---+ Evolution Experiments ---++Introduction An evolution experiment typically consists of evolving one or more ancestor strains in laboratory culture and assessing the types and impacts of acquired mutations on that organism's ability to reproduce, consume or produce compounds of interest, or resist particular stress conditions. The topics discussed on this page assume a general bacterial evolution experiment where a bacterial strain is serially passaged to fresh media over time, however there are other types of evolution experiments including evolving eukaryotes or viruses, passaging pathogens and symbionts between hosts, and evolving multiple strains in co-culture that require additional considerations. Broadly speaking there are two kinds of evolution experiments: adaptive evolution and mutation accumulation. In adaptive evolution experiments cells within a population compete with each other and generally trend towards increasing fitness. These experiments are useful for identifying mutations that improve a particular function, such as growth on a particular compound or resistance to a particular stress. Individuals with detrimental mutations are usually out-competed, thus the mutations evolved in these experiments are biased towards beneficial mutations. Mutation accumulation experiments, on the other hand, reduce or remove selection, allowing the persistence of all types of mutations and are thus useful for calculating overall mutation rates or examining the spectrum of possible mutations. In these experiments the effect of selection is usually limited by randomly selecting colonies to streak from one plate to another. Since colonies are grown initially from individual cells this also represents the smallest possible bottleneck size, supporting genetic drift. The discussion below applies predominantly to adaptive evolution experiments, you can read more about mutation accumulation experiments [[ProcedureBacterialMutationAccumulation][here.]] For an example protocol and overview of a specific adaptive evolution experiment see [[https://www.jove.com/t/65342/daily-transfers-archiving-populations-measuring-fitness-long-term][this paper]] describing the Long-term Evolution Experiment with <i>E. coli</i>. ---++Experimental design An overall evolution experiment will likely consist of four phases: *1)* inoculating populations, *2)* routine transfers, *3)* isolating clones and freezing stocks, and *4)* analysis. *1) Inoculating populations* Regardless of whether they are adaptive or mutation accumulation experiments, evolution experiments should be started by streaking the ancestral stock onto an agar plate and then initiating each population from a separate isolated colony. This reduces the impact of mutations that may already be present in the ancestral stock and ensures that the populations begin as single genotypes. It's a good idea to start more populations than you need in case some become contaminated later. Also set up a separate tube or flask with an uninoculated control to check for contamination. After the first day of growth populations should be frozen so they can be sequenced and referred to later, in particular when determining whether identical mutations shared between samples evolved during the experiment or were present in the initial culture. *2) Routine transfers* Over the course of an evolution experiment the populations will be transferred to fresh media to continue growing after their initial media has been used up. Between transfers the pipette should be wiped with ethanol to prevent cross-contamination between samples. Designing an effective evolution experiment requires taking into account the impact of: * *Timing of transfers* Cultures should be transferred at regular intervals after they have reached saturation, usually every 24 hours (+/- one hour). Particularly in the beginning of an experiment, before the populations have adapted to the culture conditions, it's possible some cultures will not reach saturation in one day. It's usually best to postpone transferring all the cultures for an additional day to allow them all to reach saturation and remain on the same transfer schedule. * *Volume of transfers and bottleneck size* The volume transferred to fresh media will determine the number of cells transferred to the next culture (i.e. the population bottleneck). Transferring a small number of cells can lead to genetic drift where mutations, including detrimental ones, are able to proliferate in the population by chance rather than because they provide an advantage. On the other hand, transferring too many cells limits the number of generations the population will be able to undergo before the culture is saturated. ~10<sup>6</sup> cells is typically a good number, the equivalent of transferring 5 µl of an overnight culture of <i>E. coli</i> grown in LB media. * *Number of generations* An evolution experiment will generally run for a set number of generations (like 300 generations for one month of ten generations per day) or, in a longer, open-ended evolution experiment, require tracking the number of generations for which each population