Fig 1.
Production of recombinant antibody fragments.
(A) Molecular design of mono- (rFab) and bivalent forms (rIgG). The light chain and full length (HC) or VH-CH1 region (Fd) of the heavy chain are depicted with reddish and bluish colors, respectively. Intermolecular disulfide bonds are drawn in orange lines, whereas the C-terminal epitope tail consisting of a c-Myc epitope (c-Myc) and a hexa-histidine tag (6 × His) is drawn in purple. Fc, fragment crystallization; CH1-3, constant domains 1–3 of heavy chain; CL, constant domain of light chain, VH, variable region of heavy chain; VL variable region of light chain. (B) Schematic illustration of two expression systems. DNA fragments for an internal ribosome entry site (IRES) and self-cleavage 2A peptide from foot-and-mouth virus (F2A) are shown in green and orange, respectively. For producing antibody fragments into culture supernatants, heavy and light chain genes possess their own signal sequences (SSH and SSL, respectively) at the 5'-terminus. The gene encoding the Fd region was amplified without a stop codon (purple triangles) to add the epitope tags c-Myc and 6 × His at the 3'-terminus (S1B Fig and S2 Table). EF, a promoter of the human elongation factor gene; Bsd, blasticidin-resistant gene. (C) Analysis of molecular assembly of antibody fragments. Immunoblot analysis was carried out using authentic mAb 132 (IgG-132, 1.25–10.0 ng/lane) and supernatants of HEK293T cells transfected with expression plasmids containing IRES (rFab, rIgG) or F2A (rIgG) under non-reducing condition. Blot was probed with an HRP-conjugated anti-mouse IgG (Fab specific) (Sigma). Corresponding pictures on the right indicate the bands of IgG, rIgG, or rFab, whereas arrowheads indicate the bands of free and dimeric LC (LC2). Molecular mass markers on the left are in kDa.
Fig 2.
Reactivity of recombinant antibodies in ELISA.
(A) Reactivity of authentic and recombinant antibodies. Twofold serially diluted antibodies (0.002–0.3 nM Fab equivalent) were added to wells coated with 100 μg/well rMoPrP. Purified IgG was used for authentic mAbs, whereas the supernatant of HEK293T cells transfected with each expression plasmid was used for recombinant antibodies after estimation of concentrations equivalent to the Fab portion by immunoblot analysis using serially diluted purified IgG as a standard. Reactivities of authentic IgG and rIgG of mAbs 132, 31C6, and 44B1 as well as their rFab fragments are shown. Mean absorbances at 450 nm and SD (three independent experiments) are plotted. The supernatant of HEK293T cells transfected without any plasmids was used as a negative control (NC). Dotted line indicates a cutoff value calculated by the average plus 3 × SD. (B) Antigen concentration-dependent reactivity of authentic and recombinant antibodies. Twofold serially diluted rMoPrP (0.8–100 ng/well) was adsorbed to wells. For detection of rMoPrP, 0.3 nM antibodies equivalent to the Fab portion were used. (C) Quantification of purified rFab-132 and rIgG-132. Concentrations of recombinant antibodies were estimated by comparing 2-fold serially diluted authentic IgG-132 (0.125–10 ng). Blot was probed with an HRP-conjugated anti-mouse IgG (Fab specific) (Sigma), and signal intensities were quantified using Image Gauge software (Fujifilm). (D) Difference in reactivity of monovalent (rFab-132) and bivalent (IgG-132 and rIgG-132) forms of mAb 132. ELISA was carried out using 100 ng/well rMoPrP as an antigen and twofold serially diluted purified rFab-132 (0.09–12 nM), IgG-132, and rIgG-132 (0.02–3 nM) as primary antibodies.
Fig 3.
Reactivity of mAb 132-EGFP fusion proteins.
(A) Molecular design of mAb 132-EGFP fusion proteins. The light and heavy chains are depicted as in Fig 1A, and the spacer between heavy chains and EGFP is indicated with red lines. The epitope tag consisting of c-Myc epitope and a 6 × His (purple lines) was added at the C-terminus of EGFP. (B) Analysis of the molecular assembly of mAb 132-EGFP fusion proteins. Expression plasmids were introduced into HEK293T cells and mAb 132-EGFP fusion proteins were detected with a HRP-conjugated anti-mouse IgG (Fab specific) (Sigma). Corresponding illustrations on the right indicate the products of rFab-EGFP, rF(ab’)2-132-EGFP, rIgG(ΔCH3)-132-EGFP and rIgG-132-EGFP, whereas arrowheads indicate possible assembly intermediates, free light chain and free light chain dimer (LC2). The square brackets indicate the molecular weight range of mono- and bivalent fusion proteins. (C) Quantification of purified rFab-132-EGPG and rIgG-132-EGFP. Concentrations of fusion proteins were estimated by comparing 2-fold serially diluted authentic IgG-132 (0.125–10 ng). Blot was probed with an HRP-conjugated anti-mouse IgG (Fab specific), and signal intensities were quantified. (D) Difference in reactivity of monovalent (rFab-132-EGFP) and bivalent (rIgG-132-EGFP) form of mAb 132-EGFP fusion proteins. ELISA was carried out using 100 ng/well of rMoPrP as an antigen and 2-fold serially diluted purified rFab-132-EGFP (0.09–12 nM), rIgG-132-EGFP, and IgG-132 (0.02–3 nM) as primary antibodies.
