Targeting 4-1BB costimulation to the tumor stroma with bispecific aptamer conjugates enhances the therapeutic index of tumor immunotherapy

Abstract

Despite the recent successes of using immune modulatory Abs in patients with cancer, autoimmune pathologies resulting from the activation of self-reactive T cells preclude the dose escalations necessary to fully exploit their therapeutic potential. To reduce the observed and expected toxicities associated with immune modulation, here we describe a clinically feasible and broadly applicable approach to limit immune costimulation to the disseminated tumor lesions of the patient, whereby an agonistic 4-1BB oligonucleotide aptamer is targeted to the tumor stroma by conjugation to an aptamer that binds to a broadly expressed stromal product, VEGF. This approach was predicated on the premise that by targeting the costimulatory ligands to products secreted into the tumor stroma, the T cells will be costimulated before their engagement of the MHC-peptide complex on the tumor cell, thereby obviating the need to target the costimulatory ligands to noninternalizing cell surface products expressed on the tumor cells. Underscoring the potency of stroma-targeted costimulation and the broad spectrum of tumors secreting VEGF, in preclinical murine tumor models, systemic administration of the VEGF-targeted 4-1BB aptamer conjugates engendered potent antitumor immunity against multiple unrelated tumors in subcutaneous, postsurgical lung metastasis, methylcholantrene-induced fibrosarcoma, and oncogene-induced autochthonous glioma models, and exhibited a superior therapeutic index compared with nontargeted administration of an agonistic 4-1BB Ab or 4-1BB aptamer.

Figure 1. Characterization of VEGF-4-1BB aptamer conjugates

A. Binding of VEGF-4-1BB aptamer conjugates to filter-immobilized targets was measured using a double-filter binding assay. ³²P-labeled 4-1BB aptamer, VEGF aptamer, or VEGF-4-1BB aptamer conjugates were passed through nitrocellulose filters immobilized with decreasing amounts of murine 4-1BB protein (m4-1BB), murine VEGF protein (mVEGF), or IgG, except for the rightmost lane that contained no protein. Bound radioactivity was visualized by exposure to X-ray sensitive film. B. 4-1BB costimulation of polyclonally activated CD8⁺ T cells. CD8⁺ T cells were activated with anti-CD3 antibody and incubated with an agonistic 4-1BB antibody or IgG control, 4-1BB aptamer, VEGF-4-1BB, or control aptamers in which the 4-1BB aptamer was replaced with a scrambled aptamer (Scram). Proliferation was measured 48 hours later by ³H-thymidine incorporation. Statistical analysis: 4-1BB antibody versus isotype antibody, p=0.01; 4-1BB aptamer versus scrambled aptamer, p=0.028, VEGF-4-1BB aptamer conjugate versus VEGF-scrambled aptamer conjugate, p=0.069.

Figure 2. Treatment of tumor-bearing mice with VEGF-4-1BB aptamer conjugates potentiate vaccine-induced protective antitumor immunity

A. Balb/c mice were implanted subcutaneously with 4T1 tumor cells, vaccinated with irradiated B7-1 and MHC class II-expressing 4T1 cells at days 7, 10 and 13 (arrows) and treated with VEGF-4-1BB aptamer conjugates or with a mixture of VEGF and 4-1BB aptamers 18 hours after each vaccination (8 mice per group). Statistical analysis: ** p < 0.01. (Days 25 and 27, p=0.0089 and 0.0027, respectively). B. C57BL/6 mice were implanted subcutaneously with B16.F10 melanoma cells, vaccinated with GVAX at days 5, 8 and 11 (arrows), and treated with VEGF-4-1BB aptamer conjugates or with a mixture of VEGF and 4-1BB aptamers 18 hours after each vaccination (8 mice per group). Statistical analysis: * p<0.05. (Days 19 and 21, p= 0.0433 and 0.0403, respectively).

