IIIa Receptor

Oct 15, 1995 - Shuang Liu,* D. Scott Edwards,* Richard J. Looby, Anthony R. Harris, Michael J. Poirier,. John A. Barrett, Stuart J. Heminway, and Timo...
0 downloads 0 Views 349KB Size
Bioconjugate Chem. 1996, 7, 63−71

63

Labeling a Hydrazino Nicotinamide-Modified Cyclic IIb/IIIa Receptor Antagonist with 99mTc Using Aminocarboxylates as Coligands Shuang Liu,* D. Scott Edwards,* Richard J. Looby, Anthony R. Harris, Michael J. Poirier, John A. Barrett, Stuart J. Heminway, and Timothy R. Carroll The DuPont Merck Pharmaceutical Company, Radiopharmaceuticals Division, 331 Treble Cove Road, North Billerica, Massachusetts 01862. Received March 23, 1995X

A series of 99mTc complexes containing a hydrazinonicotinamide-conjugated cyclic IIb/IIIa receptor antagonist, cyclo(D-Val-NMeArg-Gly-Asp-Mamb-(hydrazinonicotinyl-5-(6-aminocaproic acid))), were synthesized in high yield using tricine or other aminocarboxylates as coligands. These 99mTc complexes have the potential to be used as thrombus imaging agents. The radiolabeling of the HYNIC-conjugated cyclic IIb/IIIa peptide (HYNICtide) was carried out by reaction with pertechnetate in the presence of excess tricine and stannous chloride at pH 4-5. The reaction time and temperature depend on the amount of the HYNICtide and pertechnetate used for the radiolabeling. Very high specific activity (g20 000 mCi/µmol) can be achieved for the complex [99mTc(HYNICtide)(tricine)2] without postlabeling purification. The complex [99mTc(HYNICtide)(tricine)2] was found by two reversed phase HPLC methods to exist as multiple species, some of which interconvert, depending on the temperature, reaction time, and pH of the reaction mixture. The presence of these multiple species is most likely due to different bonding modalities of either the hydrazine moiety of the HYNICtide or the two tricine coligands. The complex [99mTc(HYNICtide)(EDDA)] (EDDA ) ethylenediamine-N,N′-diacetic acid) was prepared either by reacting the cyclic IIb/IIIa HYNICtide with pertechnetate, excess EDDA, and stannous chloride at pH 4-5 and 75 °C for 30 min or by reacting excess EDDA with [99mTc(HYNICtide)(tricine)2]. The complex [99mTc(HYNICtide)(EDDA)] was found to be stable for at least 12 h in the reaction mixture. Three major species were detected in the radio-HPLC chromatograms, presumably due to the more limited number of possible coordination isomers. Similar results were obtained using other polydentate aminocarboxylates (such as HEDTA, N-(2-hydroxyethyl)ethylenediaminetriacetic acid) as coligands. It is clear that the replacement of tricine by other polydentate aminocarboxylates produces 99mTc-HYNICtide complexes with higher stability and fewer coordination isomers.

INTRODUCTION

There is currently considerable interest in labeling proteins and small peptides with 99mTc for the development of target specific imaging agents (1-4). In the last decade, a large number of radiolabeling techniques have been developed and extensively reviewed (5-10). They can be classified into three main categories: direct labeling, the preformed chelate approach, and the indirect labeling approach. The direct labeling approach uses a reducing agent to convert a number of disulfide linkages in a protein into free thiols, which are able to bind the Tc very efficiently. The advantage of this approach is that it is easy to carry out. However, very little is known about the coordination chemistry of the Tc. There is little control over the stability of the 99mTc complex or the nonspecific binding. In addition, this method applies only to proteins or their fragments because many small peptides do not have any disulfide bonds, or in some cases the disulfide bond is too critical for maintaining their biological properties to be reduced. The preformed chelate approach involves formation of the 99mTc complex with a bifunctional coupling agent (BFCA) and conjugation of the 99mTc-BFCA complex to a protein or peptide in a separate step on the tracer level. In this approach, the chemistry is better defined, and the peptide or protein is not exposed to the sometimes forcing * To whom correspondence should be addressed. Tel: (508) 671-8696 (S.L.) or (508) 671-8311 (D.S.E.). Fax: (508) 436-7500. X Abstract published in Advance ACS Abstracts, October 15, 1995.

1043-1802/96/2907-0063$12.00/0

conditions used to prepare the complex. However, the multistep tracer level synthesis limits its clinical application. In the indirect labeling approach, a BFCA is first attached to the peptide or protein to form a BFCApeptide (protein) conjugate. The radiolabeling can be achieved either by direct reduction of 99mTcO4- in the presence of BFCA-peptide (protein) conjugate or by ligand exchange with a intermediate 99mTc complex such as 99mTc-glucoheptonate. This approach combines the ease of direct labeling with the well-defined chemistry of the preformed chelate approach. Therefore, the indirect labeling is the most practical approach for the development of peptide-based target specific radiopharmaceuticals. We have been actively pursuing a research program toward developing thrombus imaging agents based on cyclic IIb/IIIa receptor antagonists (11-13). Since these small peptides are very potent, the 99mTc labeling must be accomplished with high specific activity. Therefore, we are particularly interested in a BFCA, which is able to form stable 99mTc complexes in high yield at very low concentrations of the BFCA-peptide conjugate. Recently, Abrams and co-workers reported the use of aromatic hydrazines in labeling polyclonal IgG with 99mTc for imaging focal sites of infection (14, 15). The HYNIC group is of particular interest because it can be easily labeled with high efficiency (rapid and high yield radiolabeling), and the resulting complexes have high in vivo stability (16). Since the HYNIC group can only occupy one or two sites in the technetium coordination sphere, a coligand such as glucoheptonate is required. It © 1996 American Chemical Society

64 Bioconjugate Chem., Vol. 7, No. 1, 1996

Figure 1. Cyclic IIb/IIIa HYNICtide.

