A culture-, amplification-independent, and rapid method for identification of pathogens and antibiotic resistance profile in bovine mastitis milk

Abstract

Introduction: Rapid and accurate diagnosis of causative pathogens in mastitis would minimize the imprudent use of antibiotics and, therefore, reduce the spread of antimicrobial resistance. Whole genome sequencing offers a unique opportunity to study the microbial community and antimicrobial resistance (AMR) in mastitis. However, the complexity of milk samples and the presence of a high amount of host DNA in milk from infected udders often make this very challenging.

Methods: Here, we tested 24 bovine milk samples (18 mastitis and six non-mastitis) using four different commercial kits (Qiagens' DNeasy^® PowerFood^® Microbial, Norgens' Milk Bacterial DNA Isolation, and Molzyms' MolYsis™ Plus and Complete5) in combination with filtration, low-speed centrifugation, nuclease, and 10% bile extract of male bovine (Ox bile). Isolated DNA was quantified, checked for the presence/absence of host and pathogen using PCR and sequenced using MinION nanopore sequencing. Bioinformatics analysis was performed for taxonomic classification and antimicrobial resistance gene detection.

Results: The results showed that kits designed explicitly for bacterial DNA isolation from food and dairy matrices could not deplete/minimize host DNA. Following using MolYsis™ Complete 5 + 10% Ox bile + micrococcal nuclease combination, on average, 17% and 66.5% of reads were classified as bovine and Staphylococcus aureus reads, respectively. This combination also effectively enriched other mastitis pathogens, including Escherichia coli and Streptococcus dysgalactiae. Furthermore, using this approach, we identified important AMR genes such as Tet (A), Tet (38), fosB-Saur, and blaZ. We showed that even 40 min of the MinION run was enough for bacterial identification and detecting the first AMR gene.

Conclusion: We implemented an effective method (sensitivity of 100% and specificity of 92.3%) for host DNA removal and bacterial DNA enrichment (both gram-negative and positive) directly from bovine mastitis milk. To the best of our knowledge, this is the first culture- and amplification-independent study using nanopore-based metagenomic sequencing for real-time detection of the pathogen (within 5 hours) and the AMR profile (within 5-9 hours), in mastitis milk samples. These results provide a promising and potential future on-farm adaptable approach for better clinical management of mastitis.

Keywords: antibiotic resistance; culture-independent sequencing; mastitis; milk; nanopore sequencing technology; rapid diagnosis; staphylococcus aureus; udder infections.

FIGURE 1

An overview of MinION sequencing results following DNA extraction from mastitis milk using PF/NG kits and different combinations. (A) Total reads (black circles) and read length (blue circles) following using different commercial kits/combinations. (B) Percentage of reads aligned with the bovine genome (green circles) and Staphylococcus aureus genome (olive circles). Each circle represents a sample, and the mean value is depicted with a red line.

FIGURE 2

An overview of PCR results following DNA extraction directly from mastitis milk. PCR bands for the nuc gene in S. aureus (left) and Bos taurus mitochondrion (right) were visualized in their expected size of 65 and 879 bp, respectively. The 100 bp ladder was also used.1: Mol Com5_cent, 2: Mol Plus_cent, 3: Mol Com5_cent–nuc, 4: Mol Plus_cent–nuc, 5: Mol Com5_{cent–ox–nuc}, 6: Mol Plus_{cent–ox–nuc}.

FIGURE 3

An overview of MinION sequencing results following DNA extraction from mastitis milk using MolYsis kits and combinations. (A) Total reads (black) and read length (blue) following using different commercial kits/combinations. (B) Percentage of reads aligned with the bovine genome (green) and S. aureus genome (olive). Each circle represents a sample, and the mean value is depicted with a red line.

FIGURE 4

An overview of identified AMR genes, following MinION DNA sequencing. Two mastitis milk samples (A,B) were subjected to bacterial DNA isolation using different kits and conditions (culture-independent) combinations. The green and red cells show the presence and absence of AMR genes, respectively. 1: Mol Com5_cent, 2: Mol Com5_cent–nuc, 3: Mol Com5_{cent–ox–nuc}, 4: Mol Plus_cent, 5: Mol Plus_cent–nuc, 6: Mol Plus_{cent–ox–nuc}.

FIGURE 5

A representative of BLAST analyses for taxon assigned reads. Data was obtained following DNA sequencing for DNA isolated using the Mol Com5_cent–nuc approach (one of the replicates).

FIGURE 6

Study workflow and comparison of the amount of data needed for AMR detection following two different base calling modes. Data was obtained following DNA sequencing for the Mol Com5_cent–nuc approach (one of the replicates). (A) Timing of experiment and time/reads required for AMR gene detection following FAST base calling and using real-time data analyses. (B) Number of AMR genes detected per amount of data, following high accuracy base calling for the same sample.