https://orcid.org/0000-0002-4136-6792
Figures & data
Two synaptic clefts are shown comparing the hypothesized effects of treatment with traditional (bivalent) vs. multivalent antibody formats. In both presynaptic neurons, alpha-Synuclein aggregation is taking place. The traditional antibody formats are not binding the small alpha-Synuclein-oligomers strong enough to stop the spreading, which leads to neurotoxicity on the postsynaptic neuron. In contrast, the multivalent antibody formats are hypothesized to bind stronger to small and large oligomers and inhibit the spreading and neurotoxicity to the postsynaptic neuron.
Figure 1. Design of recombinantly produced antibodies.
(a) Schematic illustrations of the three antibodies SynO2, TetraSynO2 and HexaSynO2. All antibodies have a murine IgG2c backbone. TetraSynO2 has single-chain variable fragments attached to the N-termini of the heavy chains. HexaSynO2 has single-chain variable fragments attached to the N-termini of both heavy and light chains. (b) SDS-PAGE with Coomassie staining of SynO2, TetraSynO2 and HexaSynO2 showing the purity and expected molecular weight of the antibodies.
(a) Schematic illustration of the parental antibody SynO2, TetraSynO2 and HexaSynO2 with scFvSynO2 recombinantly fused to the N-termini of heavy and/or light chains of SynO2. (b) SDS-PAGE with Coomassie staining of antibodies loaded under non-reducing and reducing (red.) conditions. Bands of intact antibodies under non-reducing conditions appear at approximately 150 kDa for SynO2, 200 kDa for TetraSynO2 and 260 kDa for HexaSynO2 (1 µg protein/lane).
Figure 2. Thermal stability of SynO2, TetraSynO2 and HexaSynO2 measured by Tycho.
Tycho measurements of the antibodies SynO2, TetraSynO2 and HexaSynO2. The first derivate of the ratio between the intrinsic fluorescence measured at 350 nm and 330 nm is plotted against the temperature. SynO2 has a major peak at 74 degrees Celsius. TetraSynO2 and HexaSynO2 have their major peaks at 70 degrees Celsius. All three antibodies have an additional minor peak at approximately 82 degrees Celsius.
First derivate of ratio between intrinsic fluorescence measured at 350 nm and 330 nm, while the antibodies were heated up linearly from 35°C to 95°C. Inflection temperatures, visible as peaks, represent major unfolding events at 74°C and 82°C for SynO2, 70°C and 81°C for TetraSynO2 and 70°C and 83°C for HexaSynO2, suggesting high structural stability for all three antibodies.
Figure 3. Characterization of SynO2Fab.
(a) Coomassie-stained SDS-PAGE shows a band at 150 kDa for biotinylated SynO2 and a band at 50 kDa for the biotinylated SynO2Fab. 1 µg protein/lane. The complete gel can be seen in Figure S6. (b) Schematic illustration of the indirect αSyn aggregate ELISA set-up with biotinylated SynO2 or biotinylated SynO2Fab binding to the αSyn HNE aggregate coating and detection by Streptavidin-horseradish peroxidase (HRP). (c) Indirect αSyn aggregate ELISA binding curves show strong binding by SynO2 to αSyn HNE aggregates and weak binding by SynO2Fab. The degree of biotinylation of SynO2 and SynO2Fab was compared by direct ELISA (Figure S7) and was verified by liquid chromatography–mass spectrometry (Figure S8-9, Tables S1-3).
(a) SDS-PAGE with Coomassie staining of SynO2, biotinylated SynO2 and biotinylated SynO2Fab showing the purity and expected molecular weight of the respective antibody or Fab. (b). Schematic illustration of the indirect alpha-Synuclein ELISA set-up where a surface is coated with alpha-Synuclein HNE aggregates. Either biotinylated SynO2 or biotinylated SynO2Fab is illustrated to bind bivalently or monovalently, respectively, to these aggregates. Streptavidin-HRP binds to the respective biotin moiety on SynO2 or SynO2Fab and the HRP produces a color upon reaction with a substrate. (c) The readout of the indirect ELISA with biotinylated SynO2 and biotinylated SynO2Fab is presented in a line graph. The absorbance at 450 nanometer wavelength is plotted against the antibody concentration. The binding curve of SynO2Fab is significantly shifted toward higher antibody concentrations, compared to the binding curve of SynO2, indicating a weaker binding of SynO2Fab to alpha-Synuclein HNE aggregates.
Figure 4. Sandwich Aβ ELISA detects cross-reactivity of SynO2 and HexaSynO2 with Aβ aggregates.
(a) Schematic illustration of the sandwich amyloid-beta ELISA set-up where a surface is coated with an antibody that can specifically bind the C-terminal of amyloid-beta. Amyloid-beta protofibrils are captured by the coating antibody. SynO2 and HexaSynO2 are illustrated to bind to the protofibrils and in turn are detected by an HRP-conjugated secondary antibody, which produces a color upon reaction with the HRP substrate. (b) The readout of the sandwich amyloid-beta ELISA is presented in a line graph. The absorbance at 450 nanometer wavelength is plotted against the antibody concentration. The binding curves of SynO2 and HexaSynO2 indicate a very weak binding to amyloid-beta protofibrils when compared with the amyloid-beta aggregate-specific antibody RmAb158.
