Adaptation at the Extremes of Life: Experimental Evolution with the Extremophile Archaeon Sulfolobus acidocaldarius

Early life on Earth may have originated in extreme environments, such as hydrothermal vents, which are characterized by extremely high temperatures and acidity¹. Microbes continue to inhabit extreme environments, including hot springs and volcanic solfatara. Characterizing the evolutionary dynamics that occur under these extreme conditions may shed light on the specialized physiological processes that enable survival under these conditions. This may have wide-ranging implications, from our understanding of the origins of biological diversity to the development of novel high-temperature enzymes with biotechnological applications.

The understanding of microbial evolutionary dynamics in extreme environments remains limited despite its critical importance. In contrast, a significant body of knowledge about evolution in mesophilic environments has been acquired through the application of a technique known as experimental evolution. Experimental evolution involves observing evolutionary change under laboratory conditions^2,3,4,5. Often, this involves a defined change environment (e.g., temperature, salinity, introduction of a toxin or a competitor organism)^7,8,9. When combined with whole-genome sequencing, experimental evolution has enabled us to test key aspects of evolutionary processes, including parallelism, repeatability, and the genomic basis for adaptation. However, to date, the bulk of experimental evolution has been performed with mesophilic microbes (including bacteria, fungi, and viruses^2,3,4,5, but largely excluding archaea). A method for experimental evolution applicable to thermophilic microbes would enable us to better understand how they evolve and contribute to a more comprehensive understanding of evolution. This has potentially wide-ranging implications, from deciphering the origins of thermophilic life on Earth to biotechnological applications involving 'extremozymes' used in high-temperature bioprocesses¹⁰ and astrobiological research¹¹.

The archaeon Sulfolobus acidocaldarius is an ideal candidate as a model organism for developing experimental evolution techniques for thermophiles. S. acidocaldarius reproduces aerobically, with an optimal growth temperature at 75 °C (range 55 °C to 85 °C) and high acidity (pH 2-3)^4,6,12,13,14. Remarkably, despite its extreme growth conditions, S. acidocaldarius maintains population densities and mutation rates comparable to mesophiles^{7,15,16,17,18}. In addition, it possesses a relatively small, well-annotated genome (strain DSM639: 2.2 Mb, 36.7% GC, 2,347 genes)¹²; S. acidocaldarius also benefits from robust genome engineering tools, allowing for a direct assessment of the evolutionary process through targeted gene knockouts¹⁹. A notable example of this is the availability of genetically modified strains of S. acidocaldarius, such as the uracil auxotrophic strains of MW001¹⁹ and SK-1²⁰, which can serve as selectable markers.

There are significant challenges with conducting experimental evolution with thermophiles like S. acidocaldarius. Extended incubation at high temperatures required for these studies imposes considerable evaporation for both liquid and solid culturing techniques. Extended operation at high temperatures can also damage the traditional shaking incubators that are commonly used in experimental evolution in liquid media. Exploring multiple temperatures necessitates a substantial financial investment for acquiring and maintaining several incubators. Furthermore, the high energy consumption required raises significant environmental and financial concerns.

This work introduces a method to address the challenges encountered in performing experimental evolution with thermophiles like S. acidocaldarius. Building upon a technique developed by Baes et al. for investigating heat shock response^14,21, the method developed here utilizes bench-top thermomixers for consistent and reliable high-temperature incubation. Its scalability allows for the simultaneous assessment of multiple temperature treatments, with reduced costs in acquiring additional incubation equipment. This enhances experimental efficiency, enabling robust statistical analysis and extensive investigation of factors influencing evolutionary dynamics in thermophiles²². Moreover, this approach significantly reduces the financial initial investment and energy consumption compared to traditional incubators, offering a more sustainable and environmentally friendly alternative.

Our method lays the groundwork for experimentally investigating evolutionary dynamics in environments characterized by extreme temperatures, which may have played a key role during the early stages of the diversification of life on Earth. Thermophilic organisms have unique properties, but their extreme growth conditions and specialized requirements have often limited their accessibility as a model system. Overcoming these barriers not only expands research opportunities for investigating evolutionary dynamics but also enhances the broader utility of thermophiles as model systems in scientific research.

1. Preparation of S. acidocaldarius growth medium (BBM⁺)

NOTE: To cultivate S. acidocaldarius, this protocol uses Basal Brock Medium (BBM⁺)²³. This is prepared by first combining the inorganic stock solutions outlined below to create BBM⁻, which may be prepared in advance. BBM⁺ is then prepared as needed by adding the organic stock solutions to BBM⁻. Stock solution recipes are also presented in Table 1. All media and stock solutions should be prepared in double-distilled H₂O (ddH₂O).