Figure 3. VEGF-mediated tumor-targeting of 4-1BB aptamers

A. VEGF expression in 4T1 and 4T07 tumors. 4T1 or 4T07 tumor cells were injected subcutaneously into Balb/c mice and when tumors reached about 0.5 cm diameter they were resected and stained for VEGF expression as described in Methods. B. Homing of systemically administered VEGF-4-1BB aptamer conjugates to VEGF-expressing tumors. Balb/c mice were implanted in opposite flanks with 1.0x10⁴ 4T1 (VEGF^high) and 5x10⁴ 4T07 (VEGF^low) tumor cells and at day 21 when tumors became readily palpable mice were administered with ³²P-labeled VEGF-4-1BB conjugates via the tail vein. As described in the Methods, only the dimeric 4-1BB was radioactively labeled. At indicated times mice were sacrificed, tumors excised, and radioactivity was determined by scintillation counting. Statistical significance 4T1 versus 4T07: 6h, p=0.2059; 24h, p=0.016; 48h, p=0.0348. C and D. VEGF-4-1BB aptamer conjugates inhibit VEGF^high 4T1, but not VEGF^low 4T07 cells. Balb/c mice were injected subcutaneously with 4T1 (C) or 4T07 (D) tumor cells and 4 days later administered with 100 pmoles of VEGF-4-1BB aptamer conjugates three times 3 days apart and tumor growth was monitored. Insert, mice were administered with 500 pmoles of unconjugated 4-1BB aptamer. (5 mice per group). Statistical significance: Panel C, days 16 and 18, p<0.001; day 20, p=0.0145, panel D, ns=not statistically significant.

Figure 4. Therapeutic index of VEGF-4-1BB aptamer conjugates

A. 4T1 tumor cells were injected subcutaneously into Balb/c mice and 4 days later administered via the tail vein with 150 pmoles of VEGF-4-1BB aptamer conjugates or with 800 pmoles of either unconjugated 4-1BB aptamer, anti-4-1BB Ab, or IgG control Ab. Aptamer and antibody administration was repeated two additional times at 3-day intervals as described in Methods, and tumor growth was monitored (5 mice per group). B. Mice shown in panel A were sacrificed at day 17, spleens, and lymph nodes weights were measured and C. CD8⁺ T cells were isolated from the spleens, lymph nodes, and livers, and were counted. The difference between the 4-1BB Ab group and all other treated groups in panels B and C was statistically significant (p< 0.001). There was no statistical difference between the control IgG, 4-1BB aptamer and VEGF-4-1BB aptamer conjugate groups. D. Tissue sections from the lung and liver were stained with hematoxylin and eosin and visualized by light microscopy.

Figure 5. 4T1 breast carcinoma post-surgical metastasis model

A. 4T1 tumor cells were injected into the abdominal mammary fat pad. At day 11, 1–2 days after tumors became palpable, tumors were surgically excised with the mice under anesthesia. Two days later mice were vaccinated with irradiated B7-1- and MHC class II-expressing 4T1 cells and 18 hours later 100 pmoles of VEGF-4-1BB aptamer conjugates or a mixture of VEGF and 4-1BB aptamers were administered via the tail vein. Vaccination and aptamer conjugate treatment was repeated twice three days apart. Mice were sacrificed when they showed signs of morbidity. (6 mice per group). B. At day 32 before the onset of signs of discomfort and morbidity 4T1 implanted, surgically resected mice were injected with ³²P-labeled VEGF-4-1BB aptamer conjugates or unconjugated mixtures of VEGF and 4-1BB aptamers via the tail vein. 24 hours later mice were sacrificed, lungs were removed, and radioactivity determined by scintillation counting.

Figure 6

Autochthonous murine tumor models. A. Methylcholanthrene-induced fibrosarcoma model. C57BL/6 mice were injected subcutaneously with 400 mg of methylcholanthrene (MCA) in castor oil. When tumors became palpable, at days 71, 78, 85 and 92, mice were injected with 100 pmoles of VEGF-4-1BB aptamer conjugates or a mixture of VEGF and aptamers via the tail vein. Mice were sacrificed when mice exhibited signs of morbidity or when tumor diameter reached 1.2 cm, whichever came first (9 mice per group.) B. Oncogene-induced high-grade glioma model. Tumor-bearing Ntv-a mice were injected via the tail vein with 150 pmoles of VEGF-4-1BB aptamer conjugates or with a mixture of VEGF and 4-1BB aptamers starting at day 21 for a total of 6 injections every 3 or 4 days. Mice were sacrificed when they showed signs of morbidity (10–20 mice per group.) Statistical significance: VEGF-4-1BB aptamer conjugate versus a mixture ofVEGF and 4-1BB aptamers, p=0.0298.