Dendritic cell based PSMA immunotherapy for prostate cancer using a CD40-targeted adenovirus vector

Abstract

Human prostate tumor vaccine and gene therapy trials using ex vivo methods to prime dendritic cells (DCs) with prostate specific membrane antigen (PSMA) have been somewhat successful, but to date the lengthy ex vivo manipulation of DCs has limited the widespread clinical utility of this approach. Our goal was to improve upon cancer vaccination with tumor antigens by delivering PSMA via a CD40-targeted adenovirus vector directly to DCs as an efficient means for activation and antigen presentation to T-cells. To test this approach, we developed a mouse model of prostate cancer by generating clonal derivatives of the mouse RM-1 prostate cancer cell line expressing human PSMA (RM-1-PSMA cells). To maximize antigen presentation in target cells, both MHC class I and TAP protein expression was induced in RM-1 cells by transduction with an Ad vector expressing interferon-gamma (Ad5-IFNγ). Administering DCs infected ex vivo with CD40-targeted Ad5-huPSMA, as well as direct intraperitoneal injection of the vector, resulted in high levels of tumor-specific CTL responses against RM-1-PSMA cells pretreated with Ad5-IFNγ as target cells. CD40 targeting significantly improved the therapeutic antitumor efficacy of Ad5-huPSMA encoding PSMA when combined with Ad5-IFNγ in the RM-1-PSMA model. These results suggest that a CD-targeted adenovirus delivering PSMA may be effective clinically for prostate cancer immunotherapy.

Figure 1. Growth characteristics of transfected cell lines in vitro.

Shown is a representative growth curve of the number of RM-1 cells (•) in culture with time compared with the number of RM-1-PSMA clone 1 cells (▾), RM-1-PSMA clone 3 cells (⧫), and RM-1-GFP clone 2 cells (▪). Each data point represents the mean ± standard error of three independent experiments.

Figure 2. Western blot analysis of PSMA protein levels in RM-1 mouse prostate cancer cell lines.

Protein extracts (20 µg) were prepared from each cell line and diluted with RIPA sample buffer. The extracts were separated by SDS-PAGE, electroblotted onto nitrocellulose and probed with an anti-human PSMA antibody directed to the C-terminal protein sequence (A) or an anti-human PSMA antibody directed to the N-terminal protein sequence (B), followed by treatment with corresponding anti-species IgG HRP conjugates. Shown are representative blots after visualization by autoradiography.

Figure 3. Western blot analysis of PSMA protein levels in adenovirus-infected dendritic cells.

(A) Protein extracts (20 µg) were prepared from cultured mouse dendritic cells after infection for 24 h with the CD40-targeted Ad5-luc1 or CD40-targeted Ad5-mPSMA expressing mouse PSMA or Ad5-huPSMA expressing human PSMA. (B) Protein extracts (20 µg) were prepared from cultured mouse dendritic cells after infection for 0, 24, or 48 h with the CD40- Ad5-huPSMA expressing human PSMA. Each cell extract was prepared using RIPA sample buffer. The extracts were separated by SDS-PAGE, electroblotted onto nitrocellulose membranes and probed with an anti-human PSMA antibody followed by treatment with the corresponding anti-mouse IgG HRP conjugates. Shown are representative blots after visualization by autoradiography. A loading control was used by co-treatment of the membranes with an anti-mouse α-tubulin antibody.

Figure 4. Flow cytometry analysis of MHC class I cell surface expression in RM-1 cells maintained in the absence and presence of IFNγ.

Evaluation of binding of anti-MHC class I (H-2Db and H-2Kb) antibodies to (A) RM-1 parental cells, (B) RM-1-GFP clone 2 cells, (C) RM-1-PSMA clone 1 cells, and (D) RM-1-PSMA clone 3 cells. Shown are histogram peaks corresponding to untreated (shaded) and cells infected with Ad5-IFNγ (unshaded).

Figure 5. Analysis of TAP expression in RM-1 cell lines induced by Ad5-IFNγ.

