Lentiviral interferon: A novel method for gene therapy in bladder cancer

Abstract

Interferon alpha (IFNα) gene therapy is emerging as a new treatment option for patients with non-muscle invasive bladder cancer (NMIBC). Adenoviral vectors expressing IFNα have shown clinical efficacy treating bacillus Calmette-Guerin (BCG)-unresponsive bladder cancer (BLCA). However, transient transgene expression and adenoviral immunogenicity may limit therapeutic activity. Lentiviral vectors can achieve stable transgene expression and are less immunogenic. In this study, we evaluated lentiviral vectors expressing murine IFNα (LV-IFNα) and demonstrate IFNα expression by transduced murine BLCA cell lines, bladder urothelium, and within the urine following intravesical instillation. Murine BLCA cell lines (MB49 and UPPL1541) were sensitive to IFN-mediated cell death after LV-IFNα, whereas BBN975 was inherently resistant. Upregulation of interleukin-6 (IL-6) predicted sensitivity to IFN-mediated cell death mediated by caspase signaling, which when inhibited abrogated IFN-mediated cell killing. Intravesical therapy with LV-IFNα/Syn3 in a syngeneic BLCA model significantly improved survival, and molecular analysis of treated tumors revealed upregulation of apoptotic and immune-cell-mediated death pathways. In particular, biomarker discovery analysis identified three clinically actionable targets, PD-L1, epidermal growth factor receptor (EGFR), and ALDHA1A, in murine tumors treated with LV-IFNα/Syn3. Our findings warrant the comparison of adenoviral and LV-IFNα and the study of novel combination strategies with IFNα gene therapy for the BLCA treatment.

Keywords: IL-6; MB49; bladder cancer; gene therapy; interferon alpha; intravesical; lentiviral vector.

Graphical abstract

Figure 1

Expression of murine IFNα protein and its target genes in syngeneic BLCA cell lines (A and B) BLCA cell lines were transduced with LV-Ctrl or LV-IFNα vectors at an MOI of 2 and murine IFNα was measured in the cell-free supernatants at 24 h (A) and 72 h (B) after viral transduction. All the three cell lines showed significant expression of IFNα protein in LV-IFNα-transduced cells compared with controls. (C) qPCR analysis of IFNα target genes IRF7, PDL-1, and TRAIL in BLCA cell lines show increased expression following treatment with LV-IFNα virus when compared with control. Not significant (ns), p > 0.05; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.

Figure 2

Cytotoxic effect of LV-IFNα on BLCA cell lines (A–C) Murine cell lines were exposed to murine recombinant IFNα (rIFNα) or transduced with LV-Ctrl or LV-IFNα and cell counts were measured using Trypan blue dye exclusion method. At 72 h, MB49 (A) and UPPL1541 (B) showed significant reduction in cell counts, while BBN975 (C) showed no change in cell numbers when compared with the saline treatment. (D) At 72 h, annexin V and propidium iodide staining showed increased apoptotic cells after 100 U rIFNα or LV-IFNα treatment in MB49 and UPPL1541 cells. Percentage of apoptotic cells in BBN975 remained unchanged. ns, p > 0.05; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. (E) Western blot for IFNα target genes shows increased expression of PD-L1, STAT1, and p-STAT1 in cell lines treated with 100 U rIFNα or LV-IFNα.

Figure 3

RNA-seq analysis of mouse BLCA cell lines treated with 100 U rIFNα or LV-IFNα (A–F) Heatmaps of significant genes (FDR cutoff 0.05 and fold change of 2) between 100 U IFNα and CTRL samples for BBN975 (A), MB49 (B), and UPPL1541 (C) and between LV-IFNα and LV-Ctrl samples for BBN975 (D), MB49 (E), and UPPL1541 (F). (G and H) Venn diagrams show significantly altered genes (FDR cutoff 0.05 and fold change of 1.5) common or unique to all the three cell lines after treatment with LV-IFNα and rIFNα. (I and J) GSEA shows enrichment of interferon alpha response pathway in MB49 cells treated with LV-IFNα and rIFNα.

Figure 4

Ingenuity pathway analysis (IPA) of canonical pathways altered after treatment with LV-IFNα IPA analysis of significant genes (FDR cutoff 0.05 and fold change of 2) was performed to identify top candidate canonical pathways with a positive Z score in BBN975 (A), MB49 (B), and UPPL1541 (C). Pathways that are common to all three cell lines are indicated by green bars, and pathways common between MB49 and UPPL1541 are shown in red.

