Article pubs.acs.org/jmc
Gastrin Releasing Peptide Receptor-Directed Radioligands Based on a Bombesin Antagonist: Synthesis, 111In-Labeling, and Preclinical Profile Panteleimon J. Marsouvanidis,†,‡ Berthold A. Nock,*,† Bouchra Hajjaj,§ Jean-Alain Fehrentz,§ Luc Brunel,§ Céline M’Kadmi,§ Linda van der Graaf,‡ Eric P. Krenning,‡ Theodosia Maina,† Jean Martinez,§ and Marion de Jong‡ †
Molecular Radiopharmacy, INRASTES, National Center for Scientific Research “Demokritos”, Ag. Paraskevi Attikis, GR-153 10 Athens, Greece § IBMM, UMR 5247, Universités Montpellier 1 & 2, CNRS, Faculté de Pharmacie, 34093 Montpellier Cedex 5, France ‡ Department of Nuclear Medicine, Erasmus MC, 3015 GD, Rotterdam, The Netherlands ABSTRACT: Novel bombesin (BBN) antagonists were synthesized by coupling the chelator 1,4,7,10-tetraazacyclododecane-1,4,7,10tetraacetic acid (DOTA) to H-D-Phe-Gln-Trp-Ala-Val-Gly-His-StaLeu-NH2 (JMV594) through linkers of increasing number of (βAla)x residues (x = 1−3). Labeling with 111In afforded the respective radiotracers in high purity and high specific activity. Bioconjugate affinity for the gastrin releasing peptide receptor (GRPR) as determined against [125I-Tyr4]BBN was high (IC50 values in the lower nanomolar range). Radioligands poorly internalized in PC-3 cells at 37 °C. Radiopeptides remained >60% intact 5 min after entering the bloodstream of healthy mice. After injection in SCID mice bearing human PC-3 xenografts all analogues showed high tumor uptake and rapid background clearance via the kidneys into urine. Interestingly, pancreatic uptake, albeit GRPR-specific, declined rapidly with time. 111In-DOTA-(βAla)2-JMV594 achieved the highest tumor values among the group (17.0 ± 2.8%ID/g vs. 8−10%ID/g, respectively, at 4 h pi) indicating that the (βAla)2linker favors in vivo interaction of radiopeptides with the GRPR.
■
INTRODUCTION Among the three bombesin receptors identified in mammals,1 the gastrin releasing peptide receptor (GRPR) is nowadays regarded as a major molecular target for radiolabeled bombesin (BBN) analogues owing to its abundant expression in many frequently occurring human cancers, including prostate cancer, mammary carcinoma, or lung cancer.2−7 As a result, several radiolabeled analogues of the amphibian tetradecapeptide BBN and its C-terminal octapeptide fragment BBN(7−14) have been developed by coupling suitable chelators to their Nterminus to allow for labeling with radiometals used in SPECT/ PET imaging or in radionuclide therapy. The resulting radiopeptides rapidly internalize into target cells after binding to the GRPR and have shown high accumulation in human GRPR-expressing xenografts in immunosuppressed mice.8−10 Hence, until recently the consensus has been to develop agonist-based radioligands, which would effectively accumulate into malignant lesions by virtue of their fast in vivo internalization into cancer cells, a prerequisite widely considered essential for optimal diagnostic imaging and radionuclide therapy of tumors. Translation of preclinical results into the clinic, however, revealed biosafety hazards linked to the intravenous (iv) © 2013 American Chemical Society
administration of potent BBN agonist-based (radio)peptides to patients, even at low doses. Most frequently observed adverse reactions were associated with abdominal smooth muscle contraction, stimulation of gastrointestinal hormone secretion, and thermoregulation, elicited by activation of bombesin receptors, which are physiologically expressed in peripheral tissues and especially in the gut.3,9,11−13 These findings had a negative impact on the clinical application of a few promising BBN agonist-based radioligands as well as on their commercialization prospects. Receptor-targeted radiotherapy, in particular, is often associated with more severe side effects, because a higher peptide amount needs to be injected into patients to reach an effective tumoricidal dose of the therapeutic radionuclide.14,15 Biosafety problems may be averted with the advent of radioligands based on GRPR-antagonists instead of GRPRagonists. In addition, the proliferative action of BBN propagated via the GRPR on cancer cells16−19 strengthens the rationale to use GRPR-antagonists as anticancer drugs. For this purpose, several BBN-receptor antagonists successfully Received: November 18, 2012 Published: February 22, 2013 2374
dx.doi.org/10.1021/jm301692p | J. Med. Chem. 2013, 56, 2374−2384
Journal of Medicinal Chemistry
Article
Figure 1. Synthesis of DOTA-(βAla)x-JMV594 conjugates, 1 (JMV4705, x = 1); 2 (JMV4168, x = 2); and 3 (JMV4706, x = 3). Fragments Ax′ and B were synthesized separately and then condensed in solution; after coupling of the chelator DOTA and removal of side chain protecting groups peptide analogues were isolated by chromatographic methods and lyophilized.
