A PSMA-Targeted Doxorubicin Small-Molecule Drug Conjugate

Abstract

We developed a model small-molecule drug conjugate (SMDC) that employed doxorubicin as a representative chemotherapeutic targeted to the cell membrane biomarker PSMA (prostate-specific membrane antigen) expressed on prostate cancer cells. The strategy capitalized on the clatherin-mediated internalization of PSMA to facilitate the selective uptake and release of doxorubicin in the target cells. The SMDC was prepared and assessed for binding kinetics, plasma stability, cell toxicity, and specificity towards PSMA expressing prostate cancer cell lines. We observed high affinity of the SMDC for PSMA (IC₅₀ 5 nM) with irreversible binding, as well as specific effectiveness against PSMA(+) cells. These findings validated the strategy for a small molecule-based approach in targeted cancer therapy.

Graphical Abstract

Prostate-specific membrane antigen (PSMA) is the hallmark biomarker for prostate cancer because it is expressed on nearly all prostate cancers and increased expression correlates with progression to castration resistance and metastatic disease.^1–4 In addition to this unique expression in prostate cancer, which has led to the development targeted delivery of imaging and therapeutic agents by our group and others, the endothelial expression of PSMA on the neovasculature of non-prostatic tumors^5–7 offers a broader opportunity for the delivery of chemotherapy and other agents to a variety of non-prostatic vascularized solid tumors.

Our efforts in the field of PSMA-targeting agents over the past several years has led to the development of a PET imaging agent (CTT1057)⁸, which will soon complete Phase III clinical trials, as well as radiotherapeutic agent that exhibited remarkable tumor uptake and in vivo efficacy.⁹ We are now focused on developing an effective PSMA-targeted chemotherapeutic agent in the context of a small-molecule drug conjugate (SMDC). We recognize the significant developments in this area of PSMA-targeted chemotherapy by other teams, but it is our understanding that key properties of such agents can be optimized to improve efficacy. The most notable of the PSMA-targeted antibody-drug conjugates (ADCs) employ the J591 antibody that recognizes an extracellular epitope of PSMA. Antitumor activity has been demonstrated for J591 conjugated to the ribosome-inactivating protein toxin saporin¹⁰ or to the antimitotic agent monomethyl auristatin E (MMAE).^{11, 12} While the latter PSMA ADC progressed to Phase II clinical trials in castration-resistant prostate cancer patients, the overall risk-to-benefit of this ADC in late-stage prostate cancer has yet to be determined.

PSMA-targeted SMDCs have also shown promise⁴⁰, the most noteworthy being EC1169 in which the anti-angiogenic agent tubulysin-B is conjugated to DUPA, a reversible urea-based PSMA-inhibitor/ligand. Tubulysin-B and DUPA are connected via a disulfide linker,¹³ however there is no clear advantage in using disulfide linkers for targeted drug-conjugate applications.¹⁴ While in vivo antitumor activity has been observed for this SMDC, uptake in PSMA(+) tumor xenografts decreases considerably after 2 h. Regardless, it has advanced to Phase I clinical trials but the trial has been narrowed to only patients that were previously treated with taxane.¹⁵ In our preliminary studies, we developed a PSMA-targeted SMDC outfitted with MMAE as the cytotoxic payload coupled to a phosphoramidate-based PSMA targeting molecule through a pH-responsive cleavable linker.³⁶ Indeed, this molecule exhibited considerable efficacy compared to MMAE alone.

In this second proof-of-concept study, we present the modular assembly and a preliminary in vitro performance evaluation of another PSMA-targeted SMDC. The design criteria we set for the development of a PSMA-targeted SMDC included the following properties: high-affinity binding to PSMA, rapid and extensive internalization into PSMA(+) cells. Like other groups in the field, we have developed high-affinity ligands to PSMA based on chemical structures that inhibit PSMA’s enzymatic activity. However, we have focused on phosphoramidate-based inhibitors that bind irreversibly to PSMA and can deliver a diverse portfolio of payloads to PSMA(+) cells.^16–27 In addition to binding irreversibly to PSMA, these inhibitors internalize rapidly and extensively into PSMA(+) cells as the enzyme-inhibitor complex.^{21–24, 28} This clatherin-mediated internalization of PSMA is an ideal characteristic for targeted drug delivery.^1–5 In terms of the linker, we chose the valine-citrulline dipeptide, a commonly utilized linker among many ADCs currently in clinical development.^37–39 The SMDC structure of this study (Figure 1) was based on the PSMA-targeting molecule (CTT1298) that we had previously developed for a PET-imaging of prostate cancer.^{8, 29} To assemble our SMDC, we selected a modular approach, which allows convenient selection of various payloads while utilizing the same PSMA-targeting platform (Figure 1). The click-ready doxorubicin-linker module 1 was prepared through a modified procedure to include a terminal azide.³⁸ Compound 2 was prepared to include a terminal-DBCO as the click-ready PSMA-targeting module.³⁰ The copper-free click reaction of Dox-SMDC was accomplished in a THF/water solution based on established methods.³⁵ This modular approach has proved successful in the assembly of our PSMA-targeted imaging and radiotherapeutic agents.^{9, 31, 32}

Figure 1.

