Characterization of a novel OX40 ligand and CD40 ligand-expressing oncolytic adenovirus used in the PeptiCRAd cancer vaccine platform

Abstract

Oncolytic viruses (OVs) have been shown to induce anti-cancer immunity and enhance cancer immunotherapies, such as immune checkpoint inhibitor therapies. OV therapies can be further improved by arming OVs with immunostimulatory molecules, including various cytokines or chemokines. Here, we have developed a novel adenovirus encoding two immunostimulatory molecules: cluster of differentiation 40 ligand (CD40L) and tumor necrosis factor receptor superfamily member 4 ligand (OX40L). This novel virus, designated VALO-D102, is designed to activate both innate and adaptive immune responses against tumors. CD40L affects the innate side by licensing antigen-presenting cells to drive CD8⁺ T cell responses, and OX40L increases clonal expansion and survival of CD8⁺ T cells and formation of a larger pool of memory T cells. VALO-D102 and its murine surrogate VALO-mD901, expressing murine OX40L and CD40L, were used in our previously developed PeptiCRAd cancer vaccine platform. Intratumoral administration of PeptiCRAd significantly increased tumor-specific T cell responses, reduced tumor growth, and induced systemic anti-cancer immunity in two mouse models of melanoma. In addition, PeptiCRAd therapy, in combination with anti-PD-1 immune checkpoint inhibitor therapy, significantly improved tumor growth control as compared to either monotherapy alone.

Keywords: CD40L; OX40L; PeptiCRAd; T cell activation; oncolytic vaccine.

Graphical abstract

Figure 1

Novel VALO-D102 oncolytic adenovirus produces high levels of biologically active human CD40 ligand (CD40L) and OX40 ligand (OX40L) (A) Schematic representation of genetic modifications in VALO-D102. The virus has a 24-base pair deletion in E1A; the CR1-alpha and gp19K genes in the E3A region have been replaced with human OX40L and CD40L genes; the 14.7K gene in the E3B region has been deleted; and finally, the adenovirus 5 knob domain has been replaced with the knob domain from adenovirus serotype 3. (B) A549 cells were infected with VALO-D102 at a MOI of 10. 72 h postinfection, supernatant was collected and added to a culture of Ramos-Blue reporter cells, and CD40 receptor activation by functional virus-produced CD40L was measured. (C) A549 cells were infected with VALO-D102 at a MOI of 10. 48 h postinfection, HEK293-OX40/NF-κB reporter cells were added to the infected A549 cells for 6 h, and OX40 activation by virus-expressed, functional membrane-bound OX40L was measured.

Figure 2

Characterization of PeptiCRAd-1 components (A) VALO-D102 stability as measured by infectious titer assay (immunocytochemistry [ICC]), viral particle (VP) titer assay (OD260), and nanoparticle tracking analysis (NTA) of freshly produced virus and virus stored at −20°C and <−60°C at indicated time points. (B) Immunological characterization of the tumor antigen-containing polylysine-modified peptides used in the PeptiCRAd-1 cancer vaccine. Purified CD8⁺ T cells from cancer patients were presensitized with peptide-pulsed (10 μg/mL), irradiated, autologous PBMCs depleted of CD4⁺ and CD8⁺ T cells. Presensitized CD8⁺ T cells were tested on days 10–12 by IFN-γ ELISpot assay for recognition of peptide-pulsed (1 μg/mL), autologous antigen-presenting cells (APCs). For the NY-ESO-1 assay, CD8⁺ T cells were presensitized with tumor antigen-containing peptides, with or without polylysine modification. APCs were pulsed with tumor antigen-containing peptides, with or without the polylysine modification. For the MAGE-A3 assay, CD8⁺ T cells were presensitized with tumor antigen-containing peptide MAGE-A3_(168−176). APCs were pulsed with tumor antigen-containing peptides, with or without the polylysine modification. A confirmatory ELISpot using a MAGE-A3_(168−176)-specific CD8⁺ T cell clone combined was also performed for MAGE-A3 peptides. APCs were pulsed with tumor antigen-containing peptides, with or without the polylysine modification. Polylysine did not affect the processing or T cell recognition of the modified peptides. (C) Averaged finite track length adjustment (FTLA) concentration/size graphs from NTA of virus only (VALO-D102), VALO-D102 coated with modified NY-ESO-1 peptide, VALO-D102 coated with modified MAGE-A3 peptide, and VALO-D102 coated individually with both modified peptides followed by mixing the mono-coated complexes together (PeptiCRAd-1).