1. Preparation of inorganic stock solutions for growth medium
   1. Prepare Trace Element Stock Solution.
      1. Prepare Trace Element Stock Solution by adding 9 g/L of sodium tetraborate decahydrate (Na₂B₄O₇·10H₂O) to ddH₂O, followed by the addition of 1:1 H₂SO₄ dropwise until it has dissolved.
      2. Sequentially introduce 0.44 g/L of zinc sulfate heptahydrate (ZnSO₄·7H₂O), 0.1 g/L of copper(II) chloride dihydrate (CuCl₂·2H₂O), 0.06 g/L of sodium molybdate dihydrate (Na₂MoO₄·2H₂O), 0.06 g/L of vanadium(IV) sulfate dihydrate (VOSO₄.2H₂O), 0.02 g/L of cobalt(II) sulfate heptahydrate (CoSO₄·7H₂O), and 3.6 g/L of manganese(II) chloride tetrahydrate (MnCl₂·4H₂O). Autoclave the solution and store at 4 °C.
   2. Prepare Fe Stock Solution (1000x) by dissolving 20 g/L of ferric chloride hexahydrate (FeCl₃·6H₂O) in ddH₂O. Sterilize the solution by filtering through a 0.22 µm filter and store at 4 °C.
   3. Prepare Brock Solution I (1000x) by dissolving 70 g/L calcium chloride (CaCl₂·2H₂O) in double-distilled water (ddH₂O). Autoclave and store at 4 °C.
      CAUTION: Dissolving CaCl₂·2H₂O in water is an exothermic process. Perform this step with care. Wear EN407-rated thermal hazard gloves to protect against injury. Do not tightly close the vessel, as pressure may build up.
   4. Prepare Brock Solution II/III by combining 130 g/L of ammonium sulfate ((NH₄)₂SO₄), 25 g/L of magnesium sulfate heptahydrate (MgSO₄·7H₂O), 28 g/L of potassium dihydrogen phosphate (KH₂PO₄), and 50 mL of the previously prepared Trace Element Stock Solution. Using 1:1 H₂SO₄, adjust the pH of the stock solution to 2-3. Autoclave the solution and store it at room temperature (RT).
2. Preparation of organic stock solutions for growth medium
   1. Prepare D-Glucose Stock Solution (100x) by dissolving 300 g/L (30% w/v) of D-glucose in ddH₂O and filter sterilize (through a 0.22 µm filter). Store at 4 °C.
      NOTE: Other carbon sources can be used in place of D-glucose according to experimental requirements (e.g., D-xylose).
   2. Prepare Peptone from Casein Stock Solution (100x) by dissolving peptone from casein at 100 g/L in ddH₂O. Autoclave to sterilize and store the stock solution at RT.
   3. Prepare Uracil Stock Solution (100x) by dissolving 2 g/L of uracil in ddH₂O and filter sterilize (through a 0.22 µm filter); store at −20 °C.
3. Preparation of BBM⁺ growth medium at 1x working concentration
   1. First, prepare 988 mL of sterile ddH₂O by autoclaving.
   2. Prepare 1 L BBM⁻ by adding 1 mL of Brock Stock Solution I, 10 mL of Brock Stock Solution II/III, and 1 mL of Fe stock solution to the 988 mL of sterile ddH₂O autoclaved previously. Adjust the medium pH to the range of 2-3 with 1:1 H₂SO₄.
      NOTE: BBM⁻ can be made in advance and stored at RT for up to 1 month.
   3. Finally, to produce 1 L BBM⁺, combine 10 mL each of D-Glucose Stock Solution, Peptone from Casein Stock Solution, and Uracil Stock Solution with 985 mL of BBM⁻. Mix well and adjust the medium pH to 2-3 with 1:1 H₂SO₄.
      NOTE: BBM⁺ should be freshly made as required.
4. Preparation of BBM⁺ solid medium
   1. Prepare 1 M stock solutions of MgCl₂ and CaCl₂; these will aid in the solidification of the BBM solid medium. For 1 M MgCl₂, add 9.5 g to 100 mL of ddH₂O. For 1 M CaCl₂, add 11 g to 100 mL of ddH₂O. Autoclave to sterilise.
      CAUTION: Dissolving CaCl₂·2H₂O in water is an exothermic process. Perform this step with care. Wear EN407-rated thermal hazard gloves to protect against injury. Do not tightly close the vessel, as pressure may build up.
   2. Mix 3% (w/v) gellan gum (e.g., 30 g/L) in ddH₂O and autoclave to sterilize.
      NOTE: Gellan gum (e.g., Gelrite) is used here as a gelling agent, as standard bacteriological agar cannot solidify at 75 °C.
   3. Separately, prepare BBM⁺ for solid medium, which is to be mixed at a 1:1 ratio with the sterile gellan gum prepared in step 1.4.1.
      1. First, prepare 1 L BBM⁻ as in step 1.3.2. Combine 887 mL of BBM⁻ with 1 mL of Fe Stock Solution and 20 mL each of D-Glucose Stock Solution, Peptone from Casein Stock Solution, and Uracil Stock Solution.
      2. Add 40 mL of 1 M MgCl₂ and 12 mL of 1 M CaCl₂. Mix well and adjust the medium pH to 2-3 with 1:1 H₂SO₄.
         NOTE: The volumes of stock solutions added here are intentionally greater than for preparation of liquid BBM⁺ in step 1.3. This is to account for mixing at a 1:1 ratio with molten gellan gum in the following step.
   4. Preheat the BBM⁺ for solid medium to approximately 75 °C and mix at a 1:1 ratio with molten gellan gum in ddH₂O. Mix well and pour into heat-stable plastic (e.g., polypropylene), glass Petri dishes, or six-well plates for the growth of S. acidocaldarius. Use the agar immediately once blended, as it will set rapidly.
      NOTE: Some polystyrene 90mm Petri dishes commonly used for microbiology will deform if incubated at high temperatures for extended periods.
      CAUTION: Take the appropriate safety precautions when dealing with hot liquids; wear EN407-rated thermal hazard gloves to protect against injury.

