Oncolytic adenovirus expressing bispecific antibody targets T-cell cytotoxicity in cancer biopsies

Abstract

Oncolytic viruses exploit the cancer cell phenotype to complete their lytic life cycle, releasing progeny virus to infect nearby cells and repeat the process. We modified the oncolytic group B adenovirus EnAdenotucirev (EnAd) to express a bispecific single-chain antibody, secreted from infected tumour cells into the microenvironment. This bispecific T-cell engager (BiTE) binds to EpCAM on target cells and cross-links them to CD3 on T cells, leading to clustering and activation of both CD4 and CD8 T cells. BiTE transcription can be controlled by the virus major late promoter, limiting expression to cancer cells that are permissive for virus replication. This approach can potentiate the cytotoxicity of EnAd, and we demonstrate using primary pleural effusions and peritoneal malignant ascites that infection of cancer cells with the BiTE-expressing EnAd leads to activation of endogenous T cells to kill endogenous tumour cells despite the immunosuppressive environment. In this way, we have armed EnAd to combine both direct oncolysis and T cell-mediated killing, yielding a potent therapeutic that should be readily transferred into the clinic.

Keywords: BiTE; adenovirus; bispecific T‐cell engager; oncolytic virus.

© 2017 The Authors. Published under the terms of the CC BY 4.0 license.

Figures

Figure 1. Characterisation of EpCAM BiTE and its effects on PBMC‐derived T cells

1. Schematic of the structure of the EpCAM‐targeted BiTE and non‐specific control BiTE. The VL and VH domains are connected with flexible peptide linkers (L) rich in serine and glycine for flexibility and solubility. SP, light chain immunoglobulin signal peptide; His, decahistidine affinity tag.

2. Induction of activation markers (i) CD69 and (ii) CD25 on CD3‐purified PBMC cultured alone or with DLD cells (5:1) in the presence of BiTE‐containing supernatants. CD69 and CD25 were measured by flow cytometry after 24 h of co‐culture. Significance was assessed versus IgG isotype.

3. Percentage of IFNγ‐positive T cells after 6 h in co‐culture with DLD cells (5:1) and BiTE‐containing supernatants.

4. Proliferation, represented by division index and percentage of parental T‐cell population entering proliferation, of CFSE‐stained T cells in co‐culture with DLD cells (5:1) and BiTE‐containing supernatants. Fluorescence was measured by flow cytometry 5 days after co‐culture. Division index was modelled using FlowJo proliferation tool.

5. Degranulation of T cells, measured by CD107a externalisation, in co‐culture with DLD cells (5:1) and BiTE‐containing supernatants. Externalisation was assessed by co‐culture with a CD107a‐specific antibody for 6 h followed by flow cytometry analysis.

6. Cytokine levels were measured by LEGENDplex human Th cytokine panel using supernatants from co‐cultures of T cells with DLD cells (5:1) in the presence of BiTE‐containing supernatants for 48 h.

Data information: Each condition was measured in biological triplicate and data represented as mean ± SD. Significance was assessed versus untreated unless stated otherwise using a one‐way ANOVA test with Tukey's post hoc analysis, *P < 0.05, **P < 0.01, ***P < 0.001.Source data are available online for this figure.

Figure EV1. Characterisation of recombinant EpCAM BiTE

1. Dot blot to estimate the quantity of EpCAM BiTE produced by transfected HEK293A cells.

2. ELISA measuring the level of EpCAM binding by controls or recombinant EpCAM or non‐specific BiTE. Each condition was measured in biological triplicate and represented as mean ± SD. Significance was assessed by comparison to empty vector control sample using a one‐way ANOVA test with Tukey's post hoc analysis, ***P < 0.001.

Source data are available online for this figure.

Figure 2. Assessment of antigen specificity of EpCAM BiTE‐mediated T‐cell cytotoxicity

1. Induction of activation marker CD25 on CD3+ T cells in co‐culture with CHO or CHO‐EpCAM cells (5:1) and BiTE‐containing supernatants, measured by FACS analysis after 24 h of co‐culture.

2. Cytotoxicity of CHO or CHO‐EpCAM cells cultured with BiTE‐containing supernatants alone or in co‐culture with T cells. Cytotoxicity was assessed by release of LDH into the culture supernatants after 24 h of incubation.

3. Cytotoxicity of multiple EpCAM‐positive carcinoma cells after 24 h in co‐culture with T cells (1:5) and BiTE‐containing supernatants. Viability was measured by MTS assay after 24 h of co‐culture.

4. Levels of EpCAM expression (N = 1) assessed by FACS analysis of EpCAM‐positive cell lines in (C), compared to background fluorescence measured by using an isotype control antibody.

