Development of Novel 68Ga- and 18F-Labeled GnRH-I Analogues

1256

Bioconjugate Chem. 2008, 19, 1256–1268

Development of Novel 68Ga- and GnRHR-Targeting Efficiency

18

F-Labeled GnRH-I Analogues with High

Margret Schottelius,* Sebastian Berger, Thorsten Poethko, Markus Schwaiger, and Hans-Jürgen Wester Nuklearmedizinische Klinik and Poliklinik, Klinikum rechts der Isar, Technische Universita¨t Mu¨nchen, 81675 Mu¨nchen, Germany. Received February 12, 2008; Revised Manuscript Received April 28, 2008

A large majority of tumors of the reproductive system express the gonadotropin releasing hormone receptor (GnRHR). Blockade and activation of this receptor with various antagonistic and agonistic analogues of native GnRH-I (pGlu1-His2-Trp3-Ser4-Tyr5-Gly6-Leu7-Arg8-Pro9-Gly10-NH2), respectively, has shown efficient suppression of tumor growth. In this study, the GnRH-receptor system has been evaluated with respect to its suitability as a target for in vivo peptide receptor targeting using radiolabeled GnRH-analogues, and in parallel, new 18F- and 68Ga -labeled GnRH analogues have been developed. In vitro radioligand binding assays performed with various GnRHR-expressing human cell lines using [125I]Triptorelin (D-Trp6-GnRH-I) as the standard radioligand revealed a very low level of GnRH receptor expression on the cell surface. Generally, total cellular activity was very low (∼3% of the applied activity), and only a small fraction (max. 40%) of cell-associated activity could be attributed to receptor-specific radioligand binding/internalization. However, substitution of fetal calf serum by NU serum in the culture medium led to increased and stable GnRHR-expression, especially in the ovarian cancer cell line EFO-27, thus allowing for a stable experimental setup for the evaluation of the new radiolabeled GnRH-I analogues. The new radiolabeled GnRH-I analogues developed in this study were all based on the D-Lys6-GnRH-I-scaffold. For 68Ga-labeling, the latter was coupled with DOTA at D-Lys6. To allow 18F-labeling via chemoselective oxime formation, D-Lys6-GnRH-I was also conjugated with Ahx (aminohexanoic acid) or β-Ala, which in turn was coupled with Boc-aminooxyacetic acid. 18F-labeling via oxime formation with 4-[18F]fluorobenzaldehyde was performed using the Boc-protected precursors. Receptor affinities of [68Ga]DOTA-GnRH-I, D-Lys6-Ahx([18F]FBOA)GnRH-I, and D-Lys6-βAla([18F]FBOA)-GnRH-I (FBOA ) fluorobenzyloxime acetyl) were determined using GnRHR-membrane preparations, and internalization efficiency of the new radioligands was determined in EFO27 cells. Both quantities were highest for D-Lys6-Ahx([18F]FBOA)-GnRH-I (IC50 ) 0.50 ( 0.08 nM vs 0.13 ( 0.08 nM for Triptorelin; internalization: 86 ( 16% of the internal reference [125I]Triptorelin), already substantially reduced in the case of the -βAla([18F]FBOA)-derivative (IC50 ) 0.86 ( 0.13 nM; internalization: 42 ( 3% of [125I]Triptorelin), while the [68Ga]DOTA-analogue showed almost complete loss of binding affinity and ligand internalization (IC50 ) 13.3 ( 1.0 nM; internalization: 2.6 ( 1.0% of [125I]Triptorelin). Generally, the lipophilic residue [18F]FBOA is much better tolerated as a modification of the D-Lys6-side chain, with receptor affinity of the respective analogues strongly depending upon spacer length between the D-Lys6-side chain and the [18F]FBOAmoiety. In summary, D-Lys6(Ahx-[18F]FBOA)-GnRH-I shows the highest potential for efficient GnRHR-targeting in vivo of the compounds investigated. Unfortunately, however, the very low cell surface expression of GnRHreceptors and thus very low radioligand uptake by GnRHR-positive tumor cells found in vitro was also confirmed by a preliminary biodistribution study in OVCAR-3 xenografted nude mice using the standard GnRHR radioligand [125I]Triptorelin. Tumor uptake was lower than blood activity concentration at 1 h p.i. (0.49 ( 0.05 vs 0.96 ( 0.13 for tumor and blood, respectively). These data seriously challenge the suitability of the GnRHR-system as a suitable target for in vivo peptide receptor imaging using radiolabeled GnRH-I derivatives, despite the availability of high-affinity radiolabeled receptor-ligands such as D-Lys6(Ahx-[18F]FBOA)-GnRH-I.

INTRODUCTION Neuropeptide receptors are overexpressed on a wide variety of human tumors. Due to their expression on the cell surface, they represent excellent molecular targets for radiolabeled neuropeptide analoguesstheir specific accumulation in receptor expressing tumors not only often allows high-contrast visualization of human malignancies, but also constitutes the basis for endoradiotherapeutic approaches. The somatostatin (sst) receptor system represents a prototype for successful peptide receptor targeting: the receptor, its function, and its expression pattern * To whom correspondence should be addressed. Margret Schottelius, Ph.D, Nuklearmedizinische Klinik and Poliklinik, Klinikum rechts der Isar, Technische Universita¨t Mu¨nchen, Ismaninger Strasse 22, D 81675 Mu¨nchen, Germany. Tel: # 49 89 4140 6492. Fax: # 49 89 4140 4897. E-mail: [email protected].

in normal and neoplastic tissues have been thorougly investigated, and an increasing number of optimized receptor-ligands labeled with a variety of radionuclides are available for both diagnostic and therapeutic applications (1–6). The success and increasing clinical impact of sst-targeting in nuclear oncology has prompted intensified research with respect to the identification of novel peptide receptors as potential candidates for in vivo tumor targeting (7, 8), and novel and/or optimized radioligands are currently being developed for–among others–bombesin (9, 10), cholecystokinin (11), and glucagon-like peptide-1 receptors (12), to mention only a few. Another receptor, which represents an interesting novel target in this field, is the gonadotropin releasing hormone receptor (GnRHR), also known as luteinizing hormone releasing hormone receptor (LHRHR). Its two subtypes, GnRHR-I and GnRHRII, have been shown to be strongly expressed on the majority

10.1021/bc800058k CCC: $40.75  2008 American Chemical Society Published on Web 05/30/2008

GnRH-I Analogues with High GnRHR-Targeting Efficiency

of tumors of the human reproductive system, either one subtype exclusively or both together: 80% of endometrial, 78% of ovarian, 52% of mammary, and 86% of prostate cancers are GnRHR-positive (13, 14). Expression of GnRHRs has also been demonstrated in human tumor specimens and cancer cell lines originating from liver (15), lanrynx (16), pancreas (17), kidney (18), brain (19), blood cells (20, 21), and skin (22, 23). The GnRH receptor system has been studied intensely as a target for peptidic drugs during the past decade. A broad palette of agonistic and antagonistic analogues derived from the native decapeptide ligand GnRH-I is clinically used for effective suppression of the pituitary-gonadal axis and consequently of the growth of hormone-dependent cancers (24). However, based on the finding that GnRH analogues also exert direct antiproliferative effects on GnRHR-expressing tumor cells (13, 24), cytotoxic analogues such as the 2-pyrrolidino-doxorubicin-coupled D-Lys6GnRH-derivative AN-207 have been developed and show high therapeutic efficiency in animal models (21, 25–27). Despite the clinical success of the pharmacological application of GnRH-derivatives (28), the GnRHR-system has not been exploited as a target for scintigraphic imaging using radiolabeled ligands yet. To date, only one 99mTc-labeled GnRH-analogue is known which has been specifically developed as a potential SPECT tracer (29). Unfortunately, the introduction of the radiometal via backbone cyclization of the peptide was accompanied by a dramatic loss of binding affinity, hampering the applicability of this analogue as an in vivo imaging agent. This study was therefore conducted to address two major open questions: First, is the GnRH-receptor system a suitable target for in vivo receptor imaging? And second, which structural prerequisites have to be met for the development of high-affinity radiolabeled GnRH analogues? To address the first question, several human cancer cell lines, i.e., the ovarian cancer cell lines EFO-27 and SKOV-3, the breast cancer cell lines MDA-MB 231 and SKBR-3, as well as the prostate cancer cell lines LNCaP and DU-145, all know to express either GnRHR-I and/or GnRHR-II (30–34), were screened for GnRHR-expression using the standard radioligands [125I]Triptorelin (pGlu-His-Trp-Ser-[125I]Tyr-D-Trp-Leu-ArgPro-Gly-NH2) for GnRHR-I (expressed on EFO-27, MDA-MB 231, SKBR-3, and LNCaP cells) or [125I]D-Lys6-GnRH-II (pGlu-His-Trp-Ser-His-D-Lys-Trp-[125I]Tyr-Pro-Gly-NH2) or GnRHR-II (expressed on SKOV-3 and DU-145 cells), and the extent of receptor-specific ligand binding was determined. Furthermore, the dependence of specific radioligand binding, reflecting the level of receptor expression, on the medium supplement, which had been previously documented for DU145 cells (35), was investigated for all six cell lines. The cells were grown in parallel both with fetal calf serum (FCS) or with Nu-Serum as serum supplements. Nu-Serum is a semisynthetic serum supplement containing, among other components, EGF, ECGS, insulin, human transferrin, triiodothyronine, progesterone, estradiol-17β, and testosterone, thus mimicking much more closely the physiological tumor cell growth conditions than FCS. In parallel, new radiolabeled GnRH analogues were developed. Previous studies had demonstrated that both the N- and the C-terminus of GnRH-ligands are responsible for receptor binding, and that their amino acid sequences are responsible for the agonistic or antagonistic profile of the peptide (36, 37). The development of a multitude of highly potent GnRH-I analogues further yielded important structure-activity relationships: first, introduction of a D-amino acid in position 6 leads to a substantial stabilization of the peptide toward degradation in vivo and an increased receptor affinity, and second, the amino acid in position 6 shows remarkable tolerance toward even bulky chemical modifications, both in GnRH agonists and antagonists (37–39).

Bioconjugate Chem., Vol. 19, No. 6, 2008 1257

Since agonism is an important prerequisite for ligand internalization, ideally leading to high intracellular tracer accumulation and enhanced tumor to nontumor ratios in vivo, the GnRH-I-agonist D-Lys6-GnRH-I (pGlu-His-Trp-Ser-Tyr-DLys-Leu-Arg-Pro-Gly-NH2) was chosen as a scaffold for radioligand development. To investigate the flexibility of the ligand with respect to the type of moiety attached to D-Lys6, two different synthetic approaches toward GnRH-I-analogues labeled with positron emitters were chosen: (A) the macrocyclic chelator DOTA (1,4,7,10-tetraazacyclododecane-N′,N′′,N′′′,N′′′′tetraacetic acid) was directly attached for 68Ga-labeling; and (B) the structure of the high-affinity cytotoxic analogue AN207 ((25); Figure 1) was used as a lead structure for the development of 18F-labeled GnRH-I-derivatives. To investigate the impact of spacer length between the peptide core and the hydrophobic moiety ([18F]FBOA (4-[18F]fluorobenzyloxim acetyl-) in the case of the 18F-analogues), two different spacers, aminohexanoic acid (Ahx) and β-alanine (β-Ala), were coupled to D-Lys6 and further functionalized with (Boc)aminooxyacetic acid for chemoselective 18F-labeling via oxime formation with 4-[18F]fluorobenzaldehyde (40) (Figure 1). The new radiolabeled compounds were evaluated in vitro, both using cell membranes to determine receptor affinity and using GnRH-I-expressing EFO-27 cells to assess internalization efficiency. In addition, an exemplary in vitro serum stability study was performed for D-Lys6(Ahx-[18F]FBOA)-GnRH-I. Furthermore, a preliminary evaluation of the OVCAR-3 cell line as a transplantable xenograft model for in vivo GnRH-Itargeting was carried out using the reference ligand [125I]Triptorelin.

