Membrane-sensitive bacterial DNA extractions and absolute quantitation of recovery efficiencies

Absolute quantification of waterborne pathogens is mandatory for microbiological risk assessment (MRA). Determination of the DNA recovery efficiency is an essential step before the quantitative molecular measurements, which has been largely ignored. In this study, we compared the DNA recovery efficiency and quality of five extraction methods, including two modified phenol-chloroform-based extractions with mechanical shearing and three commercial kits for the extraction of DNA from indigenous mixed-bacteria culture of river water. All of the methods gave relatively satisfying results from the pelleted sample through centrifugation. However, the commercial kits provided surprisingly low DNA yields for membrane-filtered samples because of DNA trapping and/or absorption on the membrane. Integrating with enzymatic lysis, bath sonication, phenol extraction, and alcohol precipitation achieved highest DNA yields and an acceptable DNA integrity for quantitative PCR. A plasmid containing the human GADPH gene fragment was demonstrated to be a suitable spiking control for determining the absolute DNA recovery efficiency. The unexpectedly low efficiencies of commercial kit extractions imply the significant underestimation of pathogenic bacteria in previous studies, which should gain enough concern in the area of pathogen monitoring in the future.