Novel prostate cancer immunotherapy with a DNA-encoded anti-prostate-specific membrane antigen monoclonal antibody

Cell lines used in this study were purchased from American Type Culture Collection (ATCC). The 293T(ATCC^®CRL-3216™) and transgenic adenocarcinoma mouse prostate (TRAMP)-C2 (ATCC^®CRL-2731™) cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco-Life Technologies) supplemented with 10% FBS and 1% penicillin/streptomycin. Lymph node carcinoma of the prostate (LNCaP) clone FGC (ATCC^®CRL-1740™) cells were maintained in RPMI-1640 (Gibco-Life Technologies) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin [19]. The commercial anti-PSMA control mAb was obtained from R&D systems.

To construct the PSMA-DMAb , the genes of both the variable heavy (V _H) and variable light (V _L) fragments of a human anti-PSMA mAb were examined, optimized, and constructed through the use of synthetic oligonucleotides with several modifications to improve expression as previously described [15]. DNA was formulated in water for subsequent administration into mice. An empty pVax1 expression vector was used as a negative control. Cells (293T) were transfected with the PSMA-DMAb plasmid and confirmation of PSMA-DMAb binding to recombinant human PSMA was carried out by Western blot analysis. Briefly, recombinant PSMA protein (R&D systems) was run on an SDS-PAGE gel and transferred to Immobilon-PVDF membrane (EMD Millipore). Membranes were blocked for 1 h in blocking buffer (Li-Cor Biosciences) and then incubated for 1 h with either commercial anti-PSMA mAb (R&D systems), pooled day 14 sera from PSMA-DMAb plasmid-injected mice, or supernatants from PSMA-DMAb plasmid-transfected 293T cells. Membranes were washed and then incubated for 1 h with a goat anti-human IgG 680RD antibody (Li-Cor Biosciences) and washed. Protein bands were visualized by scanning membranes with a Li-Cor Odyssey CLx scanner [19].

Animal experiments were conducted in accordance with the University of Pennsylvania Animal Care and Use Committee guidelines. B6.Cg-Foxn1 ^nu /J (C57BL/6 nude) and C57BL/6 (both from Jackson Laboratory) mice were administered 100 µg of PSMA-DMAb or pVax1 plasmid in a single 50 µl intramuscular injection into the quadriceps, followed by in vivo electroporation [15]. For quantifying human immunoglobulin G1 (IgG1) levels, ELISA plates were coated with 1 µg/well of goat anti-human IgG-Fc fragment antibody (Bethyl) overnight at 4 °C. The following day, plates were washed with phosphate-buffered saline with 0.1% Tween-20 (PBS-T), blocked with 10% FBS in PBS-T for 2 h at room temperature, washed, incubated for 1 h at room temperature with the respective samples that were diluted with 1% FBS in PBS-T, washed, and incubated for 1 h at room temperature with HRP-conjugated goat anti-human kappa light chain antibody (Bethyl). SIGMAFAST OPD (Sigma-Aldrich) solution was added to wells and plates kept in dark for at least 10 min for color to develop. The enzymatic reaction was stopped with 1 N H₂SO₄ and plates were read at 450 nm. A standard curve was generated using purified human IgG/Kappa (Bethyl) [15]. Binding ELISA to evaluate antibody affinity followed a similar procedure except plates were coated overnight with recombinant human PSMA and a HRP-conjugated goat anti-human IgG (H + L) (Bethyl) was used as a secondary antibody.

To detect cell surface PSMA, tubes of 1.0 × 10⁶ LNCaP or TRAMP-C2 cells were washed with phosphate-buffered saline (PBS), stained with live/dead fixable violet dead cell stain (Life Technologies) for 15 min, and then washed twice with FACS buffer (PBS + 1% FBS). Cells were next incubated for 30 min at room temperature with a 1:4 dilution of day 14 sera from PSMA-DMAb plasmid-injected mice and then washed. Finally, cells were incubated in the dark for 30 min with a 1:100 dilution of PE-conjugated anti-human Fc IgG (Biolegend), followed by a final wash with FACS buffer. Samples were resuspended in 1× stabilizing fixative (BD) and analyzed the following day on an LSR18 flow cytometer (BD Biosciences). FACS analysis was performed on a gated low forward scatter and side scatter with Annexin-V FITC and PI (Thermo Fisher) following kit protocol for the effects of PSMA-DMAb sera on LNCaP cell death.