Fig 4.
PrPSc-specific detection by direct immunofluorescence staining.
Prion-uninfected N2a-3 cells (A–E) and prion 22L strain-infected Na-3 cells (ScN2a-3-22L) (F–J) cells were directly stained using rIgG-132-EGFP (B, C, G, and H) and rFab-132-EGFP (D, E, I, and J) at the indicated concentrations equivalent to the Fab portions. The leftmost images (A, F) show negative controls for EGFP fusion proteins. Cell nuclei were stained with DAPI (blue).
Fig 5.
Binding affinity of authentic mAbs.
(A) Relationship between ligand and analyte. IgG-132 or IgG-31C6 was directly immobilized at approximately 1,000 RU on the surface of a CM5 sensorchip as a ligand. The rMoPrP was passed over the surface of the chip as an analyte. (B) Multi-cycle kinetics analysis. SPR sensorgrams for the binding of anti-PrP mAb to rMoPrP were obtained using 2-fold serial dilutions of rMoPrP (1.25–20 nM for IgG-31C6 and 6.3–50 nM for IgG-132).
Fig 6.
Binding affinity of mono- and bivalent authentic and recombinant antibody fragments.
(A) Schematic illustrations for analytes and ligands. The SPR sensorgrams were measured using two different binding types: mono- or bivalent antibody fragments were injected to the sensorchip with immobilized rMoPrP (left), or rMoPrP was injected to the sensorchip with each IgG (at approximately 500 RU) captured by anti-MoIgG (blue) immobilized on the surface (right). (B) Representative sensorgrams for the binding of authentic and recombinant antibody fragments of mAb 132 to rMoPrP by single-cycle kinetics. All sensorgrams were obtained using 2-fold serial dilutions of antibodies or rMoPrP: rFab-132, 0.63–10 and 1.25–20 nM; IgG-132, 0.31–5 and 0.625–10 nM; and rMoPrP, 3.13–50 and 6.25–100 nM. The dissociation constant was calculated from three independent experiments (mean ± SD). (C) Representative sensorgrams for the binding of authentic and recombinant antibody fragments of mAb 31C6 to rMoPrP by single-cycle kinetics. All sensorgrams were obtained using 2-fold serial dilutions of antibodies or rMoPrP: rFab-31C6, 0.06–1 and 0.13–2 nM; IgG-31C6, 3.13–50 and 6.25–100 pM; and rMoPrP, 0.16–2.5 and 0.31–5 nM.
Table 1.
Summary of equilibrium and kinetic constantsa).
Table 2.
Antigen density-dependent binding of mAb 132.
Fig 7.
Detection of PrPC on the surface of N2a-3 cells.
N2a-3 cells were incubated with DMEM containing 10 nM IgG-31C6 (B, E) and IgG-44B1 (C, F) for 2 days. The cells were stained with 1 μg/ml Alexa Fluor 647-labeled mAb 132 after treatment with (D–F) or without (A–C) 5 M GdnSCN (green). The leftmost images show negative controls for antibody-treatment (A, D). Cell nuclei were stained with DAPI (blue).
Fig 8.
Detection of PrPSc in frozen brain sections.
IFA was performed using the frozen brain sections of Chandler strain-infected (Chandler) and mock-infected (Mock) mice at 120 dpi with (Gdn(+)) or without (Gdn(-)) 2.5 M GdnSCN pretreatment. PrP molecules (green) were stained with mAbs 31C6, 44B1, and 132 followed by Alexa Fluor 488 F(ab’)2 fragment of goat anti-mouse IgG (H+L) (Invitrogen) as a secondary antibody. Anti-feline panleukopenia virus mAb P2-284 was used as a negative control mAb. Cell nuclei were counterstained with DAPI (blue). The fluorescent images were acquired from the coronal sections containing thalamus. Scale bar: 10 μm.
Fig 9.
Mechanism for PrPSc-specific detection by mAb 132.
(A, B) Binding of mAb 132 to rPrP and PrPC molecule. mAb 132 inefficiently binds to PrP molecule if density of PrP molecules is lower than that allows bivalent binding because of low affinity of monovalent binding of mAb 132 (A), whereas mAb 132 binds stably to PrP molecules by avidity effect if the density of PrP molecules is high enough to allow bivalent binding (B). Orange ovals: rPrP or PrPC; black lines on the orange ovals: epitope for mAb 132. Circles indicates the area where bivalent binding of intact mouse IgG1 takes place. (C, D) Binding of mAb 132 to PrPSc oligomer/aggregate. mAb 132 does not bind to PrPSc without guanidinium salt pretreatment (C, Gdn (-)), whereas epitopes for mAb 132 are exposed by the guanidinium salt pretreatment (D, Gdn (+)), multiple epitopes for mAb 132 on the PrPSc oligomer/aggregate allow bivalent binding of mAb 132 (Gdn (+)).