has been reported that when glucoheptonate is used as coligand, a specific activity of >100 000 mCi/µmol can be obtained after HPLC purification on a routine basis in labeling chemotactic peptides with 99mTc (4, 16). It was also found that the nature of the coligand has significant effects on the biodistribution of the radiolabeled chemotactic peptides (17). Tricine, tris(hydroxymethyl)methylglycine, has also been reported as a coligand in labeling HYNIC-conjugated polyclonal IgG with 99mTc (18, 19). The specific activity (>150 mCi/mg protein) of the radiolabeled HYNIC-IgG conjugate using tricine as coligand is higher than that (18 MΩ quality. Synthesis and biological properties of the cyclic IIb/IIIa receptor HYNICtide will be reported elsewhere (20). Analytical Methods. Three radio-HPLC methods were used. The TFA method used a Hewlett-Packard Model 1050 instrument and a Vydac C18 column (4.6 mm × 25 cm) at a flow rate of 1.0 mL/min with a gradient mobile phase from 100% A (0.1% TFA in H2O) to 100% B (0.1% TFA in acetonitrile) over 25 min. The phosphate method used a Hewlett-Packard Model 1050 instrument and a Vydac C18 column (4.6 mm × 25 cm) at a flow rate of 1.0 mL/min with a gradient mobile phase from 100% A (0.01 M pH 6 phosphate buffer) to 30% B (acetonitrile) at 15 min and 75% B at 25 min. The ion-pair method used a Hewlett-Packard Model 1090 HPLC instrument and a Zorbax Rx C18 column (4.6 mm × 25 cm) at a flow

Liu et al.

rate of 1.0 mL/min with a gradient mobile phase from 100% A (95:5 (0.025 M pH 3.7 phosphate buffer/0.005 M TBAP):acetonitrile) to 90% A and 10% B (20:80 (0.025 M pH 3.7 phosphate buffer/0.005 M tetrabutylammonium phosphate):acetonitrile) at 20 min, 40% B at 30 min and 100% B at 33 min. The ITLC method used Gelman Sciences silica gel paper strips and 1:1 mixture of acetone and saline as eluant. N,N′-Et2EDDA‚2HCl. To a suspension of potassium carbonate (15 g, 0.11 mol) in dry acetonitrile (150 mL) was added N,N′-diethylethylenediamine (5.8 g, 0.05 mol) followed by tert-butyl bromoacetate (20 g, 0.10 mol). The reaction mixture was stirred at room temperature for 24 h. Solvent was removed under reduced pressure, and the residue was extracted with dichloromethane (2 × 150 mL). The organic phases were combined, washed with H2O (2 × 50 mL), and dried over anhydrous MgSO4. Removal of dichloromethane afforded a pale yellow liquid. The liquid (tert-butyl ester of Et2EDDA) was dissolved in concentrated HCl (100 mL) and the resulting solution stirred at room temperature for 5-6 h. Solvent was removed under the reduced pressure to give a white solid, which was dried under vacuum overnight. The yield was 10.2 g (68%). Anal. Calcd (found) for C10H20N2O4‚2HCl‚ 1.5H2O: C, 36.15 (36.35); H, 7.58 (7.39); N, 8.43 (8.54). IR (cm-1, KBr disk): 3500-2000 (br, νΟΗ), 1740 (vs, νCdO). FAB-MS: m/z ) 233 (M + 1, [C10H21N2O4]+). 1H NMR (D2O): 1.20 (t, 6H, CH3, 3JH-H ) 7.3 Hz); 3.29 (q, 4H, CH2, 3JH-H ) 7.2 Hz); 3.65 (s, 4H, CH2CH2); and 4.07 (s, 4H, CH2COOH). N-MeEDTA‚2HCl. N-Methylethylenediaminetriacetic acid dihydrochloride (N-MeEDTA‚2HCl) was prepared in a fashion similar to that for Et2EDDA‚2HCl using N-methylethylenediamine (1.5 g, 0.02 mol), potassium carbonate (8.5 g, 0.06 mol), and tert-butyl bromoacetate (12 g, 0.06 mol). The yield was 5.6 g (87%). Anal. Calcd (found) for C9H16N2O6‚2HCl‚0.25C4H9OH: C, 35.36 (35.70); H, 6.08 (5.87); N, 8.25 (8.00). IR (cm-1, KBr disk): 35002000 (br, νΟΗ), 1740 (vs, νCdO). FAB-MS: m/z ) 249 (M + 1, [C9H17N2O6]+). 1H NMR (D2O): 2.88 (s, 3H, CH3); 3.56 (m, 4H, CH2CH2); and 3.85-4.20 (m, 6H, CH2COOH). N-BzlEDTA‚2HCl. N-Benzylethylenediaminetriacetic acid dihydrochloride (N-BzlEDTA‚2HCl) was prepared similarly to Et2EDDA‚2HCl using N-benzylethylenediamine (3.0 g, 0.02 mol), potassium carbonate (8.5 g, 0.06 mol), and tert-butyl bromoacetate (12 g, 0.06 mol). The yield was 7.0 g (88%). Anal. Calcd (found) for C15H20N2O6‚2HCl: C, 45.35 (45.42); H, 5.58 (5.58); N, 7.05 (7.00). IR (cm-1, KBr disk): 3500-2000 (br, νΟΗ), 1745 (vs, νCdO). FAB-MS: m/z ) 325 (M + 1, [C15H21N2O6]+). 1H NMR (D O): 3.30, 3.50 (m, 4H, CH CH ); 3.67 (s, 4H, 2 2 2 CH2COOH); 4.08 (s, 2H, CH2COOH); 4.43 (s, 2H, PhCH2); and 7.42 (m, aromatic H). Radiolabeling of Cyclic IIb/IIIa HYNICtide Using Tricine as Coligand in One Step. To a 10 mL vial was added 0.5 mL of 99mTcO4- solution (20-200 mCi/mL) in saline, followed by 0.5 mL of tricine solution (20-75 mg/mL, pH ∼5.0) in H2O, 0.1 mL of HYNICtide solution (5-500 µg/mL) in H2O, and 20 µL of SnCl2.2H2O solution (0.5-5.0 mg/mL) in 0.1 N HCl. The reaction mixture was allowed to stand at room temperature for 20-60 min depending on the amount of HYNICtide used for the radiolabeling and then analyzed by radio-HPLC. ITLC was used to determine the amount of 99mTc-colloid. Radiolabeling of Cyclic IIb/IIIa HYNICtide Using Tricine as Coligand in Two Steps. To a 10 mL vial was added 0.5 mL of 99mTcO4- solution (20-200 mCi/mL) in saline, followed by 0.5 mL of tricine solution (20-75 mg/mL, pH ∼5.0) in H2O, and 10 µL of SnCl2‚2H2O

