Fig 1.
Agonistic effect of ICOS antibodies on T cells.
(A) Purified human CD4 T cells were stimulated with serial dilutions of 37A10 or hamster IgG isotype control in soluble (left) or plate-bound (right) format in the presence of suboptimal anti-CD3 OKT3 antibody (1.0 μg/ml). Cell division was determined by CFSE dilution using flow cytometry. IFNγ production was analyzed by Cytokine Bead Arrays. (B) Repeat of A using JTX-2011 or anti-RSV-hIgG1 isotype control. (C) Induction of p-AKT in CD4 T cells. Cells were activated with anti-CD3/CD28 for 24 hrs, rested for 24 hrs, and then washed. Cells were incubated with JTX-2011 or Herceptin® (hIgG1 isotype control) on ice, and then incubated at 37ºC in the presence or absence of a cross-linking secondary antibody. p-AKT was measured by flow cytometry, and percentage of p-AKT+ cells at 2, 5 or 30 min after stimulation with JTX-2011 or Herceptin® is shown. Each symbol indicates CD4 T cells from independent donors (N = 4). (D) Percentage of p-AKT+ cells at 2, 5 or 30 min after stimulation with 37A10 or hamster IgG. Each symbol indicates CD4 T cells from independent donors (N = 2). (E) ICOS staining on CD4 T cells by flow cytometry following 24 hours of stimulation of with PBMCs with anti-CD3 antibody OKT3 at 37˚C. Data are plotted as mean ± standard deviation (SD).
Fig 2.
Lack of superagonism of ICOS antibodies.
(A-B) CD4 T cells were stimulated with anti-CD28 clones in soluble form (sol), JTX-2011, 37A10 or anti-RSV-hG1 isotype control in soluble or plate-bound (pb) format at 10 μg/ml in the presence (white) or absence (gray) of anti-CD3 OKT3 antibody (1.0 μg/ml). Proliferation was measured by CFSE dilution using flow cytometry. (C) IFNγ (top) and IL-6 (bottom) secretion after 24 hr incubation of PBMC with CD28.1, OKT3, Herceptin® or JTX-2011 (100, 1.0 or 0.01 μg/ml; 666.7, 6.67 or 0.067 nM, respectively) in solid-phase or soluble-phase, or whole blood with antibodies in soluble-phase. Cytokines in supernatants were analyzed from a total of 6 donors for each condition. Data are plotted as mean ± SD.
Fig 3.
Regression of tumors and long-term protection following ICOS antibody treatment in Sa1/N model.
(A) Dosing regimen in Sa1/N model. One million cells were injected s.c. and mice were randomized on day 7 when average tumor volume reached ~100 mm3. Mice received ICOS or isotype control antibodies by i.p. injections on days 7, 10, 14 and 17, and tumor growth was measured. (B) Tumor growth curves of mice treated with 37A10-mG2a or mIgG2a isotype (0.3 mg/kg). N = 10 per group. (C) Tumor growth curves of mice treated with JTX-1011-mG2a (0.25 mg/kg) or mIgG2a isotype (0.25 mg/kg). N = 10 per group. (D) Tumor-free mice that were previously treated with JTX-1011-mG2a were re-challenged s.c. on the opposite flank with Sa1/N tumors ~45 days after the final dose of antibody was administered (N = 7 for each). Naïve mice were challenged with Sa1/N as a positive control (N = 10). Kaplan-Meier tumor-free survival curves shown. *** P<0.001 compared to isotype. (E) Tumor-growth curves of mice treated with 37A10-mG2a (0.25 mg/kg) or 37A10-mIgG1-agly (0.25 mg/kg), or mIgG2a and mIgG1 isotype (5.0 mg/kg each). N = 10 per group. In tumor growth curves, each line indicates an individual mouse, and number of tumor-free mice out of 10 at the end of study is indicated for each group.
Fig 4.
A-D: Effect of combination treatment with ICOS and anti-PD-1 antibodies in multiple syngeneic tumor models. Mice received 0.25 mg/kg of ICOS antibody (ICOS mAb) and/or 10 mg/kg of anti-PD-1 antibody RMP1-14 (PD-1 mAb) administered i.p. Treatment in CT26 started on day 3 after tumor challenge when tumors were palpable, and antibodies were administered twice weekly for two weeks. Treatment in the 4T1 model started on day 3 after tumor challenge when tumors were palpable, and antibodies were administered once weekly for two weeks. Treatment in the B16-SIY and MC38 models started on day 6 and 11, respectively, after tumor challenge when tumors reached an average volume of ~100 mm3, and antibodies were administered once weekly for two weeks. E. Effect of combination treatment with ICOS and anti-CTLA-4 antibodies in the MC38 syngeneic tumor model in huCTLA-4 knock-in mice. Treatment in MC38 began when tumors reached an average volume of 80mm3; ipilimumab was administered i.v. once at the time of randomization, and ICOS antibody was administered i.p. at day of randomization and 7 days post-randomization. Each line indicates an individual mouse. Numbers of tumor-free mice out of 10 at the end of study are indicated for each group.
Fig 5.
Changes in T cell subsets in tumors following ICOS antibody administration in vivo.
(A) Dosing regimen for TIL studies in Sa1/N model. Dosing frequency and timing are the same as in Fig 3A. Single cell suspensions from spleens and tumors analyzed by flow cytometry 2 days after the second dose. (B) Frequencies of T cell subsets in CD3+ cells in spleens (left), frequencies of T cell subsets in CD45+CD3+ cells within tumors (right) (N = 5 per group). (C) Number of Tregs (left), CD4 Teff (middle) and CD8 T cells (right) per milligram (mg) of tumor. (D) CD8:Tregs (left) and CD4 Teff:Tregs (right) ratios in tumors after treatment. Each symbol indicates an individual mouse or donor; lines indicate mean ± SD.