Skip Navigation

Microbes overcome natural antibiotic

November 2, 2005 By Charlotte Schubert This article courtesy of Nature News.

Might we create superbugs that resist our bodies' ancient defences?

Bacteria can develop resistance to at least one experimental drug based on naturally occurring antibiotics in sweat and mucus. The finding confounds the predictions of many experts, and sounds a cautionary note for drug developers.

In people, natural bacteria-fighting compounds kill microbes in places such as the mouth, eyes and skin. More than 800 such compounds, dubbed antimicrobial peptides, have been identified in humans, plants, frogs and other organisms.

Several drugs based on antimicrobial peptides are being developed, in part because of the rising problem of antibiotic resistance to conventional drugs such as penicillin.

Antibiotic resistance develops after exposure to antibiotics culls susceptible bacteria, leaving the most hardy, resistant ones to survive. Several experts have claimed that antimicrobial peptides are largely immune to this problem, pointing out that they still work after the long time span of human evolution. This is partly because they kill bacteria particularly fast, and partly because multiple peptides are deployed at once, confounding microbes.

If something can happen in a test tube it is very likely that it can happen in the real world
Michael Zasloff
Georgetown University Medical Center, Washington DC
Among those experts was Michael Zasloff, a professor at Georgetown University in Washington, DC. In a prominent article in 2002 he said that the development of resistance to drugs based on antimicrobial compounds was "improbable"1.

But now Zasloff has changed his mind.


After the release of his 2002 article, Zasloff was challenged to a scientific duel by evolutionary biologist Graham Bell at McGill University in Montreal. Bell argued that under certain conditions bacteria could evolve resistance to an antimicrobial peptide. He dared Zasloff to test the theory with him, and Zasloff accepted the challenge.

The collaborating researchers chose a peptide that had made it through several rounds of human testing, called pexiganan. Previous attempts to induce pexiganan resistance in common bugs such as Escherichia coli and Staphylococcus aureus had been unsuccessful.

To study whether resistance would ever evolve the researchers applied a very small amount of the peptide to separate populations of E. coli and Pseudomonas fluorescens in a test tube. The amount was too small to kill the microbes, but it was enough to cause the faster-growing, more resistant bacteria to thrive. It took a long time, but after 600-700 generations, while slowly increasing the dose, the researchers created pexiganan-resistant bacteria. They report their results in Proceedings of the Royal Society B2.

This has converted Zasloff. "If something can happen in a test tube it is very likely that it can happen in the real world," he says.

The findings could change how researchers think about pexiganan and similar drugs. "The study stops this dogma that resistance doesn't occur to these types of molecules," says Kim Brogden, a microbiologist at the University of Iowa in Iowa City.

Public good?

The spectre of resistance to antimicrobial peptides also raises a crucial question for public health, says Brogden. "If antibiotic resistance is generated than we are really opening up a big part of our defence system to attack," he says. If such resistant bacteria became widespread, maybe even minor cuts and scrapes would not heal, he says.

The researchers plan to test in mice whether the resistant strains are more likely to cause infection.

Neither Brogden nor the investigators involved in the study believe that development of drugs based on antimicrobial peptides should stop. But they say research should continue more carefully. Regulatory agencies should demand studies testing resistance, they say, and doctors using such drugs should institute procedures to minimize the risk of antibiotic resistance.


  1. Zasloff M., Nature, 415. 389 - 395 (2002).
  2. Perron G. G., et al. Proceedings of the Royal Society B, doi:10.1098/rspb.2005.3301 (2005).


Need Assistance?

If you need help or have a question please use the links below to help resolve your problem.