Formalin-fixed paraffin-embedded (FFPE) human tumor tissue sections (UMass Cancer Center Tissue and Tumor Bank, Massachusetts, MA) were deparaffinized with xylene and rehydrated. Antigen retrieval was performed using a 1× working solution of citrate buffer, pH 6.0 (Sigma-Aldrich), at 100 °C for 15 min. Tissue sections were blocked with 1× PBS containing 5% normal goat serum (Cell Signaling Technology) and 0.3% Triton X-100 in a humid chamber. Tissues were washed in 1× PBS and incubated with pooled day 14 PSMA-DMAb plasmid-administered mice sera diluted 1:100 in antibody diluent. Tissues were washed in 1× PBS and incubated with a 1:500 dilution of Alexa Fluor 488-conjugated goat anti-human IgG (H + L) secondary antibody (Thermo Fisher Scientific) in antibody diluent for 1 h. Cell nuclei were counterstained with Hoechst reagent (Sigma-Aldrich). Images were acquired using the Leica TCS SP8 confocal laser scanning microscope at the cell and developmental biology microscopy core, University of Pennsylvania, PA, USA. Paraffin-embedded mouse prostate tissue was subjected to antigen retrieval and deparaffinized. Slides were then fixed with acetone and washed with PBS and sections blocked using normal goat serum followed by staining with human PSMA antibody, followed by a biotinylated goat anti-mouse and completion of immunohistochemical procedure according to the manufacturer’s instructions (Vector Labs).

ADCC activity of PSMA-DMAb was examined using Promega’s ADCC Reporter Bioassay Kit. Briefly, target LNCaP cells were incubated for 6 h at 37 °C with the engineered Jurkat effector cells and pooled day 14 sera from PSMA-DMAb plasmid-injected mice. Luciferase activity was measured by luminescence to determine ADCC activity as recommended by the manufacturer. All sera samples were tested in triplicate.

For tumor implantation, C57BL/6 male mice were injected subcutaneously with 1 × 10⁶ TRAMP-C2 cells in the right hind flank. The experimental mice were divided into treatment groups (n=10). Animals were monitored for tumor growth. As tumors became detectable, electronic calipers were used to measure the length and width of the tumor and the tumor volumes were calculated by applying the following equation: \({\text {volume\;{(V)}}} \; = {\frac{4}{3}\times{3.14159}\times\left(\frac {{\text{length}}}{2}\times\frac {{\text{width}}}{2}\times\frac {{\text{width}}}{2}\right)}\). Under the University of Pennsylvania Animal Care and Use Committee guidelines mice are sacrificed when tumor diameter reaches 2 cm, or when tumors became ulcerated. Survival differences between groups were analyzed by Students t test, p > 0.05 is considered significant.

Mice were treated for NK cell depletion on day −1 (before tumor challenge) and at days +2 and +4 after tumor inoculation with intravenous injection of 100 μl (25 μg) of either control IgG or anti-Asialo GM1 IgG (Wako Chemicals, Richmond, VA, USA) diluted in PBS. Cells were stained with anti-NK1.1 and anti-CD3 monoclonal antibodies and analyzed by flow cytometry to verify the depletion of the CD3⁻/NK1.1+ (NK) cell population in the anti-Asialo GM1-treated animals.

GraphPad Prism 6 (GraphPad Software, Inc.) program was used for statistical analysis of the data. The data from ELISA assays are expressed as mean ± SD and are representative of at least three different experiments. Comparisons between individual data points were made using Student’s t test. p values < 0.05 were considered to be statistically significant.

Human PSMA is a type II integral membrane glycoprotein that is highly expressed in prostate secretory-acinar epithelium as well as in several extra-prostatic tissues, and it possesses 86% identity and 91% similarity to mouse PSMA [20]. A plasmid capable of directing in vivo antibody production was designed by (1) creating a cassette consisting of the full-length coding sequences for the variable heavy (V _H) and light (V _L) immunoglobulin (Ig) chains from the published sequence of an anti-PSMA mAb driven off a CMV promoter; (2) optimizing the cassette sequence to improve its expression; and (3) cloning the cassette into a pVax1plasmid (Fig. 1a). Antibodies targeting PSMA produced from this optimized DNA plasmid will henceforth be referred to as PSMA-DMAb.

