Assessing the diagnostic importance of nonviable bacterial cells in respiratory infections

Abstract

Identification of bacteria in clinical samples is fundamental to combating infections. Modern molecular genetic approaches exploit nucleic acids signals from clinical samples. However, DNA-derived signals can originate from nonviable bacterial cells and, therefore, generate data that could be misinterpreted. Terminal restriction fragment length polymorphism profiling of cystic fibrosis sputum samples was combined with propidium monoazide (PMA) photo-induced cross-linking. PMA is highly membrane impermeant and is excluded from viable bacteria but readily penetrates dead cells. Exposure to a light source renders DNA in permeable cells incapable of contributing to polymerase chain reaction. PMA treatment was shown to effectively prevent dead bacteria, spiked into sputum samples, from contributing to profiles. Comparison of treated and untreated clinical samples indicated that dead bacterial cells significantly bias untreated profiles. These findings highlight the significant contribution that nonviable bacteria can make to DNA-based diagnostic analysis of clinical samples while providing a simple and effective means of avoiding such bias.