2. Reviving S. acidocaldarius from a freezer stock culture

1. Preparation of ventilated growth tubes to ensure sufficient aeration and oxygen availability.
   1. In advance, prepare pierced 2 mL microcentrifuge tubes for Sulfolobus growth. Carefully heat blunt wire (>0.7 mm, such as a straightened paper clip) using a Bunsen burner flame and pierce the lid of the tube, creating a small hole of approximately 1 mm.
   2. Place pierced tubes in an autoclavable vessel (e.g., empty pipette tip box) and autoclave to sterilize.
      CAUTION: Wear EN407-rated thermal hazard gloves to protect against thermal injury to hands from heated metal. Heat the wire for only a short time (1-2 s) to prevent overheating. Only metals with a high melting temperature (steel) are to be used.
2. Inoculate starter cultures of S. acidocaldarius.
   1. Fill the autoclaved pierced 2 mL tubes with 1.5 mL of prewarmed BBM⁺ (as prepared in section 1, pre-incubated at 75 °C for at least 15 min).
   2. Inoculate the BBM⁺-filled tubes with a scraping (equivalent to 50 - 60 mg) from a glycerol stock (25% v/v glycerol, stored at -80 °C). Here, S. acidocaldarius strain DSM639 is used. Fill an additional tube with 1.5 mL of BBM+ as a negative control.
3. To prevent any aerosols from escaping from the pierced 2 mL tube lid and contaminating the growth environment (and any other samples) and to reduce the variability of culture evaporation within the tubes, place a gas-permeable membrane on top of the lid that can withstand elevated temperatures (65-85 °C) and block microbial escape/infiltration.
   NOTE: The utilization of gas-permeable membranes for sealing tubes considerably reduces evaporation. During a 48 h period at 75 °C, sealed tubes exhibit an average volume loss of 8.63% (range: 3.29-16.45%, n = 24 technical replicates, repeated twice), in contrast to 25.05% (range: 11.92-62.91%, n = 24 technical replicates, repeated twice) for unsealed tubes.
4. Incubate the inoculated tubes in a thermomixer at 75 °C with constant shaking at 400 revolutions per minute (RPM) for 48-72 h or until OD_600nm of 0.3 - 0.8 is reached as measured by spectrophotometer.
   NOTE: The pierced lid of the 2 mL tubes and shaking allow appropriate aeration for optimal growth. The lid of the thermomixer should be in place to ensure a consistent temperature is maintained and to reduce evaporation.
5. After the incubation period, give the revived cultures a second growth phase (preconditioning).
   1. Pellet down the cultures by spinning the tubes at 5000 x g for 2 min at RT. Then, discard the supernatant and resuspend the cell pellet in fresh 1.5 mL warm BBM⁺.
      NOTE: This experiment uses the S. acidocaldarius wild-type strain DSM639, but these methods of growth and experimental evolution can be applied to any Sulfolobus strain and potentially other thermophiles.

3. Determining population density, doubling time, and exponential growth phase for S. acidocaldarius

1. Determine population density and time to exponential growth phase for chosen selection conditions. For each environmental condition to be tested, prepare three starter cultures following steps 2.1-2.5.
2. Dilute the three cultures in BBM⁺ to an OD_600nm of 0.01. Make at least 20 mL from each culture. These three cultures represent technical replicates.
3. Perform replicate growth curves of OD_600nm over time using destructive sampling.
   1. Prepare 24 pierced 2 mL tubes from step 2.1, separate them into three sets of eight, and label these technical replicates 1, 2, and 3 (e.g., eight tubes labeled as replicate 1 etc.).
   2. From each of the three diluted cultures from step 3.2, pipette 1.5 mL into seven prepared tubes. Fill the remaining tube with 1.5 mL of BBM+ as a negative control.
4. Incubate all 24 tubes in a thermomixer at 75 °C and 400 RPM. Seal with a gas-permeable membrane. At regular intervals (such as 0, 4, 8, 12, 24, 36, 48 h), remove one tube from each of the three replicates and measure OD_600nm using a spectrophotometer, then discard the tube. Replace the gas-permeable membrane after removing a tube.
5. In parallel, determine the relationship between OD_600nm and population density. Perform serial dilutions (1 x 10^-1 to 1 x 10^-6) in BBM⁺ medium for at least three time points (ideally, start, middle, and end). Inoculate 100 µL of each dilution on solid BBM⁺ medium (prepared as in step 1.4).
   NOTE: To achieve consistent colony growth and accurate counts, it is best to pipette the dilute culture on to solid media in a spot without performing mechanical spreading, such as using an L-shaped spreader or glass beads. Mechanical spreading appears to adversely impact colony formation.
6. Incubate the plates in a static incubator at 75 °C until colonies appear (5-7 days).
   1. Over this time period, keep the incubator humid to avoid excessive drying of the plates, or colonies will not form.
   2. Achieve this by placing the plates in a closed polypropylene box lined with a damp cloth. Pierce the lid of the box at least three times to enable aeration.
   3. Place buckets of water in the incubator to increase the humidity. Refill the buckets daily.
7. Once all OD_600nm measurements have been collected, determine the doubling time and time to reach the exponential growth phase by fitting a logistic curve to the relationship between OD_600nm and time, e.g., using SummarizeGrowth from the growthcurver package in R.
8. Once visible colonies have formed on solid media, count colony forming units (CFUs), aiming to count from the dilution factor, producing 50-100 colonies for the best accuracy.
9. Calculate the cell density in the culture medium using the formula: CFU/mL = number of colonies/(volume plated × dilution factor).
10. Perform log-log linear regression to determine the relationship between log10(CFU) and log10(OD_600nm). Thus, calculate the predicted CFU/mL for a given OD from the slope and intercept of the regression model: predicted CFU = 10^{[slope × log10(OD600nm) + intercept]}.
11. (Optional) Also, perform steps 3.1-3.10 in a shaking incubator to determine whether comparable growth is being achieved in thermomixers.

