Experimental drugs called bicyclic nitroimidazoles kill non-replicating as well as replicating bacteria, making them promising candidates for the treatment of latent tuberculosis. One such compound, PA-824, does this by a mechanism that suggests a new avenue for anti-infective drug design, report Clifton Barry of the US National Institute of Allergy and Infectious Diseases in Bethesda, Maryland, and his colleagues.
PA-824 inhibits cell division by interfering with the manufacture of bacteria's sticky outer layers. Barry and his colleagues report that it also generates nitric oxide, which is toxic regardless of whether or not a cell is dividing. PA-824 should not have this effect on mammalian cells or on all bacteria — only on bacteria such as Mycobacterium tuberculosis that produce a compound known as F420. This is needed for the drug to trigger nitric oxide release.