Synthesis and evaluation of novel bifunctional chelating agents based on 1,4,7,10-tetraazacyclododecane-N,N′,N″,N‴-tetraacetic acid for radiolabeling proteins

Nuclear Medicine and Biology 30 (2003) 581–595

www.elsevier.com/locate/nucmedbio

Synthesis and evaluation of novel bifunctional chelating agents based on 1,4,7,10-Tetraazacyclododecane-N,N⬘,N⬙,N⵮-Tetraacetic acid for radiolabeling proteins L.L. Chappell, D. Ma, D.E. Milenic, K. Garmestani, V. Venditto, M.P. Beitzel, M.W. Brechbiel* Radioimmune & Inorganic Chemistry Section, ROB, CCR, NCI, NIH, Bethesda, MD, 20892-1002, USA Received 1 November 2002; received in revised form 27 January 2003; accepted 15 February 2003

Abstract Detailed synthesis of the bifunctional chelating agents 2-methyl-6-(p-isothiocyanatobenzyl)-1,4,7,10-tetraazacyclododecane-1,4,7,10tetraacetic acid (1B4M-DOTA) and 2-(p-isothiocyanatobenzyl)-5, 6-cyclohexano-1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetate (CHX-DOTA) are reported. These chelating agents were compared to 2-(p-isothiocyanatobenzyl)-1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetic acid (C-DOTA) and 1, 4, 7, 10-Tetraaza-N-(1-carboxy-3-(4-nitrophenyl)propyl)-N⬘, N⬙, N⵮-tris(acetic acid) cyclododecane (PA-DOTA) as their 177Lu radiolabeled conjugates with Herceptin™. In vitro stability of the immunoconjugates radiolabeled with 177Lu was assessed by serum stability studies. The in vivo stability of the radiolabeled immunoconjugates and their targeting characteristics were determined by biodistribution studies in LS-174T xenograft tumor-bearing mice. Relative radiolabeling rates and efficiencies were determined for all four immunoconjugates. Insertion of the 1B4M moiety into the DOTA backbone increases radiometal chelation rate and provides complex stability comparable to C-DOTA and PA-DOTA while the CHX-DOTA appears to not form as stable a 177Lu complex while exhibiting a substantial increase in formation rate. The 1B4M-DOTAmay have potential for radioimmunotherapy applications. Published by Elsevier Inc. All rights reserved. Keywords: DOTA; Bifunctional Chelating Agent; Lutetium-177; monoclonal antibody; biodistribution; stability

1. Introduction Use of radiolabeled monoclonal antibodies continues to be an attractive and burgeoning area of investigation as evidenced by the recent approval of the anti-CD20 targeting Zevalin (90Y radiolabeled-Rituxan) for the treatment of non-Hodgkins lymphoma (NHL) [15]. One critical variable that influences the effectiveness of radioimmunotherapy (RIT) is the choice of the radionuclide and its associated emission characteristics [30]. Equally important is the choice of the chemical means by which the radionuclide is bound to the protein [11]. Therapy of solid tumors has been exacted with ␤- -emitting radionuclides such as 90Y (T1/2 ⫽ 2.67 d, Etotal ⫽ 939.1 keV) and 177Lu (T1/2 ⫽ 6.71 d, Etotal ⫽ 146.7 keV) [33]. The optimal range of the ␤- particles emitted by these isotopes for effective killing of cells, as * Corresponding author. E-mail address: [email protected] (M.W. Brechbiel).

defined by the physical properties of range and energy deposition, has been reported to be 28 – 42 mm and 1.2–3.0 mm, respectively, and the majority of energy deposition does not occur immediately along the emission track [24]. For RIT applications, 90Y or 177Lu must be linked as a metal complex to a monoclonal antibody (mAb) or immunoprotein via a suitable bifunctional chelating agent that possesses acceptable thermodynamic and kinetic stability to minimize release of the isotope and hence in vivo toxicity [11]. Derivatives of 1, 4, 7, 10-tetraazacyclododecane-N, N⬘, N⬙, N⵮-tetraacetic acid (DOTA) conjugated to monoclonal antibodies have been shown to form exceedingly stable complexes with the lanthanide metal ion radionuclides for in vivo applications [20,23,27]. This has resulted in the development of several bifunctional derivatives of DOTA for radiolabeling proteins, including 2-(p-isothiocyanatobenzyl)-1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetic acid (C-DOTA) [20,23,27] and 1, 4, 7, 10-tetraaza-N-(1-carboxy-3-(4-nitrophenyl)propyl)-N⬘, N⬙, N⵮-tris(acetic acid)

0969-8051/03/$ – see front matter Published by Elsevier Inc. All rights reserved. doi:10.1016/S0969-8051(03)00033-7

582

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

Fig. 1. Structure of C-DOTA and PA-DOTA.

cyclododecane (PA-DOTA) [4]. However, the formation kinetics associated with the DOTA chelating agent have also been noted as less than optimal requiring either lengthy radiolabeling protocols and/or the use of elevated temperatures to achieve acceptable yields and specific activities [16,29]. Alternate approaches for using DOTA have resulted in the development of numerous derivatives wherein modifications of the DOTA framework have been explored to address this deficiency. This strategy has been executed by either the addition of external chelating moieties, [31] the conversion of one of the carboxylates to an amide for conjugation purposes, [16,17,28] or alteration of the carbon chain length of the carboxylate. [14] While all of these investigations have met with varying levels of success and utilization, the inherent obstacle of slow formation kinetics and resultant low radiolabeling efficiency remains unsolved. One aspect that all of these previously reported derivatives have in common is that the tetraaza ring was retained without any modification. There have been numerous detailed studies of the mechanism of metal ion complex formation with DOTA as well as with amido and phosphorus analogues [9,13]. These studies generally propose a twostep mechanism of electrostatic capture of the metal followed by encapsulation during which there is deprotonation of the amines and an associated energy cost due to arrangement of the carboxylates in the proper geometries for metal binding [13]. One aspect of this process also includes arranging the 12-membered ring itself into the proper spatial geometry, a process that also has an associated energy cost. The final geometry of the lanthanide DOTA complexes has been well reported [9,13]. Forsberg and co-workers have reported on the preferential ring geometry of these complexes via modeling the tetra-amido DOTA complexes [9]. Thus, a study of DOTA derivatives wherein the macrocyclic ring was pre-arranged or pre-organized in some aspect was proposed to lower the energy barrier to complex formation thereby potentially increasing the rate of complex formation. To this end, we have designed and synthesized two derivatives, 2-methyl-6-(p-isothiocyanatobenzyl)1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (1B4M-DOTA) and 2-(p-isothiocyanatobenzyl)-5, 6-cyclohexano-1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetate (CHX-DOTA) wherein additional alkyl substitution has been added to influence the 12-membered ring conformation. Herein, the detailed syntheses of 1B4M-DOTA and

CHX-DOTA and the preparation of their conjugates with monoclonal antibody Herceptin™ are reported. This study also includes their radiolabeling with 177Lu and an assessment of relative in vitro stability and labeling kinetics. Lastly, the in vivo stability of the radioimmunoconjugates (RICs) of 1B4M-DOTA and CHX-DOTA as compared to that of the 177Lu labeled C-DOTA- and PA-DOTA-Herceptin™ immunoconjugates were evaluated in mice bearing human colon carcinoma (LS-174T) xenografts.

2. Materials and methods 2.1. Synthesis (d)-p-Nitrophenylalanine-(R,R)-trans-1, 2-aminocyclohexyl amide dihydrochloride, p-nitro-phenylalanine-2aminopropylamide, and tert-butyloxtcarbonyl-iminodiacetic acid disuccinimidyl ester were prepared as described in the literature [2,20,33]. The tert-butyl bromoacetate was purchased from Fluka (Ronkonkoma, NY). All other reagents were purchased from Aldrich (Milwaukee, WI), Sigma (St. Louis, MO), or Fluka (Ronkonkoma, NY) and used without further purification. Ion-exchange resins were obtained from Bio-Rad Laboratories (Richmond, CA). 1 H and 13C NMR were obtained using a Varian Gemini 300 instrument (Palo Alto, CA). Chemical shifts are reported in ppm on the scale relative to TMS, TSP, or solvent. Proton chemical shifts are annotated as follows: ppm (multiplicity, integral, coupling constant (Hz)). Chemical ionization mass spectra (CI-MS) were obtained on a Finnegan 3000 instrument (San Jose, CA). Fast atom bombardment mass spectra (FAB-MS) were acquired on an Extrel 400 (Pittsburgh, PA). Exact mass FAB-MS was obtained on a JOEL SX102 spectrophotometer (Peabody, MA). The exact mass measurements in FAB were obtained using an accelerating voltage of 10 kV with the samples being desorbed from a matrix using 6 keV xenon atoms. Mass measurements were performed at 10,000 resolution using electric field scans with the sample peak bracketed by two poly(ethylene glycol) reference ions. Elemental analyses were performed by Atlantic Microlabs (Atlanta, Georgia). Optical rotations were measured using a Perkin elmer 341 instrument with a 100 mm cell at 20°C. The reverse-phase HPLC (RP-HPLC) system components (Beckman Instruments, Fullerton, CA) were as follows: a pair of 114M pumps, a

