Preclinical Evaluation of a New Series of Albumin-Binding ¹⁷⁷Lu-Labeled PSMA-Based Low-Molecular-Weight Radiotherapeutics

Abstract

Prostate-specific membrane antigen (PSMA)-based low-molecular-weight agents using beta(β)-particle-emitting radiopharmaceuticals is a new treatment paradigm for patients with metastatic castration-resistant prostate cancer. Although results have been encouraging, there is a need to improve the tumor residence time of current PSMA-based radiotherapeutics. Albumin-binding moieties have been used strategically to enhance the tumor uptake and retention of existing PSMA-based investigational agents. Previously, we developed a series of PSMA-based, β-particle-emitting, low-molecular-weight compounds. From this series, ¹⁷⁷Lu-L1 was selected as the lead agent because of its reduced off-target radiotoxicity in preclinical studies. The ligand L1 contains a PSMA-targeting Lys-Glu urea moiety with an N-bromobenzyl substituent in the ε-amino group of Lys. Here, we structurally modified ¹⁷⁷Lu-L1 to improve tumor targeting using two known albumin-binding moieties, 4-(p-iodophenyl) butyric acid moiety (IPBA) and ibuprofen (IBU), and evaluated the effects of linker length and composition. Six structurally related PSMA-targeting ligands (Alb-L1–Alb-L6) were synthesized based on the structure of ¹⁷⁷Lu-L1. The ligands were assessed for in vitro binding affinity and were radiolabeled with ¹⁷⁷Lu following standard protocols. All ¹⁷⁷Lu-labeled analogs were studied in cell uptake and selected cell efficacy studies. In vivo pharmacokinetics were investigated by conducting tissue biodistribution studies for ¹⁷⁷Lu-Alb-L2–¹⁷⁷Lu-Alb-L6 (2 h, 24 h, 72 h, and 192 h) in male NSG mice bearing human PSMA+ PC3 PIP and PSMA− PC3 flu xenografts. Preliminary therapeutic ratios of the agents were estimated from the area under the curve (AUC_0-192h) of the tumors, blood, and kidney uptake values. Compounds were obtained in >98% radiochemical yields and >99% purity. PSMA inhibition constants (K_is) of the ligands were in the ≤10 nM range. The long-linker-based agents, ¹⁷⁷Lu-Alb-L4 and ¹⁷⁷Lu-Alb-L5, displayed significantly higher tumor uptake and retention (p < 0.001) than the short-linker-bearing ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L3 and a long polyethylene glycol (PEG) linker-bearing agent, ¹⁷⁷Lu-Alb-L6. The area under the curve (AUC_0-192h) of the PSMA+ PC3 PIP tumor uptake of ¹⁷⁷Lu-Alb-L4 and ¹⁷⁷Lu-Alb-L5 were >4-fold higher than ¹⁷⁷Lu-Alb-L2, ¹⁷⁷Lu-Alb-L3, and ¹⁷⁷Lu-Alb-L6, respectively. Also, the PSMA+ PIP tumor uptake (AUC_0-192h) of ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L3 was ~1.5-fold higher than ¹⁷⁷Lu-Alb-L6. However, the lowest blood AUC_0-192h and kidney AUC_0-192h were associated with ¹⁷⁷Lu-Alb-L6 from the series. Consequently, ¹⁷⁷Lu-Alb-L6 displayed the highest ratios of AUC(tumor)-to-AUC(blood) and AUC(tumor)-to-AUC(kidney) values from the series. Among the other agents, ¹⁷⁷Lu-Alb-L4 demonstrated a nearly similar ratio of AUC(tumor)-to-AUC(blood) as ¹⁷⁷Lu-Alb-L6. The tumor-to-blood ratio was the dose-limiting therapeutic ratio for all of the compounds. Conclusions: ¹⁷⁷Lu-Alb-L4 and ¹⁷⁷Lu-Alb-L6 showed high tumor uptake in PSMA+ tumors and tumor-to-blood ratios. The data suggest that linker length and composition can be modulated to generate an optimized therapeutic agent.

1. Introduction

Prostate cancer is the most commonly diagnosed non-cutaneous cancer among men in the United States and globally [1]. Death from prostate cancer occurs mainly in patients with aggressive, androgen-insensitive, metastatic disease [2,3]. Prostate-specific membrane antigen (PSMA) is a tumor-associated antigen overexpressed in prostate adenocarcinoma cells, regardless of androgen status, in the neovasculature of solid tumors, and has a low expression in benign and extra-prostatic tissues [4,5]. Studies have shown that radiopharmaceutical therapy targeting PSMA with β-particle-emitting ¹⁷⁷Lu (half-life 6.7 d) is a life-prolonging treatment option for patients with metastatic castration-resistant prostate cancer [6]. Recently, a low-molecular-weight agent, ¹⁷⁷Lu-PSMA-617, demonstrated a low toxicity profile compared to the standard of care and received regulatory approval [6,7]. Although highly promising, the median overall survival among randomized patients was 15.3 months in the treatment vs. 11.3 months in the control group, suggesting that further improvements are needed [6,8]. One way to achieve this is through the pharmacokinetic optimization of agents similar to ¹⁷⁷Lu-PSMA-617.

