DNA extraction protocols cause differences in 16S rRNA amplicon sequencing efficiency but not in community profile composition or structure.
Bottom Line: These results indicate that the concentration necessary for dependable sequencing is around 10,000 copies of target DNA per microliter.Exoskeletal pulverization and tissue digestion increased the reliability of extractions, suggesting that these steps should be included in any study of insect-associated microorganisms that relies on obtaining microbial DNA from intact body segments.Although laboratory and analysis techniques should be standardized across diverse sample types as much as possible, minimal modifications such as these will increase the number of environments in which bacterial communities can be successfully studied.
Affiliation: Committee on Evolutionary Biology, University of Chicago, Chicago, Illinois; Department of Science and Education, Field Museum of Natural History, Chicago, Illinois.Show MeSH
Mentions: The effect of extraction methodology on the concentration of the bacterial 16S rRNA gene was highly significant (Repeated measures ANOVA, F = 61.5, P < 2 × 10−16; Fig. 1A). Multifactor ANOVA revealed that all other variables and several interactions were also significant (Table 2). Two-way ANOVAs on samples within each species confirmed that both extraction methodology and life stage were significantly associated with the number of 16S rRNA gene copies recovered in all species but Crematogaster rochai (Table S2). Qiagen, phenol–chloroform, and modified PowerSoil all had significantly larger numbers of 16S rRNA gene copies than the unmodified PowerSoil protocol (Bonferroni-corrected paired t-tests, P < 4.0 × 10−10; Figs. 1A, 2A, 3). Concordantly, sequencing success was significantly associated with extraction methodology (χ2-test, χ = 24.4, P = 2.0 × 10−5; Fig. 2A) and both the number of copies of the bacterial 16S rRNA gene found in each extraction (t-test, t = 14.04, P < 2.2 × 10−16; Fig. 2A) and the DNA concentration of the sample (t-test, t = 3.37, P = 0.001; Fig. 2B). In adult samples, pooling was an effective strategy for boosting 16S rRNA gene concentration as the extractions with three individuals had significantly greater quantities of bacterial DNA than single individual extractions (paired t-test, t = 4.6, P = 1.9 × 10−5; Fig. 3), but this was not the case in larvae (t = 1.45, P = 0.16; Fig. 3). In general, adult abdomens hosted significantly more bacterial DNA than whole larvae overall (paired t-test, t = 3.20, P = 0.003; Fig. 3), but not when using the unmodified PowerSoil extractions (Fig. 3). Total DNA concentration including both ant and microbe DNA was greatest in phenol–chloroform and Qiagen extractions and lowest in PowerSoil extractions (Bonferroni-corrected paired t-tests, P < 1 × 10−4; Fig. 1B). The number of copies of the bacterial 16S rRNA gene was significantly correlated with DNA concentration (Pearson's correlation, t = 6.70, P = 2.8 × 10−10; Fig. 2B), suggesting that effective extraction techniques are useful for all types of DNA present in these samples. The number of copies of the bacterial 16S rRNA gene was also correlated with the number of reads sequenced from each sample (Pearson's correlation, t = 3.88, P = 2.4 × 10−4; Fig. 2C). Different host species had significantly different ratios of bacterial 16S rRNA gene copy number to total DNA quantity (ANOVA, F = 59.21, P < 2 × 10−16), implying that overall bacterial load differs between these species (Fig. 3).
Affiliation: Committee on Evolutionary Biology, University of Chicago, Chicago, Illinois; Department of Science and Education, Field Museum of Natural History, Chicago, Illinois.