Milk metabolomics reveals potential biomarkers for early prediction of pregnancy in buffaloes having undergone artificial insemination

Abstract

This study aimed to identify potential biomarkers for early pregnancy diagnosis in buffaloes subjected to artificial insemination (AI). The study was carried out on 10 pregnant and 10 non-pregnant buffaloes that were synchronized by Ovsynch-Timed Artificial Insemination Program and have undergone the first AI. Furthermore, milk samples were individually collected ten days before AI (the start of the synchronization treatment), on the day of AI, day 7 and 18 after AI, and were analyzed by LC–MS. Statistical analysis was carried out by using Mass Profile Professional (Agilent Technologies, Santa Clara, CA, USA). Metabolomic analysis revealed the presence of several metabolites differentially expressed between pregnant and non-pregnant buffaloes. Among these, a total of five metabolites were identified by comparison with an online database and a standard compound as acetylcarnitine (3-Acetoxy-4-(trimethylammonio)butanoate), argininesuccinic acid hydrate, 5’-O-{[3-({4-[(3aminopropyl)amino]butyl}amino)propyl]carbamoyl}-2’deoxyadenosine, N-(1-Hydroxy-2-hexadecanyl)pentadecanamide, and N-[2,3Bis(dodecyloxy)propyl]-L-lysinamide). Interestingly, acetylcarnitine was dominant in milk samples collected from non-pregnant buffaloes. The results obtained from milk metabolic profile and hierarchical clustering analysis revealed significant differences between pregnant and non-pregnant buffaloes, as well as in the metabolite expression. Overall, the findings indicate the potential of milk metabolomics as a powerful tool to identify biomarkers of early pregnancy in buffalo undergoing AI.