Several approaches involving the structural manipulation of PSMA-targeting ureas are currently under intense investigation to improve the efficacy of ¹⁷⁷Lu-PSMA-617 [8]. Examples of such optimization include the development of bivalent or high-affinity agents and the incorporation of pharmacophores to enhance their blood circulation [9,10,11,12,13,14]. Along this direction, incorporating an albumin-binding moiety into PSMA-based, low-molecular-weight agents has gained significant attention [11,13,15,16,17,18,19,20,21]. Well-characterized, small-molecule moieties (e.g., 4-(para-iodophenyl) butyric acid (IPBA), truncated Evans blue, ibuprofen (IBU), and fatty acids), referred to as “blood half-life extenders”, non-covalently bind to serum albumin and have been utilized extensively [22]. Several agents of this class have entered clinical trials, as shown in Figure 1.

The majority of these agents are derived from the structure of PSMA-617 [20,23,24]. Preclinical and clinical data have revealed increased tumor targeting of such agents. However, increased radiation dose delivery to the salivary glands and kidneys was also observed partly due to PSMA expression in these tissues [13,18,23]. In addition, increased off-target radiotoxicity, specifically hematologic toxicity, has been demonstrated in clinical studies [13,23]. Accordingly, current preclinical efforts are focused on optimizing the PSMA-targeting moiety [20,25], linker composition [12,26], and albumin-binding motif [20,26] to maximize the tumor-to-kidney radiation dose ratios. Improving tumor targeting and weakening albumin binding to reduce blood, kidney, or salivary gland uptake is the primary rationale for newly developed agents [18,25,26].

We recently developed a new series of PSMA-based agents with reduced off-target toxicity, and one of which was translated into the clinic [27]. The lead agent, L1, was also investigated using alpha (α)-particle-emitting radionuclides, including ²²⁵Ac, ²¹³Bi, ²¹²Pb, and ²¹¹At [28,29,30]. In the current study, we evaluated six albumin-binding agents by conjugating IBU and IPBA using two bifunctional linkers. Our primary goal was to generate an agent with enhanced tumor uptake and retention compared to ¹⁷⁷Lu-L1. We investigated two albumin-binding moieties and 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid mono-amide (DOTA-monoamide) as a chelating unit to generate six ligands bearing either a short (A) (¹⁷⁷Lu-Alb-L1–¹⁷⁷Lu-Alb-L3) or long linker (B) (¹⁷⁷Lu-Alb-L4–Alb-L6) (Figure 2). Our data revealed that the linker length and composition might be critical in optimizing PSMA− and albumin-based agents for clinical use.

2. Results

2.1. Synthesis and Binding Affinities of Ligands Alb-L1 to Alb-L6

We designed and synthesized six ligands using (i) PSMA-targeting Glu-Lys (N-p-bromo-benzyl) urea moiety, (ii) short linker (Structure A) or long linker (Structure B), and (iii) albumin-binding moieties IPBA and IBU, as shown in Figure 2. Standard solution-phase peptide conjugation chemistry was employed to synthesize these new compounds, as detailed in the Supplementary Materials. After evaluating ¹⁷⁷Lu-Alb-L1 to ¹⁷⁷Lu-Alb-L5, we studied ¹⁷⁷Lu-Alb-L6, containing a long polyethylene glycol (PEG) linker, to reduce the blood uptake of these agents.

2.2. In Vitro Characterization

Binding Affinity, Cell Uptake, and Internalization

The new ligands displayed high binding affinity to PSMA with K_i values ranging from 0.12 nM to 11.24 nM, as listed in Figure 2. Ligands, Alb-L4, and Alb-L5 from Structure B construct and Alb-L3 displayed significantly higher affinity (p < 0.05) than the rest of the ligands. The albumin-binding properties of the agents revealed a substantially higher affinity for the long-linker-based agents, ¹⁷⁷Lu-Alb-L4 and ¹⁷⁷Lu-Alb-L5, compared to ¹⁷⁷Lu-Alb-L6 and the short-linker-based ¹⁷⁷Lu-Alb-L2, as well as ¹⁷⁷Lu-Alb-L3 (Supplementary Table S3). The cell uptake and internalized fraction of the ¹⁷⁷Lu-labeled agents are shown in Figure 3. Except for ¹⁷⁷Lu-Alb-L1 (without a p-bromo-benzyl group on the PSMA-targeting unit), all of the agents demonstrated high cell uptake in the PSMA+ PC3 PIP cells, i.e., in the range of 28% to 75% of the incubated dose at 2 h, and this was slightly increased up to 30% to 77% at 24 h incubation. From the series, ¹⁷⁷Lu-Alb-L4 and ¹⁷⁷Lu-Alb-L5 displayed high uptake, 60% and 75%, respectively, at 2 h, and 63% and 77% at 24 h of incubation in PSMA+ PC3 PIP cells. The percentage of internalization for the agents in PSMA+ PC3 PIP cells was in the range of 32–47% at 2 h. The uptake of the radioligands in the PSMA− PC3 flu cells was significantly lower (~100-fold) than that of the PSMA+ PC3 PIP cells, revealing the PSMA-specific binding of the agents. A PSMA expression blocking study was performed using an excess of ZJ43 [31], a known PSMA inhibitor, which caused a significantly low uptake of the radiolabeled agents in PSMA+ PC3 PIP cells (Figure 3). The data further confirmed the PSMA-mediated uptake of ¹⁷⁷Lu-Alb-L1–¹⁷⁷Lu-Alb-L6 in PSMA+ PC3 PIP cells.