99mTc

Bioconjugate Chem., Vol. 7, No. 1, 1996 65

Label for Cyclic IIb/IIIa Receptor Antagonist

solution (10 mg/mL) in 1 N HCl. After the mixture was allowed to stand at room temperature for 10 min, 0.1 mL of HYNICtide solution (100 µg/mL) in H2O was added. Upon standing at room temperature for another 30-60 min, the reaction mixture was analyzed by radio-HPLC and ITLC. The complex prepared in two steps was found to be the same as that prepared in one step. Radiolabeling of Cyclic IIb/IIIa HYNICtide Using EDDA as Coligand in One Step. To a 10 mL vial was added 0.4 mL of 99mTcO4- solution (50-125 mCi/mL) in saline, followed by 0.5 mL of EDDA solution (10 mg/mL, pH 7-7.5) in H2O, 0.1 mL of HYNICtide solution (100 µg/mL) in H2O, and 10 µL of SnCl2‚2H2O solution (10 mg/ mL) in 1 N HCl. The reaction mixture was heated at 75 °C for 30 min and then analyzed by radio-HPLC and ITLC. The same procedure was used for the N-substituted EDDA analogs. Radiolabeling of Cyclic IIb/IIIa HYNICtide Using EDDA as Coligand in Two Steps. To a 10 mL vial was added 0.4 mL of 99mTcO4- solution (50-125 mCi/mL) in saline, followed by 0.5 mL of EDDA solution (10 mg/ mL, pH 7-7.5) in H2O and 10 µL of SnCl2‚2H2O solution (10 mg/mL) in 1 N HCl. After the mixture was allowed to stand at room temperature for 5 min, 0.1 mL of HYNICtide solution (100 µg/mL) in H2O was added to the reaction mixture. It was heated at 75 °C for 30 min and then analyzed by radio-HPLC and ITLC. The complex prepared in two steps was found to be the same as that prepared in one step. Radiolabeling of Cyclic IIb/IIIa HYNICtide Using Polydentate Aminocarboxylates by Exchange Labeling. The complex [99mTc(HYNIC)(tricine)2] was prepared in one step as described above. To resulting [99mTc(HYNIC)(tricine)2] kit was added 1.0 mL of aminocarboxylate solution (∼20 mg/mL, pH ∼8). The mixture was heated at 75 °C for 30 min. The radio-HPLC analysis showed that the tricine complex [99mTc(HYNICtide)(tricine)2] was completely converted into [99mTc(HYNICtide)(L)] (L ) EDDA, N,N′-Et2EDDA, N-BzlEDTA, N-MeEDTA) in 90-95% yield. RESULTS AND DISCUSSION

The high specificity of biologically active molecules, such as antibodies and peptides, has been used in the development of target specific radiopharmaceuticals in both diagnostic and therapeutic nuclear medicine (2, 3). Compared to antibodies with low molecular weight, naturally occurring peptides may better serve as target molecules for radiopharmaceutical development because peptides are necessary elements in more fundamental biological processes than any other class of molecule (3) and in many cases the affinities of peptides for their receptors are significantly greater than that of monovalent antibody fragments (3). Radiopharmaceuticals based on small peptides may not suffer some common problems, such as slow blood clearance and low target-to-background ratio, encountered with radiolabeled antibodies. The faster blood clearance makes it practical to use 99mTc, which has a half-life of 6.02 h and is the preferred radionuclide for diagnostic nuclear medicine. The preferred technetium radiolabeling method is the one-step synthesis, in which the biologically active molecule conjugated with a BFCA is labeled by reduction of pertechnetate with a reducing agent such as stannous chloride. For proteins and their fragments, the conjugation and radiolabeling do not significantly change their chemical and biological properties because the 99mTc complex is attached to only a small portion of these macromolecules. For small peptides, however, modification with a BFCA and radiolabeling may dramatically

Scheme 1

change their chemical and biological properties. In addition, due to the high potency of many peptides, it is often critical to have high specific activity, which is dependent on the BFCA used. Therefore, the choice of BFCA is very important to the success of the radiolabeling. Various BFCAs have been used in labeling proteins and peptides with 99mTc. These include DTPA (21), N2S2 diamidedithiols (22, 23), N3S triamidethiols (24), BATOs (25, 26), and N2S2 diaminedithiols (27). Among them, only N2S2 diaminedithiols show radiolabeling efficiency comparable to the HYNIC group (24, 27). When HYNIC is used as the bonding group for the Tc, the necessity of a coligand offers a wide a range of options to modify the biological characteristics of 99mTc-HYNICtide complexes, such as improving the blood clearance of the imaging agent. This, in return, may result in the improved targetto-background ratios. The cyclic IIb/IIIa HYNICtide (Figure 1) is composed of three parts: a cyclic IIb/IIIa platelet receptor antagonist, a linker, and the HYNIC bonding unit. The cyclic IIb/IIIa receptor antagonist itself has very high potency, the IC50 for the inhibition of platelet aggregation is 20 nM and 6 nM against fibrinogen binding to human platelets (28). Therefore, it must be labeled in high specific activity in order to be used without having a biological effect. The use of the 6-aminocaproic acid linker keeps the Tc center far from the peptide backbone, to minimize loss of binding affinity toward the receptor. Labeling the Cyclic IIb/IIIa HYNICtide with 99mTc Using Tricine as Coligand. The cyclic IIb/IIIa HYNICtide can be easily labeled (Scheme 1) either in one step by reaction of 99mTcO4- with a reducing agent such as stannous chloride in the presence of excess tricine or in two steps by reacting the cyclic IIb/IIIa HYNICtide with 99mTc-tricine. The reaction is carried out at room temperature for 30-60 min (depending on the amount of HYNICtide). Radiolabeling yields are g90% using 5 µg (4.69 × 10-9 mol) of HYNICtide (FW ) 1066 for HYNICtide‚2TFA) and 100 mCi of pertechnetate (∼7 × 10-10 mol of 99mTc and 99Tc for 24 h generator) resulting in a specific activity of g20 000 mCi/µmol. The HYNICtide/Tc ratio is only ∼7:1 under these conditions. Since the 99mTc-labeled IIb/IIIa HYNICtide requires no postlabeling purification, this procedure is amenable for clinical application. In order to fully characterize the 99mTc-HYNICtidetricine complex, we used three different radio-HPLC methods: the TFA method, which uses acetonitrilewater (containing 0.1% TFA) as the mobile phase and a flow rate of 1.0 mL/min, the phosphate gradient method, which uses 10 mM phosphate buffer (pH 6)-acetonitrile as the mobile phase, and the ion-pair method, which uses tetra(n-butyl)ammonium cation as the ion pair agent and 30 mM phosphate buffer (pH 3.7)-acetonitrile as the

