Cholecystokinin-2 Receptor Targeting with Novel C-terminally Stabilized HYNIC-Minigastrin Analogs Radiolabeled with Technetium-99m

The high overexpression of cholecystokinin-2 receptors (CCK2R) in tumors, such as medullary thyroid carcinoma, allows for highly specific diagnostic and therapeutic targeting with radiolabeled peptide probes derived from natural ligands for the receptor. Based on the ideal imaging characteristics, high availability and low cost of technetium-99m (99mTc)-labeled radiopharmaceuticals we have developed two hydrazinonicotinic acid (HYNIC) conjugated minigastrin analogs allowing labeling at high specific activity. The CCK2R targeting peptide conjugates show specific amino acid substitutions in the C-terminal receptor-specific sequence with the aim to increase stability and tumor targeting. The CCK2R affinity and the cell uptake of the new radioligands were analyzed using A431 human epidermoid carcinoma cells stably transfected with human CCK2R and mock transfected cells. Metabolic studies in BALB/c mice revealed a high resistance against enzymatic degradation for both radioligands. Biodistribution studies in tumor-xenografted athymic BALB/c nude mice at 1 h and 4 h p.i. showed that the two 99mTc-labeled compounds showed varying uptake in receptor expressing organs, stomach and pancreas (1.3–10.4% IA/g), as well as kidneys, the main route of excretion (7.8–19.9% IA/g). The tumor uptake in A431-CCK2R xenografts was 24.75 ± 4.38% IA/g for [99mTc]Tc-HYNIC-MGS5 and 42.48 ± 6.99% IA/g for [99mTc]Tc-HYNIC-MGS11 at 4 h p.i., whereas the tumor-to-kidney ratio was comparable (2.6–3.3). On demand availability and potential application for radioguided surgery of a 99mTc-labeled minigastrin analog support the further evaluation of these highly promising new compounds.


Introduction
Regulatory peptides exhibiting high target specificity are suitable lead structures, particularly in the field of oncology, for the development of analogs for radionuclide imaging and therapy [1,2]. Such peptide analogs have the advantage of an easy production via well-established solid phase peptide synthesis (SPPS) allowing the conjugation of a chelator for radiolabeling and the introduction of different modifications into the peptide sequence to optimize pharmacokinetics. Additionally, peptide analogs show low immunogenicity and toxicity and are therefore preferable candidates as new targeting agents [3,4]. A highly promising molecular target to develop radiopeptides for diagnostic imaging and targeted radiotherapy (TRT) of medullary thyroid carcinoma (MTC), small cell lung cancer (SCLC), astrocytoma, stromal ovarian cancer, as well as carcinoids and other tumors of neuroendocrine 2 ([ 99m Tc]Tc-N 4 -Gly-dGlu-(Glu) 5 -Ala-Tyr-Gly-Trp-Met-Asp-Phe-NH 2 ) contains an open chain tetraamine chelator forming a monocationic complex [23,24]. [ 99m Tc]Tc-EDDA/HYNIC-MG11 ([ 99m Tc]Tc-EDDA/HYNIC-dGlu-Ala-Tyr-Gly-Trp-Met-Asp-Phe-NH 2 ) conjugated to the monodentate ligand hydrazinonicotinic acid (HYNIC) needs additional coligands such as tris(hydroxymethyl)methylglycine (tricine) or ethylenediamine-N,N'-diacetic acid (EDDA) to complete the coordination sphere [25]. The administration of these two 99m Tc-labeled MG analogs to patients was well tolerated showing no to only mild side effects. With [ 99m Tc]Tc-Demogastrin 2 all known lesions in nine MTC patients could be visualized. In a comparative study with [ 99m Tc]Tc-EDDA/HYNIC-MG11 and [ 99m Tc]Tc-EDDA/HYNIC-TOC the potential additional information which can be obtained in MTC patients using gastrin receptor scintigraphy was pointed out. The same MG analogs conjugated to DOTA showed drawbacks related to high kidney uptake or low in vivo stability, requiring further improvement to develop a CCK2R targeting peptide analog with optimal tumor targeting and biodistribution profile [8].
Various research groups have worked on the development of metabolically stable MG analogs. Different strategies such as cyclization, dimerization and substitutions of amino acids mainly in the N-terminal part of the peptide sequence were investigated [8]. However, due to rapid C-terminal enzymatic degradation the need of alternative stabilization strategies was suggested [26]. Recently we could present different amino acid substitutions introduced into the C-terminal receptorspecific sequence of MG analogs improving the stability against enzymatic degradation and the biodistribution profile [15,16]. After an intense preclinical evaluation of different substitutions, we discovered that most promising results in terms of improved in vivo stability and enhanced tumor targeting could be achieved when substituting methionine (Met) with N-methyl-norleucine ((N-Me)-Nle) and phenylalanine (Phe) with 1-naphtyl-alanine . With the new MG analog DOTA-dGlu-Ala-Tyr-Gly-Trp-(N-Me)Nle-Asp-1-Nal-NH 2 (DOTA-MGS5) radiolabeled with Gallium-68, Indium-111 and Lutetium-177 a very promising targeting profile was achieved [16]. In the present study we have conjugated the clinically well-established HYNIC ligand to MGS5 to develop a 99m Tc-labeled MG analog, suitable for SPECT and radioguided surgery. Furthermore the HYNIC-conjugate dGlu-Ala-Tyr-Gly-Trp-(N-Me)Nle-Asp-(N-Me)1-Nal-NH 2 (HYNIC-MGS11) with additional N-methylation of the peptide bond between Asp and 1-Nal was synthesized to evaluate if a further stabilizing effect can be achieved. The two 99m Tc-labeled MG analogs were characterized in vitro and in vivo, including receptor affinity and cell uptake assays, as well as metabolic and biodistribution studies in tumor-xenografted BALB/c nude mice.