Figure 3

Oncolytic potency of VALO-D102 is identical to the parental strain and is not affected by coating the virus with tumor antigens (A–E) Oncolytic potency of VALO-D102 and PeptiCRAd-1, a cancer vaccine platform consisting of VALO-D102 coated with two modified tumor antigens, was compared to the Ad5/3-D24 parental strain in a (A) human A375 melanoma cell line, (B) human SK-MEL-2 melanoma cell line, (C) human A549 lung carcinoma cell line, (D) human SW982 synovial sarcoma cell line, and (E) human HCC70 triple-negative breast cancer cell line. (F) Oncolytic potency of VALO-D102 and PeptiCRAd-1 was assessed in vivo in A549 tumor-bearing, immunodeficient female athymic nude-Foxn1-nu mice. 1 × 10⁹ VP of each virus was given intratumorally triweekly starting at day 11. Replication-deficient Ad5/3luc1 was used as a control virus. The numbers of mice in each group were 11−12. Statistical analysis was performed with two-way ANOVA. ∗∗∗p < 0.001.

Figure 4

Virus-encoded OX40L and CD40L improve anti-tumor efficacy and induce robust infiltration of tumor-specific CD8⁺ T cells into the tumor in a syngeneic mouse model of B16.OVA melanoma (A) 1 × 10⁹ VP of PeptiCRAd Ad5/3-D24-OVA or PeptiCRAd VALO-mD901-OVA was given intratumorally 6, 8, and 20 days post-tumor implantation. Average tumor growth curves for all treatment groups are shown. (B) Immunological analysis of tumors and tumor-draining lymph nodes of treated mice. Lymph nodes from all mice from each treatment group were pooled in order to get enough cells for the flow cytometric analysis. The number of mice in the mock group was 7 and in both PeptiCRAd groups, was 10. Statistical analysis was performed with one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01.

Figure 5

PeptiCRAd improves tumor growth control and induces systemic tumor-specific T cell responses in a syngeneic mouse model of B16.F10.9/K1 (A) 1 × 10⁹ VP of VALO-mD901 or PeptiCRAd VALO-mD901-Trp2 was given intratumorally 6, 7, 8, 9, 10, 22, and 34 days post-tumor implantation. Average tumor growth curves for all treatment groups are shown. (B) Immunological analysis of tumors and tumor-draining lymph nodes of treated mice. Lymph nodes from all mice from each treatment group were pooled in order to get enough cells for the flow cytometric analysis. (C) Systemic tumor-specific T cell responses were analyzed with the ELISpot assay from the spleens of treated mice. The numbers of mice in each group were 8−10. Statistical analysis was performed with two-way ANOVA. ∗∗∗p < 0.001.

Figure 6

PeptiCRAd, in combination with anti-PD1, improves tumor growth control and survival compared to either monotherapy and triggers a systemic anti-tumor memory response in a syngeneic mouse model of B16.F10.9/K1 (A) Anti-PD-1 immune checkpoint inhibitor alone (200 μg/dose given intraperitoneally), 1 × 10⁹ VP of PeptiCRAd VALO-mD901-Trp2 alone, or in combination with anti-PD-1 immune checkpoint inhibitor were given intratumorally 9, 10, 11, 12, 13, and 26 days post-tumor implantation. Average tumor growth curves for all treatment groups are shown. (B) Kaplan-Meier survival curve for all treatment groups. (C) Individual tumor volumes for rechallenged mice at day 29 after secondary tumor engraftment. The numbers of mice in each group were 8−9 in (A) and (B) and 3−5 in (C). Statistical analysis was performed with two-way ANOVA for (A) and with log-rank test for (B). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001.