With increasing antibiotic resistance and the spread of SARS-CoV-2 variants, it is becoming increasingly clear that microorganisms are constantly evolving to adapt to their environment. This process, which relies on genetic mutations, offers an exciting but worrying insight into how these organisms can change over time. Understanding the mechanisms of these adaptations represents a major challenge for medical research.
This phenomenon is closely linked to the appearance of mutations and the impact they have on the organism. When these mutations weaken the organism that hosts them, they tell us about the functions that are crucial or essential to the survival of the microorganism in its environment. Conversely, when mutations confer an advantage to the carrier over the microorganisms from the same species, they provide us with information about adaptations to adjust to local conditions, such as the adaptation to the human host observed in the case of Sars-CoV-2. Thus, adaptation depends largely on how the effects of available mutations are distributed.

The question then arises as to how the effects of these mutations change over the course of adaptation and to what extent is this process predictable?
To explore these questions, Olivier Tenaillon and his collaborators used a well-known experiment called “Long-Term Experimental Evolution,” initiated by Richard Lenski. As part of this experiment, twelve populations derived from a bacterial strain of Escherichia coli have been grown and propagated in the laboratory since 1988, each accumulating more than 70,000 generations of evolution. The researchers then carefully analyzed the evolution of the distribution of mutation effects during adaptation within this renowned experiment.

To characterize the distribution of mutation effects, the researchers used a quantitative approach involving the random insertion of a mobile genetic element into the genome. By analyzing a hundred thousand mutants, they were able to see the effects associated with the disruption of each gene and therefore construct a distribution of the effects of these mutations.
On a global scale, despite considerable adaptation, the distribution remains virtually unchanged. However, significant change occurs within the small fraction of beneficial mutations. While in the ancestor, numerous mutations increase the reproductive capacity of the bacteria by more than 5%, no more mutations with a similar effect are observed after 50,000 generations, nor surprisingly after 2,000 generations. But how are these global changes reflected on the scale of specific mutations?

To address this question, the researchers first examined the essential genes, whose function is essential for the proper functioning of the organism. Under the conditions used, 550 genes are essential. However, after 50,000 generations, 77 new genes are essential in at least one lineage and 97 are no longer essential in at least one other. The notion of essentiality therefore evolves and reveals changing functional interdependencies during adaptation.
Regarding beneficial mutations, rapid changes in the identity of the mutations involved were observed. Thus, most of the advantageous mutations in the ancestor are no longer advantageous, or even are deleterious, in a genome evolved 2,000 generations ago. These changes suggest low predictability of the process. However, most of the mutations observed in lineages during early phases of adaptation are beneficial in the ancestor. The further we advance in adaptation, the less the predictive power of the distribution of the effects of mutations is important. Thus, if the first stages of adaptation are relatively predictable from the genome of the ancestor, the subsequent adjustments become more and more unpredictable.