IgE-neutralizing UB-221 mAb, distinct from omalizumab and ligelizumab, exhibits CD23-mediated IgE downregulation and relieves urticaria symptoms

Over the last 2 decades, omalizumab is the only anti-IgE antibody that has been approved for asthma and chronic spontaneous urticaria (CSU). Ligelizumab, a higher-affinity anti-IgE mAb and the only rival viable candidate in late-stage clinical trials, showed anti-CSU efficacy superior to that of omalizumab in phase IIb but not in phase III. This report features the antigenic-functional characteristics of UB-221, an anti-IgE mAb of a newer class that is distinct from omalizumab and ligelizumab. UB-221, in free form, bound abundantly to CD23-occupied IgE and, in oligomeric mAb-IgE complex forms, freely engaged CD23, while ligelizumab reacted limitedly and omalizumab stayed inert toward CD23; these observations are consistent with UB-221 outperforming ligelizumab and omalizumab in CD23-mediated downregulation of IgE production. UB-221 bound IgE with a strong affinity to prevent FcԑRI-mediated basophil activation and degranulation, exhibiting superior IgE-neutralizing activity to that of omalizumab. UB-221 and ligelizumab bound cellular IgE and effectively neutralized IgE in sera of patients with atopic dermatitis with equal strength, while omalizumab lagged behind. A single UB-221 dose administered to cynomolgus macaques and human IgE (ε, κ)–knockin mice could induce rapid, pronounced serum-IgE reduction. A single UB-221 dose administered to patients with CSU in a first-in-human trial exhibited durable disease symptom relief in parallel with a rapid reduction in serum free-IgE level.


UB-221 does not bind to the FcԑRI-bound IgE on ELISA
A representative ELISA assay, where the human IgE was absorbed to the plate-immobilized FcԑRI, to show no significant IgE binding by UB-221, c8D6 (human-murine chimerized version of UB-221), omalizumab, and human IgG (negative control) at the indicated concentration range. The 5H2, a mouse anti-human IgE mAb (AbD Serotec) was used as a positive control that caused the strong binding to the FcԑRI-bound IgE. 5 Figure S3. UB-221 does not bind RBL XL-38 cells to trigger allergic degranulation RBL SX-38 is a rat basophilic leukemia cell line expressing the , , and  chains of human FcԑRI. UB-221, c8D6 (human-murine chimerized version of UB-221), and omalizumab were investigated in the cellular binding by flow cytometry and degranulation by assay of hexosaminidase release, where human IgG or trastuzumab (anti-her2 mAb) was used as a negative control, and murine anti-IgE mAb 5H2 or anti-FcԑRI mAb was used as a positive control. (A) On FcԑRI + -RBL SX-38 cells with human IgE pre-loaded, UB-221, c8D6, and omalizumab in a free form did not bind to the cells, while 5H2 and anti-FcԑRI did. (B) On IgE-preloaded FcԑRI + -RBL SX-38 cells, UB-221, trastuzumab, and omalizumab in a free form did not activate the cells to degranulate, while 5H2 did. (C) In a pre-formed IgE immune complex at a 1:1 molar ratio, c8D6:IgE, UB-221:IgE, and omalizumab:IgE did not bind to the FcԑRI + -RBL SX-38 cells and (D) so did not activate the cells to degranulate, while 5H2:IgE did bind the cells and activate degranulation. 6 Figure S4. UB-221 can interact with IgE:CD23 but omalizumab cannot.
Interaction with IgE:CD23 can influence the CD23-mediated functions including regulation of IgE production. (A) On ELISA with recombinant human CD23 immobilized and IgE preabsorbed, UB-221, c8D6 (the human murine chimerized version of UB-221), and 5H2 (a murine anti-IgE mAb as positive control) were found to bind dose-dependently to the IgE, while omalizumab and human IgG (negative control) did not. (B) Analyzed on flow cytometry with CD23-expressing SKW6.4 cells, anti-CD23 mAb (positive control), and UB-221 and c8D6 in a pre-formed complex (UB-221:IgE and c8D6:IgE) at 1:1 molar ratio were shown to bind to the cells, while the omalizumab:IgE complex did not interact with CD23. 7

Figure. S5. Time-dependent IgE protein synthesis in human PBMCs of healthy volunteers
A study model of IgE neo-synthesis in PBMCs was conducted using fresh blood of healthy donors (HD-001 to HD-016) under co-stimulation with IL-4 and an anti-40 antibody. The PBMCs isolated from blood samples were incubated in a serum-free culture medium for indicated days and the detected IgE represents a de novo IgE production derived from the stimulated IgE-producing B cells. The study model exhibits a time-dependent, individualdependent IgE increase starting from Day 7 that plateaued at Days 11-14. IgE was normally detectable about one week after the start of incubation. The IgE concentrations fell between 79.8 ng/ml to 1856 ng/mL, which increased 3.5 to 45 folds in comparison to the IgE concentrations measured on day 7. The levels of total IgE were assayed on ELISA as a measure reflecting the IgE neo-synthesis started from Day 0. There was an apparent betweenindividual difference in the level of IgE production. The method was established for comparative effects of anti-IgE agents on the IgE synthesis, in addition to IgE neutralization.  Supplementary Table 1a, which were obtained using IgE levels of the respective untreated cells set as 100%. Data are shown as Mean±SEM. Different treatments were compared relative to the untreated group using two-way ANOVA with Tukey's multiple comparison referenced to untreated controls. *P < 0.05, **P <0.01, ***P < 0.001.   Tables   Table S1. %Reduction of IgE protein produced in human PBMCs incubated with UB-221, ligelizumab 1 , or omalizumab

