Characterization of reactivity and Ig sequence features from mAbs of GABA_AR encephalitis CSF repertoire. (A) Gating strategy in fluorescence-activated cell sorting is shown for isolation of CSF single cells for recombinant mAb cloning. CD3⁻CD14⁻CD16⁻DAPI⁻ lymphocytes (top left) were gated for CD138⁺ antibody-secreting cells (top right) or CD20⁺ B cells (bottom left), further differentiated into CD27⁺ MBCs and CD27⁻ NMBCs (bottom right). (B) Immunofluorescence stainings of recombinant human mAbs (green, as indicated in column caption) to HEK cells overexpressing the α1β3- or α1β3γ2-subunits of GABA_AR or untransfected controls (as indicated in row caption). Costaining with commercial α1-specific antibody is shown in red and nuclei staining with DAPI in blue. Representative scale bar indicates 20 µm. (C) Ig subclass distributions per mAb source cell type from GABA_AR encephalitis CSF repertoire. (D and E) Absolute frequencies of GABA_AR-reactive (GABA_AR⁺) and GABA_AR-negative (GABA_AR⁻) mAbs per Ig subclass (D) and mAb source cell type (E). (F) Comparison of SHM counts in the variable domain V genes between mAbs of different source cell types, analyzed using ordinary one-way ANOVA followed by post hoc Tukey’s multiple comparison (**, P ≤ 0.01; or not shown when P > 0.05). Each dot indicates one mAb, n = 5–46 mAbs per group. Bars indicate mean ± SD (G) Comparison of SHM counts in the variable domain V genes between GABA_AR⁺ and GABA_AR⁻ mAbs. Each dot indicates one mAb, n = 5–62 mAbs per group. Bars indicate mean ± SD. (H) Relative frequencies of SHM per nucleotide within CDRs and FRs of GABA_AR⁺ mAb genes, shown as mean ± SEM; n = 5. (I) Mean ratios of replacement to silence (R/S) mutations within CDRs and FRs for all SHMs of all GABA_AR⁺ mAb genes combined. (J–O) Concentration-dependent binding of GABA_AR⁺ mAbs to indicated polyreactivity-defining antigens in an ELISA-based assay in comparison to controls of strongly polyreactive ED38, weakly polyreactive eiJB40, and nonpolyreactive #mGO53. Bars indicate mean ± SEM from duplicate measurements of n = 2 independent experiments. All stainings were replicated at least twice on tissue or neurons from two different animals. dsDNA, double-stranded DNA; FSC, forward scatter; ssDNA, single-stranded DNA.