Antibiotics and Applied Evolution

The emergence of antibiotic resistance in bacteria has become a textbook example of evolution in action, as well as a major problem for medical practice. The conventional solution has been to "hit early and hit hard", attacking the bacteria with a cocktail of drugs in the hope of knocking them out before resistant strains appear. The idea is that it's harder to become resistant to several different drugs, and the best approach is supposed to be a "synergistic" combination which is more potent than either drug alone. Recent research challenges this notion, however, and even suggests that such a combination may be the least effective in the long term.

The researchers used a combination of mathematical modelling, experimental evolution, and high-throughput sequencing to study how bacteria evolve resistance to drug treatments. "Experimental evolution" means that they tracked how bacteria evolved resistance to a synergistic cocktail of antibiotics in controlled conditions in the lab. To do this, the team grew the bacteria Escherichia coli in nutrient broth with and without a synergistic pair of antibiotics. Every 24 hours, they transferred some of the bacteria to fresh broth so they could continue growing; at the same time, they froze some of the growing bacteria to store them for sequencing later. They therefore had a record of the bacteria spanning the course of the experiment. By sequencing the genome of samples from different times, the researchers could follow the bacteria's evolution over the course of the experiment.

That enabled the researchers to track the evolutionary changes in the bacteria, but how could they tell if the bacteria had evolved resistance to the drugs? Since bacteria make liquid turn cloudy, they could evaluate how well the bacteria were growing by regularly measuring how clear the broth was. The more clear the broth, the fewer bacteria it contained, meaning the drugs were working; by contrast, a cloudier broth meant the bacteria were doing well despite any drug treatments. Although the synergistic cocktail was initially the most effective, by the second day the bacteria had evolved resistance to the combination. The combined drug treatment didn't just become less effective; surprisingly, it became the least effective treatment — less potent than either of the drugs on their own!

What's going on here? Why would the drugs become worse as a combination than they were alone? How does the cocktail go from being synergistic to antagonistic? It's not just a question of dosage. While a higher dose makes the combination more effective, it also leads to even stronger antagonism between them later.

One explanation offered by the researchers is competitive release, where the resistant bacteria thrive in the multi-drug treatment because they no longer have to compete with their susceptible kin. Like all living things, bacteria compete with one another for the resources they need to grow. The multi-drug treatment is designed to hit the bacteria hard; it kills nearly all the susceptible bacteria, leaving the few that are resistant without much competition and with ample resources for growth and reproduction. This doesn't happen in the single drug treatments because they're less effective at killing the bacteria. This means that more susceptible bacteria will survive the treatment and compete with the resistant ones. The researchers dubbed the shift from synergy to antagonism in the multi-drug treatment the "smile-frown" transition.



spacer

By comparing the bacteria's genome sequence at the start and end of the experiment, the researchers were able to identify the changes responsible for the loss of synergy. They discovered that resistant bacteria had duplicated a portion of their genome which encoded several molecular pumps, allowing them to more efficiently pump the antibiotics out of themselves. The researchers then repeated the experiment in a different E. coli strain which lacked these molecular pumps. They found that the drugs remained synergistic for longer, and, although the synergy eventually disappeared, it never turned into antagonism.

It's important to remember that these experiments were conducted in vitro and things might be a bit different in the course of an actual infection. Nevertheless, they clearly show the shortcomings of using a static approach when trying to control evolving systems. Evolution is an ongoing process which we have to continually account for in our attempts to control pests and pathogens. Its outcomes aren't always intuitive, and effects like the smile-frown transition demonstrate the value of studying how these complex, dynamic systems evolve over time.

Reference
Pena-Miller R, Laehnemann D, Jansen G, Fuentes-Hernandez A, Rosenstiel P, et al. (2013) When the Most Potent Combination of Antibiotics Selects for the Greatest Bacterial Load: The Smile-Frown Transition. PLoS Biol 11(4): e1001540. doi:10.1371/journal.pbio.1001540 (Open Access)

Image credit
The images are modified from Figure 2 and 3 of the paper.

Full disclosure: I know one of the authors of this paper personally.