Antitumor efficacy and reduced toxicity using an anti-CD137 Probody therapeutic

Abstract

Costimulation via CD137 (4-1BB) enhances antitumor immunity mediated by cytotoxic T lymphocytes. Anti-CD137 agonist antibodies elicit mild liver inflammation in mice, and the maximum tolerated dose of Urelumab, an anti-human CD137 agonist monoclonal antibody, in the clinic was defined by liver inflammation-related side effects. A protease-activated prodrug form of the anti-mouse CD137 agonist antibody 1D8 (1D8 Probody therapeutic, Pb-Tx) was constructed and found to be selectively activated in the tumor microenvironment. This construct, which encompasses a protease-cleavable linker holding in place a peptide that masks the antigen binding site, exerted antitumor effects comparable to the unmodified antibody but did not result in liver inflammation. Moreover, it efficaciously synergized with both PD-1 blockade and adoptive T-cell therapy. Surprisingly, minimal active Pb-Tx reached tumor-draining lymph nodes, and regional lymphadenectomy did not abrogate antitumor efficacy. By contrast, S1P receptor-dependent recirculation of T cells was absolutely required for efficacy. The preferential cleavage of the anti-CD137 Pb-Tx by tumor proteases offers multiple therapeutic opportunities, including neoadjuvant therapy, as shown by experiments in which the Pb-Tx is given prior to surgery to avoid spontaneous metastases.

Keywords: 4-1BB; CD137; Probody; cancer immunotherapy.

Fig. 1.

An anti-mCD137 Pb-Tx that is activated by protease digestion in the tumor microenvironment. (A) Schematic representation of the strategy for functional tumor targeting by protease activation of an anti-mouse CD137 Pb-Tx in the tumor microenvironment. (B) Binding to immobilized mouse recombinant CD137 protein of matriptase-digested or undigested native 1D8 and 1D8 Pb-Tx determined by ELISA. (C) FACS analysis testing binding of the 1D8 Pb-Tx to anti-CD3 + CD28 activated mouse CD8 or CD4 T splenocytes following digestion with recombinant matriptase (blue). Unmasked 1D8 is used as a positive control as indicated (green). (D) Similar experiments as in C, in which CD45.1⁺-activated splenocytes were stained with 1D8 Pb-Tx or native antibody, precultured overnight at 37 °C with cell suspensions obtained from MC38-engrafted tumors to mimic the tumor tissue microenvironment. Data in C and D are representative of at least three independent experiments in triplicate wells.

Fig. 2.

1D8 Pb-Tx exerts antitumor effects without liver inflammation. (A) Treatment of BALB/c mice bearing 6-d established subcutaneous tumors derived from the syngeneic CT26 colon cancer cell line, treated when indicated by dotted lines with mIgG1 control antibody, 1D8, or the 1D8 Pb-Tx. (B) Long-term survival of mice in A. Data in A and B are representative of at least two independent experiments with n = 6 mice per group. Survival differences were analyzed by log rank tests. *P < 0.05. (C) Circulating ALT in serum from treated mice on day +18 in experiments as in A. (D) CD8⁺ T-cell content in the liver referred as total cells, assessed in the liver cell suspensions, 7 d following treatments as in A. (C and D) Data are representative of at least two independent experiments with n = 6 mice per group and expressed as mean ± SEM. (E) Representation of experimental setup for in vivo imaging of the surgically exposed livers of hCD2-RFP transgenic mice pretransferred with activated anti-OVA GFP⁺ OT1 T cells and bearing established B16OVA tumors. (F) Intravital multiphoton representative frame images of the surgically exposed livers. Mice were treated with three doses of mIgG1 control antibody, 1D8, or Pb-Tx antibodies as indicated. Endogenous T cells are visualized in red, transferred OT1 lymphocytes in green, and sinusoid vessels in blue upon in vivo staining by perfusion of a fluorescence-labeled anti-CD31 mAb. (G and H) Quantification of extravascular T-cell clusters in multiple videos (Movie S1). Data in F, G, and H are representative of two independent experiments with n = 3 mice per group. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 3.

The 1D8 anti-CD137 Pb-Tx is activated in the tumor and fails to functionally reach tdLNs, which are not necessary for efficacy. (A) An AF488 conjugated version of the 1D8 antibody or the Pb-Tx were given to mice bearing B16OVA tumors. Tumor, spleen, and tdLNs cell suspensions were studied by FACS to comparatively detect binding to CD4 and CD8 T lymphocytes. (B) OT-1 CD45.1⁺ T cells activated with CD3 + CD28 to express high-surface CD137 levels. (C) OT1 T cells as in B were injected to B16OVA tumor-bearing mice that were intraperitoneally given AF488-labeled versions of 1D8 antibody or the Pb-Tx as indicated. Cell suspensions from tumor, tdLNs, and spleen were analyzed by FACS for CD137 staining on CD45.1⁺ gated cells. (A–C) Data are representative of at least two independent experiments with n = 6 mice per group and expressed as mean ± SEM (D) Similar experiments as in A performed in Her2/NeuT female transgenic mice bearing spontaneously developed breast carcinomas. Data are shown as mean ± SEM with n = 15 mice per group. (E) Similar experiments as in C, where Violet track–marked CD3 + CD28 activated polyclonal T cells were adoptively transferred to Her2/NeuT female mice bearing spontaneously developed breast tumors to provide T cells readily expressing CD137. Data are expressed as mean ± SEM with n = 6 mice per group. (F) Treatment with control antibody, 1D8, or 1D8 Pb-Tx of locally lymphadenectomized mice on day +6 after engraftment with the CT26 cell line. Mice undergoing mock surgery were used as a control. (G) Treatment of 6-d established subcutaneous tumors derived from CT26 cell line treated with S1P inhibitor FTY720 given daily from day +5. Control antibody, 1D8, or the Pb-Tx treatment days are indicated by dotted lines. (F and G) Data are represented as individual tumor growth with n = 6 mice per group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 4.

The 1D8 anti-CD137 Pb-Tx in combination with anti-PD1 blockade effectively protects against 4T1 lung metastasis in a neoadjuvant treatment setting. (A) Schematic representation of the in vivo experiments in 4T1 mCherry breast cancer model treated with anti–PD-1 combined with 1D8 or 1D8 Pb-Tx antibody in adjuvant or neoadjuvant regimens with respect to surgery as outlined in the scheme. (B) Representative low-magnification fluorescent microscopy images of the lungs to assess metastases present in treated mice euthanized 37 d after 4T1 mCherry orthotopic implantation and subsequent surgical removal of the primary tumors on day +16. (C) Violin plot showing the quantification of the fluorescent microscopy images of the lungs from experiment in B. Results are shown as the percentage of the area occupied by mCherry fluorescence. Dots represent three distinct lung area quantifications per mouse, with n = 6 mice per experimental group. (D) Schematic representation of the survival in vivo experiments in 4T1 mCherry breast cancer model treated with anti–PD-1 combined with 1D8 or 1D8 Pb-Tx antibody in a neoadjuvant setting. (E) Survival following neoadjuvant treatment in the orthotopic 4T1 breast cancer model performed as indicated in “d”. Primary tumor surgery day is indicated by a dashed line. Survival percentage over time with n = 9 mice per group is represented. Survival differences were analyzed by log rank tests. *P < 0.05, **P < 0.01, ***P < 0.001.