2.3. Clonogenic Survival Assay

Clonogenic assays were performed to assess the cellular efficacy of ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L5 and were compared with ¹⁷⁷Lu-L1, as shown in Figure 4. The data revealed a decrease in the survival of PSMA+ PC3 PIP cells with increasing concentrations of radioactivity, with complete cell killing reached at a concentration of 370 kBq/mL for ¹⁷⁷Lu-L1, 185 kBq/mL for ¹⁷⁷Lu-Alb-L2, and 37 kBq/mL for ¹⁷⁷Lu-Alb-L5, respectively. The activity to reduce cell survival to 37% (A₀) was 18.5 kBq/mL for ¹⁷⁷Lu-L1, ~20 kBq/mL for ¹⁷⁷Lu-Alb-L2, and ~10 kBq/mL for ¹⁷⁷Lu-Alb-L5, respectively.

2.4. Biodistribution

Tissue biodistribution studies of ¹⁷⁷Lu-Alb-L2–¹⁷⁷Lu-Alb-L6 were performed at 2 h, 24 h, 48 h, and 192 h after injection to evaluate the tumor targeting and clearance of the agents from blood and normal tissues. The biodistribution of ¹⁷⁷Lu-L1 was conducted at 24 h post-injection for the agents for direct comparison. The biodistribution studies of ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L5, as well as ¹⁷⁷Lu-L1 (24 h), were performed in one experiment to keep the experimental variations minimal (for example, tumor size, animals’ age, and specific activity of ¹⁷⁷Lu). Similarly, biodistribution studies of ¹⁷⁷Lu-Alb-L3 and ¹⁷⁷Lu-Alb-L4, as well as ¹⁷⁷Lu-L1 (24 h), were conducted head-to-head in one experiment. The biodistribution study of ¹⁷⁷Lu-Alb-L6 was performed separately using tumors of a similar size.

The results of the new agents and ¹⁷⁷Lu-L1 are presented in Figure 5 and Supplementary Tables S4–S10. Biodistribution data of ¹⁷⁷Lu-L1 (3–192 h) were acquired from our previously reported studies [27]. A fast tumor accumulation was observed for ¹⁷⁷Lu-Alb-L2, which reached 17.45 ± 6.51%ID/g at 2 h post-injection and displayed higher uptake, 26.41 ± 6.73%ID/g at 24 h and 22.00 ± 4.71%ID/g at 48 h, respectively. At 192 h post-injection, tumor uptake was low, 3.39 ± 1.03%ID/g. ¹⁷⁷Lu-Alb-L3 displayed similar pharmacokinetics as ¹⁷⁷Lu-Alb-L2, 34.05 ± 1.98%ID/g at 2 h, 30.55 ± 7.44%ID/g at 24 h, 19.61 ± 4.46%ID/g at 48 h, and 5.94 ± 1.38%ID/g at 192 h.

The long-linker-based agents (Structure B), ¹⁷⁷Lu-Alb-L4 and ¹⁷⁷Lu-Alb-L5, displayed significantly higher tumor uptake and retention than ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L3. The tumor uptake of ¹⁷⁷Lu-Alb-L4 was 40.89 ± 4.73%ID/g at 2 h, 88.04 ± 16.51%ID/g at 24 h, 87.74 ± 14.09%ID/g at 48 h, and 42.22 ± 14.05%ID/g at 192 h, respectively. For ¹⁷⁷Lu-Alb-L5, it was 28.78 ± 7.25%ID/g at 2 h, 111.47 ± 13.85%ID/g at 24 h, 127.44 ± 22.85%ID/g at 48 h, and 70.96 ± 2.34%ID/g at 192 h, respectively. ¹⁷⁷Lu-Alb-L6 displayed moderate uptake and fast clearance: 38.73 ± 1.26%ID/g at 2 h, 13.72 ± 3.55%ID/g at 24 h, 9.62 ± 0.96%ID/g at 48 h, and 2.22 ± 0.37%ID/g at 192 h. The tumor uptake values of ¹⁷⁷Lu-L1 at 24 h were 12.86 ± 0.98–14.47 ± 1.32%ID/g. The tumor uptake and clearance of ¹⁷⁷Lu-L1 were in the same range as ¹⁷⁷Lu-Alb-L6 (30.81 ± 2.86%ID/g, 15.67 ± 6.25%ID/g, 11.94 ± 3.83%ID/g, 9.33 ± 3.18%ID/g, and 4.21 ± 1.45%ID/g at 192 h). The uptake in the PSMA− PC3 flu tumors was low for the agents. The highest uptake in the PSMA− PC3 flu tumor was associated with ¹⁷⁷Lu-Alb-L5, 3.87 ± 0.70%ID/g at 2 h, 4.55 ± 0.76%ID/g at 24 h, 3.95 ± 0.50%ID/g at 48 h, and 1.87 ± 0.23%ID/g at 192 h post-injection, respectively.

The blood uptake levels were high at 2 h for ¹⁷⁷Lu-Alb-L2, with 19.59 ± 6.06%ID/g; ¹⁷⁷Lu-Alb-L3, with 10.29 ± 0.41%ID/g; ¹⁷⁷Lu-Alb-L4, with 15.48 ± 7.12%ID/g; ¹⁷⁷Lu-Alb-L5, with 22.36 ± 4.72%ID/g; and ¹⁷⁷Lu-Alb-L6, with 2.97 ± 0.45%ID/g, respectively. ¹⁷⁷Lu-Alb-L6 displayed the fastest clearance from the circulation owing to the long PEG linker, and the uptake was <0.03 ± 0.01%ID/g at 24 h. However, the blood uptake values of ¹⁷⁷Lu-L1 (0.18 ± 0.16 at 3 h, 0.01 ± 0.01 at 24 h, 0.03 ± 0.03 at 48 h, 0.02 ±0.02 at 72 h, and 0.00 ±0.00 at 192 h) were significantly lower than those of ¹⁷⁷Lu-Alb-L6 and the new albumin-binding agents at all time points. ¹⁷⁷Lu-Alb-L5 displayed the slowest clearance from the blood at 192 h post-injection, 2.77 ± 0.19%ID/g, followed by ¹⁷⁷Lu-Alb-L4 (2.19 ± 0.46%ID/g), ¹⁷⁷Lu-Alb-L3 (0.24 ± 0.12%ID/g), and ¹⁷⁷Lu-Alb-L2 (≤0.01%ID/g).