4. Initiation of independent lineages for experimental evolution

1. Day 1: To generate single colonies, first perform all steps of section 2 above to generate a starting culture.
2. Day 3: Measure OD_600nm of the culture to ensure it has reached sufficient density in the exponential phase (OD_600nm 0.3-0.8 by spectrophotometer).
3. Create serial dilutions of the S. acidocaldarius DSM639 culture from 1 x 10^-1-1 x 10^-6 and spread 100 µL from each 1 x 10^-5 and 1 x 10^-6 dilution into wells of a 6-well plate containing solid BBM⁺ medium (Section 1 and Table 1) in several replicates. Incubate plates at 75 °C in a static incubator as in step 3.5 for 5-7 days for single colonies to emerge.
4. Day 8-10 (5-7 days post incubation):
   1. Establish the number of evolving lineages from single colonies. For each lineage, pick one colony at random from the plates using an inoculation loop. Resuspend the colony in a pierced 2 mL tube containing 1.5 mL of prewarmed BBM⁺ medium. Appropriately label the tube.
   2. Repeat for the desired number of lineages; here, seven lineages were established and labeled SA1-7. Fill an additional tube with 1.5 mL of BBM+ as a negative control.
   3. Seal the tops of the tubes with a breathable membrane and incubate in a thermomixer at 75 °C, 400 rpm for 48-72 h, until the cultures become turbid (equivalent to OD_600nm 0.3-0.8 by spectrophotometer).
5. Day 10-12 (7-9 days post inoculation):
   1. In a bench-top centrifuge, spin the clonal cultures at 5,000 x g for 1 min at RT. Discard the supernatant and resuspend the pellet with 200 µL of liquid BBM⁺.
   2. Mix the 200 µL cell suspensions with 200 µL of 50% glycerol (25% v/v glycerol) to create glycerol stocks of the seven independent lineages and store at -80 °C. These are the ancestral populations for the evolution experiments.

5. Performing the temperature evolution experiment

NOTE: A conceptual diagram outlining the main aspects of the experiment protocol is given in Figure 1.