Peptide Synthesis and Radiolabeling
Following straightforward SPPS HYNIC-MGS5 and HYNIC-MGS11 were synthesized using 30 µmol of resin, 150 µmol of each Fmoc-protected amino acid and 90 µmol of HYNIC. After purification by RP-HPLC and lyophilization the peptide conjugates were obtained in~10% yield with a chemical purity ≥95% as confirmed by RP-HPLC and MALDI-TOF MS. The amino acid sequences and chemical structures of both MG analogs are presented in Figure 1. HYNIC-conjugate dGlu-Ala-Tyr-Gly-Trp-(N-Me)Nle-Asp-(N-Me)1-Nal-NH2 (HYNIC-MGS11) with additional N-methylation of the peptide bond between Asp and 1-Nal was synthesized to evaluate if a further stabilizing effect can be achieved. The two 99m Tc-labeled MG analogs were characterized in vitro and in vivo, including receptor affinity and cell uptake assays, as well as metabolic and biodistribution studies in tumor-xenografted BALB/c nude mice.

Peptide Synthesis and Radiolabeling
Following straightforward SPPS HYNIC-MGS5 and HYNIC-MGS11 were synthesized using 30 μmol of resin, 150 μmol of each Fmoc-protected amino acid and 90 μmol of HYNIC. After purification by RP-HPLC and lyophilization the peptide conjugates were obtained in ~10% yield with a chemical purity ≥95% as confirmed by RP-HPLC and MALDI-TOF MS. The amino acid sequences and chemical structures of both MG analogs are presented in Figure 1.   [22,27]. The main peak occurring in the radio-HPLC chromatogram after labeling indicates complete conversion of the initial tricine complex into the EDDA complex, as already described previously [28,29]. Minor hydrophilic impurities, related to free pertechnetate and 99m Tc-coligands, could be efficiently removed by solid phase extraction (SPE) resulting in labeling with radiochemical purity >95%. Peptide related side products with relative retention of 0.9-1.1 were below 5%. Representative radiochromatograms are displayed in Figure 2.  Radiolabeling with technetium-99m using the exchange labeling approach from tricine, used as an intermediate coligand, to EDDA yielded in [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 at high molar activity of 35-40 GBq/µmol comparable to previously published results [22,27]. The main peak occurring in the radio-HPLC chromatogram after labeling indicates complete conversion of the initial tricine complex into the EDDA complex, as already described previously [28,29]. Minor hydrophilic impurities, related to free pertechnetate and 99m Tc-coligands, could be efficiently removed by solid phase extraction (SPE) resulting in labeling with radiochemical purity >95%. Peptide related side products with relative retention of 0.9-1.1 were below 5%. Representative radiochromatograms are displayed in Figure 2.

Characterization in Vitro
The stability of the two radiopeptides was analyzed in PBS, confirming a high complex stability obtained by exchange labeling. After 24 h incubation, still a high percentage of intact radiopeptide was present ([ 99m Tc]Tc-HYNIC-MGS5: 95.1%; [ 99m Tc]Tc-HYNIC-MGS11: 94.4%). MG analogs missing the penta-Glu motif are generally known for their low in vitro serum stability regardless of whether they are conjugated to DOTA [26] or HYNIC [30]. This can be changed by introducing site-specific amino acid substitutions into their C-terminal peptide sequence [15,16]. For [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 showing such substitutions, a high stability with no evidence of enzymatic degradation was found in human serum (>95% intact radiopeptide after 24 h incubation). These results are in accordance with previous results reported for 111 In-, 68 Ga-or 177 Lu-labeled DOTA-MGS5 (96-98% intact radiopeptide) incubated under the same conditions [16].

