Anti-PD-1/Her2 Bispecific Antibody IBI315 Enhances the Treatment Effect of Her2-Positive Gastric Cancer through Gasdermin B-Cleavage Induced Pyroptosis

Abstract

The majority of patients with human epidermal growth factor receptor 2 (Her2)-positive gastric cancer develop refractory to Her2-targeted therapy, where upregulation of immune checkpoints plays an essential role. Herein, a recombinant fully human IgG1 bispecific antibody IBI315 targeting both PD-1 and Her2 is developed and its antitumor efficacy as well as the underlying mechanism is investigated. IBI315 crosslinks the physical interaction between Her2-positive tumor cells and PD-1-positive T cells, resulting in significantly enhanced antitumor effects compared to each parent antibody or their combination, both in vitro and in vivo mouse tumor models reconstituted with human immune cells using patient-derived xenografts and organoids. Moreover, IBI315 treatment also induces the recruitment and activation of immune cells in tumors. Mechanistically, IBI315 triggers gasdermin B (GSDMB)-mediated pyroptosis in tumor cells, leading to the activation and recruiments of T cells. The activated T cells secret IFNγ, enhancing GSDMB expression and establishing a positive feedback loop of T cell activation and tumor cell killing. Notably, GSDMB is found to be elevated in Her2-positive gastric cancer cells, providing a rationale for IBI315's efficacy. IBI315 is supported here as a promising bispecific antibody-based immunotherapy approach for Her2-positive gastric cancer in preclinical studies, broadening the therapeutic landscape of this patient population.

Keywords: GSDMB; Her2; bispecific antibodies; immunotherapy; pyroptosis.

Figure 1

IBI315 induces the association of PD‐1 and Her2 and mediates T cell‐mediated cytotoxicity against Her2‐positive gastric cancer cells. A–C) N87 cells were labeled with cell tracker Deep Red and activated T cells were labeled with Carboxyfluorescein succinimidyl ester (CFSE). They were cocultured at a 1:1 ratio and treated with IBI315, its parent antibody, or their combination for 30 min. The association rate between N87 cells and activated T cells was analyzed using flow cytometry. A) A schematic model depicts the association of IBI315 with PD‐1 on T cells and Her2 on tumor cells. B) Representative flow cytometry images show the association of N87 cells (stained with cell tracker Deep Red) and activated T cells (stained with CFSE) treated with T+P or with IBI315. C) Quantification of associated cells (cells in the upper right quadrant of the scatter plot) with the indicated treatments is presented based on flow cytometry analysis. D,E) The impact of IBI315, parental PD‐1 antibody, parental trastuzumab, and the combination of parental drugs on the secretion of D) IFNγ and E) IL‐2 by human T cells was investigated using a mixed lymphocyte reaction (MLR). F–I) Her2‐positive gastric cancer cells (N87 and SNU‐216) were treated with IBI315, its parent antibody, or their combination for 24 h in the presence of T cells. Representative images of F) the N87/T cell and G) SNU‐216/T cell coculture systems are shown, with the white arrow indicating T cells associated with tumor cells. Quantification of tumor cell death, measured by LDH release in H) the N87/T cell and I) SNU‐216/T cell coculture systems, is presented for the indicated treatments. Scale bar 50 µm. All statistical results were obtained from three independent experiments and expressed as means ± SEM. Abbreviations: TZB: trastuzumab, αPD1: anti‐PD1 antibody (sintilimab), T+P: trastuzumab+sintilimab. Statistical significance is indicated as **P < 0.01 and ***P < 0.001.

Figure 2

IBI315 exhibits potent antitumor efficacy in Her2‐positive gastric cancers and organoids. A–H) Human immune cell‐reconstituted NOG mice bearing N87 tumors or patient‐derived xenografts (PDX‐1) were subjected to treatment with IBI315, its parent antibody, or their combination. A) Photographic documentation of N87 tumors on day 31 and B) the corresponding average tumor volumes were obtained. C) Additionally, the change in body weight for each treatment group was monitored. D) Immunohistochemical analysis revealed representative images of CD3⁺, CD8⁺, and CD4⁺ T cell infiltration within N87 tumors for each treatment group, E) which were further quantified. F) Histopathological examination of PDX‐1 included H&E and Her2 staining, along with G) photographic documentation on day 39 and H) assessment of average tumor volumes. I–K) A coculture system of organoid PDXO‐1 and T cells was established to evaluate tumor cell death induced by IBI315, its parent antibody, or their combination, as determined by measuring LDH release. I) Histological analysis of organoid PDXO‐1 included H&E staining, Her2 staining, and bright field (BF) imaging. J) Representative images of the PDXO‐1/T cells coculture system under indicated treatments were captured, with a specific focus on IBI315‐treated PDXO‐1, showcasing the association between T cells and PDXO‐1. K) The extent of PDXO‐1 cell death was quantified by measuring LDH release after a 24‐hour incubation with T cells under indicated treatments. Scale bar 50 µm. All statistical results were obtained from three independent experiments and expressed as means ± SEM. Abbreviations: TZB: trastuzumab, αPD1: anti‐PD1 antibody(sintilimab), T+P: trastuzumab+sintilimab. Statistical significance is indicated as **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 3

IBI315 induces pyroptosis in Her2‐positive gastric cancer cells in the presence of T cells. A–C) N87 or SNU‐216 cells were prelabeled with CFSE and cocultured with T cells in the presence of IBI315, its parent antibody, or their combination for 6 h. Fluorescent images were captured to illustrate the morphological changes of tumor cells, and the levels of IL‐18 in the supernatant of the tumor cell/T cell coculture system were measured. A) Representative images of N87 or SNU‐216 cells with arrowheads indicating pyroptotic cells. B) The quantification of pyroptotic cells following treatment with the indicated antibodies is presented in a column chart. C) The levels of IL‐18 in the supernatant of the tumor cell/T cell coculture system under the indicated treatments are shown. D–G) The expression of GSDMB in N87, SNU‐216 cells, and PDXO‐1 was examined and its cleavage was induced by IBI315 in the presence of T cells for 6 h. D) Coexpression of Her2 (green) and GSDMB (red) in N87 cells was visualized by fluorescent staining. E) GSDMB expression in PDX‐1 and PDXO‐1 was assessed by immunohistochemical staining. F) Immunoblotting was performed to evaluate the expression of Her2 and GSDMB in N87 and SNU‐216 cells. G) Immunoblotting was also conducted to assess GSDMB cleavage in the tumor cell/T cell coculture system treated with IgG or IBI315 for 6 h. H–J) The expression pattern of granzyme A, an enzyme involved in GSDMB cleavage‐induced pyroptosis, was evaluated in tumor‐infiltrating lymphocytes (TILs) from tumors treated with IBI315, its parent antibody, or their combination (following the same administration protocol as in Figure 2). H) Immunohistochemical staining revealed the expressions of CD3, CD8, CD4, and granzyme A at the same site in TILs of N87 tumors treated with IBI315. I) The PDX‐1 tumors were digested into single cells and the TILs were analyzed. The proportion of granzyme A⁺ cells in CD8⁺ TILs was determined by flow cytometry, with representative graphs depicting IgG or IBI315 treatment in PDX‐1, and their quantification in each group. J) Immunohistochemical staining images and quantification of granzyme A⁺ TILs in N87 tumors for each treatment group. Scale bar 100 µm. All statistical results represent the means ± SEM of three independent experiments. Abbreviations: TZB: trastuzumab, αPD1: anti‐PD1 antibody (sintilimab), T+P: trastuzumab+sintilimab. Statistical significance is indicated as **P < 0.01, ****P < 0.0001.

Figure 4

The tumoricidal effect of IBI315 depends on GSDMB‐cleavage‐induced pyroptosis. GSDMB expression in N87 and SNU‐216 cells was downregulated using siRNA (siGSDMB‐1 and siGSDMB‐2). The cells were then labeled with CFSE and cocultured with T cells in the presence of either IBI315 or IgG. Fluorescent imaging was performed to visualize the morphological changes associated with pyroptosis in tumor cells with or without GSDMB knockdown (±GSDMB) and the levels of IL‐18 in the supernatant of the tumor cell (±GSDMB)/T cell coculture system were measured. Anti‐GSDMB immunoblotting was conducted in A) N87 and C) SNU‐216 cells before and after treatment with siGSDMB‐1 and siGSDMB‐2 to evaluate the efficacy of siRNA. Representative images of B) prelabeled N87 or D) SNU‐216 cells in the tumor cell (±GSDMB)/T cell coculture system were captured, with arrowheads indicating pyroptotic cells. E,F) Quantification of pyroptotic cells and LDH release was performed. G) The levels of IL‐18 in the supernatant of the tumor cell (±GSDMB)/T cell coculture system under the indicated treatments were measured. H) GSDMB knockdown in N87 cells was achieved using shRNA (N87 Sh‐GSDMB). NOG mice reconstituted with human immune cells and bearing N87 or N87 Sh‐GSDMB tumors were treated with IBI315 or IgG. Anti‐GSDMB immunoblotting was conducted in N87 cells, and GSDMB and Her2 staining were performed in N87 tumors to confirm the efficacy of shGSDMB. I) Photographic documentation of N87 tumors on day 29 and J) the corresponding average tumor volumes were obtained. K) Immunohistochemical analysis revealed representative images of CD3⁺, CD8⁺, and CD4⁺ T cell infiltration within N87 tumors for each treatment group, L) which were further quantified. All statistical results were obtained from three independent experiments and expressed as means ± SEM. Scale bar 100 µm. Statistical significance is indicated as ns P > 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 5

IBI315‐mediated tumor cell pyroptosis activates lymphocytes and triggers a positive loop of tumor inhibition. A–F) GSDMB expression in N87 and SNU‐216 cells was downregulated using siRNA (siGSDMB‐1 and siGSDMB‐2). The cells, with or without GSDMB knockdown (±GSDMB), were then cocultured with T cells in the presence of either IBI315 or IgG for 24 h and the supernatant was collected and used as conditioned media. Peripheral blood mononuclear cells (PBMCs) were cultured in the corresponding conditioned media for 24 h and the activation of immune cells was analyzed. A,B) Representative images and C,D) quantification of the CD25⁺ T cell population within CD8⁺ T cells of PBMCs cultured in the conditioned media from A,C) N87 (±GSDMB) or B,D) SNU‐216 (±GSDMB) cells/T cell coculture system with the indicated treatments were shown. IFNγ secretion by PBMCs treated with the conditioned media from E) the indicated N87 or F) SNU‐216 cells/T cell coculture system was measured. G–O) Human immune cell‐reconstituted NOG mice bearing PDX‐1 were subjected to treatment with IBI315, its parent antibody, or their combination (following the same administration protocol as in Figure 2). On Day 39, mice were sacrificed and tumors were collected and digested into single cells. The activation of TILs in PDX‐1 of each treatment group was analyzed by flow cytometry. Representative images and quantification of the indicated TIL populations in PDX‐1 from each group were presented. All statistical results were obtained from three independent experiments and expressed as means ± SEM. Statistical significance is indicated as *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6

A schematic model depicting the mechanism underlying the antitumor effect of IBI315‐mediated tumor cell pyroptosis. IBI315 facilitates the formation of immune synapses between T cells and Her2‐positive tumor cells. Granzyme A might cleave GSDMB in tumor cells to generate the N‐terminal fragment, which forms Gasdermin transmembrane pores in the cell membrane and triggers pyroptosis and release of inflammatory cytokines from tumor cells. These cytokines can activate and recruit T cells and the IFNγ secreted by activated T cells can further increase the expression of GSDMB as well as PD‐L1 in tumor cells. The elevated expression of GSDMB in tumor cells enhances the occurrence of IBI315‐mediated cell pyroptosis. The blocking effect of IBI315 on PD‐1 of T cells also inhibits the interaction between PD‐1 and PD‐L1. The PD‐1 blocking function of IBI315 enhances T cells activation and induces pyroptosis in more tumor cells, thereby forming a positive feedback loop of T cell activation and tumor cell pyroptosis. All components of the mechanism are labeled in the figure for easy understanding.