Hepatocellular carcinoma (HCC) is a world leading cause of cancer-related mortality, and currently no curative treatment for advanced HCC is available. Glypican-3 (GPC3) is an attractive target for HCC immunotherapy. This study explored the efficacy of six GPC3-targeted bispecific antibodies, alone or in combination with chemotherapeutic drug Irinotecan, for the treatment of HCC. The bispecific antibodies were constructed using three different structures, knob-into-hole (KH), scFv-scFv-hFc, and scFv-hFc-scFv, where CD3-targeting mAb OKT3 (scFv) was paired with two representative GPC3 mAbs hYP7 (scFv) and HN3 (VH only) that target different epitopes. The In vitro cell killing assay revealed that all bispecific antibodies efficiently killed GPC3 positive cancer cells, with hYP7-KH, hYP7-OKT3-hFc, and HN3-KH being most potent. In vivo xenograft mouse studies demonstrated that all bispecific antibodies suppressed tumor growth similarly, with hYP7-OKT3-hFc performing slightly better. Combination of hYP7-OKT3-hFc with Irinotecan dramatically improved the efficacy and arrested tumor growth of HepG2, Hep3B, and G1 in xenograft mice. Our results demonstrated that the cell surface proximal bispecific antibody hYP7-OKT3-hFc was superior in terms of potency and the GPC3-targeted bispecific antibody combined with Irinotecan was much potent to control HCC growth.

Glypican-3 (GPC3) has become a compelling target for immunotherapy of hepatocellular carcinoma, including antibody-drug conjugate (ADC), and ADC-like immunotoxin. To investigate the impact of epitopes on the potency of ADCs, current study generated a large panel of chicken monoclonal antibodies (mAbs) that targeted 12 different and over-lapping epitopes on GPC3. These mAbs demonstrated a very high affinity with Kd values in the range of 10 -9 -10 -14 M, and the highest affinity (Kd value of 0.0214 pM) was 40-fold higher than the previously generated high-affinity mAb YP7 (Kd value of 0.876 nM). Additionally, these mAbs exhibited excellent thermostability with Tm values in the range of 45-82 °C. As a proof-of-concept study for ADC, we made immunotoxins (scFv fused with PE24, the 24-kDa cytotoxic domain of Pseudomonas exotoxin A) based on these mAbs, and we found that immunotoxins targeting the N-lobe of GPC3 were overall much more potent than those targeting the C-lobe and other locations. One representative N-lobe-targeting immunotoxin J80A-PE24 demonstrated 3 to 13-fold more potency than the hitherto best immunotoxin HN3-PE24 that was previously developed. J80A-PE24 could suppress tumor growth much greater than HN3-PE24 in a xenograft mouse model. Combination of J80A-PE24 with an angiogenesis inhibitor FGF401 showed additive effect, which dramatically shrank tumor growth. Our work demonstrated that, due to high affinity, excellent thermostability and potency, chicken mAbs targeting the N-lobe of GPC3 are appealing candidates to develop potent ADCs for immunotherapy of liver cancer.

No abstract

The coreceptor binding site of gp120 plays an important role in HIV entry into host cell. X5 and 17b are typical coreceptor binding site antibodies with the ability to broadly neutralize HIV. Thus, here, to study the neutralizing mechanism of two antibodies and identify the source of two antibodies with different neutralizing ability, we performed molecular dynamics simulations for the complexes of X5 and 17b with gp120 and CD4. The simulation results indicate X5 and 17b mainly affects CD4 and coreceptor binding sites. Specifically, for CD4 binding site (CD4bs), the binding of antibodies has different effects on CD4bs with and without CD4. However, for coreceptor binding sites, the binding of the antibodies has consistent influence on the region adjacent to loop V3 despite of the simulated systems with or without CD4. The binding of the antibodies enhances the interactions of gp120 region adjacent to loop V3 with other region of gp120, which are unfavorable for conformational rearrangements of the region adjacent to loop V3 and further binding the coreceptor. Additionally, the interactions of loop V3 and bridging sheet with X5 lead to the close motion of loop V3 in X5 bound form, which further influences the rearrangements in gp120.

Purpose: EGFRvⅢ is an established target for immunotherapy of glioblastoma (GBM). Current study aims to explore the efficacy of EGFRvⅢ-targeted immunotoxin combined with temozolomide (TMZ) or T cell-engaged bispecific antibody for the treatment of GBM.Methods: We generated three rabbit monoclonal antibodies (R1, R2, and R6) that specifically bound to EGFRvⅢ, but not EGFR, with high affinity. Immunotoxins were made by fusing the scFv of these antibodies with engineered Pseudomonas exotoxin PE24. The in vitro cytotoxicity and specificity of the immunotoxins was rigorously validated by EGFRvIII and EGFR-expressed cell lines. The in vivo efficacy of immunotoxin monotherapy and in combination with TMZ or EGFRvⅢ-targeted bispecific antibody was evaluated in orthotopic and subcutaneous xenograft mouse models.Results: EGFRvIII immunotoxins potently killed U87, U251 and GL261 cells that were forcefully expressing EGFRvIII, with IC50 values bellow 1.2 ng/ml. In a subcutaneous model, multiple intratumoral injections of immunotoxin at a dose of 2 mg/kg resulted in complete tumor regression in 3/5 of mice. In a C57BL/6 orthotopic glioma model transplanted with GL261 cells that expressed a mouse version of EGFRvIII, two injections of 10 micrograms of immunotoxin in the lateral ventricles significantly improved the survival, with 2/5 mice being completely cured. Furthermore, in a subcutaneous xenograft model transplanted with EGFRvIII-expressed U87 cells, a single intratumoral injection of immuntoxin followed by i.v. injections of TMZ or EGFRvⅢ-targeted bispecific antibody achieved complete regression in mice.Conclusion: EGFRvIII immunotoxins combined TMZ or T cell-engaged bispecific antibody offers promise for curative treatment of GBM.