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

165 dual-wavelength variable uv detector, operated through a 406 analogue interface module, controlled by System Gold software, and a Beckman 4.6 ⫻ 25 cm C-18 ultrasphere ODS 5␮m column. A 25 min gradient from 100% 0.05M Et3N/HOAc to 100% MeOH at 1mL/min was employed for all analytical synthesis HPLC chromatography. The size-exclusion HPLC (SE-HPLC) system components included a Dionex (Sunnyvale, CA) gradient pump with Waters Corp. (Milford, MA) 717 plus autosampler, a Gilson Inc. (Middleton, WI) 112 UV detector (280 nm) and an in-line IN/US Systems, Inc. (Tampa, FL) ␥-Ram Model 2 radiodetector using a polymer based size-exclusion HPLC column (TSK-G3000PW, 7.5 mm ⫻ 30 cm, TosoHaas, Japan) with a guard column (TSK Guard PWH, 7.5 mm ⫻ 7.5 cm) eluted isocratically at 1 mL/min with phosphate buffered saline (PBS, pH ⫽ 7.4) obtained from Mediatech (Herndon, VA). 2.1.1. Synthesis of 1-N-tert-butyloxycarbonyl-5-methyl-9(p-nitrobenzyl)-3, 8, 12-tri-oxo-1,4,7,10tetraazacyclododecane (3) 1,4-Dioxane (3.5 L) was heated to 90°C in a 5 L 3-necked Morton flask, p-nitro-phenylalanine-2-aminopropylamide (3) (5.32 g, 20 mmol) was dissolved in anhydrous DMF (40 mL) and taken up into a gas-tight syringe. Additional DMF was added to bring the final volume to 50 mL. BOC-iminodiacetic acid disuccinimidyl ester (8.54 g, 20 mmol) was dissolved in DMF, taken up into a gas-tight syringe and DMF was added to bring the final volume to 50 mL. The syringes were loaded onto a Sage Model M362 syringe pump (Orion Research, Beverly, MA) and the two solutions were added to the hot 1,4-dioxane such that the addition was complete within 24 hr. Three more additions of 20 mmol of each reactant were added with a fifth and final addition of 18 mmol over the following 5 d. After completion of the final addition, the reaction was heated for an additional 18 hr and then cooled to room temperature. The reaction was concentrated to a thick oil under vacuum and the brown residue taken up in CHCl3 (300 mL). Addition of CHCl3 resulted in a suspension of an off-white precipitate that was collected after cooling, washed with CHCl3, and dried under vacuum. The filtrate was washed with 1 M HCl (2 ⫻ 100 mL), saturated NaCl solution (2 ⫻ 100 mL), 1 M NaHCO3 (2 ⫻ 200 mL) and water (2 ⫻ 200 mL) while adding more CHCl3 (100 –150 mL) to counter additional precipitate. This was also collected by filtration and after drying over anhydrous Na2SO4, the CHCl3 solution was reduced to form additional precipitate that was also collected. TLC of the crude precipitated material indicated one major product followed closely by a faint second product (Rf ⫽ 0.14, 0.11 silica gel, 5% MeOH in CHCl3). Extensive silica gel chromatography using a gradient from 0-5% MeOH in CHCl3 was able to provide the major product free from the trailing material (11.8 g, 26%). Only trace amounts of the minor product could be isolated in good purity, with the bulk always contaminated with the major product.

583

H NMR (dmso-d6) ␦ 1.01 (d, 3H, J ⫽ 6.6), 1.376, 1.386 (2s, 9H), 2.63 –2.86 (m, 1H), 2.938 (dd, 1H, J ⫽ 13.65, 9.9), 3.02 –3.19 (m, 1H), 3.25 –3.41 (m, 2H), 3.744 (dd, 1H, J ⫽ 16.0, 12.0), 3.92 –3.98 (m, 1H), 4.02 – 4.20 (m, 1H), 4.38 – 4.50 (m, 1H), 7.135 (dt, 1H, J ⫽ 21.9, 5.4), 7.529 (dd, 2H, J ⫽ 8.7, 2.4), 7.772 (dd, 1H, J ⫽ 16.0, 9.0)), 8.138 (d, 2H, J ⫽ 8.7), 8.561 (d, 1H, J ⫽ 8.4); 13C NMR (dmso-d6) ␦ 17.64, 27.85, 24.83,34.92 (1C), 42.81,42.93 (1C), 43.96,44.17 (1C), 48.99,49.55 (1C), 49.77,50.31 (1C), 54.20,54.31 (1C), 79.68, 123.21, 130.42, 146.19, 146.38,146.45 (1C), 154.74,154.88 (1C), 168.68,168.84 (1C), 168.91,169.08 (1C), 170.53,170.67 (1C); FAB-MS (glycerol) m/e 464 (M⫹⫹1); Anal. Calcd. for C21H29N5O7: C, 54.41; H, 6.32; N, 15.11. Found: C, 54.19; H, 6.36; N, 14.95. 1

2.1.2. Synthesis of 2-methyl-6-(p-nitrobenzyl)-1,4,7,10tetraazacyclododecane-tetrahydrochloride (1B4M-cyclen tetrahydrochloride) (5) Anhydrous 1, 4-dioxane (200 mL) was chilled in an ice bath and saturated with HCl(g). The above product, the major isomer, (4.40 g, 9.50 mmol) was added and HCl(g) was bubbled through the reaction mixture for an additional 1 h. The reaction mixture was stirred at room temperature for 18 h. Et2O (approx. 150 mL) was added and the mixture was chilled in the freezer for 1 d. The precipitate was collected by vacuum filtration and washed with Et2O (400 mL). The light tan powder was vacuum dried at 60°C overnight. This product was suspended in anhydrous THF (50 mL), the flask was chilled in an ice bath and 1 M BH3.THF (58 mL) was added. The mixture was warmed to room temperature and heated at 50°C overnight (18 h) with the temperature being controlled by an I2R Thermowatch L7-110SA (Cheltenham, PA). Progress of the reaction was monitored by quenching an aliquot with MeOH, heating it with conc. HCl for 2 hours, and then after removal of the solvents, analysis by HPLC. After 72 hr, the reaction was deemed completed without formation of any aniline side-product and after cooling to room temperature the reaction was quenched with excess MeOH. The solution was stirred for an additional 24 hr after which it was rotary evaporated to a gummy residue that was held under vacuum for 18 hr. This residue was then taken up in 100% EtOH (120 mL) and while cooling in an ice bath, saturated with HCl(g), followed by a vigorous reflux for 6 hr. After cooling to room temperature, the suspension was chilled at 4°C for 24 hr. The product was collected by filtration, washed with ether and dried under vacuum (2.86 g, 67%). 1 H NMR (D2[r]O) ␦ 1.14 –1.53 (m, 3H), 2.80 – 3.98 (complex m, 16H), 7.50 –7.58 (m, 2H), 8.18 – 8.29 (m, 2H); Anal. HPLC tR ⫽ 10.37 min; FAB-MS (glycerol) m/e 322 (M⫹⫹1); Anal. Calcd. For C16H27N5O2(HCl)3(H2O): C, 42.81; H, 6.08; N, 15.61. Found: C, 42.71; H, 6.16; N, 15.71.