MATERIALS AND METHODS General Conditions. Most Fmoc (9-fluorenylmethoxycarbonyl-) amino acids, N-Boc-aminooxyacetic acid, and Rink amide resin (100-200 mesh) were purchased from Novabiochem (Merck Biosciences, Schwalbach, Germany). Boc-PyrOH (N-Boc-pyroglutamic acid) and Boc-β-Ala-OH were supplied by Bachem (Heidelberg, Germany). 6-Boc-Ahx-OSu was purchased from Fluka (Neu-Ulm, Germany). DOTA (1,4,7,10tetraazacyclododecane-N′,N′′,N′′′,N′′′′-tetraacetic acid) was obtained from Macrocyclics (Dallas, Texas, USA). Solvents and all other organic reagents and were purchased from Merck Eurolab (Darmstadt, Germany), Aldrich or Fluka (Neu-Ulm, Germany). Solid-phase peptide synthesis was carried out manually using a flask shaker (St. John Associates Inc., USA). Analytical RP-HPLC was performed on a Nucleosil 100 C18 (5 µm, 125 × 4.0 mm) column (CS Chromatographie Service, Langerwehe, Germany) using a Sykam gradient HPLC System (Sykam GmbH, Fu¨rstenfeldbruck, Germany). The peptides were eluted applying different gradients of 0.1% (v/v) TFA (trifluoroacetic acid) in H2O (solvent A) and 0.1% TFA (v/v) in acetonitrile (solvent B) at a constant flow of 1 mL/min (specific gradients are cited in the text). UV detection was performed at 220 nm using a 206 PHD UV-vis detector or an UVIS 200 photometer (Linear Instruments Corporation, Reno, NV). Preparative RP-HPLC was performed on the same HPLC system using a YMC J’Sphere ODS-H80 (150 × 20 mm, S-4 µm, 8 nm) column (YMC Europe, Schermbeck, Germany) at a constant flow of 9 mL/min. Both retention times tR as well as the capacity factors K′ are cited in the text. K′ is calculated as follows: K′ )

tR - t0 t0

tR ) retention time dead volume of the column [mL] t0 ) flow [mL × min -1]

1258 Bioconjugate Chem., Vol. 19, No. 6, 2008

Schottelius et al.

Figure 1. Structures of some selected GnRHR-binding compounds and of the new radiolabeled GnRH-I analogues developed in this study. FBOA ) fluorobenzyloxime acetyl, DOTA ) 1,4,7,10-tetraazacyclododecane-N′,N′′,N′′′,N′′′′-tetraacetic acid.

Mass spectra were recorded on the LC-MS system LCQ from Finnigan (Bremen, Germany) using the Hewlett-Packard series 1100 HPLC system. Peptide Synthesis. Triptorelin (D-Trp6-GnRH-I). Coupling of the first amino acid, Fmoc-Gly-OH, to the resin was performed as specified by the manufacturer. Briefly, the resin was allowed to preswell in NMP (N-methylpyrrolidon) for 30 min. Then, the N-terminal Fmoc-protecting group was cleaved off using 20% piperidine in DMF (dimethyl formamide). After thoroughly washing the resin with NMP, Fmoc-Gly-OH was coupled to the resin using HOBt (1-hydroxybenzotriazol) and TBTU (O-(1H-benzotriazol-1-yl)-N,N,N′,N′-tetramethyluroniumtetrafluoroborate) as coupling reagents and DIPEA (N-ethyl diisopropylamine) as a base. The subsequent assembly of the peptide sequence pGlu(Boc)-His(Trt)-Trp(Boc)-Ser(tBu)-Tyr(tBu)D-Trp(Boc)-Leu-Arg(Pbf)-Pro- on the resin-bound amino acid was performed according to standard Fmoc-protocol. Cleavage from the resin was carried out using a mixture of 95% TFA, 2.5% water, and 2.5% triisobutylsilane (v/v/v). The fully deprotected crude peptide was precipitated using diethyl ether and dried. A small fraction of crude Triptorelin was then purified via preparative HPLC (22-37% B in 20 min). Triptorelin (40 mg) was obtained in a purity (220 nm) >98%. Triptorelin: HPLC (10 f 35% B in 20 min): tR ) 12.3 min; K′ ) 6.65. Calculated monoisotopic mass (C64H82N18O13): 1310.6. found: m/z ) 1312.1 [M+H]+, 1333.9 [M+Na]+. 6 6 D-Lys -GnRH-I. D-Lys (Dde)-GnRH-I was synthesized via SPPS (solid phase peptide synthesis) in analogy to Triptorelin with the sole difference that Fmoc-D-Trp(Boc)-OH was replaced

by Fmoc-D-Lys(Dde)-OH during SPPS. After cleavage from the resin, D-Lys6(Dde)-GnRH-I was precipitated using diethyl ether, dried, and stored at -20 °C. For further derivatization, the Ddeprotecting group in D-Lys6(Dde)-GnRH-I was removed using 2% hydrazine hydrate (30%) in DMF. After 10 min, deprotected 6 D-Lys -GnRH-I was precipitated using diethyl ether, washed with ether, and dried. The crude peptide was used for subsequent derivatizations without further purification. D-Lys6-GnRH-I: HPLC (10 f 35% B in 20 min): tR ) 7.1 min; K′ ) 3.94. Calculated monoisotopic mass (C59H84N18O13): 1252.7. found: m/z ) 1253.8 [M+H]+, 1275.7 [M+Na]+. 6 6 D-Lys -GnRH-II. The synthesis of D-Lys (Dde)-GnRH-II (pGlu-His-Trp-Ser-His-D-Lys-Trp-Tyr-Pro-Gly-NH2) was carried out entirely in analogy to D-Lys6(Dde)-GnRH-I. After Ddedeprotection, the crude peptide was purified via preparative RPHPLC. D-Lys6-GnRH-II: HPLC (15 f 45% B in 20 min): tR ) 9.0 min; K′ ) 4.86. Calculated monoisotopic mass (C64H78N18O13): 1306.6. found: m/z ) 1307.9 [M+H]+, 1330.0 [M+Na]+. 6 D-Lys (DOTA)-GnRH-I. The macrocyclic chelator DOTA was coupled to the side chain of D-Lys6 in D-Lys6-GnRH-I according to a previously published protocol (41). Briefly, DOTA (2 equiv, 0.37 mmol) was preactivated with NHS (N-hydroxysuccinimide, 2.5 equiv), EDCI (1-ethyl-3-(3′-dimethylaminopropyl)-carbodiimide, 2.5 equiv) and DIPEA (5 equiv) in water for 30 min. Then, D-Lys6-GnRH-I (1 equiv, 0.18 mmol) dissolved in DMF was slowly added under vigorous stirring. After 90 min, the solvents were evaporated in vacuo, the residue was resuspended in methanol, and the suspension was centrifuged. The super-


natant methanolic solution was concentrated in vacuo and added dropwise to an excess of diethyl ether. The precipitate (crude 6 D-Lys (DOTA)-GnRH-I) was dried and subsequently purified using preparative HPLC (14-30% B in 20 min). D-Lys6(DOTA)GnRH-I (47 mg) was obtained in 16% yield and a purity (220 nm) of 97.6%. D-Lys6(DOTA)-GnRH-I: HPLC (10 f 35% B in 20 min): tR ) 10.9 min; K′ ) 5.19. Calculated monoisotopic mass (C75H110N22O20): 1638.8. found: m/z ) 1640.0 [M+H]+, 821.1 [(M+2H)/2]2+, 1662.0 [M+Na]+, 1677.9 [M+K]+. 6 D-Lys (Ahx-AOA(Boc))-GnRH-I. Coupling with Boc-AhxOH and Deprotection. Boc-Ahx-OSu (1.5 equiv, 0.52 mmol), HOBt (1.5 equiv, 0.52 mmol), and TBTU (1.5 equiv, 0.52 mmol) were dissolved in DMF; DIPEA (4 equiv, 2.08 mmol) was added, and the mixture was preactivated for 15 min. Then, this solution was added to a solution of D-Lys6-GnRH-I (1 equiv, 0.35 mmol) in DMF. After 30 min of stirring at RT (room temperature), the crude product was precipitated using diethyl ether, washed, and dried (D-Lys6(Ahx(Boc))-GnRH-I: HPLC (20 f 70% B in 15 min): tR ) 10.3 min; K′ ) 4.35). For Bocdeprotection, the dried residue was redissolved in a 1:1 mixture (v/v) of TFA and dichloromethane. Usually, deprotection was complete within 15 min. The deprotected peptide was then precipitated using diethyl ether, washed, and redissolved in water/MeOH for subsequent preparative HPLC purification (16-27% B in 20 min). D-Lys6(Ahx)-GnRH-I: HPLC (20 f 70% B in 15 min): tR ) 4.8 min; K′ ) 1.83. Calculated monoisotopic mass (C65H95N19O14): 1365.6. found: m/z ) 1366.7 [M+H]+, 684.2 [(M+2H)/2]2+, 1388.7 [M+Na]+. Coupling with N-(Boc)-Aminooxyacetic Acid. N-(Boc)-aminooxyacetic acid (AOA(Boc)) was coupled to D-Lys6(Ahx)GnRH-I under optimized coupling conditions (42). Briefly, AOA(Boc) (1.2 equiv, 0.030 mmol) was preactivated for 15 min with HOAt (1-hydroxy-7-azabenzotriazole, 1.2 equiv, 0.030 mmol), DIC (diisopropylcarbodiimide, 1.2 equiv, 0.030 mmol), and pyridine (2 equiv, 0.052 mmol) in DMF. This solution was added to a solution of D-Lys6(Ahx)-GnRH-I (1 equiv, 0.026 mmol) in DMF. After stirring at RT for 35 min, the product peptide was precipitated using diethyl ether, washed, and isolated using preparative HPLC purification (24-52% B in 20 min). 6 D-Lys (Ahx-AOA(Boc))-GnRH-I was obtained in 50% yield and a purity > 90%. D-Lys6(Ahx-AOA(Boc))-GnRH-I: HPLC (20 f 70% B in 15 min): tR ) 9.2 min; K′ ) 4.21. Calculated monoisotopic mass (C72H106N20O18): 1538.5. found: m/z ) 1539.6 [M+H]+, 1439.7 [M-Boc+H]+. 6 D-Lys (β-Ala-AOA(Boc))-GnRH-I. Coupling with Boc-βAla-OH and Deprotection. Coupling of Boc-β-Ala-OH to 6 D-Lys -GnRH-I as well as subsequent Boc-deprotection were performed in analogy to the Ahx-analogue. After preparative HPLC purification (16-23% B in 20 min), D-Lys6(β-Ala)GnRH-I was obtained in 21% yield. D-Lys6(β-Ala)-GnRH-I: HPLC (20 f 70% B in 15 min): tR ) 2.8 min; K′ ) 0.58. Calculated monoisotopic mass (C62H89N19O14): 1323.7. found: m/z ) 1324.8 [M+H]+, 1346.7 [M+Na]+. Coupling with N-(Boc)-Aminooxyacetic Acid. N-(Boc)-aminooxyacetic acid (AOA(Boc)) was coupled to D-Lys6(β-Ala)GnRH-I as described for the Ahx-derivative. After preparative HPLC purification (22-46% B in 20 min), D-Lys6(β-AlaAOA(Boc))-GnRH-I was obtained in 49% yield and a purity > 90%. D-Lys6(β-Ala-AOA(Boc))-GnRH-I: HPLC (20 f 70% B in 15 min): tR ) 8.5 min; K′ ) 3.73. Calculated monoisotopic mass (C69H100N20O18): 1496.5. found: m/z ) 1497.6 [M+H]+, 1397.7 [M-Boc+H]+. Synthesis of the Reference Compounds. D-Lys6(GaDOTA)-GnRH-I. D-Lys6(DOTA)-GnRH-I was dissolved in 0.1 N NaOAc (pH ) 4.5), and an equimolar amount of GaNO3 was added. After heating to 90 °C for 30 min, the reaction mixture was cooled to RT, and D-Lys6(Ga-DOTA)-GnRH-I was