66 Bioconjugate Chem., Vol. 7, No. 1, 1996

Figure 2. Radio-HPLC chromatogram for [99mTc(HYNICtide)(tricine)2] by the TFA gradient method (top), the phosphate gradient method (middle), and the ion-pair method (bottom).

mobile phase. The advantage of using the ion-pair method is that the cationic ion pair agent enhances the resolution of species that differ in charge by preferentially retaining anionic species. We found that when the TFA method was used only one peak was detected at 9.6 min (Figure 2). However, the same sample analyzed by the phosphate method shows a cluster of peaks between 11 and 14 min (Figure 2). By the ion pair method, the radioHPLC chromatogram shows at least 10 partially resolved peaks with retention times ranging from 20 to 28 min (Figure 2). In order to determine the number of HYNICtides in the 99mTc-HYNICtide-tricine complex, a mixed-ligand experiment was carried out. In this experiment, the cyclic IIb/IIIa HYNICtide and a model compound, 2-hydrazinopyridine (HYPY), were used in the same reaction mixture. If there is only one HYNICtide in the 99mTcHYNICtide-tricine complex, the radio-HPLC chromatogram is expected to show two sets of peaks, one set due to the 99mTc-HYNICtide-tricine complex and the other due to the 99mTc-HYPY-tricine complex. If there were two HYNICtides bonding to the Tc center, a third set of peaks from the mixed-ligand complex, 99mTc-HYNICtideHYPY-tricine, is expected. The presence of only two sets of peaks from the 99mTc-HYNICtide-tricine complex and the 99mTc-HYPY-tricine complex in the radio-HPLC chromatogram (Figure 3) demonstrates clearly that there is only one HYNIC group bonded to the Tc. If one assumes that the coordination geometry around the Tc is distorted octahedral and HYNICtide occupies only one site by forming either a Tc(V)-hydrazido or Tc(V)-diazenido bond, it requires at least two tricine ligands to complete the coordination sphere. This assumption is supported by the work from Abrams and co-workers (19), who found that the 99mTc-tricine complex, which is a precursor for the synthesis of the 99mTc-HYNICtidetricine complex, is chromatographically and electrophoretically equivalent to the complex [99TcO(tricine)2]prepared from the reaction of TcOCl4- with excess

Liu et al.

Figure 3. Radio-HPLC chromatograms for [99mTc(HYNICtide)(tricine)2], [99mTc(HYPY)(tricine)2], and their mixture (bottom) by the phosphate gradient method.

tetrabutylammonium salt of tricine. Therefore, the composition for the 99mTc-HYNICtide-tricine complex is most likely [99mTc(HYNICtide)(tricine)2] (Scheme 1). The presence of the multiple species is most likely attributed to the resolution of some of the many possible isomers that can result from different bonding modalities of the hydrazine functionality of the HYNICtide and the two tricine coligands (Figure 4). Both Tc-hydrazido and Tc-diazenido bonds have been previously reported and characterized by X-ray crystallography (29-33). These different bonding modalities from the tricine and hydrazine ligands are expected to produce numerous structural and stereo isomers. Therefore, it is not surprising that the radio-HPLC chromatogram of [99mTc(HYNICtide)(tricine)2] by the ion-pair method shows at least 10 partially resolved peaks. The observation of a single peak in the radio-HPLC chromatogram in Figure 2 is a result of the low resolution of the TFA method. The labeling studies also show that the retention times for different 99mTc-labeled HYNICtides vary systematically with lipophilicity while the peak patterns in the radio-HPLC chromatograms remain the same. This strongly suggests that the formation of the multiple species is, in fact, due to coordination isomerism and not due to peptide related rearrangements or degradation or the formation of radioimpurities. This conclusion is supported by the fact that replacement of tricine in [99mTc(HYNICtide)(tricine)2] by other polydentate aminocarboxylates dramatically decreases the number of coordination isomers. The ratios of the peak areas in the radio-HPLC chromatograms change with time, reaction temperature, and pH. Similar phenomena were observed for other 99mTc-labeled HYNICtides using tricine as coligand. This is probably caused by the conversion from a kinetically favored mixture of coordination isomers to a more thermodynamically favored mixture of isomers. In order to prove this assumption, we have successfully isolated

99mTc

Label for Cyclic IIb/IIIa Receptor Antagonist

Bioconjugate Chem., Vol. 7, No. 1, 1996 67

Figure 4. Possible coordination isomers from the tricine coligand and the hydrazine moiety.

several major peaks (Figure 5) and found that most of them interconvert and the rate of interconversion depends on the matrix. The fact that reinjection of the isolated peaks 8 and 9 (E in Figure 5) produces most other peaks in the radio-HPLC chromatogram (A in Figure 5) indicates clearly that the speciations in the radio-HPLC chromatograms are actually due to the isomerism in the complex [99mTc(HYNICtide)(tricine)2] and none are artifacts of the chromatographic conditions. We also found [99mTc(HYNICtide)(tricine)2] to be unstable in dilute solutions (Figure 6). The freshly prepared [99mTc(HYNICtide)(tricine)2] kit was diluted 1:10 and 1:100 with 10 mM phosphate (pH 7.4), 10 mM sodium acetate (pH 5.5), and 2% tricine in 10 mM sodium acetate solution (pH 5.5). The resulting solutions were analyzed by radio-HPLC and ITLC. It was found that the complex [99mTc(HYNICtide)(tricine)2] is stable only in the 2% tricine solution (Figure SI, supporting information). In either 10 mM phosphate solution (pH 7.4) or 10 mM sodium acetate solution (pH 5.5) with 100-fold dilution, 10-15% of the complex [99mTc(HYNICtide)(tricine)2] decomposes to unidentified hydrophilic 99mTc species over a period of 2 h (Figure 6). Apparently, a large excess of tricine in the diluted solutions is needed to stabilize the 99m Tc-HYNICtide core. A platelet-binding assay (34) was used to determine if all these species contain the 99mTc-labeled HYNICtide and bind to the platelets. We isolated five fractions (instead of single peaks) of the complex [99mTc(HYNICtide)(tricine)2] with retention times from 20 to 30 min using the ion pair method. We found that all five fractions are able to bind to the platelets, but with various amounts of platelet-binding. These data clearly demonstrated that all five fractions of the complex [99mTc(HYNICtide)(tricine)2] contain the radiolabeled HYNICtide. This is consistent with our in vivo studies in two canine thrombosis models for the whole kit (13). Like the individual peaks, all five isolated fractions are not quite stable in the HPLC mobile phase, particularly when these frac-

