Experimental evolution can be used to increase survival of yeast under freeze-thaw stress.

A Experimental design of selection for freeze-thaw tolerance. B Survival fraction of cells after freeze-thaw over the course of evolution. Each point on the thick purple line is the average of four biological replicates and two technical replicates for each biological replicate. Error bars indicate standard deviation. Each light coloured line represents a biological replicate. The evolution was performed thrice and the data from one evolution experiment is shown here. Inset: Average number of viable cells after every freeze- thaw-growth cycle. Purple - evolving population. Pink - control population. C Basal intracellular concentrations of trehalose relative to the control measured over the course of evolution. Each point on the thick purple line is the average of 4 biological replicates and 12 technical replicates for each biological replicate. Error bars indicate standard deviation. Each light coloured line represents a biological replicate. This measurement was made during one evolution experiment.

Freeze-thaw tolerant cells undergo reduced membrane damage and survive membrane damage.

A Flow cytometry measurements on control cells. Each point represents a single cell and the plot contains data for 10000 cells. The x-axis is the fluorescence intensity in the 5-CFDA channel. The y-axis is the fluorescence intensity in the propidium iodide channel. The contour lines represent changes in density of cells, starting from 17 with increments of 17. The cluster on top comprises membrane damaged cells and the cluster at the bottom comprises cells with intact membranes. B Flow cytometry measurements on freeze-thaw tolerant cells. Each point represents a single cell and the plot contains data for 10000 cells. The x-axis is the fluorescence intensity in the 5-CFDA channel. The y-axis is the fluorescence intensity in the propidium iodide channel. The contour lines represent changes in density of cells, starting from 17 with increments of 17. The cluster on top comprises membrane damaged cells and the cluster at the bottom comprises cells with intact membranes. C The 2-dimensional data from A and B are used to compute a 1-dimensional distribution of “membrane damage” in cells. Area under the normalized curves: Evolved (Membrane intact) = 0.6347, Evolved (Membrane damaged) = 0.3127, Control (Membrane intact) = 0.2331, Control (Membrane damaged) = 0.7597. D Viability of cells measured after sorting the two clusters identified from the data in A and B. A 0.3% sorting error was measured by sorting beads. The entire experiment described in this figure was performed four times and the data from one experiment is shown here.

Cells of the population adapted to survive freeze and thaw are smaller, denser and exhibit reduced budding.

A Brightfield image of cells. Scale bar - 10µm. B Projected area of cells measured from images as shown in A. On each box, the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually as points scattered in the x-direction. Sample size (number of cells measured): Control = 1107, Evolved = 4281, Welch’s t-test: P = 1.3 × 10−213. This experiment was performed twice and the data from one experiment for one evolved population is shown here. C Single cell mass density measured using a percoll gradient column. On each box, the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers (there are no outliers). Welch’s t-test: P = 3.2 × 10−7. This experiment was performed twice and the data from both experiments for eight technical replicates of one biological replicate was pooled. D Fraction of cells budding in stationary phase measured from images as shown in A. On each box, the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers (there are no outliers). Sample size (number of cells measured): Control = 309, Evolved = 422, Welch’s t-test: P = 2.59 × 10−6. This experiment was performed twice and the data from one experiment for one evolved population is shown here. E Distribution of normalized fluorescence intensities of DNA in cells stained with SYTOX Green measured using flow cytometry. Area under the normalized curves: Evolved (Single DNA intensity) = 0.8726, Evolved (Double DNA intensity) = 0.1274, Control (Single DNA intensity) = 0.8330, Control (Double DNA intensity) = 0.1670. Sample size (number of cells measured): Control = 199682, Evolved = 399113. This experiment was performed twice and the pooled data for four biological replicates from one experiment is shown here. F Design of the microfluidic channel through which cells are flowed to measure deformation with an image of cells leaving the microfluidic channel. Scale bar - 15µm. G Eccentricity and area of cells measured in the narrow channel. Flow rate: Control - 5.5cm/s, Evolved - 5.6cm/s. Sample size (number of cells measured): Control = 5825, Evolved = 6192. This experiment was performed four times for one evolved population and the data from one experiment is shown here. H Fraction of intact cells over time measured by turbidity of cells placed in water after Zymolyase treatment. This experiment was performed five times for one evolved population and three technical replicates and the pooled data for all five experiments is shown here. I Image of Genetically Encoded Multimers (GEMs) in the cytoplasm of cells. Scale bar - 5µm. J Mean squared displacement (MSD) of GEMs tracked in the cytoplasm of cells, plotted on logarithmic axes. Inset: MSD plotted on linear axes. More than 10,000 trajectories were used to calculate the average. K Distribution of the exponent of time, α. More than 10,000 trajectories were used to plot the histogram.

