Dynamic single-cell transcriptomic analysis of peripheral blood mononuclear cells identifies novel biomarkers for autoimmune encephalitis.

Autoimmune encephalitis (AE) is a severe neurological disease characterized by diverse clinical symptoms. Current routine imaging and laboratory assessments show low disease specificity, resulting in a high misdiagnosis rate. We dynamically characterized the single-cell transcriptome of peripheral immune cells in the acute phase of AE and six-month sequential treatment. Depletion of natural killer T (NKT) cells, specifically of the NKT-6 subpopulation, was an AE hallmark that effectively differentiated AE patients from healthy controls. Exhausted NKT in AE was independently validated to be disease-specific using an independent clinical cohort of neurological and psychiatric conditions, including neuromyelitis optica spectrum disorder, multiple sclerosis, viral encephalitis, and major depressive disorder. In addition, diminished NKT cells and NKT-6 subpopulation levels gradually reversed with clinical improvement during a 6-month follow-up. Moreover, gene modules derived from NKT cells reflected disease progression. Our findings strongly support the notion that decreased NKT cell and NKT-6 subpopulation levels are potential novel biomarkers for AE diagnosis and disease progression.