Characterization in Vitro
The stability of the two radiopeptides was analyzed in PBS, confirming a high complex stability obtained by exchange labeling. After 24 h incubation, still a high percentage of intact radiopeptide was present ([ 99m Tc]Tc-HYNIC-MGS5: 95.1%; [ 99m Tc]Tc-HYNIC-MGS11: 94.4%). MG analogs missing the penta-Glu motif are generally known for their low in vitro serum stability regardless of whether they are conjugated to DOTA [26] or HYNIC [30]. This can be changed by introducing site-specific amino acid substitutions into their C-terminal peptide sequence [15,16]. For [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 showing such substitutions, a high stability with no evidence of enzymatic degradation was found in human serum (>95% intact radiopeptide after 24 h incubation). These results are in accordance with previous results reported for 111 In-, 68 Ga-or 177 Lu-labeled DOTA-MGS5 (96-98% intact radiopeptide) incubated under the same conditions [16].

Characterization in Vitro
The stability of the two radiopeptides was analyzed in PBS, confirming a high complex stability obtained by exchange labeling. After 24 h incubation, still a high percentage of intact radiopeptide was present ([ 99m Tc]Tc-HYNIC-MGS5: 95.1%; [ 99m Tc]Tc-HYNIC-MGS11: 94.4%). MG analogs missing the penta-Glu motif are generally known for their low in vitro serum stability regardless of whether they are conjugated to DOTA [26] or HYNIC [30]. This can be changed by introducing site-specific amino acid substitutions into their C-terminal peptide sequence [15,16]. For [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 showing such substitutions, a high stability with no evidence of enzymatic degradation was found in human serum (>95% intact radiopeptide after 24 h incubation). These results are in accordance with previous results reported for 111 In-, 68 Ga-or 177 Lu-labeled DOTA-MGS5 (96-98% intact radiopeptide) incubated under the same conditions [16].

Receptor Binding and Cell Internalization Studies
Saturation binding experiments on A431-CCK2R cells revealed a high affinity to the human CCK2R. The mean values for the dissociation constant (K d ) calculated by fitting the data from two experiments to a one-site binding model (r 2 ≥ 0.99) were 13.7 ± 1.1 nM for [ 99m Tc]Tc-HYNIC-MGS5 and 14.7 ± 1.2 nM for [ 99m Tc]Tc-HYNIC-MGS11. In Figure 4 a representative saturation binding curve is displayed for both radioligands.

Receptor Binding and Cell Internalization Studies
Saturation binding experiments on A431-CCK2R cells revealed a high affinity to the human CCK2R. The mean values for the dissociation constant (Kd) calculated by fitting the data from two experiments to a one-site binding model (r² ≥ 0.99) were 13.7 ± 1.1 nM for [ 99m Tc]Tc-HYNIC-MGS5 and 14.7 ± 1.2 nM for [ 99m Tc]Tc-HYNIC-MGS11. In Figure 4 a representative saturation binding curve is displayed for both radioligands.  In the internalization assays performed on A431-CCK2R cells a high receptor mediated cell uptake was observed. With [ 99m Tc]Tc-HYNIC-MGS5 13.1 ± 0.4% of the totally added radioactivity was internalized already after 15 min and this value increased over time reaching 62.0 ± 1.6% after 2 h incubation. Even though [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 showed very similar Kd values, the cell uptake of [ 99m Tc]Tc-HYNIC-MGS11 was distinctly lower. After 15 min 4.0 ± 0.7% of the radioactivity was internalized and this value increased to 24.4 ± 1.9% after 2 h incubation. These unexpected differences between the two MG analogs indicate a possible different binding mode to human CCK2R of MGS5. Also for DOTA-conjugated MGS5 labeled with different radiometals a similarly enhanced cell uptake was found [16]. The results of the internalization experiments are displayed in Figure 5.  In the internalization assays performed on A431-CCK2R cells a high receptor mediated cell uptake was observed. With [ 99m Tc]Tc-HYNIC-MGS5 13.1 ± 0.4% of the totally added radioactivity was internalized already after 15 min and this value increased over time reaching 62.0 ± 1.6% after 2 h incubation. Even though [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 showed very similar K d values, the cell uptake of [ 99m Tc]Tc-HYNIC-MGS11 was distinctly lower. After 15 min 4.0 ± 0.7% of the radioactivity was internalized and this value increased to 24.4 ± 1.9% after 2 h incubation. These unexpected differences between the two MG analogs indicate a possible different binding mode to human CCK2R of MGS5. Also for DOTA-conjugated MGS5 labeled with different radiometals a similarly enhanced cell uptake was found [16]. The results of the internalization experiments are displayed in Figure 5.

