Identification and characterization of GABA(A) receptor autoantibodies in autoimmune encephalitis

Abstract

Autoimmune forms of encephalitis have been associated with autoantibodies against synaptic cell surface antigens such as NMDA- and AMPA-type glutamate receptors, GABA(B) receptor, and LGI1. However, it remains unclear how many synaptic autoantigens are yet to be defined. Using immunoproteomics, we identified autoantibodies against the GABA(A) receptor in human sera from two patients diagnosed with encephalitis who presented with cognitive impairment and multifocal brain MRI abnormalities. Both patients had antibodies directed against the extracellular epitope of the β3 subunit of the GABA(A) receptor. The β3-subunit-containing GABA(A) receptor was a major target of the patients' serum antibodies in rat hippocampal neurons because the serum reactivity to the neuronal surface was greatly decreased by 80% when the β3 subunit was knocked down. Our developed multiplex ELISA testing showed that both patients had similar levels of GABA(A) receptor antibodies, one patient also had a low level of LGI1 antibodies, and the other also had CASPR2 antibodies. Application of the patients' serum at the time of symptom presentation of encephalitis to rat hippocampal neuron cultures specifically decreased both synaptic and surface GABA(A) receptors. Furthermore, treatment of neurons with the patients' serum selectively reduced miniature IPSC amplitude and frequency without affecting miniature EPSCs. These results strongly suggest that the patients' GABA(A) receptor antibodies play a central role in the patients' symptoms. Therefore, this study establishes anti-GABA(A) receptor encephalitis and expands the pathogenic roles of GABA(A) receptor autoantibodies.

Keywords: GABAA receptor; autoantibody; autoimmune encephalitis; cognitive impairment; seizure; thymoma.

Figure 1.

Identification of GABA_A receptor autoantibodies in a patient with autoimmune encephalitis. A, Immunoprecipitation of cell surface target proteins with patient's serum antibodies. The immunoprecipitates of serum antibodies bound to rat hippocampal neurons were analyzed by SDS-PAGE with silver staining. The specific band at 48 kDa (arrowhead) was analyzed by the LC-MS/MS. B, MS/MS spectra of a peptide unique for the GABA_A receptor α1 subunit (m/z value of the parent ion, 509.74) obtained from the trypsinized protein band shown in A (arrowhead). The matched fragment y⁺ ion series is indicated in red. Identified peptides in the amino acid sequence of GABA_A receptor α1 are indicated in red. The accession number is P62813. The patient serum used came from the initial episode of encephalitis of Patient 1. C, Western blotting with the subunit specific antibodies showed that the α1, β3, and γ2 subunits of the GABA_A receptor were present in the immunoprecipitate by the patient serum antibodies. Asterisks indicate the position of the human IgG heavy chain (A, C). D, Patient serum antibodies bind to the inhibitory GABA_A receptors at the cell surface of rat hippocampal neurons. The serum reactivity (red; human IgG) was well overlapped with surface-expressed γ2 subunits of GABA_A receptor (green), which were apposed to gephyrin scaffold (blue; marked by arrowheads). Magnified view of the region indicated by a white square. Scale bars, 10 μm (1 μm, magnified). IP, immunoprecipitation; Contr, control; Pt, patient; GABA_AR, GABA_A receptor; WB, Western blotting; IF, immunofluorescence; hIgG, human Ig.

Figure 2.

Patients' GABA_A receptor antibodies are directed to extracellular epitope of β3 subunit. A, B, COS7 cells were transfected (TF) to surface express the indicated GABA_A receptor subunits. Transfected cells were fixed and doubly stained with the patient sera (Patient 1 or Patient 2; red, human IgG) together with the antibodies specific to the individual expressed subunits (green, insets). Nuclear DNA was stained by Hoechst 33342 (blue) to distinguish untransfected cells. To clearly show the weak binding of the serum from Patient 1 to the γ2 subunit, the detector gain of the red channel is enhanced upon image acquisition (right, middle). The ratio of the human IgG intensity to the GABA_A receptor subunit intensity was graphed (B). Error bars indicate SEM; n = 10 transfected cells. C, COS7 cells were transfected to surface express the indicated heteromeric GABA_A receptors and tested for binding of serum antibodies (red). Transfected cells were detected by staining with the individual α subunit (green) and γ2 subunit (blue) antibodies. Merged images are shown in insets. Scale bars, 20 μm.

Figure 3.

