Effects of direct-fed microbial supplement on ruminal and plasma metabolome of early-lactation dairy cows: Untargeted metabolomics approach

We examined the effects of 2 multispecies direct-fed microbial (DFM) supplements on ruminal and plasma metabolome of early-lactation dairy cows using a high-coverage untargeted metabolomics approach. A total of 45 multiparous Holstein cows (41 +/- 7 DIM) were enrolled for the 14-d pre-experimental and 91-d experimental period and were a subset from a lactation performance study, which used 114 cows. Cows were blocked using pre-experimental energy-corrected milk yield and randomly assigned within each block to 1 of 3 treatments: (1) corn silage-based diet with no DFM supplement (control; CON), (2) basal diet top-dressed with a mixture of Lactobacillus animalis and Propionibacterium freudenreichii at 3 x 10(9) cfu/d (PRO-A), or (3) basal diet top-dressed with a mixture of L. animalis, P. freudenreichii, Bacillus subtilis, and Bacillus licheniformis at 11.8 x 10(9) cfu/d (PRO-B). The basal diet was fed ad libitum daily as a TMR at 0600 and 1200 h for a duration of 91 d. Rumen fluid and blood samples were taken on d -3, 28, 49, 70, and 91 and immediately stored at -80degree celsius. Before analysis, ruminal and plasma samples from d 28, 49, 70, and 91 were composited. An in-depth, untargeted metabolome profile of the composite rumen and plasma samples and the d -3 samples was developed by using a chemical isotope labeling/liquid chromatography-mass spectrometry (LC-MS)-based technique. Differentially abundant metabolites (taking into account fold change [FC] values and false discovery rates [FDR]) were identified with a volcano plot. In the rumen, compared with the CON diet, supplemental PRO-A increased (FC >= 1.2; FDR <= 0.05) the relative concentrations of 9 metabolites, including 2-hydroxy-2,4-pentadienoic acid, glutaric acid, quinolinic acid, and shikimic acid, and PRO-B increased relative concentrations of 16 metabolites, including 2-hydroxy-2,4-pentadienoic acid, glutaric acid, 16-hydroxypalmitic acid, and 2 propionate precursors (succinic and methylsuccinic acids). Relative to PRO-A, supplemental PRO-B increased (FC >= 1.2; FDR <= 0.05) relative rumen concentrations of 3 metabolites, 16-hydroxypalmitic acid, indole-3-carboxylic acid, and 5-aminopentanoic acid, but reduced relative rumen concentrations of 13 metabolites, including carnitine, threonic acid, and shikimic acid. Compared with the CON diet, relative concentrations of 13 plasma metabolites, including myxochelin A and glyceraldehyde, were increased (FC >= 1.2; FDR <= 0.05) by PRO-A supplementation, whereas those of 9 plasma metabolites, including 4-(2-aminophenyl)-2,4-dioxobutanoic acid, N-acetylornithine, and S-norlaudanosolin, were reduced (FC <= 0.83; FDR <= 0.05). Supplemental PRO-B increased (FC >= 1.2; FDR <= 0.05) relative concentrations of 9 plasma metabolites, including trans-o-hydroxybenzylidenepyruvic acid and 3-methylsalicylaldehyde, and reduced relative concentrations of 4 plasma metabolites, including beta-ethynylserine and kynurenine. Pathway analysis of the differentially abundant metabolites in both rumen and plasma revealed that these metabolites are involved in AA and fatty acid metabolism and have antimicrobial and immune-stimulating properties. The results of this study demonstrated that dietary supplementation with either PRO-A or PRO-B altered the plasma and ruminal metabolome. Notably, ruminal and plasma metabolites involved in the metabolism of AA and fatty acids and those with immunomodulatory properties were altered by either or both of the 2 microbial additives.