Integrating system biology and intratumor gene therapy by trans-complementing the appropriate co-stimulatory molecule as payload in oncolytic herpes virus

Abstract

Systems biology has been applied at the multi-scale level within the cancer field, improving cancer prevention, diagnosis and enabling precision medicine approaches. While systems biology can expand the knowledge and skills for oncological treatment, it also represents a challenging expedition due to cancer complexity, heterogeneity and diversity not only between different cancer indications, but also in its evolution process through space and time. Here, by characterizing the transcriptional perturbations of the tumor microenvironment induced by oncolytic, we aimed to rationally design a novel armed oncolytic herpes virus. We found that intratumor oncovirotherapy with HSV-1 induces T-cell activation signatures and transcriptionally activates several costimulatory molecules. We identified differentially expressed costimulatory receptors and binding partners, where inducible co-stimulators (ICOS) resulted in the potentially most beneficial targeted therapy. Through an ex-vivo transcriptomic analysis, we explored the potential of arming an oncolytic virus as a combination therapy strategy; in particular, we engineered a targeted herpes virus encoding ICOSL (THV_ICOSL), which resulted in a significant improvement in tumor size control compared to unarmed parental virus. Also, combination with a PD-1 inhibitor enhanced antitumor efficacy as predictable by upregulation of PD-1 and ligands pair (PD-L1/PD-L2) upon oncolytic virus injection. Generation of the human version of this virus encoding hICOSL orthologue effectively and specifically activated human T cells by triggering the ICOS pathway. Our data support the data-driven generation of armed oncolytic viruses as combination immunotherapeutic with checkpoint inhibitors.

Fig. 1. Analysis of tumor microenvironment perturbation after OV treatment.

A Schematic representation of in vivo schedule of treatment. Subcutaneously implantation of LLC1-hHER2 into human HER2 tolerant mice. Mice were injected with R-LM113 OV or PBS. The day after injection, mice were killed, and tumor bulk RNA was analyzed by Nanostring Ncounter. B Funnel analysis of tumors immunological signature. Canonical activation markers were assessed and reported in the first set of heatmap. Only statistically significant differentially regulated genes were represented. Co-stimulus receptors were reported in the second set of heatmap. All the costimulatory receptors included in PanCancer immune profiling Nanostring panel were represented. Those resulting significantly differentially regulated were flagged by asterisk. Those statistically significantly upregulated (ICOS and CD30) were highlighted by black boxes. Matching ligand expression for CD30 and ICOS were analyzed and reported in the third panel of heatmaps. Broad institute’s Morpheus interface was used for heatmap representation. C, D Differential expression assessed by Nanostring was confirmed by Real-Time PCR from retrotranscribed cDNAs. One asterisk is used for p value < 0.005 for the heat map. In panels C and D, the asterisks represent p value < 0.05.

Fig. 2. Generation of THV_mICOSL.

A Schematic representation of immunological synapses required for fully competence of antitumor T cell response. A′ T cell encounters a canonical costimulus-expressing cell leading to complete activation of T effector cell and cancer cell death. A″ T cell effector encounters a cell which is deficient for the expression of costimulus; T cell is not completely activated. The ectopic expression of costimulatory ligands on the surface of tumor cells, mediated by oncolytic vectors, leads to a complete activation of T lymphocytes. (B) THV_ICOSL BAC DNAs were transfected in SKOV3 cells to produce infectious viral particles. Representative single plaque peaking. C Purification of THV_mICOSL through iodixanol gradient; the thick band indicates good productivity. D The murine ICOSL coding sequence was inserted into intergenic region US1-US2 in the background of R-LM113. The scFv targeting hHER2 is shown in glycoprotein D. © THV_mICOSL viral yield over ten passages (p2, p5 and p10). Differences in viral yield were not statistically significant. E Viral yield at different passages.

Fig. 3. FACS assessment of mICOSL on THV_mICOSL infected SKOV3 cell line.

A Not infected SKOV3 cells shows an absent ICOSL expression after staining with anti-mouse ICOSL. B Infected SKOV3 cells not stained with anti-ICOSL Ab in FSC-A value underlines a high percentage of infected, dead cells. C THV_ICOSL-infected SKOV3 cells were stained with anti-mouse ICOSL. Cells in P2 expressing mICOSL correspond to hundred percent of population. P1 population underlines a high percentage of infected, dead cells.

Fig. 4. Comparison of cytotoxic effect of unarmed R-LM113 vs. THV_mICOSL on cancer cells.

In (A) SKOV3 cells were respectively infected at 0.1, 1, 10 MOI and cell viability was assessed over 5 days post infection. In (B) the same experiment was performed in CT26 cells expressing hHER2. As expected, cytotoxicity was comparable between the unarmed and armed virus underlining the absence of a detrimental effect of arming. Percent of viable cells was calculated as percentage over not infected cells.

Fig. 5. THV-ICOSL in combination with αPD1 slows LLC1-hHER2 tumor growth.

Expression of CTLA4, PD1, PDL1, PDL2 was analyzed in the dataset described in Fig. 1 and reported as heatmap (A). (B) Efficacy data of depicted treatment in hHER2-transgenic mice challenged SC with LLC1-hHER2 cells. Intratumor treatment (IT) with PBS (Untreated), unarmed R-LM113 (5x 1E + 08), THV_mICOSL or R-LM113 (3x 5E + 07) alone or in combination with αPD1. Tumor volume was monitored over time. Lines in the graphs represent each individual tumor in a mouse. CR complete responder (also fibrotic residue).

Fig. 6. THV-ICOSL in combination with αPD1 improves mice survival.

Survival curves of LLC1-hHER2 tumor bearing mice receiving the reported treatment. Mice were killed when tumor reached >1500 mm³. PBS n:5; R-LM113 n:5; R-LM113 + PD1 n:6; THV-ICOSL n:7; THV-ICOSL + PD1 n:7.

Fig. 7. In vitro validation of biological activity of human ICOSL armed THV.

A Titers of transfection (P0) and purified THV_hICOSL virus on iodixanol gradient. B The human ICOSL coding sequence is inserted into intergenic region US1-US2 in the background of unarmed THV. The scFv targeting hHER2 is shown in glycoprotein D. C Schematic representation of ICOS Blockade Bioassay used to measure the potency and stability of ligands that binds ICOS receptor. Jurkat T cells, endogenously express TCR/CD3 and are engineered to express human ICOS and a NanoLuc luciferase reporter driven by ICOS and TCR/CD3 pathway-dependent response elements. D Activation of Jurkat T cells in coculture with SKOV3 tumor cells infected with THV empty (R-LM113) and THV_hICOSL at MOI 0,1 pfu/cell.