Access this dataset on Dryad DOI:10.5061/dryad.w9ghx3g2c

Introduction

Experimental evolution has demonstrated that mesophilic microbes readily adapt to increases in temperature. However, many microbes are psychrotolerant and resistant to cold, which is associated with physiological specializations, suggesting constraints in thermal adaptation. We hypothesized that constraints would limit adaption differently in a mesophilic species (B. thuringiensis) compared to its psychrotolerant relative B. mycoides – with adaptation at cooler temperatures and adaptation at higher temperatures being constrained in each species respectively. To test this hypothesis, we imposed 140 generations of selection at temperatures at and below the optimum for productivity for both species. The fitness and thermal performance of evolved bacteria showed ancestral thermal niche plays a role in thermal adaptation over this timescale, in support of our hypothesis of adaptive constraints. Temperature-dependent trade-offs appeared common in B. mycoides, with fitness gains associated with decreases in operational niche width; fitness gains at one temperature caused a decrease in the range of temperatures that the bacterium showed appreciable growth. Genome resequencing showed that variation in mutation supply and selection strength could not explain temperature-dependent responses to selection. Importantly, metabolic theory only held true for mesophilic B. thuringiensis, showing abundant but less studied psychrotolerant species could follow different adaptive trajectories.

The dataset contains data and methods to reproduce the results of White et al., as detailed in their manuscript "Constraints in temperature adaptation reinforce differences in thermal niche between mesophilic and psychrotolerant Bacillus cereus group species".

We hypothesized that constraints would limit adaption differently in a mesophilic species (B. thuringiensis) compared to its psychrotolerant relative B. mycoides – with adaptation at cooler temperatures and adaptation at higher temperatures being constrained in each species respectively.

We conducted experimental evolution using a mesophilic Bacillus thuringiensis strain and a psychrotolerant Bacillus mycoides which both belong to the Bacillus cereus group. After ~140 generations of adaptation to temperatures close to and far below thermal optimum, we used a three-step analytical approach.

Competition assays were used to test for fitness changes via direct competition with the strain’s ancestor.

SNP variant calling was used to identify genetic changes that emerged during experimental evolution.

Description of the data and file structure

Cell_size_checks.zip; Contains the raw image files collected to determine the cell size of each strain/lineage.

Cell size of strains was determined by suspending cells in saline prior to staining (incubation at 4°C in dark) with BacLight bacterial viability and counting staining kit. Images were acquired using a confocal fluorescence microscope (Leica AF6000) set up as follows: objective ×63 plan apochromatic; SYTO9 excitation wavelength 488 nm, emission wavelength 493-547 nm; propidium iodide excitation wavelength 552 nm, emission wavelength 567-725 nm. Analysis of images used Leica software (LAS X version 3.5.1.18803) to measure cells over five independent fields of view per strain. Dimensions were recorded as pole-equator lengths in μm for a total of 30 cells/strain in focus and planarity. Cell volume was calculated assuming an ellipsoid form (cell vol = 4⁄3 π × ρ/2× ϵ ) where ρ is the polar long axis and ϵ is the equatorial short axis.

Image files are arranged in series, each containing fifteen files with the numbering system Series001-060. Identities for lineage to number is as follows;

• In folder brhwcellsize_Series 1.zip
  brhwcellsize_Series001.tif to brhwcellsize_Series005.tif; Bacillus mycoides ancestor
  brhwcellsize_Series006.tif to brhwcellsize_Series010.tif; Bacillus mycoides, lineage 1 in 30°C selection regime
  brhwcellsize_Series011.tif to brhwcellsize_Series015.tif; Bacillus mycoides, lineage 2 in 30°C selection regime
• In folder brhwcellsize_Series 2.zip
  brhwcellsize_Series016.tif to brhwcellsize_Series020.tif; Bacillus mycoides, lineage 3 in 30°C selection regime
  brhwcellsize_Series021.tif to brhwcellsize_Series025.tif; Bacillus mycoides, lineage 4 in 30°C selection regime
  brhwcellsize_Series026.tif to brhwcellsize_Series030.tif; Bacillus mycoides, lineage 5 in 30°C selection regime
• In folder brhwcellsize_Series 3.zip
  brhwcellsize_Series031.tif to brhwcellsize_Series035.tif; Bacillus thuringiensiss ancestor
  brhwcellsize_Series036.tif to brhwcellsize_Series040.tif; Bacillus thuringiensis, lineage 1 in 30°C selection regime
  brhwcellsize_Series041.tif to brhwcellsize_Series045.tif; Bacillus thuringiensis, lineage 2 in 30°C selection regime
• In folder brhwcellsize_Series 4.zip
  brhwcellsize_Series046.tif to brhwcellsize_Series050.tif; Bacillus thuringiensis, lineage 3 in 30°C selection regime
  brhwcellsize_Series051.tif to brhwcellsize_Series055.tif; Bacillus thuringiensis, lineage 4 in 30°C selection regime
  brhwcellsize_Series056.tif to brhwcellsize_Series060.tif; Bacillus thuringiensis, lineage 5 in 30°C selection regime

Alongside these image files, there are two files.

