2024 Volume 72 Issue 2 Pages 136-142
In the aging global population, prostate cancer is a worldwide health problem because the incidence rate of this disease increases at advanced ages. Although early-stage prostate cancer can be treated by total prostatectomy, the surgery causes side effects, such as incontinence and dysuria, that lower QOL. Once the disease progresses to metastatic castration-resistant prostate cancer (mCRPC), there are no effective chemotherapeutic agents without systematic side effects. Therefore, targeted therapies for mCPRC are urgently needed. Traditional antibody–drug conjugate treatments for prostate cancer have been tested in clinical trials and several side effects have been observed. Meanwhile, small-molecule drug conjugates (SMDCs) have certain advantages over antibody drug conjugates in terms of non-immunogenicity, reproducibility, and permeability. In this review, prostate-specific membrane antigen-targeted SMDCs for treating prostate cancer are summarized.
Prostate cancer is the second most frequent cancer among men and the fifth leading cause of cancer-related death, with an estimated almost 1.4 million new cases and 375000 deaths worldwide in 2020.1) It is the most frequently diagnosed cancer among men in over half of countries, including in Northern and Western Europe, the Caribbean, Australia/New Zealand, Northern America, Southern Africa, and Japan.1,2) As long as the cancer is limited to the prostate, it can be managed by surgery, radical prostatectomy, and radiotherapy.3–5) Once metastasized, androgen deprivation therapy is effective for the metastatic disease. However, the therapy can only delay progression and the vast majority of patients with metastatic cancer eventually develop metastatic castration-resistant prostate cancer (mCRPC), which has a high mortality rate.6) Conventional chemotherapeutic agents, such as docetaxel, have been used for mCRPC patients.7) However, cytotoxic agents cause multidrug resistance and undesirable systematic side effects, and there is a pressing need for more specific and safe chemotherapeutic agents.8) The three classical diagnostic methods for prostate cancer are a prostate-specific antigen (PSA) test, digital rectal examination, and prostate biopsy. However, these methods have major disadvantages: the PSA test frequently gives false positives by detecting benign prostatic hyperplasia; the accuracy of digital rectal examination depends on the experience of the doctor; and biopsies are invasive and can result in bacterial infection and prostatitis.9–12)
1.1. Prostate-Specific Membrane Antigen (PSMA)-Targeted Positron Emission Tomography (PET) Imaging and Radiopharmaceutical Therapy (RPT)PSMA-targeted PET imaging for prostate cancer may overcome the shortcomings of classical diagnostic methods. PSMA is overexpressed on prostate cancer cells compared with normal prostate cells and a high level of PSMA expression was observed in endothelial cells in other solid tumor neovasculature.13,14) The expression level of PSMA on the cancer cells increases with tumor aggressiveness.15) Therefore, PSMA is an attractive target for prostate cancer imaging and treatment. Small-molecule PSMA ligands with radioisotopes, 68Ga-PSMA-11 (1) and 18F-DFCPyL (2), were approved by the United States Food and Drug Administration (FDA) as PET imaging agents for the diagnosis of prostate cancer16,17) (Fig. 1). The other promising strategy using PSMA ligands is RPT. 177Lu-PSMA-617 (3), which contains a therapeutic radionuclide that emits beta particles, was approved by FDA for the treatment of metastatic prostate cancer.18) Although RPT showed promising results on patients who are refractory to conventional therapies, improving blood PSA value with minimal side effects, patients tended to be retreated after a median time of 6 months.19–21) Furthermore, the same PSMA ligand containing an alpha-particle emitter, 225Ac-PSMA-617 (4), showed remarkable antitumor activity in a patient who experienced tumor progression during beta-emitting treatment. However, alpha-particle emitters caused side effects that were more severe, such as partially irreversible xerostomia, with a loss of QOL in some patients.22,23) RPT also requires specialized facilities to generate therapeutic radionuclides and these drawbacks have prompted the development of PSMA-targeted drug conjugates.
