Fig 1.
The modified plasmid pYD4 and the library sorting strategy.
A. Map of pYD4 plasmid. B. The poly linker of pYD4. Compared with the yeast display vector pYD2, a new linker with 20 amino acids different from the (G4S)3 linker upstream of the scFv gene was built in the plasmid pYD4, flanked by restriction enzyme sites for cloning of VH or VL genes individually, rather than as a scFv. An HA tag was introduced between the (G4S)3 linker and scFv as an alternative detection marker besides the SV5 tag downstream of the final scFv construct. C. Dot plots of the results of a typical three-round FACS sorting of a scFv library constructed in pYD4. For each round of FACS sorting, fluorescence-conjugated secondary antibody bound to BoNT/A LC (50 nM) was used to stain yeast libraries BoNT/A LC binding is shown on the Y-axis and the scFv display level on the X-axis after staining with anti-SV5 tag antibody. The yeast population with BoNT/A LC bound was collected by using the indicated gate settings shown for the yeast labeled green in the dot plots.
Table 1.
Yeast display libraries constructed and used for BoNT/A LC mAb generation.
Table 2.
Binding and endopeptidase inhibition characteristics of BoNT/A LC specific scFv mAbs.
Table 3.
KD values of IgGs measured using KinExA 1
Fig 2.
Classification of scFv epitopes on BoNT/A LC based on ability of scFv to bind simultaneously.
Soluble BoNT/A LC was captured by yeast-displayed scFv and the ability of phage-displayed scFv to bind BoNT/A LC was determined by flow cytometry. The 19 scFv were clustered into 4 groups (I–IV), 14 of which bound epitope I.
Fig 3.
mAb inhibition of BoNT/A LC endopeptidase activity.
A. SDS-PAGE-based substrate cleavage assay: BoNT/A LC (25 nM) and 20 molar excess of mAb were mixed in Tris buffer (50 mM, pH 8.0). GST-SNAP-25 (141–206) peptide substrate (5 μM) was added to initiate the reaction. Images of Coomassie-stained SDS-PAGE gels of the results after 5 min (the upper panel) or 15 min (lower panel) incubation with the intact substrate and cleaved product indicated by arrows. Ability to inhibit SNAP-25 cleavage was scored as positive (+) or negative (-). Additional bands in the SDS-PAGE gel likely represent GST-SNAP25 breakdown products or impurities from the mAbs. B. Plot of the results from the FRET screening assay for inhibition of substrate cleavage. The YsCsY substrate (2 μM) was mixed with each of the indicated mAbs (200 nM) and BoNT/A LC. (400 pM). The mean (± standard deviation) of the ratio of emissions at 527 nm to 480 nm after 5 min or 15 min are shown. Ratios >0.8 at 15 minutes were interpreted to indicate inhibition of BoNT/A-mediated cleavage by the mAb, denoted by (+). The epitope clusters (I-IV, Fig 2) are shown below each scFv. Nine mAbs in epitope cluster I (1C10, 1D8 1G11, 1H5, 10B12, 10C9, 10F9, 10H10 and 11D8) strongly inhibited BoNT/A LC cleavage.
Table 4.
IC50 and KD values of IgGs.
Fig 4.
Inhibition of SNAP25 cleavage by mAbs in neuronal cells.
Upper panel: Representative western blots (WB) of inhibition of BoNT/A-mediated SNAP25 cleavage by mAbs, indicating that 10 nM BoNT/A1 cleaved more than 50% of SNAP25 in Neuro-2a cells. Lower panel: Semi-quantitative analysis of the WB results. The experiments were repeated four times, and the percentage of non-cleaved SNAP25 was determined, shown as the mean ± SD. Compared with BoNT/A1 treated groups, CR2 (97.6 ± 0.6 vs. 28.2 ± 12.5, *P = 0.00003) and ING2 (65.7 ± 15.1 vs. 28.2 ± 12.5, *P = 0.0086) significantly reduced SNAP25 cleavage, while 5A20.4 did not (34.9 ± 18.9 vs. 28.2 ± 12.5, P = 0.58. Statistical significance was determined using the Holm-Sidak method for multiple t-tests with alpha = 5%, and no assumption of a consistent standard deviation using Graphpad Prism version 6.0 (La Jolla, CA).
Fig 5.
Fine epitope mapping of BoNT/A LC mAbs by alanine-scanning mutagenesis.
(A-F) Amino acids near the mAb-binding sites on BoNT/A LC for representative mAbs (10F9, 10B4, and 5A20.4) or BoNT/A LC-HN for representative mAbs (7C8, 12A11 and ING2) were mutated individually to alanine, and displayed on yeast (S2 Table). The KD values of each mAb binding to the mutants and the wild type fragment were determined in triplicate. Epitopes modeled on the surface of the crystal structure of BoNT/A (pdb ID 3BTA) for 7C8, 12A11, ING2 and 5A20.4, or BoNT/A LC–SNAP25 complex (pdb ID 1XTG) for 10F9 and 10B4 using Pymol software. The ΔΔG for each pairing was then determined to evaluate the contribution of each amino acid in the epitope. (G) Maps of each of the epitopes on the surface of the BoNT/A LC structure. The differential contributions of amino acids in the epitope were colored using a gradient from red (denoting the greatest ΔΔG) to beige (denoting the smallest ΔΔG). *the maximum ΔΔG of each mAb as shown in S6 Table (7C8, ΔΔGmax = 1.0; 12A11, ΔΔGmax = 1.7; ING2, ΔΔGmax = 1.9; 10F9, ΔΔGmax >3.5; 10B4, ΔΔGmax = 3.6; 5A20.4, ΔΔGmax = 3.4).