Receptor Binding and Cell Internalization Studies
Saturation binding experiments on A431-CCK2R cells revealed a high affinity to the human CCK2R. The mean values for the dissociation constant (Kd) calculated by fitting the data from two experiments to a one-site binding model (r² ≥ 0.99) were 13.7 ± 1.1 nM for [ 99m Tc]Tc-HYNIC-MGS5 and 14.7 ± 1.2 nM for [ 99m Tc]Tc-HYNIC-MGS11. In Figure 4 a representative saturation binding curve is displayed for both radioligands.  In the internalization assays performed on A431-CCK2R cells a high receptor mediated cell uptake was observed. With [ 99m Tc]Tc-HYNIC-MGS5 13.1 ± 0.4% of the totally added radioactivity was internalized already after 15 min and this value increased over time reaching 62.0 ± 1.6% after 2 h incubation. Even though [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 showed very similar Kd values, the cell uptake of [ 99m Tc]Tc-HYNIC-MGS11 was distinctly lower. After 15 min 4.0 ± 0.7% of the radioactivity was internalized and this value increased to 24.4 ± 1.9% after 2 h incubation. These unexpected differences between the two MG analogs indicate a possible different binding mode to human CCK2R of MGS5. Also for DOTA-conjugated MGS5 labeled with different radiometals a similarly enhanced cell uptake was found [16]. The results of the internalization experiments are displayed in Figure 5.  In A431-mock cells lacking CCK2R expression a low and comparable non-specific uptake of radioactivity (≤0.1%) was observed for both radiopeptides at each time point. The receptor specificity of the uptake in A431-CCK2R cells could be confirmed by additional blocking studies with 1 µM pentagastrin showing a clear blockage of the CCK2R mediated uptake to values ≤1.0% after 2 h incubation for both radiopeptides (data not shown).

In Vivo Stability in BALB/c Mice
Metabolic studies in BALB/c mice were performed to further evaluate the enzymatic stability of the radiopeptides. In these studies differences in the stability against enzymatic degradation could be observed between the two radiopeptides. For [ 99m Tc]Tc-HYNIC-MGS5 >65% intact radiopeptide was found in the blood and liver of BALB/c mice at 10 min after injection. Interestingly, the stability against metabolic degradation was further increased for [ 99m Tc]Tc-HYNIC-MGS11. The values of intact radiopeptide at 10 min p.i. were 96.0 ± 1.4% in blood and 95.2 ± 0.7% in liver. A faster degradation was observed in the excretory system. Much lower levels of intact radiopeptide were detectable in kidneys ([ 99m Tc]Tc-HYNIC-MGS5: 13.9 ± 0.5%; [ 99m Tc]Tc-HYNIC-MGS11: 40.0 ± 2.0%) and urine ([ 99m Tc]Tc-HYNIC-MGS5: 3.0 ± 1.6%; [ 99m Tc]Tc-HYNIC-MGS11: 9.8 ± 2.2%). Peptide related impurities already present after radiolabeling and with a relative retention of 0.9-1.1 were not considered as metabolites formed during digestion. Representative radio-HPLC profiles of the different analyzed samples are presented in Figure 6. For [ 99m Tc]Tc-HYNIC-MGS5 two main metabolites were observed in blood and liver, an early eluting hydrophilic metabolite with retention time (t R ) of~3.5 min and a second more hydrophobic metabolite eluting at t R~1  In A431-mock cells lacking CCK2R expression a low and comparable non-specific uptake of radioactivity (≤0.1%) was observed for both radiopeptides at each time point. The receptor specificity of the uptake in A431-CCK2R cells could be confirmed by additional blocking studies with 1 μM pentagastrin showing a clear blockage of the CCK2R mediated uptake to values ≤1.0% after 2 h incubation for both radiopeptides (data not shown).

In Vivo Stability in BALB/c Mice
Metabolic studies in BALB/c mice were performed to further evaluate the enzymatic stability of the radiopeptides. In these studies differences in the stability against enzymatic degradation could be observed between the two radiopeptides. For [ 99m Tc]Tc-HYNIC-MGS5 >65% intact radiopeptide was found in the blood and liver of BALB/c mice at 10 min after injection. Interestingly, the stability against metabolic degradation was further increased for   Nle had a stabilizing effect and only cleavage between Nle and Asp occured [18]. In kidney and urine a much higher degree of enzymatic degradation was observed for [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11. In kidneys mainly the hydrophilic metabolite with t R~3 .5 min not corresponding to pertechnetate eluting at t R 2.6 min was observed, whereas in urine a variety of additional metabolites with t R < 15 min was detected.