Data information: (A–C) Each condition was measured in biological triplicate and represented as mean ± SD. Significance was assessed versus untreated or T cell‐only controls using a one‐way ANOVA test with Tukey's post hoc analysis, *P < 0.05, **P < 0.01, ***P < 0.001. Source data are available online for this figure.

Figure 3. Identification of which T cells are responsible for BiTE‐mediated cytotoxicity

1. BiTE‐mediated T‐cell activation of CD4 and CD8 cells 24 h after co‐culture of CD3 T cells with DLD cells (5:1) and BiTE‐containing supernatant. Activation was assessed by surface expression of CD69 and CD25 and measured by flow cytometry.

2. Proliferative response of CFSE‐stained CD4 and CD8 T cells in co‐culture with DLD cells and incubated with BiTE‐containing supernatants. Fluorescence was measured after 5‐day incubation, by FACS analysis.

3. Degranulation of CD4 and CD8 cells following 6‐h co‐culture with DLD cells and BiTE‐containing supernatants. A CD107a‐specific antibody is added to the culture media for the duration of the co‐culture, and degranulation is assessed by flow cytometry.

4. Cytotoxicity by either the CD4 or CD8 T‐cell subset is assessed by LDH release into supernatant, following 24‐h incubation of DLD cells with CD4‐ or CD8‐purified T cells (1:5) and BiTE‐containing supernatant.

Data information: Each condition was measured in biological triplicate and represented as mean ± SD. EpCAM BiTE treatment was compared to control BiTE unless stated otherwise, and significance was assessed using a one‐way ANOVA test with Tukey's post hoc analysis, **P < 0.01, ***P < 0.001. Source data are available online for this figure.

Figure 4. Characterisation of oncolytic virus EnAd expressing EpCAM BiTE using cell lines and PBMC‐derived T cells

1. Schematic representation of EnAdenotucirev engineered to express the BiTE transgenes downstream of a CMV promoter or splice acceptor site (P) followed by a polyadenylation sequence (pA).

2. DLD cells were infected with parental EnAd or recombinant virus (100 vp/cell; 2.5e7 vp/ml) and wells harvested at 24 or 72 h. Replication was assessed by measuring genomes using qPCR against viral hexon.

3. Cytotoxicity of DLD cells infected with EnAd or recombinant virus at increasing concentrations of virus. Cytotoxicity was measured by MTS assay after 5 days of infection.

4. Supernatants from day 3 uninfected or virus‐infected HEK293A cells were assessed for transgene expression by immunoblot analysis and probed with an anti‐His antibody.

5. Induction of activation marker CD25 of CD3‐positive T cells cultured with CHO or CHO‐EpCAM (E:T 5:1) and diluted HEK293A supernatants from (D). Activation was measured by surface expression of CD25 by flow cytometry.

6. Cytotoxicity of CHO or CHO‐EpCAM cells incubated with HEK293A supernatants from (D) alone or in co‐culture with CD3‐purified PBMC (E:T 5:1). HEK293A supernatants were diluted 300‐fold. Cytotoxicity was assessed by LDH released into the supernatant after 24‐h incubation.

Data information: Each condition was measured in biological triplicate and represented as mean ± SD. Significance was assessed using a one‐way ANOVA test with Tukey's post hoc analysis with each condition compared to untreated, ***P < 0.001. Source data are available online for this figure.

Figure EV2. Cytotoxicity of EnAd expressing EpCAM BiTE in SKOV3 cells

1. A

   A quality control of viruses used in this study. The virus titre (vp/ml) was measured by PicoGreen assay and infectious particles (PFU/ml) by TCID50.

2. B, C

   SKOV3 cells were incubated with EnAd or recombinant viruses in the absence (B) or presence (C) of T cells and cytotoxicity was measured by LDH release at the specified time points. Each condition was measured in biological triplicate and represented as mean ± SD. Significance was assessed by comparison to uninfected control wells using a one‐way ANOVA test with Tukey's post hoc analysis, **P < 0.01, ***P < 0.001.

Source data are available online for this figure.

Figure 5. Superior potency of EnAd expressing EpCAM BiTE in partially EnAd‐resistant cancer cell line

1. A, B

   Viability of SKOV3 cells was monitored in real time over 160 h by xCELLigence‐based cytotoxicity assay. SKOV3 cells were seeded and infected with EnAd or BiTE‐armed EnAd viruses at 0 h, with uninfected cells serving as a negative control. In (B), CD3‐purified PBMCs (5:1) were added 2 h post‐infection and impedance was measured at 15‐min intervals.