To confirm that the plasmid directs production of fully assembled IgG, human embryonic kidney 293T cells were transfected with either empty pVax1 or PSMA-DMAb plasmid. Supernatants collected from cells at 48 h post-transfection were assayed by ELISA to quantify total human IgG levels. A concentration of nearly 800 ng/ml of human IgG was measured in supernatants of PSMA-DMAb plasmid-transfected cells (Fig. 1b). A binding ELISA performed on the same supernatants indicated that the IgG produced from PSMA-DMAb plasmid-transfected cells bound to recombinant human PSMA with high affinity (Fig. 1c). Western blot analysis further confirmed the specificity of PSMA-DMAb plasmid-derived antibodies for binding to recombinant human PSMA protein (Fig. 1d). The results indicate that the PSMA-DMAb plasmid can direct the production of anti-PSMA-specific antibodies in vitro.

The ability of the PSMA-DMAb plasmid to direct antibody production in vivo was evaluated in both immune-deficient B6.Cg-Foxn1 ^nu /J (C57BL/6 nude) and immune-competent C57BL/6J mice. Groups of five mice received a single 100 μg injection of PSMA-DMAb plasmid intramuscularly in their quadriceps muscle followed by EP for enhanced delivery [16]. Injected mice were bled at various time points post-injection to obtain sera that was evaluated by ELISA to quantitate human IgG levels. Human IgG became detectable in sera of injected mice beginning on day 5 post-injection, with peak levels achieved at day 14 post-injection in both C57BL/6 nude (1.17 ± 0.41 μg/ml, Fig. 2a) and C57BL/6 (0.82 ± 0.11 μg/ml, Fig. 2b) mice. While elevated human IgG levels persisted in C57BL/6 nude mice beyond 50 days, the levels in C57BL/6 mice dropped to baseline values by day 35 post-injection likely due to the mouse anti-human antibody response [21, 22]. Serum collected at day 14 post-injection from PSMA-DMAb plasmid-injected C57BL/6 nude mice was evaluated by ELISA (Fig. 2c) and Western blot (Fig. 2d) to evaluate the affinity and specificity of serum IgG for recombinant human PSMA. Both assays show that the IgG in day 14 sera recognized human PSMA, but not irrelevant HIV envelope protein with high affinity and specificity, suggesting that the IgG are properly folded and functional PSMA-DMAb.

In vivo distribution of PSMA-DMAb in prostate tissue was studied in mice by harvesting tissues 7 days post-plasmid injection and performing ELISA and immunohistochemistry for IgG quantification. Prostate tissue from mice administered the PSMA-DMAb plasmid exhibited higher levels of human IgG compared to prostate tissue from empty pVax1 plasmid-injected mice as measured by ELISA of tissue homogenates (Fig. 2e). Further, prostate tissues were evaluated by immunohistochemistry staining for anti-human-Fc expression. A strong immunostaining signal was detected on the cell membranes and within the prostate for the PSMA-DMAb plasmid-injected mice, but not pVax1-treated controls (Fig. 2f). Together, these findings demonstrated that the PSMA-DMAb plasmid can direct the production of robust levels of PSMA-specific human IgG in vivo.

We next evaluated the ability of PSMA-DMAb in mouse sera to bind PSMA on tumor cells and tissues. Two PSMA-expressing prostate cancer cell lines were chosen for the initial studies: (1) LNCaP cells, derived from human prostate adenocarcinoma cells; and (2) transgenic adenocarcinoma mouse prostate (TRAMP)-C2 cells derived from a heterogeneous 32-week tumor grown in the TRAMP mouse model. Both cell lines were incubated sequentially with day 14 sera from pVax1 or PSMA-DMAb plasmid-injected C57BL/6 nude mice followed by a fluorescently labeled anti-human IgG secondary antibody. Histograms (Fig. 3a) and mean fluorescent intensity MFI (Fig. 3b) obtained from flow cytometry analysis of stained cells show that in vivo produced PSMA-DMAbs bind to both PSMA-positive tumor cell lines. No staining was observed on PSMA-negative PC3 cells (data not shown).

In addition to normal and cancerous prostate cells, several studies have reported PSMA expression on a wide variety of tumors, especially on tumor neovasculature [23, 24]. Immunofluorescence assays were used to evaluate the ability of PSMA-DMAb to bind to PSMA expressed on tissue sections of human bladder and kidney tumors (Fig. 4). The results show that PSMA-DMAb was able to stain cells in the bladder and kidney tumor tissue sections, but not cells in normal ovarian tissues, confirming previous reports of PSMA expression in these tumors [25]. Furthermore, the staining shows that PSMA distribution is homogeneous throughout the bladder and kidney tumor sections. This data confirms that a PSMA-DMAb retains specificity for PSMA and specifically binds PSMA on the surface of human tumor cells.