immediately purified using preparative HPLC (14-30% B in 20 min). The peptide was obtained in 97% purity (220 nm). 6 D-Lys (Ga-DOTA)-GnRH-I: HPLC (10 f 35% B in 20 min): tR ) 7.6 min; K′ ) 3.61. Calculated monoisotopic mass (C75H108N22O20Ga): 1705.7. found: m/z ) 1705.9 [M]+, 854.0 [(M+H)/2]2+, 570.2 [(M+2H)/3]3+. 6 6 D-Lys (Ahx-FBOA)-GnRH-I and D-Lys (β-Ala-FBOA)-GnRHI. Immediately prior to the reaction with 4-fluorobenzaldehyde, 6 6 D-Lys (Ahx-AOA(Boc))GnRH-I and D-Lys (β-Ala-AOA(Boc))GnRH-I were deprotected using a 1:1 mixture (v/v) of TFA and dichloromethane. After 15 min, the deprotected peptides were precipitated using diethyl ether, washed, and redissolved in water/MeOH. Then, an equimolar amount of 4-fluorobenzaldehyde was added. After 30 min at 60 °C, the respective coupling product was purified using RP-HPLC (D-Lys6(AhxFBOA)GnRH-I: 20-66% B in 20 min, D-Lys6(β-Ala-FBOA)GnRH-I: 22-60% B in 20 min). The purity (220 nm) of both the Ahx- and β-Ala-reference compound was >98%. D-Lys6(AhxFBOA)GnRH-I: HPLC (20 f 70% B in 15 min): tR ) 10.3 min; K′ ) 4.84. Calculated monoisotopic mass for D-Lys6(AhxFBOA)GnRH-I (C74H101N20O16F): 1544.5. found: m/z ) 1545.9 [M+H]+, 773.9 [(M+2H)/2]2+, 1567.7 [M+Na]+. D-Lys6(βAla-FBOA)GnRH-I: HPLC (20 f 70% B in 15 min): tR ) 9.6 min; K′ ) 4.55. Calculated monoisotopic mass for D-Lys6(βAla-FBOA)GnRH-I (C71H95N20O16F): 1502.5. found: m/z ) 1503.8 [M+H]+, 752.8 [(M+2H)/2]2+, 1525.5 [M+Na]+. Radiolabeling. Instrumentation. Radio-HPLC was performed on a Nucleosil 100 C18 (5 µm, 125 × 4.0 mm) column using a Sykam gradient system (Sykam GmbH, Fu¨rstenfeldbruck, Germany) and an UVIS 200 photometer (Linear Instruments Corporation, Reno, NV). For radioactivity measurement, the outlet of the UV-photometer was connected to a Na(Tl) welltype scintillation counter Ace Mate 925-Scint (EG&G Ortec, Mu¨nchen, Germany). [125I]Triptorelin and D-Lys6-[125I]GnRH-II. Iodogen Method. A solution of approx 200 µg of peptide in 200 µL of TRISHCl buffer (25 mM, 0.4 M NaCl, pH 7.5) was transferred to an Eppendorf tube coated with 50 µg of Iodogen (Pierce, Rockford, USA). After the addition of 5-20 µL (18.5-74 MBq) of n.c.a. [125I]NaI in 0.05 M NaOH (Hartmann Analytic, Brauschweig, Germany), the cap was vortexed, and the labeling reaction was allowed to proceed for 15 min at RT. The peptide solution was then removed from the insoluble oxidizing agent, and the radiolabeled peptide was separated from unreacted precursor via RP-HPLC using an isocratic solvent mixture of 24% (DLys6-[125I]GnRH-II) or 32% ([125I]Triptorelin) EtOH (0.5% AcOH) in water (0.5% AcOH). Radiochemical yields after purification ranged from 35% to 65%. Chloramine T Method. Triptorelin (100 µg) was dissolved in 100 µL of TRIS-HCl buffer (25 mM, 0.4 M NaCl, pH 7.5). After addition of 10 µL (37 MBq) of n.c.a. [125I]NaI in 0.05 M NaOH and 50 µL of a Chloramine T solution (1 mg/mL), the labeling reaction was allowed to proceed for 15 min at RT. Then, 10 µL of a Na2S2O5 solution (1 mg/mL) and 120 µL of TRISHCl buffer were added, and [125I]Triptorelin was purified via RP-HPLC as described above. Radiochemical yields after purification ranged from 79% to 87%. For use in the in vitro experiments, the HPLC eluate was evaporated to dryness, and [125I]Triptorelin was redissolved in RPMI-1640 medium containing 5% BSA (activity concentration of approx 100 000 cpm/10 µL). [125I]Triptorelin: 32% EtOH isocratic: tR (range) ) 6.6-8.5 min; K′ (range) ) 3.77-4.83. [125I]D-Lys6-GnRH-II: 24% EtOH isocratic: tR (range) ) 7.7-9.5 min; K′ (range) ) 4.13-5.33. 6 68 68 D-Lys ([ Ga]DOTA)-GnRH-I. GaCl3 was eluted with 0.1 N HCl from an in-house OBNINSK 68Ge/68Ga-generator (1.48 GBq; Chemotrade Chemiehandelsgesellschaft mbH, Leipzig,


Germany). To 1.3 mL of the generator eluate were added 150 µL of HEPES (2.5 M) and 7 µg (4.25 nmol) of D-Lys6(DOTA)GnRH-I in 7 µL of water, and the reaction mixture was heated to 95 °C for 20 min. After cooling to RT, the reaction mixture was diluted with 2 mL of water, and D-Lys6([68Ga]DOTA)GnRH-I was immobilized on a Sep-Pak Plus C-18 cartridge (Waters, Eschborn, Germany), washed with 10 mL of water, and eluted with 2 mL of ethanol. The solvent was then evaporated in vacuo and the product was redissolved in RPMI1640 medium (Biochrom, Berlin, Germany) containing 5% BSA (bovine serum albumin) and diluted to an activity concentration of approx 500 000 cpm/10 µL for use in internalization studies. Quality control was performed using a standard TLC system. 6 18 6 18 D-Lys (Ahx-[ F]FBOA)-GnRH-I and D-Lys (β-Ala-[ F]FBOA)-GnRH-I. The synthesis of 4-[18F]fluorobenzaldehyde from 4-formyl-N,N,N-trimethylanilinium triflate was performed as described previously (40). For peptide labeling, 10 µL of a 10 mM solution of D-Lys6(Ahx/β-Ala-AOA(Boc))-GnRH-I in water was added to 20-40 µL of a freshly prepared methanolic 4-[18F]fluorobenzaldehyde solution. Then, 1 µL of TFA was added. After 15 min at 75 °C and product isolation via RPHPLC (15-50% B in 20 min, analytical column), both [18F]FBOA-GnRH-I-analogues were obtained in radiochemical yields of 74-85% (based on 4-[18F]fluorobenzaldehyde) and an overall synthesis time of approx 60 min. For use in in vitro studies, the HPLC eluate was evaporated to dryness, and the [18F]FBOA-peptides were redissolved in RPMI-1640 medium containing 5% BSA (activity concentration of approx 250 000 cpm/10 µL). D-Lys6(Ahx-[18F]FBOA)GnRH-I: HPLC (15 f 50% B in 20 min): tR ) 16.7 min; K′ ) 9.14. D-Lys6(β-Ala[18F]FBOA)GnRH-I: HPLC (15 f 50% B in 20 min): tR ) 15.2 min; K′ ) 8.77. Lipophilicity. To a solution of approx 2 kBq of radiolabeled peptide in 500 µL of PBS (pH 7.4), 500 µL of octanol were added (n ) 6). Vials were vortexed vigorously for 3 min. To achieve quantitative phase separation, the vials were centrifuged at 14 600 × g for 6 min in a Biofuge 15 (Heraeus Sepatech, Osterode, Germany). The activity concentrations in 100 µL samples of both the aequous and organic phases were measured in a γ-counter. Both the partition coefficient Pow, which is defined as the molar concentration ratio of a single species A between octanol and water at equilibrium and log Pow, which is an important parameter used to characterize lipophilicity of a compound (43), were calculated. Cell Culture. The ovarian cancer cell line EFO-27 was purchased from DSMZ (Braunschweig, Germany). OVCAR-3 ovarian cancer cells were a kind gift from Dr. A. Kru¨ger, Institute of experimental oncology, Klinikum rechts der Isar, TU Mu¨nchen, Germany. The prostate cancer cell lines DU-145 and LNCaP as well as the breast cancer cell lines MDA-MB 231 and SKBR-3 were kindly supplied by the group of Prof. R. Senekowitsch-Schmidtke, Department of Nuclear Medicine, Klinikum rechts der Isar, TU Mu¨nchen, Germany. All cell lines except SKBR-3 were grown in RPMI-1640 medium supplemented with either 10% FCS (fetal celf serum; Biochrom, Berlin, Germany) or 10% Nu-Serum (BD Bioscience, Heidelberg, Germany). SKBR-3 cells were grown in Gibco DMEM/ Nutrition Mix F-12 with Glutamax-I (1:1) (invitrogen, Karlsruhe, Germany) supplemented with the same two sera mentioned above. Cultures were maintained at 37 °C in a 5% CO2/ humidified air atmosphere. For cell counting, a CASY1-TT cell counter and analyzer system (Scha¨rfe System GmbH, Reutlingen, Germany) was used. Determination of the Serum Dependence of [125I]Triptorelin Binding. One day prior to the experiment, cells that had been grown in medium containing either FCS or Nu-Serum were harvested using Trypsin/EDTA (0.05% and 0.02%) in PBS,

Schottelius et al.