Kidney uptake was relatively high for ¹⁷⁷Lu-Alb-L5 (33.45 ± 6.41%ID/g at 2 h, 70.23 ± 16.64 at 24%ID/g, 48.17 ± 15.62%ID/g at 48 h, and 5.06 ± 0.99%ID/g at 192 h) and ¹⁷⁷Lu-PSMA-Alb-L4 (64.44 ± 8.75%ID/g at 2 h, 48.81 ± 19.86%ID/g at 24 h, 23.46 ± 7.75%ID/g at 48 h, and 1.34 ± 0.82 at 192 h). ¹⁷⁷Lu-Alb-L3 displayed the highest kidney uptake at 2 h, 73.28 ± 22.85%ID/g, and the fastest clearance, 2.77 ± 1.56%ID/g at 24 h. ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L6 displayed low kidney uptake of ~17.86%ID/g at 2 h. ¹⁷⁷Lu-Alb-L6 displayed fast renal clearance resulting in <1%ID/g after 24 h; in contrast, ¹⁷⁷Lu-Alb-L2 showed a slow and steady wash-out from the kidneys from ~7.44 ± 2.64% IA/g at 24 h post-injection to 3.74 ± 1.05%ID/g at 48 h post-injection. The kidney uptake and clearance values of ¹⁷⁷Lu-L1 were in the range of ¹⁷⁷Lu-Alb-L6 (5.16 ± 2.38%ID/g at 3 h, 0.28 ± 0.18%ID/g at 24 h, 0.22 ± 0.10%ID/g at 72 h, and 0.01 ± 0.00%ID/g at 192 h).

The activity levels in the salivary and lacrimal glands of the agents were comparable to the blood levels and were relatively high for ¹⁷⁷Lu-Alb-L5 (salivary glands, 5.83 ± 1.06%ID/g at 2 h, 4.35 ± 0.66%ID/g at 24 h, 3.20 ± 0.24 h %ID/g at 48 h, and 1.18 ± 0.09%ID/g at 192 h) and ¹⁷⁷Lu-Alb-L4 (salivary glands, 5.40 ± 0.33%ID/g at 2 h, 0.86 ± 0.21%ID/g at 24 h, 0.62 ± 0.07%D/g at 48 h, and 0.18 ± 0.06%ID/g at 192 h). Similarly, the uptake of the other normal tissues, including the lung and liver, were high for ¹⁷⁷Lu-Alb-L5 and decreased continuously over time. For the agents ¹⁷⁷Lu-Alb-L2, ¹⁷⁷Lu-Alb-L3, ¹⁷⁷Lu-Alb-L4, and ¹⁷⁷Lu-Alb-L6, normal tissue clearance was below the blood levels and mostly ≤2%ID/g after 24 h.

The selected tumor-to-normal tissue ratios of the agents are shown in Figure 5B. A high blood pool was observed for all of the agents at 2 h after injection, resulting in a low tumor-to-blood ratio (1.0 ± 0.2), and it increased over time. The tumor-to-blood ratios were ~4–8 during 24 h and 48 h post-injection for ¹⁷⁷Lu-Alb-L2 and ~200–220 after 192 h injection. In contrast, ¹⁷⁷Lu-Alb-L3 displayed fast blood clearance, ~200–250 at 24–48 h, and remained in that range until 192 h. For ¹⁷⁷Lu-Alb-L4, the ratios were in the range of ~50 during 24–48 h and ~100–150 at 192 h. The slowest clearance was observed for ¹⁷⁷Lu-Alb-L5; the tumor-to-blood ratios were ~10–20 during 24–48 h and ~30 during 192 h. The tumor-to-blood values were >300 for ¹⁷⁷Lu-L1 at all time points.

The tumor-to-kidney ratio was <1 for ¹⁷⁷Lu-Alb-L2–¹⁷⁷Lu-Alb-L5 at 2 h and remained at ≤10 during 24–48 h and increased significantly, ~20–50 at 192 h after injection. The lowest tumor-to-kidney ratio was associated with ¹⁷⁷Lu-Alb-L5. In contrast, the tumor-to-kidney values were >4 for ¹⁷⁷Lu-L1 at 24 h and remained >50 at all time points. A similar time course was observed for the tumor-to-liver, tumor-to-salivary, and tumor-to-lacrimal glands, as well as the tumor-to-bone ratios.

