The serum proteome and metabolome of dairy cows clustered for body fat content from a large dairy herd

Abstract

The transition from late pregnancy to early lactation implies metabolic and endocrine changes for accomplishing the adaptation to the rapid increase of milk production. Voluntary feed intake can usually not cover the energy and nutrient requirements in the first weeks of lactation, and dairy cows thus need to mobilize body reserves, mainly from adipose tissue. The extent of this mobilization that can be assessed by recording backfat thickness (BFT), varies between animals but is commonly more pronounced in cows that are over-conditioned at calving. Over-conditioned cows are at greater risk for developing metabolic disorders, such as ketosis, and thus for compromised welfare and performance than cows of normal or lean body condition. Making use of a large dataset including also health records from a herd with 1,709 multiparous Holstein cows, the objectives of this thesis work were (1) to characterize the variation in pre-calving back fat thickness (BFT) and the subsequent BFT loss during early lactation, and to relate it to milk production, health condition, and selected blood variables, (2) to perform an untargeted metabolomics analysis for comparing the metabolome in blood serum of selected subgroups differing in body condition loss, health status and in dietary methionine (Met) supply, and (3), to undertake proteome analyses in other subgroups of animals that were either lean or over-conditioned before calving but were otherwise not differing in health status and Met supply. Animals from which serum samples and BFT records were available both at day 25 ante partum (ap) and day 30 post partum (pp) were selected (n =713) and subjected to K-means cluster analyses. Five clusters were obtained each considering the BFT-ap and the difference between BFT-ap and BFT pp (ΔBFT). The clusters were validated and the serum samples analysed for non-esterified fatty acids (NEFA), ß-hydroxybutyrate (BHB), for two adipokines, i.e., adiponectin and leptin, and for one inflammation marker (Haptoglobin). In confirmation of the literature, cows in the clusters with greater ΔBFT underwent more intense lipolysis and ketogenesis than cows with smaller ΔBFT. Cows categorized as very fat ap had lesser milk yields than other clusters. No differences in the serum metabolome at day 30 pp were detectable in cows with different ΔBFT, health status, and Met supply (n = 184). Even though the subset was further limited to fat versus lean cows (n = 30 in total) that were all healthy and did not receive supplemental Met for the proteome analysis, no differences were observed between the two groups. The findings about the classical variables recorded were largely confirmatory whereas the multivariate results from metabolomics and proteomics could not further extend the current knowledge about the relationship between body condition, fat mobilization, and metabolism.