Bioconjugate Chem. 2001, 12, 559−568
559
90Y
and 177Lu Labeling of a DOTA-Conjugated Vitronectin Receptor Antagonist Useful for Tumor Therapy Shuang Liu,* Eric Cheung, Marisa C. Ziegler, Milind Rajopadhye, and D. Scott Edwards Medical Imaging Division, DuPont Pharmaceuticals Company, 331 Treble Cove Road, North Billerica, Massachusetts 01862 . Received November 30, 2000; Revised Manuscript Received March 30, 2001
The 90Y and 177Lu complexes (RP697 and RP688, respectively) of a DOTA-conjugated vitronectin receptor antagonist (SU015: 2-(1,4,7,10-tetraaza-4,7,10-tris(carboxymethyl)-1-cyclododecyl)acetyl-Glu(cyclo{Lys-Arg-Gly-Asp-D-Phe})-cyclo{Lys-Arg-Gly-Asp-D-Phe}) were prepared by reacting SU015 with the radiometal chloride in ammonium acetate buffer (pH > 7.2) in the presence of an antioxidant (sodium gentisate, GA). Through a series of radiolabeling experiments, it was found that there are many factors influencing the rate of 90Y chelation and the radiolabeling efficiency of SU015. These include the purity of SU015, the pH, reaction temperature, and heating time, as well as the presence of trace metal contaminants, such as Ca2+, Fe3+, and Zn2+. The chelation of 90Y by SU015 is slow, so that heating at elevated temperatures (50-100 °C) is needed to complete the 90Y-labeling. The rate of 90Y chelation is also dependent on the pH of the reaction mixture. Under optimized radiolabeling conditions (pH 7.2-7.8 and heating at 50-100 °C for 5-10 min), the minimum amount of SU015 required to achieve 95% RCP for RP697 is ∼25 µg for 20 mCi of 90YCl3 corresponding to a SU015:90Y ratio of ∼30:1.
INTRODUCTION
We have been interested in the development of diagnostic radiopharmaceuticals based on small molecule receptor ligands (1-15). We have been using a ternary ligand system for the 99mTc labeling of a variety of hydrazinonicotinamide-conjugated biomolecules (HYNICBM), including a GPIIb/IIIa receptor antagonist for thrombus imaging (1-8), and chemotactic peptides (9) and LTB4 receptor antagonists for imaging infection and inflammation (12-14). In a separate communication (15), we report the 99mTc labeling of an HYNIC-conjugated vitronectin receptor antagonist (SQ168: 2-sulfonatobenzenesulfonic acid hydrazone of N-(6-hydrazinonicotinamido)-Glu(cyclo{Lys-Arg-Gly-Asp-D-Phe})-cyclo{Lys-ArgGly-Asp-D-Phe}) and biological properties of the complex [99mTc(SQ168)(tricine)(TPPTS)] (RP593: TPPTS ) trisodium triphenylphosphine-3,3′,3′′-trisulfonate). In the cNeu Oncomouse tumor model (15), RP593 shows high tumor uptake and long tumor residence time (3.4% ID/g at 2 h and 1.5% ID/g at 24 h postinjection) and is excreted predominantly via the renal system. This suggests that the corresponding 90Y and 177Lu complexes have the potential as target-specific radiopharmaceuticals for tumor therapy. As a continuation of our efforts in developing new peptide-based diagnostic and therapeutic radiopharmaceuticals, a DOTA-conjugated vitronectin receptor antagonist (Figure 1: SU015 ) 2-(1,4,7,10-tetraaza-4,7,10tris(carboxymethyl)-1-cyclododecyl)-acetyl-Glu(cyclo{LysArg-Gly-Asp-D-Phe})-cyclo{Lys-Arg-Gly-Asp-D-Phe}) has been synthesized. In this report, we present the synthesis and characterization of its corresponding 90Y and 177Lu complexes (Figure 1: RP697 and RP688). Through a series of radiolabeling experiments, we developed a for* To whom correspondence should be addressed: Tel: 978671-8696 (S.L.); FAX: 978-436-7500; e-mail: shuang.liu@ dupontpharma.com.
