I have ‘what’ in my sample?
Reagent contamination in PCR has long been documented – not only by operators but also from commercial sources long before those operators ever laid DNA-ridden hands on them. While contamination of some/many Taq polymerase enzymes with residual host E. coli DNA has had attention, little attention has been paid to the ubiquitous DNA extraction kits. Yet contamination of columns with legionella DNA was first reported back in 2002 (van der Zee et al) and confirmed with a New Zealand report a year late that had identified and characterised the same issue (Evans et al., 2003). I have heard other anecdotal reports from other kits over the years and observed them myself for various qPCR applications – using high resolution melting to differentiate the no template controls from an actual sample amplification. Comparing a range of column-based DNA extraction kits, they serially diluted a salmonella culture (not previously identified on extraction columns), extracted the dilutions and then submitted them to next generation sequencing (NGS). By using dilutions, the step-wise increase of ‘other’ bacterial species showed the contribution of the extraction process to the resulting NGS profile. As might be expected, a wide range of bacterial genera were detected with the possible source of some of them provided (e.g. nitrogen-fixing bacteria from ultrapure water tanks). Interesting to note the difference among the different manufacturer’s kits – possibly indicating the different sites of manufacture of the columns and their various water contaminants (manufacture of the columns is reported to use a large quantity of water to flush the columns) Also of note is the checklist to guard against erroneous bacterial contributions in your NGS data – such as sequencing your ‘blank extractions’ (empty DNA extractions through the same column process). While the Geneaid Presto Mini gDNA Bacteria kit (ah oh, here comes the plug) contains gamma-irradiated columns to guard against this contamination, another route would be to avoid columns altogether. Contact us if you want some suggestions or assistance. Meanwhile, let us know if you’d had some unusual species from 16S NGS data that might be attributable to columns. References: Evans et al.J. Clin. Micro. July 2003, 3452-3453 Van der Zee et al. J. Clin Micro. March 2002, 1126