Targeted therapy can deliver cytotoxic agents into cancer cells resulting in fewer systemic toxic effects compared with conventional indiscriminate chemotherapy. Another benefit of targeting PSMA is that PSMA ligands can enter PSMA-expressing cells by endocytosis after binding, thereby releasing the cytotoxic agents inside the cells.24) Several cleavable linkers have been examined in PSMA-targeted drug conjugates for efficient drug release. A similar strategy has been tested with PSMA antibody drug conjugates (ADCs), and clinical trials showed some promising results, although neutropenia and peripheral neuropathy were also observed.25) The toxicity may have been caused by the slow blood clearance rate of the antibody, which could lead to premature drug release.26) In addition, ADCs can be immunogenic, are expensive to produce due to the need for clinical-grade manufacturing, and often have difficulty penetrating deeply into solid tumors because ADCs are large.27–29) Therefore, PSMA ligand small-molecule drug conjugates (SMDCs) have the advantages of a shorter residence time, penetrative ability, non-immunogenicity, and reproducible synthesis. This paper introduces the SMDCs that have been reported to date.
SMDCs reported for prostate cancer have a cleavable linker between the PSMA ligand and chemotherapeutic agent. The linkers are categorized by the cleavage conditions for releasing the chemotherapeutic agents into the cancer cells.
2.1. Cleavage by AmidasesKozikowski and colleagues, who developed urea-based PSMA ligands, reported the first SMDC 5, which consisted of urea-based PSMA ligands derived from glutamate and p-aminophenylalanine linked to doxorubicin by glutaric acid30) (Fig. 2). The cleavage mode of the linker of the SMDC was expected to be hydrolysis of the amide bond of the doxorubicin pyranose ring by an amidase, such as PSA. 5 blocked the binding of 3H-ZJ24, a previously reported PSMA ligand, with an IC50 value of 40.8 ± 1.6 nM. However, 5 showed almost no cytotoxicity against both PSMA-positive C4-2 and PSMA-negative PC3 prostate cancer cells. Doxorubicin itself exhibited IC50 values of 32 nM against the C4-2 cells and 223 nM against the PC3 cells. 5 may not have undergone the appropriate enzymatic processing required to release doxorubicin.
Ivanenkov et al. also used doxorubicin as the anticancer moiety for their SMDCs but with a hydrazone linker that was more susceptible to hydrolysis in the natural environment than the pyranose amide bond.31) They synthesized several SMDCs with linkers of different lengths and reported that 6, which contained two phenylalanine fragments, was localized mainly in the nuclei of LNCaP cells, thereby providing high selectivity in contrast to doxorubicin (Fig. 3). They also synthesized 7, in which the hydrazone linker was replaced with an amide linker, similar to 5.30) 7 showed no cytotoxic effects on both cell lines, consistent with the results for 5. The authors suggested that the lack of activity was also related to the high stability of the amide bond, which was insensitive to amidase-driven cleavage. 6 exhibited higher cytotoxicity against PSMA-positive LNCaP cells (50% cytotoxicity concentration (CC50) = 95 nM), close to that observed for doxorubicin (CC50 = 93 nM), than against PSMA-negative PC3 cells (CC50 = 926 nM).
There are several examples of SMDCs with linkers cleaved by intracellular glutathione (GSH). Low and colleagues reported prodrugs in which various anticancer drugs were linked to PSMA ligands via disulfide bonds.32,33) After the disulfide was cleaved by intracellular GSH, the disulfide linker generated ethylene sulfide, carbon dioxide (major route), and 1,2-oxythiolane-2-one (minor route), and released the anticancer drug (Fig. 4). Their SMDCs were linked to various anticancer drugs: vinca alkaloids, which inhibit microtubules (8, 9); topoisomerase I inhibitors (10, 11); and protein synthesis inhibitors (12, 13) (Fig. 5). Except for 12, these SMDCs exhibited high cytotoxicity against PSMA-positive LNCaP cells. However, in 12, the authors speculated that the hydrolase in the endosome might open the epoxy ring or lactone ring of the anticancer agent, eliminating the cytotoxicity. 