Figure 5. Radio-HPLC chromatograms for [99mTc(HYNICtide)(tricine)2] (A), and the isolated peaks (peak 1 (B), peaks 2 and 3 (C), peak 5 (D), peaks 8 and 9 (E), and peak 10 (F)).

Figure 6. Instability of [99mTc(HYNICtide)(tricine)2] in dilute solution (100-fold).

tions are diluted (5-10-fold) with saline before the platelet-binding assay. In the process of isolation and dilution, parts of the isolated fractions decompose to pertechnetate and parts of them interconvert. The differences in the solution instability of these five fractions may contribute to their different level of plateletbinding. Labeling Cyclic IIb/IIIa HYNICtide with 99mTc Using EDDA as Coligand. Since the complex [99mTc(HYNICtide)(tricine)2] is not stable and exists in many

68 Bioconjugate Chem., Vol. 7, No. 1, 1996

Liu et al.

Scheme 2

isomeric forms, some of which interconvert depending the reaction conditions, it would be difficult to develop for clinical use, even though the biological data from two animal models show that [99mTc(HYNICtide)(tricine)2] can be used to image arterial, and deep venous thrombus (13). Therefore, our research is directed toward preparing 99mTc-HYNICtide complexes with higher stability and a minimum number of coordination isomers using other polydentate aminocarboxylates as coligands. EDDA is of particular interest as coligand because it is potentially tetradentate and is expected to form a more symmetrical and stable complex. The higher symmetry should result in fewer coordination isomers. We have prepared a 99mTc-HYNICtide complex using EDDA as coligand either by a one-step synthesis (Scheme 2) or by a two-step approach, in which a 99mTc-EDDA precursor is formed first and is then allowed to react with the cyclic IIb/IIIa HYNICtide. Using the one-step approach, the reaction was carried out by direct reduction of 99mTcO4in the presence of the cyclic IIb/IIIa HYNICtide and excess EDDA with heating at 75-80 °C for 30 min. The use of shorter reaction times (particularly at HYNICtide concentrations e20 µg/mL) dramatically decreases the labeling yield. If the reaction temperature is too high (g95 °C), the 99mTc-HYNICtide complex is not stable and decomposes by forming 99mTc-colloid and 99mTcO4-. The specific activity (e10 000 mCi/µmol) obtained using EDDA as coligand is slightly lower than that using tricine (g20 000 mCi/µmol). This is probably due to the higher stability of the 99mTc-EDDA precursor. The EDDA concentration also affects the radiolabeling of the cyclic IIb/IIIa HYNICtide. If the EDDA concentration is too high, it will require a higher HYNICtide concentration for a successful labeling. If the EDDA concentration is too low, there is not enough EDDA to stabilize the 99mTcHYNICtide complex and to prevent the formation of 99mTc-colloid. It was found that the lowest acceptable EDDA concentration is ∼5 mg/mL. Using the phosphate method, the radio-HPLC chromatogram shows two main peaks at 11.3 and 12.2 min (Figure 7). By the ion-pair method, three major peaks are observed (Figure 7). This method clearly demonstrates that there are at least three HYNICtide related species in the [99mTc(HYNICtide)(EDDA)] kit. Considering the potential isomerism from the hydrazine moiety and the possibility of a ligand X either trans or cis to the Tc-hydrazido bond, it is not surprising that there are still three or more coordination isomers for the complex [99mTc(HYNICtide)(EDDA)]. However, compared to the >10 isomeric species from [99mTc(HYNICtide)(tricine)2],

Figure 7. Radio-HPLC chromatogram for [99mTc(HYNICtide)(EDDA)] by the phosphate gradient method (top) and the ionpair method (bottom).

Figure 8. Radio-HPLC chromatograms for [99mTc(HYNICtide)(EDDA)] in the original kit (top), the HPLC purified [99mTc(HYNICtide)(EDDA)] in saline at t ) 20 min (middle), and the HPLC purified [99mTc(HYNICtide)(EDDA)] in saline at t ) 180 min (bottom).

the dramatic decrease in coordination isomers for [99mTc(HYNICtide)(EDDA)] is clearly a big improvement. Another advantage of using EDDA as coligand is the higher stability of the corresponding 99mTc complex. This is demonstrated by the fact that the complex [99mTc(HYNICtide)(EDDA)] is stable for at least 12 h in the kit matrix and the peak ratios remain relatively constant. The major isomer can be easily isolated and is stable for 2 h in saline solution (Figure 8) and 3 h in 1% EDDA solution. No decomposition was observed. However, there is evidence of conversion to the other species as seen by the growing shoulder on the main peak (Figure

99mTc

Bioconjugate Chem., Vol. 7, No. 1, 1996 69

Label for Cyclic IIb/IIIa Receptor Antagonist

Figure 10. N-Substituted EDDA analogs.

Figure 9. Three possible bonding modalities of EDDA.

