Hybrid Chelator-Based PSMA Radiopharmaceuticals: Translational Approach

(1) Background: Prostate-specific membrane antigen (PSMA) has been extensively studied in the last decade. It became a promising biological target in the diagnosis and therapy of PSMA-expressing cancer diseases. Although there are several radiolabeled PSMA inhibitors available, the search for new compounds with improved pharmacokinetic properties and simplified synthesis is still ongoing. In this study, we developed PSMA ligands with two different hybrid chelators and a modified linker. Both compounds have displayed a promising pharmacokinetic profile. (2) Methods: DATA5m.SA.KuE and AAZTA5.SA.KuE were synthesized. DATA5m.SA.KuE was labeled with gallium-68 and radiochemical yields of various amounts of precursor at different temperatures were determined. Complex stability in phosphate-buffered saline (PBS) and human serum (HS) was examined at 37 °C. Binding affinity and internalization ratio were determined in in vitro assays using PSMA-positive LNCaP cells. Tumor accumulation and biodistribution were evaluated in vivo and ex vivo using an LNCaP Balb/c nude mouse model. All experiments were conducted with PSMA-11 as reference. (3) Results: DATA5m.SA.KuE was synthesized successfully. AAZTA5.SA.KuE was synthesized and labeled according to the literature. Radiolabeling of DATA5m.SA.KuE with gallium-68 was performed in ammonium acetate buffer (1 M, pH 5.5). High radiochemical yields (>98%) were obtained with 5 nmol at 70 °C, 15 nmol at 50 °C, and 60 nmol (50 µg) at room temperature. [68Ga]Ga-DATA5m.SA.KuE was stable in human serum as well as in PBS after 120 min. PSMA binding affinities of AAZTA5.SA.KuE and DATA5m.SA.KuE were in the nanomolar range. PSMA-specific internalization ratio was comparable to PSMA-11. In vivo and ex vivo studies of [177Lu]Lu-AAZTA5.SA.KuE, [44Sc]Sc-AAZTA5.SA.KuE and [68Ga]Ga-DATA5m.SA.KuE displayed specific accumulation in the tumor along with fast clearance and reduced off-target uptake. (4) Conclusions: Both KuE-conjugates showed promising properties especially in vivo allowing for translational theranostic use.


