DNA extraction method influences human milk bacterial profiles.
Author(s): Cheema AS, Stinson LF, Lai CT, Geddes DT, Payne MS
Publication: J Appl Microbiol, 2021, Vol. 130, Page 142-156
PubMed ID: 32654260 PubMed Review Paper? No
Purpose of Paper
This paper compared the DNA concentration, source profile, and bacterial 16S rRNA next-generation sequencing (NGS) data of DNA that was extracted from breast milk using four different kits.
Conclusion of Paper
Only, the Qiagen MagAttract Microbial DNA Isolation (QM) Kit and the Norgen Milk Bacterial DNA Isolation (NM) Kit yielded quantifiable DNA from all eleven breast milk specimens. A trend toward higher DNA yield with the NM kit compared to the QM kit was observed. The percentage of DNA that was human (β-globin) was significantly higher when extraction was with the NM kit than the QM kit. While the assumed yield (subtraction) of bacterial DNA was comparable between kits, there was more amplifiable 16S rRNA when extraction was with the NM than the QM kit. Overall, the bacterial profiles of the extracts were comparable at the phyla level. While more species were detected following extraction with NM than QM kit, differences were limited and could be explained by contamination, which differed between kits. In hierarchical clustering, only four of the eleven specimen pairs clustered based on volunteer source.
Studies
-
Study Purpose
This study compared the DNA concentration, source profile, and bacterial 16S rRNA sequencing data of DNA that was extracted from breast milk using four different kits. Milk was collected into falcon tubes from a single breast of each of eleven healthy breastfeeding mothers using an electric breast pump. The milk was transported on ice, aliquoted, and frozen at -80°C. DNA was extracted from whole milk using the TRIzol LS Reagent Kt and from the pellet following centrifugation at 40,000 g for 5 min using the Qiagen MagAttract Microbial DNA Isolation Kit, Norgen Milk Bacterial DNA Isolation Kit, and Qiagen MagAttract Microbiome DNA/RNA Isolation Kit. The manufacturer’s recommended elution volume was 100 µL for all kits except the TRIzol LS Reagent, which used 300 µL. DNA was quantified using the Qubit dsDNA High Sensitivity Assay. 16S rRNA and human (β-globin) DNA were quantified by real-time PCR. The presence of PCR inhibitors was investigated by real-time PCR using DNA extracts that were spiked with Streptococcus agalactiae DNA. 16S rRNA DNA was amplified and purified using NucleoMag NGS magnetic beads and then used to construct Pacific Biosciences SMRTbell Libraries, which were sequenced on a X. Data were analyzed using the Greenfield Hybrid Analysis Pipeline.
Summary of Findings:
While quantifiable DNA was extracted from the pellet of all eleven breast milk samples using the Qiagen MagAttract Microbial DNA Isolation (QM) Kit or the Norgen Milk Bacterial DNA Isolation (NM) Kit, only two of eleven specimens yielded quantifiable DNA using the Qiagen MagAttract Microbiome DNA/RNA Isolation Kit or the TRIzol LS Reagent Kit. The QM kit also yielded detectable DNA from the negative control (water), but 6S rRNA DNA was not detected in the negative control extractions regardless of extraction method. The average yield of DNA (fluorometer) was slightly, but not significantly, higher using the NM kit than the QM kit (0.68±1.51 ng/µL versus 0.55±1.09 ng/µL, respectively). The percentage of DNA that was human (β-globin) was significantly higher when extraction was with the NM kit than the QM kit (44.09%±6.19% versus 28.40%±18.18%, P=0.035). When the amount of bacterial DNA was quantified by subtracting the amount of human DNA from total DNA, the amounts were comparable between the two kits (0.37 ng/µL versus 0.36 ng/µL), but the CT values for 16S rRNA were higher when DNA was extracted with the QM than the NM kit (29.2±09 versus 28.4 ± 2.10, P=0.017). PCR inhibition was comparable between the QM and NM kit extracts (12.6 versus 12.7%). Firmicutes, Actinobacteria, and Proteobacteria were detected in DNA extracted from breast milk using both kits, and the abundance of each was similar between extraction methods (74.92% versus 61.41%, 20.13% versus 17.96% and 4.73% versus 20.52%, respectively). A total of 20 bacterial species were detected when extraction was with the QM kit, but in addition to those 20, 6 more species were detected when extraction was with the NM kit. For the majority of species, no significant differences were found, but R. glycinis was more abundant when extraction was with the NM than the QM kit (P=0-00002); however, the authors state this species was also detected in the negative control when extraction was with NM kit, raising the possibility that this difference was due to contamination. Contamination detected in DNA isolated with the QM kit was from a single species, but DNA isolated using the NM kit mapped to six species of bacteria. Some differences between kits were also observed in the levels of bacterial species between matched specimens, but they did not reach significance in univariate analysis. In hierarchical clustering, four specimens clustered based on volunteer source, but the others did not.
Biospecimens
Preservative Types
- Frozen
Diagnoses:
- Normal
Platform:
Analyte Technology Platform DNA Fluorometry Cell count/volume Next generation sequencing DNA Next generation sequencing DNA Real-time qPCR Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Real-time qPCR Specific Targeted nucleic acid B-globin
16S rRNA
Analyte Extraction and Purification Analyte isolation method Qiagen MagAttract Microbial DNA Isolation Kit (QM)
Norgen Milk Bacterial DNA Isolation Kit (NM)
Qiagen MagAttract Microbiome DNA/RNA Isolation Kit
TRIzol LS Reagent