The area under the curve (AUC_0-192h) values and AUC ratios of tumor-to-blood and tumor-to-kidney over 192 h of the new agents and ¹⁷⁷Lu-L1 are listed in Figure 6A,B. Tumor AUC_0-192h of ¹⁷⁷Lu-Alb-L5 was significantly higher than ¹⁷⁷Lu-Alb-L4 (p < 0.002). In addition, tumor AUC of ¹⁷⁷Lu-Alb-L5 and Alb-L4 were ~4-fold higher compared to ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L3 and ~6-fold higher than ¹⁷⁷Lu-Alb-L6 and ¹⁷⁷Lu-L1. The tumor-to-blood AUC_0-192h ratio of ¹⁷⁷Lu-Alb-L4 was ~2-fold higher than ¹⁷⁷Lu-Alb-L5, while the tumor-to-kidney AUC_0-192h ratios were in the same range. ¹⁷⁷Lu-Alb-L6 and ¹⁷⁷Lu-L1 displayed nearly similar tumor AUC_0-192h and kidney AUC_0-192h, and consequently both agents had a similar tumor-to-kidney AUC_0-192h (~8). However, the tumor-to-blood AUC_0-192h ratios of ¹⁷⁷Lu-L1 were significantly higher (~867) than the tumor-to-blood AUC_0-192h ratios of ¹⁷⁷Lu-Alb-L6 (~48). The tumor-to-kidney AUC_0-192h ratios of ¹⁷⁷Lu-Alb-L6 were ~2-fold higher than the other albumin-binding agents from the series. The tumor-to-blood AUC_0-192h of ¹⁷⁷Lu-Alb-L6 was comparable to ¹⁷⁷Lu-Alb-L4; it was ~1.14-fold higher.

3. Discussion

Here, we investigated a new series of albumin-binding ¹⁷⁷Lu-labeled agents based on the structure of ¹⁷⁷Lu-L1, our previously reported lead agent [27]. As reported by others and us, small-molecule organic moieties with low and reversible binding to serum albumin (Mol wt. 67 kDa) were utilized as a possibility to extend the circulating time of the PSMA-based agents, providing prolonged exposure to the tumors. Such modification significantly increased tumor uptake for the agents, relative to our original compound, ¹⁷⁷Lu-L1, through an increased blood half-life. These agents were designed and derived from our previously linker-based targeting platform [32,33]. Only two structural features were investigated, linker length and the attachment of two albumin-binding moieties, as shown in Figure 2. Nevertheless, several structure–activity relationships were derived from this small series of compounds, as described below.

We observed significantly higher (p < 0.05) binding affinity and PSMA+ cell binding both in vitro and in vivo for new agents, specifically ¹⁷⁷Lu-Alb-L4 and ¹⁷⁷Lu-Alb-L5 (Structure B), compared to ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L3 (Structure A). ¹⁷⁷Lu-Alb-L6 displayed low binding affinity, likely due to the long PEG linker, as we noted earlier with a similar construct [34]. Proof-of-concept cell efficacy data further confirmed the effect of enhanced cellular uptake and internalization of ¹⁷⁷Lu-Alb-L5 (Structure B) compared to ¹⁷⁷Lu-L1 or ¹⁷⁷Lu-Alb-L2 (Structure A). Enhanced cellular internalization is critical to radiation-induced DNA damage and PSMA-expressing cancer cell death.

The tumor uptake of ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L3 was significantly lower than our previously reported long-linker-based albumin-binding agent, ¹⁷⁷Lu-L14 (Figure 2) (tumor AUC_0-192h of 5740 ± 520%ID/g.h), although ≥1.5-fold higher than ¹⁷⁷Lu-L1 (tumor AUC_0-192h of 1734 ± 130%ID/g.h) (Figure 6A,B) [27]. The long-linker-based agents, ¹⁷⁷Lu-Alb-L4 (tumor AUC_0-192h of 12,857 ± 1469%ID/g.h) and ¹⁷⁷Lu-Alb-L5 (tumor AUC_0-192h of 18,842 ± 1693%ID/g.h) demonstrated a > 2-fold improvement in tumor uptake compared to ¹⁷⁷Lu-L14. The blood uptake values of ¹⁷⁷Lu-Alb-L4 (blood AUC_0-192h of 307 ± 80) were in the range of ¹⁷⁷Lu-L14 (blood AUC_0-192h of 314 ± 37), indicating a 2-fold improvement in the tumor-to-blood ratios of ¹⁷⁷Lu-Alb-L4 relative to ¹⁷⁷Lu-L14. Notably, the kidney uptake of ¹⁷⁷Lu-Alb-L4 (AUC_0-192h of 3895 ± 631) was 1.5-fold higher than ¹⁷⁷Lu-L14 (AUC_0-192h of 2550 ± 347). In contrast, the tumor and PSMA-expressing typical healthy tissue uptake values were significantly lower for ¹⁷⁷Lu-Alb-L6, most likely owing to low PSMA-binding and albumin-binding properties. Notably, the tumor and kidney AUCs of ¹⁷⁷Lu-Alb-L6 were in the range of ¹⁷⁷Lu-L1; however, the blood AUC was ~30-fold higher for ¹⁷⁷Lu-Alb-L6 than¹⁷⁷Lu-L1. The in vivo performance of ¹⁷⁷Lu-Alb-L6 was consistent with the earlier reports, notably, a fast kidney clearance of long PEG-linker-based ¹⁷⁷Lu-Alb analogs, published by Kelly et al. [11,12].

By comparing the efficacy (in terms of therapeutic index) of the different agents to the absorbed dose thresholds of radiotoxicity for kidneys (~28 Gy) and blood (~2 Gy), the results for the maximum absorbed dose to the tumors of the agents were determined. Figure 6C shows these results. The maximum tumor absorbed dose was estimated to be the lowest of the two values, obtained from the kidney maximum absorbed dose and the blood maximum absorbed dose. For all of the compounds, the dose-limiting organ was found to be the blood. ¹⁷⁷Lu-Alb-L6, and to a lesser extent, ¹⁷⁷Lu-Alb-L4, appear to be the more viable compounds with potential tumoricidal absorbed doses able to be delivered to the tumor.