centrifuged, resuspended with culture medium, transferred into 24-well plates (1 mL/well), and placed in the incubator overnight. On the day of the experiment (cell count: 160 000-220 000 cells/well), the culture medium was removed and the cells were washed once with 250 µL of medium (unsupplemented RPMI1640) before being left to equilibrate in 190 µL of medium at 37 °C for a minimum of 15 min before the experiment. Then, 50 µL per well of medium containing 5% BSA (n ) 3, control experiment, nca conditions) or of a 25 µM solution of Triptorelin in the same medium (n ) 3, determination of nonspecific binding) were added. This was followed by the addition of ∼100 000 cpm [125I]Triptorelin in 10 µL of medium (5% BSA). The cells were then incubated for 60 and 120 min, respectively, at 37 °C in a 5% CO2/humidified air atmosphere. Incubation was terminated by removal of the incubation medium. Cells were thoroughly rinsed with 250 µL of fresh medium. The wash medium was combined with the supernatant of the previous step. This fraction represents the amount of free radioligand. To determine the amount of bound radioligand, cells were lysed by the addition of 300 µL of 1 N NaOH, transferred to vials, and combined with 250 µL of PBS used for rinsing the wells. Quantification of the amount of free and bound activity was performed in a γ-counter (Wallach, Turku, Finland). Internalization Studies. In general, the internalization experiments were performed in strict analogy to the experiments described above. However, because all internalization experiments were performed as dual tracer experiments, cells were not only incubated with the reference [125I]Triptorelin, but with a mixture of radioligands, i.e., of ∼100 000 cpm [125I]Triptorelin and either ∼250 000 cpm of the respective [18F]FBOA-GnRHI-analogue or ∼500 000 cpm D-Lys6([68Ga]DOTA)-GnRH-I. After the addition of the radioligands, the cells were incubated for 10, 30, 60, and 120 min at 37 °C. After removal of the incubation medium and washing with medium, an additional acid wash step was performed to remove surface-bound (acidreleasable) radioactivity. Cells were incubated with 250 µL of ice cold acid wash buffer (0.02 M NaOAc buffered with AcOH to pH ) 5), and after removal of the acid wash buffer, the cells were thoroughly rinsed with another 250 µL of ice cold acid wash buffer. Both acid wash fractions were combined. The internalized activity was released by incubation with 250 µL of 1 N NaOH as described above. Quantification of the amount of free, acid-releasable, and internalized activity was performed in a γ-counter. Determination of IC50 Using Cell Membranes. Cell membrane preparations of cells expressing the human GnRHR were obtained from Chemicon International (Hampshire, UK). Assembly of the competition binding curves was performed in close analogy to the manufacturer’s protocol. Briefly, solutions of unlabeled competitor (Triptorelin, D-Lys6(Ga-DOTA)-GnRHI, D-Lys6(Ahx-FBOA)-GnRH-I, and D-Lys6(β-Ala-FBOA)GnRH-I; 20 µL, concentration range from 1 · 10-11 to 1 · 10-5 M, n ) 3, per concentration and peptide) in binding buffer (50 mM HEPES, 5 mM MgCl2, 1 mM CaCl2, 0.2% BSA, buffered to pH 7.4 using 1 N NaOH) were pipetted into a 96-well plate. To each well, 20 µL of [125I]Triptorelin in binding buffer (approx 100 000 cpm) was added. Then, 10 µg of cell membrane in 60 µL of binding buffer was added and incubated for 90 min at RT. The final radioligand concentration in the incubation mixture was 0.42 nM. Membranes were then aspired onto a glass fiber filtermat (Printed Filtermat B, Wallach, Turku, Finland), which had been preconditioned with 0.33% polyethylenimine for 30 min and then washed five times with 200 µL (per well) of wash buffer (50 mM HEPES, 500 mM NaCl, 0.1% BSA, buffered to pH 7.4 using 1 N NaOH), using a Mach II M Harvester 96 (Tomtec CE, Etten-Leur, The Netherlands).


Membranes were washed 15 times with 100 µL (per well) of wash buffer. The filtermat was then cut into squares containing the membrane from the respective wells, and membrane bound activity was counted in a γ-counter. Serum Stability. To four 500 µL samples of freshly prepared human serum was added 50 µL of a solution containing 4.6 MBq of D-Lys6(Ahx-[18F]FBOA)GnRH-I in PBS (pH ) 7.5). The mixtures were incubated in a 5% CO2 environment at 37 °C. After 10, 30, 60, and 120 min, respectively, the samples were removed from the incubator and diluted with 10 mL of water. The solution was applied to a Sep-Pak Plus C-18 cartridge, and the eluate was collected (fraction I). Then, the cartridge was washed with 10 mL of water. The eluate was again collected (fraction II). After drying with 10 mL of air, the cartridge was eluted with 2 mL of 70% EtOH containing 1% (v/v) 0.1 N HCl (fraction III). The radioactivity in all eluates (fractions I-III) as well as the activity remaining bound to the cartridge material were counted in a Capintec. The activity collected in the EtOH-eluate was further analyzed via analytical RP-HPLC (5 f 90% B in 20 min). Biodistribution Study. To establish tumor growth, OVCAR-3 cells were detached from the surface of the culture flasks using 1 mM EDTA in PBS, centrifuged, and resuspended in serumfree culture medium. Concentration of the cell suspension was 5 · 106 cells/100 µL medium. Nude mice (Swiss nu/nu, female, 6-8 weeks; Charles River, Sulzfeld, Germany) were injected with 100 µL of the cell suspension subcutaneously into the flank. Six weeks after tumor transplantation, all mice showed solid palpable tumor masses (tumor weight 0.15-0.8 g) and were used for the experiments. For biodistribution studies, mice (n ) 3) were intravenously injected with 0.37 MBq [125I]Triptorelin in 100 µL PBS into the tail vein. To determine nonspecific tracer accumulation, 25 µg of unlabeled Triptorelin was coinjected (n ) 3). At 60 min after radioligand injection, mice were sacrificed and dissected. The organs of interest were removed, weighed, and counted in a γ-counter (Wallach, Turku, Finland). Data are expressed as percent injected dose per gram tissue (%iD/g).

RESULTS Peptide Synthesis. Triptorelin (pGlu-His-Trp-Ser-Tyr-D-TrpLeu-Arg-Pro-Gly-NH2), D-Lys6-GnRH-I (pGlu-His-Trp-Ser-Tyr6 D-Lys-Leu-Arg-Pro-Gly-NH2) as well as D-Lys -GnRH-II (pGluHis-Trp-Ser-His-D-Lys-Trp-Tyr-Pro-Gly-NH2) were assembled on Rink Amide resin via a standard Fmoc-protocol. Cleavage from the resin was carried out using 95% trifluoroacetic acid (TFA), leading to concomitant side chain deprotection. Triptorelin and D-Lys6-GnRH-II were then purified via preparative HPLC and were obtained in purities > 98% (220 nm). 6 6 D-Lys -GnRH-I was further derivatized on the Lys -side chain, either by direct coupling with unprotected DOTA (41) to yield D-Lys6(DOTA)-GnRH-I or by coupling with different spacers (aminohexanoic acid (Ahx) or β-Ala), deprotection, and subsequent coupling with (Boc)aminooxyacetic acid ((Boc)Aoa) (Figure 1). After workup and product isolation via preparative HPLC, 6 D-Lys (DOTA)-GnRH-I was obtained in 16% yield. Due to impurities in the starting peptide and resulting formation of side products, necessitating an intermediate HPLC purification step, yields of D-Lys6((Boc)Ahx)- and D-Lys6((Boc)β-Ala)GnRH-I were also comparably low (approx 20%). The respective (Boc)Aoa-derivatives, however, were obtained in 50% yield after preparative HPLC isolation. Purities of these compounds were somewhat lower than for the other peptides synthesized in this study (90 vs >98%), which is occasioned by unavoidable cleavage of the Boc-protecting group from the Aoa-functionality during solvent evaporation after preparative HPLC, the 10%


impurity consisting of the respective Ahx-Aoa- and β-Ala-Aoapeptides. However, since the Aoa-deprotected peptides are the reactive species during 18F-labeling and are in any case generated in situ during the labeling reaction, no attempt was made to further purify the seemingly impure D-Lys6(Ahx(Boc)Aoa)- and D-Lys6(β-Ala-(Boc)Aoa)GnRH-I preparations, and they were used as such for the labeling reaction. Radiolabeling. Radioiodination of Triptorelin and D-Lys6GnRH-II was initially carried out using the Iodogen method. However, under these conditions, radiochemical yields were comparably low (35-65% after HPLC purification) due to the formation of radiolabeled side products, and an optimized protocol based on a method described in the literature (44) was developed. Using chloramine T as an oxidant, the incorporation of radioiodide into the peptide was improved, and the formation of radioiodinated side products was successfully suppressed, leading to radiochemical yields of [125I[Triptorelin of 82-89% after HPLC isolation. Because unreacted precursor was completely separated from both radioiodinated peptides, respectively, via HPLC, the specific activity of the radioiodinated compounds was assumed to be that of the radioiodide used for their preparation, i.e., 74 TBq/mmol. 18 F-labeling via oxime formation with 4-[18F]fluorobenzaldehyde was performed in analogy to a previously published method (40), but with one major modification. Instead of first preparing and purifying the deprotected aminooxy-functionalized peptides and then using them for the labeling reaction, they were generated in situ from the protected (Boc)Aoaprecursors in the acidic (pH ) 2) reaction milieu at 75 °C. Compared to the original labeling protocol, this necessitated a higher peptide concentration in the reaction mixture (2.5 vs 0.5 mM), since Boc-deprotection was not quantitative under the reaction conditions. However, D-Lys6(Ahx-[18F]FBOA)-GnRHI and D-Lys6(β-Ala-[18F]FBOA)-GnRH-I were obtained in high yields (74-85% based on 4-[18F]fluorobenzaldehyde) and high radiochemical purities (>99%) in an overall synthesis time of approx. 60 min (including RP-HPLC purification). Identity of the [18F]FBOA-peptides was confirmed via coinjection with the 19 F-reference compounds (Figure 2). The variance in peak symmetry between the UV and the radiochromatogram, i.e., tailing of the activity peak, is due to suboptimal lead shielding of the detector. 68 Ga-labeling of D-Lys6(DOTA)-GnRH-I was carried out according to a standard radiometalation strategy. After purification via a RP-18 cartridge, D-Lys6([68Ga]DOTA)-GnRH-I was obtained in 94% radiochemical yield and >99% radiochemical purity and a specific activity of 18.1 TBq/mmol. Lipophilicity. For all radiolabeled peptides, the partition coefficient P between n-octanol and PBS (0.01 M, pH 7.4) and log P as a standard quantity to describe ligand lipophilicity (43) were determined. Lipophilicities of [125I]Triptorelin and D-Lys6(Ahx-[18F]FBOA)-GnRH-I were similar ([125I]Triptorelin: log P ) -0.247 ( 0.016, D-Lys6(Ahx-[18F]FBOA)-GnRH-I: log P ) -0.395 ( 0.016), whereas it was lower in the case of 6 18 D-Lys (β-Ala-[ F]FBOA)-analogue due to the reduced length of the aliphatic chain compared to its Ahx-counterpart (log P ) -0.686 ( 0.060). With a log P of -3.703 ( 0.019, DLys6([68Ga]DOTA)-GnRH-I was the most hydrophilic of the compounds investigated. In Vitro Studies. Serum Dependence of [125I]Triptorelin/ 6 125 D-Lys -[ I]GnRH-II Binding. All cell lines investigated (EFO27, SKOV-3, LNCaP, DU-145, MDA-MB 231, SKBR-3) were cultured in parallel with different medium supplements, i.e., fetal calf serum (FCS) and Nu-Serum, a semisynthetic serum supplement. To evaluate the influence of culture conditions on GnRHR-expression, cells, that had been grown in the presence of the different serum supplements, were incubated with the