Selection based on trehalose linked entry into a quiescence-like state recapitulates the adaptation dynamics.

A Optical density of the cell cultures inoculated into fresh media after a freeze-thaw. B Average doubling time of cells in the population calculated from the growth curves in A. C Average lag phase duration of cells in the population calculated from the growth curves in A. D Survival fraction of cells after freeze-thaw simulated over the course of evolution. Each light coloured line represents an iteration of the simulation. The thick purple line indicates the average for 25 iterations. Error bars indicate standard deviation. E Basal intracellular concentrations of trehalose relative to the control simulated over the course of evolution. Each light coloured line represents the average of the population in an iteration of the simulation. The thick purple line indicates the average for 25 iterations. Error bars indicate standard deviation of 25 iterations.

The mechano-chemical adaptation to freeze-thaw tolerance is convergent.

A Experimental design to measure survival after refreezing and for the evolution of refreezing tolerance. B Number of viable cells after consecutive freeze-thaws. Error bars indicate standard deviation. This experiment was performed five times with two technical replicates and the pooled data from all experiments is shown here. The point marked in blue denotes the selection pressure for (kF T = 1) evolution. The point marked in green denotes the selection pressure for (kF T = 3) evolution. Basal amounts of intracellular trehalose measured in the stationary phase. On each box, the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers (there are no outliers). P (kF T = 1) = 7.1 ×10−8, P (kF T = 3) = 1.6 ×10−7. DFraction of cells budding measured in the stationary phase. On each box, the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually as points scattered in the x-direction. This experiment was performed once and the data for one evolved population and six technical replicates is shown here. Sample size: Control (kF T = 1) = 309, Evolved (kF T = 1) = 422, P (kF T = 1) = 2.5 × 10−6, Control (kF T = 3) = 357, Evolved (kF T = 3) = 459, P (kF T = 3) = 8.1 × 10−5. E Mean squared displacement (MSD) of GEMs tracked in the cytoplasm of cells, plotted on logarithmic axes. Inset: MSD plotted on linear axes. More than 10,000 trajectories were used to calculate the average. F Distribution of the exponent of time, α. More than 10,000 trajectories were used to plot the histogram.

A listing of the genetic mutations across all replicates including both the kFT = 1 and kFT = 3 selection regimes, coloured boxes indicate the identified mutations.

Replicates of populations of cells adapted to freeze-thaw stress do not have any specific genetic mutation in common. FS - Frame Shift, SC - Stop Codon, NS - Non-Synonymous, S - Synonymous, IG - Inter-genic region.

Survival fraction of cells measured over different freeze durations Error bars represent standard deviation of biological and technical replicates.

Design of the deformability cytometry device

Parameters for spin coating.

Sketch of the experimental design to probe the digestion of the cell wall.

BME - beta-mercaptoethanol.

Map of cloned GEMs Plasmid.

Parameter values for simulations.

Average survival fraction of the one freeze-thaw evolved populations propagated without stress over several generations.

On each box, the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered. outliers (there are no outliers)

Sequenced samples.

Sketch of the analysis of sequencing reads.

A Representation of samples collected for sequencing. Sample collection points are displayed on a plot of the the number of viable cells in the control and evolving population across freeze-thaw-growth cycles. B Mapping of reads to a reference genome followed by variant calling. C A few representative variants filtered and clustered based on their frequency across all samples. Error bars denote 95% confidence intervals. D Trajectories of frequencies of representative variants from parent strain till the end point of selection. Each line represents a sample.

Replicates of populations of cells adapted to freeze-thaw stress do not have any duplicated genes in common

List of duplicated genes across all replicates including both the kFT = 1 and kFT = 3 selection regimes. Coloured cells denote a duplication.