1. Use the seven ancestral populations generated in section 4 for the evolution experiment.
   1. Carry out all experimental evolution assays in sterile, pierced 2 mL tubes sealed with a permeable membrane (described above, section 2).
   2. Alongside these populations, maintain separate pierced 2 mL tubes with 1.5 mL of BBM⁺ as negative controls free from culture to monitor for (cross-) contamination and to set a threshold for OD indicating no growth of culture.
2. Establish temperature treatments: Using thermomixers allows for multiple temperature treatments to be established. Here, use the following temperature treatments: constant 75 °C, constant 65 °C, and gradually dropping from 75- 65 °C by decreasing the temperature by 1 °C every 4 days ('temperature drop treatment').
   NOTE: These treatments can be tailored to specific experimental requirements. A separate thermomixer is required for each temperature treatment.
3. Day 1: Revive the ancestral populations from their glycerol stocks (prepared in step 4.5) following the steps outlined in section 2.
4. Day 3:
   1. Dilute these cultures to an OD_600nm of 0.01 (measured by spectrophotometer) to a total volume of at least 6 mL of prewarmed BBM⁺. Aliquot 1.5 mL from each of the seven dilute ancestral population cultures into three separate pierced 2mL tubes, one for each temperature treatment established in step 5.2.
      NOTE: This aliquotting step ensures that any genetic variation present in each ancestral population is present in all temperature treatments at the outset of the evolution experiment. These cultures will act as transfer 0 (Tx0) to begin the temperature evolution experiment.
   2. Seal the tubes with the breathable membrane and place each set of seven ancestral populations into separate thermomixers. Set each thermomixer to the required temperature and 400 rpm to begin the evolution experiment.
5. Day 5:
   1. After 48 h, carry out transfer 1 (Tx1) by inoculating 1.5 mL warmed BBM⁺ medium in a pierced 2 mL tube with 15 µL of culture from each of the Tx0 cultures (1:100 dilution). For speed, carry out this step with a variable-width multichannel pipette to complete multiple transfers from a single experiment in unison, reducing the time the cultures are out of the thermomixer.
   2. As before, seal the tubes before putting them back in randomized positions into their respective thermomixers. Carry out randomization manually or through the iMetaLab 96 well randomizers.
6. Measure the OD_600nm of Tx0 in a plate-based spectrophotometer (200 µL in 96 well plates; a same volume of BBM⁺ is used as a blank) to track the growth of the independent lineages.
   NOTE: The OD_600nm should be measured after performing the transfer to a fresh medium to prevent contamination. Optical density can also be measured by other means, such as with a standard spectrophotometer. The use of a plate-based spectrophotometer enables simultaneous measurement of multiple cultures, streamlining the process.
7. Repeat steps 5.5 and 5.6 on days 7, 9, 11, etc., corresponding to Tx2, Tx3, Tx4, etc. Hold the temperature constant for the 75 °C and 65 °C treatments. For the temperature drop treatment, decrease the temperature by 1 °C every second transfer (i.e., on Tx2, Tx4, Tx6, etc.).
8. Prepare glycerol stocks (step 3.2.3) of populations after every 10^th transfer (i.e., Tx10, Tx20, etc.), labeling the tube with the ancestral population identifier (e.g., SA1 for the 1^st independent population) and the transfer number (e.g., Tx10 for the 10^th transfer). Use these glycerol stocks later to revive populations to study how populations changed over time or to restart the experiment if required (e.g., due to contamination or unexpected incidents resulting in loss of cultures).
9. Let the experiment proceed either for a predetermined number of transfers or until a predetermined number of generations has elapsed. On the final transfer, prepare glycerol stocks of all the descendent lineages.
   NOTE: Here, the experiment proceeded until the temperature drop treatment reached 65 °C, which occurred on Tx22. This corresponds to approximately 150 generations (calculated based on the dilution factor of 100 in 4.6: log₂(100) = 6.64 generations per transfer). The length of evolution experiments can be adjusted according to the experimental requirements.

6. Post-evolution experiment growth assays: ancestral vs. evolved lineages

NOTE: A conceptual diagram outlining the growth/fitness assay protocol is given in Figure 2.

1. Prepare pierced 2 mL tubes with 1.5 mL BBM⁺. Prepare enough tubes to grow all the descendent lineages plus each of the ancestral strains.
2. Revive ancestral populations, and each descendent population stored in step 5.9 after the final transfer of the evolution experiment using the method outlined in steps 2.2-2.5, but with incubation at the temperature each population experienced for the final two transfers of the evolution experiment, i.e., 75 °C for the 75 °C constant treatment, and 65 °C for the 65 °C constant and temperature drop experiments. To achieve comparable population densities, conduct the 75 °C incubation for 72 h and the 65 °C incubation for 120 h.
3. Measure OD_600nm of the grown cultures via a plate-based spectrophotometer.
4. Perform growth assays of each descendent population and each ancestral strain at both temperatures (i.e., 75 °C and 65 °C) in three replicates. Prepare pierced 2 mL centrifuge tubes with 1.5 mL of BBM⁺. Prepare six tubes for each evolved lineage and each ancestral strain. Label the tubes with the lineage name (SA1-7 or ancestor), the growth temperature (75 °C or 65 °C), and the replicate ID (1, 2, or 3).
   NOTE: If fewer than six thermomixers are available, the growth assay can be replicated over multiple days across several experimental blocks. In this instance, it is important to measure each lineage and ancestor in each experimental block so that block effects can be estimated and accounted for statistically.
5. Inoculate the fresh medium in the six prepared tubes with 15 µL of culture from each descendent lineage and ancestral strain.
6. Set three thermomixers to 75 °C and three thermomixers to 65 °C, designating each thermomixer to a replicate ID. Place the tubes in the thermomixer corresponding to the temperature and replicate ID. Within each thermomixer, ensure that the lineages and ancestors are randomly placed to control for heterogeneity.
7. Seal each set of 24 tubes with a permeable membrane. Incubate for 48 h.
8. Transfer 200 µL from each culture into a 96-well plate. Measure OD_600nm using a spectrophotometer.
   NOTE: A variable-width multi-channel pipette can be used to facilitate transferring between microcentrifuge tubes and 96-well plates.
9. Plot and analyze data using statistical software (e.g., R).

