Crucial Role of Foxp3 Gene Expression and Mutation in Systemic Lupus Erythematosus, Inferred from Computational and Experimental Approaches

Abstract

The impaired suppressive function of regulatory T cells is well-understood in systemic lupus erythematosus. This is likely due to changes in Foxp3 expression that are crucial for regulatory T-cell stability and function. There are a few reports on the correlation between the Foxp3 altered expression level and single-nucleotide polymorphisms within the Foxp3 locus. Moreover, some studies showed the importance of Foxp3 expression in the same diseases. Therefore, to explore the possible effects of single-nucleotide polymorphisms, here, we evaluated the association of IVS9+459/rs2280883 (T>C) and −2383/rs3761549 (C>T) Foxp3 polymorphisms with systemic lupus erythematosus. Moreover, through machine-learning and deep-learning methods, we assessed the connection of the expression level of the gene with the disease. Single-nucleotide polymorphisms of Foxp3 (IVS9+459/rs2280883 (T>C) and −2383/rs3761549 (C>T)) were, respectively, genotyped using allele-specific PCR and direct sequencing and polymerase chain reaction-restriction fragment length polymorphism, in 199 systemic lupus erythematosus patients and 206 healthy age- and sex-matched controls. The Statistical Package for the Social Sciences version 19 and Fisher’s exact and chi-square tests were used to analyze the data. Moreover, six machine-learning models and two sequential deep-learning models were designed to classify patients from normal people in the E-MTAB-11191 dataset through the expression level of Foxp3 and its correlated genes. The allele and genotype frequencies of both polymorphisms in question were found to be significantly associated with an increased risk of systemic lupus erythematosus. Furthermore, both of the two single-nucleotide polymorphisms were associated with some systemic-lupus-erythematosus-related risk factors. Three SVM models and the logistic regression model showed an 81% accuracy in classification problems. In addition, the first deep-learning model showed an 83% and 89% accuracy for the training and validation data, respectively, while the second model had an 85% and 79% accuracy for the training and validation datasets. In this study, we are prompted to represent the predisposing loci for systemic lupus erythematosus pathogenesis and strived to provide evidence-based support to the application of machine learning for the identification of systemic lupus erythematosus. It is predicted that the recruiting of machine-learning algorithms with the simultaneous measurement of the applied single nucleotide polymorphisms will increased the diagnostic accuracy of systemic lupus erythematosus, which will be very helpful in providing sufficient predictive value about individual subjects with systemic lupus erythematosus.

Systemic lupus in the era of machine learning medicine

“…Reports using genomic and genetic expression datasets identified several important biomarker for LN including C1QA, C1QB, MX1, RORC, CD177, DEFA4, and HERC5 for LN. 118 For non-renal SLE, FOXP3, 88 MX2, 106 HLA-DQA1, 90 HLA-DQB1, 90 HLA-DRB1, 90 neutrophil extracellular trap-related genes (HMGB1, ITGB2 and CREB5), 70 ABCB1, 120 IFI27 120 and PLSCR1 120 have been reported. Other types of biomarkers included proteomics (IFIT3, MX1, TOMM40, STAT1, STAT2 and OAS3), 101 metabolomics, 91 lipidomics 94 and microRNA profiles.…”