1.4 The model organism Pseudomonas fluorescens SBW25 10
1.4.4 The Reverse-‐Evolution Experiment (REE) 20
The experiment presented here took advantage of the ability of P. fluorescens to quickly diversify in a static microcosm. New variants (WS) arise by mutations in genes that are involved in the production of cellulose (see section 1.4.3). The polymer enables the cells to colonise the air-‐liquid interface. Within this niche oxygen and nutrients are equally available and provide a benefit enabling WS to increase in frequency (see section 1.4.2). Under shaking conditions, however, this novel trait is lost due to the costs of cellulose production without it providing any benefits. One of the fundamental questions that were addressed by the REE was how repeatable and predictable is evolution. Can a novel trait (WS), once it is lost under shaking conditions, repeatedly evolve when the bacteria are grown again in a static environment? What is the genetic basis of repeated WS evolution?
Dr. Hubertus Beaumont and Professor Paul Rainey initiated the REE in 2006 with 12 replicate lineages of the bacterium P. fluorescens SBW25. The lineages were kept in glass microcosms containing 6 ml liquid KB media. All lines were treated equally and were exposed to two alternating environmental conditions – static and shaken. Under static conditions the microcosms were left undisturbed, whereas under shaken conditions the microcosms grew under constant shaking at 160 rpm.
The experiment started by growing 12 replicate cultures of P. fluorescens SBW25 in the static environment. After the first selection round (three days), the static cultures were diluted and plated on KB agar plates, which were incubated for two days. The plates were then examined for colonies with morphologies different to that of the ancestral colony phenotype. One colony of the most common new type was picked from the agar plate, transferred into a fresh microcosm and subsequently exposed to shaken conditions. After three days (second selection round) the procedure described above was repeated, and the most common new colony type now growing under shaken conditions was selected and transferred into a new microcosm. In essence, every time a new type was found, it was transferred into new media, and kept under the opposite environmental condition
Figure 1.8: Experimental design of the REE of one of the 12 replicated lines in P. fluorescens SBW25. The initial inoculation was carried out with 12 replicates of SBW25. Cultures were kept in microcosms under static conditions. After one selection round (three days), the cultures were plated on agar plates and screened for new colony types. The most common new type was used to inoculate new KB microcosms that were kept under shaken conditions. After three days the cultures were plated and screened for the most common new colony type. The new type was then used to inoculate the new static microcosms starting a new cycle of evolution. All 12 replicate lines went through three to eight cycles of evolution during the REE (Beaumont et al., 2009).
Where no new type could be detected after three days, 6 µl of the bacterial culture
were transferred into fresh microcosms. These were kept under the original conditions parallel to the plating step (see above), either static or shaken, for additional three days (same selection round). Cultures were diluted, plated and screened for new types. Where no new colony morphology type was found on the plates, the procedure described above was repeated until a new colony type was observed. If no new type was observed over a long period of time (10 transfers within the same selection round) the lineage was terminated. Every newly evolved type was stored in small tubes containing a saline-‐glycerine solution at -‐80°C. The
‘frozen fossil collection’ was later revived and used for further analysis (Beaumont
1.4.4.1 Whole-‐genome sequencing of evolutionary endpoints
Sequencing the entire genome of an organism has become an indispensable and powerful tool for evolutionary biologists. Primarily it has been used to discover changes in the DNA sequence of an evolutionarily derived genome by comparing it with the known reference ancestral genome sequence. Solexa sequencing is an efficient technology that yields high quality sequences and was used in this experiment. Here the genotypes of the evolutionary endpoint in the 12 lineages that did not produce any more new colony types was revived from the freezer stock and used for Solexa sequencing. This way it was possible to reconstruct the evolutionary pathway that was taken by each lineage depending of the environmental conditions (static or shaken). The phenotype was mapped to the genotype by identifying all causative mutations and the order in which they appeared.
1.5
The rise of a stochastically switching phenotype
The experimental design of the REE made it possible to follow real-‐time evolution of 12 parallel lineages of P. fluorescens that evolved alternately in a static and shaken environment, starting from the same clonal founder population (Fig. 1.8; see section 1.4.4). Increased selection pressure under static conditions, due to the lack of oxygen in the media, favours diversification into multiple new types, usually different WS phenotypes. Propagation, typically of a WS type, in a shaken environment often reverses the phenotype to one that resembles the ancestral SM type (Fig. 1.1; Rainey & Travisano, 1998). Each phenotypic change was caused by a mutational modification. In two of the 12 lineages, Line 1 and Line 6, a novel phenotype evolved independently after nine rounds of selection. This new type showed distinctive colony morphology on agar plates and was not detected in any of the 12 lineages during earlier selection rounds. Phenotypic and genetic analysis revealed that the new type had properties of a stochastically switching genotype (Beaumont et al., 2009; Gallie, 2009).