Renal radiotoxicity has not proved to be a significant issue for ¹⁷⁷Lu-PSMA-targeted therapy, possibly due to optimal low linear energy transfer radiations of ¹⁷⁷Lu (β_max 0.5 MeV, 1.7 mm). Furthermore, a minimal expression of PSMA has been found in human kidneys. It has been reported that murine PSMA, with 91% similarity to the human PSMA sequence, is overexpressed in the proximal microtubules of the murine renal cortex [35]. The high kidney uptake observed in mice upon the administration of urea-based PSMA therapeutic agents is likely due to the binding to the PSMA mouse isoform. However, most preclinical development of PSMA-based albumin-binding radiotherapeutics has been focused on reducing tumor-to-kidney AUC values [12,20,23]. Furthermore, recent in vitro studies revealed that these albumin-binding compounds are associated with significantly higher binding to human blood plasma than mouse plasma [16,36,37]. Accordingly, tumor-to-blood ratios appeared to be critical in optimizing the albumin-binding agents in patient studies, as reported by Kramer et al. [23]. Their analysis from clinical studies revealed that the kidney-absorbed dose of ¹⁷⁷Lu-PSMA-617 was ~4-fold lower than the albumin-based agents, ¹⁷⁷Lu-EB-PSMA-617 and ¹⁷⁷Lu-PSMA-ALB-56. In addition, the tumor-to-red-marrow values of ¹⁷⁷Lu-EB-PSMA-617 and ¹⁷⁷Lu-PSMA-ALB-56 were >10-fold and >5-fold higher than ¹⁷⁷Lu-PSMA-617, respectively [23]. As a result, the tumor dose at maximum injectable activity was ¹⁷⁷Lu-EB-PSMA-617 (60.1 Gy), and ¹⁷⁷Lu-PSMA-ALB-56 (96 Gy) was significantly lower than ¹⁷⁷Lu-PSMA-671 (131 Gy). Although speculative, projecting mouse data to human data, the estimated tumor radiation doses of ¹⁷⁷Lu-Alb-L4 (84 Gy) and ¹⁷⁷Lu-Alb-L6 (94 Gy), respectively, were in the range of the reported agents, considering red marrow as the dose-limiting organ. Following similar dosimetry logistics, the tumor radiation dose of ¹⁷⁷Lu-L1 (224 Gy) is estimated to be significantly higher than the developed albumin-based agents, considering kidney as the dose-limiting organ.

Furthermore, we compared these AUC values with our previously reported PSMA-based antibody ¹¹¹In-DOTA-5D3, which was carried out using the same tumor models and same time points [38]. The estimated values are tumor AUC_0-192h is 4484 ± 790%ID/g.h, blood AUC_0-192h is 2015 ± 251%ID/g.h, and kidney AUC_0-192h is 880 ± 38%ID/g.h. The data revealed that the tumor uptake values of the low-molecular-weight albumin-binding agents, such as ¹⁷⁷Lu-Alb-L4, are significantly (~3-fold) higher than those of the large antibody-based agent, as anticipated. In comparison, the blood AUC of ¹⁷⁷Lu-Alb-L4 is significantly lower (>6-fold) than ¹¹¹In-DOTA-5D3, albeit with increased kidney uptake (>4-fold). The data suggest that an optimized albumin-binding agent might be a superior option for PSMA-based radiopharmaceutical therapy compared to antibody-based agents.

Although many reported preclinical studies using albumin-binding PSMA-based ¹⁷⁷Lu-labeled therapeutics were conducted using the PSMA+ PC3 PIP tumor model, these studies mainly used athymic nude mice for tumor implantation. In contrast, our studies were performed using NSG mice because of our institutional availability of this strain. The other notable variables could be related to the high specific activity of the agents used in our studies obtained through HPLC purification and the relatively large tumor used for the biodistribution studies (Supplementary Table S9). We anticipate that blood and normal tissue uptake data could be a rational indicator to compare the performance of the radiotherapeutics, as revealed in a recent report by Tschan et al. [39]. IBU-based ¹⁷⁷Lu-Alb-L4 displayed a low binding affinity to blood. Our data suggest that an optimized PEG linker, in combination with an IBU-based moiety, may improve the pharmacokinetics of this class of agents.

There are a few limitations to the study. We used a transfected cell line (PSMA+ PC3 PIP) which may have displayed an unrealistically high PSMA expression and may not have reflected the natural abundance and heterogeneity of PSMA in human cancer. However, the PSMA+ PC3 PIP and PSMA− PC3 flu cells have the advantage of being isogenic and androgen-independent cell lines and are anticipated to display similar biological factors for evaluating tumor pharmacokinetics, except for the PSMA expression levels. Many reported studies currently use the same tumor models for developing similar structure–activity relationship data because of the fast and predictable growth rate. There might be some variability associated with tumor sizes, although the data related to the pharmacokinetic performance are expected to be similar.

4. Materials and Methods

General Methods. All reagents and solvents were purchased from Sigma-Aldrich or Fisher Scientific unless specified and are listed in Supplementary Table S1. These were directly used without further purification. Amino acid derivatives were received from Chem-Impex International. 1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetraacetic acid mono-N-hydroxysuccinimide ester (DOTA-NHS-ester) was purchased from Macrocyclics Inc, Dallas. (2S)-2-[[(1S)-1-carboxy-3-methylbutyl]carbamoylamino]pentanedioic acid (ZJ43) was synthesized in-house following that reported in [31]. All new compounds were synthesized using standard solution-phase chemistry based on our well-established methods [27]. Column chromatography of intermediates was performed using Biotage Isolera Flash Chromatography with SNAP Ultra C₁₈ Sep-Pak columns. Purification of final compounds was performed using an Agilent (Santa Clara, CA, USA) high-performance liquid chromatography (HPLC) system equipped with a model 1200 quaternary pump and a model 1200 UV absorbance detector using a 250 mm × 10 mm Phenomenex Luna C18 column. Spectral characterization data of the new agents are included in Supplementary Materials. ¹⁷⁷LuCl₃ was supplied by the US Department of Energy Isotope Program. All ¹⁷⁷Lu-labeled radioligands were purified via HPLC to remove unreacted ligand from the radiolabeled material to ensure high specific radioactivity. Animal studies were undertaken according to the guidelines set forth by Johns Hopkins Animal Care and Use Committee.

