Bacterial resistance to antibiotics can evolve spontaneously, as bugs divide under selective pressure from an antibiotic. But usually, resistance is a feat a bacterium borrows from its neighbors: It can spread both within and between strains via so-called conjugation. And such transfer "is the main route by which bacteria acquire resistance," Matt Redinbo said. "It's a big problem."

Redinbo is professor of chemistry, biochemistry and biophysics at the University of North Carolina in Chapel Hill and senior author of a paper in the July 24, 2007 issue of the Proceedings of the National Academy of Sciences that enlists an unexpected ally to nip bacterial conjugation - or at least its drug-resistance consequences - in the bud: bisphosphonates, which prevent the breakdown of bone by an unrelated mechanism and are used to prevent bone loss in cancer and treat osteoporosis.

The first step of plasmid transfer is fairly nonspecific: Bacteria simply sidle up to one another, and a hole opens in both bacterial walls that allows the plasmid to be transferred. Bisphosphonates target relaxase, the enzyme that starts and stops the process of plasmid transfer by cutting the plasmid's DNA for replication, and joining its ends afterward. "Relaxase is the gatekeeper, and it is also the Achilles' heel of the resistance process," Redinbo said.

The researchers began with X-ray crystallography studies of relaxase, and found that it contains two DNA binding sites, and that both sites interact with the same magnesium ion. Bisphosphonates bind competitively to the same magnesium ion, and the scientists tested 11 different bisphosphonates, including the FDA-approved drugs etidronic acid (Didronel, Procter & Gamble Pharmaceuticals Inc.), pamidronic acid (Aredia, Novartis Inc.), neridronic acid (Nerixia, Abiogen Pharma) and clodronic acid (Benefos, Aventis Pharma, or Clasteon, Oryx Pharmaceuticals).

As Redinbo and his team predicted, several of the bisphosphonates were effective in preventing the transfer of plasmids between cultured E. coli. But unexpectedly, that was not their only mechanism of action. Plasmid-containing E. coli cells also died at much lower concentrations of the bisphosphonate imidodiphosphate than plasmid-negative ones, suggesting that something about having a relaxase sensitizes E. coli to a toxic effect of the drug.

Why exactly that is the case is not yet clear, beyond the very general trend that the relaxase might be involved in bacterial DNA replication in general, and "if you screw up DNA manipulation, you tend to kill the cell," as Redinbo put it. But the results suggested that bisphosphonates could be effective either by themselves or as combination treatments. "If you could selectively kill the bugs that are already resistant and capable of making their neighbors resistant, that could be a powerful tool," Redinbo said.

To date, Redinbo and his group have done all their experiments in E. coli, so which bacterial species can be stopped in their tracks by bisphosphonates is still unclear. "We don't want people to think we have a magic bullet right now," Redinbo cautioned.

Plasmid-swapping has not been documented in all bacterial species - it's never been seen in tuberculosis bacteria, for example. But it's unclear whether those species can't take up plasmids, or whether they simply haven't yet. And plasmid-mediated resistance occurs in a wide variety of bacteria and clearly contributes to the transfer of resistance across species, including between Gram-positive and Gram-negative bacteria. Redinbo named the acquisition of vancomycin resistance by Staphylococcus aureus from enterococcus bacteria as a well-documented example.

Redinbo is scientific founder of a biotech company seeking to develop the findings. Exigent Pharmaceuticals, Inc. was founded in June and is based in Research Triangle Park, NC. Exigent has seed funding from local venture capital group Golden Pine Ventures, and is doing animal studies to determine whether bisphosphonates are effective beyond the confines of the petri dish, and the E. coli bacterium. But for now, the ink is hardly dry on the incorporation. The PNAS office decided a conflict-of-interest statement was unnecessary, because the company had not yet been founded when the paper was accepted.