Optimization of DNA extraction for advancing coral microbiota investigations

Abstract

Background: DNA-based sequencing approaches are commonly used to identify microorganisms and their genes and document trends in microbial community diversity in environmental samples. However, extraction of microbial DNA from complex environmental samples like corals can be technically challenging, and extraction methods may impart biases on microbial community structure.

Methods: We designed a two-phase study in order to propose a comprehensive and efficient method for DNA extraction from microbial cells present in corals and investigate if extraction method influences microbial community composition. During phase I, total DNA was extracted from seven coral species in a replicated experimental design using four different MO BIO Laboratories, Inc., DNA Isolation kits: PowerSoil®, PowerPlant® Pro, PowerBiofilm®, and UltraClean® Tissue & Cells (with three homogenization permutations). Technical performance of the treatments was evaluated using DNA yield and amplification efficiency of small subunit ribosomal RNA (SSU ribosomal RNA (rRNA)) genes. During phase II, potential extraction biases were examined via microbial community analysis of SSU rRNA gene sequences amplified from the most successful DNA extraction treatments.

Results: In phase I of the study, the PowerSoil® and PowerPlant® Pro extracts contained low DNA concentrations, amplified poorly, and were not investigated further. Extracts from PowerBiofilm® and UltraClean® Tissue and Cells permutations were further investigated in phase II, and analysis of sequences demonstrated that overall microbial community composition was dictated by coral species and not extraction treatment. Finer pairwise comparisons of sequences obtained from Orbicella faveolata, Orbicella annularis, and Acropora humilis corals revealed subtle differences in community composition between the treatments; PowerBiofilm®-associated sequences generally had higher microbial richness and the highest coverage of dominant microbial groups in comparison to the UltraClean® Tissue and Cells treatments, a result likely arising from using a combination of different beads during homogenization.

Conclusions: Both the PowerBiofilm® and UltraClean® Tissue and Cells treatments are appropriate for large-scale analyses of coral microbiota. However, studies interested in detecting cryptic microbial members may benefit from using the PowerBiofilm® DNA treatment because of the likely enhanced lysis efficiency of microbial cells attributed to using a variety of beads during homogenization. Consideration of the methodology involved with microbial DNA extraction is particularly important for studies investigating complex host-associated microbiota.

Keywords: Amplicon sequencing; Coral microbiota; DNA extraction; Optimization; SSU ribosomal RNA gene.

Fig. 1

Overview of experimental design. During phase I, DNA extraction treatments were performed on subdivided tissue, with efficiency of SSU gene amplification assessed using gel screening of PCR products. The green check mark and red X indicate that amplicons from the treatment were and were not chosen for sequencing, respectively. During phase II, well-performing PB, VG, PG, and VGl extracts were amplified and sequenced for microbial community analysis. PS PowerSoil, PP PowerPlant Pro, PB PowerBiofilm, VG UC Vortex Garnet, PG UC Powerlyzer Glass, VGl UC Vortex Glass

Fig. 2

Boxplots of total DNA concentrations grouped by treatment (n = 19–22 individual extractions per treatment). A and B letters differentiate significantly different groups (Tukey’s test p < 0.05). Medians are indicated by the solid black lines, and the 25 and 75% quartiles are represented by the lower and upper bounds of the box. Outliers are indicated as black circles and represent samples falling outside the 10 and 90% quartiles. PS PowerSoil, PP PowerPlant Pro, PB PowerBiofilm, VG UC Vortex Garnet, PG UC Powerlyzer Glass, VGl UC Vortex Glass

Fig. 3

Number of sequences before and after quality filtering and removal of low-quality sequences. Samples are grouped by DNA extraction treatment nested within coral species. Different letters (A, B, and C) denote statistically significant differences between species (one-tailed t test or Mann-Whitney ranked sums test, p < 0.05). PG UC Powerlyzer Glass, VGl UC Vortex Glass, PB PowerBiofilm, VG UC Vortex Garnet

Fig. 4

nMDS ordination of SSU rRNA gene sequences recovered from the different DNA extraction treatments and compared using Bray-Curtis distances for a all species, b O. faveolata, c O. annularis, and d A. humilis. In a, species groupings are designated by colors. In b–d, samples from the same coral colony are designated by number

Fig. 5

Heatmaps displaying the presence or absence of dominant MED nodes that ranked within the top 50% of the dataset for each species for a O. faveolata, b O. annularis, and c A. humilis. ‘Rep’ designates technical replicates of identical O. annularis DNA extracts tagged with different barcodes during PCR. The colors designate different colonies of that species