7. Whole genome sequencing of evolved lineages and identification of mutations

1. Revive descendent lineages stored in step 5.9 in pre-warmed BBM⁺ at 75 °C of 48-72 h by inoculating a scrape of ice from each frozen population into BBM⁺ as in section 2, steps 2.2-2.5. Prepare between 1-10 mL of culture volume to obtain a sufficient quantity of genomic DNA. The exact volume required depends on the option chosen for sequencing in steps 7.2 or 7.3 (below).
   NOTE: It is good practice to also sequence the strain used to initiate the ancestral populations. This is to correctly determine whether any mutations detected in the descendent lineages arose during the evolution experiment and not before.
2. (Option 1) Extract genomic DNA and prepare Illumina sequencing libraries for Illumina sequencing.
   1. Extract and purify genomic DNA using a commercial genomic DNA extraction kit for bacteria.
   2. Ensure the extracted genomic DNA meets the quantity and quality requirements of the sequencing provider by using commercial kits recommended by the provider.
   3. Perform genomic DNA library preparation using a commercial kit according to the manufacturer's protocol. A 250 bp paired-end protocol is recommended. Store extracted genomic DNA at -20 °C until ready to submit samples to the sequencing provider.
3. (Option 2) Alternatively, send the samples to a commercial provider that performs DNA extraction, genomic DNA quality checks, Illumina library preparation, and Illumina sequencing as a combined service.
4. Analyze sequenced genomes to identify mutations.
   1. Receive FASTQ files and, if not already performed by the sequencing provider, perform adapter trimming using Trimmomatic with a sliding window quality cutoff of Q15.
   2. Using the trimmed FASTQ files, run the breseq pipeline (version 0.38.1) to detect mutations, comparing the genomic DNA reads to the reference sequence for the chosen strain. For S. acidocaldarius DSM639, this is RefSeq ID NC_007181.

8. (Optional) Assessment of thermomixer vs. incubator energy consumption

1. To compare the energy consumption of using an incubator versus a thermomixer for growth, measure the energy consumption of each device over 24 h using an energy monitoring plug.
2. Use two plugs to measure the energy consumption of both devices at the same time. Alternatively, measure energy consumption on successive days.
3. Disconnect the incubator or thermomixer from the electrical socket. Plug in the energy monitoring plug into the socket and initialize it according to the manufacturer's instructions, including installing the monitoring and control software.
4. Plug the incubator or thermomixer into the energy monitoring plug and allow it to run for a minimum of 2 h (but ideally for 24 h or longer) to achieve a good estimation of typical energy consumption. Record the energy consumption of each device.

Growth curve measurements
Growth curves for S. acidocaldarius DSM639 are shown in Figure 3A. Growth was found to be similar when comparing incubation using thermomixers with that in conventional incubators. Average growth rate parameters were estimated by fitting a logistic curve to each replicated growth curve and calculating the mean and standard error. Times to mid-exponential phase on the thermomixer and incubator were 27.2 h ± 1.1 h and 31.1 h ± 1.9 h, respectively. Estimated initial doubling times for the thermomixer and incubator at 75 °C were 4.29 h ± 0.28 h and 4.19 h ± 0.44 h, respectively, which is consistent with previously published values²⁴. The relationship between log₁₀(OD_600nm) and log₁₀(CFU) was well characterized by a linear model (adjusted R² = 0.82, F(1,22) = 104, p < 0.00001, slope = 1.73 ± 0.17, intercept = 9.73 ± 0.14). The relationship between OD_600nm and CFU is thus given by the formula CFU = 10^{(1.73 × log10(OD600nm) + 9.73)}. An OD_600nm of 0.3 thus corresponds to approximately 6.7 × 10⁸ CFU/mL (Figure 3B).

Temperature evolution experiment
Three temperature conditions, constant 75 °C, constant 65 °C, and temperature drop (75-65 °C, decreasing by 1 °C every two transfers), were initiated using seven independent lineages derived from S. acidocaldarius DSM639. OD_600nm measurements were taken following each transfer over the 45 days of the experiment (approximately 150 generations at 75 °C) and are shown in Figure 4. OD_600nm measurements taken across days are inherently noisy, as they may be subject to subtle differences in, for example, growth period, temperature, etc. However, measurements taken across days can still be useful to assess population viability, as well as give an indication as to whether fitness is improving over time. Lineages from the constant 75 °C condition increased in OD_600nm from an initial range of 0.125-0.3 to a range of 0.248-0.471 by the end of the experiment. This suggests that fitness has improved with this treatment. In contrast, lineages from the constant 65 °C treatment displayed a drop in OD_600nm, from an initial range of 0.018-0.087 to 0.008-0.04 at the final time point. This suggests that populations were not able to adapt to the constant 65 °C temperature, although the fact that viable organisms could be recovered shows that the populations were not washed out through successive dilutions, suggesting some degree of adaptation. Finally, populations in the temperature drop treatment increased from the initial OD_600nm range of 0.099-0.279 to 0.3-0.39 at Tx6 (corresponding to 288 h and 73 °C in this treatment), followed by a steady decrease to a range of 0.003-0.024 at the final time point.

Growth/fitness assays
Fitness assays were performed for each descendent population following the evolution experiments. OD_600nm was assayed after 48 h growth for all seven independent lineages, followed by fitting linear models in R for each assay temperature, with 'selection environment' as a main effect and 'replicate/thermomixer' as a block effect. Growth of the ancestral strain was used as the reference level for treatment contrasts. The data are shown in Figure 5.

