Serum Metabolomics Fingerprinting during the Dry Off Period Identifies Metabolite Signatures That Can Predict the Risk of Metritis

The objective of this study was to screen transition dairy cows during the dry off period for identification of metabolite signatures in the serum that can be used for prediction of risk of metritis and give insights into the pathobiology of the disease. Blood samples were collected from coccygeal vein at -8 and -4 wks prepartum, disease diagnosis week, and +4 and +8 wks postpartum. Gas Chromatography Mass Spectrometry was used to identify and quantify 29 metabolites in the serum of 20 healthy control cows (CON) and 6 cows that were diagnosed with metritis. Data were analyzed using univariate and multivariate analysis. Results showed that 16, 12, 14, 15, and 10 metabolites were significantly altered at -8, -4, disease diagnosis, +4, and +8 wks around calving in cows diagnosed with metritis versus healthy CON. The multivariate analyses indicated consistent disease-dependent clustering with fairly similar set of metabolites distinguishing premetritis cows from healthy controls at -8 and -4 wks. The utility of these metabolites as biomarkers of risk of disease was assessed by the area under the curve (AUC) and AUC values of 1.0 and 0.969 were observed at -8 and -4 wks, respectively. Overall, results of this study indicated that selected metabolites can be used to early predict the risk of metritis in transition dairy cows. Results indicated significant (P<0.05) metabolite alterations at -8, -4 and disease diagnosis in premitritis cows and those that developed metritis as compared to healthy CON. The multivariate analyses also demonstrated consistent disease-dependent clustering with fairly similar set of metabolites distinguishing premetritis cows from healthy controls at -8 and -4 wks. Among the metabolites that distinguished premetritic cows from healthy CON at -8 wks; oxalate, ornithine, pyroglutamic acid, glutamic acid, and D-mannose were ranked as the top 5 in variable importance in the projection. A similar set of metabolites (except oxalate substituted by phosphoric acid) were ranked as the top 5 at -4 wks indicating that those top 5 metabolites can be used as predictive biomarkers at -8 and -4 wks before the incidence of postpartum metritis. Intriguingly, multiple metabolites like galactose, phosphoric acid, oleic acid, urea, and oxalate were identified to be different (P<0.5) even at +4 and +8 wks after parturition. The significant alterations of serum metabolite concentrations and disease-dependent clustering around parturition indicates the potential of these metabolites to track the progression and development of metritis in dairy cattle.