Establishing a New Platform to Investigate the Efficacy of Oncolytic Virotherapy in a Human Ex Vivo Peritoneal Carcinomatosis Model

Abstract

Oncolytic virotherapy constitutes a promising treatment option for many solid cancers, including peritoneal carcinomatosis (PC), which still represents a terminal stage of many types of tumors. To date, the in vitro efficacy of oncolytic viruses is mostly tested in 2D-cultured tumor cell lines due to the lack of realistic 3D in vitro tumor models. We have investigated the feasibility of virotherapy as a treatment option for PC in a human ex vivo peritoneum co-culture model. Human HT-29 cancer cells stably expressing marker genes GFP and firefly luciferase (GFP/luc) were cultured on human peritoneum and infected with two prototypic oncolytic viruses (GLV-0b347 and MeV-DsRed). Both viral constructs were able to infect HT-29 cells in patient-derived peritoneum with high tumor specificity. Over time, both GFP signal and luciferase activity decreased substantially, thereby indicating successful virus-induced oncolysis. Furthermore, immunohistochemistry stainings showed specific virotherapeutic infections of HT-29 cells and effective tumor cell lysis in infected co-cultures. Thus, the PC model established here provides a clinically relevant screening platform to evaluate the therapeutic efficacy of virotherapeutic compounds and also to investigate, in an autologous setting, the immunostimulatory potential of oncolytic viruses for PC in a unique human model system superior to standard 2D in vitro models.

Keywords: ex vivo peritoneal culture model; oncolytic measles vaccine virus; oncolytic vaccinia virus; peritoneal carcinomatosis; virotherapy.

Figure 1

Virotherapeutic treatment of GFP/luc-labeled human HT-29 tumor cells in cell culture with GLV-0b347. (A) Schematic illustration of the three-step virotherapeutic process and associated detection capabilities: (1) GFP/luc-labeled HT-29 tumor cells were seeded into a 24-well cell culture plate. Successful plating of the cells can be verified by the determination of GFP fluorescence. (2) The treatment of HT-29 cells with oncolytic viruses encoding a red-fluorescent marker protein. The successful infection of the tumor cells can be verified by the determination of red fluorescence. (3)/(4) The viral oncolysis can be determined by a decrease in GFP as well as in luciferase activity. Over time, enhanced red fluorescence indicates an increasing number of tumor cells being infected by the red-fluorescence marker gene encoding virotherapeutic compounds. (B) The fluorescence images of HT-29 GFP/luc-labeled cells at 72 h postinfection (hpi) with GLV-0b347 at different multiplicities of infection (MOIs), as depicted. BF, brightfield; OL, overlay of GFP and TurboFP635 signal.

Figure 2

Virotherapeutic treatment of GFP/luc-labeled human HT-29 tumor cells in cell culture with MeV-DsRed. Fluorescence images of HT-29 GFP/luc-labeled cells at 72 h postinfection (hpi) with MeV-DsRed at different multiplicities of infection (MOIs), as depicted. BF, brightfield; OL, overlay of GFP and DsRed signal.

Figure 3

Comparison of different detection options for the oncolytic activity of virotherapeutic compounds GLV-0b347 (A, images to the left) and MeV-DsRed (B, images to the right) in HT-29 GFP/luc tumor cells. HT-29 GFP/luc cells were infected with GLV-0b347 (A) or MeV-DsRed (B) at different multiplicities of infection (MOIs) ranging from 0.0001 to 1 for GLV-0b347, from 0.001 to 10 for MeV-DsRed, or remained uninfected (MOCK). At 72 h postinfection (hpi), remaining tumor cell masses were determined by either (i) SRB viability assays, (ii) the measurement of the luciferase activity, or (iii) the quantification of the GFP or red-fluorescence intensity. Each measurement was calculated relative to the MOCK control. The mean ± SD of at least two independent experiments performed in triplicate is shown. ANOVA test relative to MOCK-infected control: * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001.