Modular assembly of PSMA-targeted doxorubicin conjugate (Dox-SMDC) and its mechanism of release upon internalization. PSMA-targeting motif in orange, click-chemistry motif in green, valine-citrulline linker in blue, and doxorubicin in red

The IC₅₀ for Dox-SMDC against PSMA and its mode-of-binding was determined as previously described,^6–8 with minor modifications.³⁰ Consistent with our other PSMA inhibitor-conjugates, the binding affinity of the targeting molecule was not reduced by the presence of the drug payload.^16–27 The IC₅₀ for Dox-SMDC against PSMA was 5 nM and exhibited an irreversible mode of binding to PSMA, which is typical for phosphoramidate-based PSMA inhibitors.^{21–24, 28, 35}. The IC₅₀ for doxorubicin alone against PSMA was determined to be greater than 10 μm, indicating negligible binding to PSMA.

Due to its intrinsic fluorescence, doxorubicin serves to be a valuable tool in research and imaging.³⁰ To determine Dox-SMDC’s selective binding and uptake, the intracellular localization of the Dox-SMDC in C4-2B and PC-3 cells were observed through laser confocal scanning microscopy (Figure 2). Fluorescence uptake was observed in the PSMA(+) C4-2B cells but not in the PSMA(−) PC3 cells. The presence of CTT1057 (a subnanomolar PSMA inhibitor) completely blocked fluorescence in C4-2B cells supporting the observation that Dox-SMDC was selective for PSMA(+) cells.

Figure 2:

Confocal images of PSMA(+) and PSMA(−) cells treated with Dox-SMDC. C4-2B and PC-3 cells were incubated with Dox-SMDC (10 μM) for 30 min. C4-2B cells were also pre-incubated with a blocker (CTT1057, 100 μM) and then incubated with Dox-SMDC (10 μM) for 30 min. All cells were treated with DAPI after fixation to detect nuclei. Representative images A (DAPI; blue) and B (Dox-SMDC; red) were merged to obtain images C.

The in vitro performance and selectivity of the SMDC was assessed using both PSMA(+) C4-2B and PSMA(−) PC-3 cell lines (Figure 2). While Dox-SMDC was less potent than free doxorubicin against C4-2B cells, it was more effective against PSMA(+) C4-2B cells in a concentration-dependent manner compared to the PSMA(−) PC-3 cells. These results confirmed the greater selective toxicity of Dox-SMDC for PSMA(+) cells. However, the apparent concentration-dependent toxicity of Dox-SMDC for the PSMA(−) PC-3 cells could not be discounted. Based on our experience with PSMA-targeted nanoparticles, non-specific uptake is not uncommon for aggressive prostate cancer cells.³³ Indeed, others have observed the non-specific spontaneous uptake of non-targeted nanoparticles by cancer cells that traffic differently depending on phenotype.³⁴ To determine whether the observed uptake for Dox-SMDC in the PSMA(−) PC3 cells could be attributed to the known non-specific uptake nanoparticles in cancer cells, we investigated the likelihood for the formation of nanoparticles by Dox-SMDC using dynamic light scattering (DLS). Indeed, we observed an average particle size distribution of 103 nm. Based on the these DLS results it is likely that mode of uptake of Dox-SMDC in PC3 cells, and to some extent in the C4-2B cells, is through mechanisms previously described.³⁴

In conclusion we have developed a PSMA-targeted SMDC outfitted with doxorubicin that exploits the specificity and internalizing nature of our irreversible PSMA inhibitors. The modular assembly of our SMDC is amenable to the preparation of subsequent analogs for which the chemotherapeutic payload can be conveniently substituted. Dox-SMDC prepared for this study maintained the high-affinity for PSMA of the targeting molecule as well as its irreversible mode of binding. While Dox-SMDC was expectedly not as potent as doxorubicin alone, it exhibited significant potency for PSMA(+) cells with additional potency due to non-specific uptake known for cancer cells. With these promising results, we are currently pursuing second-generation analogs to identify a clinically relevant targeted chemotherapeutic agent for prostate cancer as well as other non-prostatic malignancies characterized by neovasculature expression of PSMA.

Figure 3.

Concentration dependent toxicity of doxorubicin (Dox) against PSMA(−) PC-3 cells and PSMA(+) C4-2B cells (A), and concentration dependent toxicity of Dox-SMDC against PSMA(−) PC-3 cells and PSMA(+) C4-2B cells (B). Cells were incubated with doxorubicin for 72 h, and then assessed for viability using CellTiter-Glo. Each assay was performed with three technical replicates and repeated with three different cell passages (n = 3). Standard deviation represented as error bars. An unpaired, Welch’s t-test (two-tailed, 95% confidence interval) was performed to determine statistical significance between different concentrations. **** p <0.0001.