Supplementary Methods
Humanization of 8D6. The amino acid sequence of VH and Vκdomains from 8D6 were aligned with those of the human germline VH1-69/JH4 and Vκ1-39/Jκ1 alleles. To make the CDR graft, the acceptor VH framework, which differs from the human germline VH1  Figure 1B).
Binding to the FcԑRI-bound IgE on ELISA. ELISA plates were coated with FcԑRIα-Fc γ at 4°C overnight and blocked with assay diluents for 1 hour at RT. The plates were then saturated with 1 μg/mL human IgE. After washing with wash buffer, the captured IgE was incubated with UB-221 (hu8D6), c8D6, Omalizumab, and a murine anti-human IgE mAb 5H2 (AbD Serotec) at 10, 1, 0.1, and 0.01 μg/mL. The human IgG was detected using HRPconjugated goat anti-human kappa light chain (GeneTex) and HRP-conjugated goat antimurine IgG.Fc (Jackson ImmunoResearch). The color developed with TMB substrate solution was detected at 450 nm on the Molecular Devices microplate reader.
Binding to IgE on RBL SX-38 basophils. RBL SX-38 cells, rat basophilic leukemic cells transfacted with genes encoding the α, β, and g chains of human FcεRI were used as a pool of cell surface FcεRI. RBL SX-38 cells at 2x10 6 /mL in FACS buffer were incubated with IgE at 3 μg/mL for 30 minutes on ice. After washing, cells of 2x10 5 cells in 100 ml FACS buffer were incubated with hu8D6, c8D6, Omalizumab, and 5H2 antibody at 10 μg/mL for 30 minutes on ice, followed by washing. Bound antibodies were detected by FITC-labeled goat IgG specific for human IgG-Fc or FITC-labeled F(ab)'2 rabbit anti-murine IgG (AbD Serotec). The stained cells were analyzed on FACS Canto II. Humanized 8D6 (UB-221), c8D6, and omalizumab could not bind to IgE-saturated RBL SX-38 cells, but 5H2 was bound to RBL SX-38 cells.

Binding of UB-211:IgE complex to and activation of rat huFcԑRI + -RBL SX-38 cells.
RBL SX-38 at 3 x 10 5 cells were seeded in a 24-well plate in 0.5 mL of culture medium overnight in a 37°C incubator. On the next day, each well was washed twice with 0.5 mL of Tyrode's buffer and then 0.25 mL of pre-warmed Tyrode's buffer containing 10 μg/mL of the  ELISA plates were coated with 1 μg/mL of goat anti-human IgM or anti-IgA. Calibration standards were prepared in the culture medium at the range of 15.6 to 1,000 ng/mL for IgM and 7.8 to 5,000 ng/mL for IgA. The captured IgM and IgA were detected with HRP conjugated goat anti-human IgM and anti-IgA antibodies, respectively.
Binding of anti-IgE mAbs to the CD23-bound IgE. The 96-well ELISA plates were coated with 100 μL of CD23 at 5 μg/mL, and the next day, 100 μL of 100 ng/mL IgE was added and incubated for 1 hour at RT. After washing, serially diluted UB-221, omalizumab, or ligelizumab (UBP lot) at 0.0001  100 μg/mL were added and incubated at RT for 1 hour. The binding of the anti-IgE mAbs was detected with an HRP-conjugated goat anti-human IgG Fc (Jackson Immuno Research).

Complement-dependent cytotoxicity (CDC) to CD23-expressing SKW6.4 B-lymphoma cells.
The CD23-overexpressed SKW6.4 B-lymphoma cells at 10 6 cells/mL were incubated and preabsorbed with 800 ng/mL human IgE for 30 min on ice. Excess IgE was washed out and the cell density was adjusted to 2x10 6 cells/mL and 50 µL cell suspension was applied per well into a microplate, to which 50 µL RPMI-1640 medium containing 20% human serum and UB-221 at a concentration range of 0.078 to 10 µg/mL was added. The anti-CD20 rituximab was used as a positive control and omalizumab as a negative control. After incubated at 37 o C for 1 hour, 100 µL CellTiter-Glo 2.0 reagent (Promega, CN: G9242/3) was added to each well and incubated at RT for 10 minutes. The intensity of luminescence was determined by a microplate reader (Molecular Devices, SpetraMax M2). Medium with cells represents the maximal luminescence, and medium without cells represents the basal luminescence. The CDC% was calculated as: