Multi-Channel Metabolomics Analysis Identifies Novel Metabolite Biomarkers for the Early Detection of Fatty Liver Disease in Dairy Cows

Abstract

Fatty liver disease, a type of metabolic disorder, frequently occurs in dairy cows during the parturition period, causing a high culling rate and, therefore, considerable economic losses in the dairy industry owing to the lack of effective diagnostic methods. Here, metabolite biomarkers were identified and validated for the diagnosis of metabolic disorders. A total of 58 participant cows, including severe fatty liver disease and normal control groups, in the discovery set (liver biopsy tested, n = 18), test set (suspected, n = 20) and verification set (liver biopsy tested, n = 20), were strictly recruited and a sample collected for their feces, urine, and serum. Non-targeted GC-MS-based metabolomics methods were used to characterize the metabolite profiles and to screen in the discovery set. Eventually, ten novel biomarkers involved in bile acid, amino acid, and fatty acid were identified and validated in the test set. Each of them had a higher diagnostic ability than the traditional serum biochemical indicators, with an average area under the receiver operating characteristic curve of 0.830 ± 0.0439 (n = 10) versus 0.377 ± 0.182 (n = 9). Especially, combined biomarker panels via different metabolic pipelines had much better diagnostic sensitivity and specificity than every single biomarker, suggesting their powerful utilization potentiality for the early detection of fatty liver disease. Intriguingly, the serum biomarkers were confirmed perfectly in the verification set. Moreover, common biological pathways were found to be underlying the pathogenesis of fatty liver syndrome in cattle via different metabolic pipelines. These newly-discovered and non-invasive metabolic biomarkers are meaningful in reducing the high culling rate of cows and, therefore, benefit the sustainable development of the dairy industry.