Biodistribution in Tumor-xenografted BALB/c Nude Mice
The results of the biodistribution studies in the A431-CCK2R/A431-mock tumor-xenograft model at 1 h and 4 h p.i. evaluating the tumor targeting and tissue uptake are displayed in Figure 7a. For selected organs additional autoradiography studies were performed at 1 h p.i. (see Figure 7b). The uptake values calculated for the tumors and different dissected tissues are summarized in the supplementary material (Table S1). Rapid clearance from the body mainly through the kidneys resulted in low non-specific uptake in most tissues and organs for [ 99m Tc]Tc-HYNIC-MGS5 as well as for [ 99m Tc]Tc-HYNIC-MGS11. The uptake of both radiopeptides significantly decreased in most tissues from 1 h to 4 h p.i. (p < 0.05) except for stomach (p = 0.07), intestine (p = 0.14) and bone (p = 0.18) in mice injected with [ 99m Tc]Tc-HYNIC-MGS5 as well as pancreas (p = 0.12) and kidneys (p = 0.18) in mice injected with [ 99m Tc]Tc-HYNIC-MGS11. [ 99m Tc]Tc-HYNIC-MGS5 showed a significantly lower uptake in blood at 1 h in comparison with [ 99m Tc]Tc-HYNIC-MGS11 (p < 0.01), however both radiopeptides showed a similar hydrophobicity and protein binding. At 4 h p.i. a similar trend was observed in blood, but this was not significant (p = 0.06). Also the non-specific uptake in lung, heart, spleen and liver was higher for [ 99m Tc]Tc-HYNIC-MGS11 at 1 h and 4 h p.i. (p < 0.05). In mouse CCK2R are primarily localized in brain and stomach and at lower expression levels also in colon, pancreas, kidney and ovary [31]. Due to the overall negative charge the two radioligands are unable to cross the blood-brain barrier, but we found a considerable uptake in stomach and pancreas, which was significantly higher for [ 99m Tc]Tc-HYNIC-MGS5 at both time points. For [ 99m Tc]Tc-HYNIC-MGS5 at 4 h p.i. a stomach uptake of 12.89 ± 2.91% IA/g and a pancreas uptake of 6.64 ± 2.21% IA/g was found, whereas [ 99m Tc]Tc-HYNIC-MGS11 displayed much lower values of 3.95 ± 0.15% IA/g in stomach (p = 0.0009) and 1.30 ± 0.42% IA/g in pancreas (p = 0.003) at the same time point. Also the intestinal uptake of [ 99m Tc]Tc-HYNIC-MGS11 with values of 0.82 ± 0.14% IA/g versus 1.39 ± 0.34% IA/g for [ 99m Tc]Tc-HYNIC-MGS5 at 4 h p.i. was significantly lower (p = 0.02). In line with the high CCK2R expression level confirmed for A431-CCK2R cells [32], an impressively high CCK2R mediated tumor uptake was observed in A431-CCK2R xenografts, whereas the uptake in A431-mock xenografts remained at very low levels (<1% IA/g). Tumor weights as determined after sacrifice were 0.17 ± 0.06 g for A431-CCK2R xenografts (n = 16) and 0.16 ± 0.12 g for A431-mock xenografts (n = 15). One mouse did not develop the A431-mock tumor. In A431-CCK2R xenografts, [ 99m Tc]Tc-HYNIC-MGS5 showed a high and persistent uptake with values of 25.09 ± 2.39% IA/g at 1 h and 24.75 ± 4.38% IA/g at 4 h p.i. The uptake values of [ 99m Tc]Tc-HYNIC-MGS11 were almost doubled (39.87 ± 7.12% IA/g at 1 h and 42.48 ± 6.99% IA/g at 4 h p.i.; p < 0.01). This improvement was rather surprising given the lower cell uptake observed in vitro (24.4 ± 1.9% for [ 99m Tc]Tc-HYNIC-MGS11 versus 62.0 ± 1.6% for [ 99m Tc]Tc-HYNIC-MGS5 after 2 h incubation). Receptor specificity was only tested by blocking studies in vitro, however, the very low uptake values observed in A431-mock tumor-xenografts (0.17-0.93% IA/g) for both radioligands confirmed that the tumor uptake was highly receptor specific. A very low and similar peptide amount of~15 pmol was injected to the animals studied with both radiopeptides to avoid possible receptor saturating effects. The divergent uptake values observed for mouse stomach and pancreas in comparison with A431-CCK2R tumor-xenografts observed with both radioligands are therefore rather related to interspecies receptor differences between mouse and human CCK2R, than varying receptor expression levels. Even though the mouse and human CCK2R share an amino acid identity of~90% [31], methylation of the peptide bond between Asp and 1-Nal in [ 99m Tc]Tc-HYNIC-MGS11 seems to affect the uptake into CCK2R-expressing organs in mouse. Interspecies differences between human and rat CCK2R have been reported also for other radiolabeled MG analogs [15,33]. The observed differences in stomach and pancreas need to be interpreted with caution as a different physiological uptake may occur in humans. However, no further experiments were carried out to investigate the affinity of the two MG analogs for mouse CCK2R.