584

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

2.1.3. Tetra-tert-butyl 2-methyl-6-(p-nitrobenzyl)-1,4,7,10tetraazacyclododecane-1,4,7,10-tetraacetate (1B4MDOTA-tBu ester) (6) The above cyclen 5 (670 mg, 1.49 mmol) was dissolved in water (10 mL) and the pH increased to approximately 13 by addition of solid NaOH. The reddish aqueous layer was extracted with CHCl3 (3 ⫻ 80 mL). The combined yellow CHCl3 layers were reduced to dryness and dried under vacuum (493 mg) 1 H NMR (CDCl3) ␦ 0.91 (m, 3H), 2.3 –2.9 (m, 20H), 7.30 (d, 2H), 8.09 (d, 2H,); 13C NMR (CDCl3) ␦ 18.47, 39.96, 45.06, 45.37, 45.55, 46.28, 49.74, 53.87, 57.33, 123.57, 129.82, 130.00, 146.58, 147.73; MS (CI/NH3) m/e 332 (M⫹⫹1). The free base cyclen (484 mg) generated above was dissolved in anhydrous DMF and chilled in an ice bath. Tert-butyl bromoacetate (1.17 g, 6.08 mmol) was added. The reaction was stirred for 45 min and warmed to room temperature. A solution of Na2CO3 in water (645 mg in 13 mL) was added and the mixture stirred for 2 h. Toluene (10 mL) was added and the mixture stirred for an additional 3 h. The reaction mixture was poured into a separatory funnel and the aqueous layer drained. The orange toluene layer was saved. The aqueous layer was extracted with CHCl3 (2 ⫻ 80 mL) and the CHCl3 layers were combined with the toluene. The combined organic layers were reduced to dryness. This residue was purified on a silica gel column (2.5 cm ⫻ 35 cm) and eluted with a gradient of 5-10% MeOH in CHCl3 and finally 10% NH4OH in MeOH. Early fractions contained high Rf materials and later fractions contained the product, as determined by mass spectrometry. The later fractions containing the product were combined and the solvent removed by rotary evaporation to leave an orange foam (501 mg, 42%) 1 H NMR (CDCl3) ␦0.8 (br, 3H), 1.37, 1.39 (singlets, 36H), 2.0 – 4.0 (m, 24H), 7.3 (m, 2H), 8.07 (m, 2H, Ar); MS (CI/NH3) m/e 779 (M⫹⫹1), 801 (M ⫹ 23); Anal. Calcd. for C40H67N5O10: C, 61.75 H, 8.68; N, 9.00. Found: C, 55.95; H, 8.05; N, 8.11. 2.1.4. 2-methyl-6-(p-nitrobenzyl)-1,4,7,10tetraazacyclododecane-1,4,7,10-tetraacetic acid (1B4MDOTA-NO2) (7) Tetra-tert-butyl ester 6 (214 mg, 0.275 mmol) was heated to reflux in conc. HCl (8 mL) for 6 h. The solvent was removed by rotary evaporation. The residue was taken up in H2O (1 –2 mL) and lyophilized to isolate the product as an orange solid (149 mg, 75%) 1 H NMR (D2O pH 1.5) ␦ 1.2 (m, 3H), 3.0 – 4.2 (m, 24H), 7.59 (m, 2H), 8.26 (d, 2H); 1H NMR (D2O pH 14) ␦ 0.55, 0.7, 0.8 (3 d, 3H), 2.2–3.6 (m, 24H), 7.4 (m, 2H, Ar), 8.18 (d, 2H); MS (CI/NH3) m/e 554 (M⫹⫹1); HPLC tR ⫽ 11.6 min; Anal. Calcd. for C24H35N5O10(HCl)4(H2O): C, 40.18; H, 5.76; N, 9.76. Found: C, 40.76; H, 5.68; N, 9.79.

2.1.5. 2-methyl-6-(p-aminobenzyl)-1,4,7,10tetraazacyclododecane-1,4,7,10-tetraacetic acid (1B4MDOTA-NH2) (8) A Schlenk flask was charged with 10% Pd/C (28.9 mg) and H2O (5 mL) and fitted onto an atmospheric hydrogenator. The apparatus was flushed with H2(g) twice to saturate the catalyst. A solution of 7 (104 mg, 0.188 mmol) in H2O (5 mL) was injected via syringe into the flask. The hydrogenation was allowed to proceed until the uptake of H2(g) ceased. The reaction mixture was filtered through a bed of Celite 577 on a medium glass frit. The filtrate was reduced to dryness by rotary evaporation and the residue taken up in water (1–2 mL). This solution was then lyophilized to give the aniline as a solid (87 mg, 88%). 1 H NMR (D2O pH 1.5) ␦ 1.1 (m, 3H), 3.0 – 4.2 (m, 24H), 7.4 (m, 4H); 1H NMR (D2O pH 14) ␦ 0.6, 0.7, 0.9 (m, 3H), 2.2–3.6 (m, 24H), 6.8 (m, 2H), 7.1 (d, 2H); FAB-MS (glyceraol) m/e 524 (M⫹⫹1); Anal. HPLC tR ⫽ 8.1 min; HRFAB M ⫹ H⫹ calcd for C24H38N5O8 524.2720 found [HRFAB] m/e ⫽ 524.2723, error ⫽ ⫹0.6 ppm. 2.1.6. 2-methyl-6-(p-isothiocyanatobenzyl)-1,4,7,10tetraazacyclododecane-1,4,7,10-tetraacetic acid (1B4MDOTA) (1) A 1 M solution of SCCl2 in CHCl3 (61 ␮L) was added to aniline 8 (35 mg, 0.055 mmol) dissolved in H2O (0.5 mL) in a 3 dram vial. The mixture was vigorously stirred with a spin vane for 2 h at room temperature. The aqueous layer was transferred with a pipet to a round bottom flask and the CHCl3 layer washed with H2O (3 ⫻ 0.5 mL). The combined aqueous layers were lyophilized to leave the 1B4M-DOTA as a yellow solid (28 mg). 1 H NMR (D2O pH 1.5) ␦ 1.2 (m, 3H), 2.9 – 4.2 (m, 24H), 7.4 (m, 4H); 1H NMR (D2O pH 14) ␦ 0.6, 0.7, 0.9 (3 m, 3H), 2.2 –3.6 (m, 24H), 6.8, 7.3 (m, 4H); MS (CI/NH3) m/e 566 (M⫹⫹1); HPLC tR ⫽ 19.7 min (minor), 20.2 min (major), HR-FAB M⫹H⫹ calcd for C25H35N5O8S 566.2285 found [HRFAB] m/e ⫽ 566.2301, error ⫽ ⫹2.9 ppm. 2.1.7. 1-Tert-butyloxycarbonyl 2-(p-nitrobenzyl)- 3, 8, 12trioxo-5, 6-cyclohexano-1, 4, 7, 10-tetraazacyclododecane (CHX-cyclen-BOC-triamide) (10) Anhydrous 1,4-dioxane (3.5 L) was heated to 90°C in a 5 L 3-necked Morton flask, (d)-p-nitrophenylalanine-(R,R)trans-1, 2-aminocyclohexyl amide dihydrochloride [33] (3.79 g, 10 mmol) was dissolved in anhydrous DMSO (35 mL) and Et3N (3 mL, 22 mmol) was added. The mixture was stirred for 30 min and the Et3N/HCl was removed by filtration. The filtrate was taken up into a gas-tight syringe and additional DMSO was added to bring the final volume up to 50 mL. BOC-iminodiacetic acid disuccinimidyl ester [20] (4.27 g, 10 mmol) was dissolved in DMF, taken up into a gas-tight syringe and DMF was added to bring the final volume up to 50 mL. The syringes were loaded onto a Sage Model M362 syringe pump and the two solutions were added to the hot 1,4-dioxane such that the addition was

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

complete within 24 hr. Three more additions of 10 mmol of each reactant were added via syringe pump over the following 5 d. After the fourth addition, the reaction was heated for an additional 18 hr and then cooled to room temperature. The reaction was concentrated down to a thick oil under vacuum and the brown residue dissolved in CHCl3 (100 mL). The CHCl3 layer was washed with 1 M HCl (2 ⫻ 100 mL), saturated NaCl solution (2 ⫻ 100 mL), 1 M HCO3- (2 ⫻ 200 mL) and water (2 ⫻ 200 mL). More CHCl3[r] (100 –150 mL) was added during the extractions. The CHCl3 layer was dried over anhydrous Na2SO4, filtered and the filtrate reduced to dryness by rotary evaporation. The crude product was divided and each portion passed over a short silica gel column with 5% MeOH in CHCl3 for gross purification. All fractions which contained product (Rf ⫽ 0.6, silica gel, 10% MeOH in CHCl3) were combined. The crude product was only slightly soluble in CHCl3 and to eliminate any precipitation problems was divided into thirds and applied as a slurry to three separate silica gel columns. For this chromatography, a gradient from 0 to 5% MeOH in CHCl3 was used, all fractions that contained product were combined, reduced to dryness and vacuum dried to give the product as a light brown solid (5.89 g, 29%) 1 H NMR (dmso-d6) ␦ 1.1–1.5 (m, s, 13 H), 1.7–2.0 (m, 4H), 3.1–3.3 (m, 3H) 3.8 – 4.4 (m, 6H), 6.92 (t, 1H), 7.58 (d, 2H), 8.21 (d, 2H), 8.34 (d, 1H), 8.57 (dd, 1H); 13C NMR (dmso-d6) ␦ 24.28, 24.70, 27.86, 31.01, 32.41, 35.69, 47.23, 48.20, 49.11, 50.63, 55.30, 56.64, 123.48, 130.46, 146.13, 146.55, 155.18, 168.65, 169.56, 171.51; MS (FAB/glycerol) m/e 504 (M⫹⫹1); [␣]23d[r] 1.1° (DMSO, c ⫽ .31); Anal. HPLC tR ⫽ 23.4 min; Anal. Calc. for C24H33N5O7: C, 56.05; H, 6.61; N, 13.91. Found: C, 56.27; H, 6.66; N, 13.55. 2.1.8. 2-(p-nitrobenzyl)- 3, 8, 12-trioxo-5, 6-cyclohexano1, 4, 7, 10-tetraazacyclododecane hydrochloride (CHXcyclen triamide hydrochloride) 1,4-Dioxane (80 mL) was chilled in an ice bath and saturated for 2 h with HCl(g). The above product (3.11 g, 6.2 mmol) was added and HCl(g) was bubbled through the reaction mixture for an additional 1 h. The reaction mixture was stirred at room temperature for 18 h. Diethyl ether (approx. 150 mL) was added and the mixture was chilled in the freezer for 1 d. The precipitate was collected by vacuum filtration and washed with Et2O (400 mL). The light brown powder was vacuum dried at 60°C overnight (2.78 g, 98%) 1 H NMR (dmso-d6) ␦ 1.26 (m, 4H), 1.70 (m, 4H) 3.10 (d, 2H, J ⫽ 8.1), 3.45 (s, 7H), 3.80 (m, 4H) 4.2 (m, 2H), 7.03 (d, 1H, J ⫽ 7.8), 7.56 (d, 2H, J ⫽ 9.0), 8.16 (d, 2H, J ⫽ 9.0), 8.60 (d, 1H, J ⫽ 9.6), 9.14 (d, 1H, J ⫽ 7.2); 13C NMR (dmso-d6) ␦ 24.27, 24.58, 31.20, 32.35, 35.45, 44.98, 45.40, 50.93, 54.27, 57.06, 123.48, 130.64, 145.94, 146.61, 165.31, 165.62, 171.08; MS (FAB/glycerol) m/e 404 (M⫹⫹1); [␣]23d[r] 21.5° (DMSO, c ⫽ .13); Anal. HPLC tR ⫽ 20.8; Anal. Calcd for C19H25N5O5(HCl)(H2O): C, 49.84; H, 6.16; N, 15.29. Found: C, 49.60; H, 6.00; N, 15.16.

