Abstract
Identification of fusobacteria from clinical specimens currently requires analysis of metabolic end products by gas-liquid chromatography in addition to certain biochemical and enzymatic tests because of the relative biochemical inactivity of these bacteria. Even the finding of pointed, thin gram-negative cells on Gram-stained slides can no longer be relied on for identification of Fusobacterium nucleatum, since at least four other species of fusobacteria have been seen to exhibit similar morphology. We examined 46 clinical isolates and six American Type Culture Collection type strains of fusobacteria by conventional methods and by the Microbial ID Systems MIDI software package for analyzing cellular fatty acid patterns measured by capillary column gas-liquid chromatography. Distinctive patterns of major fatty acids could be used to reliably identify most clinical isolates to the species level. The MIDI system identified 89% of the isolates correctly and provides an alternative to conventional methods.