Figure 5. Anti-LigB mAb bactericidal properties.

(A, B) Leptospira survival assays were used to assess mAb bactericidal activity. (A) After incubation with individual mAbs, the survival rate of L. interrogans Pomona was observed using dark-field microscopy (B) while the survival rate of bioluminance-producing L. interrogans Manilae was measured using a 96-well luminometer-based assay (full dataset Figure 5—figure supplement 1). Each value represents the mean ± S.D. from three individual trials of two replicates. Statistically significant (t-test; p<0.05) differences were calculated from the comparisons between mAb-treated groups and the control group (*) or between mAb-treated groups and the pAb group (#). (C) A schematic is shown for the SAXS-derived LigB structure. (D) For the set of mAbs, the ELISA antigen binding value is plotted against the LD₅₀. Points for mAbs are colored to match the three domain regions defined in the schematic. Statistically-determined outliers are labelled and data was fit with and without outliers (dashed line, R²=0.508 and solid line, R²=0.773, respectively). (E) For mAbs measured for cell surface binding, the FACS cell binding value is plotted against the LD₅₀ and data is fit with and without outlier, mAb C23 (dashed line, R²=0.312 and solid line, R²=0.466, respectively).

Figure 5—figure supplement 1. Serum bactericidal activity characterization of anti-LigB mAb library.

(A) Time-dependence of mAb serum bactericidal activity. Bioluminance-producing L. interrogans Manilae were incubated with individual mAbs at 100 μg/ml. The leptospiral survival rates in the presence of mAbs and human serum complements were measured using a luminometer at different time points (60, 90, and 120 min). Negative controls were measured in the presence of no antibody (control) and an isotypic mouse IgG (unrelated Ab). Survival measurements for hamster anti-LigB polyclonal antibodies (anti-LigB pAb) are also shown. The data were fit using a standard logistic equation. (B) The median lethal time (LT₅₀) was calculated for individual mAbs from the fitted logistic curve using Origin software. (C) Dose-dependence of mAb serum bactericidal activity. Bioluminance-producing L. interrogans Manilae were incubated with individual mAbs at various mAb concentrations (25, 50, and 100 μg/ml). The leptospiral survival rates in the presence of mAbs and human serum complements were measured at 90 min. The data were fit using a dose inhibition equation using Origin software. (D) The median lethal dose (LD₅₀) was calculated for individual mAbs from the fitted dose inhibition curve.

Figure 5—figure supplement 2. Statistically determined outliers for scatter plots.

Cook’s distance (D_i) demonstrates the influence of a point on a defined correlation. For the linear correlations of LOG(K_D) vs. LD₅₀ (Figure 5D) and FACS Cell Binding vs. LD₅₀ (Figure 5E), D_i was determined for each data point. D_i values above the threshold of 4/n (red) deviate from the expected relationship between binding and bactericidal activity and were considered to be outliers.

Figure 5—figure supplement 3. Binding of library C mAbs to LigB7-12.

Library C mAbs were generated from hybridoma cell lines exposed to LigB1-7. The ability of library C mAbs to bind LigB7-12 was also assessed by colorimetic (λ₆₃₀) TMB-ELISA. LigB7-12 binding was identified for C12 and V10 (positive control). Each value represents the mean ± S.D. from two individual trials of two replicates.

Figure 5—figure supplement 4. Comparison of bactericidal activity values, LT₅₀ and LD₅₀.

(A) The Cook’s Distance (D_i) was determined for the linear correlation of LT₅₀ and LD₅₀ (R² = 0.867). C23 (red) was the only mAb with a D_i value above the outlier threshold. (B) Plots of K_D vs. LD₅₀ and K_D vs. LT₅₀ are overlaid and centered on fits with C12 and C23 removed (solid line: LD₅₀R² = 0.773, dashed line: LT₅₀R² = 0.872). The LT₅₀ value of C23 (open red) is shifted farther away from the expected value than the LD₅₀ value of C23 (solid red).