The biological activity of PSMA-DMAb was next evaluated by using an antibody-dependent cell-mediated cytotoxicity (ADCC) mechanism of action assay [26, 27]. The assay involves incubating PSMA-expressing LNCaP cells with effector cells for 6 h in the presence of different concentrations of serum from pVax1 or PSMA-DMAb plasmid-injected mice. The effector cells are Jurkat cells that stably express high-affinity V158 FcγRIIIa and a gene for firefly luciferase driven off a nuclear factor of activated T cell (NFAT) response element [28]. The assay readout is based on activation of gene transcription in effector cells as measured by firefly luciferase production. As indicated in Fig. 5a, day 14 serum from PSMA-DMAb plasmid-injected mice mediates an ADCC effect.

As a second demonstration of the biological activity of PSMA-DMAb, flow cytometry was used to measure apoptosis and necrosis of LNCaP cells that were co-cultured with human PBMCs in the presence of sera from pVax1 or PSMA-DMAb plasmid-injected mice. The results (Fig. 5b) show that there was a statistically significant increase in apoptosis (Q3 section of the histogram) as well as necrosis (Q2 section in the histogram) for LNCaP cells co-cultured with human PBMCs in the presence of PSMA-DMAb in comparison to control pVax1 sera. Combined, these findings show that the synthetic PSMA-DMAb can bind Fc receptors and mediate an ADCC effect on tumor cells [29].

In vivo functional activity of PSMA-DMAb was assessed using a TRAMP-C2 tumor challenge mouse model [30]. For this assay, C57BL/6 mice were subcutaneously implanted with 1 × 10⁶ TRAMP-C2 tumor cells and then injected 1 week later with 100 μg of either pVax1 or PSMA-DMAb plasmid by intramuscular injection with enhanced EP [30]. Mice were followed for up to 56 days with regular measurements of tumor size made on each mouse during this period (Fig. 6a). Tumors in the pVax1-treated mice began to grow at day 7–10 post-implantation, while tumors were not detectable in PSMA-DMAb-treated mice until days 15–17. Rapid tumor growth was noted for the control groups (pVax1), but the PSMA-DMAb-treated group exhibited an obvious suppression of tumor growth due to the antibody-mediated tumor-protective immunity. Over the course of the 56-day observation period, there was a statistically significant reduction in average tumor volumes (p = 0.0201) (Fig. 6b) and a significant improvement in survival (p = 0.0280) in mice receiving the PSMA-DMAb construct compared to the control mice (Fig. 6c). It is likely that this effect might be further enhanced in the absence of the mouse anti-human antibody response. Visual inspection of tumors (Fig. 6d) that developed in each group revealed that the tumors in the PSMA-DMAb group were impacted early and remained small and subdermal, while tumors in the pVax1 control group protruded out of the skin and became ulcerated.

The anti-tumor activity of many therapeutic antibodies including ADCC and antibody-dependent cellular phagocytosis (ADCP) is dependent on the interaction of the IgG-Fc domain with Fc gamma receptors (FcγRs) on effector cells. Natural killer cells express high levels of FcγRs, therefore we also examined the contribution of NK cells to the observed effects of PSMA-DMAb on tumor growth. Previous studies have reported that human IgG can bind to all activating mouse FcγRs and can induce ADCC/ADCP with mouse NK cells and mouse macrophages [29]. Groups of mice were treated with either control IgG or the NK cell-depleting anti-AGM1 IgG antibody and then implanted with TRAMP-C2 cells. One week later, mice were given a single injection of either pVax1 or PSMA-DMAb plasmid and were subsequently evaluated for tumor growth up to 56 days. There was a rapid onset of tumor development, accelerated tumor growth, and decreased survival in PSMA-DMAb-immunized, NK cell-depleted mice (p = 0.0019, Fig. 6e), but not in those pretreated with the control IgG. Taken together, these data demonstrate that PSMA-DMAb can exert a profound therapeutic effect on a PSMA-expressing tumor in vivo, supporting the possible application of this therapy for the treatment of prostate cancer.