Untargeted Metabolomics And Mechanism Exploration Of Incident Stroke Among Chinese Hypertensive Population

Objective Stroke is the second leading cause of death after ischemic heart disease in most western countries,but in China,stroke has already become the leading cause of death.About 1.6 million people die of stroke every year.Stroke has become a significant public health issue that causes huge economic and health burden to our residents.Hypertension is the most important risk factor for stroke,the population base of which is very large in China.Therefore,the hypertensive population is the most important target population for the primary prevention of stroke.With the development of technology,metabolomics has shown its potential in the field of cardiovascular and cerebrovascular diseases.A growing number of studies have found a significant association between plasma metabolites and hypertension or stroke.If one or a group of sensitive and specific metabolic markers can be found in hypertensive population in China to predict the occurrence of stroke,it will have a great impact on the primary prevention of stroke.The development from hypertension to stroke is a long process,during which a variety of changes may occur in the cardiovascular and cerebrovascular systems,leading to stroke.Among them,oxidative stress is considered as one of the most crucial mechanisms of stroke,and intervention in oxidative stress plays a role in the prevention or treatment of stroke.However,there are no metabolomics studies on stroke in hypertensive patients in the literature and no reports on the relevant mechanisms.The mechanism of hypertension related stroke is complex,which is different from non-hypertensive stroke.Moreover,the incidence of stroke in Chinese hypertensive patients is significantly higher than that in western populations.The existing evidence cannot fully explain this difference.Therefore,the current study first explores epidemic trend of stroke in hypertensive population from Northeast China during 2004-2016,using our established cohorts.We then perform non-targeted metabonomics through a nested case-control study taking stroke as the primary end point with the aim to find a set of plasma metabolites that associates stroke in hypertensive patients.Finally,we perform experiments in cells and animals to explore whether metabolites affect the onset of stroke through oxidative stress mechanism.Methods Part I:Using previously established two cohorts in rural areas of Liaoning province,we analyzed data from three follow-ups?2004-2008,2008-2010 and2013-2016?.New cases of stroke were collected,and age-adjusted person-year incidence of stroke was calculated.Using Cox proportional hazards model,we calculated age-adjusted hazard ratio?HR?and 95%confidence Interval?CI?of incidence of stroke to analyze the incidence trend of stroke among hypertensive population in Northeast China from 2004 to 2016.Part ?:From the eligible hypertensive patients in our previously established cohort,90 cases of hypertension related stroke were randomly selected and matched with a 1:1ratio.Controls were the eligible subjects with the same follow-up period and no stroke from the cohort.Using ultra-high performance liquid chromatography-tandem mass spectrometry,non-targeted metabolomics were performed on baseline plasma from all patients in modes of positive and negative ions.After the conversion of original data,univariate and multivariate statistical analyses were carried out.Receiver operating characteristic curve?ROC?analysis,pathway enrichment analysis and other methods were used to screen metabolites and metabolic pathways related to the incidence of hypertension related stroke.Part ?:According to the screening results of the previous part,phytosphingosine was selected for further investigation.Human umbilical vein endothelial cells?HUVEC?were treated with phytosphingosine at concentrations of 5?M,10?M and 50?M for 24h,and cell activity was detected by CCK8.Cells were treated with phytosphingosine at5?M,10?M and 50?M for 24 hours and H₂O₂ at 500?M for 12 hours.Cell activity was detected again by CCK8 and expression levels of oxidative stress marker malondialdehyde?MDA?and endoplasmic reticulum stress marker glucose regulated protein 78?GRP78?were detected by western blot.Male spontaneously hypertension rat stroke-prone?SHRSP?rats aged 8 weeks were selected for further study.A total of 16SHRSP rats were randomly divided into two groups.The experimental group was gavaged normal saline+phytosphingosine?0.01mg/g?