Figure S3.
Characterization of GABAAR mAb binding to native receptors. (A) Overview of binding from patient’s polyclonal samples and derived recombinant GABAAR mAbs in CBAs using COS7 cells overexpressing individual or multiple GABAAR subunits as indicated. Examples are shown for α1β3γ2, α1β3, β3, and γ2 in Fig. 2 A. Patient’s CSF and PPE showed CBA-binding patterns similar to the derived CSF mAbs with prominent α1 reactivity and additional weak binding to COS7 cells expressing α2β3, α5β3 (both), and β3γ2 (PPE only). #113-101, #113-115, and #113-198 selectively target α1, as detecting cells expressing α1β3 and α1β3γ2, but not β3 or γ2 alone (Fig. 2 A). Specificity was confirmed by the absence of reactivity to α2β3γ2, α3β3γ2, α4β3γ2, and α5β3γ2. #113-175 detects GABAAR only when α1 and γ2 are coexpressed, but not in any other combination containing a different α-subunit (α2–α5) with β3γ2 coexpression. mAb #113-201 binds the subunits α1 and γ2 independently (Fig. 2 A). The α-subunit–mediated reactivity is specific to α1, as CBA overexpressing different α-subunits (α2–α5) in combination with β3 was not detected. GABA-negative control #mGO53 showed no binding. All GABAAR-negative mAbs from patient’s CSF repertoire with anti-neuronal reactivity (exemplarily shown for #113-109) and negative control #mGO53 showed no binding. Note that β3 alone is expressed on the cell surface, but α1 alone is not (Ebert et al., 1999). (B–G) A flow cytometry approach analyzing HEK cells transfected with rat or human α1β3γ2 GABAAR and EGFP was used to complement tissue-based quantification of GABAAR mAb binding to native receptors (Fig. 2 B). (B) Selection of cells for analysis was based on sequential gating on a homogeneous cell population in forward scatter (FSC) and sideward scatter (SSC), single cells, live cells (negative for DAPI), and lastly the population of 30% highest EGFP signal as a marker for transfection. (C) Concentration-dependent GABAAR mAb #113-115 binding to EGFP-positive and thus GABAAR HEK cells in comparison to EGFP-negative HEK cells. Similar data were obtained from all GABAAR mAbs. (D and E) Relative MFI values were used to model binding to HEK-expressed rat GABAAR (D) or human GABAAR (E) using nonlinear regression models for one site-specific binding (Table S2). Bars indicate mean ± SEM from n = 3 experiments. (F and G) Regression model–derived Half Max concentrations (50% of saturation binding = MFImax) are compared between the flow cytometry assays using rat or human α1β3γ2 GABAAR and the murine tissue–binding assay (Fig. 2 A): The correlations indicate generally similar binding between assays (R = −0.70, P = 0.23 in F; R = −0.21, P = 0.77 in G), with the exception of #113-201, which bound weakly to murine and rat GABAAR but was the strongest binder to human GABAAR. (H) For analysis of competitive binding (in addition to Fig. 2 C), fluorophore-coupled GABAAR mAb #113-101 was stained on unfixed murine brain tissue in combination with GABAAR mAbs as full IgG or Fab in excess as indicated. Quantified mean MFIs as relative values to noncompetition conditions are shown as a heat map, each from 30 ROIs from two independent experiments. Receptor binding competition is visualized in black and signal enhancement in yellow. (I) An ELISA assay was used to exclude binding of #113-101 to the variable domain of the #113-115. High-binding plates were coated with commercial anti-human IgG MT145 or Fab fragments of control mAb #mGO53 or GABAAR mAb #113-115 as indicated in column group titles above. Sample mAbs (human IgG or Fab) were then applied as indicated in column labels below and detected using commercial detection antibody MT78-ALP. Note that MT145 and MT78-ALP are Fc specific; thus, Fabs could not be detected when used as sample mAb. Bars indicate mean + SD from triplicates of n = 2 experiments. (J) A control ELISA assay was used to confirm successful coating of Fab fragments to high-binding plates. Plates were coated as above before application of commercial mouse anti-FLAG antibody (as sample mAb) for capture by Fab fragments. Bars indicate mean + SD from triplicates of n = 2 experiments.

