Expression of HMGB1 and TLR4 in neuropsychiatric systemic lupus erythematosus patients with seizure disorders

Qin Huang, Shuqun Shen, Hang Qu, Yu Huang, Danni Wu, Haishan Jiang, Chao Yuan

Abstract

Background: Previous studies show that the high‐mobility group box protein 1 (HMGB1) and the toll‐like receptor 4 (TLR4) participate in systemic lupus erythematosus (SLE). The two molecules contribute to the occurrence and persistence of seizures in various disease conditions, such as epilepsy. Since seizures are one of the most severe complications associated with neuropsychiatric SLE (NPSLE), the current study aimed at investigating whether HMGB1 and TLR4 play any role in NPSLE related seizures.
Methods: Data from 291 SLE patients and 100 healthy controls (HC) were prospectively collected from 2013 to 2018. The ELISA test was used to determine serum levels of HMGB1 for all patients and HC and cerebrospinal fluid (CSF) levels of NPSLE patients. The expression levels of TLR4 by the peripheral blood monocytes (PBMCs) were determined by real-time PCR of TLR4 mRNA. Binary logistic regression and ROC curve analysis were used to predict NPSLE.
Results: Among the 291 SLE patients, 188 had active disease and were grouped into two, NPSLE (N=86) and Non-NPSLE (N=102) groups. Among the NPSLE patients, 21 had seizure disorders. Serum HMGB1 levels were increased in NPSLE (8.73±0.29 ng/mL) and were associated with disease activity (r=0.6527, P=0.000). Both serum and CSF HMGB1 levels in NPSLE patients with seizure disorders (9.59±0.63 and 2.90±2.29 ng/mL, respectively) were higher than in patients with other neuropsychiatric symptoms (8.45±0.33 and 2.56±1.70 ng/mL, respectively), though without significance. The gene expression of mRNA TLR4 in PBMCs was similar to serum HMGB1 in the investigated groups. Independent predictors of NPSLE were SLEDAI-2k (OR 1.25; 95% CI: 1.155–1.353), serum HMGB1 (OR 1.659; 95% CI: 1.266–2.175), and anti- Rib-P Ab (OR 3.296; 95% CI: 1.013–10.725). ROC curves for the above predictors had a large AUC (95% CI) of 0.936 (0.900–0.971), indicating a good prediction of NPSLE occurrence.
Conclusions: The expression of HMGB1 and TLR4 was increased in NPSLE, but HMGB1 and TLR4 had minimal effect on NPSLE related seizures. The serum levels of HMGB1 were positively correlated with disease activity, and could, therefore, be a potential biomarker of NPSLE for use in future clinical practice.