Local delivery of interleukin 7 with an oncolytic adenovirus activates tumor-infiltrating lymphocytes and causes tumor regression

Abstract

Cytokines have proven to be effective for cancer therapy, however whilst low-dose monotherapy with cytokines provides limited therapeutic benefit, high-dose treatment can lead to a number of adverse events. Interleukin 7 has shown promising results in clinical trials, but anti-cancer effect was limited, in part due to a low concentration of the cytokine within the tumor. We hypothesized that arming an oncolytic adenovirus with Interleukin 7, enabling high expression localized to the tumor microenvironment, would overcome systemic delivery issues and improve therapeutic efficacy. We evaluated the effects of Ad5/3-E2F-d24-hIL7 (TILT-517) on tumor growth, immune cell activation and cytokine profiles in the tumor microenvironment using three clinically relevant animal models and ex vivo tumor cultures. Our data showed that local treatment of tumor bearing animals with Ad5/3- E2F-d24-hIL7 significantly decreased cancer growth and increased frequency of tumor-infiltrating cells. Ad5/3-E2F-d24-hIL7 promoted notable upregulation of pro-inflammatory cytokines, and concomitant activation and migration of CD4+ and CD8 + T cells. Interleukin 7 expression within the tumor was positively correlated with increased number of cytotoxic CD4+ cells and IFNg-producing CD4+ and CD8+ cells. These findings offer an approach to overcome the current limitations of conventional IL7 therapy and could therefore be translated to the clinic.

Keywords: Oncolytic virus; adenovirus; immunotherapy; interleukin 7.

Figure 1.

Ad5/3-E2F-d24-IL7 functionality in vitro and in vivo. (a) A schematic presentation of chimeric 5/3 oncolytic adenovirus armed with human interleukin 7. (b) Relative human cancer cell viability after addition of 1, 10, 100 or 1000 VP/cell at day 4 after infection in A549 and RD cell lines. (c) Relative hamster cancer cell viability after addition or 100, 1000, 5000 or 10000 VP/cell at day 5 after infection in HT100 and DDT1-MF2 or, at day 8 after infection in HapT1 cell lines. (d) Level of IL7 in the infected cancer cell supernatants. (e) IL7 bioactivity measured via application of infected cells supernatant on IL7-dependent murine cell line 2E8 in different dilutions. Recombinant murine IL7 (rmIL7) and murine IL7 (rhIL7) were used as controls at 20 ng/ml. (f) Human IL7 cross-reactivity in hamster splenocytes proliferation. All in vitro experiments were carried out in triplicates and repeated at least twice. Data is presented as mean ± SEM. (g) Schematic representation of syngeneic hamster experimental model. (h) Tumor growth until day 14. Tumor volumes were normalized against day 0. Data is presented as mean ± SEM. (i) Immune cells infiltrating treated tumors: CD4 + T cells, CD8 + T cells, MHC II expressing cells, GM-1+ cells (marker of NK cells), and Mac-2+ cells (marker of monocytes/macrophages). Data is presented as mean ± SEM. (j) Immune-related gene expression changes in tumor digests. Data is presented as mean ± SEM.

Figure 2.

Lytic capability and replication of IL7 armed adenovirus in cancer patient ex vivo tumor cultures (a) Viability of tumor digests HUSOV4, OvCaS and HUSOV5 from ovarian cancer. Cell viability data is normalized against the uninfected mock. Experiments were performed in triplicates. Statistical significance is represented as *p < .05, **p < .01, ***p < .001, and ****p < .0001. (b) Ad5/3 replication evaluation through quantitative real-time PCR from ovarian samples. Viral copy number was normalized against human β-actin. PCRs were carried out in duplicates. (c) IL7 protein concentration in supernatants from ovarian samples, measured by Cytokine Bead Array (CBA) assay. Experiment was carried out in triplicates. All data is presented as mean ± SEM.

Figure 3.