internalize 99mTc-Demobesin 1 displayed unprecedentedly high and GRPR-specific accumulation in human prostate cancer PC-3 xenografts in mice and a rapid background clearance, even from the strongly GRPR-positive mouse pancreas, via the kidneys into urine. This pharmacokinetic profile in mice was found to be superior to that of agonist-based 99m Tc-Demobesin 4 ([99mTc−N40,Pro1,Tyr4,Nle14]BBN) in a direct comparison study.25 Most importantly, the favorable
inhibiting the growth of human xenografts in nude mice have been developed.20−23 The advantages of exploiting such receptor antagonists as molecular vehicles of radionuclides to GRPR-expressing cancer were first evident with our analogue 99m Tc-Demobesin 1.24 This radiotracer was generated by coupling an acyclic tetraamine chelator to D-Phe6 of the potent GRPR-antagonist [H-D-Phe6,LeuNHEt13,desMet14]BBN(6−14) followed by 99mTc-labeling. Despite its poor ability to 2375
dx.doi.org/10.1021/jm301692p | J. Med. Chem. 2013, 56, 2374−2384
Journal of Medicinal Chemistry
Article
Table 1. Analytical Data for 1, 2, 3 HPLC tR (min), UV trace compound
sequence
MW calculated/found (m/z)
system 1b
system 2c
1 2 3
DOTA-(βAla)-JMV594a DOTA-(βAla)2-JMV594 DOTA-(βAla)3-JMV594
1570.83/[M + 2H+]/2 = 786.18 1641.90/[M + 2H+]/2 = 821.5 1712.98/[M + 2H+]/2 = 857.2
1.35 1.37 1.37
16.3 17.2 17.3
JMV594: D-Phe-Gln-Trp-Ala-Val-Gly-His-Sta-Leu-NH2. bSystem 1: RP-HPLC with UV detection; Column, Chromatolith Speed ROD (50 mm × 4.6 mm); solvents, A = 0.1% TFA(aq) (v/v) and B = 0.1% TFA (v/v) in MeCN; linear gradient, 0% B to 100% B in 4 min; flow rate, 5 mL/min. c System 2: RP-HPLC with UV detection; Column, C18 XSelect CSH (5 μm, 4.6 mm × 150 mm); solvents, A = 0.1% TFA in H2O (v/v) and B = MeCN; linear gradient, 20% B to 25% B in 10 min followed by isocratic elution; flow rate, 1 mL/min. a
clearance and excellent tolerability of 99mTc-Demobesin 1 were reproduced in a small number of prostate cancer patients.26 It is interesting to note that radiolabeled somatostatin receptor antagonists have likewise shown high lesion uptake and fast background clearance, even from tissues physiologically expressing somatostatin receptors.27 Furthermore, results originally acquired with 99mTc-Demobesin 1 were soon corroborated by data obtained by a broader spectrum of BBN antagonist-based radioligands modified with similar or different chelators and labeled with a wide range of radiometals.28,29 Nevertheless, the exact molecular mechanism(s) shaping the superior in vivo profile of both bombesin and somatostatin receptor antagonists as compared to agonists have not been completely understood yet, although binding of antagonists to a higher number of receptors expressed on target cells has been suspected. Several classes of potent bombesin receptor antagonists have been developed by structural interventions in the C-terminal Leu13-Met14NH2 dipeptide fragment of frog BBN,20 as for example development of desmethionine Leu13-alkylamide analogues (as in 99mTc-Demobesin 1),30 by replacement of the peptide bond between the two last amino acid residues by a pseudopeptide bond,31 or by replacement of Leu13 by a γ amino acid statyl residue (Sta = statine, (3S,4S)4-amino-3-hydroxy-6methylheptanoic acid).32 Several of these peptides were able to recognize BBN binding sites on rat pancreatic acini in vitro and to antagonize BBN stimulated amylase secretion in the nanomolar range, while in vivo they were shown to inhibit the BBN-induced pancreatic enzyme secretion in rodents. Among these analogues the linear nonapeptide JMV594 (H-DPhe-Gln-Trp-Ala-Val-Gly-His-Sta-Leu-NH2) has attracted our attention. JMV594 was previously developed by us from the frog BBN(6−14) sequence after replacement of the C-terminal dipeptide fragment by Sta13-Leu14NH2 and substitution of Asn6 by D-Phe6 and has shown high binding affinity and potent antagonistic properties at the GRPR.32 In this study we have coupled the universal chelator DOTA (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid) to the N-terminus of JMV594 via a series of linkers containing an increasing number of βAla-residues, thereby generating the following DOTA-(βAla) x -JMV594 peptide-conjugates: JMV4705, x = 1; JMV4168, x = 2; and JMV4706, x = 3 (Figure 1). Coupling of DOTA will allow labeling with a wide range of diagnostic (PET: 68Ga; and SPECT: 111In, 67Ga) and therapeutic (177Lu, 90Y, 213Bi) radiometals. On the other hand, linker length variations will demonstrate the metal-chelate position leading to radiopeptides of improved biological features, especially of higher tumor uptake.33 After labeling with 111In BBN analogues are directly compared in GRPRpositive in vitro models and in SCID mice bearing human prostate cancer PC-3 xenografts as potential candidates for
application in the diagnostic imaging of GRPR-expressing tumors in man. Furthermore, 111In-radiotracers may serve as surrogates during early analogue-screening approaches for the development of the respective 177Lu/90Y/213Bi-radioligands intended for therapeutic applications.
