Abstract
Bovine mastitis is an inflammation of the mammary gland that negatively impacts milk production, animal welfare, and is a significant economic burden to the dairy industry. Many factors, including age, stage of lactation, housing, and milking procedures can affect incidence of mastitis on dairy farms worldwide. Mastitis typically occurs in the first few weeks postpartum and is generally caused by a bacterial infection. Mastitis can be clinical (CM; presence of clinical signs such as flakes or clots in the milk and/or redness or swelling of the mammary gland) or subclinical [SCM; milk somatic cell count (SCC) ≥200,000 cells/mL and no clinical signs]. Milk components, such as lactose, protein, sodium (Na), and potassium (K) concentrations; and the ratio of Na to K (Na/K) are distinctly altered during mastitis. The overall aim of the research presented in this dissertation was to comprehensively evaluate bovine mastitis utilizing different technologies, such as 16S rRNA full-length sequencing, proteomics, metabolomics, and lipidomics, on the same milk samples collected daily (over the first 22 d postpartum) from healthy cows and cows with naturally occurring CM or SCM on four commercial dairies in southern Idaho. The first objectives of this dissertation were to use SCC, concentrations of lactose, protein, Na and K, and Na/K to identify CM (in colostrum and milk) and SCM (in milk only) in individual mammary gland quarters of cows in early lactation and to test the diagnostic ability of components for mastitis. We hypothesized that components would accurately identify CM (in colostrum and milk) and SCM (in milk only) in individual mammary gland quarters. We found that diagnostic thresholds of SCC and K concentration could distinguish colostrum produced by quarters with CM from colostrum produced by quarters without CM. Furthermore, diagnostic thresholds of Na/K and Na and K concentrations were capable of accurately identifying milk produced by quarters with CM. Diagnostic thresholds of Na/K and lactose and Na concentrations accurately identified milk produced by quarters with SCM. These components in both colostrum and milk aid in the detection of CM and SCM in individual mammary gland quarters of cows in early lactation. The second objectives were to identify differences in milk bacterial communities among samples collected from quarters with CM, quarters with SCM, and healthy quarters between two timepoints: before the matched event (1-3 d prior to first observance of clinical signs) and during the matched event (first observance of clinical signs). The milk samples collected from quarters with SCM and from healthy quarters were matched, as closely as possible based on dairy, parity, and d postpartum, to milk samples collected from quarters with CM. We hypothesized bacterial communities would not be different among milk samples collected from quarters with CM, milk samples collected from quarters with SCM, and milk samples collected from healthy quarters before the matched event. We further hypothesized bacterial communities would be different among milk samples collected from quarters with CM, milk samples collected from quarters with SCM, and milk samples collected from healthy quarters during the matched event. Finally, we hypothesized bacterial communities in milk collected from quarters with CM or SCM would exhibit distinct shifts between the two timepoints but that the bacterial communities in milk samples collected from healthy quarters would not exhibit such shifts. Staphylococcus was more abundant in milk samples collected from quarters with SCM relative to milk samples collected from healthy quarters and quarters with CM. An interesting aspect of this study was that Bifidobacterium was identified in high relative abundance across all milk samples. Accounting for environmental influences may provide important information to assist our understanding of mastitis and the milk microbiome. The third objective was to perform proteomic, metabolomic, and lipidomic analyses on milk samples collected from quarters with CM, milk samples collected from quarters with SCM, and milk samples collected from healthy quarters at two timepoints (before the matched event and during the matched event). A final objective was to identify specific proteins, metabolites, and lipids that may predict CM in early lactation, aiding in the management of mammary gland health. We hypothesized that proteins, metabolites, and lipids would be differentially abundant in milk collected from quarters with CM or SCM compared to milk collected from healthy quarters. Furthermore, we hypothesized that a few proteins, metabolites, and lipids would be early predictors of CM. We found differences in specific proteins, metabolites, and lipids in milk produced by quarters with CM or SCM compared to milk produced by healthy quarters among cows and within the same cow. Proteins and metabolites related to immune function were at higher levels in milk produced by quarters with CM or SCM relative to milk produced by healthy quarters. Milk produced by quarters with CM exhibited higher levels of lipids abundant in cell membranes and lower levels of main milk lipids relative to milk produced by healthy quarters. Several specific proteins (VPS37B subunit of ESCRT-I, transitional endoplasmic reticulum ATPase, fructose-bisphosphatase, complement C8 gamma chain, cartilage acidic protein 1, alpha-L-iduronidase, osteoclast stimulating factor 1, N-acetyl-alpha-glucosaminidase, ras-related protein Rab-18, nucleobindin 1, HRas proto-onco GTPase, ß-casein, NSF attachment protein gamma, filamin A, glycosylation-dependent cell adhesion molecule 1, transforming growth factor ß, MIA SH3 domain endoplasmic reticulum export factor 3, syntaxin-19, xanthine dehydrogenase/oxidase, and tartrate-resistant acid phosphatase type 5), metabolites (citric acid, L-leucine, L-tyrosine, palatinose, 3-indolelactic acid, L-isoleucine, L-lysine, and trans-3-hydroxyl-L-proline), and lipids (one sphingomyelin species, four phosphatidylcholine species, one triglyceride species, one phosphatidylinositol species, and one phosphatidylethanolamine species) were identified in milk via machine learning to accurately predict the occurrence of CM in individual mammary gland quarters of cows in early lactation. Overall, this research will contribute to the advancement of bovine mastitis research and commercial milk production.