**Background:** Neovascular age-related macular degeneration (nAMD), characterized by choroidal neovascularization (CNV), is a leading cause of vision loss. While anti-VEGF antibody therapy is the standard treatment, 5–15% of patients show poor or no response, and resistance can develop over time. The Notch signaling pathway, particularly the ligand Jagged1, is known to promote angiogenesis and inflammation. This study investigates whether antibody-mediated blockade of Jagged1 can attenuate CNV, either alone or in combination with anti-VEGF.
**Methods:** CNV was induced in female C57BL/6J mice (8–10 weeks old) by laser photocoagulation (532-nm laser, 80 mW, 50 µm spot, 0.1 s duration). A phage-display selected mouse IgG2a monoclonal antibody against Jagged1 (anti-JAG1.b70, KD 6 nM for mouse, 0.79 nM for human Jagged1) was administered intraperitoneally at 5 mg/kg on days 0 and 5 post-laser. Controls included isotype control IgG, anti-VEGF (B20-4.1.1), and a hamster anti-Jagged1 antibody (HMJ1-29). Vascular leakage was assessed by fundus fluorescein angiography (FFA) at days 5 and 10. CNV lesion size was quantified by ICAM-2 immunofluorescence on RPE-choroid-sclera flat mounts at day 10. IBA-1-positive mononuclear phagocytes were counted and classified as ramified or ameboid. Retinal cytokine/chemokine levels were measured by multiplex assay (31-plex) at day 4. DLL4 and Jagged1 expression were analyzed by confocal microscopy. Plasma half-life was determined over 264 hours. Effector-negative PGLALA variants (P329G, L234A, L235A) were generated and tested. In vitro, HUVECs were stimulated with VEGF (0, 10, or 100 ng/mL) and Jagged1/DLL4 signaling effects on VEGFR2 expression were assessed by Western blot.
**Key Results:** Anti-JAG1 treatment significantly reduced vascular leakage at day 10 (p < 0.05, unpaired t-test) and reduced CNV lesion area by approximately 40% compared to control IgG (p < 0.01). A similar reduction was observed with the HMJ1-29 antibody. Anti-VEGF alone reduced lesion size by 80%. Jagged1 blockade resulted in a 60% reduction in total IBA-1-positive cells within lesions (p < 0.01), a 2.0-fold increase in the fraction of ramified (resting) cells, and a 3.9-fold decrease in ameboid (activated) cells. Retinal concentrations of CX3CL1, CXCL16, and TNF-α were reduced after anti-JAG1 treatment. In the light-induced retinal degeneration model, anti-JAG1 showed no significant effect on outer nuclear layer thickness or TUNEL-positive cells. Jagged1 expression in CNV lesions was not altered by antibody treatment. Importantly, Jagged1 blockade led to a 7-fold upregulation of DLL4 in endothelial cells within CNV lesions (p < 0.05). In vitro, DLL4 signaling reduced VEGFR2 levels by approximately 70% in the presence of VEGF. Plasma half-life of anti-JAG1 at 5 mg/kg was 4.5 days (with CNV) vs. 4.4 days (without CNV), comparable to anti-VEGF (4.9 and 5.1 days, respectively). At 0.5 mg/kg, anti-JAG1 half-life was ~1.5-fold shorter than anti-VEGF. Combined treatment with 5 mg/kg anti-JAG1 and 0.5 mg/kg anti-VEGF reduced lesion size by ~86%, compared to ~53% for anti-VEGF alone and ~73% for anti-JAG1 alone (p < 0.01 for combination vs. either alone). The PGLALA effector-negative variant of anti-JAG1 showed a non-significant reduction in CNV size compared to wild-type, indicating the therapeutic effect is dependent on target blockade, not Fc-mediated effector functions.
**Clinical Implications:** This study identifies Jagged1 as a novel therapeutic target for nAMD. The finding that Jagged1 expression is independent of VEGF signaling supports its potential utility in patients who are non-responsive or resistant to anti-VEGF therapy. The additive effect of combined Jagged1 and VEGF blockade suggests that dual targeting could enhance treatment efficacy. The demonstration that the therapeutic effect is solely dependent on target blockade (not effector functions) allows for engineering of effector-silenced antibodies, potentially reducing Fc-mediated adverse events such as platelet aggregation observed with some anti-VEGF agents. The absence of toxicity to retinal vascular tissue and the lack of effect on pericyte coverage or retinal vascular morphology further support the safety profile. However, the authors note that translation to clinical practice requires testing in larger animal models (e.g., non-human primates) with intravitreal administration, as the current study used IP injections in mice.