Introduction
Prostate-specific membrane antigen (PSMA) has become a very popular target in the diagnosis and treatment of prostate cancer in the last decade. PSMA is a glycoprotein with several functions originating from its glutamate-carboxypeptidase activity. In the central nervous system, PSMA acts as NAALADase, which cleaves the glutamate moiety from the neurotransmitter N-acetyl aspartyl glutamate. However, in the proximal small intestine, this enzyme, called folate hydrolase FOLH1, releases glutamate residues from poly-glutamated folate [1,2]. Besides these physiological functions, PSMA seems to play an important role in prostate carcinogenesis since it is highly expressed in prostate tumor cells.
This expression correlates with the aggressiveness and invasiveness of the tumor [3][4][5], and is a major reason for choosing PSMA as a molecular target in the management of prostate cancer (PC).
Prostate cancer is the second most common cancer among men and the fifth leading cause of death worldwide [6,7]. However, early detection of PC in a localized stage can significantly reduce its mortality, leading to a 5-year survival rate of more than 90% [8]. In contrast, late-stage tumors are aggressive and almost resistant to available therapies. Metastatic castration-resistant prostate cancer (mCRPC) is one of the most aggressive forms of prostate cancer, with poor outcomes and restricted therapy options [9]. One of the most promising approaches herein is PSMA-targeted radioligand diagnosis and therapy. The unique characteristics of PSMA as a molecular target in combination with the small-molecule PSMA inhibitors as target vectors paved the way for the development of highly sensitive radiopharmaceuticals like the PET radioligand [ 68 Ga]Ga-PSMA-11 and its therapeutic counterpart [ 177 Lu]Lu-PSMA-617 [10,11].
One of the challenges in designing appropriate PSMA inhibitors for theranostic use is balancing the reduction of off-target accumulation in order to minimize the exposure and irradiation of excretory organs and other tissues where physiological PSMA expression is known, such as the salivary glands and the kidneys [12][13][14][15], with the development of PSMA ligands which can be easily synthesized and effectively labeled. To address some of these concerns, we developed AAZTA 5 .SA.KuE and DATA 5m .SA.KuE.
Like all PSMA ligands, the herein described PSMA radiopharmaceuticals consist of three parts: chelator, linker moiety, and a KuE-based PSMA-targeting vector.
The DATA 5m chelator has optimal labeling properties at mild conditions for the generator-based PET nuclide gallium-68. Furthermore, the AAZTA and DATA chelators are also suitable for instant kit-labeling applications with e.g., lutetium-177 and gallium-68 [17,23].
Lysine-urea-glutamate (KuE) has been established as a PSMA inhibitor. KuE consists of lysine and glutamate which are both linked to each other via a urea unit. KuE is based on the natural PSMA-substrate NAAG, but cannot be cleaved by the enzyme [10]. Both PSMA-617 and PSMA-11 carry this structural unit as the PSMA-binding entity [24].
The third structural element found in PSMA radiopharmaceuticals is the linker moiety, connecting the chelator to the urea-based target vector. In addition to the function of coupling, these linkers are usually designed to improve the pharmacokinetics of the compounds [10]. These moieties can interact with the aromatic-binding region of the PSMA binding pocket, leading to an increase in the affinity of the PSMA ligand [25]. The coupling of KuE is achieved via the side-chain amine of the lysine. Usually, amide coupling reactions are used for this purpose. Alternatively, conjugation can be achieved by using square acid diethyl esters (SADE). This group allows two amines to be selectively coupled via asymmetric amidation, forming a squaramide. This simplifies the synthesis in so far as, for example, no protective group chemistry is required, as is the case with standard amide couplings. The coupling reaction is selective with amines only and by controlling the amidation of both squaric acid esters via pH [17,[26][27][28][29]. The control of the asymmetric amidation via the pH value can be explained by the different aromaticity and thus the different mesomeric stabilities of the individual intermediates at the different pH values ( Figure 2) [30][31][32]. Lysine-urea-glutamate (KuE) has been established as a PSMA inhibitor. KuE consists of lysine and glutamate which are both linked to each other via a urea unit. KuE is based on the natural PSMA-substrate NAAG, but cannot be cleaved by the enzyme [10]. Both PSMA-617 and PSMA-11 carry this structural unit as the PSMA-binding entity [24].
The third structural element found in PSMA radiopharmaceuticals is the linker moiety, connecting the chelator to the urea-based target vector. In addition to the function of coupling, these linkers are usually designed to improve the pharmacokinetics of the compounds [10]. These moieties can interact with the aromatic-binding region of the PSMA binding pocket, leading to an increase in the affinity of the PSMA ligand [25]. The coupling of KuE is achieved via the side-chain amine of the lysine. Usually, amide coupling reactions are used for this purpose. Alternatively, conjugation can be achieved by using square acid diethyl esters (SADE). This group allows two amines to be selectively coupled via asymmetric amidation, forming a squaramide. This simplifies the synthesis in so far as, for example, no protective group chemistry is required, as is the case with standard amide couplings. The coupling reaction is selective with amines only and by controlling the amidation of both squaric acid esters via pH [17,[26][27][28][29]. The control of the asymmetric amidation via the pH value can be explained by the different aromaticity and thus the different mesomeric stabilities of the individual intermediates at the different pH values ( Figure 2) [30][31][32]. With regard to PSMA radiopharmaceuticals, the use of squaric acid shows another advantage. Squaric acid has an aromatic character and can therefore interact with the aromatic binding region in the PSMA binding pocket resulting in an increased affinity. Greifenstein et al. recently demonstrated that a square amide containing DOTAGA-KuE derivative is comparable to the standard compounds PSMA-617 and PSMA-11 in terms of in vitro binding affinity, tumor accumulation, and in vivo kinetics [28]. The synthesis of the DATA chelator is based on the synthesis described by Farkas et al. [16] and Greifenstein et al. [17]. It was synthesized according to Figure 3. With regard to PSMA radiopharmaceuticals, the use of squaric acid shows another advantage. Squaric acid has an aromatic character and can therefore interact with the aromatic binding region in the PSMA binding pocket resulting in an increased affinity. Greifenstein et al. recently demonstrated that a square amide containing DOTAGA-KuE derivative is comparable to the standard compounds PSMA-617 and PSMA-11 in terms of in vitro binding affinity, tumor accumulation, and in vivo kinetics [28]. The synthesis of the DATA chelator is based on the synthesis described by Farkas et al. [16] and Greifenstein et al. [17]. It was synthesized according to Figure 3.