mulation for routine preparations of RP697 and RP688. This formulation may be used for the 90Y- and 177Lulabeling of other DOTA-conjugated biomolecules. EXPERIMENTAL SECTION
Materials. Acetic acid (ultrapure), ammonium hydroxide (ultrapure), diethylenetriaminepentaacetic acid (DTPA), and sodium gentisate were purchased from either Aldrich or Sigma Chemical Co. and were used as received. 90YCl3 (in 0.05 N HCl) was purchased from NEN Life Sciences, N. Billerica, MA. High specific activity 177 LuCl3 was obtained from University of Missouri Research Reactor, Columbia, MO. The cyclic pentapeptide, cyclo(Arg-Gly-Asp-D-Phe-Lys), as its trifluoroacetic acid (TFA) salt was prepared according to the literature method (16). Synthesis of 2-(1,4,7,10-Tetraaza-4,7,10-tris(tertbutoxycarbonylmethyl)-1-cyclododecyl)acetyl-Glu(cyclo{Lys-Arg-Gly-Asp-D-Phe})-cyclo{Lys-Arg-GlyAsp-D-Phe}. To a solution of tris(tert-butyl)-1,4,7,10tetraazacyclododecane-1,4,7,10-tetraacetic acid (45 mg, 0.078 mmol) in DMF (2 mL) were added HBTU (18 mg, 0.047 mmol) and Hunig’s base (20 µL). The mixture was stirred for 15 min. To this were added a solution of Glu(cyclo{Lys-Arg-Gly-Asp-D-Phe})-cyclo{Lys-Arg-Gly-AspD-Phe} (63 mg, 0.038 mmol) and Hunig’s base (20 µL) in DMF (1 mL). The reaction mixture was allowed to stir under nitrogen at room-temperature overnight. The solvent was removed in vacuo, and the residue was triturated in ethyl acetate (10 mL), filtered, and washed with ethyl acetate to give the desired product (75.2 mg). Purification of the DOTA conjugate is achieved by preparative HPLC (Vydac reverse phase C18 column; solvent A: 0.1% TFA in water; solvent B: 0.1% TFA 90% aq. ACN). Removal of the mobile phase gave the product as a lyophilized solid. ES-MS: m/z ) 1873.9 for [M + H]+ (C87H138N23O23) and 937.2 for [M + 2H]+2. The analytical HPLC method for the DOTA-conjugate used
10.1021/bc000146n CCC: $20.00 © 2001 American Chemical Society Published on Web 06/12/2001
560 Bioconjugate Chem., Vol. 12, No. 4, 2001
Liu et al.
Figure 1. Structures of SU015 and its yttrium and lutetium complexes.
a HP-1050 HPLC system with a Diode array detector (λ ) 220 nm) and a Vydac C18 column (4.6 mm × 250 mm, 80 Å pore size). The column temperature was 50 °C. The flow rate was 1 mL/min with the mobile phase starting 98% solvent A (0.1% TFA in water) and 2% solvent B (0.1% TFA in 90% acetonitrile) to 63.2% solvent A and 36.8% solvent B at 16 min. The retention time was 20.0 min, and the purity was 99%. Synthesis of 2-(1,4,7,10-Tetraaza-4,7,10-tris(carboxymethyl)-1-cyclododecyl)-acetyl-Glu(cyclo{LysArg-Gly-Asp- D -Phe})-cyclo{Lys-Arg-Gly-Asp- D Phe} (SU015). A solution of 2-(1,4,7,10-tetraaza-4,7,10tris(tert-butoxycarbonylmethyl)-1-cyclododecyl)acetyl-Glu(cyclo{Lys-Arg-Gly-Asp-D-Phe})-cyclo{Lys-Arg-Gly-AspD-Phe} (4.5 g, 2.4 mmol) in TFA (110 mL) was stirred at room temperature under nitrogen for 2 h. The solution was concentrated in vacuo, and the residue was purified by preparative HPLC (Vydac C18 column; solvent A: 0.1% TFA in water; solvent B: 0.1% TFA 90% aq. ACN). Removal of the mobile phase gave the product as a lyophilized solid (1.5 g, tetra TFA salt). ES-MS: m/z ) 1703.8 for [M + H]+ (C75H114N23O23) and 853.0 for [M + 2H]2+. The analytical HPLC method for SU015 used a HP-1050 HPLC system with a Diode array detector (λ ) 220 nm) and a Vydac C18 column (4.6 mm × 250 mm, 80 Å pore size). The column temperature was 50 °C. The flow rate was 1 mL/min with the mobile phase starting 98% solvent A (0.1% TFA in water) and 2% solvent B (0.1% TFA in 90% acetonitrile) to 63.2% solvent A and 36.8% solvent B at 16 min. The retention time was 13.4 min, and the purity was 97%.