has been evolved. The number of generations each day will depend on the dilution factor (df) of each transfer, as the populations will regrow to approximately the same density at saturation. The number of generations for a culture each day is equal to log(df)/log(2), for example diluting a culture 1000-fold and allowing it to grow to saturation will result in 10 generations. * *Stress conditions* If the goal is to improve resistance to a particular stress condition it may be beneficial to increase the stress over time as the strains adapt to increasing concentrations and a high level of stress may be too strong to overcome until the populations have acquired a few mutations. * *Back-up cultures* After transferring cultures to fresh media, the previous day's cultures should be stored at 4°C until the next day. Each day a fresh, uninoculated control should be set up to confirm the media remains uncontaminated. If an uninoculated control shows contamination the back-up cultures should be taken out of the fridge, warmed up, vortexed, and transferred from instead. Note that this does not count as progressing the total number of generations again as the previous day's transfer was unsuccessful. If contamination is a concern, for example with slow-growing populations in rich media, it's a good idea to streak out the populations partway through an evolution period to confirm they have not been contaminated (diversity in colony size is to be expected, however, as the strains within a population evolve). *3) Isolating clones and freezing stocks* At the end of the transfer period the evolved strains should be [[ProtocolsFreezingStrains][frozen]] for further analysis. This can be done as populations by freezing culture media from the last transfer and/or as isolates by streaking each population on a plate and inoculating a large, isolated colony into fresh media and then freezing it. Isolates can also taken from frozen populations after the experiment by thawing and streaking out the stock. *4) Analysis* The end result of an evolution experiment is likely isolated, evolved clones that have acquired additional mutations relative to their ancestor. The goals, then, are usually to identify these mutations in their genomes and determine what their functional consequences are, if any. * *Identifying mutations* Mutations can be identified by sequencing the isolates' genomes and comparing them to the ancestor with [[https://barricklab.org/twiki/pub/Lab/ToolsBacterialGenomeResequencing/documentation/][<i>breseq</i>.]] Sequencing isolates from replicate populations allows the identification of parallelism, or similar mutations arising independently in multiple cultures, further supporting their importance. If identical mutations are identified in multiple cultures (for example the same base substitution at the same position in the genome) the original cultures frozen after the first day of evolution should be checked for the presence of the same mutation to determine whether it arose independently or was present from the initial cultures. * *Functional analysis* The functional consequences of evolution are typically quantified with [[ProceduresLongTermCompetitions][fitness assays]] or other functional assays (i.e. [[ProtocolsGrowthRates][growth curve,]] minimum inhibitory concentration, or biochemical assays, as appropriate). Performing these assays on evolved isolates can confirm that the isolates have evolved the desired trait, however to tie the phenotype to a specific mutation candidate mutations should be reconstructed in the ancestor strain and then tested. These assays should be performed under the same conditions in which the populations were evolved (i.e. the same media) as the mutations may only be beneficial under those conditions. ---++Complications Some potential difficulties that may arise during evolution experiments include: * *Evolved strains without improved fitness* It's possible that measuring the fitness of evolved strains will not demonstrate a substantial increase in fitness. This can occur for multiple reasons. First, mutations may only be beneficial in the specific environment in which they evolved. This can include, for example, the presence of other strains within the population if a strain has evolved mutualistic or cheating interactions with other strains that are not available when performing assays on isolates. Some mutations, such as tandem duplications, are also easily reverted and may be lost during isolation if the same selective pressure is not maintained. * *Hypermutators* Strains may become hypermutators, usually by mutating proofreading genes like <i>mutS</i> or <i>mutL</i>. This leads to a substantial increase in the mutation rate, however many of the mutations will likely be neutral, such as synonymous codon changes. Inactivation of specific proofreading genes results in mutational spectra with biases towards specific base substitutions.
This topic: Lab
>
WebHome
>
ProtocolList
>
ProceduresEvolutionExperiment
Topic revision: r9 - 2024-07-12 - IsaacGifford
Copyright ©2025 Barrick Lab contributing authors. Ideas, requests, problems?
Send feedback