Organic Synthesis
N,N -dibenzylethyldiamines were first reacted with tert-butyl bromoacetate to give the di-alkylated compound 1. The benzyl protecting groups were then removed by reduction. The diazepane 3 was formed by a double Mannich reaction. For this purpose, 2-nitrocyclohexanone was used, the ring of which was opened using the anion exchanger Amberlyst ® A21. In the following Mannich reaction, this ring-opened intermediate reacted with 2 to form the desired diazepane 3.
After reduction of the nitro group (4), tert-butyl bromoacetate was added in an undercurrent to give the mono-alkylated compound 5. The secondary amine of 5 was then methylated in a reductive amination. This led to the protected chelator DATA 5m 6. In order to functionalize 6 with the target vector, however, it was necessary to introduce an ethylenediamine bridge. For this purpose, the methyl ester of 6 was saponified using lithium hydroxide (compound 7) and the mono Boc-protected ethylenediamine was linked via an amide coupling to get 8. After an acidic deprotection compound 9 could be conjugated to the target vector using squaric acid. N,N′-dibenzylethyldiamines were first reacted with tert-butyl bromoacetate to give the di-alkylated compound 1. The benzyl protecting groups were then removed by reduction. The diazepane 3 was formed by a double Mannich reaction. For this purpose, 2-nitrocyclohexanone was used, the ring of which was opened using the anion exchanger Amberlyst ® A21. In the following Mannich reaction, this ring-opened intermediate reacted with 2 to form the desired diazepane 3.
After reduction of the nitro group (4), tert-butyl bromoacetate was added in an undercurrent to give the mono-alkylated compound 5. The secondary amine of 5 was then methylated in a reductive amination. This led to the protected chelator DATA 5m 6. In order to functionalize 6 with the target vector, however, it was necessary to introduce an ethylenediamine bridge. For this purpose, the methyl ester of 6 was saponified using lithium hydroxide (compound 7) and the mono Boc-protected ethylenediamine was linked via an amide coupling to get 8. After an acidic deprotection compound 9 could be conjugated to the target vector using squaric acid.
The PSMA inhibitor lysine-urea-glutamate (KuE) was synthesized and coupled to 3,4-dibutoxycyclobut-3-en-1,2-dione (SADE) according to Figure 4. The PSMA inhibitor lysine-urea-glutamate (KuE) was synthesized and coupled to 3,4-dibutoxycyclobut-3-en-1,2-dione (SADE) according to Figure 4.  For the introduction of the urea unit, the amino group of the protected lysine was transformed into an isocyanate using triphosgene. The isocyanate was then reacted with tert-butyl protected glutamate and the protected PSMA inhibitor lysine-urea-glutamate 10 was obtained and followed by reductive deportation of the lysine side chain, yielding 11. This compound was then coupled to SADE in phosphate buffer at pH 7. Acidic deprotection of the protected compound 12 led to the couplable PSMA inhibitor lysine-ureaglutamate-squaric acid monoester 13 (KuE.SAME).
The free primary amine of DATA 5m (9) was then coupled to the free coupling side of KuE.SAME (13) in 0.5 M phosphate buffer at pH 9 to obtain the final compound For the introduction of the urea unit, the amino group of the protected lysine was transformed into an isocyanate using triphosgene. The isocyanate was then reacted with tert-butyl protected glutamate and the protected PSMA inhibitor lysine-urea-glutamate 10 was obtained and followed by reductive deportation of the lysine side chain, yielding 11. This compound was then coupled to SADE in phosphate buffer at pH 7. Acidic deprotection of the protected compound 12 led to the couplable PSMA inhibitor lysine-urea-glutamatesquaric acid monoester 13 (KuE.SAME).
For the introduction of the urea unit, the amino group of the protected lysine was transformed into an isocyanate using triphosgene. The isocyanate was then reacted with tert-butyl protected glutamate and the protected PSMA inhibitor lysine-urea-glutamate 10 was obtained and followed by reductive deportation of the lysine side chain, yielding 11. This compound was then coupled to SADE in phosphate buffer at pH 7. Acidic deprotection of the protected compound 12 led to the couplable PSMA inhibitor lysine-ureaglutamate-squaric acid monoester 13 (KuE.SAME).
The free primary amine of DATA 5m (9) was then coupled to the free coupling side of KuE.SAME (13) in 0.5 M phosphate buffer at pH 9 to obtain the final compound DATA 5m .SA.KuE (14) ( Figure 5).

