Management of germ free animals has changed little since the beginning of the 20th century. The current upswing in their use, however, has led to interest in improved methods of screening and housing. Traditionally, germ free colonies are screened for bacterial colonization by culture and examination of Gram stained fecal samples, but some investigators have reported using PCR-based methods of microbial detection, presumably because of perceived increased sensitivity. The accuracy and detection limit for traditional compared to PCR-based screening assays are not known. The purpose of this study was to determine the limit of detection of bacterial contamination of mouse feces by aerobic and anaerobic culture, Gram stain, and qPCR, and to compare the accuracy of these tests in the context of a working germ free mouse colony. We found that the limit of detection for qPCR (approximately 10(5) cfu/g of feces) was lower than for Gram stain (approximately 10(9) cfu/g), but that all 3 assays were of similar accuracy. Bacterial culture was the most sensitive, but the least specific, and qPCR was the least sensitive and most specific. Gram stain but not qPCR detected heat-killed bacteria, indicating that bacteria in autoclaved diet are unlikely to represent a potential confounding factor for PCR screening. We conclude that as a practical matter, bacterial culture and Gram stain are adequate for screening germ free mouse colonies for bacterial contaminants, but that should low numbers of unculturable bacteria be present, they would not be detected with any of the currently available means.
Keywords: PCR; animal models; germ-free; gnotobiotic; mice; test verification.