Screening Of Metabolic Markers For Diagnosis Of Fatty Liver Disease In Perinatal Cows

Fatty liver is a kind of metabolic disorder disease which is easy to occur in the perinatal period of dairy cows.It leads to the reduction of production and affects the subsequent reproductive performance of dairy cows,resulting in the high elimination rate after delivery,which has caused huge economic losses to the dairy industry.Due to the large size of dairy cattle,B ultrasound and other auxiliary diagnostic methods are not accurate.The lack of effective diagnostic methods has caused huge economic losses to the dairy industry.The only reliable diagnosis is liver biopsy.However,liver biopsy will further damage the health of cattle and have a negative impact on milk production.Of course,in production practice,people also use the current serum biochemical indicators(such as ast,alt,Glu,etc.)to assist the diagnosis of fatty liver in dairy cows,but there are also large inaccuracies.Therefore,it is necessary to develop reliable,economic,practical and harmless diagnostic methods in dairy production.Therefore,we intend to screen the possible diagnostic markers by using the metabolites of cow feces,urine and serum.Holstein dairy cows were liver biopsied and serum collected during perinatal period and diagnosed as normal group(fat content=13.17%±8.20%,n=8)and fatty liver diseased group(fat content=84.71%±4.83%,n=10),named as discovery set.Choose to collect blood,urine,and stool samples on an empty stomach before morning feeding one day,and perform GC-TOF / MS detection.The detection data were screened by multi-dimensional V-plot volcano map and one-dimensional T test to obtain potential biomarkers.At the same time,traditional serum biochemical indicators were tested as a control for subsequent biomarker diagnostic ability testing.Blood,urine,and feces were collected on an empty stomach in the same way,and the validation set was screened in the same way for the second round of screening.The screened markers and the potential biomarkers obtained from the discovery set were intersected to obtain the final biomarkers for verification.Other 20 dairy cows with similar biochemical,B-ultrasound and other veterinary diagnosis results indexes with diagnosed normal and/or diseased cattle,were divided into suspected normal group(n=10)and suspected diseased group(n=10),named as test set forvalidation.Firstly,candidate biomarkers were screened by one-dimensional and multi-dimensional detection methods in discovery set.Then,the same method for secondary round screening were carried out in the test group.Samples were collected through GC-TOF/MS detection and carried out on Xplore MET platform,data are processed by Xplore MET platform and SPSS software.QC quality control and principal component analysis showed that the data model of the diseased group and the normal group was successful,and the diseased group significantly differentiate from the normal group based on the volcanic map and OPLS-DA analysis.Candidate markers expressed in the diseased group and the normal group were identified in the discovery set.Violin map and ROC curve of these potential biomarkers verified their diagnosibility and their reliability The differential expression levels of these candidate markers were also validated by the test set.It was found that there were 23 potential biomarkers in feces,7 in urine and 24 in blood.After intersecting the markers obtained by screening with the test set,two biomarker in faces were finally obtained: L-α-aminobutyric acid and behenic acid;one in urine:3-nitrotyrosine;In the blood,seven were retained: L-asparagine,palmitoleic acid,L-serine,stearic acid,undecanoic acid,petrolatum,heptadecanoic acid.After the obtained biomarkers are verified by the independent sample set test set described above,the method verification is continued.The ROC curve was used to test the diagnostic capabilities of biomarkers and traditional biochemical indicators,and it was confirmed that biomarkers have significantly more diagnostic accuracy than traditional serum biochemical indicators(AST,GLU,etc.).In addition,our serum,feces,and urine metabolomics analysis shows that compared with normal cows,dairy cows with fatty liver have abnormal metabolic regulation in bile acids,amino acids,lipids and other biological processes.As we all know,these processes are closely related to the disorder of fat metabolism and fat deposition,so these abnormal metabolic regulation processes may be an intrinsic factor in the pathogenesis of fatty liver in dairy cows.It is worth mentioning that the biomarkers identified in this study are significantly related to these metabolic pathways,which provides a prospective theoretical guide for us to study the pathogenesis of fatty liver in dairy cows.In summary,significant differentially expressed molecules were identified and validated as effective faces,urine,and serum biomarkers to diagnose dairy cows with fatty liver disease.The diagnostic effect of novel identified biomarkers is much better than traditional serum biochemical indicators.It is of great significance of these biomarkers for the diagnosis and detection of fatty liver diseased dairy cows,therefore playing potential important role in production improvement and sustainable development of dairy industry.