585

2.1.9. 2-(p-nitrobenzyl)-5, 6-cyclohexano-1, 4, 7, 10tetraazacyclododecane hydrochloride (CHX-cyclen) (11) The above hydrochloride (820 mg, 1.79 mmol) was combined with anhydrous THF (30 mL), the flask was chilled in an ice bath and 1 M BH3 in THF (15 mL, 15 mmol) was added. The mixture was warmed to room temperature and heated at 50°C for 18 h. MeOH (200 mL) was added dropwise to the reaction mixture and the solution was reduced to dryness. The yellow solid was transferred to a 250 mL round bottom flask with absolute EtOH (50 mL) and chilled in an ice bath. HCl(g) was bubbled through the reaction mixture for 2 h and then heated to reflux for 12 h. Et2O (approx. 150 mL) was added and the mixture chilled in the freezer overnight. The hygroscopic pale green-yellow precipitate was collected by filtration, washed with Et2O, and vacuum dried (650 mg, 72%). 1 H NMR (D2O, pH ⫽ 1) ␦ 1.4 –2.0 (m, 7H), 2.0 – 4.2 (m, 16H), 7.68 (m, 2H), 8.32 (m, 2H); Anal. HPLC tR ⫽ 15.1 min; MS (CI/NH3) m/e 362 (M⫹⫹1); [␣]23d[r] ⫺30° (H2O, c ⫽ .203);Anal. Calcd. for: C19H31N5O2(HCl)3.5(H2O): C, 45.00; H, 7.25; N, 13.81. Found: C, 45.93; H, 7.20; N, 13.24. 2.1.10. Tetra-tert-butyl 2-(p-nitrobenzyl)-5, 6-cyclohexano1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetate (CHX-DOTA-tBu-ester) (12) The CHX-cyclen 11 (650 mg, 1.28 mmol) was dissolved in water (10 mL) and the pH raised to ⬃13 with NaOH pellets. The aqueous layer was extracted with CHCl3 (3 ⫻ 50 mL). The combined CHCl3 layers were dried over anhydrous Na2SO4, filtered and the filtrate reduced to dryness. The yellow solid was vacuum dried (492 mg, ⬃ 100%). 1 H NMR (CDCl3) ␦ 0.9 –1.3 (m, 6H), 1.7 (m, 4H), 1.8-3.0 (m, 17H), 7.37 (d, 2H), 8.15 (d, 2H); 13C NMR (CDCl3) ␦ 24.97, 25.27, 30.98, 32.26, 39.72, 43.55, 46.40, 47.13, 47.56, 48.47, 57.82, 60.97, 123.39, 130.07, 146.40, 147.73; MS (CI/NH3) m/e 362 (M⫹⫹1). The free base of 11 (473 mg, 1.31 mmol) was dissolved in DMF (5 mL) and chilled in an ice bath. Tert-butyl bromoacetate (1.04 g, 5.33 mmol) was added and the reaction mixture stirred for 30 min. A solution of Na2CO3 in water (557 mg/11 mL) was added and this was stirred for 1.5 h. Toluene (5 mL) was added and the reaction mixture was stirred for another 2.5 h. The reaction mixture was poured into a separatory funnel and the aqueous layer was drained and the orange toluene layer retained. The aqueous layer was extracted with CHCl3 (2 ⫻ 40 mL) and the CHCl3 layers were combined with the toluene solution. The combined organic layers were reduced to dryness. This residue was purified on two consecutive silica gel column (2.5 cm ⫻ 35 cm) eluted with 5% MeOH in CHCl3. Early fractions contained high Rf materials and middle fractions (Rf ⫽ 0.54) contained the product. The combined fractions were concentrated by rotary evaporation to yield the product as an orange oil (450 mg, 42%) 1 H NMR (dmso-d6) ␦ 1.46 (m, 44H), 1.6 –3.6 (m, 23H),

586

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

7.43 (d, 1H, J ⫽ 8.7), 7.78 (d, 1H, J ⫽ 7.8), 8.14 (d, 1H, J ⫽ 8.7), 8.20 (d, 1H, J ⫽ 8.7); 13C NMR (dmso-d6) ␦ 22.42, 22.90, 23.63, 24.36, 25.21, 31.34, 32.01, 43.12, 44.46, 47.98, 48.28, 50.71, 51.80, 52.05, 55.51, 55.63, 61.28, 62.25, 81.49, 81.61, 81.80, 123.63, 130.00, 130.43, 146.28, 148.58, 172.63, 172.93, 173.11; MS (CI/NH3) m/e 818 (M ⫹ H⫹); Anal. Calcd. for: C43H71N5O10: C, 63.12; H, 8.76; N, 8.56. Found: C, 62.85; H, 8.77; N, 8.41. 2.1.11. 2-(p-Nitrobenzyl)-5, 6-cyclohexano-1, 4, 7, 10tetraazacyclododecane- 1, 4, 7, 10-tetraacetic acid (CHXDOTA-NO2) (13) Tetra-tert-butyl ester (450 mg, 0.55 mmol) was refluxed in conc. HCl (aq) (10 mL) for 6 h. The solvent was removed by rotary evaporation and the residue taken up in H20 (1 –2 mL) and freeze dried. The product was isolated as a light brown solid (390 mg, 94%). 1 H NMR (D2O, pH ⫽ 1.5) ␦ 1.2 –2.0 (m, 8H), 2.0 – 4.2 (m, 23H), 7.55 (d, 2H, J ⫽ 8.7), 8.25 (d, 2H, J ⫽ 9.0); MS (CI/NH3) m/e 594 (M⫹⫹1), 616 (M⫹⫹23); HPLC tR ⫽ 13.3 min; [␣]23d[r] –55.3° (H2O, c ⫽ .0525); Anal. Calcd. for C27H39N5O10(HCl)4(H2O): C, 42.81; H, 5.99; N, 9.25. Found: C, 42.62; H, 5.93; N, 9.18. 2.1.12. 2-(p-Nitrobenzyl)-5, 6-cyclohexano-1, 4, 7, 10tetraazacyclododecane- 1, 4, 7, 10-tetraacetic acid (CHXDOTA-NO2) (13) (Alternate Purification) Tetra-amine 11 (1.55 g, 3.06 mmol) was alkylated as described above. The crude ester was treated with trifluoroacetic acid (50 mL) for 18 h. After elimination of the acid by rotary evaporation, the residue was dried under vacuum for 24 h. The residue was taken up in minimal H2O and loaded onto cation ion-exchange resin (2.6 ⫻ 30 cm, AG50wX8, 200 – 400 mesh, H⫹ form) and washed with H2O until the eluant was above pH 5.0. The crude product was eluted from the column with 2 M NH4OH (1L). The basic solution was rotary evaporated to leave the crude product as a brown solid after drying under vacuum for 24 h. This material was taken up in minimal H2O and loaded onto an anion ion-exchange resin column (1.6 ⫻ 30 cm, AG1x8, 200 – 400 mesh, HOAc form) and eluted with a 0.0 –1.5 M HOAc linear gradient (2L total) collected in 18 ⫻ 150 test tubes. The reaction by-product arising from trialkylation as determined by FAB-MS was eluted first (tubes 10 –17) while the tetraacetic acid product eluted later (tubes 25–36). The contents of the tubes were combined and concentrated to ⬃50 mL after which the product was isolated as a white powder after lyophilization (860 mg, 47%). This product was identical in all regards to that isolated as the tetra-ester followed by acidic deprotection of the esters. 2.1.13. 2-(p-Aminobenzyl)-5, 6-cyclohexano-1, 4, 7, 10tetraazacyclododecane-1, 4, 7, 10-tetraacetate (CHXDOTA-NH2) (14) A Schlenk flask was charged with 10% Pd/C (29 mg) and H2O (5 mL) and fitted onto an atmospheric hydrogenator.