GABA_A-receptor-containing β3 subunit is a major target of the patients' antibodies in hippocampal neurons. A, Validation of miRNA constructs for the GABA_A receptor β3 subunit. HEK293T cells were cotransfected with the indicated knock-down (miR) and β3 expression vectors. Three days after the transfection, the cell lysates were analyzed by Western blotting with GABA_A receptor β3 and β-catenin antibodies. miR-LacZ, Control miRNA targeting to LacZ; resβ3, miR-211-resistant β3. B, Effective knock down of the endogenous β3 subunit. Cultured rat hippocampal neurons were transfected with the miR-β3 expression vector at 10 DIV. Cell surface GABA_A receptor β2/β3 (red) and γ2 (blue) subunits were stained at 15 DIV. MicroRNA-transfected neurons were reported by the GFP expression (green). C, Binding of serum antibodies (Patient 1 and Patient 2; red) was examined in neurons in which the β3 subunit was knocked down (green). Magnified view of the region indicated by a white square (B, C). Arrows indicate the soma and dendrites of the neuron in which β3 was knocked down (B, C). Scale bars, 10 μm (2 μm, magnified). D, Neurons were cotransfected with the indicated miR and the knock-down-resistant construct (resβ3) or GST (for mock). The number of clusters labeled by β3 antibody or human IgG of patients' serum (Patient 1 and Patient 2) was counted and graphed. One-way ANOVA with Scheffe's post hoc analysis, ***p < 0.001 compared with miR-LacZ; †††p < 0.001 compared with miR-β3–211 + resβ3. Error bars indicate SEM. The number of neurons examined is indicated in parentheses.

Figure 4.

Identification of coexisting antibodies with GABA_A receptor antibodies in the patient serum. A, B, Patient 1 and Patient 2 sera from their initial episodes of encephalitis were tested by cell-based binding assay (A) and cell-based ELISA tests (B) against the GABA_A receptor β3 subunit, LGI1, CASPR2, and DCC. Additional sera were tested: from Patient A with limbic encephalitis (LE), Patient B with myasthenia gravis (MG) and thymoma (TH), and Patient C with neuromyotonia (NMT) and thymoma. Contr, Serum sample from a control patient with a neurodegenerative disease. Scale bar, 20 μm in A. Average values from triplicate measurements of the individual serum are shown in B.

Figure 5.

Patients' GABA_A receptor antibodies specifically reduce synaptic and cell surface GABA_A receptor density. A, Cultured rat hippocampal neurons were incubated with serum from Patient 1 and Patient 2, serum from Patient C with thymoma that contained both LGI1 and CASPR2 antibodies (Fig. 4), or a control serum for 2 d. Representative images of surface GABA_A receptor clusters in neurons treated with the control or the serum from Patient 1 are shown (left). Bottom, Magnified view of the region indicated by a white square. Synaptic GABA_A receptors, which were γ2 (red) or β3 (data not shown) subunit-positive clusters adjacent to both gephyrin (green) and vGAT (blue), were counted. Surface-expressed GABA_A receptor clusters labeled by the γ2 subunit antibody and gephyrin clusters were also independently counted. In addition, synaptic GluA1 clusters adjacent to both PSD-95 and vGluT1 and surface-expressed GluA1 clusters were counted. Scale bars, 20 μm (1 μm, magnified). Statistical analyses were performed by one-way ANOVA with Scheffe's post hoc analysis (γ2 clusters); or by Student's t test (β3 and GluA1 clusters). *p < 0.05; **p < 0.01; ***p < 0.001 compared with control; †p < 0.05, †††p < 0.001 compared with Patient C. Error bars indicate SEM. The number of separate cultures is indicated in parentheses. B, Surface biotinylated and total proteins of the serum from Patient 1 treated hippocampal neurons were analyzed by Western blotting with the indicated antibodies. Student's t test; n.s., not significant. Error bars indicate SEM; n = 3 separate cultures. C, D, Serum samples of Patient 1 at two different time points, from the episode of thymoma and MG (before onset of encephalitis) and from the episode of encephalitis, were analyzed by cell-based ELISA (C), and their effects on synaptic GABA_A receptors were investigated (D). Average values for GABA_A receptor β3 and LGI1 ELISA from triplicate measurements of the individual serum samples are shown (C). Titer of serum antibodies against AchR is shown (bar graph in C; Miyazaki et al., 2012). Synaptic GABA_A receptors were counted as in A. One-way ANOVA with Tukey's post hoc analysis; ***p < 0.001; n.s., not significant. Error bars indicate SEM; n = 3 separate cultures.

Figure 6.

Patients' GABA_A receptor antibodies selectively decrease mIPSCs. A, Representative traces of mIPSCs (V_H = −20 mV, top) and mEPSCs (V_H = −80 mV, bottom) recorded from cultured rat hippocampal neurons, which were incubated with the serum from Patient 1 and Patient 2 or a control individual for 1 d. B, D, Treatment of neurons with the patient serum significantly decreased the amplitude and frequency in mIPSCs (B), but did not affect those in mEPSCs (D). Statistical analyses were performed by one-way ANOVA with Scheffe's post hoc analysis. *p < 0.05; **p < 0.01; ***p < 0.001 compared with control. C, Cumulative distribution of the mIPSC amplitude. Note the significant leftward shift in the amplitude distribution in the patients' serum-treated neurons (F_{(2, 540)} = 62.9, p < 0.001; two-way ANOVA). Error bars indicate SEM; Control, n = 9; Patient 1, n = 10; Patient 2, n = 11 (B, C); Control, n = 9; Patient 1, n = 9; Patient 2, n = 12 (D). n values indicate the number of neurons examined from two separate cultures.