"cellsize.txt": this file contains the measurements of cell size obtained from the image files listed above. Variables are:

• Species: The species examined (either Bm for B. mycoides or Bt for B. thuringiensis).
• Lineage: Lineage of the strain/lineage tested (either "ancestor" for ancestral strain, or "30X", showing lineage was result of experimental evolution at 30°C (X being a number between 1-5 indicating a different replicate lineage).
• polar: polar long axis length in μm
• equator: equatorial short axis length in μm
• vol_cyl: Cell volume, calculated assuming an ellipsoid form (cell vol = 4⁄3 π × ρ/2× ϵ ) where ρ is the polar long axis and ϵ is the equatorial short axis.

This data was analysed using the R script "hugh_cell_size.R".

Raw_thermal_profiling_data.zip;

This file contains the raw OD data used to generate thermal performance curves in this study.

Raw data is arranged by the temperature each strain was grown at, each with its own folder (15°C, 17°C, 20°C, 23°C, 27°C, 29°C, 33°C). In each X folder;

26 files beginning "Hw start od . . ."; each containing the output of the plate reader at a given timepoint during the experiment. There are five essential sheets for these files;

• Absorbance results; the optical density reading (at wavelength of 600nm) for each well at the timepoint
• Raw data 600nm; the raw data formatted and presented in Absorbance results. Headers are; Well (cell row and number of 96-well plate), Wavelength(s) [nm] (wavelength using for reading, in nanometres), Raw absorbance (measure of growth in data, unitless data) and measurement time (seconds).
• General information; Information regarding the session information of the Multiskan Sky OD reader (including execution time, device name, serial number, firmware version, software version, set temperature and whether cuvette and incubator are on) and protocol parameters (including assay type, plate template, wavelength, measurement type, shake (to homogenise the samples) and shake duration.
• Plate layout; Layout of plate. None specified here, can be ignored.
• Run log; the optical density reading (at wavelength of 600nm) for each well at the timepoint. Can be ignored.

"Data formatting template X.xlsx"; Taking information from the sheet *Absorbance results in the above files, and entering it into cells A1-M9 (for replicate 1-2) and A11-M19. Colours are used to show layout (cells containing water in blue, cells containing blank controls in yellow, and strains containing samples in red). The table underneath details

• Strain: Strain examined. Either BtA (B. thuringiensis, ancestor), BmA (B. mycoides, ancestor), Bt30/Bt15 (B. thuringiensis selected at 30°C or 15°C) and Bm30/Bm15 (B. mycoides selected at 30°C or 15°C)
• Temp: Experimental temperature (°C)
• Rep: Biological replicate (1, 2, 3 or 4)
• ID; concatenated Strain and Rep information.
• Time: Experimental timepoint (in hours, changed manually)
• Raw absorbance: Absorbance at wavelength 600nm of blanks
• Average OD of blanks: Average absorbance at wavelength 600nm of blanks

Data from each "Hw start OD ..." file is copied and pasted into cells A1-M9 or A11-M19. Time can be used to discriminate plates, as the plate occupying A1-M9 was always read first at each timepoint.
The collated results are stored in the sheet "collated data" within this file, and in "X output.csv"

These "X output.csv" results are processed to produce the file "X well values" which contains parameters from the growth curves extracted using R package GrowthCurver

• IDnames: concatenated Strain and Rep information e.g. Bm15-1 means replicate 1 of B. mycoides selected at 15°C.
• Rcurve: maximum growth rate of sample at experimental temperature
• Kcurve: maximum carrying capacity of sample at experimental temperature
• auc1curve: area under the logistic curve obtained by taking the integral of the logistic equation
• aucEcurve: empirical area under the curve which is obtained by summing up the area under the experimental curve from the measurements in the input data
• temp: experimental temperature

We recommend using the "X well values STATIONARY.csv" for analysis, as these record only to stationary phase of bacterial growth and so avoid die-off which could skew results.

"All well R and auc1 STATIONARY PHASE REDUCED.csv" contains concatenated results of "X well values STATIONARY.csv".

Checking_activation_energy.zip

"Activation energy check rTPC 11 5 2024.csv"

Output from growth curve data to obtain the activation energy (e) from thermal performance curves for each strain/lineage.

• Strain: Species of bacteria (either B. thuringiensis or B. mycoides)
• Type: Identifies strain as the ancestral strain ("Anc") or a lineage derived from experimental evolution ("evo").
• Temp: The selection regime for experimental evolution the strain is from (either Anc for ancestor strains that did not undergo any experimental evolution, 15°C or 30°C).
• e: activation energy, defined as the minimum energy required for growth to occur.

