Fig. 6: Structure-based humanization of hTAAB.
a Immunoblot analysis of the relative phosphorylation ratios of Tie2 and Akt after treatment of HUVECs with hTAAB in the form of mouse IgG1 or human chimeric IgG1, IgG2, or IgG4 (1 μg/ml). b Densitometric analyses of the p-Tie2/Tie2 and p-Akt/Akt ratios. Data from five independent experiments (n = 5) were analyzed and expressed as mean ± SD (*P = 0.0257 (p-Tie2/Tie2), 0.0397 (p-Akt/Akt) vs. control, ***P < 0.001 vs. control; ###P < 0.001 hIgG1 vs. hIgG2; $$P = 0.0078 (p-Tie2/Tie2), 0.0013 (p-Akt/Akt) hIgG2 vs. hIgG4). P values by one-way ANOVA test followed by Sidak’s multiple comparisons test. c Quantification of the heterogeneous composition of the Tie2 dimeric mutant/hTAAB IgG subtype complex. The hTAAB IgGs (hIgG1, hIgG2, or hIgG4) were incubated with purified hTie2 Ig3–Fn3 N691C/D682C, followed by size-exclusion chromatography for negative-stain EM analysis. Based on the shape of the complex on the image, we classified the complexes into six categories (linear, <tetragonal, tetragonal, pentagonal, hexagonal, and >hexagonal), and the proportion of each category is indicated (total counted particles n = 1381, 1162, and 1277 for IgG1, IgG2, and IgG4, respectively). d Alignment of sequences of heavy- and light-chain variable regions of hTAAB and humanized TAAB with the closest human germline gene. Dots in human germline and humanized TAAB sequences indicate residues that are identical to parental hTAAB. CDRs of hTAAB and grafted CDRs in humanized TAAB are colored in blue. Back-mutated residues to parental hTAAB are colored in green and are indicated with green triangles. Secondary structure elements (β-strands) are noted above alignment as arrows. e Homology model structure of hzTAAB-H1L1 Fv superimposed on the chimeric hTAAB Fv structure, adapted from the crystal structure of the Tie2 Fn2–Fn3/chimeric hTAAB Fab complex. The heavy and light chains of chimeric hTAAB are colored in dark orange and light orange, respectively, and those of hzTAAB-H1L1 are shown in purple and pink. f–h Rationales for structure-based humanization of the Tie2-activating antibody. Residues in the light chain critical for maintaining VH–VL pairing and CDR conformation (f) as well as affinity for Tie2 Fn3 (g) were back-mutated to the mouse hTAAB sequence. Residues in the heavy chain critical for maintaining CDR conformation were back-mutated to the mouse hTAAB sequence (h). Mutations are indicated with arrows, and residues are colored the same as in (e). i Binding kinetics of humanized TAABs (hzTAAB-H1L1, hzTAAB-H2L1, hzTAAB-H1L2, and hzTAAB-H2L2) for hTie2 ECD measured by SPR analysis. The equilibrium dissociation constant (KD, M) was calculated as the ratio of off-rate to on-rate (koff/kon). Kinetic parameters were determined with the global fitting function of Biacore Insight Evaluation Software using a 1:1-binding model. Vertical dashed lines indicate the start of the dissociation phase.