Radiolabeling
Radiolabeling of AAZTA 5 .SA.KuE with scandium-44 and lutetium-177 was performed according to the literature [17]. DATA 5m .SA.KuE was radiolabeled with gallium-68 in ammonium acetate buffer (1 M, pH 5.5), varying amounts of precursor (5 nmol to 60 nmol), and different temperatures (room temperature to 70 °C). Labeling was carried out in triplicate with 30-50 MBq of gallium-68. Figure 6A shows the kinetic studies of the gallium-68-radiolabeling of DATA 5m .SA.KuE. The lower the quantity of precursor used, the higher the temperature required to obtain quantitative radiochemical yields (RCY). Labeling of 10 nmol at 50 °C only achieved a RCY of 56% after 15 minutes. The increase to 15 nmol at 50 °C results in quantitative RCY (>99%). The increase of temperature even allowed the quantitative labeling (>99% RCY) of 5 nmol. Furthermore, 50 µg (60 nmol) can be radiolabeled in yields of over 99% with gallium-68 even at room temperature. The high radiochemical yield and high purity of [ 68 Ga]Ga-DATA 5m .SA.KuE was confirmed by radio-HPLC ( Figure 6B).

Radiolabeling
Radiolabeling of AAZTA 5 .SA.KuE with scandium-44 and lutetium-177 was performed according to the literature [17]. DATA 5m .SA.KuE was radiolabeled with gallium-68 in ammonium acetate buffer (1 M, pH 5.5), varying amounts of precursor (5 nmol to 60 nmol), and different temperatures (room temperature to 70 • C). Labeling was carried out in triplicate with 30-50 MBq of gallium-68. Figure 6A shows the kinetic studies of the gallium-68-radiolabeling of DATA 5m .SA.KuE. The lower the quantity of precursor used, the higher the temperature required to obtain quantitative radiochemical yields (RCY). Labeling of 10 nmol at 50 • C only achieved a RCY of 56% after 15 minutes. The increase to 15 nmol at 50 • C results in quantitative RCY (>99%). The increase of temperature even allowed the quantitative labeling (>99% RCY) of 5 nmol. Furthermore, 50 µg (60 nmol) can be radiolabeled in yields of over 99% with gallium-68 even at room temperature. The high radiochemical yield and high purity of [ 68 Ga]Ga-DATA 5m .SA.KuE was confirmed by radio-HPLC ( Figure 6B).
Studies of the complex stability were performed in human serum (HS) and phosphate buffered saline (PBS). In both media, [ 68 Ga]Ga-DATA 5m .SA.KuE showed a stability of >98% over a period of 120 minutes ( Figure 6C).

