Edited by: Carl James Yeoman, Montana State University, USA
Reviewed by: Pierre Germon, National Institute for Agricultural Research, France; Kenneth James Genovese, United States Department of Agriculture, USA
Specialty section: This article was submitted to Veterinary Infectious Diseases, a section of the journal Frontiers in Veterinary Science
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Mastitis is one of the most costly diseases affecting the dairy industry, and identification of the causative microorganism(s) is essential. Here, we report the use of next-generation sequencing of bacterial 16S rRNA genes for clinical mastitis diagnosis. We used 65 paired milk samples, collected from the mastitic and a contralateral healthy quarter of mastitic dairy cattle to evaluate the technique as a potential alternative to bacterial culture or targeted PCR. One large commercial dairy farm was used, with one trained veterinarian collecting the milk samples. The 16S rRNA genes were individually amplified and sequenced using the MiSeq platform. The MiSeq Reporter was used in order to analyze the obtained sequences. Cattle were categorized according to whether or not 1 of the 10 most abundant bacterial genera in the mastitic quarter exhibited an increase in relative abundance between the healthy and mastitic quarters equal to, or exceeding, twofold. We suggest that this increase in relative abundance is indicative of the genus being a causative mastitis pathogen. Well-known mastitis-causing pathogens such as
Mastitis is one of the most important diseases in dairy herds worldwide, compromising animal welfare and causing considerable economic loses (
An increase in the use of the culture-independent alternatives to identify bacterial DNA in milk samples has overcome some of the limitations of bacterial culture being rapid (results in 1–2 days), unaffected by antibiotic administration pre-sampling and having increased the sensitivity of detection of known mastitis-causing organisms, as well as enabling the investigation of potential new pathogens. Advances in next-generation sequencing allow the in-depth investigation of clinical samples’ microbiomes and determining their taxonomic composition including unculturable species (
Our group has previously used metataxonomics and described the microbial diversity in bovine mastitic and healthy milk; this was a cross-sectional study of 136 samples of mastitic milk and 20 samples of uninfected milk as defined by having a low cell count. Results were compared to results obtained by culturing (
The use of the Illumina MiSec sequencing platform and the MiSeq Reporter for sequences analysis could further decrease the cost of metataxonomic studies facilitating at the same time a speedier analysis of the obtained sequences. Here, we use a metataxonomic approach in order to identify potential clinical mastitis pathogens and further evaluate its potential uses as a clinical diagnostic tool.
The research protocol was reviewed and approved by the Cornell University Institutional Animal Care and Use Committee (protocol number 2013-0056). The methods were carried out in accordance with the approved guidelines.
The study was conducted using cows from a commercial dairy herd near Ithaca, NY, USA, milking approximately 2,800 cows. Primiparous and multiparous cows were housed separately in free-stall barns, the concrete stalls being bedded using mattresses and manure solids. Cows were fed a total mixed ration to meet or exceed the nutrient requirements of a 650 kg lactating Holstein cow producing 45 kg/day of milk containing 3.5% fat and 3.2% protein and assuming a dry matter intake of 25 kg/day (
Cows with clinical mastitis were identified using the parlor computer system, which identified those with a significant reduction in milk production; these animals were further examined, and if visual assessment of milk revealed flakes, clots, or serous milk, a sample for on-farm culture was taken by trained farm personnel and the animal moved to the hospital pen. Additionally, cows identified as having abnormal milk during routine fore stripping in the milking parlor were similarly sampled and moved to the hospital pen.
Milk samples for metataxonomic analysis were collected aseptically by a trained veterinarian, following the recommendations of the National Mastitis Council mastitis handbook, during the morning milking the day after the cows entered the hospital pen. Teat ends were cleaned with routine pre-dipping technique and disinfected with 70% ethanol, and the first streams of milk were discarded. Sixty-five cows were sampled, 10-ml milk being extracted from both the mastitic quarter and a contralateral non-mastitic quarter. The samples were transported on ice for DNA extraction.
DNA was extracted from each collected sample separately. Also, 10 ml of milk was centrifuged at 4°C and 9,000 rpm for 30 min. The fat and majority of supernatant were removed by suction and 300 μl supernatant retained to resuspend the pellet. The milk pellet and the remaining supernatant were vortexed and transferred to a sterile micro centrifuge tube using a sterile transfer pipette, before being incubated at 40°C for 12 h with 180 μl of tissue lysis buffer ATL (Qiagen, Valencia, CA, USA), 40 μl of proteinase K (IBI Scientific), and 20 μl of lysozyme solution (10 mg/ml) to maximize bacterial DNA extraction.
Isolation of genomic DNA was performed on 250 μl of post-incubation mixture pipetted into PowerBead Tubes (PowerSoil® DNA Isolation kit, MO BIO Laboratories, Inc., Carlsbad, CA, USA) and settled in a Mini-Beadbeater-8 (Biospec Products, Battersville, OK, USA) for microbial cell disruption. DNA extraction was performed using a PowerSoil DNA Isolation Kit (MO BIO Laboratory Inc.) following the manufacturer’s recommendation. DNA concentration and purity were evaluated by optical density using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Rockland, DE, USA) at wavelengths of 230, 260, and 280 nm.
For amplification of the V4 hypervariable region of the bacterial 16S rRNA gene, primers 515F and 806R were used according to a previously described method (
Amplicon aliquots were standardized to the same concentration and then pooled. Final equimolar libraries were sequenced using the MiSeq reagent kit V2 for 300 cycles on the MiSeq platform (Illumina, Inc., San Diego, CA, USA). Gene sequences were processed using the 16S Metagenomics workflow in the MiSeq Reporter analysis software version 2.5 based on quality scores generated by real-time analysis during the sequencing run. Quality-filtered indexed reads were demultiplexed for generation of individual FASTQ files and aligned using the banded Smith–Waterman method of the Illumina-curated version of the Greengenes database for taxonomic classification of milk microbes. The output of this workflow was a classification of reads at multiple taxonomic levels: kingdom, phylum, class, order, family, genus, and species. To calculate relative abundance, we divided the number of sequences belonging to a specific species by the total number of sequences obtained from the specific sample. The same was done with information obtained at the bacterial genus (instead of species) level.