Protein extracts (20 µg) were prepared from each cell line and diluted with RIPA sample buffer. The extracts were separated by SDS-PAGE, electroblotted onto nitrocellulose and probed with (A) an anti-mouse TAP1 antibody or (B) an anti-mouse TAP2 antibody, followed by treatment with a corresponding anti-mouse IgG HRP conjugates. Shown are representative blots after visualization by autoradiography. A loading control was used by co-treatment of the membranes with an anti-mouse GAPDH antibody.

Figure 6. Effect of cell viability on cell lines after infection in vitro with Ad5-IFNγ.

Shown is a representative curve of the cell viability in (A) RM-1 cells or (B) RM-1-PSMA clone 3 cells infected in culture with increasing concentrations of Ad5-IFNγ compared with uninfected cells. Effect of cell viability was determined at day 1 (•), day 2 (▾), day 3 (⧫), and day 4 (▪) after initiation of infection with Ad5-IFNγ. Each data point represents the mean ± standard error of three independent experiments.

Figure 7. CTL assays of T-cells from mice after ex vivo infection of DCs with the CD40-targeted Ad5 vectors expressing PSMA or luciferase followed by intraperitoneal administration.

Mice were immunized as described in Material and Methods with dendritic cells infected with the CD40-targeted Ad5-luc1 or with the CD40-targeted Ad5-huPSMA. At day 24 after initiation of the experiment, the animals were euthanized, and CTL activity was measured in cells harvested from the spleens. Target cells used were RM-1 parental (•), RM-1-PSMA clone 1 cells (▾), RM-1-PSMA clone 3 cells (⧫), and RM-1-GFP clone 2 cells (▪). Target cells were either untreated (A and C) or pre-treated by infection with Ad5-IFNγ (B and D). Each data point represents the mean ± standard error of four replicate wells.

Figure 8. CTL assays of T-cells from mice after direct intraperitoneal administration of CD40-targeted vectors expressing PSMA or luciferase.

Mice were immunized as described in Material and Methods with the CD40-targeted Ad5-luc1 or with the CD40-targeted Ad5-huPSMA. At day 24 after initiation of the experiment, the animals were euthanized, and CTL activity was measured in cells harvested from the spleens. As a control, CTL activity was measured in cells harvested from spleens of naïve (untreated) mice. Target cells used were RM-1 parental (•), RM-1-PSMA clone 1 cells (▾), RM-1-PSMA clone 3 cells (⧫), and RM-1-GFP clone 2 cells (▪). Target cells were either untreated (A, C, and E) or pre-treated by infection with Ad5-IFNγ (B, D, and F). Each data point represents the mean ± standard error of four replicate wells.

Figure 9. Assessment of sensitivity to NK cell-mediated cytotoxicity.

(A) Cytotoxic activity of NK effector cells on YAC-1 and RM-1 target cells was determined by a ⁵¹Cr-release assay at indicated E:T ratios. Target cells used were YAC-1 (•), RM-1 parental (▾), RM-1 parental cells pre-treated by infection with Ad5-IFNγ (▪), RM-1-PSMA clone 3 cells (⧫), or RM-1-PSMA clone 3 cells pre-treated by infection with Ad5-IFNγ (▴). Each data point represents the mean ± standard error of four replicate wells. (B) Flow cytometry analysis of YAC-1 and RM-1 cells stained with PE labeled anti-H60, anti-MULT-1, or anti-Rae-1 antibodies or with an isotype-matched control antibody. Numbers below the histograms correspond to mean fluorescence intensity of each peak.

Figure 10. Assessment of tumor growth after vaccine treatment.

Antitumor effect of a CD40-targeted Ad5-huPSMA vaccine was determined using the RM-1-PSMA mouse model. Mice were immunized as described in Material and Methods with the CD40-targeted Ad5-luc1 or CD40-targeted Ad5-huPSMA. At day 24 after initiation of the experiment, each mouse received 4×10⁶ RM-1 parental cells or 4×10⁶ RM-1-PSMA clone 1 cells injected subcutaneously. Three days later, the treatment groups were injected at the site of tumor cell injection with 1×10⁸ ifu of Ad5-IFNγ or with normal saline. Beginning at the time of tumor cell challenge (day 0), the tumors were measured and volumes calculated by the formula: tumor volume = ½ × (length×width ²) where length is the longest distance of the tumor. Each data point represents the mean volume of 15 tumors ± standard error.