Table 1. Radiolabeling Results Using EDDA and Its N-Substituted Analogs complexa

8), suggesting that there is an equilibrium between different isomeric forms. The interconversion also shows that these peaks in the radio-HPLC chromatogram are actually from the complex [99mTc(HYNICtide)(EDDA)], not artifacts of the chromatographic conditions. On the basis of these data, it is quite clear that the use of EDDA to replace tricine as coligand produces a 99mTc complex with higher solution stability and less isomerism. Using EDDA as Coligand by the Exchange Labeling Approach. The major advantage of using tricine as coligand is the high labeling efficiency and favorable labeling kinetics. The labeling efficiency using EDDA as coligand is not as high as that obtained using tricine even though it forms more stable 99mTc-HYNICtide complexes with fewer coordination isomers. In order to achieve higher specific activity using EDDA as coligand while retaining the stability of the 99mTc-HYNICtide complex, the exchange labeling approach was used to label the cyclic IIb/IIIa HYNICtide. In this approach, the tricine complex [99mTc(HYNICtide)(tricine)2] is formed first and is then allowed to undergo ligand exchange (Scheme 2) with a large excess of EDDA (EDDA:tricine g 2) and prolonged (g30 min) heating at 75 °C. The radio-HPLC shows that the complex [99mTc(HYNICtide)(EDDA)] prepared by this approach is the same as that prepared by the one-step synthesis. The exact nature of the bonding between the TcHYNICtide core and EDDA is not quite clear. Very little is known about the coordination chemistry at the carrieradded (99Tc) level. It has been reported that polydentate aminocarboxylates such as EDTA and TCTA (1,4,7triazacyclononane-N,N′,N′′-triacetate) form stable Tc(III) and Tc(IV) complexes (35, 36). However, EDTA is used as a transfer ligand in Neurolite kits (37). In the labeling process, the complex [99mTcO(EDTA)]n- is formed first and then undergoes ligand exchange with a stronger chelator, ECD (N,N′-ethylenediylbis-L-cysteinate diethyl ester), to form a neutral complex [99mTcO(ECD)]. It is difficult to imagine that the Tc is reduced to Tc(III) or Tc(IV) and then reoxidized back to Tc(V))O for ligand exchange in the presence of excess reducing agent. Since the complex [99mTc(HYNICtide)(EDDA)] can be prepared by the two-step approach, it is reasonable to assume that the oxidation state for technetium is Tc(V), with only one HYNIC group bonding to the Tc center. If EDDA acts as a tetradentate ligand, another monodentate ligand X (such as Cl-) is required to complete the Tc coordination sphere. Figure 9 shows several possible bonding modalities for the EDDA ligand. Labeling the Cyclic IIb/IIIa HYNICtide with 99mTc Using HEDTA as Coligand. HEDTA (N-(2-hydroxyethyl)ethylenediaminetriacetic acid) is potentially pentadentate using its N2O3 donor atoms in bonding to the Tc to form a six-coordinated 99mTc-HYNICtide complex. Either hydroxyl-O or carboxylate-O may coordinate to the Tc center trans to the Tc-hydrazido or Tc-diazenido

99mTc(HYNICtide)(EDDA) 99mTc(HYNICtide)(HEDTA) 99mTc(HYNICtide)(N,N′-Et

2EDDA) 99mTc(HYNICtide)(N-BzlEDTA) 99mTc(HYNICtide)(N-MeEDTA)

yield (%)

tRb (min)

90-95 90-93 90 93 85

11.3, 12.2 11.5, 12.8, 13.5 12.1, 12.5, 13.0 15.6, 16.5 13.1, 13.8

a 99mTc-HYNICtide complexes were prepared by the one-step approach. b Peaks (g2%) observed in radio-HPLC chromatograms (by the phosphate method).

bond. Radiolabeling studies show that it can be used as a coligand to label the cyclic IIb/IIIa HYNICtide either by a one-step reaction or by the ligand exchange labeling method (Scheme 2). The labeling efficiency is comparable to that of EDDA by the one-step approach. The radioHPLC data by both the phosphate (Figure SII, supporting information) and ion-pair methods show more isomerism for the complex [99mTc(HYNICtide)(HEDTA)] probably due to the asymmetric character of HEDTA. N-Substituted EDDA and HEDTA Analogs. The results from studies on [99mTc(HYNICtide)(EDDA)] and [99mTc(HYNICtide)(HEDTA)] clearly demonstrate that these polydentate aminocarboxylates are promising coligands because of the increased stability and less isomerism. 1H NMR studies on vanadium(V) complexes showed that in aqueous solution the complex [VO2(EDDA)]exists as a mixture of isomers (R-cis and β-cis) at room temperature while complexes [VO2(L)]- (L ) N,N′-Me2EDDA and EDTA) exist only in the R-cis configuration because of the steric effect of the N-substituents (38, 39). In order to explore the effect of N-substituents on isomerism in 99mTc-HYNICtide complexes, we have prepared several N-substituted EDDA analogs (Figure 10) and used them to label the cyclic IIb/IIIa HYNICtide. Synthesis of N-substituted EDDA analogs is straightforward. The alkylation is carried out by reacting N-substituted ethylenediamine with tert-butyl bromoacetate in the presence of anhydrous potassium carbonate in acetonitrile. The acid hydrolysis of the tert-butyl ester can be easily achieved by dissolving the corresponding tert-butyl ester in concentrated hydrochloric acid. These ligands were isolated as their hydrochloride salts and have been characterized by IR, FAB-MS, 1H NMR, and elemental analysis. Radiolabeling studies (Table 1) clearly show that N-substituted EDDA analogs can be used as coligands to synthesize 99mTc-HYNICtide complexes. Like EDDA, they typically give lower labeling efficiency. The introduction of different N-substituents produces 99mTcHYNICtide complexes with different hydrophilicity, but it offers no improvement with respect to the reduction of the number of coordination isomers. CONCLUSION

In conclusion, the cyclic IIb/IIIa HYNICtide, cyclo(DVal-NMeArg-Gly-Asp-Mamb(hydrazinonicotinyl-5-(6-ami-