Figure 5

Caspase-mediated cell death in IFNα-sensitive cells (A and C) Cell counts by Trypan blue dye exclusion method in MB49 (A) and UPPL1541 (C) cells treated with recombinant TRAIL (rTRAIL, 100 ng/mL) in the presence or absence of caspase 8 inhibitor (50 μM). rTRAIL reduced cell counts in a dose-dependent manner and was partially rescued by addition of caspase 8 inhibitor. (B and D) Percentage of apoptotic cells measured by annexin V/propidium iodide (PI) staining showed increased apoptosis in MB49 (B) and UPPL1541 (D) cells treated with rTRAIL and was rescued in cells treated with rTRAIL and caspase 8 inhibitor. (E) MB49 cells treated with increasing concentrations of tunicamycin, ER stress inducer, showed dose-dependent reduction in cell counts. (F) Caspase 12 inhibitor rescues cell death induced by rIFNα. (G and H) IL-6 protein expression measured by ELISA in cell-free supernatants of MB49 (G) and UPPL1541 (H) cells is shown. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.

Figure 6

Electron microscopy and transduction of bladder urothelium treated with LV vectors (A) Normal mouse urothelium at 50,000×. (B) Mouse urothelium 4 h post-LV-IFNα instillation demonstrating one vacuole containing a virus particle inside the cytoplasm (red circle) at 100,000× is shown. (C) Twenty-four hours post-instillation of LV-IFNα demonstrates four vacuoles with LV vectors within them at 50,000×. (D) Ninety-six hours post-LV-IFNα instillation, four vacuoles containing LV vectors adjacent to the nucleus at 75,000× are shown. (E) After five weekly instillations, at 10,000×, there are numerous vector-filled vacuoles outside the nucleus in a second cell layer (∗ denotes vacuoles and arrow denotes cell layer). (F) Mouse urothelium after 3 months of BBN followed by monthly instillations of LV-IFNα for 3 months is shown. At 10,000×, a dark nucleus with shrunken cytoplasm is visible within an apoptotic cell. (G) In the same mouse as in image (F) at 100,000×, viral vectors are seen within a darkened nucleus, apoptotic. (H) At 10,000×, viral vectors (marked by V) are seen near nuclei deep to the connective tissue (marked by C). (I) Whole mounts of mouse bladders instilled with LV-βgal vector showing β-gal-positive blue cells in the urothelium are shown. (J and K) Tissue sections showing β-gal-positive cells in the urothelium are shown. Insets (J′) and (K′) show higher magnifications. Scale bars, 100 μm. (L and M) Measurement of IFNα11 levels by ELISA in urine of mice treated with LV-Ctrl or LV-IFNα (L) or tissue (M) at 24, 48, 72, and 96 h post-instillation is shown.

Figure 7

Efficacy of LV-IFNα in murine MB49 intravesical model (A) Kaplan-Meier plot showing percent survival in C57Bl/6, MB49 intravesical disease model treated with vehicle (Syn3), LV-Ctrl, and LV-IFNα (p = 0.0004). (B) H&E staining of whole-mount bladder tissues with corresponding Ki67 staining in MB49 is shown. (C) IHC for CD4 and CD8 T cells on MB49 intravesical tumors treated with Syn3 (vehicle), LV-Ctrl, and LV-IFNα vectors is shown. (D) Quantification of CD4 and CD8 cells in the MB49 intravesical model is shown. (E) Heatmap of 190 significantly altered genes at an FDR cutoff 0.05 and log2 fold change of one between LV-IFNα and LV-Ctrl groups. (F) GSEA analysis shows enrichment of apoptosis, natural-killer-cell-mediated cytotoxicity, and Fc-gamma-receptor-mediated cytotoxicity (top panel) and EGFR tyrosine kinase inhibitor resistance, platinum drug resistance in LV-IFNα-treated tumors, and enrichment of bladder cancer gene set in LV-Ctrl group.

Figure 8

Expression of cell-type-specific immune cell markers in MB49 tumors Heatmap of expression comparing LV-IFNα versus LV-Ctrl (A) and LV-IFNα versus vehicle (B) in MB49 tumors. Asterisk indicates significantly altered genes.