Blanks indicate missing data.

The data was analysed using the script "Activation energy check script 11 5 2024".

Fig_1_and_supp_fig_1.zip

All well values output ALL STRAINS BOTH 24 08 23.csv

• IDnames: concatenated Strain and Rep information e.g. Bm15-1 means replicate 1 of B. mycoides selected at 15°C.
• Rcurve: maximum growth rate of sample at experimental temperature
• Kcurve: maximum carrying capacity of sample at experimental temperature
• auc1curve: area under the logistic curve obtained by taking the integral of the logistic equation
• aucEcurve: empirical area under the curve which is obtained by summing up the area under the experimental curve from the measurements in the input data
• Strain: Species of bacteria (either B. thuringiensis or B. mycoides) with lineage (either A for ancestral or XYY to reflect lineage number (X; 1-5) and experimental evolution treatment (YY; 30°C or 15°C).
• Selection temperature; The selection regime for experimental evolution the strain is from (either Anc for ancestor strains that did not undergo any experimental evolution, 15°C or 30°C).
• Rep: Replicate number
• Growth temperature: Growth temperature (°C) at which Rcurve/Kcurve/auc1curve/aucEcurve were calculated.

This file was fed into "plotting rTPCs 4 9 25 FIGURE 1.R" and "plotting Supplementary figure 1.R" to produce figure 1 and Supplementary figure 1 respectively.

Fig_2.zip

Scripts and data for plotting figure 2 from the article "Constraints in temperature adaptation reinforce differences in thermal niche between mesophilic and psychrotolerant Bacillus cereus group species". Analysis focuses on determining responses to selection at 15°C and 30°C in Bacillus thuringiensis and Bacillus mycoides by competition assays. Contains raw data and processed data for analysis. 

WT results for analysis LINEAGES 27 2 24.csv

• Genotype: Species (Bt or Bm).
• SelectionTemp: Selection regime (either "A" if ancestor, or 30°C or 15°C).
• Lineage: Number 1-5 indicating the experimental lineage. Note that only one lineage exists for ancestral strain, marked as 3.
• Speclineage: Genotype, SelectionTemp and Lineage concatenated into a single variable.
• Line: Binary variable defining sample as ancestral (A) or selected (S).
• Facet: When plotting, indicates where data should be presented (either in the Bt30 facet, or in the Bm30 facet).
• CompTemperature; Temperature in which competition experiment took place (30°C or 15°C)
• Replicate: Biological replicate for competition experiment.
• Selectionrateconstant: Selection rate constant of wildtype strain. Average across three replicates. Defined as the difference between the Malthusian parameters of the wildtype strain of interest and the antibiotic resistant standard competitor.
• Relativefitness: Relative fitness of wildtype strain. Average across three replicates. Defined as the ratio between the Malthusian parameters of the wildtype strain of interest and the antibiotic resistant standard competitor.

NA = missing data, where there was a general die-off in the media during competition.

WT results for analysis LINEAGES 27 2 24.csv is fed into "Analyses 27 2 24 STATISTICS.R" for analysis.

Analysis data for 15c with means 10 1 23.csv/Analysis data for 30c with means 1 23.csv

Files containing the individual fitness parameters for each wildtype strain replicate, and the means used in "WT results for analysis LINEAGES 27 2 24.csv"

• Strain: Species (Bt or Bm) and lineage information (A for ancestor, 1-5 for lineages from experimental evolution and 30 or 15 for experimental evolution temperature).

• Temperature: Competition temperature for these strains (15 or 30°C)

• Replicate: Biological replicate (1, 2 or 3)

• Selectionrateconstant: Selection rate constant of wildtype strain. Average across three replicates. Defined as the difference between the Malthusian parameters of the wildtype strain of interest and the antibiotic resistant standard competitor.

• Relativefitness: Relative fitness of wildtype strain. Average across three replicates. Defined as the ratio between the Malthusian parameters of the wildtype strain of interest and the antibiotic resistant standard competitor.

  Further columns detail the mean values

• Strain: Species (Bt or Bm) and lineage information (A for ancestor, 1-5 for lineages from experimental evolution and 30 or 15 for experimental evolution temperature).

• Mean Selection Rate constant: Average selection rate constant for Strain across replicates.

• Mean relative fitness: Average selection rate constant for Strain across replicates.

Fig_3.zip

Scripts and data for plotting figure 3 from the article "Constraints in temperature adaptation reinforce differences in thermal niche between mesophilic and psychrotolerant Bacillus cereus group species".