The apparatus was flushed with H2(g) two times in order to fully saturate the catalyst. A solution of 13 (102 mg, 0.13 mmol) in H2O (5 mL) was injected via syringe into the flask. The hydrogenation was allowed to proceed until the uptake of H2(g) ceased. The reaction mixture was filtered through a bed of Celite 577 packed in a medium glass frit. The filtrate was reduced to dryness by rotary evaporation and the residue taken up in water (1–2 mL). This was lyophilized to leave the aniline as a pale yellow solid (97 mg, ⬃100%). 1 H NMR (D2O, pH ⫽ 1.5) ␦ 1.2–2.0 (m, 8H), 2.0 – 4.2 (m, 23H), 7.44 (m, 4H); 1H NMR (D2O, pH ⫽ 14) ␦ 0.8 –1.8 (m, 8H), 1.8 –3.6 (m, 23H), 6.83 (m, 2H), 6.08 (m, 2H); FAB-MS (glycerol) m/e 564 (M⫹⫹1); [␣]23d[r] –51.5° (H2O, c ⫽ .05); Anal. HPLC tR ⫽ 15.8 min; HR-FAB M ⫹ H⫹ calcd for C27H41N5O8Na 586.2853 found [HRFAB] m/e ⫽ 586.2841, error ⫽ ⫺2.1 ppm. 2.1.14. 2-(p-Isothiocyanatobenzyl)-5, 6-cyclohexano-1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetate (CHX-DOTA) (2) A 1 M solution of SCCl2 in CHCl3 (55 ␮L) was added to CHX-DOTA-NH2 (30.5 mg, 0.049 mmol) dissolved in H2O (0.5 mL) in a 3 dram vial. The mixture was stirred rapidly for 2 h at room temperature. The aqueous layer was decanted with a pipette into a round bottom flask and the CHCl3 layer washed with H2O (3 ⫻ 0.5 mL). The combined aqueous layers were lyophilized to give CHX-DOTA as a yellow solid (34.6 mg, 94%). 1 H NMR (D2O, pH ⫽ 1) ␦ 1–1.6 (m, 4H), 1.6 –2.0 (m, 3H), 2.0 –2.6 (m, 5H), 2.– 4.0 (m, 18H), 7.38 –7.51 (m, 4H); 1 H NMR (D2O, pH ⫽ 14) ␦ 0.8 –1.4 (m, 4H), 1.4 –3.7 (m, 27H), 7.16 –7.31 (m, 4H); MS (FAB/glycerol) m/e 606 (M⫹⫹1); IR (Nujol) 2150 cm-1; HPLC tR ⫽ 21.45 min; M - H⫹ calcd for C28H38N5O8S 604.2441 found [HRFAB] m/e ⫽ 604.2448, error ⫽ ⫹1.2 ppm. 2.2. Conjugation of Herceptin™ with C-DOTA, PADOTA, 1B4M-DOTA, CHX-DOTA The Herceptin™ was generously provided by Dr. R. Altemus (Radiation Oncology Branch, NCI). The Herceptin™ (5 mg/mL) was conjugated with either C-DOTA, PA-DOTA, 1B4M-DOTA, or CHX-DOTA employing the linkage methods for aryl isothiocyanato groups that have been well described in the literature [22]. Unreacted or “free” ligand was separated from the conjugated antibody by dialysis in 0.15 M NH4OAc. The average number of chelates per antibody for the conjugation products were 1.4, 3.3, 3.8, and 2.4 for C-DOTA, PA-DOTA, 1B4M-DOTA, CHX-DOTA, respectively, as determined by the appropriate spectrometric method for these chelating agents [7]. Protein concentration was determined using the Lowry method with a standard of bovine serum albumin [18].

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

2.3. Cell lines The human colon carcinoma (LS-174T) cell line was grown in supplemented Eagle’s minimum essential medium as previously described [6,32]. The human gastric carcinoma cell line (N87) was kindly provided by Dr. Raya Mandler, Metabolism Branch, NCI. The N87 cell line, [26] which expresses high levels of HER2, was maintained in RPMI 1640 (Quality Biologicals, Gaithersburg, MD) medium supplemented with 10% fetal bovine serum (FBS; Gemini Bio-Products, Woodland, CA) and 0.01 mM nonessential amino acids (Quality Biologicals). The human melanoma (SK-Mel) cell line obtained from ATCC (Manassas, VA) was utilized as a HER2 negative control. 2.3.1. Radiolabeling and comparative radiolabeling of the C-DOTA, PA-DOTA, 1B4M-DOTA, CHX-DOTA Herceptin™ Immunoconjugates The 177Lu (1–3 mCi in 10 –20 ␮L 0.1 M HCl) (Missouri University Research Reactor (MURR), Columbia, Missouri) was added to ⬃200 ␮L 0.15 M NH4OAc buffer (pH 5.0-5.5) containing 300 – 400 ␮g of each of the DOTAHerceptin™ immunoconjugates. The reaction mixtures were incubated at 37°C for 1–2.5 hours. The reaction kinetics were followed by taking aliquots at different times and analyzed the components using ITLC developed in 10 mM EDTA/0.15 M NH4OAc. The reactions were halted by adding 5 ␮L of 0.1 M DTPA and the reaction yields were determined by the ITLC method described previously [19]. The 177Lu-DOTA-Herceptin™ conjugates were purified through a 10-DG desalting column (Bio-Rad, Hercules, CA) eluted using PBS and the antibody peaks were collected. Purity of the 177Lu radiolabeled DOTA-Herceptin™ radiolabeled immunoconjugates was determined using ITLC and/or size exclusion HPLC (SE-HPLC) Radio-iodination of Herceptin™ with Na 125I was performed as described using the Iodogen method [10]. The product was purified using a desalting column (PD-10; Amersham Biosciences, Piscataway, NJ).

587

ing was calculated for each dilution. The values presented are an average of the serial dilutions. To confirm the specific reactivity of the RIC, cells were incubated with ⬃200,000 cpm of the RIC along with an excess (10 ␮g) of unlabeled HerceptinTM. 2.5. In vitro stability determinations The purified 177Lu-DOTA-Herceptin™ conjugates (2 mL each, 600 –1200 ␮Ci) were mixed with 2 mL human serum (Gemini Bioproducts, Woodland, CA). The mixtures were maintained in a 5% CO2 incubator at 37°C. At different time points, 50 ␮L aliquots were taken, mixed with 5 ␮L of 0.1 M DTPA and incubated at 37°C for 30 min. The percentage of 177Lu associated with the immunoconjugate was analyzed by both ITLC and SE-HPLC. 2.6. Tumor model and in vivo studies The radioimmunoconjugates were compared in vivo using athymic mice bearing human colon adenocarcinoma xenografts. Female athymic mice (nu/nu), obtained from Charles River Laboratories (Wilmington, MA) at 4-6 weeks of age, were injected subcutaneously on the flank with 2 ⫻ 106 LS-174T cells in 0.2 mL of RPMI-1640. At approximately 10 –14 days, when the tumors measured between 0.4 – 0.6 cm in diameter, the mice received the 177Lu-labeled Herceptin™ and 125I-labeled HerceptinTM. The mice were injected with each RIC (⬃5 ␮Ci of each) intravenously (i.v.) via the tail vein. Mice (n ⫽ 5) were sacrificed by exsanguination at 24, 48, 72, and 168 hr. Blood, tumor and the major organs were collected, wet-weighed, and counted in a ␥-counter (Minaxi-␥; Packard, Downers Grove, IL). The percent injected dose per gram (%ID/g) was determined for each tissue as well as the radiolocalization indices (%ID/g in tumor divided by the %ID/g in the normal tissue). The mean and standard deviation for each tissue was also calculated. 2.7. Statistical analysis

2.4. Radioimmunoassay The immunoreactivity of the RICs were assessed in a live-cell radioimmunoassay (RIA) as detailed elsewhere [12]. HER2 positive cells (N87) and HER2 negative cells (SK-Mel) were harvested, pelleted at 1,000 ⫻ g (Allegra 6KR; Beckman Coulter, Palo Alto, CA), re-suspended in PBS (pH 7.2) containing 1% BSA and added to 12 ⫻ 75 mM polypropylene tubes (1 ⫻ 106 cells in 100 ␮L). Serial dilutions of the radiolabeled Herceptin™ preparations (⬃200,000 cpm -12,500 cpm in 50 ␮L) were then added in duplicate and gently shaken. Following an overnight incubation at 4°C, the cells were washed once with 4 mL of 1% BSA in PBS, pelleted at 1,000 ⫻ g for 5 min and the supernatant decanted. The pelleted cells were then counted in a ␥-scintillation counter (Packard) and the percent bind-

The data was analyzed by application of Students t-test using Sigmaplot version 4.01 (SPSS, San Raphael, CA).