Figure 2. Analytical chromatograms of the coinjection of D-Lys6(βAla-[18F]FBOA)-GnRH-I (upper panel) and D-Lys6(Ahx-[18F]FBOA)GnRH-I (lower panel) with their respective 19F-analogues, i.e., D-Lys6(βAla-FBOA)-GnRH-I and D-Lys6(Ahx-FBOA)-GnRH-I. RP-HPLC was performed using a Nucleosil 100 C18 column and a gradient of 15 f 50% B in 20 min (A ) 0.1% TFA, B ) 0.1% TFA in acetonitrile).

respective standard radioligand for 60 and 120 min, and total as well as nonspecific radioligand uptake in the presence of an excess of cold competitor (5 µM) at the respective time point were determined. For EFO-27, LNCaP, MDA-MB 231, and SKBR-3 cells, which express the GnRH-I-receptor (14, 30, 31, 33, 34), [125I]Triptorelin was used as the radioligand. The cell lines SKOV-3 and DU-145 have been shown to express the putative GnRH-II-receptor (30, 32), and thus the GnRH-II analogue D-Lys6-[125I]GnRH-II, was used as the radioligand to evaluate whether potential effects of the serum supplement on receptor expression could also be observed in the case of the receptor type II. Data obtained for the cells grown with Nu-Serum and FCS are shown in Figure 3. Generally, cellular uptake of [125I]Triptorelin or D-Lys6-[125I]GnRH-II was low (e3.5% of total activity), and the extent of nonspecific binding was high (55-80% of total binding). This is reflected by the low percentage of specific ligand binding (in % of total binding) shown in Figure 3. However, substitution of FCS in the culture medium by Nu-Serum increased the fraction of specifically bound ligand in all cell lines and independently of the receptor subtype, indicating stronger GnRHR expression under these culture conditions. The observed increase in specific binding was highest (increase by a factor of ∼2) in EFO-27, LNCaP, and MDA-MB 231 cells. Of the cell lines investigated, EFO27 cells showed the highest absolute radioligand uptake (3.01 ( 0.19% and 3.11 ( 0.18% of the applied activity after 60 and 120 min, respectively; LNCaP, 2.49 ( 0.15% and 2.55 ( 0.18%; MDA-MB, 1.53 ( 0.07% and 2.15 ( 0.18%), and thus EFO-27 cells cultured with Nu-Serum were chosen as the in vitro model for the evaluation of the new radioligands developed in this study. Internalization Studies. Since high cellular uptake of radiolabeled peptide agonists greatly enhances the tumor to background ratio in peptide receptor imaging, the new 68Ga- and 18 F-labeled GnRH-I analogues were also evaluated with respect to their internalization into EFO-27 cells. All studies were

Schottelius et al.

Figure 3. Dependence of radioligand uptake into different human GnRHR-positive cell lines on the serum supplement. Cells were grown using either fetal calf serum (FCS) or semisynthetic Nu-Serum for several weeks. They were then transferred into 24-well plates and incubated for 60 min with either the GnRHR-I ligand [125I]Triptorelin or the GnRHR-II ligand [125I]D-Lys6-GnRH-II (GnRHR-II expressing cells are labeled with an asterisk (*)). Data are means of triplicate experiments (n ) 3) and are given as specifically bound radioligand () total - nonspecifically bound) in % of total binding. Nonspecific binding was determined in the presence of an excess (5 µM) of cold ligand.

performed in triplicate as dual tracer experiments with the respective 68Ga- and 18F-labeled peptides and [125I]Triptorelin as an internal reference. This experimental setup assures comparability of the data obtained for the 68Ga- and 18F-labeled peptides, because the differences in specific activity between the radiometallated (c.a.) and radiofluorinated (n.c.a.) peptides are significant and may lead to large variations in cellular ligand uptake. However, uptake data for [125I]Triptorelin in all experiments showed that the internalization of the reference peptide is unchanged by the different final peptide concentrations in the assay wells (1.8 nM for D-Lys6([68Ga]DOTA)-GnRH-I, subnanomolar for the 18F-labeled peptides). Furthermore, performing the internalization studies as dual tracer experiments also allows elimination of systematic errors due to variable cell count or cell viability. In the case of slowly internalizing receptors such as GnRHR (45, 46), it is of particular importance to be able to discriminate between internalized and receptor-bound radioligand. Thus, an acid wash was performed after the incubation step to release receptor-bound radioligand before lysis of the cells. Then, both acid-releasable (receptor-bound) and internalized activity were quantified. Of the new compounds investigated, D-Lys6(Ahx[18F]FBOA)-GnRH-I showed the highest internalization into EFO-27 cells. However, its cellular uptake was slightly lower than that of the reference [125I]Triptorelin (85.8 ( 16.0% and 84.9 ( 5.6% of the internalization found for [125I]Triptorelin in the same experiment at 60 and 120 min, respectively) (Table 1). The D-Lys6(β-Ala-[18F]FBOA)-analogue, however, showed less than half of the internalization found for [125I]Triptorelin, and for D-Lys6([68Ga]DOTA)-GnRH-I almost no internalization was detected. As already observed in the initial uptake studies (serum depencence of GnRHR-I expression), nonspecific receptor binding and internalization in the presence of an excess of unlabeled Triptorelin (5 µM) was high. For the two 18F-labeled analogues, only approx 10% of the acid-releasable activity was specifically bound to the cell membrane at 60 min ([125I]Triptorelin: 17%), and only 20-30% ([125I]Triptorelin: 37%) of the internalized activity was found to be taken up specifically. In the case of D-Lys6([68Ga]DOTA)-GnRH-I, no specific receptor binding or internalization was observed.



Table 1. Internalization of the Different Radiolabeled GnRH-I-Analogues into GnRHR-Expressing EFO-27 Ovary Carcinoma Cellsa peptide

internalization [% of [125I]Triptorelin]

[125I]Triptorelin 6 (Ahx-[18F]FBOA)-GnRH-I 6 18 D-Lys (β-Ala-[ F]FBOA)-GnRH-I 6 68 D-Lys ([ Ga]DOTA)-GnRH-I

100 86 ( 16 42 ( 3 2.6 ( 1.0

D-Lys

a Cells were coincubated with both the 18F- or 68Ga-labeled compound and the reference [125I]Triptorelin for 60 min at 37° C (n ) 3). Data are given in % of the internalization of the reference [125I]Triptorelin in the same experiment and are means ( SD (triplicate determination).

Table 2. Binding Affinities of the Reference Triptorelin and the Three New GnRH-I-Analogues to the Human GnRHRa peptide

IC50 [nM]

Triptorelin 6 (Ahx-FBOA)-GnRH-I 6 D-Lys (β-Ala-FBOA)-GnRH-I 6 D-Lys (Ga-DOTA)-GnRH-I

0.13 ( 0.01 0.50 ( 0.08 0.86 ( 0.13 13.27 ( 1.01

D-Lys

a

Competitive binding studies were performed using a membrane preparation of GnRHR-expressing cells (10 µg membrane/sample) with [125I]Triptorelin as the radioligand (n ) 3 for each peptide concentration). Experiments were carried out in triplicate. Data are means ( SD.

Figure 5. Stability of D-Lys6(Ahx-[18F]FBOA)-GnRH-I in human serum. Data represent the amount of intact tracer found in serum after incubation at 37 °C for 30 to 240 min. A one-phase exponential fit yielded a calculated serum half-life of 254 ( 18 min for D-Lys6-(Ahx[18F]FBOA)-GnRH-I. Table 3. Biodistribution of [125I]Triptorelin in OVCAR-3 Tumor Bearing Nude Mice 60 min p.i. (n ) 3)a

a

Figure 4. Exemplary IC50-curves obtained for the different new GnRH-I analogues and the reference Triptorelin using GnRHR-membrane preparations (10 µg/well, n ) 3; error bars are omitted for better clarity) and [125I]Triptorelin as the radioligand. Data represent one of three separate determinations.

The internalization data for D-Lys6(Ahx-[18F]FBOA)-GnRHI were also confirmed in the ovarian cancer cell line OVCAR3, which in contrast to the EFO-27 cell line is a well-established tumor model and was chosen as a model for the potential future in vivo evaluation of the new GnRH-I analogues. In OVCAR-3 cells, D-Lys6(Ahx-[18F]FBOA)-GnRH-I showed 85.3 ( 15.8% and 89.1 ( 15.0% of the internalization of [125I]Triptorelin at 60 and 120 min, respectively. In OVCAR-3 cells, however, the fraction of nonspecifically bound/internalized ligand (approx 90% in both cases) is even higher than in EFO-27 cells. Determination of IC50. The GnRH-receptor affinities of Triptorelin, D-Lys6(Ahx-FBOA)-GnRH-I, D-Lys6(β-Ala-FBOA)GnRH-I, and D-Lys6(Ga-DOTA)-GnRH-I were determined using a commercially available membrane preparation of GnRHR-I-expressing cells and [125I]Triptorelin as the radioligand. Results are summarized in Table 2. While the Ga-DOTAanalogue shows a loss of binding affinity by a factor of ∼100 compared to the reference Triptorelin, both FBOA-analogues display only slightly reduced receptor affinities, which are still in the subnanomolar range (Figure 4). The order of affinities, however, reflects the trend already observed for the internalization efficiencies of the respective GnRHR-ligands, i.e., Trip-

organ

activity accumulation [%iD/g]

blood heart lung liver intestine stomach kidney spleen pancreas muscle OVCAR-3 tumor

0.96 ( 0.13 0.54 ( 0.01 1.49 ( 0.28 2.39 ( 1.03 13.56 ( 3.57 2.64 ( 0.64 2.82 ( 0.67 0.54 ( 0.10 0.52 ( 0.10 0.22 ( 0.05 0.49 ( 0.05

Data are given in % injected dose per gram tissue (%iD/g).

torelin > D-Lys6(Ahx-FBOA)-GnRH-I > D-Lys6(β-Ala-FBOA)GnRH-I . D-Lys6(Ga-DOTA)-GnRH-I. Serum Stability. The new GnRH-I-analogue with the highest receptor affinity and internalization efficiency, D-Lys6(Ahx[18F]FBOA)-GnRH-I, was investigated with respect to its stability in human serum (Figure 5). The radiolabeled peptide was incubated with human serum for up to 120 min, followed by extraction of the radioactivity using a RP-18 cartridge. The cartridge eluate was analyzed using RP-HPLC. Over time, an increasing amount of a hydrophilic metabolite with no retention on the RP-18-material of the extraction cartridge was formed (5.4%, 9.9%, 14.4%, and 29.3% of the total activity applied to the cartridge after 10, 30, 60, and 120 min incubation, respectively). This metabolite was always eluted in the washing step with water and was not further characterized. At all time points investigated, however, the activity immobilized on the cartridge (elution with a slightly acidified ethanolic solution) was analyzed using RP-HPLC and was found to consist exclusively of intact tracer. Formation of further metabolites was not observed. The decrease of the amount of intact D-Lys6(Ahx18 [ F]FBOA)-GnRH-I over time is illustrated in Figure 5. A onephase exponential fit of the data yielded a serum half-life of 254 ( 18 min for this ligand. In Vivo Biodistribution Study. To evaluate the potential of OVCAR-3 xenografts as an in vivo model for GnRHR-imaging, a preliminary biodistribution study for the reference ligand [125I]Triptorelin was performed in OVCAR-3 tumor bearing nude mice. Data are summarized in Table 3. The highest accumulation of [125I]Triptorelin was observed in the organs of the gastrointestinal tract, probably due to a predominant hepatobiliary clearance of the tracer, and in the kidneys. Unfortunately, uptake of [125I]Triptorelin in the


Figure 6. Tumor to organ ratios for [125I]Triptorelin in OVCAR-3 tumor bearing nude mice at 60 min p.i.. White bars represent the control experiment (tracer only, n ) 3); gray bars represent blocking conditions (coinjection of 20 µg Triptorelin, n ) 3).