4.1. Radiochemistry

Radiolabeling was performed under standard labeling conditions in ammonium acetate buffer (0.2 M) at pH ~4.5 following our reported method [27]. The radiolabeling was performed in a radiochemistry microwave chamber at 90 °C for 5 min at 40 watts (Resonance Instruments Inc., Skokie, IL, USA), and the reaction solution was purified using reverse-phase HPLC. An isocratic HPLC method was developed in each case, as listed in Supplementary Table S2, to remove the unreacted ligand from the radiolabeled material to ensure high specific activity. L-Ascorbic acid was added to the isolated radiolabeled compounds in the final formulation to maintain stability and was used for in vitro and in vivo experiments. All ¹⁷⁷Lu-labeled compounds were stable for up to 4 h at room temperature without significant radiolysis and were stable at 4 °C for 7 d at a concentration of 37 MBq/mL.

4.2. Measurement of Partition Coefficients

The partition coefficient of the ¹⁷⁷Lu-labeled agents was determined in 1-octanol and phosphate-buffered saline (PBS) (pH 7.4). 1-Octanol (3 mL) and PBS (3 mL) were pipetted into a 15 mL test tube containing 370 kBq of the test compound. The test tube was vortexed for 2 min and then centrifuged (4000× g, 5 min). Aliquots (0.1 mL) from the 1-octanol and PBS phases were transferred into two test tubes for counting. The amount of radioactivity in each test tube was measured using the automated γ-counter. The partition coefficient was calculated using the following equation: log P_ow = log[counts_1-octanol/counts_PBS], and the data are listed in Supplementary Table S3.

4.3. In Vitro Assays

4.3.1. Competitive Inhibition Assays

NAALADase Assay. Binding affinities of all new ligands were measured according to a previously described competitive fluorescence-based assay [40]. In brief, cell lysates of LNCaP cell extracts were incubated with PSMA-targeted agents (0.01 nm–100 μM) in the presence of 4 μM NAAG at 37 °C for 2 h and the reference PSMA inhibitor, ZJ43 (0.01 nm–100 μM) [31]. The amount of released glutamate from NAAG was measured by incubating it with a working solution of the Amplex Red glutamic acid kit (Molecular Probes Inc., Eugene, OR, USA) at 37 °C for 60 min. Fluorescence was measured with excitation at 535 nm and emission at 590 nm using a microplate reader. Inhibition curves were determined using semi-log plots. Data were analyzed using a one-site total binding regression using GraphPad Prism version 9 for Windows (GraphPad Software, San Diego, CA, USA). The IC₅₀ values were determined as the concentration at which enzymatic activity was inhibited by 50%. Assays were performed in triplicate, with the entire inhibition study repeated at least once. Enzyme inhibitory constants (K_i values) were generated using the Cheng–Prusoff conversion [41].

4.3.2. Protein-Binding Assay

Albumin-binding properties of the compounds were analyzed following a reported method [42]. A solution of ¹⁷⁷Lu-Alb-L2–¹⁷⁷LuAlb-L5 (74 kBq in 10 μL PBS) was added to 190 μL human serum albumin solution (45 mg HSA in 1 mL of PBS). After mixing, the solution was incubated at 37 °C for 60 min. Then, a 100 μL of the reaction solution was loaded onto a gel filtration column (Thermo Scientific™ Zeba™ Spin 7K MWCO size exclusion spin columns, Waltham, MA, USA), previously equilibrated with 0.1 M acetate buffer (pH 6.0), followed by centrifugation (1500× g, 2 min). The radioactivity of the column and eluate was then measured using an automated γ-counter. The data are listed in Supplementary Table S3.

4.3.3. Cell Culture

Androgen-independent PSMA-high (PSMA+) PC3 PIP cells and PSMA-low (PSMA−) PC3 flu cell lines are an isogenic subline pair of human PC3 cell lines (androgen-independent PSMA-negative bone metastatic prostate carcinoma). These cell lines were generously provided by Warren Heston (Cleveland Clinic). According to the literature reports, PC3 PIP cells were initially generated via the transfection of PC3 cells, employing VSV-G pseudo-typed lentiviral-vector-expressing human PSMA [43,44,45]. As reported previously, flow cytometry and Western blot assays are routinely used to evaluate PSMA expressions of PSMA+ PC3 PIP cells and PSMA− PC3 flu cells [27]. Selected data generated for the studies of this report are provided in Supplementary Figure S1. These cells were cultured in RPMI-1640 cell culture medium supplemented with 10% fetal calf serum, L-glutamine, antibiotics, and puromycin (2 µg/mL) to maintain expression, and were used for in vitro cell uptake studies and in vivo tumor generation. All cell cultures were maintained at 5% carbon dioxide at 37 °C in a humidified incubator. Authentication of the cell lines were performed routine by JHU GRCF (https://grcf.jhmi.edu/biorepository-cell-center/bioprocessing/cell-line-authentication). We tested mycoplasma contaminations of the cell line cultures every two weeks using the MycoAlert PLUS mycoplasma detection kit (Lonza).