Pharmaceuticals 2018, 11, x 8 of 16 seems to affect the uptake into CCK2R-expressing organs in mouse. Interspecies differences between human and rat CCK2R have been reported also for other radiolabeled MG analogs [15,33]. The observed differences in stomach and pancreas need to be interpreted with caution as a different physiological uptake may occur in humans. However, no further experiments were carried out to investigate the affinity of the two MG analogs for mouse CCK2R.   The structural difference of the two radioligands also had a strong impact on the kidney uptake. The kidney uptake of [ 99m Tc]Tc-HYNIC-MGS11 (19.90 ± 2.09% IA/g at 1 h and 17.17 ± 2.93% IA/g at 4 h p.i.) was two times higher in comparison with [ 99m Tc]Tc-HYNIC-MGS5 (10.56 ± 1.15% IA/g at 1 h and 7.80 ± 1.47% IA/g at 4 h p.i.). Despite CCK2R-related uptake, reabsorption and retention of radiolabeled peptides in kidneys is mainly driven by multiple transport mechanisms, involving megalin and cubulin [34]. In megalin-deficient mice a significantly reduced renal reabsorption was confirmed for a radiolabeled minigastrin analog, with renal uptake values reduced to 37-49% when compared to wild-type mice [35]. It has been shown for different radiolabeled MG analogs that the uptake in stomach and CCK2R-expressing tumor-xenografts can be efficiently blocked by co-injection of a 1000-fold molar excess of unlabeled peptide, whereas no considerable effect occurs in kidneys [36]. In this study no additional blocking studies were performed in vivo to confirm the receptor-specific uptake in the different organs.
The tumor targeting profile of [ 99m Tc]Tc-HYNIC-MGS5 well compares with previous results obtained with [ 111 In]In-DOTA-MGS5, showing a similar tumor uptake of 19.53 ± 5.42% IA/g and 23.49 ± 1.25% IA/g at 1 h and 4 h p.i., respectively [15]. This high and persistent tumor uptake is clearly superior when compared to other 99m Tc-labeled MG analogs previously studied [37]. For [ 99m Tc]Tc-Demogastrin-2 evaluated in A431-CCK2R xenografted SCID mice at a similar injected peptide amount of 10 pmol a tumor uptake of 12.89 ± 4.69% ID/g at 4 h p.i. was reported, which was however, connected with a very high kidney uptake (58.62 ± 8.98% ID/g). The metabolic stability of this compound (60% intact radiopeptide at 5 min p.i.) was comparable to [ 99m Tc]Tc-HYNIC-MGS5 [37]. In our previous studies with radiolabeled DOTA-MGS5 we concluded that a combination of increased protein binding and stabilization against enzymatic degradation might be responsible for the highly improved targeting profile. The enhanced protection against metabolic degradation of [ 99m Tc]Tc-HYNIC-MGS11 led to a further doubling in tumor uptake. Such an improvement could not be achieved by enzymatic stabilization alone, as exemplified by the co-injection of protease inhibitors [37]. [ 99m Tc]Tc-Demogastrin-2 coinjected with 300 µg phosphoramidon showed a similar enzymatic stability in vivo (85% intact radiopeptide at 5 min p.i.) as compared to [ 99m Tc]Tc-HYNIC-MGS11, but the tumor uptake was clearly inferior (18.21 ± 5.97% ID/g at 4 h p.i.). When comparing the tumor-to-organ activity ratios of [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11, somewhat lower tumor-to-blood ratios were observed for [ 99m Tc]Tc-HYNIC-MGS11. Thus, [ 99m Tc]Tc-HYNIC-MGS11 also showed increased non-specific uptake in most organs at both investigated time points. Due to the concomitant increase of the uptake in A431-CCK2R xenografts and kidneys observed for [ 99m Tc]Tc-HYNIC-MGS11, a similar tumor-to-kidney ratio was found for both radiopeptides (2.4-3.3 for [ 99m Tc]Tc-HYNIC-MGS5 and 2.0-2.6 for [ 99m Tc]Tc-HYNIC-MGS11). The respective tumor-to-organ activity ratios calculated for blood, kidney, intestine, pancreas and stomach for both radioligands at different time points are displayed in Table 1. To our knowledge this is the first report on 99m Tc-labeled CCK2R targeting peptide analogs showing such an astonishingly improved tumor uptake along with clearly reduced kidney uptake. [ 99m Tc]Tc-HYNIC-MGS5 showed a very similar targeting profile when compared to DOTA-MGS5 radiolabeled with different radiometals suitable for SPECT, PET and TRT [15]. The tumor uptake of [ 99m Tc]Tc-HYNIC-MGS11 was further improved, however, connected with a concomitant increase in kidney uptake. Recently, two 99m Tc-labeled non-peptidic radioligands have been described showing high tumor uptake in a mouse tumor model based on human epithelial cells transfected with CCK2R, however, tumor-to-kidney ratio was clearly inferior [38,39].
These developments give high promise that in the near future a kit for 99m Tc-labeling will be available allowing the localization and staging of CCK2R expressing tumors. Gastrin receptor scintigraphy might show a lower sensitivity when compared to PET imaging with a 68 Ga-labeled CCK2R targeting MG analog, but additionally allows for radioguided surgery. The concept of surgical guidance with conventional gamma probes for intraoperative identification and removal of metastatic lesions has already been successfully introduced into clinical practice by the use of 99m Tc-labeled somatostatin analogs in patients with neuroendocrine tumors [40] as well as 99m Tc-labeled ligands targeting prostate specific membrane antigen in patients with prostate cancer [41]. Due to its high tumor uptake and tumor retention combined with high clearance from other tissues [ 99m Tc]Tc-HYNIC-MGS11 might be favorable for imaging and radioguided surgery.