3. Results Syntheses of the bifunctional DOTAs were envisioned as an expansion and convergence of previously reported syntheses of C-DOTA and two acyclic DTPA chelating agents [2,20,33]. In brief, the previously reported methodology [20] that provided a singly substituted bifunctional 12-membered ring could be exploited to introduce an array of alkyl substituents, limited to the availability of the appropriate diamines with which to form the triamine cyclization component. The other cyclization component could also be

588

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

Fig. 2. Synthesis of the Bifunctional Chelating Agent 1B4M-DOTA.

employed to introduce substituents to the ring through the use of substituted iminodiacetic acid derivatives. While compounds are readily available from amino acids, this portion of the molecule was held constant in this current course of study. Application of this methodology was expected to produce the desired substituted 12-membered rings, albeit in somewhat reduced yields due to the additional ring substituents providing steric hindrance to the ring closure reaction. This was expected to be significant in the case of the inclusion of the trans-cyclohexyl ring. However, some benefit was also predicted in this instance due to the geometric restriction of the ring potentially lowering some of the entropy barriers to the cyclization. Synthesis of the 1B4M-DOTA was initiated from the previously reported p-nitrophenylalanine 2-aminoproylamide. Careful analysis of this amide has shown that its synthesis by direct aminolysis of the methyl ester results in ⬃ 90 –95% of the desired regioisomer with ⬃ 5–10% being the 1-methylpropyl amide [3]. Despite efforts to separate these isomers, no success has been reported. Therefore, this mixture was carried forward into the cyclization reaction with plans to either devise an acceptable separation procedure after this step or then carry the synthesis through to completion as a mixture. This diamine was reacted with bis(succinimidyl) ester of BOC-iminodiacetic acid under relatively high-dilution conditions. Precise equimolar addition of the two components was performed via syringe pump, a convenient and accurate method for such reactions [5,20]. Isolation of the cyclization product was both simplified and then complicated by its extreme insolubility. Upon completion of the reaction and the subsequent removal of the DMF that had been introduced with the two reactants, a highly insoluble solid was obtained. This material not only possessed a mass spectrum

consistent for the product, but also appeared to be comprised of two products, one major and one faint minor spot by TLC with a very small ⌬Rf (0.03). Due to very low solubility of the crude product and the small ⌬Rf, purification procedures required performing chromatography on small separate portions after which the “pure” fractions were pooled and then re-chromatographed until the major product could be isolated. These efforts were also complicated by apparent increasing insolubility with increasing purity. However, only very small amounts of inadequately pure minor product could be isolated. Both components had identical mass spectra, however, comparative NMR spectra of the two products clearly indicated that they possessed very similar structures. On this basis, the minor product was tentatively assigned to the structure that would be generated by reaction of the minor regiochemical component of the amino acid amide starting material with the major product being as depicted in Fig. 2. The carbamate protecting group was cleaved with HCl in dioxane as it had been noted previously that use of other acids resulted in complicating the subsequent borane reduction, i.e. competing reduction of the aryl nitro group to an aniline. This effect was rarely observed to occur with HCl present, yet it was remained essential to perform this BH3 reduction with rigid temperature control to eliminate concurrent reduction of the aryl nitro group [33]. The cyclic polyamine was directly alkylated with excess tert-butyl bromoacetate and the tetra-ester was isolated by chromatography. The tert-butyl groups were cleaved to generate the tetra-acetic acid. The nitro group was hydrogenated producing the desired aniline that was then immediately treated with thiophosgene to generate the final product isothiocyanate in ⬃7.5% overall. One should note that while this isolated product is pure by regiochemistry, no resolution of the four stereochemical

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

589

Fig. 3. Synthesis of the Bifunctional Chelating Agent CHX-DOTA.

components was performed. As such, NMR spectra and HPLC results were frequently complicated by the presence of two diastereomeric pairs of enantiomers. This effect was evident in the analytical HPLC chromatography with peaks occasionally being “split” or there being a major and minor component. However, this effect was capricious and never adequately consistent to permit preparative separations. That this compound could be isolated in good purity by silica gel chromatography is evidence that this effect from stereochemical differences on the chromatography was minimal. Synthesis of the CHX-DOTA (Fig. 3) was initiated from the previously reported chiral diamine 9 [33]. The diamine was reacted with bis(succinimidyl) ester of BOC-iminodiacetic acid as described for 4 [5,20]. Isolation of this cyclization product was considerably simplified as compared to 4. After removal of the solvents that had been introduced with the two reactants, a chloroform solution of the crude product was obtained. Silica gel chromatography cleanly provided the product without undue complication. The carbamate protecting group was cleaved with HCl in dioxane as previously described and the amides were then reduced with BH3/THF. The polyamine was directly alkylated with excess tert-butyl bromoacetate. The tetra-ester was isolated by chromatography and afterwards, the tertbutyl groups were cleaved to generate tetra-acetic acid 13. Alternately, crude tetra-tert-butyl ester was deprotected with acid and the tetra-acetic acid isolated by executing a sequence of cation and anion ion-exchange chromatography protocols. While no clear advantage to either route seems evident, the ability to exercise options for future compounds that are not amenable to silica gel chromatography seems of value. The nitro group was reduced to produce aniline 14

that was then immediately treated with thiophosgene to generate the final isothiocyanate 2 [2]. The C-DOTA, PA-DOTA, 1B4M-DOTA, and CHXDOTA were each conjugated to Herceptin™ using established methodology [22]. The number of chelates to protein was determined using a spectrophotometric method to accurately measure the number of functional metal binding sites available on each conjugate [7]. The immunoreactivity results as assayed via a cell binding assay for each of the C-DOTA, PA-DOTA, 1B4M-DOTA, and CHX-DOTAconjugates is presented in Table 1. Radiolabeling with 177Lu of the four immunoconjugates was performed as previously reported [21]. Typical reaction yields were 97%, 99%, 98%, and 81% for C-DOTA, PADOTA, 1B4M-DOTA, and CHX-DOTA-Herceptin™ conjugates, respectively. The results obtained for the reaction kinetics is illustrated in Fig. 4. The reaction kinetics for the CHX-DOTA-Herceptin™ was rapid, however the yield was low. To achieve a similar reaction yield for all four immunoconjugates, the ranking for the rate of reaction was

Table 1 Radioimmunoassay data for studied radioimmunoconjugatesa Conjugate

HER2⫹

HER2⫺

Spec Activity (␮Ci/␮g)

C-DOTA PA-DOTA 1B4M-DOTA CHX-DOTA 125 I

50.8 43.1 47.2 45.8 58.5

7.1 8.8 6.7 4.3 11.2

7.2 16.5 13.7 14.8 5.7

a

Immunoreactivity of the 177Lu-labeled Herceptin™ conjugates was assessed using a cell binding radioimmunoassay as described in Materials and Methods.

590

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

Fig. 6. In Vitro Serum Stability (HPLC) of Radioimmunoconjugates. Fig. 4. Radiolabeling Formation Kinetics of Radioimmunoconjugates.

177

177

Lu DOTA-Herceptin™

Lu DOTA-Herceptin™

1B4B-DOTA⬎PA-DOTA⬎C-DOTA. The immunoconjugates were radiolabeled yielding the following specific activities of 7.1 ␮Ci/␮g, 16.4 ␮Ci/␮g, 5.4 ␮Ci/␮g, and 2.4 ␮Ci/␮g, for the 177Lu-C-DOTA, PA-DOTA, 1B4MDOTA, and CHX-DOTA-Herceptin™ conjugates after purification, respectively. An in vitro serum stability study was performed with all four of the above RIC with the measurements being determined by ITLC and SE-HPLC methods over four weeks. The results for these studies are presented in Figs. 5 and 6. The evaluation by ITLC indicated that there was a loss of radioactivity of less than 1% from the RICs after a 5 h incubation in serum. At 1 day there is an apparent difference among the RICs with 1.8, 0.2, 0.4, and 2.7% of the radioactivity lost from the 177Lu-C-DOTA, PA-DOTA, 1B4MDOTA, and CHX-DOTA-Herceptin™ conjugates, respectively. At the end of the 28 day study, he 177Lu-PA-DOTAHerceptin™ conjugate demonstrated superior stability among the four different DOTA immunoconjugates that

Fig. 5. In Vitro Serum Stability (ITLC) of Radioimmunoconjugates.