■
RESULTS Synthesis of DOTA-(βAla)x-JMV594 Conjugates. We chose a fragment condensation approach for the synthesis of the conjugates and combined solid phase peptide synthesis (SPPS) and synthesis in solution. In fact, solid phase synthesis of peptides containing a statine residue is challenging, as acylation of the hydroxyl side chain of statine can occur during elongation of the peptide with large excess of activated Nαprotected amino acids. Furthermore, epimerization during fragment condensation cannot occur in this case, because fragment A is terminated with a glycine residue. First, fragments Ax′ (Fmoc-(βAla)x-D-Phe-(Trt)Gln-(Boc)Trp-Ala-Val-GlyOH; x = 1−3) were synthesized on the solid support and fragment B′ (Fmoc-His(Trt)-Sta-Leu-NH2) in solution, as shown in Figure 1. Fragment Ax′ (x = 1−3) was released from the 2-chlorotrityl chloride resin and the Fmoc-group was removed from the N-terminus of fragment B′ with diethylamine in DMF. The two fragments were condensed in DMF using BOP in the presence of DIEA as coupling agent to produce the respective constructs Ax′-B′ (Fmoc-(βAla)x-D-Phe(Trt)Gln-(Boc)Trp-Ala-Val-Gly-His(Trt)-Sta-Leu-NH2; x = 1−3); the Fmoc groups were removed from the Nα-amino acids by diethylamine in DMF. At the final stage, coupling of the DOTA-tris(tBut)ester chelator to the Nα-amino acid of individual Ax′-B construct was achieved in high yields using HATU as a coupling agent. After treatment with TFA to remove side chain protecting groups, the conjugates DOTA(βAla)x-JMV594: 1 (JMV4705, x = 1); 2 (JMV4168, x = 2); and 3 (JMV4706, x = 3) were isolated by preparative HPLC and lyophilized (Figure 1). BBN analogues were obtained with an average yield of ∼40% and a >97% purity as confirmed by RP-HPLC; ES/MS spectra were in accordance with the expected formulas (Table 1). Labeling with 111In. Incorporation of 111In by the DOTA framework was achieved by heating a mixture of each peptideconjugate and 111InCl3 in acidic medium for 20 min at 90 °C (>95% yield).34 As demonstrated by RP-HPLC analysis of the labeling reaction mixtures, all three 111In-radiotracers eluted as a single radiochemical species (>95% purity), and typical specific activities of ≈4−7 MBq/nmol. Elution times (tR) determined by HPLC (system 3, column III) for the three radiopeptides were [111In]1 ([111In]JMV4705), 18.2 min; [111In]2 ([111In]JMV4168), 19.6 min; [111In]3 ([111In]JMV4706), 20.0 min. A representative radiochromatogram for [111In]2 is shown in Figure 2. 2376
dx.doi.org/10.1021/jm301692p | J. Med. Chem. 2013, 56, 2374−2384
Journal of Medicinal Chemistry
Article
Figure 4. Intracellular Ca2+ mobilization induced in GRPR-positive PC-3 cells by 1 h incubation at 37 °C with increasing concentrations of BBN (Δ) BBN (EC50 = 0.73 ± 0.14 nM); Ca2+ release inhibition curves by increasing concentrations of (■) 1 (EC50 = 11.5 ± 1.6 nM), (◆) 2 (EC50 = 130 ± 34 nM), and (▼) 3 (EC50 = 137 ± 16 nM) in the presence of 1 nM BBN; effects induced by BBN at 10 μM were set as maximum.
Figure 2. RP-HPLC analysis of labeling reaction mixture of [111In]2 with chemical structure included (system 3, column III).
Affinity of Peptide Conjugates to Human GRPR. Binding affinities of peptide conjugates for the human GRPR were determined by competition binding assays at 4 °C in prostate adenocarcinoma PC-3 cells35 against [125I-Tyr4]BBN. As shown in Figure 3, all analogues displaced the radioligand
Binding of 111In-Radiopeptides to PC-3 Cells. During 1 h incubation at 37 °C, all three 111In-radiopeptides were bound to PC-3 cells adhered as confluent monolayers in six-well plates (36% of total added activity was associated to PC-3 cells for [111In]2 and ≈27% for [111In]1 and [111In]3; ≈1.5 × 106 cells/ well were measured). Only 6% ([111In]2) and 4.5% ([111In]1 and [111In]3) of total-added activity internalized into the cells and the rest of radioactivity was bound to the cell-membrane, as observed also for other BBN-receptor radioantagonists.25,28 Total association to cells dropped to base levels (