2. C, D

   CD3‐purified PBMCs were cultured with SKOV3 cells (5:1) that were infected with parental EnAd or recombinant armed viruses. At each time point, T cells were harvested and analysed for surface expression of CD69 (C) or CD25 (D) by flow cytometry.

3. E

   Time‐lapse sequences showing co‐cultures of SKOV3 carcinoma cells (unstained), NHDF (red) and CD3‐purified PBMC (blue), infected with EnAd, EnAd‐CMV‐EpCAMBiTE or uninfected. Apoptosis was visualised using CellEvent Caspase 3/7 detection reagent (green). Images were taken on a Nikon TE 2000‐E Eclipse inverted microscope at intervals of 15 min covering a period of 72 h. Representative images were recorded at the times displayed; original magnification ×10; scale bar, 100 μm. Full time‐lapse sequences are displayed in Movies EV1, EV2 and EV3.

Data information: (A–D) Each condition was measured in biological triplicate and represented as mean ± SD (in A and B, solid line and dotted line, respectively). Significance was assessed by comparison to uninfected control using a one‐way ANOVA test with Tukey's post hoc analysis, *P < 0.05, ***P < 0.001. Source data are available online for this figure.

Figure EV3. Cytotoxicity and T‐cell activation by EnAd expressing EpCAM BiTE in DLD cells

1. A, B

   Cytotoxicity for infected DLD cells in the absence (A) or presence of T cells (B). DLD cells were infected and co‐cultured with T cells, and cytotoxicity was measured by LDH release at the specified time points.

2. C, D

   T cells from (B) were harvested and stained for activation markers CD69 (C) or CD25 (D) and analysed via flow cytometry.

3. E

   Quantification of EpCAM BiTE produced from DLD cells infected with recombinant viruses. Standard curve of LDH released (Abs) of DLD cells in co‐culture with CD3+ cells and serial dilutions of a known quantity of recombinant EpCAM BiTE (Ei). In parallel, co‐cultures were incubated with diluted supernatants (10,000‐fold) from 3‐day infected DLD cells (Eii). Standard curve allowed the approximate determination of EpCAM BiTE produced at 165 and 50 μg per million DLD cells for EnAd‐CMV‐EpCAMBiTE and EnAd‐SA‐EpCAMBiTE, respectively.

Data information: Each condition was measured in biological triplicate (A–D) or duplicate (E) and represented as mean ± SD (A–D). Significance was assessed by comparison to uninfected control wells using a one‐way ANOVA test with Tukey's post hoc analysis, **P < 0.01, ***P < 0.001.Source data are available online for this figure.

Figure 6. EnAd expressing EpCAM BiTE can selectively kill primary human tumour cells from chemotherapy‐pretreated patients

1. A, B

   Cytotoxicity of EpCAM+ cells (A) or FAP+ fibroblasts (B), first isolated from three patients' ascites and expanded ex vivo, then incubated with recombinant BiTE, or infected with EnAd or recombinant virus. Cytotoxicity was measured by flow cytometry after 5 days.

2. C

   Induction of activation marker CD25 on CD3‐positive T cells cultured with ascites‐derived EpCAM+ and FAP+ cells from (A and B).

Data information: Each condition was measured in biological triplicate and represented as mean ± SD. Significance was assessed by comparison to untreated using a one‐way ANOVA test with Tukey's post hoc analysis, *P < 0.05, **P < 0.01, ***P < 0.001. Source data are available online for this figure.

Figure 7. EnAd expressing EpCAM BiTE can overcome immune‐suppressive effects of pleural effusion and ascites fluid

1. A

   Quantity of human IL‐10 in primary malignant exudates. NS, normal serum; A, ascites fluid; P, pleural effusion.

2. B, C

   PBMC‐derived T cells were incubated with anti‐CD3/CD28 beads in normal serum or exudate fluid (50%) from 12 cancer patients. Induction of T‐cell activation markers CD69 and CD25 at 24 h (B) and degranulation (CD107a externalisation) at 6 h (C) were analysed using flow cytometry.

3. D, E

   CD3‐purified PBMCs were co‐cultured with SKOV3 (5:1) and EpCAM or control BiTE in the presence of normal serum or either pleural effusion or ascites fluid (50%) from seven patients. CD69 and CD25 dual positivity (D) and CD107a externalisation (E) on CD3+ T cells were measured by flow cytometry at 24 and 6 h, respectively.

4. F

   Viability of SKOV3 cells was monitored in real time over 130 h by xCELLigence‐based cytotoxicity assay. SKOV3 cells were seeded and incubated with EnAd‐SA‐ControlBiTE or EnAd‐SA‐EpCAMBiTE at 16 h. Uninfected cells served as a negative control. CD3‐purified PBMCs (5:1) were added 2 h post‐infection in the presence of culture medium or either pleural effusion or ascites fluid. Impedance was measured at 10‐min intervals.