PSMA Binding Affinity
The PSMA binding affinity of DATA 5m .SA.KuE and AAZTA 5 .SA.KuE, as well as PSMA-11, was determined in a competitive radioligand assay using PSMA-positive LNCaP cells that were incubated with 0.75 nM [ 68 Ga]Ga-PSMA-10 in the presence of 12 increasing concentrations of the non-labeled SA-conjugated compounds. The measured radioactivity values were plotted against the concentrations of the SA conjugates ( Figure 7). IC 50 values were determined using GraphPad Prism 9 (Table 1). AAZTA 5 .SA.KuE and DATA 5m .SA.KuE showed similar binding affinities while PSMA-11 seems to have two-fold higher affinity in vitro.  Studies of the complex stability were performed in human serum (HS) and phos phate buffered saline (PBS). In both media, [ 68 Ga]Ga-DATA 5m .SA.KuE showed a stability of >98% over a period of 120 minutes ( Figure 6C).

PSMA Binding Affinity
The PSMA binding affinity of DATA 5m .SA.KuE and AAZTA 5 .SA.KuE, as well as PSMA-11, was determined in a competitive radioligand assay using PSMA-positive LNCaP cells that were incubated with 0.75 nM [ 68 Ga]Ga-PSMA-10 in the presence of 12 in creasing concentrations of the non-labeled SA-conjugated compounds. The measured radi oactivity values were plotted against the concentrations of the SA conjugates ( Figure 7). IC5 values were determined using GraphPad Prism 9 (Table 1). AAZTA 5 .SA.KuE and DATA 5m .SA.KuE showed similar binding affinities while PSMA-11 seems to have twofold higher affinity in vitro.
Molecules 2021, 26, 6332 8 uptake of [ 68 Ga]Ga-DATA 5 .SA.KuE were found to be PSMA-specific, since they coul blocked by co-injection of PMPA as seen in Figure 10.

Animal Studies
In order to evaluate the in vivo behavior of the SA.KuE conjugates, an LNCaP-xenograft model was used. Labeling of AAZTA 5 .SA.KuE with the different nuclides scandium-44 and lutetium-177 seemed to have no impact on the pharmacokinetic properties of the conjugates, since there were no significant differences observed in the biodistribution data ( Figure 9). Tumor accumulation values of all four compounds were similar, 3.92 ± 0.50% ID/g, 5.41 ± 0.83% ID/g, 4.43 ± 0.56% ID/g and 5.52 ± 0.75% ID/g for [ 44 5 .SA.KuE were found to be PSMA-specific, since they could be blocked by coinjection of PMPA as seen in Figure 10.  To further understand the pharmacokinetics of the developed PSMA ligands, we performed µPET-scans with the same xenograft model (Figure 11). Tumor accumulation of all three compounds was very similar. The kidney uptake of [ 68 Ga]Ga-DATA 5 .SA.KuE was remarkably lower than the reference compound [ 68 Ga]Ga-PSMA-11. This finding correlates with the results obtained from the time-activity curves of both compounds ( Figure  12). Herein, the radioactivity concentration of [ 68 Ga]Ga-DATA 5m .SA.KuE decreased con- To further understand the pharmacokinetics of the developed PSMA ligands, we performed µPET-scans with the same xenograft model (Figure 11). Tumor accumulation of all three compounds was very similar. The kidney uptake of [ 68 Ga]Ga-DATA 5

.SA.KuE
Molecules 2021, 26, 6332 9 of 18 was remarkably lower than the reference compound [ 68 Ga]Ga-PSMA-11. This finding correlates with the results obtained from the time-activity curves of both compounds ( Figure 12). Herein, the radioactivity concentration of [ 68 Ga]Ga-DATA 5m .SA.KuE decreased continuously 10 min p.i. while the concentration in the tumor remained constant. However, the radioactivity concentration of [ 68 Ga]Ga-PSMA-11 remained at a higher level during the period of the scan. As demonstrated in the µPET scans, uptake in the tumor as well as in the kidney was PSMA-specific. After co-injection of PMPA, almost no radioactivity could be detected (Figure 11). To further understand the pharmacokinetics of the developed PSMA ligands, we performed µPET-scans with the same xenograft model (Figure 11). Tumor accumulation of all three compounds was very similar. The kidney uptake of [ 68 Ga]Ga-DATA 5 .SA.KuE was remarkably lower than the reference compound [ 68 Ga]Ga-PSMA-11. This finding correlates with the results obtained from the time-activity curves of both compounds ( Figure  12). Herein, the radioactivity concentration of [ 68 Ga]Ga-DATA 5m .SA.KuE decreased continuously 10 min p.i. while the concentration in the tumor remained constant. However, the radioactivity concentration of [ 68 Ga]Ga-PSMA-11 remained at a higher level during the period of the scan. As demonstrated in the µPET scans, uptake in the tumor as well as in the kidney was PSMA-specific. After co-injection of PMPA, almost no radioactivity could be detected (Figure 11).