OVCAR-3 tumors was as low as nonspecific tracer accumulation in heart, spleen, and pancreas (∼0.5%iD/g). This leads to very low tumor-to-organ ratios of 0.51, 0.21, 0.04, 0.17, and 2.22 for blood, liver, intestine, kidney, and muscle, respectively (Figure 6). Furthermore, a blocking study was performed, in which 20 µg of unlabeled Triptorelin was coinjected. In this study, absolute tracer accumulation in all organs was approx threefold higher than observed in the control experiment (data not shown), which might be occasioned by a slowed tracer excretion due to some pharmacological effect caused by Triptorelin. However, tumor to organ ratios were nearly identical to those found in the control experiment (Figure 6), demonstrating that tracer accumulation in OVCAR-3 tumors is not receptor-mediated.

DISCUSSION Due to their high incidence on human tumors of the reproductive system, GnRH-receptors have been successfully used for direct targeting with potent GnRH-analogues, leading to tumor growth suppression (24). However, to our knowledge only one attempt to exploit this receptor system as a target for in vivo peptide receptor imaging has been reported to date (29). One possible reason for this circumstance is the fact that almost no data are available concerning the level of GnRHreceptor protein expression in human tumor cell lines, since for most studies investigating, e.g., GnRH receptor function transiently transfected cells are used (45, 47), and receptor expression in cancer cell lines or tumor specimens is almost exclusively determined using mRNA blotting techniques or hormonal response studies (34). For the evaluation of GnRH-radiotracers, however, receptor expression on the respective human tumor cell lines is a crucial quantity, because high cellular radioligand uptake and thus the possibility for high-contrast GnRHRimaging depends on high intrinsic membrane presentation of the target receptor. Furthermore, the capacity of the receptor to internalize the radioligand is helpful to achieve high activity accumulation in tumor tissue. Among all G-protein coupled receptors that have been cloned to date, the type I mammalian GnRHR is the only receptor fully lacking the C-terminal cytoplasmatic tail, which is highly important for effective receptor internalization (45, 46). Nevertheless, the human GnRHR-I is internalized (45, 48), albeit slowly. In a very recent study, it has been demonstrated that this internalization is constitutive and not–as previously assumed–agonist-induced, and also cannot be stimulated above constitutive level in the presence of various GnRH-analogues (49). In contrast, the type II mammalian and nonmammalian

Schottelius et al.

GnRHRs, which possess intracellular C-tails of varying length, are rapidly internalized (45, 46). From this point of view, the type II GnRHR seems more suitable for targeting with radiolabeled GnRH-analogues than the type I receptor. However, it is still a matter of ongoing debate whether a functional human GnRHR-II exists at all. Some studies suggest GnRHR-II mediated actions of ant/agonistic GnRH-I or GnRH-II analogues on GnRHR-expressing cancer cells (50, 51), others indicate ligand-selective active conformations of GnRHR-I, explaining the different ligand-induced signaling observed for GnRH-I and -II analogues (52), and others demonstrate modulation of GnRHR-I function by expression of GnRHR-II gene fragments (53). Therefore, the focus of this study was directed toward the evaluation of the GnRH-I receptor as a potential target for radiolabeled GnRH-I-analogues as well as the parallel development of high-affinity GnRH-I analogues labeled with positronemitting radionuclides (18F, 68Ga) for potential use in GnRHRimaging via PET. First, several GnRHR-expressing cell lines derived from human ovarian (EFO-27, SKOV-3, OVCAR-3), breast (MDAMB 231, SKBR-3), and prostate cancer specimens (LNCaP, DU145) were screened for GnRHR-expression using a radioligand binding assay. While EFO-27, MDA-MB 231, SKBR-3, and LNCaP cells have been reported to express GnRHR-I only (14, 30, 31, 33, 34), OVCAR-3 cells express both the human type I and type II GnRHR (30), and DU-145 and SKOV-3 cells have been shown to express the type II GnRHR (30, 32). Although, based on the argumentation above, the focus of this study was the evaluation of the GnRH-I receptor as a potential target for peptide receptor imaging, the latter two cell lines were also included in this study and screened for GnRHR-II expression using the corresponding standard radioligand D-Lys6[125I]GnRH-II. This allows comparison of two receptor subtypes, GnRHR-I and the putative GnRHR-II, with respect to the level of receptor expression and efficiency of radioligand uptake, and thus identification of the more suitable target receptor for the intended purpose. Unfortunately, GnRHR-expression and thus radioligand uptake was low in all cell lines investigated (e3.5% of the applied activity after a 120 min incubation) and independent of the receptor subtype. Furthermore, the level of nonspecific binding, which was determined in the presence of an excess of unlabeled peptide (5 µM), was >55% for all cell lines. These data are consistent with a previous study, in which GnRHR-expression on different breast cancer cell line membranes was investigated using [125I][D-Ala6-NRMeLeu7-Pro9-NEt]-GnRH-I as the radioligand (31). Similar results were obtained using membranes of SKBR-3 and MDA-MB 231 cells: very low receptor binding (∼1.5% of applied activity) and a high level (g50%) of nonspecific binding was found. Another problem encountered during the in vitro studies presented here was that absolute radioligand uptake did not correlate with the cell count in the respective assay in most cases, indicating variable levels of receptor expression or receptor presentation on the cell surface (data not shown), even when cell culture conditions were kept constant for the different assays. However, reliability of the in vitro system can be partially improved by the choice of the appropriate serum supplement used for cell culture. Substituting fetal calf serum (FCS) by a semisynthetic serum supplement, Nu-Serum, leads to an increase in the fraction of specifically bound radioligand in all cell lines, independently of the GnRH receptor type investigated, indicating enhanced receptor expression under these conditions (Figure 3). This had already been documented previously for DU-145 cells (35), and seems, based on the data obtained in this study, a suitable means to enhance and stabilize


GnRHR-expression in cell culture, thus providing reliable in vitro systems for the evaluation of new radiolabeled GnRHRligands. In this study, EFO-27 cells were chosen as the optimal in vitro model due to their comparably high and constant receptor expression. Generally, the low GnRH receptor protein expression, that has been observed in this study, is explainable by the results of recent studies elucidating the regulation of human GnRHRexpression. Obviously, a large proportion of the human GnRHR protein is inefficiently processed by the cell under physiological conditions and is retained at the site of production, the endoplasmatic reticulum, or an intracellular pool (49). This inefficient routing to the plasma membrane seems to be due to a high degree of protein misfolding which can be partly overcome by the action of so-called “pharmacoperones”, small membrane-permeable antagonistic molecules that stabilize the correct receptor conformation and thus allow efficient routing to the plasma membrane (54–56). In contrast to the GnRHRs of other species, that show high plasma membrane expression, the human receptor seems to have specifically evolved by point mutation to be inefficiently folded and thus retained intracellularly. However, despite its very low membrane expression (approx 2000 receptors/cell in MCF7 and LNCaP cells (56)) signaling of the human GnRHR is very robust, suggesting that a proportion of the intracellular receptors might traffic to the cell surface during stimulation and thus providing a rationale for the successful application of GnRH-analogues for suppression of tumor growth (24). In parallel to the evaluation of the different cell lines as potential in vitro models, three novel GnRH-I analogues were developed (Figure 1): D-Lys6([68Ga]DOTA)-GnRH-I, D-Lys6(Ahx[18F]FBOA)-GnRH-I, and D-Lys6(β-Ala-[18F]FBOA)-GnRH-I. While radiolabeling of the first was performed according to a standard protocol, the strategy for 18F-labeling via oxime formation with 4-[18F]fluorobenzaldehyde was modified as compared to a previous method (40). Instead of the deprotected aminooxy-functionalized peptides, their respective Boc-protected precursors, i.e., D-Lys6(Ahx-(Boc)AoA)-GnRH-I and D-Lys6(βAla-(Boc)AoA)-GnRH-I, were used for the labeling reaction. One substantial problem with using the deprotected peptides is the very high reactivity of the aminooxy group toward ketones and aldehydes of any kind, leading to the formation of unwanted conjugates with even traces of acetone or formaldehyde and softeners during workup or storage of the purified peptides. This problem can be avoided using the Boc-protected precursors. In this study, deprotection of the aminooxy group was carried out in situ during the labeling reaction, and the [18F]FBOA-peptides were obtained in high radiochemical yields, identical to those obtained when the deprotected precursors were used (40). It is important to note, however, that a 5-fold higher peptide concentration was needed for efficient labeling as compared to deprotected aminooxy-peptides. However, this potential drawback is balanced by the much greater ease of peptide precursor preparation and handling as well as facilitated HPLC purification of the corresponding [18F]FBOA-peptides due to a significantly lower fraction of unlabeled impurities in the case of the Bocprotected analogues. Invitroevaluationof D-Lys6([68Ga]DOTA)-GnRH-I, D-Lys6(Ahx[18F]FBOA)-GnRH-I, and D-Lys6(β-Ala-[18F]FBOA)-GnRH-I yielded unexpected results with respect to the sensitivity of the 6 D-Lys -GnRH-I scaffold toward chemical modification of the 6 D-Lys -side chain. While data from the literature had indicated that even very large molecules such as proteins (37, 39) could be conjugated to this position without significantly affecting receptor affinity, the modifications introduced in this study had substantial effects on receptor affinity and ligand internalization. Interestingly, introduction of a highly hydrophilic moiety, i.e.,


