Changes in Susceptibility to Oncolytic Vesicular Stomatitis Virus during Progression of Prostate Cancer

Abstract

A major challenge to oncolytic virus therapy is that individual cancers vary in their sensitivity to oncolytic viruses, even when these cancers arise from the same tissue type. Variability in response may arise due to differences in the initial genetic lesions leading to cancer development. Alternatively, susceptibility to viral oncolysis may change during cancer progression. These hypotheses were tested using cells from a transgenic mouse model of prostate cancer infected with vesicular stomatitis virus (VSV). Primary cultures from murine cancers derived from prostate-specific Pten deletion contained a mixture of cells that were susceptible and resistant to VSV. Castration-resistant cancers contained a higher percentage of susceptible cells than cancers from noncastrated mice. These results indicate both susceptible and resistant cells can evolve within the same tumor. The role of Pten deletion was further investigated using clonal populations of murine prostate epithelial (MPE) progenitor cells and tumor-derived Pten(-/-) cells. Deletion of Pten in MPE progenitor cells using a lentivirus vector resulted in cells that responded poorly to interferon and were susceptible to VSV infection. In contrast, tumor-derived Pten(-/-) cells expressed higher levels of the antiviral transcription factor STAT1, activated STAT1 in response to VSV, and were resistant to VSV infection. These results suggest that early in tumor development following Pten deletion, cells are primarily sensitive to VSV, but subsequent evolution in tumors leads to development of cells that are resistant to VSV infection. Further evolution in castration-resistant tumors leads to tumors in which cells are primarily sensitive to VSV.

Importance: There has been a great deal of progress in the development of replication-competent viruses that kill cancer cells (oncolytic viruses). However, a major problem is that individual cancers vary in their sensitivity to oncolytic viruses, even when these cancers arise from the same tissue type. The experiments presented here were to determine whether both sensitive and resistant cells are present in prostate cancers originating from a single genetic lesion in transgenic mice, prostate-specific deletion of the gene for the tumor suppressor Pten. The results indicate that murine prostate cancers are composed of both cells that are sensitive and cells that are resistant to oncolytic vesicular stomatitis virus (VSV). Furthermore, androgen deprivation led to castration-resistant prostate cancers that were composed primarily of cells that were sensitive to VSV. These results are encouraging for the use of VSV for the treatment of prostate cancers that are resistant to androgen deprivation therapy.

FIG 1

Primary cultures of mouse prostate epithelial cells express cytokeratins. Primary cultures of mouse prostate epithelial cells were cultured from normal Pten^L/L prostates, Pten^−/− prostate tumors, and castration-resistant Pten^−/− prostate tumors. Cells were seeded into imaging plates and labeled with DAPI (blue) as a marker for cellular DNA and immunolabeled for cytokeratin (red) as a marker for epithelial cells. Using the DAPI label to define the region of interest to denote single cells, the percentages of cells immunolabeled with cytokeratin were determined and are shown in the graph. Murine embryonic fibroblasts (MEFs) were used as a negative control for cytokeratin immunolabeling.

FIG 2

Prostate tumors contain a mixture of resistant and susceptible cells. Primary cultures of prostate epithelial cells were established from Pten^−/− prostate tumors of 3-month-old mice and their littermate normal Pten^L/L controls. Cells were seeded into imaging plates and were infected with rwt-GFP or rM51R-GFP viruses (green). Cells were fixed, permeabilized, immunolabeled for cytokeratin (red) and stained with DAPI to label cellular DNA (blue), and then analyzed with a high-content imaging system. Representative fluorescence images are shown in panels A and B. The numbers of cells that were positive for GFP and cytokeratin were expressed as a percentage of total cytokeratin-positive cells (C and D). The data shown are averages of at least three experiments ± the standard errors of the mean (SEM).

FIG 3

Castration-resistant tumor cells are primarily susceptible to VSV infection. Pten^−/− mice were castrated at 3 months of age and by 6 months of age had developed castration-resistant tumors. Primary cultures of prostate epithelial cells were established from Pten^−/− prostate tumors of 6-month-old mice, castration-resistant tumors, and their littermate normal Pten^L/L control prostates. Cells were seeded into imaging plates and infected with rwt-GFP and rM51R-GFP viruses (green). Cells were fixed, permeabilized, immunolabeled for cytokeratin (red) and stained with DAPI to label cellular DNA (blue), and then analyzed with a high-content imaging system. Representative fluorescence images are shown in panels A and B. The numbers of cells that were positive for GFP and cytokeratin were expressed as a percentage of total cytokeratin-positive cells (C and D). The data shown are averages of at least three experiments ± the SEM.