When assayed at 75 °C, there were on average significant increases in fitness over the ancestral strain for lineages from the constant 75 °C (t-test: t₂₁₀=3.64, p=0.0003) and constant 65 °C (t-test: t₂₁₀=2.8, p=0.005) treatments, but not from the temperature drop treatment (t-test: t₂₁₀=−0.87, p=0.38). When assayed at 65 °C, on average, lineages from all treatments displayed an increase in fitness (constant 75 °C lineages; t-test: t₂₁₀=4.68, p<0.0001, constant 65 °C lineages; t-test: t₂₁₀,=4.24, p<0.0001, temperature drop lineages; t-test: t₂₁₀=3.15, p=0.002). However, for both assay temperatures, there were considerable differences between lineages in their fitness (Figure 5). Some lineages did not differ considerably from the ancestral strain or had decreased in fitness; this was particularly evident for lineages from the temperature drop treatment.

It is worth noting that the relationship between OD_600nm and CFU/mL might have changed during the evolution experiment. This can be assessed by determining growth parameters for evolved lineages (following steps 3.1-3.10).

Whole genome sequencing results
Whole genome analysis was carried out using breseq (version 0.38.1)²⁵ on the seven lineages from the constant 75 °C condition using the reference genome of S. acidocaldarius DSM639 (RefSeq accession NC_007181.1). A variety of mutations were revealed involving insertions, deletions, and single nucleotide polymorphism (SNP) across the genomes of all the descendent lineages (Table 2). Multiple insertion and nonsynonymous mutations were in protein-coding genes, as well as intergenic regions which may influence gene expression due to frameshifts in genes' promoter regions. Some of the mutations were consistent across multiple descendent lineages in genes involved in various functions like cell wall biosynthesis, transcription, metabolism, cellular transport, and catalytic activity (Table 2). Among these mutations was a large deletion of 54,667 base pairs in five out of the seven lineages; this was confirmed by a missing coverage evidence plot for each population (frequency ranging from 93.2% to 100%). The deleted region equates to a loss of 53 genes; the roles of these genes in adaptation will be investigated in future studies. Some differences were noted between the used isolate of DSM639 and the published reference sequence (shown in Supplementary Table 1).

Energy consumption of shaking incubator versus thermomixer
The energy consumption of a shaking incubator was compared to a thermomixer using a commercially available energy monitoring smart plug across a range of common incubation temperatures and the 75 °C temperature used here. At 75 °C, the thermomixer consumed approximately 1/40^th the energy of a traditional shaking incubator (Figure 6), suggesting thermomixers as a potential means of reducing the carbon footprint associated with experimental evolution.

Figure 1: Flow chart illustrating the evolution experiment protocol across three temperature treatments. Please click here to view a larger version of this figure.

Figure 2: Flow chart illustrating the steps of the growth/fitness assay protocol. Please click here to view a larger version of this figure.

Figure 3: Determination of key growth parameters and comparison of incubation devices for S. acidocaldarius DSM639. Three replicate cultures were grown at 75 °C in 7 separate tubes. Destructive sampling was used to measure (A) OD_600nm (means ± standard error of n=3 technical replicates; some error bars are smaller than plotting symbols); curves represent fitted logistic growth models and (B) colony forming units (CFUs); lines represent fitted log-log linear regression models). Please click here to view a larger version of this figure.

Figure 4: Representative results for optical densities (OD_600nm) obtained during evolution experiments. Optical densities of independent lineages measured during the evolution experiments under three temperature treatments (constant 65 °C, constant 75 °C, drop 75 °C-65 °C) conducted over approximately 150 generations. Curves depict Loess smooths over time for each independent lineage. Please click here to view a larger version of this figure.

Figure 5: Representative results for growth assays. Growth assays for independent lineages derived from S. acidocaldarius DSM639 following temperature evolution experiments (constant 65 °C, constant 75 °C, drop 75 °C-65 °C) in comparison to the ancestor strain. For all lineages, growth was assayed at 65 °C and 75 °C. Colored points show the mean ± standard error of technical replicates (shown in grey, n = 12 for the ancestor and n = 3 for each evolved lineage). The grey bar denotes the mean ± standard error of the ancestral fitness. Please click here to view a larger version of this figure.

Figure 6: Energy consumption of traditional incubator versus thermomixer devices. Energy consumption was recorded using a commercially available energy-monitoring smart plug over 2 h. Please click here to view a larger version of this figure.

Table 1: Media recipes and stock solutions required to grow S. acidocaldarius. All media and stock solutions should be made with double-distilled H₂O (ddH₂O) and then sterilized either by autoclaving or filter sterilizing through a 0.22 µm filter, as indicated. Please refer to section 1 of the protocol for a detailed description of how to prepare all media and stock solutions. Please click here to download this Table.

Table 2: Representative results for whole genome sequencing of descendent lineages. Mutations found in descendent lineages descendant from S. acidocaldarius DSM639 from constant 75 °C treatment. Mutations indicate changes relative to the direct ancestor of the lineage, which possesses several changes relative to the reference sequence for S. acidocaldarius DSM639 (RefSeq NC_007181; mutations in the ancestor are shown in Supplementary Table 1). n indicates the number of lineages in which a mutation was found. → gene on the 'forward reading frame'; ← gene on the 'reverse reading frame'. †These changes are relative to a (A)10→11 frameshift present in the isolate of S. acidocaldarius DSM639 (Supplementary Table 1). Please click here to download this Table.

