Integrated multi-omic analyses of bovine milk identify biomarkers of negative energy balance

Abstract

Dairy cows are susceptible to negative energy balance, which can lead to metabolic disorders such as ketosis. Negative energy balance (NEB) often occurs in early lactation, but can also be due to food scarcity. Its quantification is difficult and prone to error, justifying the need to identify biomarkers instead. The effect of NEB on milk composition is known to be directly related to its intensity, impacting major and minor milk constituents. As such, one promising approach may be to identify non-invasive biomarkers in milk. To identify potential biomarkers of NEB, we performed an integrative multi-omic study of milk production and composition in two feed restriction trials of different lengths and intensities. Multivariate data integration using a redundancy analysis enabled an exploration of the linear relationships between variation in energy balance and milk production and composition. A highly correlated multi-omic signature of NEB was then identified using a multi-block partial least squares discriminant analysis. Early and late integration of data from the two feed restriction trials enabled the identification of a robust multi-omic panel of biomarkers of NEB. Taken together, these analyses showed that feed restrictions led to consistent decreases in milk yield, lactose content and uric acid concentration, as well as increased isocitrate and serotransferrin concentrations and differentially abundant microRNAs in both whole milk and milk fat globules. These findings are promising for the development of a panel of non-invasive biomarkers for monitoring animal energy status, and enhance our understanding of adaptations to NEB.