Characterization of GABAAR mAb binding to native receptors. (A) Overview of binding from patient’s polyclonal samples and derived recombinant GABAAR mAbs in CBAs using COS7 cells overexpressing individual or multiple GABAAR subunits as indicated. Examples are shown for α1β3γ2, α1β3, β3, and γ2 in Fig. 2 A. Patient’s CSF and PPE showed CBA-binding patterns similar to the derived CSF mAbs with prominent α1 reactivity and additional weak binding to COS7 cells expressing α2β3, α5β3 (both), and β3γ2 (PPE only). #113-101, #113-115, and #113-198 selectively target α1, as detecting cells expressing α1β3 and α1β3γ2, but not β3 or γ2 alone (Fig. 2 A). Specificity was confirmed by the absence of reactivity to α2β3γ2, α3β3γ2, α4β3γ2, and α5β3γ2. #113-175 detects GABAAR only when α1 and γ2 are coexpressed, but not in any other combination containing a different α-subunit (α2–α5) with β3γ2 coexpression. mAb #113-201 binds the subunits α1 and γ2 independently (Fig. 2 A). The α-subunit–mediated reactivity is specific to α1, as CBA overexpressing different α-subunits (α2–α5) in combination with β3 was not detected. GABA-negative control #mGO53 showed no binding. All GABAAR-negative mAbs from patient’s CSF repertoire with anti-neuronal reactivity (exemplarily shown for #113-109) and negative control #mGO53 showed no binding. Note that β3 alone is expressed on the cell surface, but α1 alone is not (Ebert et al., 1999). (B–G) A flow cytometry approach analyzing HEK cells transfected with rat or human α1β3γ2 GABAAR and EGFP was used to complement tissue-based quantification of GABAAR mAb binding to native receptors (Fig. 2 B). (B) Selection of cells for analysis was based on sequential gating on a homogeneous cell population in forward scatter (FSC) and sideward scatter (SSC), single cells, live cells (negative for DAPI), and lastly the population of 30% highest EGFP signal as a marker for transfection. (C) Concentration-dependent GABAAR mAb #113-115 binding to EGFP-positive and thus GABAAR HEK cells in comparison to EGFP-negative HEK cells. Similar data were obtained from all GABAAR mAbs. (D and E) Relative MFI values were used to model binding to HEK-expressed rat GABAAR (D) or human GABAAR (E) using nonlinear regression models for one site-specific binding (Table S2). Bars indicate mean ± SEM from n = 3 experiments. (F and G) Regression model–derived Half Max concentrations (50% of saturation binding = MFImax) are compared between the flow cytometry assays using rat or human α1β3γ2 GABAAR and the murine tissue–binding assay (Fig. 2 A): The correlations indicate generally similar binding between assays (R = −0.70, P = 0.23 in F; R = −0.21, P = 0.77 in G), with the exception of #113-201, which bound weakly to murine and rat GABAAR but was the strongest binder to human GABAAR. (H) For analysis of competitive binding (in addition to Fig. 2 C), fluorophore-coupled GABAAR mAb #113-101 was stained on unfixed murine brain tissue in combination with GABAAR mAbs as full IgG or Fab in excess as indicated. Quantified mean MFIs as relative values to noncompetition conditions are shown as a heat map, each from 30 ROIs from two independent experiments. Receptor binding competition is visualized in black and signal enhancement in yellow. (I) An ELISA assay was used to exclude binding of #113-101 to the variable domain of the #113-115. High-binding plates were coated with commercial anti-human IgG MT145 or Fab fragments of control mAb #mGO53 or GABAAR mAb #113-115 as indicated in column group titles above. Sample mAbs (human IgG or Fab) were then applied as indicated in column labels below and detected using commercial detection antibody MT78-ALP. Note that MT145 and MT78-ALP are Fc specific; thus, Fabs could not be detected when used as sample mAb. Bars indicate mean + SD from triplicates of n = 2 experiments. (J) A control ELISA assay was used to confirm successful coating of Fab fragments to high-binding plates. Plates were coated as above before application of commercial mouse anti-FLAG antibody (as sample mAb) for capture by Fab fragments. Bars indicate mean + SD from triplicates of n = 2 experiments.

This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal