Baculovirus-assisted Reovirus Infection in Monolayer and Spheroid Cultures of Glioma cells

Abstract

The mammalian orthoreovirus Type 3 Dearing has great potential as oncolytic agent in cancer therapy. One of the bottlenecks that hampers its antitumour efficacy in vivo is the limited tumour-cell infection and intratumoural distribution. This necessitates strategies to improve tumour penetration. In this study we employ the baculovirus Autographa californica multiple nucleopolyhedrovirus as a tool to expand the reovirus' tropism and to improve its spread in three-dimensional tumour-cell spheroids. We generated a recombinant baculovirus expressing the cellular receptor for reovirus, the Junction Adhesion Molecule-A, on its envelope. Combining these Junction Adhesion Molecule-A-expressing baculoviruses with reovirus particles leads to the formation of biviral complexes. Exposure of the reovirus-resistant glioblastoma cell line U-118 MG to the baculovirus-reovirus complexes results in efficient reovirus infection, high reovirus yields, and significant reovirus-induced cytopathic effects. As compared to the reovirus-only incubations, the biviral complexes demonstrated improved penetration and increased cell killing of three-dimensional U-118 MG tumour spheroids. Our data demonstrate that reovirus can be delivered with increased efficiency into two- and three-dimensional tumour-cell cultures via coupling the reovirus particles to baculovirus. The identification of baculovirus' capacity to penetrate into tumour tissue opens novel opportunities to improve cancer therapy by improved delivery of oncolytic viruses into tumours.

Figure 1

RV’s cellular receptor JAM-A is present on the BV envelope. (a) Western blot analysis of Sf9 cells infected with BV^JAM. The blot was reacted with mAb α-HA followed by antibody-alkaline phosphatase, to detect the presence of JAM-A in infected cells. (b–d) Electron microscopic images of BV^JAM, negatively stained using uranyl acetate and incubated with mouse mAb α-JAM-A (i = sc-53623, ii = ab17261) followed by α-mouse antibody tagged with 10-nm colloidal gold. (b) Example of a BV^JAM field on grid showing low background labelling. (c,d) Enlargements of BV^JAM virions bearing respectively a single (c) or multiple (d) gold grains.

Figure 2

RV and BV^JAM associate and form a biviral complex. (a) Schematic representation of the BV^JAM-RV complex. RV attachment protein σ1 bound to JAM-A expressed on the BV envelope. (b,c) Electron microscopy images of BV^JAM-RV complexes. The virions were negatively stained with uranyl acetate. Most complexes consisted of one BV^JAM and one RV virion (b), some complexes showed other combinations of single or multiple BV^JAM and RV virions (c). (d) Flow cytometry analyses of recognition of JAM-A on HER911 cells and U-118MG cells as negative control, by α-JAM-A antibodies ab17261 and sc-53623. (e) The RV yields from HER911 cells and culture medium upon incubation with α-JAM-A antibodies ab17261 and sc-53623 or as controls unrelated antibodies of the same provider as controls. The error bars represent the standard deviation (n = 3), p values: ** ≤ 3E-4 (f) The RV yields from U-118 MG cells and culture medium after incubation with BV^JAM or BV^JAM exposed to α-JAM-A antibodies sc-53623 or unrelated control antibodies of the same provider. The dashed line represents RV input. The error bars represent the standard deviation (n = 3), p values: * ≤ 4.4E-2, ** ≤ 8.8E-3.

Figure 3

Complexing BV^JAM and RV increases RV infection of U118-MG cells. (a) Flow cytometry analyses of the fraction of RV infected U-118 MG cells, stained for RV capsid protein σ3, upon incubation with either RV or with different ratios of BV^JAM-RV complexes 40 h post-infection. The error bars represent the standard deviation (n = 2). (b) RV yields from U-118 MG cells upon infection with either RV or with different ratios of BV^JAM-RV complexes as determined by plaque assays. The error bars represent the standard deviation (n = 3), for BV^JAM-RV ratios 0:5 and 1,000:5 (n = 2).

Figure 4

Enhanced RV-mediated killing of U-118 MG cells by apoptosis upon infection with BV^JAM-RV. (a) Analyses of the cell viability of U-118 MG cells by WST-1 assay at four days post-infection with BV^JAM, RV, or BV^JAM-RV complexes. Uninfected cells were set at 100% viability. The error bars represent the standard deviation (n = 9), p values: * = 3.9E-2, ** = 4.8E-3, *** < 1E-4. (b) Caspase-3 and -7 analysis in the lysate of U-118 MG cells upon infection with BV^JAM, RV or BV^JAM-RV complexes. Measurements were normalized to the caspase-3 and -7 values in uninfected cells. The error bars represent the standard deviation (n = 3), p values: ** ≤ 2.1E-3, *** < 1E-4.

Figure 5

Improved penetration and spread of RV in U-118 MG spheroids upon infection with BV^JAM-RV. (a) Example of a graph assigning the midpoint of a spheroid. After the spheroids (25,000 cells at start) were fixed and sectioned, the diameter of each of the spheroid sections was measured. These values were combined in a graph as represented here and the trend line through the data points was used to determine the spheroid section with the largest diameter. (b) Representative images of RV penetration analysis in U-118 MG spheroids, either as single agent (10,000 vp/cell) or in combination with BV^JAM in a RV:BV^JAM ratio of 1:2.5. Two, four and six days post-infection spheroids were fixed, sliced and stained for the presence of RV capsid protein σ3 by immunocytochemistry. (c) Uninfected area of spheroids infected as in (b) with RV only or the BV^JAM-RV complex at six days post-infection. For each spheroid the uninfected area of three to five middle sections was measured and averaged. The error bars represent the standard deviation (n = 4 spheroids), p value: ** = 9.2E-3. (d) Representative images of RV titration experiment on spheroids infected with RV only or in the presence of a surplus of BV^JAM at six days post-infection through immunocytochemistry staining for RV capsid protein σ3.

Figure 6

Improved cell killing of U-118 MG spheroids upon infection with BV^JAM-RV. Spheroids of 25,000 (a) or 5,000 (b) cells at start were either mock treated or transduced with BV^JAM only, RV only or RV pre-incubated with BV^JAM or BV-GFP. At six days post-infection the viability of the spheroids was assessed by WST-1 assay. Mock-treated cells were set at 100% viability. The error bars represent the standard deviation (n = 8), p values: * = 3.1E-2, ** = 1.6E-2, *** = 2E-4.