Data information: (A–E) Each condition was measured in biological triplicate and represented as mean ± SD. Significance was assessed by comparison to normal serum using a one‐way ANOVA test with Tukey's post hoc analysis. *P < 0.05, **P < 0.01, ***P < 0.001.Source data are available online for this figure.

Figure 8. EnAd expressing EpCAM BiTE can activate endogenous T cells to kill endogenous tumour cells within culture media and malignant exudate fluid

1. A

   Representative spectra (ascites sample, patient 1) demonstrating screening of exudate fluid for its cellular composition, as assessed by flow cytometry. Proportion of each cell type is documented in the table.

2. B

   Endogenous unpurified ascites cells (patient 1) were incubated in RPMI culture medium or 50% ascites fluid in the presence of free EpCAM or control BiTE. After 24 h, the total cell population was harvested, and the number of CD3+CD69+CD25+ cells measured by flow cytometry.

3. C–E

   Total ascites cells (patient 1) were incubated in RPMI culture medium or 50% ascites fluid and treated with free BiTE, EnAd or EnAd expressing BiTE. After 5 days, CD25 expression of endogenous T cells (C) and the quantity of CD3+ (D) and EpCAM+ (E) cells were measured by flow cytometry.

Data information: Each condition was measured in biological triplicate and represented as mean ± SD. Significance within each fluid treatment was assessed by comparison to untreated using a one‐way ANOVA test with Tukey's post hoc analysis. Significance between fluid treatments was assessed by two‐way ANOVA test with Bonferroni post hoc analysis, *P < 0.05, **P < 0.01, ***P < 0.001. Source data are available online for this figure.

Figure EV4. Replication (SA promoter)‐dependent GFP transgene is selectively expressed in EpCAM + tumour cells

1. A, B

   Total cells from a peritoneal ascites sample were incubated with EnAd‐CMV‐GFP or EnAd‐SA‐GFP, with uninfected cells serving as a negative control. After 3 days, wells were imaged by bright‐field or fluorescence microscopy. Original magnification ×10; scale bar, 100 μm (A). Total cells were harvested, and the proportion of GFP+ cells that were CD3‐, CD11b‐ or EpCAM‐positive determined using flow cytometry (B). Each condition was measured in biological quadruplicate and represented as mean ± SD. Significance was assessed by comparison to uninfected control wells using a one‐way ANOVA test with Tukey's post hoc analysis, *P < 0.05, **P < 0.01, ***P < 0.001.

Source data are available online for this figure.

Figure 9. EnAd‐EpCAMBiTE shows reproducible activity within an expanded cohort of patient malignant peritoneal and pleural exudates

1. A–C

   Unpurified total cells from ascites or pleural effusions (from seven different patients; pleural effusion, blue; peritoneal ascites, red) were incubated in 50% fluid from the same exudate sample in the presence of free BiTE, EnAd or recombinant virus. After 5 days, the total cell population was harvested, and the number of (A) CD3+ T cells and those which were (B) CD25+ was quantified. (C) The number of EpCAM+ cells was measured using flow cytometry.

2. D

   Representative microscopy images (magnification ×10; scale bar, 100 μm) and flow cytometry analysis of pleural effusion cells of patient 3 (cancer cells and lymphocytes) following treatment with EnAd or EnAd‐SA‐EpCAM BiTE. Pink, total cells; blue, CD3+ cells; red, CD3+CD25+ cells; black, EpCAM+ cells.

3. E

   At 5 days, cytokine levels were measured by LEGENDplex human Th cytokine panel using pleural effusion cultures following incubation with free recombinant BiTE or infection with EnAd or recombinant virus. E:T (CD3+:EpCAM+) ratio upon receipt of patient samples were 0.51 (patient 1), 37.6 (patient 2), 0.39 (patient 3), 4.44 (patient 4), 0.14 (patient 5), 6.12 (patient 6) and 30.6 (patient 7).

Data information: Each condition was measured in biological triplicate and represented as mean ± SD. Significance was assessed by comparison to untreated control samples using a one‐way ANOVA test with Tukey's post hoc analysis, *P < 0.05, **P < 0.01, ***P < 0.001. Source data are available online for this figure.

Figure EV5. Expression of PD‐1 on PBMC and malignant exudate T cells

1. The expression of PD‐1 by endogenous T cells following their initial isolation from PBMC, ascites and pleural effusions was assessed by flow cytometry. PD‐1 expression on CD3 cells within each donor sample was measured once and represented as mean ± SD of all measured samples.

Source data are available online for this figure.