Single-cell immunology of autoimmune encephalitis: from antibody-defined syndromes to compartmentalized immune responses

Autoimmune encephalitis is diagnosed by the clinical syndrome and the presence of antibodies against neuronal surface or intracellular antigens, yet patients who share an antibody differ widely in severity, treatment response, and outcome. Antibody titre explains only part of this variation, because it does not describe the cells that produce the antibody, where they reside, or which other effectors are engaged. Single-cell RNA sequencing, paired single-cell B-cell and T-cell receptor sequencing, and single-cell multi-omics now allow these questions to be addressed directly in blood, cerebrospinal fluid, and brain tissue. We review evidence generated by these methods across antigen-defined autoimmune encephalitis. Anti-NMDAR studies describe B-cell expansion alongside compartment-specific T-cell states. In LGI1-antibody and CASPR2-antibody encephalitis, antigen-reactive B cells are enriched in cerebrospinal fluid, with evidence consistent with prior peripheral maturation and accompanying T-cell alterations. Ri- and GAD65-associated syndromes show distinct neuron-reactive or stem-cell-like memory T-cell populations, while antigen-reactive B cells may coexist. The review also considers compartment-specific responses, clonal maturation, change during treatment, and the practical constraints on clinical use. Most cohorts remain small, single-centre, and cross-sectional, and few findings have been validated prospectively. Single-cell immunology currently refines mechanistic understanding rather than replacing antibody-based diagnosis, and its near-term value lies in selecting candidates for functional testing and in defining variables for future trials.