Background

Adaptive laboratory evolution experiments provide a powerful tool for readily-replicated, precisely-controlled, long-term, real-time study of evolutionary paths to antibiotic resistance that needs to be considered for the development of sustainable new antimicrobial therapies. The simplest protocol for an experimental evolution is passaging cells in batch cultures using the method of serial transfer (Gresham and Dunham, 2014). During the serial passage, the bacterial population is cultured for 24 h covering populations at exponential to stationary phases, followed by transfer of small bacterial inoculum from sub-inhibitory medium to fresh culture medium restarting the entire cycle anew (Jansen et al., 2013). This protocol can be parallelized using microtiter plates and multiplex liquid handling, which enable the simultaneous analysis of hundreds of populations. The long-term selection can also be performed by continuous chemostat culturing techniques. The chemostat protocol implies the maintenance of a bacterial population continuously in exponential phase by regular inflow of fresh media (that may be supplemented with the desired dose of an antimicrobial) and simultaneous outflow of the waste culture.

Jansen and co-authors (2013) stated that each method covers certain situations in the patient and/or clinical environment. They pointed out that serial passage reflects patient-to-patient transfer or spread across tissues within a single patient, where the transmitted inoculum is initially low and can outgrow after the successful invasion. The chemostat would rather represent infection within a single tissue or systemic bacteremia. When the evolving pathogen population in the chemostat is also continuously challenged by a temporal gradient of an antimicrobial, the chemostat system may simulate bacterial evolution in the host environment, in response to antimicrobial treatment. During this adaptive evolution, mutations may arise, and those that are beneficial under the antimicrobial selection pressure will be fixed over time in the population. Cryopreservation of regular samples from different time points will allow the generation of a living fossil record of the evolving population (Jansen et al., 2013; Gresham and Hong, 2015). After the adaptation, understanding what genetic changes confer the resistance is desirable. Therefore, combining the experimental evolution study with whole-genome sequencing of the evolved population, one can study the evolution of antimicrobial resistance in real-time and identify resistance mechanisms to a novel antimicrobial.

Despite the advantages, the chemostat continuous devices have been limited in their utility first of all because of difficulties posed by the construction, operation and maintenance. Aside from these, only very few publications provide relevant detailed instructions. One other drawback with chemostat systems is that they require large amounts of media and are expensive to build or purchase as commercially available options. Additionally, the replication requiring the operation of several chemostats in parallel over extended periods is technically challenging and rather laborious. We have developed a multiplex, low-cost, small-scale continuous culture device that can operate at a working volume of 25 ml and allows the operation of 8 chemostats in parallel by a single person (Figure 1). Herein, we provide a detailed technical information how to build and operate it. Another challenge with continuous culture devices is that for longer experiments their use is often limited by biofilm formation on the culture chamber walls. Our solution to this problem may be simply aseptically transferring the culture to a new polypropylene culture tube to prolong the experiment each time biofilm formation becomes obvious.

Here, we also validate the use of our DIY continuous culture system to generate and study the resistance emergence in bacteria. The drug-sensitive model bacterium is challenged by progressively increasing concentration of a novel antimicrobial/common antibiotic, thus generating drug-resistant mutants and keeping them constantly challenged. This device and method can be used ultimately to forecast antimicrobial treatment outcomes, including the likelihood of resistance, cross-resistance and collateral-sensitivity. We anticipate that this affordable system will become a useful tool to address the resistance and to develop non-antibiotic based therapies capable of bypassing resistance acquisition. Alternatively, we propose its use to assess the adaptative evolution of photosynthetic microorganisms, such as cyanobacteria and microalgae, to extreme environmental changes derived from anthropogenic pollution (i.e., herbicides, heavy metals, petroleum contamination and others).