Natural selection sometimes works better if you 'start small'

Published on
November 21, 2017

Bedrijven die voor hun bioreactoren op zoek willen naar gisten, bacteriën of schimmels die nieuwe nuttige producten produceren, of dit doen bij minder goede groeiomstandigheden, kunnen voortaan beter niet meer automatisch kiezen voor selectie met een grote beginhoeveelheid micro-organismen. Ze zouden anders het risico kunnen lopen om de evolutionaire verbeter-mogelijkheden niet goed te benutten. Dat blijkt uit onderzoek van Wageningen University & Research dat gepubliceerd is in het wetenschappelijke tijdschrift PNAS.

Evolutionary possibilities

Companies that are searching for yeasts, bacteria or fungi for their bioreactors to produce new products, or to produce products under less optimal conditions, should not always start the selection process with a large number of microorganisms. Otherwise they could fail to fully utilise the possibilities for evolutionary improvement. This was shown recently in a study at Wageningen University & Research, which was published in the journal PNAS.

Companies that produce certain new products using bacteria, fungi or yeasts constantly endeavour to achieve the highest possible efficiency. If researchers at these companies are looking for cells with improved or new characteristics, such as the production of a specific enzyme or substance, they often experiment with strains of cells that have these characteristics. In that case, spontaneous mutation and natural selection for the fittest microorganisms is a simple and robust strategy.

Smart selecting

The obvious approach for achieving this result is to start with the largest possible number of microorganisms. Indeed, in this case it is most likely that you have genetic variants in the population that can take a leap forward. However, experiments in Wageningen have now shown that if you start with a large number of microorganisms, you run the risk of evolutionary stagnation and do not fully exploit evolutionary possibilities.

Arjan de Visser is the coordinator of the study. De Visser: “If you begin with a large number of microorganisms, you indeed have a high probability of quickly finding an improvement. The microorganism that wins the natural selection process is probably the one that has made the biggest leap forward in fitness due to a mutation. The winning microorganism will quickly outcompete other organisms in the culture that have made smaller improvements in fitness. However, natural selection is not always linear. In that case, a subsequent step towards improved fitness requires a type of microorganism with smaller improvements or even a small decline in fitness. But these cells lose the race in the cell culture.”

That risk of missing opportunities for further improvement primarily occurs when the improvement is caused by only one gene or a small number of genes. De Visser: “Mutations in these few genes can then have strong interaction effects. In such cases, it is better to choose a design with multiple, smaller populations of cells. Then you are not betting on just one horse – the one that starts the fastest – and you incur a lower risk of stagnation after a single large improvement.” 

De Visser expects that this phenomenon could also be important in plant breeding. Plant breeders, who select for characteristics such as drought resistance, would perhaps be cleverer if they did not always select from a very large number of plants.