Background: Ewing sarcoma (EWS) is a rare cancer of the bone and soft tissue, most prevalent in children and young adults. The treatment of EWS has progressed relatively little in over 30 years. Survival rates for patients, particularly those with metastatic and/or relapsed disease remain poor, highlighting the urgent need for innovative treatment options. Methods: Here, we have explored the therapeutic potential of the oncolytic Maraba virus strain MG1 using various in vitro models of EWS, including established cell lines, doxorubicin-resistant derivatives, spheroid cultures and primary patient-derived Ewing sarcoma cell cultures. We examined the direct oncolytic activity of MG1 and its ability to stimulate the immune-mediated killing of EWS by human healthy donor peripheral blood mononuclear cells. Results: We show that MG1 undergoes productive replication and exerts direct oncolysis of established EWS cell lines, doxorubicin-resistant EWS cell lines and patient-derived Ewing sarcoma cell cultures more recently established from tumours. In contrast, primary mesenchymal stem cells (the likely cell of origin of EWS) were resistant to MG1, with IFN-I being a major determinant of tumour cell selectivity. MG1-treated PBMC produced IFN-I and killed EWS cells in vitro, in a natural killer (NK) cell-dependent manner. Conclusions: The ability of MG1 to kill EWS cells directly and stimulate NK cell cytotoxicity against this tumour suggests that MG1 may provide therapeutic benefit for EWS patients where the efficacy of conventional treatments is currently limited.

Ewing sarcoma (EWS) is a rare cancer of the bone and soft tissue, most prevalent in children and young adults. Treatment of EWS has progressed relatively little in over 30 years. Survival rates for patients, particularly those with metastatic and/or relapsed disease, remain poor, highlighting the urgent need for innovative treatment options. We have explored the therapeutic potential of the oncolytic Maraba virus strain MG1. We show that MG1 undergoes productive replication and exerts direct oncolysis of established EWS cell lines, doxorubicin-resistant EWS cell lines and patient-derived Ewing sarcoma (PDES) cell cultures more recently established from tumours. In contrast, primary mesenchymal stem cells (the likely cell of origin of EWS) were resistant to MG1, with IFN-I being a major determinant of tumour cell selectivity. MG1 treated PBMC produced IFN-I and killed EWS cells in vitro, in a natural killer (NK) cell-dependent manner. The ability of MG1 to kill EWS cells directly and to stimulate NK cell cytotoxicity against this tumour suggests that Maraba virus MG1 may provide therapeutic benefit for EWS patients where the efficacy of conventional treatments is currently limited.

A major barrier to effective immunotherapy in osteosarcoma (OS) is the highly immunosuppressive tumour microenvironment (TME), which limits immune recognition and elimination of tumour cells. We evaluated a panel of oncolytic herpes simplex viruses (oHSVs) for their direct oncolytic activity, immune‑modulatory properties and capacity to counteract OS‑associated immunosuppression using immunologically relevant in vitro models. We demonstrate that established OS cell lines, primary cell cultures and dissociated OS cells from freshly resected tumour samples are susceptible to direct oncolysis by three oHSVs; HSV1716, HSV1716‑GMCSF, and HSV47Δ, although susceptibility levels varied. Treatment of peripheral blood mononuclear cells from healthy donors and OS patients with oHSVs enhanced natural killer (NK) cell activation and promoted immune‑mediated killing of OS cell lines and primary OS cell cultures. Among the three viruses, HSV1716‑GMCSF exhibited the strongest immune‑stimulatory effects and was uniquely capable of reducing the abundance of CD163+CD206+ immunosuppressive TAMs; use of this oHSV was therefore prioritised. We developed a multicellular spheroid model of OS, incorporating OS cells, mesenchymal stem cells and TAMs, which exhibits resistance to immune-mediated killing, better reflecting the immunosuppressive TME in patients. In this model, pairing HSV1716‑GMCSF treatment with either anti‑GD2 or anti‑EGFR monoclonal antibodies (mAbs), selected according to OS tumour antigen expression, significantly increased immune‑mediated tumour cell killing. These findings suggest that personalised combination strategies pairing oHSVs with appropriate mAbs provide a promising therapeutic approach for OS by integrating direct oncolysis, remodelling of the immunosuppressive TME and enhanced immune‑mediated tumour destruction.

Most adenovirus (Ad) vectors are based on the genome of human Ad type 5 (Ad5), which targets their entry to cells that express the Coxsackie and adenovirus receptor (CAR or CXADR). However, certain human Ads do not use CAR; for example, Ad35 interacts with cell surface CD46 and Ad3 uses desmoglein 2 (DSG2) for cell entry. In this study, a comparison of different Ad receptors using transcriptomic and proteomic databases showed that CD46 is widely expressed across human cells and tissues whereas CAR and DSG2 are more restricted to epithelial cells. We have used a hybrid virus, Ad5F35, that comprises an Ad5 genome in which the Ad5 fibre was replaced with that of Ad35, thus retargeting the virus from CAR- to CD46-expressing cells and enabling transduction of primary human NK and T cells. However, lymphocytes required approximately 10-to-20-fold-more Ad5F35 particles per cell (ppc) compared to A549 epithelial cells to achieve a similar level of transduction. Consistent with this, quantitation of the cell-surface density of CD46 molecules revealed approximately 100 CD46 molecules per µm² in primary NK cells compared with approximately 2000 CD46 per µm² in HeLa cells, a 20-fold difference. Cell surface CD46 density was reduced by approximately 95% by RNA interference in HeLa cells to levels that approximate those found on NK cells. Lower CD46 density reduced transduction by Ad5F35 but this could be compensated for with increased MOI. Our results identify the density of cell surface CD46 as a critical determinant of Ad5F35 transduction and demonstrate that Ad5F35 is an efficient vector for gene delivery in primary human NK cells.

Most adenovirus (Ad) vectors are based on the genome of human Ad type 5 (Ad5), which targets their entry to cells that express the Coxsackie and Adenovirus Receptor (CAR or CXADR). However, certain human Ads do not use CAR, for example Ad35 interacts with cell-surface CD46 and Ad3 uses desmoglein 2 (DSG2) for cell entry. In this study, a comparison of different Ad receptors using transcriptomic and proteomic databases showed that CD46 is widely expressed across human cells and tissues whereas CAR and DSG2 are more restricted to epithelial cells. We have used a hybrid virus, Ad5F35, that comprises an Ad5 genome in which the Ad5 fibre was replaced with that of Ad35, thus retargeting the virus from CAR- to CD46-expressing cells and enabling transduction of primary human NK and T cells. However, lymphocytes required approximately 10 to 20-fold more Ad5F35 particles per cell (ppc) compared to A549 epithelial cells to achieve a similar level of transduction. Consistent with this, quantitation of the cell-surface density of CD46 molecules revealed approximately 100 CD46 molecules per μm ² in primary NK cells compared with approximately 2000 CD46 per μm ² in HeLa cells, a 20-fold difference. Cell-surface CD46 density was reduced by approximately 95% by RNA interference in HeLa cells to levels that approximate those found on NK cells. Lower CD46 density reduced transduction by Ad5F35 but this could be compensated for with increased MOI. Our results identify the density of cell surface CD46 as a critical determinant of Ad5F35 transduction and demonstrate that Ad5F35 is an efficient vector for gene delivery in primary human NK cells.