A General Procedure for Synthesis of RP688 and RP697. To a shielded, clean 5 mL vial containing 100 µg of SU015 and 1-10 mg of sodium gentisate (GA) dissolved in 0.5 mL of 0.5 M ammonium acetate buffer (pH ) 7.0, 7.5, or 8.0) was added 10-40 µL of 90YCl3 or 177 LuCl3 solution (∼ 20 mCi) in 0.05 N HCl. The reaction mixture was allowed to stand at room temperature or was heated at 50-100 °C for 5-30 min. For RP697, a sample of the resulting solution was diluted 20-fold with 4 mM DTPA solution (pH ) 5) and then analyzed by HPLC (injection volume ) 1-2 µL) and by ITLC. For RP688, the resulting solution was used for HPLC analysis without dilution (injection volume ) 3-5 µL). The HPLC method for analysis of RP688 and RP697 used a HP-1100 HPLC system with a UV/visible detector (λ ) 220 nm), an IN-US radio-detector, and a Zorbax C18 column (4.6 mm × 250 mm, 80 Å pore size). The flow rate was 1 mL/min with an isocratic mobile phase from 0 to 18 min using 87% solvent A (0.025 M ammonium acetate buffer, pH 6.8) and 13% solvent B (acetonitrile), followed by an isocratic wash using 40% of solvent A and 60% solvent B from 19 to 25 min. The ITLC method used Gelman Sciences silica gel ITLC paper strips and a 1:1 mixture of acetone and saline as eluant. By this method, RP697 and RP688 migrate to the solvent front while [90Y]/ [177Lu]colloid and [90Y]/[177Lu]acetate remain at the origin. Synthesis of [175Lu]RP688 for LC-MS. To a clean 5 mL vial containing 5 mg of SU015, 10 mg of GA, and 10 mg of Lu(NO3)3‚6H2O was added 1.0 mL of 0.5 M ammonium acetate buffer (pH ) 7.5). The reaction mixture was heated at 100 °C for 20 min. After cooling
90Y
and
177Lu
Labeling of Vitronectin Receptor Antagonist
Figure 2. Typical radio (top) and UV/vis (bottom) HPLC chromatograms for the reaction mixture containing RP697. The UV/vis trace was recorded by injecting the reaction mixture without dilution. There is a retention time difference (∼1.0 min) for the two detectors since they were not synchronized.