Productivity rTPC parameter correlations by lineage WOP 7 12 24

• Species: Species (Bt or Bm).
• Regime: Selection regime (either "Anc" if ancestor, or 30°C or 15°C).
• Topt: Temperature at which productivity is highest, derived from thermal performance curve of each lineage.
• Bpk: Maximum productivity derived from thermal performance curve of each lineage.
• Wop: Operational thermal niche width, over which productivity is at least half Bpk, derived from thermal performance curve of each lineage.

rTPC parameter correlations by lineage WOP 7 12 24

• Species: Species (Bt or Bm).
• Regime: Selection regime (either "Anc" if ancestor, or 30°C or 15°C).
• Topt: Temperature at which growth rate is highest, derived from thermal performance curve of each lineage.
• Bpk: Maximum growth rate derived from thermal performance curve of each lineage.
• Wop: Operational thermal niche width, over which growth rate is at least half Bpk, derived from thermal performance curve of each lineage.

Figure 3 in the manuscript is produced using "rTPC correlations script FIGURE 3 4 9 25.R".

Fig_4_and_supp_fig_2.zip

Scripts and data for plotting figure 4 and supplementary figure 2 from the article "Constraints in temperature adaptation reinforce differences in thermal niche between mesophilic and psychrotolerant Bacillus cereus group species".

Productivity rTPC parameter correlations by lineage WOP 7 12 24

• Species: Species (Bt or Bm).
• Regime: Selection regime (either "Anc" if ancestor, or 30°C or 15°C).
• Topt: Temperature at which productivity is highest, derived from thermal performance curve of each lineage.
• Bpk: Maximum productivity derived from thermal performance curve of each lineage.
• Wop: Operational thermal niche width, over which productivity is at least half Bpk, derived from thermal performance curve of each lineage.

rTPC parameter correlations by lineage WOP 7 12 24

• Species: Species (Bt or Bm).
• Regime: Selection regime (either "Anc" if ancestor, or 30°C or 15°C).
• Topt: Temperature at which growth rate is highest, derived from thermal performance curve of each lineage.
• Bpk: Maximum growth rate derived from thermal performance curve of each lineage.
• Wop: Operational thermal niche width, over which growth rate is at least half Bpk, derived from thermal performance curve of each lineage.

Figure 4 and supplementary figure 2 in the manuscript are produced using "rTPC correlations script GROWTH AND PRODUCTIVITY PLOTS" and data is analysed using "rTPC correlations ANALYSIS.R"

Fig_5.zip

Scripts and data for plotting figure 5 from the article "Constraints in temperature adaptation reinforce differences in thermal niche between mesophilic and psychrotolerant Bacillus cereus group species".

dn ds ratios 8 9 22.csv

• Strain: Species, either B. thuringiensis (Bt) or B. mycoides (Bm)
• Phenotype: Rifampicin resistance phenotype, either susceptible (wildtype, WT) or resistant.
• Lineage: Experimental lineage, numbered 1-5.
• Selection Temp: Temperature (°C) at which selection took place.
• dN: Number of non-synonymous single nucleotide polymorphisms (SNPs) differentiating the strain from its ancestor.
• dS: Number of synonymous single nucleotide polymorphisms (SNPs) differentiating the strain from its ancestor.
• dN/dS: Ratio of non-synonymous SNPS to synonymous SNPS in each strain, indicator of strength and direction of selection.

Data is analysed using "dn ds analysis PLOTS 7 18 24.R" and used to produce the plots.

Fig_6.zip

Scripts and data for plotting figure 6 from the article "Constraints in temperature adaptation reinforce differences in thermal niche between mesophilic and psychrotolerant Bacillus cereus group species".

Venn for r Bm WT(in).csv

• *Bm15: Product of genes showing convergent single-nucleotide polymorphisms within B. mycoides lineages selected at 15°C. Convergence is defined as the same gene possessing a non-synonymous SNP in at least 3 lineages within the given selection regime.
• *Bm30: Product of genes showing convergent single-nucleotide polymorphisms within B. mycoides lineages selected at 30°C. Convergence is defined as the same gene possessing a non-synonymous SNP in at least 3 lineages within the given selection regime.

Venn for r Bt only WT(in).csv

• *Bt15: Protein ID of genes showing convergent single-nucleotide polymorphisms within B. mycoides lineages selected at 15°C. Convergence is defined as the same gene possessing a non-synonymous SNP in at least 3 lineages within the given selection regime.
• *Bt30: Protein ID of genes showing convergent single-nucleotide polymorphisms within B. mycoides lineages selected at 30°C. Convergence is defined as the same gene possessing a non-synonymous SNP in at least 3 lineages within the given selection regime.

Sharing/Access information

Link to Bioproject containing sequences used to derive data for Figure 5 and Figure 6:

• https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA826440

R is required to run all scripts.
Annotations are provided throughout the script through 1) library loading, 2) dataset loading and cleaning, 3) analyses, and 4) figure creation.