FIG 4

Tumor-derived Pten^−/− cells support very little viral gene expression even at a high MOI. Vector control Pten^L/L cells and acutely deleted Pten^−/− cells (A) or nontransduced control Pten^L/L cells and tumor-derived Pten^−/− cells (B) were infected with rwt-dsRed virus at MOIs of 1, 10, and 50. Bright-field and fluorescence images were obtained at 24 h postinfection. The images shown are representative from two to three independent experiments.

FIG 5

Viral protein expression in control Pten^L/L cells and acutely deleted Pten^−/− cells versus tumor-derived Pten^−/− cells. (A) Control Pten^L/L, acutely deleted Pten^−/−, and tumor-derived Pten^−/− cells were infected with rwt virus at an MOI of 10. Cell lysates were harvested at 6 and 12 h postinfection and probed for viral matrix protein (M) protein expression by immunoblotting. Representative immunoblots of at least three experiments are shown. (B) Viral matrix (M) protein expression was quantified and expressed as a percentage of M protein expression in control Pten^L/L cells at 12 h postinfection. *, P < 0.05. (C) Vector control Pten^L/L and acutely deleted Pten^−/− cells were infected with rwt virus at an MOI of 10 or mock infected (M). At 6, 12, and 24 h postinfection, the cells were pulse-labeled with [³⁵S]methionine. Cell lysates were harvested and resolved on SDS-PAGE gels. Representative phosphorimages are shown. Viral proteins L, G, N, P, and M are indicated on the right. (D) Nontransduced control Pten^L/L and tumor-derived Pten^−/− cells were infected with rwt virus at an MOI of 10 or mock infected (M). At 6, 12, and 24 h postinfection, cells were pulse-labeled with [³⁵S]methionine and analyzed as in panel C.

FIG 6

In a single cycle infection, tumor-derived Pten^−/− cells are more resistant to virus-induced cell death than control Pten^L/L and acutely deleted Pten^−/− cells. Vector control Pten^L/L and acutely deleted (A and B) or nontransduced control Pten^L/L and tumor-derived Pten^−/− cells (C and D) were infected with rwt (A and C) or rM51R (B and D) viruses at an MOI of 10. Cell viability at the indicated times was determined by using an MTT assay and is expressed as a percentage of the mock-infected cells. The data show averages of at least three experiments ± the SEM. *, P < 0.05.

FIG 7

In a multiple cycle infection, tumor-derived Pten^−/− cells are highly resistant to virus-induced cell death. Vector control Pten^L/L and acutely deleted (A and B) or nontransduced control Pten^L/L and tumor-derived Pten^−/− cells (C and D) were infected with rwt (A and C) or rM51R (B and D) viruses at an MOI of 0.1. Cell viability was determined using MTT assay and is expressed as a percentage of the mock-infected cells. The data show averages of at least three experiments ± the SEM. *, P < 0.05.

FIG 8

Tumor-derived Pten^−/− cells constitutively express high levels of STAT1, which is phosphorylated in response to infection with VSV with either wt or mutant M protein. Control Pten^L/L, acutely deleted, and tumor-derived Pten^−/− cells were infected with rwt (A) or rM51R (B) viruses at an MOI of 10. Cell lysates were harvested at the indicated times and probed for phospho-STAT1 and total STAT1 expression. Representative immunoblots for at least three experiments are shown. The levels of phospho-STAT1 and total STAT1 were quantified and are expressed as the percentage of levels in tumor-derived Pten^−/− cells at 6 h postinfection. The data show averages of at least three experiments ± the SEM. *, P < 0.05.

FIG 9

Acutely deleted Pten^−/− cells respond poorly to IFN stimulation. Control Pten^L/L and acutely deleted Pten^−/− cells were pretreated with the indicated concentrations of IFN and then infected with rwt virus at an MOI of 10. Cell viability was determined by using an MTT assay at 72 h postinfection and is expressed as a percentage of mock-infected cells. *, P < 0.05.

FIG 10

Model for the development of VSV-susceptible and VSV-resistant cells in prostate cancers resulting from Pten deletion.