to room temperature, a sample of the resulting solution was filtered, and the filtrate was analyzed by LC-MS. No further purification was performed for the LC-MS analysis. The LC-MS spectra were collected using a HP1100 LC/ MSD system with API-electrospray interface. The LCMS method used a Zorbax C18 column (4.6 mm × 150 mm, 3.5 µm particle size) at a flow rate of 1 mL/min and a gradient mobile phase starting from 92% solvent A (10 mM ammonium acetate buffer, pH 7.0) and 8% solvent B (methanol) to 100% B at 23 min. The MSD parameters are as follows: detection mode: positive mass range: 600-2000; gain: 1.0; fragmentor: 30 V; gas temperature: 350 °C; drying gas flow: 13 L/min; nebullizer pressure: 60 psig (max.); V capillary: 4000 V; UV/vis detector: λ ) 220 nm. RESULTS
Synthesis of RP688 and RP697. RP697 was prepared by reacting SU015 with 90YCl3 in an ammonium acetate buffer (pH ) 6.5-8) at 50-100 °C. GA was used as a stabilizer to prevent radiolytic degradation of RP697. The total volume was always 0.5 mL, and the 90YCl3 concentration was 40 mCi/mL. We typically used 100 µg of SU015 for 20 mCi of 90YCl3 corresponding to a SU015: 90 Y ratio of ∼130:1. The use of excess SU015 is to compensate for the presence of other trace metal contaminants. Under optimized conditions, RP697 can be prepared in high yield with the RCP > 95%. RP688 was prepared in a similar fashion by reacting SU015 with 177LuCl (100 µg of SU015, 10 mg of GA for 20 mCi of 3 177LuCl , pH > 7.2, heating at 100 °C for 5 min). 3 HPLC Characterization. Figures 2 and 3 show typical HPLC (radio and UV) chromatograms for RP688 and RP697, both of which show only one radiometric peak. The retention time for RP688 and RP697 is almost identical under the same chromatographic conditions. There are several small peaks due to radioimpurities in the HPLC chromatogram of both RP688 and RP697. Since these radioimpurities are less than 1.0%, no further characterization was performed. Many attempts were made to use LC-MS for the characterization of the
Bioconjugate Chem., Vol. 12, No. 4, 2001 561
Figure 3. Typical radio (top) and UV/vis (bottom) HPLC chromatograms for the reaction mixture containing RP688. There is a retention time difference (∼1.0 min) for the two detectors since they were not synchronized.
reaction mixture containing RP697, but the results were often difficult to interpret, due to dilute concentration of RP697. LC-MS Characterization of RP688. We have used LC-MS to determine the composition of RP688. To identify the peak due to RP688 in the UV profile, we also prepared the corresponding “cold” complex Lu(SU015) by reacting SU015 with excess of lutetium(III) nitrate in a 0.5 M ammonium acetate buffer (pH ) 7.5). Figure 4 shows the UV/vis and MSD profiles for RP688 and 175Lu(SU015). Figure 5 shows the LC-MS spectra of RP688 (bottom) and 175Lu(SU015) (top). The LC-MS patterns of both complexes are identical except the difference in isotope. The LC-MS spectral data support the proposed structure (Figure 1). There are several other intense peaks due to complexes of trace metal contaminants in the reaction mixture containing RP688. The molecular weights (Table 1) corresponding to these peaks match those from Ca(SU015) (tR ) 10.85-10.91 min), Co(SU015) (tR ) 10.05-10.08 min), Zn(SU015) (tR ) 10.11-10.21 min), and Fe(SU015) (tR ) 10.22-10.32 min). Buffer Concentration. In this experiment, we used 100 µg of SU015 and 10 mg of GA, 20 mCi of 90YCl3, and ammonium acetate buffer (0.1, 0.25, or 0.5 M) to prepare RP697. The pH in the mixture was 7.2 after addition of 90 YCl3. The reaction mixture was heated at 100 °C for 10 min. Figure 6 shows that the buffer concentration does not have a significant impact on RCP as long as the pH is well controlled. The 0.5 M ammonium acetate buffer was chosen for the remaining radiolabeling experiments mainly due to its higher buffering capacity. 90Y-Chelation Kinetics at Room Temperature. In this experiment, we explored the possibility of preparing RP697 by reacting 100 µg of SU015 with 20 mCi of 90YCl3 in the presence of GA (10 mg) at room temperature. The pH in the reaction mixture was 6.6, 7.2, or 7.8 after addition of 90YCl3. Samples of the reaction mixture were analyzed by radio-HPLC at 5, 30, 60, and 120 min. The RCP data are summarized in Figure 7. It is obvious that the rate of 90Y chelation depends on the pH of the reaction mixture. At pH 6.6, the RCP for RP697 was only ∼60% after 2 h at room temperature. It is possible to prepare RP697 with RCP > 90% at room temperature at pH >7.0,
562 Bioconjugate Chem., Vol. 12, No. 4, 2001
Liu et al.
Figure 4. UV/vis (λ ) 220 nm) and MSD profiles of the reaction mixture containing [175Lu]RP688 (top) and RP688 (bottom).