[68Ga]DOTA, on the D-Lys6-side chain led to a hundredfold decrease in receptor affinity compared to the reference peptide Triptorelin, while the effect of conjugation with the lipophilic residue [18F]FBOA via different spacers (Ahx, β-Ala) was very small, i.e., a loss of binding affinity by a factor of 4 to 6 (Table 2). These data indicate first that only lipophilic modifications are well-tolerated in this position, and second that spacer length between the D-Lys6-side chain and the lipophilic moiety has significant impact on the receptor affinity of D-Lys6-modified GnRH-I analogues. The receptor binding data obtained via membrane binding assays correlate well with the internalization data from studies using EFO-27 cells (Table 1). Receptor-mediated internalization of the [68Ga]DOTA-analogue was negligible, corresponding to the very low receptor affinity of this compound, while cellular uptake of D-Lys6(Ahx-[18F]FBOA)-GnRH-I was close to that of the reference [125I]Triptorelin, reflecting the similar receptor affinity of these radioligands. Because D-Lys6(Ahx-[18F]FBOA)-GnRH-I showed the most promising in vitro profile of the new compounds evaluated, its in vitro stability in human serum was investigated in preparation for future in vivo evaluation. The calculated serum half-life of this compound is >4 h, indicating both sufficient stability of the peptide scaffold as well as of the oxime bond between the aminooxy-functionalized peptide and 4-[18F]fluorobenzaldehyde and thus suitable stability for in vivo application for imaging purposes. Another step toward the future in vivo evaluation of D-Lys6(Ahx18 [ F]FBOA)-GnRH-I was the search for an appropriate in vivo tumor model. Because EFO-27 cells are not tumorigenic in mice, internalization of D-Lys6(Ahx-[18F]FBOA)-GnRH-I and of the standard ligand [125I]Triptorelin were also investigated in OVCAR-3 cells, a well-established GnRHR-expressing transplantable tumor model. For both peptides, absolute radioligand uptake was identical to that observed in EFO-27 cells, and the difference in internalization efficiency between the two peptides was also identical in both cell lines, indicating the suitability of OVCAR-3 cells as a transplantable tumor model for the future in vivo evaluation of new radiolabeled GnRH-I analogues. To confirm this, a preliminary biodistribution study was performed in OVCAR-3 tumor bearing nude mice. Because [125I]Triptorelin has an even higher receptor affinity than 6 18 D-Lys (Ahx-[ F]FBOA)-GnRH-I and therefore can be viewed as a “gold standard”-GnRH-ligand, [125I]Triptorelin was used as the radioligand for the evaluation of the OVCAR-3 tumor model. The comparably high lipophilicity of [125I]Triptorelin leads to predominant hepatobiliary clearance of the radioligand, as illustrated by the very high uptake in intestine and also substantial hepatic accumulation (Table 3). Because the lipophilicity of D-Lys6(Ahx-[18F]FBOA)-GnRH-I is similar to that of [125I]Triptorelin, the excretion patterns of both radioligands will probably be very similar. Consequently, a high extent of hepatobiliary excretion and thus high background activity can also be expected for the fluorinated peptide. Therefore, based on the experience from studies with different glycosylated somatostatin receptor-ligands, which have demonstrated that only ligands with a log P < 1.6 show a suitable excretion pattern for high-contrast peptide receptor imaging (57, 58), pharmacokinetic ligand optimization is warranted before performing in vivo studies with D-Lys6(Ahx-[18F]FBOA)-GnRH-I. With respect to receptor-specific tracer accumulation, the biodistribution data of [125I]Triptorelin (Table 3) parallel the in vitro data found for all GnRHR-expressing cell lines investigated in this study: the level of GnRHR-expression on the respective tumor cells seems to be very low. Tumor uptake of [125I]Triptorelin was as low as nonspecific tracer accumulation in organs such as heart, spleen, and pancreas (∼0.5%iD/g) and is even


lower than blood activity concentration at the same time point (0.96%iD/g at 60 min p.i.). A blocking study revealed the absence of receptor-mediated tumor accumulation (Figure 6). However, to confirm this observation and to be able to quantify GnRH-I-receptor expression in OVCAR-3 tumors more precisely, ex vivo GnRHR-autoradiography or immunohistochemistry after excision of the implanted tumor would be necessary. Nevertheless, this observation seriously challenges the suitability of the GnRH receptor system as a target for peptide receptor imaging, since the obvious lack of receptor-mediated tracer accumulation will not allow high-contrast imaging and sensitive detection of GnRHR-positive lesions, even if overall tracer pharmacokinetics are substantially improved.

CONCLUSION The in vitro evaluation of the three new radiolabeled GnRH-I analogues investigated in this study, D-Lys6([68Ga]DOTA)GnRH-I, D-Lys6(Ahx-[18F]FBOA)-GnRH-I, and D-Lys6(β-Ala[18F]FBOA)-GnRH-I, yielded interesting insights into the structural requirements for the development of high-affinity radiolabeled GnRH-I agonists. Obviously, in this study only the conjugation of lipophilic residues to the D-Lys6-side chain is well-tolerated by the peptide scaffold without significant loss of binding affinity, and spacer length between the D-Lys6-side chain and the hydrophobic moiety has substantial impact on the receptor affinity of the radioligand. Of the new compounds investigated, D-Lys6(Ahx-[18F]FBOA)-GnRH-I shows the highest receptor affinity and internalization efficiency into GnRHR-expressing tumor cells, bearing the highest structural homology to a previously developed high-affinity cytotoxic GnRH-I analogue. D-Lys6(Ahx[18F]FBOA)-GnRH-I illustrates again the versatility and ease of the newly developed 18F-labeling method via chemoselective oxime formation with 4-[18F]fluorobenzaldehyde, especially when the Boc-protected aminooxy-peptide is used as the labeling precursor, and represents the first radiolabeled GnRH-I analogue with suitable binding characteristics for efficient GnRHRtargeting. Unfortunately, both the in vitro radioligand binding assays performed with various GnRHR-expressing human cancer cell lines as well as the initial in vivo biodistribution study carried out using OVCAR-3 tumor-bearing nude mice indicated a very low level of GnRHR expression on the cell surface. While this obviously does not seem to hamper application of nonradiolabeled GnRH-analogues for the effective suppression of tumor growth, as demonstrated in a variety of studies, it has a deleterious impact on the suitability of the GnRHR system as a suitable target for peptide receptor imaging using radiolabeled GnRH-analogues. In our opinion, GnRH receptors represent a highly interesting target for pharmacological interventions, but do not fulfill the prerequisites for successful peptide receptor imaging, even when suitable receptor-ligands as developed in this study are available.

ACKNOWLEDGMENT This work was supported by a KKF (Kommission für Klinische Forschung am Klinikum rechts der Isar) grant (No 48-03) and by a MMW grant (Wissenschaftliches Herausgeberkollegium der Münchner Medizinischen Wochenschrift).

LITERATURE CITED (1) Ginj, M., Chen, J., Walter, M. A., Eltschinger, V., Reubi, J. C., and Maecke, H. R. (2005) Preclinical evaluation of new and highly potent analogues of octreotide for predictive imaging and targeted radiotherapy. Clin. Cancer Res. 11, 1136–1145.

Schottelius et al. (2) Esser, J. P., Krenning, E. P., Teunissen, J. J., Kooij, P. P., van Gameren, A. L., Bakker, W. H., and Kwekkeboom, D. J. (2006) Comparison of [177Lu-DOTA0,Tyr3]octreotate and [177LuDOTA0,Tyr3]octreotide: which peptide is preferable for PRRT? Eur. J. Nucl. Med. Mol. Imaging 33, 1346–1351. (3) Schottelius, M., Poethko, T., Herz, M., Reubi, J. C., Kessler, H., Schwaiger, M., and Wester, H. J. (2004) First 18F-labeled tracer suitable for routine clinical imaging of sst receptorexpressing tumors using positron emission tomography. Clin. Cancer Res. 10, 3593–3606. (4) Meisetschlager, G., Poethko, T., Stahl, A., Wolf, I., Scheidhauer, K., Schottelius, M., Herz, M., Wester, H. J., and Schwaiger, M. (2006) Gluc-Lys([18F]FP)-TOCA PET in patients with SSTRpositive tumors: biodistribution and diagnostic evaluation compared with [111In]DTPA-octreotide. J. Nucl. Med. 47, 566–573. (5) Decristoforo, C., Mather, S. J., Cholewinski, W., Donnemiller, E., Riccabona, G., and Moncayo, R. (2000) 99mTc-EDDA/ HYNIC-TOC: a new 99mTc-labelled radiopharma-ceutical for imaging somatostatin receptor-positive tumours; first clinical results and intra-patient comparison with 111In-labelled octreotide derivatives. Eur. J. Nucl. Med. 27, 1318–1325. (6) Forrer, F., Waldherr, C., Maecke, H. R., and Mueller-Brand, J. (2006) Targeted radionuclide therapy with 90Y-DOTATOC in patients with neuroendocrine tumors. Anticancer Res. 26, 703– 707. (7) Reubi, J. C. (2003) Peptide receptors as molecular targets for cancer diagnosis and therapy. Endocrine ReV. 24, 389–427. (8) Reubi, J. C., and Waser, B. (2003) Concomitant expression of several peptide receptors in neuroendocrine tumors: molecular basis for in vivo multireceptor tumour targeting. Eur. J. Nucl. Med. Mol. Imaging 30, 781–793. (9) Zhang, H., Chen, J., Waldherr, C., Hinni, K., Waser, B., Reubi, J. C., and Maecke, H. R. (2004) Synthesis and evaluation of bombesin derivatives on the basis of pan-bombesin peptides labeled with indium-111, lutetium-177, and yttrium-90 for targeting bombesin receptor-expressing tumors. Cancer Res. 64, 6707–6715. (10) Waser, B., Eltschinger, V., Linder, K., Nunn, A., and Reubi, J. C. (2007) Selective in vitro targeting of GRP and NMB receptors in human tumours with the new bombesin tracer 177LuAMBA. Eur. J. Nucl. Med. Mol. Imaging 34, 95–100. (11) Nock, B. A., Maina, T., Behe, M., Nikolopoulou, A., Gotthardt, M., Schmitt, J. S., Behr, T. M., and Macke, H. R. (2005) CCK-2/gastrin receptor-targeted tumor imaging with 99m Tc-labeled minigastrin analogs. J. Nucl. Med. 46, 1727–1736. (12) Wild, D., Be´he´, M., Wicki, A., Storch, D., Waser, B., Gotthardt., M., Keil, B., Christofori, G., Reubi, J. C., and Ma¨cke, H. R. (2006) [Lys40(Ahx-DTPA-111In)NH2]Exedin-4, a very promising ligand for glucagon-like peptide-1 (GLP-1) receptor targeting. J. Nucl. Med. 47, 2025–2033. (13) Schally, A. V. (1999) Luteinizing hormone-releasing hormone analogs: their impact on the control of tumorigenesis. Peptides 20, 1247–1262. (14) Halmos, G., Arencibia, J. M., Schally, A. V., Davis, R., and Bostwick, D. G. (2000) High Incidence of receptors for luteinizing hormone releasing hormone (LHRH) and LHRH receptor gene expression in human prostate cancers. J. Urol. 163, 623– 629. (15) Pati, D., and Habibi, H. R. (1995) Inhibition of human hepatocarcinoma cell proliferation by mammalian and fish gonadotropin-releasing hormones. Endocrinology 136, 75–84. (16) Krebs, L. J., Wang, X., Nagy, A., Schally, A. V., Prasad, P. N., and Liebow, C. (2002) A conjugate of doxorubicin and an analog of Luteinizing Hormone-Releasing Hormone shows increased efficacy against oral and laryngeal cancers. Oral Oncol. 38, 657– 663. (17) Friess, H., Buchler, M., Kiesel, L., Kruger, M., and Beger, H. G. (1991) LH-RH receptors in the human pancreas. Basis for antihormonal treatment in ductal carcinoma of the pancreas. Int. J. Pancreatol. 10, 151–159.