177

Lu DOTA-Herceptin™

were investigated with only 1.6% loss of radioactivity from the RIC. Both the 177Lu-C-DOTA and the 1B4M-DOTAHerceptin™ conjugates exhibited similar stability with a loss of 5.4 and 5.2% of the radioactivity. The 177Lu-CHXDOTA-Herceptin™ conjugate appeared to be less stable with 9.8% of the radioactivity no longer associated with the RIC. The HPLC analysis (Fig. 6) yielded similar results. Following a 5 h incubation in serum, the CHX-DOTA RIC already demonstrated some instability in that 1.1% of the radioactivity had dissociated from the RIC while less than 1% had been lost from the other three RICs. After 1 day of incubation, the amount of radioactivity released from the RICs was 1.3, 0.58, 2.4, and 4.2% for the 177Lu-C-DOTA, PA-DOTA, 1B4M-DOTA, and CHX-DOTA-Herceptin™ conjugates. After two weeks these values increased to 5.0, 3.4, 8.3, and 11.6% and by the end of the 28 days were up to 6.1, 7.0, 8.5, and 15.6%, respectively. The in vivo stability of the 177Lu-C-DOTA, PA-DOTA, 1B4M-DOTA, and CHX-DOTA-Herceptin™ conjugates (Table 2) were then compared. Athymic mice bearing human colon carcinoma (LS-174T) xenografts were injected via the tail with each of the RIC along with 125I-labeled HerceptinTM. The mice were then euthanized at 24, 48, 72, 96 and 168 h post-injection of the RIC. In studies designed to evaluate the in vivo stability and properties of a RIC, specific attention should be given to the normal tissue for which the radionuclide exhibits high affinity [25]. In the case of a lutetium-labeled RIC, the normal organ of interest is bone [8]. The femur %ID/g for each of the RIC is presented in Fig. 7 along with the %ID/g calculated for the 125 I-labeled HerceptinTM. The RIC comprised of the CHX-DOTA ligand resulted in the highest values of femur uptake throughout the study. At 24 h, the femur %ID/g was 3.52 ⫾ 0.67 which then peaked at 96 h at 4.04 ⫾ 0.88. In contrast, the RIC consisting of the PA-DOTA ligand yielded the lowest femur %ID/g with 1.98 ⫾ 0.17 and then declined to 0.97 ⫾ 0.17 and 1.04 ⫾ 0.16 at 96 and 168 h, respectively. The 177LuC-DOTA and 1B4M-DOTA Herceptin™ conjugates were

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595 Table 2 Biodistribution of Herceptin™ radiolabeled with Conjugate

177

Lu using bifunctional chelates after intravenous injection: Percent Injected dose/grama

Tissue 24

C-DOTA

PA-DOTA

1B4M-DOTA

CHX-DOTA

125

I

Blood Tumor Liver Spleen Kidney Lung Blood Tumor Liver Spleen Kidney Lung Blood Tumor Liver Spleen Kidney Lung Blood Tumor Liver Spleen Kidney Lung Blood Tumor Liver Spleen Kidney Lung

b

591

c

17.22 (1.25) 17.08 (3.36) 6.63 (1.24) 5.93 (0.68) 5.20 (0.56) 6.26 (0.76) 13.26 (2.98) 17.69 (6.23) 7.66 (0.76) 7.24 (1.32) 4.38 (0.45) 5.28 (1.14) 13.88 (1.71) 17.48 (5.09) 5.09 (1.11) 4.05 (0.62) 3.88 (0.29) 5.58 (0.72) 9.40 (2.11) 16.32 (4.94) 7.80 (0.81) 5.67 (2.21) 4.87 (0.67) 4.70 (1.22) 12.21 (1.62) 12.60 (3.85) 2.95 (0.46) 3.21 (0.64) 2.73 (0.35) 5.39 (0.90)

48

Time (hr) 72

96

168

13.31 (2.75) 36.31 (12.32) 8.18 (1.28) 6.04 (0.66) 5.13 (0.95) 4.42 (1.92) 9.22 (2.60) 23.85 (7.73) 7.92 (1.17) 7.74 (1.60) 3.59 (0.70) 3.46 (0.82) 13.87 (1.17) 42.46 (12.35) 4.48 (0.50) 4.19 (0.48) 3.94 (0.45) 4.95 (0.47) 9.60 (0.83) 26.78 (3.25) 5.48 (0.53) 4.50 (0.56) 4.23 (0.88) 3.49 (0.33) 11.28 (1.18) 19.25 (4.44) 2.58 (0.38) 2.76 (0.35) 2.48 (0.47) 4.14 (0.75)

11.62 (1.91) 37.38 (14.62) 5.86 (1.01) 5.46 (1.37) 5.02 (0.77) 4.58 (0.72) 7.70 (1.97) 32.30 (10.41) 7.40 (0.91) 5.50 (1.22) 3.43 (0.7) 3.23 (0.73) 11.65 (1.44) 39.22 (6.50) 5.34 (1.09) 5.07 (1.62) 4.35 (0.70) 5.32 (1.09) 7.58 (2.01) 24.23 (11.96) 6.60 (1.07) 4.32 (0.69) 3.56 (0.36) 3.72 (1.08) 9.55 (1.67) 17.33 (5.90) 2.39 (0.41) 2.34 (0.63) 2.21 (0.39) 4.05 (0.79)

11.23 (1.19) 25.95 (6.76) 4.84 (0.33) 4.59 (0.57) 3.68 (0.42) 4.98 (0.69) 4.88 (2.00) 20.77 (4.11) 6.87 (2.28) 4.47 (0.52) 2.91 (0.94) 2.27 (0.86) 10.40 (1.41) 27.06 (5.12) 4.14 (0.98) 4.43 (0.64) 3.11 (0.31) 4.21 (0.50) 6.84 (0.92) 21.28 (5.51) 5.71 (0.80) 4.03 (0.37) 3.22 (0.62) 3.12 (0.39) 8.28 (1.14) 12.52 (3.37) 1.87 (0.28) 1.91 (0.30) 1.61 (0.26) 3.58 (0.47)

6.12 (2.17) 24.64 (5.89) 4.82 (0.75) 4.72 (2.13) 2.63 (0.42) 2.86 (0.78) 3.54 (0.90) 18.15 (7.14) 5.80 (0.62) 4.47 (0.79) 1.89 (0.20) 1.62 (0.27) 5.12 (2.88) 22.28 (9.81) 4.79 (0.90) 3.33 (0.94) 2.40 (0.58) 2.65 (1.09) 3.82 (1.51) 18.14 (9.22) 6.65 (3.05) 3.68 (0.80) 2.79 (0.37) 2.65 (0.83) 5.44 (1.80) 9.11 (3.87) 1.40 (0.43) 1.31 (0.48) 1.19 (0.33) 2.42 (0.70)

Athymic mice bearing s.c. human colon carcinoma (LS-174T) xenografts were co-injected i.v. with approximately 2-5 ␮Ci of 177Lu-labeled immunoconjugates and 125I-Herceptin. The mice (n ⫽ 5) were sacrificed by exsanguination as described in Materials and Methods. The blood, tumor and major organs were collected, wet-weighed and the radioactivity measured. The %ID/g was calculated for each tissue as well as the average deviation. b The values represent the average %ID/g (percent injected dose/gram). c Values in parentheses are the standard deviation of the %ID/g. a

not appreciably different from each other and both were intermediate in the femur %ID/g as compared to the CHXand PA-DOTA RIC. Differences among the 177Lu-labeled RICs were also evident in the %ID/g calculated for the tumor xenografts and the other normal tissues that were collected (Table 2). The greatest uptake in tumor by a RIC was observed with the 177Lu-1B4M-DOTA Herceptin. At 48 and 72 h the tumor %ID/g was 42.46 ⫾ 12.35 and 39.22 ⫾ 6.50, respectively. The lowest values, 26.78 ⫾3.25 and 24.23 ⫾ 11.96, were obtained with the CHX-DOTA RIC at the same time points. Among the normal tissues, the liver presented with the highest %ID/g for each of the RIC. The CHX-DOTA had the highest initial value (7.80 ⫾ 0.81) at 24 h which declined to 5.48 ⫾ 0.53 at 48 h and then fluctuated, ending with the highest value at 168 h (6.65 ⫾ 3.05) of the RICs. The conjugate containing the PA-DOTA had an initial value of 7.66 ⫾ 1.17 at 24 h which peaked at 48 h (7.92 ⫾ 1.17) and then declined to 5.80 ⫾ 0.62 by 168 h. The

C-DOTA RIC had a similar profile in which the liver %ID/g peaked at 48 h (8.18 ⫾ 1.28) and decreased to 4.82 ⫾ 0.75 at 168 h. The conjugate containing the 1B4MDOTA RIC resulted in the lowest liver %ID/g values throughout the entire study period. In other normal tissues, the C-DOTA RIC resulted in the highest kidney %ID/g at each of the study time points with the exception of 168 h, in which a value of 2.79 ⫾ 0.37 was obtained with the CHX-DOTA RIC. The spleen %ID/g was the greatest with the PA-DOTA RIC from 24-72 hr while higher values were determined at 96 and 168 h with the C-DOTA RIC. The lowest spleen %ID/g values were obtained with 177Lu-1B4M-DOTA-Herceptin at 24, 48 and 168 h and with the CHX-DOTA RIC at 72 and 96 h. Tissue-to-blood ratios were also calculated to more accurately gauge and compare the in vivo stability of each of the 177Lu-labeled immunoconjugates. If radioactivity in the tissues were due to specific accumulation, then the tissueto-blood ratio would be expected to increase as a function of time. If the radioactivity in the tissues was a result of what