Figure 6. Effect of the buffer concentration on RCP of RP697.
Figure 7. Room temperature 6.6, 7.2, and 7.8.
Figure 5. LC-MS spectra for [175Lu]RP688 (top) and RP688 (bottom). The y-axis represents intensity of the signal. The x-axis represents m/z ratio.
but the RCP of RP697 could not be further improved. The elevated temperatures are needed to complete the 90Y chelation. Preparation of RP697 at 50 °C. In this experiment, we prepared RP697 using 100 µg of SU015 and 10 mg of GA for 20 mCi of 90YCl3. The pH in the mixture was 7.2 or 7.8 after addition of 90YCl3. After heating at 50 °C for 5, 10, and 30 min, a sample of the reaction mixture was analyzed by radio-HPLC and ITLC. It was found (Figure 8) that at pH 7.8 heating at 50 °C for 5 min is sufficient
90Y
chelation kinetics at pH )
to achieve 95% RCP for RP697. The maximum RCP was obtained by heating the reaction mixture at 50 °C for 10 min. Preparation of RP697 at 100 °C. In this experiment, we prepared RP697 using 100 µg of SU015 and 10 mg of GA for 20 mCi of 90YCl3 at pH 6.6, 7.2 or 7.8. The reaction mixture was heated at 100 °C for 5, 10, 15, and 30 min. Figure 9 shows the effect of heating time on the RCP of RP697 at different pH values. It is clear that 5 min heating is sufficient to complete the 90Y chelation at pH 7.8. Prolonged heating is detrimental due to hydrolysis and radiolytic decomposition of the radiolabeled peptide. 90Y-Labeling Efficiency of SU015. We studied the 90 Y-labeling efficiency of SU015 by determining the minimal amount of SU015 required to achieve 95% RCP for RP697. In this experiment, the heating temperature was 100 °C and the heating time was 10 min. Figure 10
90Y
and
177Lu
Labeling of Vitronectin Receptor Antagonist
Figure 8. Effect of heating (at 50 °C) time on RCP of RP697 at pH 7.2 and 7.8.
Figure 9. Effect of heating (at 100 °C) time on RCP of RP697 at pH 6.6, 7.2, and 7.8.
Figure 10. Effect of the pH on radiolabeling efficiency.
shows RCP data for RP697 prepared using 25, 50, or 100 µg of SU015, 10 mg of GA, 20 mCi of 90YCl3, at pH 6.6, 7.2, or 7.8. Obviously, the pH has dramatic impact on the 90Y-labeling efficiency of SU015. At pH 7.2-7.8, the minimum amount of SU015 required to achieve 95% RCP for RP697 is about 25 µg for 20 mCi of 90YCl3 corresponding to a SU015:90Y ratio of ∼30:1. DISCUSSION
Therapeutic radiopharmaceuticals are radiolabeled molecules designed to deliver therapeutic doses of ionizing radiation to specific disease sites (17). Most systemically administered therapeutic radiopharmaceuticals are small organic or inorganic compounds with definite composition. They can also be macromolecules such as monoclonal antibodies and antibody fragments labeled with a therapeutic radionuclide. Ideally, a therapeutic radiopharmaceutical should localize at the tumor sites in high concentration and deliver sufficient radiation dose to tumor cells, while clearing rapidly from the blood to minimize radiation damage to normal tissues.
