Screening Biomarkers For Diagnosis Of Sepsis-associated Acute Kidney Injury Based On Oxidative Lipidomics

Background: Sepsis-associated acute kidney injury(SA-AKI)is a common complication in critically ill patients in the intensive care unit(ICU)and has become a major public health problem affecting millions of patients worldwide,with its development significantly affecting patient prognosis and increasing hospital costs and length of stay.In addition to short-term mortality,SA-AKI is also associated with an increased risk of chronic kidney disease(CKD),end stage renal disease(ESRD)and long-term mortality.AKI is of great importance to improve the prognosis of patients.However,due to the complex pathophysiological mechanisms of SA-AKI,the understanding of SA-AKI at the molecular level still needs to be further enriched.The field of histology has flourished in recent years,and recent advances in quantitative and spatial mass spectrometry methods offer new opportunities to assess changes in renal metabolomics in SA-AKI.Recent studies have confirmed that lipid metabolism is significantly abnormal in patients with sepsis and that changes in patient lipid quantification and type correlate significantly with patient disease severity and prognosis.Basic studies have shown that during sepsis,a large number of inflammatory mediators are produced,promoting lipid peroxidation,and oxidized lipids,as powerful inflammatory regulators,are inextricably linked to the progression of inflammation.At the same time,excessive accumulation of oxidized phospholipids or oxidized free fatty acids may be closely associated with multi-organ dysfunction,including the kidney,during sepsis.Thus,we envision that the production of large amounts of inflammatory mediators in sepsis patients induces reprogramming of circulating oxidized lipids and promotes the development and progression of AKI.Purpose: Identification of circulating oxidized lipid characteristics and potential pathophysiological pathways in SA-AKI patients by oxidative lipidomics provides new predictive indicators and intervention targets for the diagnosis and treatment of SA-AKI.Methods: A total of 87 study subjects were included in this study,divided into AKI stage I-II group,AKI stage III group and healthy controls(HCs).Early morning fasting blood samples were collected and plasma samples from SA-AKI patients and healthy controls were analyzed for oxidized lipid substances by liquid chromatography tandem mass spectrometry(LC-MS/MS).Principal component analysis(PCA)and orthogonal partial least-squares discrimination analysis(OPLS-DA)were used to construct models based on subgroups,and then to identify The main differences in oxidized lipids between different subgroups were identified.The Fold change(FC)and Variable importance in projection(VIP)values were combined with univariate analysis to further filter out the differential oxidized lipids.Least absolute shrinkage and selection operator(LASSO)and random forest(RF)were used to approximate the variables.Logistic regression was used to construct the prediction model,and leave-one-out(LOO)method was used for cross-validation,thus avoiding overfitting.The Kyoto encyclopedia of genes and genomes(KEGG)database was used to find the metabolic pathways associated with them.SPSS 25.0 was applied to statistically analyze the clinical and demographic characteristics of the included subjects,and the R language program was used to compare whether there were any statistical differences between the above indicators.Results: 1.Serum creatinine(Scr)and urea nitrogen(BUN)were higher in SA-AKI stage III patients than in SA-AKI stage I-II patients,but partial pressure of carbon dioxide(PCO2)was significantly lower than in SA-AKI stage I-II patients.2.Total ion current(TIC)plots of the same QC sample(Quality control,QC)showed highly overlapping curves,high aggregation on PCA plots and PC1 control plots,Pearson correlation analysis showed that |r| was between 0.9996 and 0.9997,and empirical cumulative distribution function(ECDF)showed that the CV values <0.2accounted for 80% of the substances.Empirical cumulative distribution function(ECDF)showed that the substances with CV values <0.2 accounted for more than 80%,indicating that the instrument was stable throughout the experiment and the reproducibility of the experiment was good.3.Twenty-one differentially oxidized lipids from polyunsaturated fatty acids(PUFA)were identified between the SA-AKI and HCs groups.Among them,2 were up-regulated and 19 were down-regulated.The ratios of 9,10-Ep OME/9,10-Di HOME and12,13-Ep OME/12,13-Di HOME were both significantly lower in SA-AKI patients compared to the HCs group.4.Random forest and LASSO regression approaches were used for variable reduction and selection.We identified 5 intersected oxidative lipids,including5(S),12(S)-Di HETE,5-iso PGF2 VI,5,6-Di HETr E,11,12-EET,and 9,10-Di HOME.Among them,9,10-Di HOME and 5,6-Di HETr E both showed optimum performance in distinguishing SA-AKI from HCs with a same AUC of 0.98,while 5-iso PGF2 VI showed good performance in different stages of SA-AKI with AUC of 0.70.The logistic regression method was used to build a predictive model with an AUC of 0.907,a leave-one-out cross-validation with an accuracy of 0.728 and a kappa value of 0.573,indicating that the diagnostic model was able to successfully differentiate between SA-AKI and HCs at different stages with high sensitivity and specificity.5.The KEGG database was used to find the affected metabolic pathways,and the results showed that two pathways,vascular smooth muscle contraction and5-hydroxytryptaminergic synapses,were significantly enriched.Among them,the overall expression changes of vascular smooth muscle contraction pathway in SA-AKI patients were more obvious and tended to be down-regulated with the enrichment of factors,with a differential abundance(DA)score of 0.67,suggesting that down-regulation of vascular smooth muscle contraction may play an important role in the pathogenesis of SA-AKI.Conclusion: This project clarified the plasma oxidized lipid profile of SA-AKI patients and further identified potential biomarkers and predictive models with good diagnostic efficiency.These findings further enrich the pathophysiological mechanisms of SA-AKI and provide new perspectives for further in-depth exploration of biomarkers and intervention targets for SA-AKI.