Figure 4

Human ex vivo peritoneum model and schematic illustration of the three-step virotherapeutic process in co-cultures with GFP/luc-labeled human HT-29 tumor cells. (A) Photographic image of the human ex vivo peritoneal model cultivated between stainless steel rings in a 24-well plate. (B) Photographic image of the peritoneum in the ex vivo model through a light microscope. (C) (1) Preparation of co-cultures of the peritoneum from noncancer patients and human GFP/luc-labeled HT-29 tumor cells. Successful plating of the cells can be verified by fluorescence microscopy. (2) Virotherapeutic treatment of co-cultures with oncolytic viruses carrying a red-fluorescent marker protein. Successful infection of the tumor cells can be verified by the determination of red fluorescence via fluorescence microscopy. (3) Viral oncolysis can be determined by a decrease in GFP and red fluorescence as well as by a decrease in luciferase activity.

Figure 5

Virotherapeutic treatment of PC models with recombinant vaccinia virus GLV-0b347. The fluorescence images of GLV-0b347-infected co-cultures ((A); 1.5 × 10⁶ plaque-forming units (PFU)) or MOCK-infected (B) co-cultures consisting of the peritoneum of noncancer patients and adherently growing GFP/luc-labeled human HT-29 tumor cells at days 2, 4, and 7 postinfection (dpi); original magnification 4×. (C) Luciferase activity of GFP/luc–HT-29 cells growing on peritoneum at 2, 4, and 7 dpi, either GLV-0b347-infected or MOCK-infected. Each measurement is normalized to the MOCK control. The mean ± SD of one experiment performed in triplicates is shown. t-test relative to MOCK-infected control: * p < 0.05 and ** p < 0.01. ns; not significant. (D) The hematoxylin and eosin staining of human peritoneal tissue with and w/o co-culture of GFP/luc-labeled HT-29 cells at 7 dpi, either GLV-0b347-infected or MOCK-infected. (E) The EpCAM staining of peritoneal tissue with the co-culture of HT-29 cells at 7 dpi, either GLV-0b347-infected or MOCK-infected. (F) The vaccinia virus staining of peritoneal tissue with co-culture of HT-29 cells at 7 dpi, either GLV-0b347-infected or MOCK-infected. Experiments were conducted with the peritoneal tissue from different patients and show representative data from at least three different experiments. P; peritoneum. Black arrows indicate intact or infected and lysed HT-29 cells on the surface of the peritoneum. Scale bars represent 100 μm.

Figure 6

Virotherapeutic treatment of PC models with recombinant measles vaccine virus MeV-DsRed. The fluorescence images of MeV-DsRed-infected co-cultures ((A); 1.5 × 10⁶ plaque-forming units (PFU)) or MOCK-infected (B) co-cultures consisting of the peritoneum of noncancer patients and GFP/luc-labeled human HT-29 tumor cells at days 2, 4, and 7 postinfection (dpi); original magnification 4×. (C) The luciferase activity of GFP/luc–HT-29 cells growing on peritoneum at 2, 4, and 7 dpi, either MeV-DsRed-infected or MOCK-infected. Each measurement is normalized to the MOCK control. The mean ± SD of one experiment performed in triplicates is shown. t-test relative to MOCK-infected control: * p < 0.05. ns; not significant. (D) The hematoxylin and eosin staining of human peritoneal tissue with and w/o co-culture of GFP/luc-labeled HT-29 tumor cells at 7 dpi with MeV-DsRed or MOCK infection. (E) The EpCAM staining of peritoneal tissue with co-culture of HT-29 cells at 7 dpi with MeV-DsRed or MOCK infection. (F) The MeV staining of peritoneal tissue with co-culture of HT-29 cells at 7 dpi with MeV-DsRed or MOCK infection. The experiments were conducted with peritoneal tissue from different patients and show representative data from at least three different experiments. P; peritoneum. Black arrows indicate intact or infected and lysed HT-29 tumor cells on the surface of the peritoneum. Scale bars represent 100 μm.