4.3.4. Cell Uptake and Internalization Study

Cell uptake studies were performed following our previously reported protocol [27]. In brief, adherent PSMA+ PC3 PIP cells and PSMA− PC3 flu cells detached using nonenzymatic buffer (Gibco) and ~1 million cells per tube were incubated in 100 µL of each radiolabeled agent (370 kBq/mL) for 2 h and 24 h at 37 °C in the 100 µL growth medium (binding buffer (1× PBS + 2mM EDTA + 0.5% FBS)). After incubation, the medium was removed at the indicated time points, and the cells were washed three times with ice-cold PBS. The collected pooled washes and the cell pellets were counted using an automated γ-counter. The radioactivity values were converted into a percentage of incubated dose (%ID) per million cells. Experiments were performed in triplicate and repeated two times.

For the PSMA blocking studies, PSMA+ PC3 PIP cells were pre-incubated with ZJ43 (10 µM final concentration) for 30 min and then washed 3 times with binding buffer followed by incubation of the radioactive dose (100 µL of 370 kBq/mL in binding buffer) for 2 h. The cell uptake studies were then conducted using the method mentioned in the previous section.

For the internalization assays, cells were detached using nonenzymatic buffer, and aliquots of 1 million cells per tube were incubated with 370 kBq/mL of each radiolabeled agent for 2 h and 24 h at 37 °C along with the 100 µL of the binding buffer, as mentioned in the cell uptake study. At the indicated time points, the medium was removed, and cells were washed with binding buffer followed by a mildly acidic buffer (50 mM glycine, 150 mM NaCl (pH 3.0)) at 4 °C for 5 min. The acidic buffer was then collected, and cells were washed twice with binding buffer. The collected pooled washes (containing cell-surface-bound ¹⁷⁷Lu-Alb-L1–¹⁷⁷Lu-Alb-L6) and cell pellets (containing internalized ¹⁷⁷Lu-Alb-L1–¹⁷⁷Lu-Alb-L6) were counted in an automated γ-spectrometer along with the standards. All radioactivity values were converted into a percentage of incubated dose (%ID) per million cells. Experiments were performed in triplicate and repeated 2 times. Data were fitted according to linear regression analysis.

4.3.5. Clonogenic Survival Assay

Cells (200–1000) were seeded in 60 mm culture dishes. Each radioligand (¹⁷⁷Lu-L1, ¹⁷⁷Lu-Alb-L2, ¹⁷⁷Lu-Alb-L5) was diluted in a prewarmed medium at different concentrations (0, 0.37, 1.85, 3.7, 18.5, 37, 185, and 370 kBq/mL) and incubated with the cells for 48 h, as we previously reported [27]. The radiolabeled compound was replaced with fresh medium, and cells were incubated for 2 weeks or until colonies had at least 50 cells. The colonies were stained with crystal violet and counted, and the surviving fraction was normalized to the control plating efficiency, as previously described [46].

4.4. In Vivo Experiments

Biodistribution

Five-to-six-week-old male NSG mice were purchased from Johns Hopkins University research animal resources. Briefly, ~14–20 days after subcutaneous inoculation of PSMA+ PC3 PIP (3 × 10⁶ cells) or PSMA− flu cells (1 × 10⁶ cells) in 100 µL HBSS solution on the upper flanks, tissue biodistribution studies were performed. Male NSG mice bearing PSMA+ PC3 PIP and PSMA− PC3 flu xenografts were injected intravenously with the respective ¹⁷⁷Lu-labeled agent, ¹⁷⁷Lu-Alb-L2–¹⁷⁷Lu-Alb-L6 (1.85 MBq) diluted in 150 µL saline. Mice were sacrificed at 2 h, 24 h, 48 h, and 192 h post-injection, and selected tissues were harvested, weighed, and measured radioactivity using an automated γ-counter. A group of 4 mice was used for each time point; the results were listed as the percentage of the injected dose per gram of tissue mass (%ID/g). The data are presented as the average ± standard deviation (SD). The biodistribution study of ¹⁷⁷Lu-L1 was performed at only 24 h during the biodistribution study of ¹⁷⁷Lu-Alb-L2 and ¹⁷⁷Lu-Alb-L5 in a single experiment. The biodistributions of ¹⁷⁷Lu-Alb-L3, ¹⁷⁷Lu-Alb-L4, and ¹⁷⁷Lu-L1 (24 h) were acquired in a separate experiment. The data are listed in Supplementary Tables S4–S10. Biodistribution data of ¹⁷⁷Lu-L1 at 3 h, 24 h, 48 h, and 72 h were obtained from our previous report [27]. The 192 h post-injection data of ¹⁷⁷Lu-L1 are unpublished and were acquired during the same study.

4.5. Statistical Analysis

All graphs, AUC calculations, and statistical analyses were created and performed using the GraphPad Prism software (version 9.0). Significant differences were evaluated using a one-way ANOVA or unpaired t-test; p-values < 0.05 were considered to be significant. p-values lower than 0.05 (p < 0.05), p < 0.01, p < 0.001, and p < 0.0001 were referred to with one (*), two (**), three (***), or four (****) asterisks, respectively.

5. Conclusions

The data suggest that ¹⁷⁷Lu-Alb-L4 could be an improved option for ¹⁷⁷Lu-L1, and further modification in the linker construct using a PEG linker may be possible to reduce blood and kidney uptake.