Supplementary Table 1: Mutations present in the isolate of S. acidocaldarius DSM639 relative to the reference sequence (RefSeq accession NC_007181.1). → gene on the 'forward reading frame'; ← gene on the 'reverse reading frame'. ‡SACI_RS04020 is annotated as a pseudogene in NC_007181.1, but the Δ1 bp frameshift mutation observed here putatively restores its function as with the mutation, it encodes a protein with 100% identity to rgy reverse gyrase gene (RefSeq accession WP_176586667.1). Please click here to download this File.

This work has developed an experimental evolution protocol for thermophiles, here tailored for the archaeon S. acidocaldarius, but adaptable to other microbes with high-temperature growth requirements. This protocol builds on techniques initially designed for mesophilic bacteria but is specifically modified to overcome the technical challenges associated with high-temperature aerobic growth^2,4,5,24. The creation of this method will enable the exploration of previously uncharacterized evolutionary dynamics that occur under extreme temperature conditions, resembling the conditions that potentially existed during the earliest stages of the evolution of life¹.

There are several critical aspects of the protocol that must be observed to achieve reliable results from evolution experiments. The first is independence between lineages, which is achieved by streaking out to single colonies before initiating the experiment. This is imperative to ensure that any common mutations found in separate lineages have independent origins. If, instead, a bulk liquid culture is aliquoted to initiate the lineages, common genetic variants may be distributed to each population. A second key consideration is to control for contamination, which is achieved through maintaining a negative control free from culture and monitoring for growth. Unlike experiments with mesophiles, extremophile evolution experiments are unlikely to be contaminated from external sources, but cross-contamination between lineages may still occur. Maintaining a balance between aeration and evaporation is a further key consideration. S.acidocaldarius is an obligate aerobe and so requires oxygen for growth^10,23. Manually piercing polypropylene microcentrifuge tubes and sealing them with gas permeable membrane^21,25, was key to maintaining aeration without excessive evaporation due. This reduction can be likely attributed to the preservation of a uniform environment within the sealed tubes (data not shown).

Two aspects of the method may require adjusting or troubleshooting according to the specific environmental conditions used for the evolution experiment: the dilution factor and the time between transfers. In this method, a 1:100 dilution was performed every transfer, necessitating an average of log₂(100) = 6.64 doublings to return to the same population size. This provides a good balance between the potential for adaptation (an optimal dilution factor is calculated by Wahl and Gerrish²⁶) and experimental tractability. In principle, a strong dilution factor would enable more generations to occur per day, as it requires more doublings to return to the same population density (e.g., 1:1000 requires log₂(1000) = 9.97 doublings on average). However, if fewer doublings are achieved than are required between transfers, the populations will gradually be washed out through dilution unless a beneficial mutation that decreases doubling time occurs. with populations transferred every 48 h, it is required that the doubling times be less than 8 h, which aligns with the estimated doubling time and previously published values for heterotrophic growth in S. acidocaldarius at 75 °C (4-6 h)^23,27. However, the gradual decrease in OD_600nm over time for the constant 65 °C and temperature drop treatments suggests that this may not be a long enough time to achieve the required number of doublings and that a longer transfer period would be beneficial.

Several important advantages of using thermomixers over traditional shaking incubators for experimental evolution are noted, particularly when investigating adaptation to temperature. Thermomixers present a cost-effective alternative to shaking incubators, with the added advantage of a reduced spatial footprint. These characteristics enable the efficient assessment of multiple temperature treatments in parallel. Thermomixers also exhibit lower energy consumption than shaking incubators, so their use can potentially reduce environmental impact when a modest number of tubes is incubated (e.g., fewer than 100). This disparity in energy consumption is particularly evident at temperatures relevant to thermophiles but is also applicable across a spectrum of more moderate temperature ranges (Figure 6), meaning thermomixers could also be adopted for reducing the environmental impact of mesophile experimental evolution. Collectively, these attributes suggest that using thermomixers in experimental evolution has the potential to improve the accessibility of this research methodology, particularly in resource-constrained environments, such as laboratories in developing countries and in primary and secondary education.

Despite providing a way forward for thermophile experimental evolution, the current method has some limitations relative to mesophile experimental evolution. Firstly, the number of replicate independent lineages that can be established using thermomixers is small relative to high-throughput methods involving microtiter plates, which are in widespread use for mesophiles⁴. Evaporation at high temperatures currently precludes the use of microtiter plates over the time window required to reach exponential growth. Secondly, estimating key growth curve parameters, such as growth rate, doubling time, and lag time, is made challenging as plate-based spectrophotometers that can operate at high temperatures (>45 °C) are not currently available for purchase. With the method outlined here, growth curve characterization at present requires labor-intensive destructive sampling and hence limits throughput. Finally, high-throughput techniques for assessing the competitive fitness of adapted isolates are not yet thoroughly developed for thermophiles. Particularly, the development of a thermostable fluorescent protein that can be used to differentiate between strains would enable high-throughput flow cytometry-based fitness assays^28,29; recent developments of thermostable fluorescent proteins with different spectra raise this as a potential future development^30,31,32. Increases in throughput are a logical next step for the further development of experimental evolution techniques for thermophiles, as it will enable evolutionary processes to be more completely described.