Materials
All commercially obtained chemicals were of analytical grade and used without further purification. Na[ 99m Tc]TcO 4 was obtained from a commercial 99 Mo/ 99m Tc-generator (Ultratechnekow, Mallinckrodt, Petten, The Netherlands) eluted with physiological saline. The A431 human epidermoid carcinoma cell line stably transfected with the plasmid pCR3.1 containing the full coding sequence for the human CCK2R (A431-CCK2R) as well as the same cell line transfected with the empty vector alone (A431-mock) were kindly provided by Dr. Luigi Aloj [42]. Both cell lines were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% (v/v) fetal bovine serum and 5 mL of a 100× penicillin-streptomycin-glutamine mixture at 37 • C in a humidified 95% air/5% CO 2 atmosphere. Media and supplements were purchased from Invitrogen Corporation (Lofer, Austria).

Analytical Systems and Methods
For preparative HPLC purification a Gilson 322 chromatography system (Gilson International, Limburg, Germany) with Gilson UV/VIS-155 multi-wavelength detector, equipped with an Eurosil Bioselect Vertex Plus C18A precolumn (300 Å, 5 µm, 30 × 8 mm) and a Eurosil Bioselect Vertex Plus C18A column (300 Å 5 µm 300 × 8 mm) (Knauer, Berlin, Germany) was used with a gradient system starting from 80% solvent A (water containing 0.1% TFA) and  For in vivo assays the radiolabeled peptides were purified by SPE. For this purpose, the labeling mixture was passed through a C18-SepPak-Light cartridge (Waters, Milford, MA, USA), followed by 5 mL saline, and the radiolabeled peptide was eluted with EtOH/H 2 O 65/35 v/v and diluted with PBS.

Evaluation of the in Vitro Properties
The resistance against degradation of [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 (1000 pmol/mL, n = 2) in human serum was studied for up to 24 h. Furthermore, the radiopeptides were incubated in PBS (1000 pmol/mL, n = 1). At each time point of 1, 2, 4 and 24 h after incubation the intact radiopeptide was assessed by analytical HPLC. Serum samples were precipitated with ACN and centrifuged to collect the supernatant and diluted with water prior to radio-HPLC. For the determination of the distribution coefficient (log D), 500 µL of the radiopeptide solutions (50 pmol/mL in PBS) were added to 500 µL octanol (1:1) vigorously vortexed (n = 8) for 15 min and centrifuged to separate the two phases. From each phase a 75 µL sample was taken, the radioactivity measured in a 2480 Wizard 2 automatic gamma-counter (PerkinElmer Life Sciences and Analytical Sciences, Wallac Oy, Turku, Finland) and the distribution of the radiopeptides calculated. The protein binding in human serum (500 pmol/mL, n = 2) was assessed by Sephadex G-50 size-exclusion chromatography (GE Healthcare Illustra, Little Chalfont, UK) for up to 24 h.

