Background

Limiting technical variation within as well as between laboratories is imperative for reproducibility and hence for scientific progress from experimental research. Within the expanding gut microbiota research community, a plethora of methodologies are used to profile the phylogenetic composition of the intestinal ecosystem. Each step in this microbiota analysis process is subject to technical variation depending on the protocol or the materials used. For instance, comparing several protocols to extract DNA from murine feces showed striking differences in the obtained results even within the same laboratory (Ferrand et al., 2014). Therefore, it is clear that standardized protocols are needed to enable meta-analyses of multiple different studies.

For analyzing the human fecal microbiota, a large international consortium of researchers recently compared the effects of numerous technical approaches in every single step of the gut microbiota analysis pipeline in several independent laboratories (Costea et al., 2017). This study identified differences in the DNA extraction method as the biggest influence on the downstream gut microbiota analysis results. Based on the obtained DNA quality as well as on the reproducibility between different laboratories, the so-called ‘Protocol Q’ was identified as the best one and was proposed as a benchmark for extracting DNA from human feces (Costea et al., 2017).

Although such multi-centered comparative studies have not been performed for murine fecal DNA extraction protocols, we recently reported a gut microbiota profiling study in mice using a protocol similar to the Protocol Q recommended for human fecal DNA extraction (Mamantopoulos et al., 2017). Like the latter, our protocol uses a combination of mechanical bead-beating and chemical lysis with the QIAGEN QiaAmp^® Stool Kit. Indeed, it has been reported that mere chemical lysis of feces results in an underrepresentation of DNA from Gram-positive bacteria that have a thicker cell wall (Salonen et al., 2010). In contrast, both Protocol Q and our protocol detailed below are expected to result in efficient lysis of both Gram-positive and Gram-negative bacteria.