Fig 1.
(A) We evolved MALs for a total of 120 (YPD) and 60 (HU and RM) single-cell bottlenecks. The ten genetic backgrounds included four diploid homozygous strains (WA—West African, WE—Wine/European, NA—North American, SA—Sake) and all six possible hybrids derived from them. For each strain, we initiated 4 replicated MALs in each condition named as “strain-condition–replicate”, with e.g., “NA/NA-YPD-1” denoting the North American homozygote, evolved in YPD, replicate 1. Green, blue and orange circles represent the MALs in YPD, HU and RM respectively. The NA/NA genetic background MALs overview is shown as example. (B) Heatmap show the cell doubling time (hours), cell yield (log2 cell yield normalized to that of the spatial control strain, NA/WA) and cell viability. Cell viability in YPD is represented as CFUYPD/counted cells, while cell viability in HU and RM is calculated as CFUHU or RM/CFUYPD). “X” indicates MALs with all four replicates going extinct. “-” indicates strains excluded because of extremely low fitness in that condition and could not be propagated. (C) Evolution of cell doubling time and yield. Boxplots show the cell doubling time and yield in drug and drug-free conditions across the single-cell bottleneck (x-axis). Center line: median; box: interquartile range (IQR); whiskers: 1.5×IQR; dots: outliers beyond 1.5×IQR. Green, blue and orange boxplots represent the MALs phenotyped in their evolved condition (YPD, HU and RM respectively) while the gray boxplots represent MALs evolved in drugs but phenotyped in drug-free condition. The p value is measured by Mann–Whitney U test by comparing phenotypes at the initial and last timepoints of evolution. For each condition, the number of MALs included in the analysis (N) is NYPD = 40, NHU = 24, NRM = 32. (D) Qualitative summary of the direction in which each fitness proxy changes during the mutation accumulation.
Fig 2.
Chromosome XII amplification compensates for rDNA contraction in RM.
(A) Aneuploidies identified in the MALs. Blue tones indicate chromosome copy number. The chromosomes are sorted by their size with larger chromosomes on the top. (B) Upper left, the short read sequencing revealed segmental CNV on chromosome XII of NA/WE-RM-4. The red arrows indicate the breakpoint of each segmental CNV. Upper right dot plot, the x-axis represents NA and WE reference genomes, y-axis represents the assembled contigs of NA/WE-RM-4 obtained from long read sequencing. Lower left, zoom-in dot plot of the chromosome XII inversion. The bar below represents chromosome XII, with black circle and red rectangle indicating centromere and rDNA repeats respectively. Lower right, CNV-associated inversion structure: the extra copy of chromosome XII right arm together with the centromere was inverted and replaced the left arm. (C) rDNA copy number per chromosome XII (upper panel) and rDNA copy number per sample (lower panel) normalized by the ancestors for all MALs binned based on the evolved condition and the copy number of chromosome XII. Boxplot: center lines = median; boxes = interquartile range (IQR); whiskers = 1.5×IQR; points = outliers beyond 1.5×IQR. (D) Cell doubling time (upper panel) and yield (lower panel) of RM-evolved MALs phenotyped with (left) or without (right) RM. The blue and red boxplots represent MALs with and without chromosome XII gain respectively. The p value was calculated to compare the cell doubling time and yield between the initial and last timepoint populations by Mann–Whitney U test.
Fig 3.
Segmental CNV and structural variation.
(A) Segmental CNV on chromosome III, XII and XIV of the SA/WE-HU-3 line detected by short read sequencing. (B) Left panel, CNV on chromosome XI and XIV of SA/WE-HU-2 line detected by short read sequencing. Right panel, the gained segment on chromosome XI right end was inserted to chromosome XIV left end. In panels (A) and (B), the grey bar below the coordinate represents the chromosome, on which the black circle and red rectangle indicate the location of centromere and rDNA respectively. The red arrows indicate the segmental CNV break points. (C) Right panel, de novo phased assembly of SA/WE-HU-2 line based on the long-read sequences with color to indicate the sequence similarity between the two parental subgenomes. Left panel, the zoomed-in dot plots highlight one single assembled contig with SA chromosome XI segment inserted in chromosome XIV. (D) Number of structural variants identified by assembly-based alignment in three long-read sequenced MALs. Circle sizes are proportional to SVs observed and sectors indicate the types of structural variants.
Fig 4.
Substitutions, INDEL rates and signatures.
(A) Substitutions rate in YPD, HU and RM. (B) Substitutions rate of YPD-evolved MALs partitioned by the four parental genetic backgrounds, e.g., boxplot at “WE” represents substitutions rates of a total of 16 MALs of WE/WE, WA/WE, NA/WE and SA/WE in YPD. Mann–Whitney U test is used for significance analysis. *p = 0.05, **p = 0.01, ***p = 0.007. (C) INDEL rates in YPD, HU and RM. (D) Ratio of GC-to-AT to AT-to-GC substitutions. Relative contribution and cosine similarity of COSMIC (E) Single Base Substitution (SBS) and (F) small insertions and deletions (ID) signatures. (G) Mutational profiles based on the relative incidences of base-substitution changes within a trinucleotide context. N indicates the number of substitutions in each condition.
Fig 5.
(A) Mitochondrial DNA copy number estimated by the sequencing depth of ATP6, COX2, COX3 genes and normalized by the sequencing depth of the nuclear genome. Violin plot: center lines = median. The dashed line shows the mitochondrial copy number before evolution. (B-C) Mitochondrial DNA copy number of hybrid NA/WE (B) and SA/WE (C) before evolution (upper panels), and for one MAL after HU evolution (lower panels). The sequencing depth was estimated for 100 bp non-overlapping windows (dots) sliding along the mitochondrial genome (x-axis) and normalized to the sequencing depth of the nuclear genome. The dashed green line indicates the mean copy number across windows based on presence and absence. (D) Substitutions rate in mtDNA is higher in HU 4.86×10−4 substitutions/line/generation compared to 3.29×10−5 in YPD and 4.73×10−5 in RM.
Fig 6.
The loss of heterozygosity landscapes.
(A) LOH fraction and genomic positions in YPD, HU and RM conditions. The heatmap shows the proportion of MALs that have LOH events using 10 kb windows across the genome. Black dots indicate the centromeres. (B) LOH rate (left panel) and the proportion of total LOH length in the genome (right panel). Violin plot: center lines = median. (C) Impact of interstitial, terminal and chromosome loss LOH type in LOH rate (upper panel), proportion of genome in LOH (middle panel) and LOH length distribution (bottom panel). Boxplot: center lines = median; boxes = interquartile range (IQR); whiskers = 1.5×IQR; points = outliers beyond 1.5×IQR.