GnRH-I Analogues with High GnRHR-Targeting Efficiency (18) Sion-Vardi, N., Kaneti, J., Segal-Abramson, T., Giat, J., Levy, J., and Sharoni, Y. (1992) Gonadotropin-releasing hormone specific binding sites in normal and malignant renal tissue. J. Urol. 148, 1568–1570. (19) van Groeninghen, J. C., Kiesel, L., Winkler, D., and Zwirner, M. (1998) Effects of luteinizing hormone-releasing hormone on nervous system tumours. Lancet 352, 372–373. (20) Chen, A., Ganor, Y., Rahimipour, S., Ben-Aroya, N., Koch, Y., and Levite, M. (2002) The neuropeptides GnRH-II and GnRH-I are produced by human T cells and trigger laminin receptor gene expression, adhesion, chemotaxis and homing to specific organs. Nat. Med. 8, 1421–1426. (21) Keller, G., Schally, A. V., Gaiser, T., Nagy, A., Baker, B., Halmos, G., and Engel, J. B. (2005) Receptors for luteinizing hormone releasing hormone (LHRH) expressed in human nonHodgkin’s lymphomas can be targeted for therapy with the cytotoxic LHRH analogue AN-207. Eur. J. Cancer 41, 2196– 2202. (22) Moretti, R. M., Montangnani-Marelli, M., van Groeninghen, J. C., and Limonta, P. (2002) Locally expressed GnRH receptors mediate the oncostatic and antimetastatic activity of GnRH agonists on melanoma cells. J. Clin. Endocrin. Metab. 87, 3791– 3797. (23) Keller, G., Schally, A. V., Gaiser, T., Nagy, A., Baker, B., Westphal, G., Halmos, G., and Engel, J. B. (2005) Human malignant melanomas express receptors for luteinizing hormone releasing hormone allowing targeted therapy with cytotoxic luteinizing hormone releasing hormone analogue. Cancer Res. 65, 5857–5863. (24) Montagnani-Marelli, M., Moretti, R. M., JanuszkiewiczCaulier, J., Motta, M., and Limonta, P. (2006) Gonadotropinreleasing hormone (GnRH) receptors in tumors: a new rationale for the therapeutical application of GnRH analogs in cancer patients? Curr. Cancer Drug Targets 6, 257–269. (25) Nagy, A., Schally, A. V., Armatis, P., Szepeshazi, K., Halmos, G., Kovacs, M., Zarandi, M., Groot, K., Miyazaki, M., Jungwirth, A., and Horvath, J. (1996) Cytotoxic analogs of luteinizing hormone-releasing hormone containing doxorubicin or 2-pyrrolinodoxorubicin, a derivative 500-1000 times more potent. Proc. Natl. Acad. Sci. U.S.A. 93, 7269–7273. (26) Nagy, A., and Schally, A. V. (2005) Targeting of cytotoxic luteinizing hormone-releasing hormone analogs to breast, ovarian, endometrial, and prostate cancers. Biol. Reprod. 73, 851–859. (27) Keller, G., Schally, A. V., Gaiser, T., Nagy, A., Baker, B., Halmos, G., and Engel, J. B. (2005) Receptors for luteinizing hormone releasing hormone expressed on human renal cell carcinomas can be used for targeted chemotherapy with cytotoxic luteinizing hormone releasing hormone analogues. Clin. Cancer Res. 11, 5549–5557. (28) Harrison, G. S., Wiermann, M. E., Nett, T. M., and Glode, L. M. (2004) Gonadotropin-releasing hormone and its receptor in normal and malignant cells. Endocrine-Related Cancer 11, 725–748. (29) Barda, Y., Cohen, N., Lev, V., Ben-Aroya, N., Koch, Y., Mishani, E., Fridkin, M., and Gilon, C. (2004) Backbone metal cyclization: novel 99mTc labeled GnRH analog as potential SPECT molecular imaging agent in cancer. Nucl. Med. Biol. 31, 921–933. (30) Gru¨ndker, C., Schlotawa, L., Viereck, V., Eicke, N., Horst, A., Kairies, B., and Emons, G. (2004) Antiproliferative effects of the GnRH antagonist Cetrorelix and of GnRH-II on human endometrial and ovarian cancer cells are not mediated through the GnRH type I receptor. Eur. J. Endocrinol. 151, 141–149. (31) Eidne, K. A., Flanagan, C. A., Harris, N. S., and Millar, R. P. (1987) Gonadotropin-releasing hormone (GnRH)-binding sites in human breast cancer cell lines and inhibitory effects of GnRHAntagonists. J. Clin. Endocrinol. Metab. 64, 425–432. (32) Enomoto, M., Endo, D., Kawashima, S., and Park, M. K. (2004) Human Type II GnRH receptor mediates effects of GnRH on cell proliferation. Zool. Sci. 21, 763–770.

Bioconjugate Chem., Vol. 19, No. 6, 2008 1267 (33) Bahk, J. Y., Hyun, J. S., Lee, H., Kim, M. O., Cho, G. J., Lee, B. H., and Choi, W. S. (1998) Expression of gonadotropinreleasing hormone (GnRH) and GnRH receptor mRNA in prostate cancer cells and effect of GnRH on the proliferation of prostate cancer cells. Urol. Res. 26, 259–264. (34) Hapgood, J. P., Sadie, H., van Biljon, W., and Ronacher, K. (2005) Regulation of expression of gonadotrophin-releasing hormone receptor genes. J. Neuroendocrinol. 17, 619–638. (35) Enomoto, M., Mori, T., and Park, M. K. (2002) Influence of serum supplements in culture medium on gonadotrolin-releasing hormone effects on colony formation. Life Sci. 71, 2153–2160. (36) So¨derha¨ll, J. A., Polymeropoulos, E. E., Paulini, K., Gu¨nther, E., and Ku¨hne, R. (2005) Antagonist and agonist binding models of the human gonadotropin-releasing hormone receptor. Biochem. Biophys. Res. Commun. 333, 568–582. (37) Beckers, T., Bernd, M., Kutscher, B., Ku¨hne, R., Hoffmann, S., and Reissmann, T. (2001) Structure-Function Studies of linear and cyclized peptide antagonists of the GnRH receptor. Biochem. Biophys. Res. Commun. 289, 653–663. (38) Letsch, M., Schally, A. V., Szepeshazi, K., Halmos, G., and Nagy, A. (2003) Preclinical evaluation of targeted cytotoxic luteinizing hormone releasing hormone analogue AN-152 in androgen-sensitive and insensitive prostate cancers. Clin. Cancer Res. 9, 4505–4513. (39) Yang, W. H., Wieczorck, M., Allen, M. C., and Nett, T. M. (2003) Cytotoxic activity of gonadotropin-releasing hormone (GnRH)-pokeweed antiviral protein conjugates in cell lines expressing GnRH receptors. Endocrinology 144, 1456–1463. (40) Poethko, T., Schottelius, M., Thumshirn, G., Hersel, U., Herz, M., Henriksen, G., Kessler, H., Schwaiger, M., and Wester, H. J. (2004) Two-step methodology for high-yield routine radiohalogenation of peptides: 18F-labelled RGD- and octreotide analogs. J. Nucl. Med. 45, 892–902. (41) Schottelius, M., Schwaiger, M., and Wester, H. J. (2003) Rapid and high-yield solution phase synthesis of DOTA-Tyr3-octreotide and DOTA-Tyr3-octreotate using unprotected DOTA. Tetrahedron Lett. 44, 2393–2396. (42) Shin, I., and Park, K. (2003) Solution phase synthesis of aminooxy peptoids in the C to N and N to C directions. Org. Lett. 4, 869–872. (43) Buchwald, P., and Bodor, N. (1998) Octanol-water partition of nonzwitterionic peptides: predictive power of a molecular sizebased model. Proteins: Struct., Funct. Genet. 30, 86–99. (44) Segal-Abramson, T., Kitroser, H., Levy, J., Schally, A. V., and Sharoni, Y. (1992) Direct effects of luteinizing hormonereleasing hormone agonists and antagonists on MCF-7 mammary cancer cells. Proc. Natl. Acad. Sci. U.S.A. 89, 2336–2339. (45) Caunt, C. J., Finch, A. R., Sedgley, K. R., Oakley, L., Luttrell, L. M., and McArdle, C. A. (2006) Arrestin-mediated ERK activation by gonadotropin-releasing hormone receptors: receptorspecific activation mechanisms and compartmentalization. J. Biol. Chem. 281, 2701–2710. (46) Millar, R. P., Lu, Z. L., Pawson, A. J., Flanagan, C. A., Morgan, K., and Maudsley, S. R. (2004) Gonadotropin-releasing hormone receptors. Endocr. ReV. 25, 235–275. (47) Hislop, J. N., Madziva, M. T., Everest, H. M., Harding, T., Uney, J. B., Willars, G. B., Millar, R. P., Troskie, B. E., Davidson, J. S., and McArdle, C. A. (2000) Desensitization and internalization of human and xenopus gonadotropin-releasing hormone receptors expressed in alphaT4 pituitary cells using recombinant adenovirus. Endocrinology 141, 4564–4575. (48) Vreci, M., Anderson, L., Hanyaloglu, A., McGregor, A. M., Groarke, A. D., Milligan, G., Taylor, P. L., and Eidne, K. A. (1998) Agonist-induced endocytosis and recycling of the gonadotropin-releasing hormone receptor: effect of beta-arrestin in internalization kinetics. Mol. Endocrinol. 12, 1818–1829. (49) Pawson, A. J., Faccenda, E., Maudsley, S., Lu, Z. L., Naor, Z., and Millar, R. P. (2008) Mammalian type I Gonadotropinreleasing hormone receptors undergo slow, constitutive, agonistindependent internalization. Endocrinology 149 (3), 1415–1422.

1268 Bioconjugate Chem., Vol. 19, No. 6, 2008 (50) Grundker, C., Schlotawa, L., Viereck, V., Eicke, N., Horst, A., Kairies, B., and Emons, G. (2004) Antiproliferative effects of the GnRH antagonist cetrorelix and of GnRH-II on human endometrial and ovarian cancer cells are not mediated through the GnRH type I receptor. Eur. J. Endocrinol. 151, 141–149. (51) Fister, S., Gunthert, A. R., Emons, G., and Grundker, C. (2007) Gonadotropin-releasing hormone type II antagonists induce apoptotic cell death in human endometrial and ovarian cancer cells in vitro and in vivo. Cancer Res. 67, 1750–1756. (52) Lu, Z. L., Gallagher, R., Sellar, R., Coetsee, M., and Millar, R. P. (2005) Mutations remote from the human gonadotropinreleasing hormone (GnRH) receptor-binding sites specifically increase binding affinity for GnRH II but not GnRH I: evidence for ligand-selective, receptor-active conformations. J. Biol. Chem. 280, 29796–29803. (53) Pawson, A. J., Maudsley, S., Morgan, K., Davidson, L., Naor, Z., and Millar, R. P. (2005) Inhibition of human type I gonadotropin-releasing hormone receptor (GnRHR) function by expression of a human type II GnRHR gene fragment. Endocrinology 146, 2639–2649. (54) Ulloa-Aguirre, A., Janovick, J. A., Brothers, S. P., and Conn, P. M. (2004) Pharmacologic rescue of conformationally-defective

Schottelius et al. proteins: implications for the treatment of human disease. Traffic 5, 821–837. (55) Janovick, J. A., Knollman, P. E., Brothers, S. P., Ayala-Yanez, R., Aziz, A. S., and Conn, P. M. (2006) Regulation of G proteincoupled receptor trafficking by inefficient plasma membrane expression: molecular basis of an evolved strategy. J. Biol. Chem. 281, 8417–8425. (56) Sedgley, K. R., Finch, A. R., Caunt, C. J., McArdle, C. A. Intracellular gonadotropin-releasing hormone receptors in breast cancer and gonadotrope lineage cells, J. Endocrinol. 191, 625636. (57) Schottelius, M., Wester, H. J., Reubi, J. C., SenekowitschSchmidtke, R., and Schwaiger, M. (2002) Improvement of pharmacokinetics of radioiodinated Tyr3-octreotide by conjugation with carbohydrates. Bioconjugate Chem. 13, 1021–1030. (58) Wester, H. J., Schottelius, M., Poethko, T., Bruus-Jensen, K., and Schwaiger, M. (2004) Radiolabelled carbohydrated somatostatin analogs: a brief review of the current status. Cancer Biother. Radiopharm. 19, 231–244. BC800058K

Development of Novel 68Ga- and 18F-Labeled GnRH-I Analogues

Recommend Documents