70 Bioconjugate Chem., Vol. 7, No. 1, 1996

nocaproic acid))), was labeled with 99mTc using polydentate aminocarboxylates as coligands. A high specific activity (g20 000 mCi/µmol) can be achieved using tricine as coligand. The complex [99mTc(HYNICtide)(tricine)2] was unstable in solution and was found to exist in multiple forms that we attribute to coordination isomerism. It should be emphasized that the isomerism becomes observable only in the radiolabeling of small peptides. For proteins such as polyclonal IgG, the chemical and biological properties are dominated by the polypeptide structure and should not be affected by coordination isomerism at the Tc center. The use of EDDA to replace tricine as a coligand produces the complex [99mTc(HYNICtide)(EDDA)] with higher solution stability and less isomerism. The N-substituted analogs of EDDA can also be used as coligands but show no superiority over EDDA with respect to the solution stability and reduction of isomers for the 99mTc-HYNICtide complexes. Acknowledgement is made to J. Pietryka for the 1H NMR spectra and to Dr. T. D. Harris, Dr. M. Rajopadhye, D. Glowacka, P. R. Damphousse, J. P. Bourque, and K. Yu for the synthesis of the cyclic IIb/IIIa HYNICtide, cyclo(D-Val-NMeArg-Gly-Asp-Mamb(hydrazinonicotinyl5-(6-aminocaproic acid))). Supporting Information Available: Diagram showing the instability of [99mTc(HYNICtide)(tricine)2] in three different media upon dilution. Radio-HPLC chromatogram for the complex [99mTc(HYNICtide)(HEDTA)] prepared by the one-step synthesis (2 pages). Ordering information is available on any current masthead page. LITERATURE CITED (1) Goldenberg, D. M. (1989) Future role of radiolabeled monoclonal antibodies in oncological diagnosis and therapy. Seminars Nucl. Med. 19, 332-339. (2) Eckelman, W. C. (1994) The application of receptor theory to receptor-binding and enzyme-binding oncologic radiopharmaceuticals. Nucl. Med. Biol. 21, 759-769. (3) Fischman, A. J., Babich, J. W., and Strauss, H. W. (1993) A ticket to ride: peptide radiophamaceuticals. J. Nucl. Med. 34, 2253-2263. (4) Fischman, A. J., Babich, J. W., and Rubin, H. R. (1993) Infection imaging with technetium-99m-labeled chemotactic peptides analogs. Semin. Nucl. Med. 24, 154-168. (5) Eckelman, W. C., Paik, C. H., and Steigman, J. (1989) Three approaches to radiolabeling antibodies with 99mTc. Nucl. Med. Biol. 16, 171-176. (6) Fritzberg, A. R., Berninger, R. W., Hadley, S. W., and Wester, D. W. (1988) Approaches to radiolabeling of antibodies for diagnosis and therapy of cancer. Pharm. Res. 5, 325334. (7) Otsuka, F. L., and Welch, M. J. (1987) Methods to label monoclonal antibodies for use in tumor imaging. Nucl. Med. Biol. 14, 243-249. (8) Hnatowich, D. J. (1990) Antibody radiolabeling, problems and promises. Nucl. Med. Biol. 17, 49-55. (9) Hnatowich, D. J. (1990) Recent developments in radiolabeling of antibodies with iodine, indium, and technetium. Semin. Nucl. Med. 20, 80-91. (10) Srivastava, S. C., and Mease, R. C. (1991) Progress in research on ligands, nuclides and techniques for labeling monoclonal antibodies. Nucl. Med. Biol. 18, 589-603. (11) Barrett, J. A., Heminway, S. J., Damphousse, D. J., Thomas, J. R.; Looby, R. J., Edwards, D. S., Harris, T. D., Rajopadhye, M., Liu, S., and Carroll, T. R. (1994) Platelet GP IIb/IIIa antagonists in the canine arteriovenous shunt: potential thrombus imaging agents. J. Nucl. Med. 35, 52P (Abstract 202). (12) Harris, T. D., Barrett, J. A., Bourque, J. P., Carroll, T. R., Damphousse, P. R., Edwards, D. S., Glowacka, D., Liu, S., Looby, R. J., Poirier, M. J., Rajopadhye, M., and Yu, K. (1994)

Liu et al. Design and synthesis of radiolabeled GPIIb/IIIa receptor antagonists as potential thrombus imaging agents. J. Nucl. Med. 35, 245P (Abstract 1005). (13) Barrett, J. A., Bresnick, M., Crocker, A., Damphousse, D. J., Hampson, J. R., Heminway, S. J, Mazaika, T. J., Kagan, M., Lazewatsky, J., Edwards, D. S., Liu, S., Harris, T. D., Rajopadhye, M., and Carroll, T. R. (1995) RP-431: a potential thrombus imaging agent. J. Nucl. Med. 36, 16P (Abstract 55). (14) Schwartz, D. A., Abrams, M. J., Hauser, M. M., Gaul, F. E., Larsen, S. K., Rauh, D., and Zubieta, J. A. (1991) Preparation of hydrazino-modified proteins and their use for the synthesis of 99mTc-protein conjugate. Bioconjugate Chem. 2, 334-336. (15) Abrams, M. J., Juweid, M., tenKate, C. I., Schwartz, D. A., Hauser, M. M., Gaul, F. E., Fuccello, A. J., Rubin, R. H., Strauss, H. W., and Fischman, A. J. (1990) Technetium-99mhuman polyclonal IgG radiolabeled via the hydrazino nicotinamide derivative for imaging focal sites of infection in rats. J. Nucl. Med. 31, 2022-2028. (16) Babich, J. W., Solomon, H., Pike, M. C., Kroon, D., Graham, W., Abrams, M. J., Tompkins, R. G., Rubin, R. H., and Fischman, A. J. (1993) Technetium-99m-labeled hydrazino nicotinamide derivatized chemotactic peptide analogs for imaging focal sites of bacterial infection. J. Nucl. Med. 34, 1967-1974. (17) Babich, J. W., and Fichman, A. J. (1995) Effect of “coligand” on the biodistribution of 99mTc-labeled hydrazino nicotinic acid derivatized chemotactic peptides. Nucl. Med. Biol. 22, 25-30. (18) Larson, S. K., Abrams, M. J., Higgins, J. D., III, Solomon, H. F., Babich, J. W., and Fischman, A. J. (1994) Technetium complexes of tricine: useful precursor for the 99mTc labeling of hydrazino nicotinamide modified human polyclonal IgG. J. Nucl. Med. 35, 105P. (19) Larson, S. K., Caldwell, G., Higgins, J. D., III, Abrams, M. J., and Solomon, H. F. (1994) Technetium complex of tricine: useful precursor for the 99mTc labeling of hydrazino nicotinamide modified proteins. J. Labeled Compds. Radiopharm. 35, 1-2. (20) Rajopadhye, M., Edwards, D. S., and Liu, S. Unpublished data. (21) Lanteingne, D., and Hnatowich, D. J. (1984) The labeling of DTPA coupled proteins. Int. J. Radiat. Isot. 35, 617-621. (22) Kasina, S., Rao, T. N., Srinivasan, A., Sanderson, J. A., Fitner, J. N., Reno, J. M., Beaunier, P. L., and Fritzberg, A. R. (1991) Development and biologic evaluation of a kit for preformed chelate technetium-99m radiolabeling of an antibody Fab fragment using a diamide dimercaptide chelating agent. J. Nucl. Med. 32, 1445-1451. (23) Fritzberg, A. R., Abrams, P. G., Beaumier, P. L., Kasina, S., Morgan, A. C., Rao, T. N., Reno, J. M., Sanderson, J. A., Srinivasan, A., and Wilbur, D. S. (1988) Specific and stable labeling of antibodies with technetium-99m with a diamide dithiolate chelating agent. Proc. Natl. Acad. Sci. U.S.A. 85, 4025-4029. (24) Liu, S., and Edwards, D. S. (1994) New N2S2 diamidedithiol and N3S triamidethiols as bifunctional chelating agents for labeling small peptides with 99mTc. In Technetium and Rhenium in Chemistry and Nuclear Medicine (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), pp 383-393, SGEditoriali, Padova. (25) Linder, K. E., Wen, M. D., Nowotnik, D. P., Malley, M. F., Goufoutas, J. Z., Nunn, A. D., and Eckelman, W. C. (1991) Technetium labeling of monoclonal antibodies with functionalized BATO’s: 1. TcCl(DMG)3 PITC. Bioconjugate Chem. 2, 160-170. (26) Linder, K. E., Wen, M. D., Nowotnik, D. P., Ramalingam, K., Sharkey, R. M., Yost, F., Narra, R. K., Nunn, A. D., and Eckelman, W. C. (1991) Technetium labeling of monoclonal antibodies with functionalized BATO’s: 2. TcCl(DMG)3CPITC labeling of B72.3 and NP-4 whole antibodies and NP-4 F(ab′)2. Bioconjugate Chem. 2, 407-414. (27) Baidoo, K. E., Scheffel, U., Finley, P., and Lever, S. Z. (1994) No carrier added (NCA) Tc-99m-diaminedithiolate derivatives of chemotactic peptides for imaging inflammation. J. Nucl. Med. 35, 19P (Abstract 67).

