Identification and validation of the potential early diagnostic biomarkers of infliximab non-response to treatment of rheumatoid arthritis based on autophagy / 西安交通大学学报(医学版)

ABSTRACT

【Objective】 The purpose of this study was to perform bioinformatics analysis of infliximab （IFX） treatment of rheumatoid arthritis （RA） non-responders and responders gene data chip based on autophagy， and to screen out the characteristic autophagy-related genes， carry out a preliminary experiment on the role of characteristic genes， and explore the possible mechanism of IFX treatment of RA non-response. 【Methods】 R language was used to analyze the differential expressions of autophagy-related genes on the selected gene expression omnibus database （GEO） data set. Univariate logistic regression， least absolute shrinkage and selection operator （Lasso） regression， and multiple logistic regression analyses were made to screen out the characteristic autophagy-related genes； receiver operating characteristic curve （ROC） was used to assess the diagnostic value of the characteristic autophagy-related genes. Real-time PCR was used to detect the expressions of characteristic autophagy-related genes in RA clinical samples， and Western blot was used to detect the role of characteristic autophagy-related genes in autophagy in RA synovial cell line MH7A cells. 【Results】 Part of autophagy-related genes were differentially expressed between non-responders and responders to IFX treatment of RA （P<0.05）， and autophagy-related gene sets were mainly expressed in non-responders. Hepatocyte growth factor-regulated tyrosine kinase substrate （HGS） and autophagy related 4 homolog B （ATG4B） were screened out as characteristic genes of non-response to IFX treatment of RA. The ROC showed that the predictive model constructed using the selected markers had good discrimination ability in the internal validation data set GSE78068 （AUC=0.708， P<0.05） and the external validation data set GSE58795 （AUC=0.719， P<0.05）. The Real-time PCR results showed that the expression of HGS was increased in peripheral blood leukocytes in the non-responders. After overexpression of HGS in MH7A cells， the ratio of LC3BⅡ/LC3BⅠ was increased while the expression of P62 was decreased， indicating that autophagy increased. 【Conclusion】 The autophagy-related gene expression profiles were different between IFX-treated RA non-responders and responders， suggesting that the non-response of IFX treatment of RA may be related to autophagy. The prediction model constructed with HGS and ATG4B had a certain ability to predict IFX-treated RA non-response， and HGS promoted autophagy in MH7A cells， which provides a new idea for future research on the molecular mechanism of IFX in treating RA non-response.