Development of 5D3-DM1: A Novel Anti-Prostate-Specific Membrane Antigen Antibody-Drug Conjugate for PSMA-Positive Prostate Cancer Therapy

Abstract

Prostate cancer (PC) is a potentially high-risk disease and the most common cancer in American men. It is a leading cause of cancer-related deaths in men in the US, second only to lung and bronchus cancer. Advanced and metastatic PC is initially treated with androgen deprivation therapy (ADT), but nearly all cases eventually progress to castrate-resistant prostate cancer (CRPC). CRPC is incurable in the metastatic stage but can be slowed by some conventional chemotherapeutics and second-generation ADT, such as enzalutamide and abiraterone. Therefore, novel therapeutic strategies are urgently needed. Prostate-specific membrane antigen (PSMA) is overexpressed in almost all aggressive PCs. PSMA is widely used as a target for PC imaging and drug delivery. Anti-PSMA monoclonal antibodies (mAbs) have been developed as bioligands for diagnostic imaging and targeted PC therapy. However, these mAbs are successfully used in PC imaging and only a few have gone beyond phase-I for targeted therapy. The 5D3 mAb is a novel, high-affinity, and fast-internalizing anti-PSMA antibody. Importantly, 5D3 mAb demonstrates a unique pattern of cellular localization to the centrosome after internalization in PSMA(+) PC3-PIP cells. These characteristics make 5D3 mAb an ideal bioligand to deliver tubulin inhibitors, such as mertansine, to the cell centrosome, leading to mitotic arrest and elimination of dividing PC cells. We have successfully developed a 5D3 mAb- and mertansine (DM1)-based antibody-drug conjugate (ADC) and evaluated it in vitro for binding affinity, internalization, and cytotoxicity. The in vivo therapeutic efficacy of 5D3-DM1 ADC was evaluated in PSMA(+) PC3-PIP and PSMA(-) PC3-Flu mouse models of human PC. This therapeutic study has revealed that this new anti-PSMA ADC can successfully control the growth of PSMA(+) tumors without inducing systemic toxicity.

Keywords: 5D3 antibody; MCC linker; anti-PSMA antibody; antibody-drug conjugates (ADC); drug delivery; mertansine (DM1); prostate cancer; prostate-specific membrane antigen (PSMA); targeted therapy.

Figure 1.

Characterization of 5D3-DM1 ADC. (A) MALDI-TOF spectra of 5D3 mAb, intermediate 5D3-MCC, and 5D3-DM1 ADC. (B) DLS analysis and size distribution of 5D3 mAb (i) and 5D3-DM1 (ii). (C) Fused flow cytometric histogram of PSMA(+) PC3-PIP and PSMA(–) PC3-Flu cells labeled with 5D3-DM1-AF-488 (i) and 5D3-AF-488 (ii), showing that there is no change in the PSMA labeling properties of 5D3 after modification. Note: Histograms were plotted in red (PC3-PIP cells) and blue (PC3-Flu cells) for both 5D3-DM1-AF-488 and 5D3-AF-488 for direct comparison.

Figure 2.

Confocal fluorescence images of 5D3 mAb and 5D3-DM1 ADC in PSMA(+) PC3-PIP cells. Internalized 5D3-AF-488 (A) and 5D3-DM1-AF-488 (B) after 1 h. (C) Perinuclear localization of 5D3-AF-488. (D) Cells treated with 5D3-AF-488 (green) and dextran–rhodamine (red, visualizing endosomes), a dividing cell exhibiting the mAbs localized to the centrosomes. (Scale bar: 50 μm).

Figure 3.

In vitro cytotoxicity in PC3-PIP and PC3-Flu cells. (A) Cytotoxicity of 5D3-DM1 in PSMA(+) PC3-PIP (IC₅₀ = 0.70 nM) and PSMA(–) PC3-Flu (IC₅₀ = 13.98 nM) cells. (B) Cytotoxicity of DM1 equivalent to 5D3-DM1 in PSMA(+) PC3-PIP (IC₅₀ = 603.8 nM) and PSMA(–) PC3-Flu (IC₅₀ = 735.2 nM) cells. IC₅₀ values were calculated using the GraphPad/Prism software.

Figure 4.

In vivo therapeutic study. (A) Dosing schedule of in vivo therapy, imaging, and toxicological study. (B) In vivo NIR fluorescence images of drug delivery. The tumor uptake of 5D3-DM1-CF-680 in PSMA(+) PC3-PIP and PSMA(–) PC3-Flu dual-tumor mouse models. (i) Untreated and (ii) treated with 2.5 mg/kg of 5D3-DM1-CF-680. Images were obtained 24 h after the administration of ADC at day 15.

Figure 5.

In vivo therapeutic effect of 5D3-DM1-CF-680 in human PC xenograft mouse models (n = 6/group). Therapeutic efficacy of 5D3-DM1-CF-680 (1.0, 2.5, and 5.0 mg/kg) in (A) PSMA(+) PC3-PIP PC tumors and (B) PSMA(–) PC3-Flu PC tumors in mouse models. (C) Relative tumor volumes at day 21 in mice treated with multiple-doses for PC3-PIP and PC3-Flu tumors. The Kaplan–Meier analysis of the surrogate survival. Kaplan–Meier graphs were plotted using the time taken to reach four times the initial tumor size as the endpoint in (D) PSMA(+) PC3-PIP PC tumors and (E) PSMA(–) PC3-Flu PC tumors in mouse models (n = 10/group).

Figure 6.

Toxicological study of 5D3-DM1-CF-680 ADC therapy (n = 4/group). (A) BUN (healthy range 16–30 mg/dL), (B) ALT (healthy range 15–60 U/L), (C) AST (healthy range 50–100 U/L) concentrations for ADC-treated and saline-treated in PSMA(+) PC3-PIP and PSMA(–) PC3-Flu dual-tumor mouse models, and (D) changes in the animal body weight during the therapy relative to the initial body weight (n = 6/group).

Figure 7.

H&E staining of tumors and organs in a 5.0 mg/kg dose-treated mouse. (A) PSMA(+) tumor at 10× (i) and 40× (ii). (B) PSMA(–) tumor at 10× (i) and 40× (ii). (C) Liver tissue (i) and kidney tissue (ii) at 40×. Study shows high necrosis in the PSMA(+) PC3-PIP tumor, low necrosis in the PSMA(–) PC3-Flu tumor, and no toxicological damage in the liver or kidney. (Scale bars: 100 μm in 10× images and 50 μm in 40× images).

Scheme 1.. Synthesis of the 5D3-DM1 ADC and Its Fluorescence Analogues^a

^aIn the first step, 5D3 mAb was functionalized by reaction with the Sulfo-SMCC heterobifunctional linker (i) and conjugated with DM1 (ii). The resulting 5D3-DM1 (iii) and 5D3 mAbs (iv) were labeled with AF-488 (for in vitro imaging and flow cytometric analysis) or CF-680 (for in vivo optical imaging).