Gastrointestinal cancer-associated fibroblasts expressing Junctional Adhesion Molecule-A are amenable to infection by oncolytic reovirus

Abstract

Gastrointestinal (GI) cancers are characterized by extensive tumor stroma that both promotes tumor progression and acts as a physical barrier for adjacent tumor cells, limiting the effect of current treatment modalities. Oncolytic virotherapy is currently investigated in clinical trials as a novel therapeutic agent for different malignancies of the GI tract, but it is largely unknown whether these viruses can also target the tumor stroma. Here, we investigated the tropism of two commonly studied OVs, adenovirus and reovirus, towards primary GI fibroblasts from human oesophageal, gastric, duodenal and pancreatic carcinomas (N = 36). GI fibroblasts were susceptible to type 3 Dearing (T3D) strain R124 and bioselected mutant reovirus (jin-3) infection but not oncolytic adenovirus (Ad5-Δ24). Efficient infection and apoptosis of human and mouse GI cancer-derived fibroblasts by these reoviruses was partially dependent on the expression of the reovirus entry receptor, Junctional Adhesion Molecule-A (JAM-A). Moreover, human GI cancer organoid-fibroblast co-cultures showed higher overall infectivity when containing JAM-A expressing fibroblasts as compared to JAM-A negative fibroblasts, indicating a potential role of JAM-A expressing fibroblasts for viral dissemination. We further show that JAM-A is not only necessary for efficient reovirus infection of fibroblasts but also partially mediates reovirus-induced apoptosis, dependent on signaling through the C-terminal PDZ-domain of JAM-A. Altogether, our data show the presence of JAM-A expressing fibroblasts in both human and murine GI cancers that are amenable to infection and induction of apoptosis by reovirus, extending the potential anti-cancer actions of reovirus with stromal targeting.

Cancer Spheroids and Organoids as Novel Tools for Research and Therapy: State of the Art and Challenges to Guide Precision Medicine

“…The results demonstrated that all oncolytic viruses significantly inhibited cell viability in organoid cultures. The tropism of two commonly used OV, adenovirus (Ad5-Δ24) and reovirus (R124 and jin-3), toward primary gastrointestinal fibroblasts derived from human esophageal, gastric, duodenal, and pancreatic carcinomas was studied [ 173 ]. Considerable cell death was observed in the majority of the fibroblast cultures infected with either R124 or jin-3 reoviruses, while Ad5-Δ24 did not induce cell death in the vast majority of fibroblasts tested.…”

Preclinical evaluation of the gorilla‐derived HAdV‐B AdV‐lumc007 oncolytic adenovirus ‘GoraVir’ for the treatment of pancreatic ductal adenocarcinoma

“…Consequently, therapeutic strategies targeting these cells have gradually emerged [ 20 , 21 ]. Recently, it was demonstrated that oncolytic mammalian orthoreovirus (reovirus) was able to infect and kill human untransformed CAFs isolated from several gastrointestinal tumour types in vitro , in contrast to an oncolytic derivative of HAdV‐C5 [ 16 ]. To investigate whether GoraVir does have the potential to infect and lyse CAFs, the pancreatic stellate cell line PS‐1 was exposed to GoraVir or HAdV‐C5 at high and low MOI and cell viability was determined at 6 days post infection (Fig.…”

“…In contrast, HAdV‐C5 showed minor cytotoxicity at MOI 10 (mean residual viability 80.6 ± 3.58%) and no cell killing at a lower MOI. To continue our exploration of GoraVir's ability to infect and kill CAFs, we exposed patient‐derived primary fibroblasts previously isolated from pancreatic cancer tissues [ 16 ] with GoraVir at MOI 10 and measured cell viability at 5 days post infection. Again, GoraVir was shown to induce cell killing in the majority of primary fibroblasts (Fig.…”

“…Another approach to target CAFs using ICOVIR15 included the incorporation of a bispecific T-cell engager directed at FAP [40]. Meanwhile, a recent paper by Harryvan et al [16] showed that the more wildtype-like HAdV-C5D24 was unable to induce cell death in human primary CAFs in contrast to wildtype and bioselected mutant reoviruses. Since the resistance to Ad infection was also observed in CAR-expressing fibroblasts, the authors attributed these differences to the D24-modification in the HAdV-C5D24 vector, which makes it not suitable for infecting and killing of quiescent cells with an intact Rb-pathway.…”

“…The pancreatic stellate cell line PS‐1 was kindly provided by prof. Kocher (Queen Mary University London, London, UK) and has been described previously [ 15 ]. Primary pancreatic cancer‐associated fibroblasts (CAFs) were isolated as described previously [ 16 ] and cultured in Dulbecco's Modified Eagle's Medium (DMEM)/F‐12 (Gibco) supplemented with 8% FBS and 1% P/S. Cells were cultured in DMEM supplemented with 8% FBS and 1% P/S unless stated otherwise.…”