A Novel Bispecific Anti-IL17/VEGF Fusion Trap Exhibits Potent and Long-Lasting Inhibitory Effects on the Development of Age-Related Macular Degeneration

Abstract

Age-related macular degeneration (AMD) is a severe eye disease in people aged 60 years and older. Although anti-VEGF therapies are effective in treating neovascular AMD (NvAMD) in the clinic, up to 60% of patients do not completely respond to the therapies. Recent studies have shown that blood-derived macrophages and their associated proinflammatory cytokines may play important roles in the development of persistent disease and resistance to anti-VEGF therapy. To address this issue, we constructed an antibody-based bispecific fusion protein that can simultaneously inhibit IL-17-induced inflammation and VEGF-mediated neovascularization. As a result, the bispecific fusion protein 17V05 effectively inhibited multiple proinflammatory cytokines and chemokines, as well as laser-induced choroidal neovascularization (CNV). More importantly, 17V05 also exhibited stronger and longer inhibitory effects than conbercept in vivo. Thus, we provide a novel and promising strategy for treating AMD patients who are not sensitive to anti-VEGF therapies.

Figure 1

The construction of the bispecific fusion protein 17V05. (a) Schematics of the 17V05 fusion protein structure. (b) SEC showed that both 17V01 and 17V05 have monomeric contents over 95% after one-step affinity purification. (c) Mass spectrometry showed that there is cleavage between M464/Y465 in D2 of the bispecific fusion trap. The 17V05 protein with an M464A mutation prevented cleavage at this site. (d) DSC showed that the T_onsets of 17V01 and 17V05 were 66.16°C and 65.83°C and the T_m1 values were 79.54°C and 79.58°C, respectively. DSC: differential scanning calorimetry, T_onsets: melting temperature at onset, T_m1: melting temperature peak 1, HC: heavy chain, and Da: Dalton.

Figure 2

The 17V05 protein simultaneously bound to IL-17A and VEGF as strongly as its parental mAbs. (a) The binding affinity of 17V05 for IL-17A was measured in duplicate by ELISAs and compared to that of the parental anti-IL17A mAb 608. Bevacizumab (Avastin) was used as a negative control (n = 2). (b) The ability of 17V05 to bind to VEGF was examined in duplicate by ELISAs and compared to that of FC-D2 (D2 was fused to the N-terminus of an Fc). mAb 608 was used as a negative control (n = 2). (c) The equilibrium dissociation constants (KDs) of 17V05 were obtained and compared to those of the parental mAbs 608 and bevacizumab (n = 3). N = the number of independent experiments. ELISA: enzyme-linked immunosorbent assay; KD: the equilibrium dissociation constant; Ka: the association constant; Kd: the dissociation constant.

Figure 3

Treatment with 17V05 potently inhibited the production of VEGF by RPE cells. (a) Treatment with 17V05 blocked the interaction of IL-17A with IL-17RA as effectively as 608 in an ELISA-based assay (n = 3). (b) The 17V05 protein more effectively (∼3-fold) blocked the signaling of VEGF through KDR than bevacizumab in a luciferase-reporter cell-based assay. In contrast, 608 failed to interfere with the signaling (n = 3). (c) The 17V05 protein was 9-fold more potent than bevacizumab in inhibiting the production of VEGF by RPE cells, while 608 slightly inhibited VEGF secretion (n = 3). All assays were performed in duplicate. KDR: kinase insert domain receptor (VEGFR2); RPE: retinal pigment epithelium.

Figure 4

Treatment with 17V05 potently inhibited the production of multiple proinflammatory cytokines and chemokines by RPE cells. (a) Treatment with 17V05 inhibited the secretion of IL-6 from RPE cells as effectively as 608, whereas bevacizumab had no effects (n = 3). (b) Treatment with 17V05 inhibited the secretion of IL-8 from RPE cells as effectively as 608, whereas bevacizumab had no effects (n = 3). (c) The 17V05 protein inhibited the secretion of CXCL1 from RPE cells as effectively as 608, whereas bevacizumab had no effects (n = 3). (d) The 17V05 protein inhibited the secretion of MCP-1 from RPE cells as effectively as 608, whereas bevacizumab had no effects (n = 3). All assays were performed in triplicate. IL-6: interleukin-6, IL-8: interleukin-8, CXCL1: the chemokine (C-X-C motif) ligand-1, and MCP-1: monocyte chemoattractant protein-1.

Figure 5

Treatment with 17V05 exhibited more potent inhibitory effects on the development of AMD than aflibercept in the laser-induced CNV model. (a) Pictures of eyes (8 mice per group, total 16 eyes per group) with laser-induced CNV before (upper panel) and 7 days after dosing with 17V05, aflibercept, and PBS (lower panel) (n = 2). Top two panels are FFA pictures before and after dosing, and the bottom two panels are corresponding IB4 staining (red) and DAPI staining (blue). (b) The percentage of changes in total leakage areas in the eyes of mice treated with 17V05, aflibercept, and PBS. Formula = (Leakage Areas_{at day 7-after-dosing} − Leakage Areas_{at day 0-before-dosing})/Leakage Areas_{at day 0-before-dosing} × 100. (c) Neovascular blood vessels in the retina were stained with isolectin GS-IB4 after being treated with 17V05, aflibercept, or PBS. ^∗∗ depicts p < 0.01, ^∗ depicts p < 0.05, and ns = not significant. IB4: isolectin GS-IB4; DAPI: 4′,6-diamidino-2-phenylindole.

Figure 6

XAB4-D2, a 17V05 surrogate, exhibited durable and potent inhibitory effects on the development of ocular neovascularization in a persistent retinal neovascularization (RNV) rat model. (a) Pictures of eyes (8 rats per group) treated with PBS (upper panel), conbercept (middle panel), and XAB4-D2 (lower panel) on Day 21 post-treatment (n = 2). (b) Upper panel: a schematic depicts the dosing schedule and the timepoints for taking FFA. Bottom panel: the leak scores were measured on Day 14 (middle) or Day 21 (right) after treatment with PBS, conbercept, and XAB4-D2. The leak score measured on Day 7 (left) was shown as baselines. ^∗∗ depicts p < 0.01; ^∗ depicts p < 0.05.