yeast cultures were distributed into CaFe feeding apparatus and offered to 3-4 clean 162 flies (see above). Flies were allowed to feed for 48 hours and then removed from the 163 vials. Measurements of total fly food consumption were taken by recording the change in 164 5 meniscus of the two capillary tubes. Flies were removed from the vials which were then 165 rinsed with 1.6 mL YPDA media and the supernatant (1.5 mL total volume because 166 some volume is reabsorbed into the agar in the vial) was collected and used as the 167 inoculate for the next round of yeast population growth. 168 For the control selection treatment yeast cultures were not distributed into the CaFe 169 feeding apparatus, being instead placed on the bench-top nearby. After 48h, each control   The main model considers the sporulation probability to be a function of the 201 treatment type (fly treatment or control) and the assay time (2.5 days or 5 days) giving where C i is the count of sporulated cells in sample i, N i is the total number of cells 203 counted in sample i. The predictor β depends on the T treatment type and the A assay 204 6 time of observation i. 205 We additionally modeled the variation that comes from replicate populations sharing 206 the same ancestry as meaning that each replicate population is considered to be sampled from a normal 208 distribution and the prior on the standard deviation is a Student's t distribution with 3 209 degrees of freedom, a location of 0 and scale of 10.

210
Here, because we assume that the parameter for each combination of strain, treatment,

295
Our results mostly support the latest view about the adaptive value of sporulation. 296 We found that strict passaging through the Drosophila digestive tract resulted in the produces from an existing demes; but such theory remains to be fully developed.

346
In the context of a competition-colonization trade-off (Tilman, 1994), a yeast strain 347 that sporulates earlier or at a higher frequency before ingestion by insects is more likely 348 to survive the process of ingestion, digestion and transfer. However, if the period of 349 competition within a deme is long, the higher sporulating genotype will reproduce more 350 slowly and eventually be displaced by genotypes that have lower sporulation rates. In