Discussion
The discovery of PSMA as molecular target in the diagnosis and therapy of prostate cancer, as well as the application of radiolabeled PSMA inhibitors, have revolutionized the management of this disease resulting in a significant improvement especially in staging and assessment of prostate cancer [38]. Although several PSMA ligands have been developed over the last decades, the search for novel tracers with optimized pharmacokinetic properties particularly for therapeutic purposes is still present, since some of the clinically used PSMA radioligand therapeutics e.g., [ 225 Ac]Ac-PSMA-617 display some severe side effects, like xerostomia [12,13,39].
To determine the effect of the chelator on the PSMA binding affinity and the internalization ratio of PSMA ligands, we synthesized two PSMA inhibitors with different hybrid chelators. In the cell-based assays, both DATA 5m .SA.KuE and AAZTA 5

Discussion
The discovery of PSMA as molecular target in the diagnosis and therapy of prostate cancer, as well as the application of radiolabeled PSMA inhibitors, have revolutionized the management of this disease resulting in a significant improvement especially in staging and assessment of prostate cancer [38]. Although several PSMA ligands have been developed over the last decades, the search for novel tracers with optimized pharmacokinetic properties particularly for therapeutic purposes is still present, since some of the clinically used PSMA radioligand therapeutics e.g., [ 225 Ac]Ac-PSMA-617 display some severe side effects, like xerostomia [12,13,39].
To determine the effect of the chelator on the PSMA binding affinity and the internalization ratio of PSMA ligands, we synthesized two PSMA inhibitors with different hybrid chelators. In the cell-based assays, both DATA 5m .SA.KuE and AAZTA 5 .SA.KuE showed similar binding affinity and internalization ratios, indicating that an exchange of DATA 5m against AAZTA 5 had no impact on either the binding affinity or on the internalization ratio in PSMA-positive LNCaP cells. These findings correlate with the results published by Sinnes et al., who investigated the influence of the exchange of DOTA chelator in DOTA-PSMA-617 against AAZTA 5 . Both DOTA-PSMA-617 and AAZTA 5. -PSMA-617 displayed similar in vitro binding affinities and internalization ratios in LNCaP cells [40]. However, the reported binding affinities and internalization ratios of AATA 5 -PSMA-617 and DOTA-PSMA-617 were higher than those of the SA.KuE conjugates. In particular, [ 44 Sc]Sc-PSMA-617 seems to display high PSMA-binding affinity as published by several groups [41,42]. Since PSMA-617 was not commercially available at the time this study was performed, PSMA-11 was used as reference.
However, PSMA-11 displayed also higher binding affinity in vitro which could be due to the better interaction with the PSMA binding pocket. In contrast, the internalization ratio of PSMA-11 was similar to these of the SA.KuE-conjugates. Interestingly, the investigated internalization fraction of [ 68 Ga]Ga-PSMA-11 was noticeably lower compared to the ratio described in literature [42,43] which could be due to differences in study design and setup. The PSMA-specificity of binding and uptake in LNCaP cells and LNCaP tumors could be demonstrated for all herein investigated PSMA-inhibitors by blocking PSMA receptors with the potent inhibitor PMPA.
In order to evaluate the pharmacokinetic behavior of our compounds and to compare them with PSMA-11, we performed animal studies using an LNCaP xenograft model. Thus, these compounds seem to display a rapid renal clearance along with a good tumor accumulation. However, the tumor uptake of the SA.KuE conjugates was lower than that of the gallium-68 and lutetium-177 labeled PSMA-617 radioligands [10,11]. Ghiani et al. recently described a novel scandium-44 labeled PSMA radioligand with even higher tumor accumulation than the PSMA-617 counterpart [41]. Nevertheless, a direct comparison between the presented results and those reported by other groups is not possible because of the differences in xenograft models and experimental setups.