The 10 most abundant bacterial species in each mastitic quarter were identified. The increase in relative abundance of these bacteria in the mastitic quarter, comparing to the healthy one was calculated (dividing the relative abundance in the mastitis quarter by the relative abundance in the healthy one). A minimum twofold increase in relative abundance was taken to indicate probable pathogenicity. Subsequently, the relative abundances in healthy and mastitic quarters of the bacteria identified as potential pathogens were compared with the use of the non-parametric Wilcoxon exact test. This was not done for putative pathogens that were only identified in one mastitis case.
In 53 of the 65 sampled cattle (81%), we were able to identify a bacterial species among the 10 most abundant in the mastitic quarter that had a relative abundance at least double that of itself in the healthy quarter. Results regarding these 53 cows are presented in Table
Species | Healthy quarter | Mastitic quarter | ||
---|---|---|---|---|
23 | 0.23 ± 0.09 | 31.93 ± 5.81 | <0.0001 | |
4 | 0.011 ± 0.0016 | 17.39 ± 8.56 | 0.01 | |
3 | 0.003 ± 0.003 | 2.10 ± 0.55 | 0.049 | |
2 | 0.01 ± 0.003 | 9.03 ± 7.73 | 0.17 | |
3 | 4.96 ± 3.01 | 11.35 ± 3.62 | 0.10 | |
2 | 0.01 ± 0.003 | 12.64 ± 6.72 | 0.16 | |
2 | 0.01 ± 0.006 | 7.60 ± 4.12 | 0.16 | |
4 | 1.01 ± 0.37 | 4.83 ± 1.69 | 0.01 | |
2 | 0.06 ± 0.03 | 35.77 ± 32.74 | 0.16 | |
1 | 0.11 | 13.91 | ||
1 | 0.25 | 2.92 | ||
1 | 0.003 | 1.73 | ||
1 | 0.02 | 7.17 | ||
1 | 0.34 | 1.32 | ||
1 | 2.65 | 8.05 | ||
1 | 1.1 | 2.29 | ||
1 | 0.33 | 2.65 |
The most prevalent bacterial genus was
If it is accepted that an increase in bacterial sequences abundance between a healthy quarter and one which is mastitic indicates pathogenicity, then most of the cows in our study exhibited increases such that the case of mastitis could be attributed to specific bacteria. We used a metataxonomic approach not in order to conduct a study on the bovine milk microbiome in health and disease as we and other research groups have done previously (
Admittedly, more research is warranted before our approach is considered as an alternative for cattle mastitis diagnostics. Additionally, certain limitations do have to be considered here. Using a 16S rRNA approach, we were only able to describe bacterial populations. Any yeast- or fungus-related mastitis would not be detected. There is also the chance that such a mastitis pathogen would have caused a disturbance to the mastitic quarter microbiome leading to differences between the mastitic and the healthy quarter and potential false positives. Inclusion of 18S rRNA sequencing can in the future alleviate this problem. Viral mastitis is also not considered here, but this is a common problem for all the diagnostic methods currently employed for every day bovine mastitis diagnostics.
The most commonly identified bacterium here was
Both CNS and coagulase positive
DNA sequencing used in this study also identified bacteria not yet acknowledged as mastitis pathogens, but present in this study at abundances, which warrant further investigation into their significance. In two study cows,
Several bacterial genera are difficult to identify quickly by culture presenting circumstances in which genomic techniques could be advantageous.
Other bacteria were identified in the study at low abundances, demonstrating an increase in relative abundance between healthy and mastitic quarters and/or being of unknown significance with regard to mastitis.
Mastitic quarters in 12 cattle were not associated with a causative bacterium for which there are several possible explanations: some bacteria, e.g.,
Admittedly, there are still some limitations to affordable metataxonomic sequencing. However, DNA sequencing technology has advanced at an incredible pace in recent years, leading to astonishing decreases in sequencing cost: at the scale of the whole human genome, the price per megabase has decreased by nearly an order of magnitude per year since 2001 (
Our metataxonomic approach enabled 80% of samples to be associated with a potential mastitis pathogen and identified lesser known pathogens, including at least one organism that may subsequently prove to be associated with mastitis in cattle (
JO analyzed data and wrote the manuscript draft. EG conducted the field study and the laboratory work and also critically revised the manuscript. SB assisted in data analysis and writing of the manuscript draft. RB conceived the study and critically revised the manuscript. GO corresponding author; conceived the study, assisted data analysis, and critically revised the manuscript. All authors approved the final version of the paper and agreed to be accountable for all aspects of the work.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors thank the owners and staff of Sunnyside Dairy (Venice Center, NY, USA) for allowing use of their cows and facilities and for their assistance during experimental procedures. GO is gratefully acknowledging support from the Wellcome Trust (Wellcome Trust ISSF non-clinical fellowship). Preliminary results were presented as an Abstract at the Annual Congress of the British Cattle Veterinary Association, Southport, October 2015.
This project was supported by Agriculture and Food Research Initiative Competitive Grant No 2013-67015-21233 from the USDA National Institute of Food and Agriculture. The authors have no conflict of interest to declare.
The Supplementary Material for this article can be found online at
The research materials supporting this publication can be accessed by contacting the corresponding author.