Comparative Evaluation of DNA Extraction Methods from Feces of Multiple Host Species for Downstream Next-Generation Sequencing

Abstract

The gastrointestinal tract contains a vast community of microbes that to this day remain largely unculturable, making studies in this area challenging. With the newly affordable advanced sequencing technology, important breakthroughs in this exciting field are now possible. However, standardized methods of sample collection, handling, and DNA extraction have yet to be determined. To help address this, we investigated the use of 5 common DNA extraction methods on fecal samples from 5 different species. Our data show that the method of DNA extraction impacts DNA concentration and purity, successful NGS amplification, and influences microbial communities seen in NGS output dependent on the species of fecal sample and the DNA extraction method used. These data highlight the importance of careful consideration of DNA extraction method used when designing and interpreting data from cross species studies.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. Fecal DNA extraction efficiency varies dependent on extraction method and host species.

Mean total amount (± standard deviation) of DNA extracted from zebrafish gastrointestinal tract (A) or a standardized mass of feces from mice (B), cats (C), dogs (D), or horses (E), using four commercially available DNA extraction kits and one manual extraction procedure (isopropanol). n = 40 individual zebra fish, and 8 individuals for mouse, cat, dog, and horse with 8 samples used for all extraction methods. Samples were extracted and total DNA was measured by fluorometry. Statistical significance determined using one way ANOVA with Student Newman-Keuls post hoc test. Significance defined by p≤0.05 and denoted by like lower case letters, i.e., samples marked with the same letter are significantly different. Whiskers denoting lower standard deviation are cropped at zero.

Fig 2. Receiver operator curves of 260/280 and 260/230 nm absorbance for all DNA extraction methods for each animal species.

Absorbance values of all DNA samples from each species (n = 40) were plotted in an ROC generated by Sigma-Plot.

Fig 3. Comparison of DNA extraction method on next generation sequencing (NGS) relative abundance at the phylum level.

Gray bars represent samples that resulted in sequencing below 10,000 reads. n = 8 samples per extraction method.

Fig 4. Comparison of DNA extraction method on next generation sequencing (NGS) relative abundance at the family level.

Gray bars represent samples that resulted in sequencing below 10,000 reads. n = 8 samples per extraction method.

Fig 5. Principal Component Analysis (PCA) of samples with successful amplification and sequencing in at least half (4/8) samples.

Colors denote extraction method: DNeasy (blue), PowerFecal (pink), CadorPathogen (purple), QIAmp Stool (green), and Isopropanol (brown). Numbers denote individual animal samples tracked across all kits tested for that species

Fig 6. Diversity of fecal microbiota.

Chao1 estimate of microbial diversity plotted by Tukey box and whisker graph. For zebrafish and horse samples, statistical significance was determined using student’s t-test. For mouse, cat, and dog samples, statistical significance was determined using ANOVA with Student Newman Keuls post hoc test. Statistical significance defined by p≤0.05 and denoted in the figure by lower case letters.