11 was tested for its tumor-suppressive effect and toxicity in prostate cancer-implanted mice, and the tumor-suppressive effect was weaker in the SMDC-treated group than in the free drug-treated group.33) However, four out of five mice in the free drug-treated group died within the test period, whereas none of the mice in the SMDC-treated group died, even at high concentrations. This suggests that SMDCs are safer, less toxic to healthy cells, and more selective to cancer cells than single administration of anticancer drugs. Kumar et al. reported 14, a PSMA ligand conjugated with both a PET-imageable chelating structure and a cytotoxic drug.34) NOTA, which is commonly used in PSMA-PET and can chelate 68Ga, was selected as the chelate structure, and DM1, a microtubule inhibitor, was used as the cytotoxic drug and was bound to the PSMA ligand via a disulfide linker. The cytotoxicity of 14 decreased in the order LNCaP > C4-2 > PC3, which correlated with the PSMA expression level of each cell. In addition, in a PET imaging test, PSMA-positive PC3-PIP cells were selectively visualized, and 14 did not accumulate in the liver, which metabolizes DM1. Lv et al. reported 15, a PSMA ligand joined via a disulfide linker to paclitaxel, which is a first-line chemotherapeutic agent for prostate cancer.35) After reductive cleavage by GSH in vivo, the disulfide at the linker site in 15 can release paclitaxel by hydrolytic ester cleavage, rather than generating ethylene sulfide and carbon dioxide as in previous disulfide linkers. In addition, 16, with no disulfide bond, was synthesized for comparison. In a cytotoxicity test using PSMA-positive 22RV1 cells, 15 (IC50 = 121.1 nM) was less toxic compared with paclitaxel (IC50 = 14.25 nM). The authors speculated that this may be due to the slow release of paclitaxel from 15. In addition, 16, with no disulfide bond, was even less toxic, presumably due to its extremely slow rate of hydrolysis. In contrast, the mice treated with 15 displayed a significant decrease in tumor volume, showing no significant difference compared with paclitaxel. Leamon et al. reported 17, in which tubulysin B hydrazide, a microtubule polymerization inhibitor, was conjugated with a PSMA ligand.36)18–20 containing D-form amino acids, showed no PSMA binding affinity, indicating that L-form amino acids were essential. Furthermore, the authors reported that the length of the glutamate moiety was optimal because the PSMA binding ability decreased when the glutamate moiety was converted to 2-aminoadipic acid or aspartic acid. Compared with docetaxel, which is often used in chemotherapy for advanced prostate cancer, 17 showed a stronger antitumor suppressive effect without causing side effects, such as weight loss, in in vivo tests. 17 is now in clinical trial phase I.
Brentuximab vedotin, an ADC approved by the FDA for treating Hodgkin’s lymphoma, contains the valine–citrulline p-aminobenzylalcohol (PAB) linker, which is enzymatically degraded by cathepsin B, a lysosomal protease commonly expressed in malignant cells, as the linker between the antibody and the drug.37) Boinapally et al. designed and synthesized 21, in which the PSMA ligand and the microtubule inhibitor monomethyl auristatin E (MMAE) were linked with a valine–citrulline PAB linker, and 22, which contained a non-cleavable linker38) (Fig. 6). 21 exhibited cytotoxicity against PSMA-positive PC3 PIP cells (IC50 = 3.9 nM) and PSMA-negative PC3 flu cells (IC50 = 151.1 nM). The weak cytotoxicity to PSMA-negative cells was due to the release of some MMAE. 22 had low cytotoxicity (IC50 = 4.8–5.8 µM) with or without PSMA expression, suggesting that release of MMAE from the SMDC was necessary to cause cytotoxicity. In an in vivo study in prostate cancer cell-implanted mice, the median survival time was significantly prolonged in the group of mice receiving 80 µg/kg of 21 compared with the non-administration group (treated group: 56 d, non-treated group: 47 d). Regarding toxicity, MMAE administration alone caused weight loss requiring euthanasia, but steady weight gain was confirmed in the group treated with 21, and no toxicity was observed. Machulkin et al. synthesized 23, consisting of an independently optimized PSMA ligand with a similar valine–citrulline PAB linker and MMAE, and evaluated its cytotoxicity, changes in intracellular reactive oxygen levels, and metabolic stability.39) 23 showed strong toxicity in both PSMA-positive 22Rv1 cells (CC50 = 29 ± 2 nM) and PSMA-negative PC3 cells (CC50 = 27 ± 3 nM). The active oxygen concentration was increased in both types of cells compared with the control. The authors investigated the activity of the main cytochrome P-450 isoforms, namely 1A2, 2C8, 2C9, 2C19, 2D6, and 3A4, in the presence of 23, and 23 predominantly inhibited only cytochrome 3A4 (IC50 = 11.2 µM). The substrate in the presence of mouse and human liver microsome had a half-life of 27.4 min. However, the metabolic enzymes could not be identified. Compared with docetaxel, a standard chemotherapeutic agent for prostate cancer, 23 had a higher therapeutic index. Alhamadsheh and colleagues. designed and synthesized 24 with a site that binds to the serum protein transthyretin to overcome the short blood retention time of SMDCs.40) As expected, 24 had a blood retention time 