Evaluation of cytokines and chemokines in tumor microenvironment of ovarian cancer ex vivo tumor histocultures (HUSOV4, OvCaS and HUSOV5) upon viral infection. (a) Level of pro-inflammatory cytokines. (b) Pooled pro-inflammatory changes. (c) Level of anti-inflammatory cytokines. (d) Pooled anti-inflammatory changes. (e) Overall ratio of pro- to anti-inflammatory cytokines. Data was normalized against mock. (f) Level of chemokines. Data was normalized against mock. (g) Absolute number of migrated cells assessed using a Transwell system (5 μM pore membrane). The supernatant with cytokines and chemokines obtained after infection was added to the lower chamber. PBMCs were plated in the top chamber, and after 24 hours the number of cells at the bottom (migrated cells) were counted using 123count eBeads (Invitrogen). (h) Migrated immune cells profiling. All experiments were performed in triplicates, and resulting data is presented as mean ± SEM. Statistical significance is represented as *p < .05, **p < .01, ***p < .001, and ****p < .0001.

Figure 4.

Evaluation of infiltrating CD4+ and CD8 + T cell activation and cytotoxicity in ovarian cancer ex vivo histocultures (HUSOV4, OvCaS and HUSOV5) upon viral infection. (a) Frequency of CD69+ cells, (b) IFNg+ cells, (c) GzmB+ cells and (d) TNF+ cells in CD4 + T cell populations measured through flow cytometry. (e) Differentially expressed genes in CD4 + T cell population. (f) Frequency of CD69+ cells, (g) IFNg+ cells, (h) GzmB+ cells and (i) Perf+ cells in CD8 + T cell populations measured through flow cytometry. (j) Differentially expressed genes in CD8 + T cell population. All flow cytometry experiments were run in duplicates, and resulting data is presented as mean ± SEM. Statistical significance is represented as *p < .05, **p < .01 and ***p < .001.

Figure 5.

Evaluation of infiltrating CD4+ and CD8 + T cell activation, cytotoxicity and IL7Ra expression in multiple ex vivo samples and correlation with IL7 levels. (a) Frequency of GzmB+CD4+ cells, (b) IFNg+CD4+ cells, (c) IFNg+CD8+ cells and (d) CD69+ CD8+ cells in tumor microenvironment measured through flow cytometry. (e) Correlation between IL7 concentration and frequency of GzmB+CD4+ cells, (f) IFNg+CD4+ cells, (g) IFNg+CD8+ cells and (h) CD69+ CD8+ cells in tumor microenvironment (i) Mechanism of IL7Ra internalization. Briefly, interaction of IL7 with IL7Ra induces the heterodimerization of and conformational changes in IL7Rα and common gamma-chain receptor (γ chain). These conformational changes bring together tyrosine kinases Jak1 and Jak3, which phosphorylate each other, thereby increasing their kinase activity. Subsequently, the activated Jak proteins phosphorylate tyrosine residue Y449 in the cytoplasmic domain of IL7Ra creating a docking site for downstream effectors, for instance, signal transducer and activator of transcription 5 (STAT5). Phosphorylated STAT5 then homodimerizes and translocates to the nucleus, where it activates the expression of SOCS proteins, which are involved in cytokine signaling silencing, e.g. targeting receptor for internalization and degradation. (j) IL7Ra expression level on the surface of CD4+ cells and (k) CD8+ cells on day 1, 3 and 7 upon viral infection, measured through flow cytometry. (l) Expression of activation markers GzmB and CD69 on the CD127+ CD4+ cells on day 1, 3 and 7 upon viral infection, measured through flow cytometry. All flow cytometry experiments were run in duplicates, and resulting data is presented as mean ± SEM. Statistical significance is represented as *p < .05, **p < .01.

Figure 6.

Ad5/3-E2F-d24-hIL7 efficacy in two distinct in vivo models bearing an ovarian PDX. (a) Schematic representation of PDX murine experimental model. (b) NOG mice tumor growth until day 15. Tumor volumes were normalized against day 0. Data is presented as mean+SEM. (c) Frequency of CD4+ and CD8+ cells and (d) CD69+ CD4+ and CD69+ CD8+ tumor infiltrating cells in NOG mice. (e) Frequency of circulating CD4+ and CD8+ in NOG mice. (f) NOG-IL2 mice tumor growth until day 15. Tumor volumes were normalized against day 0. Data is presented as mean+SEM. (g) Frequency of CD4+ and CD8+ cells and (h) CD69+ CD4+ and CD69+ CD8+ tumor infiltrating cells in NOG-IL2 mice. (i) Frequency of circulating CD4+ and CD8+ in NOG-IL2 mice. All flow cytometry experiments were run in duplicates, and resulting data is presented as mean ± SEM. Statistical significance is represented as *p < .05, **p < .01, ***p < .001, and ****p < .0001.