News :: story cache

Game theory in the popular press.

The line of least resistance - from the print edition

Understanding the biochemistry of bacteria will lead to more relaxed bacteria—and healthier humans

3 May 2001
text is a cache of

GENETICISTS have long promised that their science will bring a revolution to medicine. Yet like all revolutions, this one has had its victims. It has been built on the corpses of legions of bacteria that have perished in the course of decades of research. Now other bacteria may reap the benefits of this sacrifice. Using information gleaned from studies of bacterial biology, researchers are designing medicines that will cure people, while also giving bacteria a bit of a rest.

That may sound odd: surely the point of medicine is to kill bacteria, not to cultivate them? But trying to exterminate bacteria has had nasty consequences. An antibiotic kills the weakest specimens in a population. Those that are resistant to the drug survive and resume breeding. Over time, the resistant strains outnumber the susceptible ones—and the antibiotic becomes useless.

Worries about antibiotic resistance now loom large. Last year, America’s Food and Drug Administration approved Zyvox, a drug that introduced a new class of antibiotics to patients for the first time in 25 years. But in April John Quinn, of the University of Illinois at Chicago, and his colleagues reported in the Lancet that some people had developed infections resistant to the new drug after using it for only three weeks.

In March, America’s Centres for Disease Control published a new set of guidelines for controlling and reducing the dosage of antibiotics in patients. Agricultural use is a problem, too. According to the Union of Concerned Scientists, based in Cambridge, Massachusetts, around 70% of the antibiotics made in America go directly to farm animals, because dosed beasts grow larger. The hordes of antibacterial soaps and detergents in the shops also increase the pressure on wild bacteria to evolve resistance.

This state of affairs has a familiar ring to economists, who know it as the “tragedy of the commons”. In the short term, each group—of doctors, farmers or vigilant housekeepers—overuses a common resource, to the detriment of all in the long term. The solution could lie in exploiting another idea beloved of economists, game theory, and tailoring it to the constraints imposed by natural selection. The idea is to slow the arms race between antibiotics and bacterial evolution, either by interfering with bacterial mechanisms of resistance or by suppressing them entirely.

Acting defensive

When a bacterium detects a dangerous chemical, it mounts a host of responses. One of the most important is to chew up the toxin with custom-made resistance enzymes. The natural precision of these enzymes has been a boon to medicine makers over the decades: chemists have been able to generate new varieties of antibiotic by tweaking the design of existing compounds just enough to fool the enzymes. If the resistance enzyme cannot recognise and destroy the new variety, the drug can do its work unhindered.

Natural selection, however, soon catches up. This has prompted researchers to look for ways to interfere with the actions of the bacterial enzymes themselves. Gerard Wright and his colleagues at McMaster University in Ontario, Canada, found that some resistance enzymes bear a resemblance to a family of molecules known as the protein kinases. Because protein kinases seem to be involved in a variety of disorders, pharmaceutical and biotechnology companies have been looking into their structures for years. The resemblance between the two groups of compounds means that inhibitors of protein kinases also inhibit bacterial resistance enzymes. Dr Wright is now trying to find a way to reverse bacterial resistance by modifying one of these protein-kinase inhibitors.

Bacteria also safeguard themselves from toxins by turning on an “efflux” system, a form of cellular garbage-disposal that ejects any offending substance without further ado. The efflux mechanism is a molecule bound to a bacterium’s outer membrane. It locks on to the offending toxin and ejects it through the membrane. Some species of bacteria have several types of efflux system. Microcide, a firm based in Mountain View, California, has found a compound that attacks three of these systems in Pseudomonas aeruginosa. As hoped, this compound augmented the potency of antibiotics in mice infected with this pathogen.

Eventually, bacteria would evolve around such gimmicks, just as they evolved around antibiotics. The only way to stop this evolution is to neutralise the threat they pose without killing them too quickly in the process. That would slow down the arms race between the bacteria and the drug makers, and Michael Alekshun and Stuart Levy of Paratek Pharmaceuticals in Boston, Massachusetts, think they have found a way to do it. They have identified a regulon (a collection of genes whose expression is regulated by a single protein) in the genome of Escherichia coli. This regulon controls the bacteria’s defences against antibiotics.

When E. coli senses a dangerous chemical, a protein called MarA activates this regulon, which is known as Mar because its activation confers “multiple antibiotic resistance”. Mar starts up the cell’s efflux system, and also stops the cell from allowing any more threatening molecules in by halting the production of porin, a membrane protein that acts as a channel into the cell. Once the threat subsides, the MarR (for “Mar repressor”) protein turns off the Mar regulon, and the cell returns to its normal state.

To disguise an antibiotic attack from a bacterium, all that is needed is an increased concentration of MarR and a lowered concentration of MarA. This month, at a meeting of the American Society of Microbiology in Orlando, Florida, Dr Alekshun and Dr Levy will unveil the crystal structure of the MarR protein, a discovery that makes it easier to find molecules that will interact with it. They have started the hunt for molecules that will alter its function, and are also analysing a set of substances that inactivate MarA.

Initially, the researchers saw controlling the Mar regulon as a means to increase or restore the potency of existing antibiotics. That would be good, but would almost certainly result in the evolution of resistance in due course. Further experiments, though, produced an unexpected result. E. coli without MarA do not form communities.

Usually, as bacteria float past a congenial surface, they adhere to it and form a mass of accumulated layers called a “biofilm”. In time, they produce a sturdy sugary coat that guards the biofilm’s tenants from antibiotics. Infections are often the result of biofilms forming on soft tissues. But in the Petri dish, bacteria without MarA did not form biofilms. Dr Alekshun and Dr Levy believe that the Mar regulon must also control some important process related to biofilm formation.

If the phenomenon occurs in bodies, as well as glassware, then inactivating MarA would stop infections forming. Bacteria could not gain a foothold, and the host’s immune system could simply flush them out of the body. Antibiotics could then be used more sparingly. By the same token, bacteria could stop racing to improve as well: because a Mar-based drug would render bacteria harmless but would not kill them, it would not impose a strong selection pressure. Just as game theory suggests, a compromise that reduces the damage done by both sides can work to their mutual benefit. Sometimes mercy is more than its own reward—even when it is shown to germs.

Copyright The Economist Newspaper Limited 2002. All rights reserved.