Immunoprecipitation and mass spectrometry-based identification: A positive CSF (patient 1) and a negative CSF (control CSF) were used. Five µL of CSF were mixed with 50 µL of protein G–conjugated agarose beads (Sigma Aldrich, Lyon, France) and completed to 500 µL with PBS. The mixture was incubated for 2 h at 4°C with rotation to allow for the coupling of CSF’s antibodies with protein G. Simultaneously, whole protein extract from one mouse brain was prepared and incubated with 50 µL of agarose beads. Nonspecific contaminant was removed from the lysate by 5-min centrifugation at 16 000 g and 4°C. Cleared lysate was subsequently used for immunoprecipitation with antibodies-conjugated agarose beads. Cell viability: To investigate the effect of patients’ CSF on neuron mortality in vitro, mouse hippocampal neurons were plated in 12-well plates that were pre-coated with poly-L-lysine at a density of 75 000 cells per well. The complete cell culture medium contained Neurobasal medium supplemented with 2% B-27, 1% L-glutamine, 0,5% penicilline/streptomycine, at 37°C and 5% CO2. At DIV 13, after washing, cells were incubated with patients’ CSF (1:50). To monitor cell death during incubation time, 0.2 µM propidium iodide (PI) were diluted in the cell culture medium. A videomicroscope Axiovert M200 with 20x objective and cooled CCD camera was used to recorded fluorescent images during experiments. The number of dead cells was counted by a person blinded to the study using in-house written macro running under ImageJ software. AK5 enzyme activity: Adenylate kinase (AK) enzyme activity was measured in mouse brains derived cytosolic fraction in the presence of the patient’s CSF or commercial anti-AK5 antibodies (Abcam, ab117927). Cytosolic fraction of rat brain homogenate was obtained as described above. In the presence of ADP, AK catalyzes the formation of ATP which was continuously assayed by using Luminescent ATP Detection Assay kit (ab113849, Abcam). Luminescence intensity was plotted against ATP concentration to generate a stander curve. AK activity is expressed in µmol of ATP formed per second per mg of protein. AK activity in each experiment was reported to the control (n≥3). Supplemental figure e-1: A: DAB Immunostaining of adult mouse brain with patient 1 and control CSF. Note a strong reactivity of patient 1’s CSF to the hippocampus (Hi), cerebellum (Cer) and cerebral cortex (Cx). Scale bare = 50 µm B: Fluorescence immunostaining of 18-days old embryonic mouse brain. Patient 1’s CSF reacts with the dentate gyrus (1) and the subplate neurons (2). Scale bare = 50 µm. Supplemental figure e-2: Immunoprecipitation of the antigens with patient 1’s CSF. Whole mouse brain protein extract was incubated with patient 1’s CSF or the control CSF. Antigen-antibody immunocomplex was then precipitated using protein G-coupled agarose bead. Upper panel shows silver staining of immunoprecipitate. Note the specifically enriched band around 70 kDa obtained after immunoprecipitation with patient1’s CSF, which was confirmed by immunoblot using the patient’s CSF (lower panel). H: whole brain homogenate, E: eluate of immunoprecipitation. Supplemental figure e-3: AK5 as an antigen of patient CSF autoantibodies: confirmation by immunohistochemistry. An adult rat brain section was immunostained with patient 1’s CSF and anti-AK5 antibody; both staining colocalized, confirming the nature of AK5 as the target of patient CSF autoantibodies. Scale bare = 200 µm Supplemental figure e-4: Axial FLAIR MRI from patient 1 (B) and coronal FLAIR MRI from patient8 at presentation (A and B) Axial FLAIR sequences from patient 4 at initial clinical symptoms (C) and after 2 months (C’) showing bilateral hippocampal atrophy. Supplemental figure e-5: Analysis of IgG subclasses of AK5 antibodies in the CSF of patients 2,3,4,5,6. Supplemental figure e-6: Epitope mapping with the CSF of patients 2,3,4,5. Supplemental figure e-7: A: Enzymatic activity of AK5 in a mouse brain. Cytosolic extract of an adult mouse brain was obtained by ultracentrifugation. A coupled-enzyme assay allows monitoring of the whole adenylate kinase activity by following the formation of ATP in the presence of 50 mM ADP. For each set of each experiment, AK activity in the presence of commercial anti-AK5 antibody or patients’ CSF was reported to the control. The mean values of at least three independent measurements were presented. B: Cytotoxicity. Cell death was monitored by using propidium iodide, which stains for the nucleus of broken cells. Neurons were incubated with control CSF (squares), patient 1’s CSF (triangles), patient 2’s CSF (circles), or without any CSF (diamonds). The quantity of cells that died during the incubation time was reported to initial cell number. The mean values of at least 3 independent experiments were presented. .