In recent years, a body of pub­li­ca­tions in the micro­bi­ology field has chal­lenged all pre­vious knowl­edge of how antibi­otics kill bac­teria. “A slew of papers came out studying this phe­nom­enon, sug­gesting that there is a gen­eral mech­a­nism of killing by antibi­otics,” said Kim Lewis, Uni­ver­sity Dis­tin­guished Pro­fessor in the Depart­ment of Biology and director of Northeastern’s Antimi­cro­bial Dis­covery Center.

The stan­dard thinking at the time was that the three main classes of bac­te­ri­cidal antibi­otics each had a unique way of killing bac­te­rial cells—like spe­cial­ized assas­sins each trained in a single type of weaponry. But this new research sug­gested that all antibi­otics work the same way, by urging bac­te­rial cells to make com­pounds called reac­tive oxygen species, or ROS, which bac­teria are nat­u­rally sus­cep­tible to.

If they were right it would have been an impor­tant finding that could have changed the way we treat patients,” said Iris Keren, a senior sci­en­tist in Lewis’ lab.

And that’s exactly how sci­ence usu­ally works, said Lewis—through chal­lenges to main­stream thinking. But recent results reported by Lewis, Keren, and their research  part­ners in an article pub­lished Friday in the journal Sci­ence sug­gest that this alter­na­tive hypoth­esis doesn’t hold up. For example, even bac­teria that are inca­pable of making ROS are still vul­ner­able to antibi­otics. Fur­ther, some antibi­otics can work their fatal magic in both aer­obic and anaer­obic conditions—but reac­tive oxygen species can only form when there’s oxygen to fuel them.

We chose to do the sim­plest and most crit­ical exper­i­ment aimed at fal­si­fying this hypoth­esis,” said Lewis. “Killing by antibi­otics is unre­lated to ROS pro­duc­tion,” the authors wrote. The find­ings were cor­rob­o­rated by Uni­ver­sity of Illi­nois researchers in another study released on Friday .

The team treated bac­te­rial cul­tures with antibi­otics in both the pres­ence and absence of oxygen. Other than the gaseous envi­ron­ment, the two treat­ments were iden­tical. There was no dif­fer­ence in cell death between the two populations.

Before per­forming these exper­i­ments, Lewis’ team first looked at sig­nals of a flu­o­res­cent dye, which pre­vious researchers had used as an indi­cator for ROS levels. The team treated bac­te­rial cells with a variety of antibi­otics and mea­sured the strength of this signal. Since antibi­otics were pre­sumed to increase ROS levels, one would have expected increased con­cen­tra­tions of antibi­otics to cor­re­late with stronger sig­nals. How­ever, Lewis’ group saw no such correlation.

But there’s a dif­fer­ence between cor­re­la­tion and direct obser­va­tion,” Keren said. In order to sup­port their obser­va­tions with unequiv­ocal data, the team mem­bers phys­i­cally sep­a­rated the cells that had stronger flu­o­res­cent sig­nals from those with weak sig­nals and treated them both with the same antibi­otics. Both pop­u­la­tions suf­fered equiv­a­lent cell death.

The research from Dr. Lewis’ group demon­strates that, con­trary to cur­rent dogma, antibi­otics appar­ently do not kill bac­teria through induc­tion of reac­tive oxygen species,” said Steven Projan, vice pres­i­dent for research and devel­op­ment at iMed and head of Infec­tious Dis­eases and Vac­cines at Med­Im­mune, both sub­sidiaries of AstraZeneca. “The results shown are rather clear but still leave us with the mys­tery as to how antibac­te­rial drugs help infected people clear bac­te­rial infections. At this point, we should prob­ably dis­pense with the ‘one size fits all’ approach to bac­te­rial killing by antibi­otics,” said Projan, who was not involved in the research.

With these results, Lewis and Keren hope the field will be able to focus its efforts on under­standing the true mech­a­nisms of how antibi­otics wipe out bac­teria in order to effec­tively address chronic bac­te­rial infec­tions, one of the most pressing issues facing public health today.