Fig 2: Blockade of clathrin-mediated endocytic pathway via Fc II/III receptors inhibits SynI antibody uptake and restores synaptic transmission.a Representative images of SynI-mAb (1.5 µg/mL)-treated hippocampal neurons incubated either with sodium azide (NaN3; 500 nM), PitStop 2 (30 µM) or a Fc?R II/III antibody at the indicated concentrations for 72 h. The control condition indicates cells exposed to SynI-mAb in the presence of the NaN3/PitStop2 solvent. After treatments, cells were fixed, permeabilized and incubated with anti-mouse fluorescent secondary antibodies (green). Double staining for ßIII-tubulin (blue) was performed after fixation. Scale bars, 10 µm. b Quantification of the immunoreactivity intensity of internalized SynI-mAb. Graph shows mean ± SEM, single neuron data points independent preparations; **p < 0.01, ***p < 0.001, one-way ANOVA/Dunnett’s tests versus control. c Mean frequency and amplitude of mEPSCs (upper panels) and mIPSCs (lower panels) in hippocampal neurons treated with either SynI-mAb (1.5 µg/mL) or vehicle for 72 h in the presence or absence of 500 nM NaN3. d Mean frequency and amplitude of mEPSCs (upper panels) and mIPSCs (lower panels) in hippocampal neurons treated with either control CSF or patient #1 CSF (1.5 µg/mL) for 72 h in the presence or absence of 500 nM NaN3. Graphs show means ± SEM, single neuron data points from three independent preparations; *p <0.05, **p < 0.01, two-way ANOVA/Bonferroni’s tests. e Expression of Fc?II/III receptors in WT and SynI KO hippocampal neurons. Representative images of primary WT and SynI KO hippocampal neurons at 14.

Fig 3: Patient-derived SynI autoantibodies decrease excitatory and inhibitory synaptic transmission and synapses.a Representative recordings of miniature excitatory postsynaptic currents (mEPSCs) in primary hippocampal neurons treated with either control CSF (1.5 µg/mL, black traces), LE-CSF (1.5 µg/mL, red traces) or CIS-SF (1.5 µg/mL, blue traces) for 72 h. b Cumulative distribution of inter-event intervals and mean mEPSC frequency. c Cumulative distribution and mean of mEPSC amplitude. d Rise and decay times of mEPSCs. e Representative recordings of miniature inhibitory postsynaptic currents (mIPSCs) in hippocampal neurons treated as in (a). f Cumulative distribution of inter-event intervals and mean frequency of mIPSCs. g Cumulative distribution and mean of mIPSC amplitude. h Rise and decay times of mIPSCs. Cumulative plots show pooled data from all neurons; graphs show means ± SEM, single neuron data points from at least three independent preparations; *p < 0.05, ***p < 0.001, Kruskal–Wallis test and Dunn’s post hoc. i Representative images of primary hippocampal neurons treated with either control CSF or LE-CSF (1.5 µg/mL) for 72 h and stained for ß-III tubulin (blue), VGLUT1 (green) and VGAT (red). Scale bars, 10 and 1 µm for low and high magnification, respectively. j, k Density of VGLUT1-positive excitatory synapses (j) and VGAT-positive inhibitory synapses (k) in control CSF and LE-CSF treated neurons. l Representative images of primary hippocampal neurons treated as in (a) and stained for ß-III tubulin (blue) and SynI (green). Scale bars, 10 and 1 µm for low and high magnification, respectively. m Quantification of the density of SynI immunoreactive synaptic puncta. Graphs show means ± SEM, single neuron data points from three independent preparations; **p < 0.01, ***p < 0.001, Mann–Whitney test.

Fig 4: Patient autoantibodies and affinity-purified SynI antibodies have a similar distribution in primary neurons and their intracellular accumulation requires SynI.Representative images of WT (a) or SynI KO (b) neurons incubated for 72 h with either LE-CSF, SynI-mAb or equal volumes of the respective vehicles. Neurons were fixed and primary antibodies were visualized with anti-human and anti-mouse fluorescently conjugated secondary antibodies (green), respectively. Double staining for ßIII tubulin (blue) was performed after fixation. Scale bars, 10 and 1 µm for low and high magnification, respectively. c Quantification of the intensity of primary antibody immunoreactivity. The graph shows means ± SEM, single neuron data points from 4 independent preparations; ***p < 0.001, two-way ANOVA/Bonferroni’s tests. d Representative TEM images of nerve terminals from cultured hippocampal neurons incubated with either control CSF or LE-CSF and processed for immunogold against human IgG antibody. Arrow indicates gold particles after gold enhancement. Scale bars, 200 nm. e Histograms representing the percentage of gold-positive synapses. Graphs show means ± SEM, n = 4–5 synapses from two independent preparations; *p < 0.05, two-tailed Student’s t-test.

Fig 5: Genetic ablation of TNFa in skeletal muscle cells delays synapse elimination.a Strategy for gene targeting. A pair of LoxP sites flanked exon 3 and 4 of murine TNFa gene. The Cre-LoxP and Flp-Frt systems were used to ablate TNFa gene and selection marker Neo from the murine genome, respectively. P3 and P4 indicate sites corresponding to primers used for genotyping. b Genotyping of conditional knockout mice using PCR method. c Absence of TNFa expression in sternocleidomastoid muscles of TNFaf/f; HSA-Cre mice at P6. Scale bar: 20 µm. d–f LAL muscles of conditional knockout mice at indicated times (P8, P14) were whole-mount stained with R-BTX (red) and antibodies against NF and Syn1 (Nerve, green). Shown are example images of NMJs of mice with indicated genotypes at indicated times. White arrows indicate poly-innervated NMJs (d). Scale bars: 20 µm. g Quantification for the percentage of poly-innervated NMJs. Data are shown as means ± SEM (P8, 1877 NMJs from 4 TNFaf/f mice, 2002 NMJs from 4 TNFaf/f; HSA-Cre mice; P14, 1036 NMJs from 6 TNFaf/f mice, 3708 NMJs from 8 TNFaf/f; HSA-Cre mice, 1372 NMJs from 3 HSA-Cre mice, 1636 NMJs from 3 TNFaf/-; HB9-Cre mice, 636 NMJs from 3 TNFaf/f; HB9-Cre mice, 1707 NMJs from 3 TNFaf/f mice, 3708 NMJs from 3 TNFaf/f; MPZ-Cre mice). Mann–Whitney test was used to determine significance. *P < 0.05, ***P < 0.001, NS no significant difference.