This research illuminates the critical importance of methodological choices, specifically the selection of DNA extraction kits, in shaping the outcomes of metagenomic analyses. In this study, we compared the performance of four extraction kits commonly used in laboratories (FastDNA spin soil kit, QIAamp DNA stool minikit,QIAamp Power Pro and DNeasy PowerSoil) for 16S microbiome studies together with a recently developed extraction kit by Origin Diagnostics and Research by employing the Illumina Miseq platform.
While a number of sequences were identifiable at genus level, substantial number of taxa were identifiable only up to family level. Previous studies have stated this observation that the microbiota contains certain proportions of unclassified bacteria at this level. The limited availability of 16S rRNA gene sequences and insufficient research on classifying animal microbiomes may account for these reasons (Chen et al. 2017). Another potential explanation for this phenomenon could stem from the existence of reads exhibiting low sequence accuracy, potentially leading to their failure to be classified into a specific genus. Despite the pipeline's high-resolution capability, the depth of 16S rRNA amplicon sequencing attained in the present study proved insufficient for precise taxonomic assignment at the species level. While many genera exhibited several distinct OTUs linked to them, discerning the unique species each OTU represented proved challenging, with only three taxa tentatively identified at the species level, representing a minor portion of the collected sequences. However, these findings offer limited insights due to concerns regarding the reliability of species-level resolution from sequencing the V3-V4 hypervariable regions of the 16S rRNA gene. Previous studies have noted instances of closely related species yielding sequences that are 100% identical, casting doubt on the accuracy of such determinations.
In terms of quantity of DNA obtained FastDNA spin soil kit provided the highest DNA concentration, likely due to its lysis efficiency. On the other hand, the lower 260/280 ratios in QIAamp DNA stool minikit and Stool DNA Kit suggest the presence of contaminants or impurities, highlighting the need for caution in kit selection to ensure high-quality DNA. The impact of DNA extraction kits on microbial richness and diversity is crucial for understanding the composition of complex microbial communities. The higher bacterial operational taxonomic unit (OTU) richness observed with QIAamp DNA stool minikit, QIAamp Power Pro, and DNeasy PowerSoil suggests these kits are more effective in preserving a broader range of microbial taxa. In contrast, the lower richness and diversity observed with Stool DNA Kit may result from specific biases in its extraction procedure, limiting its ability to capture the full microbial diversity present in the samples. The clustering observed in the PCoA analysis based on Bray-Curtis dissimilarity emphasizes that the choice of DNA extraction kit significantly influences bacterial community composition. The distinct separation of Stool DNA Kit samples from those extracted with other kits indicates the ability of the kits to retrieve bacterial communities that maybe crucial for the study. The minor differences between QIAamp DNA stool minikit, QIAamp Power Pro, and DNeasy PowerSoil suggest some similarities in their ability to preserve microbial community composition, reinforcing the importance of kit selection in shaping downstream analysis results.
The DESeq2 analysis revealed specific biases in the abundance of bacterial families associated with different DNA extraction kits. For instance, the lower abundance of Kiritimatiellae with QIAamp DNA stool minikit suggests a selectivity against this bacterial family. Stool DNA Kit, on the other hand, exhibited a unique profile with lower Prevotellaceae and Bacteroidales_RF16_group but higher Oscillospiraceae and Peptostreptococcaceae, indicating a preference for certain microbial groups. These kit-dependent biases should be considered when interpreting results related to the abundance of specific taxa.
The variation in the ability to identify spiked bacteria among different extraction kits is crucial when designing experiments with known microbial additions. Kits with bead beating steps, such as DNeasy PowerSoil and FastDNA spin soil kit, demonstrated better efficiency in extracting spiked bacteria, suggesting their suitability for accurate representation of added microbial strains. This finding is particularly relevant for studies involving targeted analyses of specific bacterial taxa.
The most prevalent bacterial groups found in feces were members of the Firmicutes and Bacteroidota phyla, similar to the findings by Young et al. It is well-established that the gastrointestinal tract of calves is colonized with a diverse microbiota before birth, which undergoes significant changes after birth, with Firmicutes, Bacteroidota, Proteobacteria, and Actinobacteriota successively becoming dominant post-weaning (Oikonomou et al. 2013; Mao et al. 2015). However, contrary to common reports, this study observed a succession from Firmicutes and Bacteroidota to Verrucomicrobiota, aligning with microbial distribution patterns reported in gut microbiota by Mtshali et al. (2022).
As noted in several prior studies, extraction kits often exhibit a tendency to favour Gram-negative bacteria, potentially leading to their overrepresentation in the resulting data. In our investigation, we observed that the DNeasy PowerSoil and FastDNA spin soil kits displayed improved Staphylococcaceae/Enterobacteriaceae ratios compared to other extraction kits. This enhancement could be attributed to the mechanical disruption steps, such as bead beating, which have been shown to efficiently lyse Gram-positive bacterial cells. However, samples extracted using the Origin kit did not yield similar results and instead exhibited an overrepresentation of Gram-negative bacteria. Neglecting key Gram-positive bacterial taxa can lead to misinterpretation of the data, particularly in studies where the balance between Gram-positive and Gram-negative bacteria is crucial for understanding microbial community dynamics or host-microbe interactions. Furthermore, overlooking Gram-positive bacteria may hinder the detection of potential pathogens or beneficial microbes, impacting the accuracy and completeness of microbiome analyses.
One possible way to mitigate extraction bias is to increase the quantity of cell lysate and extend the duration of the bead beating step. Additionally, optimizing the lysis buffer composition to enhance the lysis efficiency of both Gram-positive and Gram-negative bacteria can help alleviate bias. Moreover, validating the efficiency of the chosen extraction method by spiking samples with known quantities of bacteria from different taxonomic groups can aid in assessing and addressing potential biases.
The collective findings underscore the need for careful consideration when selecting a DNA extraction kit. Both DNA yield and quality, as well as their impact on microbial community composition, should be considered (Zielińska et al. 2017). Researchers must weigh the trade-offs between DNA quantity and potential biases introduced by different kits. Differences in bacterial community composition and efficiency in lysis may significantly impact downstream applications, particularly in metagenomic studies, where accurate representation of microbial diversity is paramount (Navgire et al. 2022). Hence, researchers should choose a kit that aligns with the goals and nuances of their specific study design.