Residual feed intake (RFI) is designed to estimate net efficiency of feed use, so low RFI animals are considered for selection to reduce feeding costs. However, metabolic profiling of cows and availability of predictive metabolic biomarkers for RFI are scarce. Therefore, this study aims to generate a better understanding of metabolic mechanisms behind low and high RFI in Jerseys and Holsteins and identify potential predictive metabolic biomarkers. Each metabolite was analyzed to reveal their associations with two RFIs in two breeds by a linear regression model. An integrative analysis of metabolomics and transcriptomics was performed to explore interactions between functionally related metabolites and genes in the created metabolite networks. We found that three main clusters were detected in the heat map and all identified fatty acids (palmitoleic, hexadecanoic, octadecanoic, heptadecanoic, and tetradecanoic acid) were grouped in a cluster. The lower cluster were all from fatty acids, including palmitoleic acid, hexadecanoic acid, octadecanoic acid, heptadecanoic acid, and tetradecanoic acid. The first component of the partial least squares-discriminant analysis (PLS-DA) explained a majority (61.5%) of variations of all metabolites. A good division between two breeds was also observed. Significant differences between low and high RFIs existed in the fatty acid group (P < 0.001). Statistical results revealed clearly significant differences between breeds; however, the association of individual metabolites (leucine, ornithine, pentadecanoic acid, and valine) with the RFI status was only marginally significant or not significant due to a lower sample size. The integrated gene-metabolite pathway analysis showed that pathway impact values were higher than those of a single metabolic pathway. Both types of pathway analyses revealed three important pathways, which were aminoacyl-tRNA biosynthesis, alanine, aspartate, and glutamate metabolism, and the citrate cycle (TCA cycle). Finally, one gene (2-hydroxyacyl-CoA lyase 1 (+HACL1)) associated with two metabolites (-α-ketoglutarate and succinic acid) were identified in the gene-metabolite interaction network. This study provided novel metabolic pathways and integrated metabolic-gene expression networks in high and low RFI Holstein and Jersey cattle, thereby providing a better understanding of novel biochemical mechanisms underlying variation in feed efficiency.
Keywords: dairy cattle; gene-metabolite network; metabolomics; residual feed intake.