99mTc

Bioconjugate Chem., Vol. 7, No. 1, 1996 71

Label for Cyclic IIb/IIIa Receptor Antagonist

(28) Jackson, S., DeGrado, W., Dwivedi, A., Parthasarathy, A., Hifley, A., Krywko, J., Rockwell, A., Markwalder, J., Wells, G., Wexler, R., Mousa, S., and Harlow R. (1994) Templateconstrained cyclic peptides: design of high-affinity ligands for GP IIb/IIIa. J. Am. Chem. Soc. 116, 3220-3230. (29) Archer, C. M., Dilworth, J. R., Jobanputra, P., Thompson, R. M., McPartlin, M., Povey, D. C., Smith, G. W., and Kelly, J. D. (1990) Development of new technetium cores containing technetium-nitrogen multiple bonds. Synthesis and characterization of some diazenido-, hydrazido- and imido- complexes of technetium. Polyhedron 9, 1497-1502. (30) Dilworth, J. R., Jobanputra, P., Thompson, R. M., Archer, C. M., Povey, D. C., Kelly, J. D., and Hiller, W. (1992) Crystal structure of a diazenidodithiocarbamate complex of technetium, [Tc(NNC6H4Cl)((CH3)2NCS2)2(PPh3)]. Z. Naturforsch. 46, 449-452. (31) Archer, C. M., Dilworth, J. R., Jobanputra, P., Thompson, R. M., McPartin, M., and Hiller, W. (1993) Technetium diazenido complexes. Part 1. Synthesis and structures of [TcCl(NNC6H4Cl-4)2(PPh3)2] and [TcCl(NNPh)(Ph2PCH2CH2PPh2)2][PF6].H2O. J. Chem. Soc., Dalton Trans., 897-904. (32) Dilworth, J. R., Jobanputra, P., Thompson, R. M., Povey, D. C., Archer, C. M., and Kelly, J. D. (1994) Technetium diazenido complexes. Part 2. Substitution chemistry of structures of [TcCl(NNC6H4Cl-4)2(PPh3)2] and the synthesis of technetium diazenido-complexes directly from [NH4][TcO4]. J. Chem. Soc., Dalton Trans., 1251-1256. (33) Abrams, M. J., Larsen, S. K., Shaikh, S. N., and Zubieta, J. (1991) Investigation of technetium-organohydrazine coor-

dination chemistry. The crystal and molecular structures of [TcCl2(C8H5N4)(PPh3)2]‚0.75C7H8 and [TcNCl2(PPh3)2]‚0.25CH2Cl2. Inorg. Chim. Acta 185, 7-15. (34) Plow, E. F., Marguerie, G., and Ginsberg, M. (1987) Fibrinogen, fibrinogen receptors and the peptides that inhibit these interactions. Biochem. Pharmacol. 36, 4035-4041. (35) Burgi, H. B., Anderegg, G., and Blauenstein, P. (1981) Preparation, characterization, and crystal, molecular, and electronic structure of (H2EDTA)99TcIV(µ-O)299Tc(H2EDTA)‚ 5H2O. A 2.33-Å Tc-Tc distance which may represent a σ2π2δ*2 bond. Inorg. Chem. 20, 3829-3834. (36) Linder, K. E., Dewan, J. C., and Davison, A. (1989) Technetium bis(µ-oxo) dimers of 1,4,7-triazacyclononaneN,N′,N′′-triacetate (TCTA). Synthesis and characterization of [(TCTA)Tc(µ-O)2Tc(TCTA)]n- (n ) 2, 3) and the crystal structure of Ba2[(TCTA)Tc(µ-O)2Tc(TCTA)](ClO4)‚9H2O. Inorg. Chem. 28, 3820-3825. (37) Dewanjee, M. K. (1990) The chemistry of 99mTc-labeled radiopharmaceuticals. Semin. Nucl. Med. 20, 5-7. (38) Amos, L. W., and Sawyer, D. T. (1972) Proton nuclear magnetic resonance studies of several polyaminocarboxylic acid complexes of vanadium(V). Inorg. Chem. 11, 2692-2697. (39) Crans, D. C., and Shin, P. K. (1994) Characterization of vanadium(V) complexes in aqueous solutions: ethanolamineand glycine-derived complexes. J. Am. Chem. Soc. 116, 1305-1315.

BC950069+