Serum Non-targeted Metabolomics Study About Spontaneously Hypertensive Intracerebral Microbleeds Mice

Background:Hypertensive cerebral microbleeds（HCMB）may be the early stage of hypertensive intracerebral hemorrhage（HICH）.Once it develops into HICH,the high mortality and disability rate will be seriously threatens human health.Therefore,if HCMB can be found and intervened in time,it may will help to improve the prognosis of patients.However,the early clinical symptoms of HCMB may not be significant and hard to check,so it is difficult to achieve the purpose of preventing HICH through early diagnosis and intervention of HCMB.Therefore,this study aims to analyze the serum non-targeted metabolomic characteristics of HCMB mice by liquid chromatography-mass spectrometry（LC-MS）,to explore the early biomarkers of HCMB and analyze its potential pathophysiological mechanism,to lay a foundation for the prevention of HICH.Methods:In this study,60 8-month-old C57 male mice were randomly divided into three groups.HCMB（n=40）group,hypertension（HTN,n=10）group and healthy control（CON,n=10）group were induced by angiotensin II,L-nitroarginiemethylester and saline.The construction status of the model was determined by 3.0T magnetic resonance imaging,hematoxylin-eosin and diaminobenzidine staining.Blood was collected through orbit,and the serum（6/group）was analyzed by non-targeted LC-MS.Orthogonal least partial square discriminant analysis（OPLS-DA）were used to verify the stability of the model between metabolite expression level and sample category stability of the model.According to the variable importance in the projection（VIP>1）obtained by the OPLS-DA model and the P-value of the t-test（P<0.05）combined with the nonparametric rank-sum test,the significant metabolites of HCMB,HTN and CON were screened.Besides,fold change（FC）was used to judge the expression of metabolites.Then,the receiver operating characteristic curve（ROC）was drawn by SPSS 26.0 to further screen the differential metabolites of HCMB with the area under the curve（AUC）>0.85 and p-value less than0.05.Finally,the KEGG database was used for multivariate statistical analysis such as enrichment pathway and functional annotation of differential metabolites.Results:（1）Model construction:compared with HTN and CON groups,the increase of blood pressure and the average precipitation under body weight in HCMB group were significantly changed（P<0.05）.Signs of cerebral microbleeds were observed in 3.0T magnetic resonance imaging,hematoxylin eosin staining and diaminobenzidine staining.It was confirmed that HCMB models were successfully established.（2）Testing of metabolites and disease classification:OPLS-DA permutation testing showed that R²Y_（+）=0.998,Q²_（+）=0.828,R²Y_（-）=0.983and Q²_（-）=0.845 in the group of HCMB and HTN;R²Y_（+）=0.992,Q²_（+）=0.829,R²Y_（-）=0.991 and Q²_（-）=0.901 in the group of HCMB and CON;R²Y_（+）=0.985,Q²_（+）=0.683,R²Y_（-）=0.999 and Q²_（-）=0.801 in the group of HTN and CON.The R²Y and Q²of the three groups were all greater than 0.5 and close to 1,suggesting that the model fit between the metabolite expression level and sample category was high and the metabolite had a good predictive ability for disease.（3）Screening biomarkers:Compared with HTN and CON groups,a total of 93 different metabolites were found in HCMB（P<0.05,VIP>1）.Through ROC curve verification,it was found that the AUC of 8significantlyup-regulateddifferentialmetaboliteswere3-formyl-6-hydroxyindole,isoalanine,citrulline,2-methylthiazolidine,glycine valine tetrapeptide,5,7-macrodiene-9-alcohol glucose,sphingosine phosphate,γ-glutamyluidine and 6 significantly down regulated differential metabolites such as octanoylglucuronide,（+/-）-hexyl carnitine,Lyso PC（18:3（6Z,9Z,12Z））,S-Methyl hexanethioate,N-monomethyl rizatriptan and 3-hydroxycetylcarnitine was greater than0.85 and P<0.01,which can be used as a potential biomarker of HCMB.Besides,after multivariate statistical analysis,it was found that arginine metabolism（P<0.05）and purine metabolism（P<0.05）in HCMB group were significant differences in metabolic pathways.Combined with arginine metabolism and purine metabolism,we found that the differential metabolites of L-guanine,citrulline,hypoxanthine,guanine and adenine were the differential metabolites affecting arginine metabolism and purine metabolic pathway respectively.The AUC of citrulline was>0.85,P<0.01,suggesting that citrulline might be the most important potential biomarker of HCMB.Conclusion:（1）z The HCMB mouse model was successfully constructed,which can highly reproduce the pathological process of clinical spontaneous HCMB.（2）z Ninety-three potential biomarkers suggested the occurrence of HCMB,among which citrulline was the most important potential biomarker.