(a) Schematic illustration of the sandwich Aβ ELISA set-up with Aβ42 protofibrils captured by an Aβ C-terminal-specific antibody coated to a plate. Binding of SynO2, HexaSynO2 and RmAb158 was detected through an HRP-conjugated secondary antibody. (b) Sandwich ELISA binding signal demonstrating unspecific binding of SynO2 and HexaSynO2 to Aβ42 protofibrils at high antibody concentrations. Nonlinear regression curves (“one site – specific binding”) were calculated in GraphPad Prism.
Figure 5. Inhibition ELISA illustrating the binding strength of SynO2 and HexaSynO2 to αSyn monomers, HNE aggregates and αSyn fibrils.
(a) Schematic illustration of the αSyn inhibition ELISA set-up, with SynO2 or HexaSynO2 pre-incubated with αSyn monomers, αSyn HNE aggregates or αSyn fibrils, and subsequently added to an αSyn HNE aggregate-coated plate. (b) Visualization of binding signals, normalized to 0% as no signal and 100% as the maximal signal of each construct, respectively. Linear regression curves (“log(inhibitor) vs. normalized response”) were calculated in GraphPad Prism. IC50 values, calculated from the regression curves, indicate the concentration of αSyn HNE aggregate or αSyn fibrils needed in solution with the antibody to inhibit 50% of the respective antibody from binding to the αSyn HNE aggregate-coated surface.
(a) Schematic illustration of the inhibition ELISA set-up, where either alpha-Synuclein monomers or aggregates are pre-incubated separately with either SynO2 or HexaSynO2. These mixtures are then added to a plate with alpha-Synuclein coating. SynO2 and HexaSynO2 are illustrated to bind to the alpha-Synuclein coating and in turn are detected by an HRP-conjugated secondary antibody, which produces a color upon reaction with the HRP substrate. (b) The readout of the inhibition ELISA is presented in a line graph. The percentage of antibody binding is plotted against the alpha-Synuclein concentration. The binding curve of HexaSynO2 with aggregates and fibrils is shifted the farthest to the left compared to SynO2 with aggregates and fibrils. This means that it requires lower alpha-Synuclein concentrations for HexaSynO2 than for SynO2 in the pre-incubation to inhibit the respective antibody from binding to the alpha-Synuclein-coated plate.
Figure 6. Kinetic evaluation of interactions of SynO2Fab, SynO2, TetraSynO2 and HexaSynO2 with αSyn HNE aggregates recorded by LigandTracer.
(a) Schematic illustration of the LigandTracer set-up with a plastic surface coated with αSyn HNE aggregates and 125I-labeled antibodies added in solution. (b) Interaction curves of SynO2Fab and SynO2, and (c) SynO2, TetraSynO2 and HexaSynO2 with αSyn HNE aggregates recorded by LigandTracer and fitting curves (red) using a one-to-one depletion corrected model (SynO2Fab, HexaSynO2) or a one-to-two model (SynO2, TetraSynO2), respectively. 100 nM coating with αSyn HNE aggregates. Two consecutive association phases (3 hours and 4 hours, respectively) with 1 nM and 3 nM of the respective 125I-labeled antibody. Recorded interaction curves were evaluated in TraceDrawer using a “one-to-one depletion corrected” (SynO2Fab and HexaSynO2) or a “one-to-two” model (SynO2 and TetraSynO2). Signal intensities of each curve were scaled to Bmax, the estimated signal intensity at saturation, with 100% representing target saturation.
(a) Schematic illustration of the LigandTracer experimental design, alpha-Synuclein aggregates are coated to Petri-dish surface and incubated with Iodine-125-labeled antibodies for two consecutive association phases of 3 and 4 hours. The unbound antibodies are washed off and a dissociation phase follows for additional 14 hours. (b) The readouts of the LigandTracer experiments with SynO2 and SynO2Fab are presented in a line graph together with the respective fitting curves. The percentage of occupied targets is plotted against the time. The interaction curve of SynO2Fab reaches an equilibrium very quickly in each association or dissociation phase, indicating a weak binding. Calculated kinetic parameters are presented in an inserted table. (c) The readouts of the LigandTracer experiments with SynO2, TetraSynO2 and HexaSynO2 are presented in a line graph together with the respective fitting curves. The percentage of occupied targets is plotted against the time. The major observation is that the interaction curve of TetraSynO2 and HexaSynO2 has almost no decline during the dissociation phase, whereas SynO2 shows a decline during in the first two hours of dissociation phase. Calculated kinetic parameters are presented in an inserted table.
Table 1. Start values for kinetic evaluations of interactions between antibodies and αSyn HNE aggregates recorded by LigandTracer. All interaction curves were individually fit with start values set at (1) global scope or (2) constant scope as indicated below, respectively.
Supplemental material