+10%DMSO?SHRSP-PHS group,n=8?,and the control group was gavaged normal saline+10%DMSO?SHRSP-Con group,n=8?for 8 weeks.Wistar rats with matching gender and age were selected as normal control group?WKY group,n=8?.All rats were fed a Japanese high salt and low potassium diet to accelerate stroke.Systolic blood pressure was measured by caudal sleeve method.Plasma MDA and superoxide dismutase?SOD?levels at 4 and 8weeks after intervention were detected by ELISA method.Clinical stroke score of rats was evaluated and recorded,and changes in the structure of brain tissue were observed by naked eye and HE staining.Results Part I:Subjects in the current study included the cohort from 2004 to 2008?13,626 hypertensive patients?,the cohort from 2008 to 2010?13,241 hypertensive patients?,and the cohort from 2013 to 2016?4,321 hypertensive patients?.Among men,the age-standardized person-year incidence of stroke increased from 686 per 100,000 in2004 cohort to 1362.6 per 100,000 in 2008 cohort,and to 1760.4 per 100,000 in 2013cohort.Similarly,the age-standardized person-year incidence of stroke in women increased from 563.3 per 100,000 in 2004 cohort to 617.2 per 100,000 in 2008 cohort,and to 1386.3 per 100,000 in 2013 cohort.Compared with 2004,stroke incidence in male hypertensive patients increased by 79.8%?HR and 95%CI:1.798,1.427-2.264?in 2008to 2010,and increased by 162.3%?HR and 95%CI:2.623,2.013-3.418?in 2013 to 2016.Among women,the stroke incidence increased 172.5%?HR and 95%CI:2.725,2.027-3.662?in 2013 to 2016,compared with 2004.Most of the increase was due to ischemic stroke.Part ?:The proportions of age,sex and baseline blood pressure grading were consistent between the case group and the control group.There were no differences in body mass index,waist circumference,total cholesterol,triglycerides,low-density and high-density lipoprotein,fasting blood glucose,smoking and drinking rates between the two groups?all P>0.05?.The overall data quality of non-targeted metabolomics was optimal.Among the established metabolites with clear qualitative information,121metabolites were screened by the CARS method in the positive mode and 66 metabolites in the negative mode.Plasma metabolites in hypertensive patients with stroke were significantly different from those without stroke.Using selection criteria of VIP value>1of PLS-DA model and P<0.05 of univariate analysis,a total of 16 endogenously differential metabolites were screened,including 7 down-regulated metabolites?2-methyl-1-pyrroline,dl-dihydrosphingosine,phytosphingosine,PC?8:0/0:0?[U],oleoyl ethyl amide,5?-Androstan-3?-ol-17-one and glycerol?and 9 up-regulated metabolites?dl-indole-3-lactic acid,PC?15:0/0:0?[U],PC?18:3/0:0?[U],ricinoleic acid methyl ester,d-pipecolinic acid,1?-hydroxy-25,26,27-trinorvitamin D3 24-carboxylic acid,glycerol1-hexadecanoate,methyl arachidonate and tetrahydroaldosterone?.The area under the ROC curve of the model to distinguish stroke cases and controls was 0.843 with 95%CI of 0.788-0.898.Relevant metabolic pathways included sphingolipid metabolism,glycerolipid metabolism,glycerophospholipid metabolism,galactose metabolism,phenylalanine metabolism and tryptophan metabolism.Subgroup analyses suggested that the screened metabolites and metabolic pathways of ischemic stroke were partly different from that of overall stroke.Part ?:cellular experiments:phytosphingosine at concentrations of 5?M,10?M and50?M had no significant effects on HUVEC cell viability.After pretreatment with phytosphingosine,cell damage induced by H₂O₂ was significantly alleviated,and cell viability was gradually increased with the increase of concentration.MDA levels in HUVEC cells increased significantly after 12h treatment with 500?M H₂O₂.Compared with the control group,MDA production in phytosphingosine intervention groups decreased as the increase of phytosphingosine concentration?5?M,10?M and 50?M?.Meanwhile,the production of GRP78 was also significantly reduced.Animal experiment:with the increase of the rats'week age,the weight of rats in each group increased.After 8weeks of gavage administratio,blood pressures of rats in SHRSP-PHS and SHRSP-Con were 228±7 mmHg and 226±9 mmHg,respectively.There was no significant difference in blood pressure between the two SHRSP groups?P>0.05?.Four weeks after the intervention,plasma levels of MDA and SOD in the SHRSP-PHS group were not significantly different from those in the SHRSP-Con group.After 8 weeks of intervention,plasma MDA level was significantly lower and SOD level was significantly higher in the SHRSP-PHS group compared with SHRSP-Con group?both P<0.05?.After 8 weeks of follow-up,no stroke occurred in WKY group,and 100%stroke occurred in SHRSP-PHS group and SHRSP-Con group.Hemorrhage and ischemic changes of brain tissue were more frequent in the SHRSP-Con group than in the SHRSP-PHS group.Clinical stroke score was lower but not significant in the SHRSP-PHS group compared with the SHRSP-Con group.Conclusions Our study found for the first time that the incidence of stroke in hypertensive population in rural areas of Northeast China increased sharply in the past decade.The stroke incidence in male was always higher than female.The total incidence has increased by about 1.6 times compared with that of 10 years ago.The incidence of ischemic stroke has increased by about 3 times compared with one decade ago,while the incidence of hemorrhagic stroke has not increased significantly.Plasma metabolic profile in hypertensive patients with future stroke is significantly different from that of patients without stroke.A variety of metabolites and metabolic pathways may be involved in the pathogenesis of hypertension related stroke,and the combination of these metabolites can optimally predict the occurrence of hypertension related stroke.Among them,phytosphingosine can improve the brain structure and function of SHRSP rats as well as the clinical symptoms of stroke.The protective role in stroke may be related to its anti-oxidative stress effect.This study shed a new light on the prevention and mechanism exploration of hypertensive related stroke.