Identification And Validation Of Endoplasmic Reticulum Stress-related Potential Biomarkers Of Periodontitis

Background:Periodontitis is an inflammatory disease that occurs in tooth-supporting tissues and has become a global public health problem.Currently,the diagnosis of periodontitis relies mainly on clinical indicators,but this has a time lag and is not ideal for early diagnosis and intervention.The endoplasmic reticulum plays an important role in cellular protein synthesis,folding,and structural maturation.When the protein folding load of the endoplasmic reticulum exceeds its working capacity,the cell can suffer from endoplasmic reticulum stress(ERS),and severe ERS may trigger cellular dysfunction or even death.It has been found that ERS is involved in the pathological process of periodontitis,but its exact mechanism of action is still unclear.The number and depth of studies on ERS-related biomarkers of periodontitis are still very limited.Objective:Using bioinformatics methods to identify and validate the ERS-related potential biomarkers of periodontitis,to investigate the roles of ERS in periodontitis,to analyze the relationship between immune cell infiltration and periodontitis,as well as potential biomarkers,and to explore the correlations between potential biomarkers and periodontal clinical indicators,aiming to facilitate the development of early diagnosis techniques for periodontitis.Methods:Three datasets containing the genetic test results of healthy periodontal tissue and periodontitis tissue samples,GSE10334,GSE16134,and GSE23586,were downloaded from the GEO database.They were merged and batch corrected,and 2/3 samples were randomly assigned into the training set and 1/3 samples were assigned into the validation set.ERS-related genes were obtained from the Gene Cards database.And the differentially expressed genes(DEGs)were identified in the training set.The least absolute shrinkage and selection operator(LASSO)and support vector machine-recursive feature elimination(SVM-RFE)machine learning algorithms were used to further filter DEGs,respectively.The protein-protein interaction(PPI)network of key DEGs was constructed using the STRING online platform.Receiver operating characteristic(ROC)curves were plotted for the key DEGs and the area under the curve(AUC)was calculated to determine the diagnostic efficacy of the genes and thus identify the potential biomarkers of periodontitis.The CIBERSORT algorithm was used to assess the infiltration of immune cells in healthy periodontal tissue and periodontitis tissue samples,and to calculate the relationships between potential biomarkers and immune cells.Thirty-three periodontally healthy patients and 31 periodontitis patients obtained from the Hospital of Stomatology,Jilin University were used as study subjects,and their periodontal clinical indicators,such as plaque index(PI),gingival index(GI),probing depth(PD)and clinical attachment loss(CAL),as well as their periodontal tissue samples,were collected.The m RNA expression levels of each potential biomarker in different samples were tested by q RT-PCR,and ROC curves were plotted to validate their diagnostic efficacy for periodontitis,and the correlations between the expression levels of different genes in periodontitis were analyzed.The periodontitis tissue samples were divided into different subgroups according to the periodontal clinical indicators,and the differences in the expression levels of potential biomarkers between the subgroups were compared.And the correlations between the expression levels of potential biomarkers and the periodontal clinical indicators were also analyzed.Results:Thirty-six DEGs were obtained by differential expression analysis.Sixteen and twenty-eight genes were screened by the LASSO regression and SVM-RFE algorithms,respectively.Eleven genes(including SERPINA1,ERLEC1,VWF,DERL3,PDIA4,FOS,CXCL8,EDEM2,APOE,KDELR1,and IL6)were obtained after taking the intersection of the two and were identified as key DEGs.The PPI network constructed by proteins encoded by key DEGs was tightly linked,with numerous interactions between the nodes.SERPINA1,ERLEC1,and VWF showed well diagnostic efficacy in both the training and validation sets,being identified as the ERS-related potential biomarkers of periodontitis.According to the results of immune cell infiltration analysis,compared to healthy periodontal tissue,resting dendritic cells,resting mast cells,memory B cells,follicular helper T cells,and M1macrophages,etc.were significantly lower in periodontitis tissue samples,while plasma cells and neutrophils,etc.were significantly more.Correlation analysis of potential biomarkers with immune cells showed that the expression levels of SERPINA1 were positively correlated with M0 macrophages and neutrophils,etc.,and negatively correlated with resting dendritic cells and resting mast cells,etc.;the expression levels of ERLEC1were positively correlated with plasma cells,etc.,and negatively correlated with resting dendritic cells,follicular helper T cells,resting mast cells,and CD8~+T cells,etc.The expression levels of VWF were positively correlated with plasma cells,etc.,and negatively correlated with resting dendritic cells,resting mast cells,follicular helper T cells,etc.The results of q RT-PCR experiments on clinical samples obtained from the Hospital of Stomatology,Jilin University showed that the expression of SERPINA1,ERLEC1,and VWF m RNAs in periodontitis tissues were significantly higher than that in healthy periodontal tissue.The three potential biomarkers had good diagnostic efficacy for periodontitis,and there were positive correlations between their expression levels in periodontitis tissues.The expression levels of the three potential biomarkers in periodontitis tissue samples were not significantly different in various PLI,GI,PD and CAL subgroups,and there was no significant correlation between the three potential biomarkers and periodontal clinical indicators.Conclusion:ERS is closely related to the pathological process of periodontitis.The expression levels of the three ERS-related genes,SERPINA1,ERLEC1,and VWF,were significantly higher in periodontitis tissues than in healthy periodontal tissues.There were differences in immune cell infiltration in periodontitis tissues and healthy periodontal tissues,and there were significant correlations between several immune cell subsets and SERPINA1,ERLEC1,and VWF.SERPINA1,ERLEC1,and VWF have well diagnostic efficacy for periodontitis and can be used as potential biomarkers for early diagnosis.The expression levels of SERPINA1,ERLEC1,and VWF were not significantly correlated with periodontal indicators including PLI,GI,PD,and CAL.