6.6-fold longer than the ligands without transthyretin. In addition, 24 was more effective than non-transthyretin-introduced ligands in cancer cell growth inhibition studies in mice engrafted with prostate cancer. This may be because the transthyretin-binding site prolongs blood retention time and exposure time. Subsequently, Alhamadsheh and colleagues. also reported 25, which contained a glutamate-valine–citrulline tripeptide linker,41) developed by Anami.42) This linker is stable in mouse serum, whereas the valine–citrulline dipeptide linker is labile in mouse serum (hydrolyzed by carboxyesterase 1c) causing unwanted toxicity in the mouse model. Even though the valine–citrulline dipeptide linker is stable in human serum, using the glutamate-valine–citrulline tripeptide linker that is stable in mouse serum is important for predicting the outcome of human clinical trials more accurately. Stability studies showed that neither 24 nor 25 were degraded in human serum after 24 h, whereas in mouse serum, only 20.9% of 24 remained and 71% of 25 remained. As a result of an in vivo toxicity test using CD-1 mice, 24 reduced the body weight of mice at 1200 nmol/kg, whereas no weight loss was observed with 25 at the same dose, indicating that the glutamate-valine–citrulline tripeptide linker is stable in mouse serum. Wang et al. synthesized 26 and 27, which were PSMA ligands conjugated with Cy5.5, which is a fluorescent dye, and MMAE for fluorescence imaging of the SMDCs.43) 26 contained a valine–citrulline linker cleaved by cathepsin B, whereas 27 contained a non-cleavable linker. Fluorescence imaging tests of both 26 and 27 confirmed localization in intracellular lysosomes in PSMA-positive PC3 PIP cells, and no fluorescence was observed in PSMA-negative PC3 flu cells. Subsequently, immunofluorescent staining tests for α-tubulin were performed to confirm whether MMAE worked, and disruption of the microtubule network was observed only when 26, which had a cleavable linker, was administered to PSMA-positive PC3 PIP cells. Therefore, selective toxicity of SMDCs required cleavable linkers. Luo et al. synthesized 28 that introduced sites capable of clustering because nanoclusters are more likely to accumulate in cancer cells, which have imperfect vascular walls, than in normal cells, due to the enhanced permeability and retention effect.44) The tyrosine and thiol-modified lysine residues of 28 allowed nanocluster formation with Au3+. To confirm that MMAE was released from the nanoclusters, 28 was exposed to cathepsin B, and 95% MMAE release was confirmed by HPLC analysis after 48 h. In addition, the radiation enhancement was investigated because gold absorbs radiation energy and damages DNA through the photoelectric effect and Auger effect, and MMAE increases sensitivity to radiation by inhibiting the G2-M phase of the cell cycle. From the administration results of 28 and its comparative compound group, significant radiosensitization by gold and MMAE was observed in PSMA-positive PC3 PIP cells. PSMA-negative PC3 flu cells did not show as much radiosensitization as PC3 PIP cells, indicating that the combination with the SMDC nanoclusters and radiation may be beneficial for prostate cancer treatment.
This review summarized targeted SMDC therapeutics, in which cytotoxic drugs are connected to small-molecule PSMA ligands with various linkers. PSMA is overexpressed in prostate cancer cells, mCRPC, and the neovasculature of solid tumors, making it an ideal target for selective delivery of anticancer drugs to cancer cells. Even though ADCs targeting PSMA have been developed, the ADCs have drawbacks, such as immunogenicity, low permeability in solid tumors, excessive blood retention time, and low production reproducibility. Small-molecule PSMA ligands are expected to overcome these shortcomings. Linkers for connecting small-molecule PSMA ligands to anticancer drugs were classified according to their cleavage modes (amidase, acidic pH, GSH, and cathepsin), and cytotoxic agents included microtubule inhibitors, protein synthesis inhibitors, and topoisomerase inhibitors. The SMDCs were weakly toxic to PSMA-negative prostate cancer cells and more toxic to PSMA-positive prostate cancer cells. When the toxic drug was administered alone, the mice lost weight to the point that they required euthanasia, but when the SMDCs were administered, the mice tended to gain weight, suggesting less systemic toxicity. Future research on SMDCs is expected to lead to the development of PSMA-targeted therapeutic agents for prostate cancers with fewer side effects that can replace systemically toxic chemotherapeutic agents.
This work was supported in part by JSPS KAKENHI Grant Number 21K15227.
The author declares no conflict of interest.