Receptor Binding and Cell Internalization Studies
The receptor affinity of the radioligands prepared using the above described labeling protocol was evaluated in saturation studies on A431-CCK2R cells. For the assay, 96-well filter plates (MultiScreen HTS -FB, Merck Group, Darmstadt, Germany) were pretreated with 10 mM TRIS/139 mM NaCl buffer, pH 7.4 (TRIS-buffer) (2 × 250 µL) and 400,000 A431-CCK2R cells per well were added in 35 mM HEPES buffer, pH 7.4, containing 10 mM MgCl 2 , 14 µM bacitracin, and 0.5% bovine serum albumin (BSA), a hypotonic solution disturbing the integrity of the cell membranes. Thereafter, increasing concentrations of the radiolabeled peptide conjugates (0.1-112 nM) were added in triplicate reaching a total volume of 200 µL. In parallel, non-specific binding was determined by co-incubation with 1 µM pentagastrin. After 1 h incubation at room temperature, the medium was removed by filtration followed by two rapid rinses with ice-cold TRIS buffer (200 µL). The filters were collected and counted in a gamma-counter. The K d value was calculated fitting the data with Origin software (Origin 6.1, OriginLab Corporation, Northampton, MA, USA) to a one-site binding model using the formula y = B max × x/(K d + x).
For internalization experiments, A431-CCK2R and A431-mock cells were seeded at a density of 1.0 × 10 6 cells per well in 6-well plates and grown to confluence for 48 h. At the day of the experiment, cells were washed twice with ice-cold internalization medium supplemented with 1% (v/v) fetal bovine serum and supplied with fresh medium before incubation with [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 at a final peptide concentration of 0.4 nM in a total volume of 1.5 mL in triplicates. At different time points for up to 2 h incubation the cell uptake was interrupted by removal of the medium and rapid rinsing with ice-cold internalization medium (two times). Thereafter, the cells were incubated twice at ambient temperature in acid wash buffer (50 mM glycine buffer pH 2.8, 0.1 M NaCl) for 5 min, to remove the membrane-bound radioligand. Finally, the cells were lyzed by treatment in 1 M NaOH and collected (internalized radioligand fraction). All collected fractions (supernatant, surface wash, lyzed cells) were measured together with a standard in the gamma counter. The radioactivity of the lyzed cells was expressed as percentage of the total radioactivity added (% of internalized radioactivity). Non-specific binding was evaluated in A431-mock cells and in additional blocking studies with 1 µM pentagastrin.

Evaluation of the in Vivo Stability and Biodistribution
All animal experiments were conducted in compliance with the Austrian animal protection laws and with the approval of the Austrian Ministry of Science (BMWFW-66.011/0075-WF/V/3b/2016).

Metabolic Stability in BALB/c Mice
Metabolic stability studies in vivo with [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 were performed in 5-6 week-old female BALB/c mice (Charles River, Sulzfeld, Germany; n = 2). Mice were injected intravenously via a lateral tail vain with 37-74 MBq of the 99m Tc-labeled HYNIC-analogs (corresponding to 2 nmol total peptide). Ten minutes post injection (p.i.) mice were euthanized and a sample of blood and urine was collected together with the liver and kidneys. Liver and kidneys were rapidly homogenized in 0.5 mL of a 20 mM HEPES buffer pH 7.3 with an Ultra-Turrax T8 homogenator (IKA-Werke, Staufen, Germany) for 1 min at RT. Before determining the percentage of intact radiopeptide by analytical radio-HPLC samples were precipitated with ACN, centrifuged and diluted with H 2 O (1:1).

Biodistribution in Tumor-xenografted BALB/c Nude Mice
Biodistribution studies evaluating the tumor uptake of [ 99m Tc]Tc-HYNIC-MGS5 and [ 99m Tc]Tc-HYNIC-MGS11 were performed in 7 week-old female athymic BALB/c nude mice (Charles River, Sulzfeld, Germany). To induce tumor-xenografts, mice were injected subcutaneously with 2 × 10 6 A431-CCK2R (right flank) and A431-mock cells (left flank). The tumor-xenografts were allowed to grow for 11-12 days reaching medium tumor weights of~0.2 g. Mice were randomly divided into groups of four and injected intravenously via a lateral tail vein with 0.3 MBq of the 99m Tc-labeled HYNIC-analogs (corresponding to~15 pmol total peptide). The groups of animals were sacrificed at 1 h and 4 h p.i., tumors and other tissues (blood, lung, heart, muscle, bone, spleen, intestine, liver, kidney, stomach and pancreas) were removed, weighed, and the radioactivity measured together with a standard in the gamma counter. Results were expressed as percentage of injected activity per gram tissue (% IA/g) and tumor-to-organ activity ratios were calculated for selected tissues. Statistical analysis was performed using independent two population t-test (significance level p = 0.05) with Origin software. The radioactivity in selected organs (stomach, pancreas, liver, kidneys, A431-CCK2R and A431-mock xenograft) was additionally visualized by autoradiography using a phosphorimager (Cyclone Plus, PerkinElmer Life Sciences and Analytical Sciences, Downers Grove, Il, USA). After dissection, the organs were exposed to a multisensitive storage phosphor screen for 40 min and image analysis was performed using OptiQuant TM software (OptiQuant 5.0, PerkinElmer Life Sciences and Analytical Sciences, Downers Grove, Il, USA). Funding: This study was funded by the Austrian Science Foundation (FWF) grant P 27844.