Bioconjugate Chem., Vol. 12, No. 4, 2001 563
While tumor imaging relies heavily on a high targetto-background ratio for the best contrast, tumor therapy depends largely on a high concentration of radioactivity for a long duration. Thus, the therapeutic radiopharmaceutical must have high and fast tumor uptake, high tumor-to-background ratio, long tumor residence time, and fast renal clearance. High tumor uptake and fast renal clearance are particularly important to a high tumor-to-background ratio and to reduce the radiation burden to other major organs (kidney, liver, and bone marrow). The new radiopharmaceutical must have high radiochemical purity (RCP g 90%) and high solution stability. Unlike diagnostic radiopharmaceuticals, which are typically made by radiopharmacists before injection, therapeutic radiopharmaceuticals have to be manufactured and released under GMP (Good Manufacturing Practice) conditions. Therefore, the radiopharmaceutical must retain its chemical and biological integrity during release and transportation. Choice of Biomolecules. Angiogenesis, the formation of new blood vessels, is a requirement for malignant tumor growth and metastasis (18-20). The angiogenic process depends on vascular endothelial cell migration and invasion, regulated by cell adhesion receptors (2129). The integrin Rvβ3 (vitronectin receptor) is such a cell adhesion receptor and interacts with proteins and peptides containing the RGD tripeptide recognition sequence. Many peptide vitronectin receptor antagonists have very high binding affinity for the vitronectin receptor and have been shown to be able to inhibit neovascularization, tumor-induced angiogenesis, and tumor growth (27-33). Recently, Kessler and co-workers (34, 35) reported two 125I-labeled cyclic pentapeptides: 3-125I-D-Tyr4cyclo(RGDyV) and 3-125I-D-Tyr4-cyclo-(RGDyK(SAA1)) (SAA ) sugar amino acid). It was found that 3-125I-D-Tyr4-cyclo(RGDyV) has fast hepatobiliary and renal excretion (34). The tumor/muscle and tumor/blood ratios for melanoma in nude mice were 5.5 and 9.5, respectively, at 60 min postinjection. Substitution of leucine with a SAA-functionalized lysine amino acid residue resulted in improved blood retention time, renal excretion, and better targetto-background ratio. A blocking study using 3 mg/kg of the Rvβ3 selective cyclo(RGDfV) demonstrated that the localization of radioactivity in the tumor is due to Rvβ3 binding. In our previous communication (15), we reported a 99mTc-labeled vitronectin receptor antagonist (RP593) and its biological properties in two spontaneous tumor models (c-Neu Oncomouse and dog). The high tumor uptake and long tumor residence time (3.4% ID/g at 2 h and 1.5% ID/g at 24 h postinjection in the c-Neu Oncomouse tumor model) led us to prepare the corresponding 90Y and 177Lu complexes, which have the potential for radiotherapy of a wide spectrum of tumors. Choice of Radionuclides. Identifying an appropriate isotope for radiotherapy is often a difficult task and requires weighing a variety of factors (36, 37). These include tumor uptake, blood clearance, duration of receptor binding, rate of radiation delivery, half-life and specific activity of the radionuclide, and the feasibility of large-scale production of the radionuclide in an economical fashion. The ultimate goal for a therapeutic radiopharmaceutical is to deliver the maximum amount of radiation dose to the tumor cells to achieve a cytotoxic effect without causing unmanageable side-effects. The physical half-life of the therapeutic radionuclide should match the biological half-life of the radiopharmaceutical at the tumor site. If the half-life of the radionuclide is too short, much of the decay will have occurred
564 Bioconjugate Chem., Vol. 12, No. 4, 2001 Table 1. LC-MS Data for
177Lu
Liu et al.
and Nonradioactive Metal Complexes
complex
formula
molecular mass
found (M + H)+
found (M + 2H)2+
RP688 175Lu(SU015) Ca(SU015) Co(SU015) Cu(SU015) Zn(SU015) Fe(SU015)
C75H110N23O23176/177Lu
1877.8 1876.8 1741.8 1760.8 1766.4 1768.2 1758.7
1878.3 1877.5 1742.9 1760.8 1766.3 1768.6 1758.5
939.3 938.8 871.9 881.4 883.9 884.8 879.8
C75H110N23O23Lu C75H111N23O23Ca C75H111N23O23Co C75H111N23O23Cu C75H111N23O23Zn C75H110N23O23Fe
Table 2. Trace Metal Contaminants in a Commerical Stock Solution and the Stability Constants for Their DOTA Complexes
90Y
metal ion
max. amount (mg/Ci Y-90)
max. amount (mmol/Ci Y-90)
stability constant for M-DOTA
Y Al Ca Ce Co Cr Cu Fe La Ni Pb Sr Zn Zr
∼2.0