Organic Synthesis
DATA 5m was synthesized according to the procedure described by Farkas et al. and Greifenstein et al. [17].

In Vitro Binding Affinity
PSMA binding affinity was determined according to the literature [39]. LNCaP-cells (purchased from Sigma-Aldrich) were cultured in RPMI 1640 (Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (Thermo Fisher Scientific), 100 µg/ml streptomycin, and 100 units/mL penicillin at 37 • C in 5% CO 2 in a humidified atmosphere. The medium was changed approximately every 3 days. Cells in exponential phase of growth were harvested by a 5 min treatment with a 0.05% trypsin-0.02% EDTA solution and neutralized with serum-containing medium prior to counting. 10 5 LNCaP cells per well were applied in MultiScreen HTS DV Filter Plates (Merck Millipore) and incubated with 0.75 nM [ 68 Ga]Ga-PSMA-10 in the presence of 12 increasing concentrations of the non-labeled SA-conjugated compounds. After incubation at room temperature for 45 min, cells bound on the filter plates were washed several times with ice-cold PBS to remove free radioactivity. The cell-bound activity was determined by punching out the filters and transferring them into individual tubes for measurement in a γ-counter (2480 WIZARD 2 Automatic Gamma Counter, PerkinElmer, Waltham, MA, USA). Data were analyzed in GraphPad Prism 9 using nonlinear regression. Experiments were replicated 4-times.

Internalization Ratio
Internalization ratio was determined according to the literature [45,46]. Prior to seeding cells, 24-well plates were coated with 0.1% poly-L-lysine (Sigma-Aldrich) in PBS for 20 min at room temperature. Subsequently, 10 5 LNCaP cells in 1 mL RPMI 1640 Medium were added in each well and incubated for 24 h at 37 • C. Then, 250 µL of the 68 Ga-labeled compounds in Opti-MEM™ I Reduced Serum (ThermoFisher) were added to each well to a final concentration of 30 nM. The plates were then incubated for 45 min at 4 • C and 37 • C respectively either with or without adding PMPA (Sigma-Aldrich) to a final concentration of 500 µM. The supernatant was removed and the cells were washed several times with ice-cold PBS. Afterwards, cells were incubated twice with 50 mM glycine buffer pH 2.8 for 5 min to remove the surface-bound radioactivity. In order to determine the internalized fraction of the compounds, cells were lysed by incubation with 0.3 M NaOH for 10 min.
LNCaP-xenografts were anesthetized with 2% isoflurane prior to i.v. injection of 0.5 nmol of the radiolabeled compounds. The specific activities of the tracers were approximately 10 MBq/nmol, 6 MBq/nmol, and 15 MBq/nmol of gallium-68-labeled compounds, MicroPET-imaging. After i.v. injection of the labeled compounds, anesthetized mice (one mouse for each group) were placed in the prone position in a nanoScan ® PET/MR (Mediso). MRI measurements were performed followed by a static PET scan with the nanoScan PET/MRI (Mediso, Budapest, Hungary). PET data were reconstructed with Teratomo 3D (four iterations, six subsets, voxel size 0.4 mm), co-registered to the MR, and analyzed with Pmod software (version 3.6) (PMOD Technologies LLC, Zürich, Switzerland) Material Map for co-registration of the PET scan; 3D Gradient Echo External Averaging (GRE-EXT), Multi Field of View (FOV); slice thickness, 0.6 mm; TE, 2 ms; TR, 15 ms; flip angle, 25 deg.

Conclusions
In summary, the synthesized hybrid chelator-based PSMAradiopharmaceuticals DATA 5m .SA.KuE and AAZTA 5  The decreased kidney uptake of the SA.KuE conjugates is noteworthy, which could be a major benefit in reducing irradiation of the kidneys, resulting in lower nephrotoxicity and improved tolerability.