Development and validation of a modified broad-range 16S rDNA PCR for diagnostic purposes in clinical microbiology.

Broad-range PCR followed by sequencing identifies bacterial pathogens, even in challenging settings such as patients receiving antibiotics or infected with fastidious or non-cultivable organisms. The major problem with broad-range PCR is the risk of sample contamination. Risk is present at every step of the procedure, starting from sample collection. Contaminating bacterial DNA may be present not only in laboratory reagents but also at the surface of plastic consumables and containers used for specimen drawing and transport to the diagnostic laboratory. Contaminating DNA is amplified efficiently, leading to false-positive results. Thus, high specificity depends on eliminating such spurious targets, an awkward problem given the abundance of such targets and a highly sensitive method that detects very small numbers of molecules. Several investigators have reported strategies for eliminating the amplification of contaminating DNA sequences. So far, none of these methods has been entirely effective and reproducible. Here we describe a method that uses Exonuclease III (ExoIII) to disable contaminating sequences from acting as templates, while maintaining the high sensitivity of PCR for pathogen DNA. We use this assay in 144 clinical specimens from normally sterile sites, identifying pathogens from 24 (17%). Conventional methods identified pathogens in only four of these specimens, all of which were positive for the same pathogen by PCR. Compared with conventional methods, broad-range PCR with ExoIII pre-treatment of reagents substantially improves the diagnostic yield of bacterial pathogen identification from normally sterile sites.