592

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

Fig. 7. Comparison of the Uptake in Bone of Herceptin™ Radiolabeled with

was present in the plasma compartment, then the tissue-toblood ratio would be expected to remain constant with time. The femur-to-blood ratios are relatively constant for the C-DOTA and PA-DOTA conjugates. The 1B4M-DOTA RIC yielded similar results up to the 168 h time point at which time there was a slight increase over the 96 h time point (0.18 to 0.52). The femur-to-blood ratios for the CHX-DOTA RIC were higher overall than the other RIC and also exhibited an increase from 0.60 at 96 h to 1.09 at 168 h. Differences in the liver-to-blood and spleen-to-blood ratios were also evident. The C-DOTA, 1B4M-DOTA and CHX-DOTA conjugates each showed an increase in the liver-to-blood ratios with CHX-DOTA demonstrating the greatest increase that did not occur until 168 h. In contrast, the PA-DOTA exhibited a steady increase in the liver-toblood ratio throughout the study period with an initial ratio of 0.62 at 24 h; peaking at 96 h with 1.91 and declined slightly to 1.77 by 168 h. Differences were also found between the PA-DOTA and the other three RIC in the spleen-to-blood ratios. Again, in this instance, the ratios obtained for the C-DOTA, 1B4MDOTA and CHX-DOTA were relatively stable and less than 1 from 24 to 96 h. It was not until the 168 h time point that an increase was observed in the ratios. In contrast to this pattern, the RIC constructed with the PA-DOTA ligand began an increase in the spleen-to-blood ratio at 96 h.

4. Discussion Synthesis of a specific stereoisomer of CHX-DOTA was based on the hypothesis that inclusion of a trans-cyclohexyl ring into the backbone structure of a bifunctional DOTA would not only add additional rigidity to the ring and also

177

Lu.

enhance the pre-organization geometry of the metal binding orbitals of the tertiary amines. A second aspect of this hypothesis was that the overall ring conformation could be influenced or controlled by the stereochemistry of the pendant substituents. Preferred geometries of metal complexes formed by DOTA and its derivatives have been extensively studied and much attention has been given to the isomeric and intra-conversional nature of the carboxylate or carboxamide arms [9,13]. The studies of Forsberg et.al. in this regard also described the geometries of the backbone ethylene linkages between the amines and the interchange between the ␭ and ␴ conformations [9]. Incorporating this information into a scheme of setting or locking the ring ethylenediamines into one of these conformations selectively, we proposed that this modification would not only enhance the pre-organization aspect of DOTA, but also improve complex formation kinetics by eliminating energy costs associated with both conformational interchange and those associated with actual complex formation. The initial portion of this investigation as stated earlier was to insert this conformational control via stereoselectivity only in part at this stage employing a previously reported stereoisomer for construction of this ring. This isomer then reflects the ␭ conformer stereochemically in three of four carbons of the 12-membered ring as depicted in the Newman projections in Fig. 8. One would expect that use of the enantiomer of this isomer would be equally acceptable. Monoclonal antibody Herceptin™ was chosen to evaluate the relative 177Lu radiolabeling kinetics and yields of the four DOTA derivatives as their respective immunoconjugates as well as to perform comparative biodistribution studies. This antibody was chosen for these studies due to our prior use and experience and the relative robustness of this protein towards conjugation [21]. However, conjugation of bifunctional ligands at excessive levels has been

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

593

Fig. 8. Newman Projections of Chelate Ring Conformation: (a) ␭ conformation of bridging ethylenediamine; (b) ␭ conformation of bridging ethylenediamine ring segment with p-nitrobenzyl substituent with the stereochemical configuration in CHX-DOTA; (c) ␭ conformation of bridging ethylenediamine ring segment with the trans-cyclohexyl ring locking substituent with the stereochemical configuration in CHX-DOTA.

shown to decrease binding and immunoreactivity of a mAb recommending that some care still be exercised to maintain the binding activity [1]. Use of 177Lu was chosen as the radionuclide for evaluating these immunoconugates specifically as we had just completed a study that employed not only this radionuclide, but also C-DOTA and PA-DOTA [21]. Furthermore, the availability, half-life (T1/2 ⫽ 6.71 d), and associated ␥-emission characteristics of 177Lu, [30] permit an accurate and extended biodistribution study to be performed to detect early, as well as late differences among the radioimmunoconjugates. Radiolabeling of the four immunoconjugates was accomplished in NH4OAc buffer at pH 5.0 –5.5 at 37°C for a half hour or more as required. From the radiolabeling kinetic study, ranking the initial formation rate appears to provide a relative formation rate where CHX-DOTA>1B4MDOTA>PA-DOTA>C-DOTA. However, the radiolabeling reaction of CHX-DOTA with 177Lu was not in high yield (Fig. 4) and the complex was not as stable as indicated by both the in vitro stability study in serum (Figs. 5, 6) and biodistribution study in vivo (vide infra). Although the reaction yield and stability of the 177Lu 1B4M-DOTA was reasonably moderate when forming just the simple nonconjugated complex (data not shown), once the 1B4MDOTA was conjugated onto the mAb, the 177Lu complex formation rate was greater than observed for both the PADOTA and C-DOTA and its in vitro stability was similar to C-DOTA. Therefore, it seems that inclusion of this specific trans-cyclohexyl moiety into the DOTA ring framework altered the DOTA and resulted in some instability of the complex. However, addition of the methyl group in the 1B4M-DOTA appeared to increase the rate of formation and provided complex stability similar to, if not marginally greater than that observed for C-DOTA and PA-DOTA. In vivo experiments were performed with 177Lu C-DOTA, PA-DOTA, 1B4M-DOTA, and CHX-DOTA conjugated to Herceptin™ in athymic mice (n⫽5) bearing

s.c. human colon carcinoma (LS-174T) xenografts. The mice were injected i.v., sacrificed at selected time points and the %ID/g as well as the tissue-to-blood ratios were calculcated. Each of the radioimmunoconjugates retained immunoreactivity (Table 1) and demonstrated excellent tumor targeting (Table 2). The Herceptin™ conjugated with the 1B4M-DOTA ligand resulted in the highest tumor %ID/g with 42.46 ⫾ 12.35. Ranking the RICs by tumor %ID/g provides 1B4M-DOTA > C-DOTA > PA-DOTA > CHX-DOTA, all of which demonstrated greater tumor targeting than the 125I-labeled Herceptin (Table 2). This ranking of the C-DOTA and PA-DOTA ligands is consistent with previously published data albeit the RIC consisted of a genetically engineered mAb in which the CH2 domain was deleted [21]. Appropriate evaluation of the in vivo stability of a RIC should include a careful examination of the data obtained from the femurs since bone is a primary organ of accretion of lanthanides [8]. The RIC containing the CHX-DOTA resulted in the highest femur %ID/g throughout the study period while the PA-DOTA yielded the lowest values. For example, the femur %ID/g at 24 h was 3.52 ⫾ 0.67 and 1.98 ⫾ 0.17 for the CHX-DOTA and PA-DOTA respectively. Ranking of RICs in vivo stability according to the accretion, or lack thereof, of radioactivity in the femur results in PA-DOTA > 1B4M > C-DOTA > CHX-DOTA. In other normal tissues, such as the liver, this ranking would be the same as determined by tumor uptake, with 1B4M-DOTA exhibiting the lowest accretion of radioactivity throughout the study, while the CHX-DOTA RIC resulted in the highest accumulation. To better ascertain the retention of the 177Lu by the four radioimmunoconjugates, tissue-to-blood ratios were calculated. An increase in the normal tissue-to-blood ratio is an indicator of the non-specific accumulation of radioactivity in the tissue and hence the release of the radioactivity from the ligand and/or the metabolism of the RIC. By these criteria, the PA-DOTA RIC appears to be the least stable;

594

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595

the liver- and the spleen-to-blood ratios steadily increase throughout the study. The 1B4M-DOTA RIC overall has the lowest ratios for all of the organs than the other RIC with the exception of the liver at 168 h, while the CHXDOTA RIC resulted in ratios that were in general higher. The in vivo results with the PA-DOTA RIC appear to contradict the data that was obtained in the in vitro serum stability and provide an argument for inclusion of in vivo studies to appropriately and accurately analyze a RIC. 5. Conclusion Two novel backbone substituted DOTA derivatives, 1B4M-DOTA and CHX-DOTA, were synthesized and characterized. The respective radiolabeling formation rates of these two chelating agents were comparatively evaluated both in vitro and in vivo with PA-DOTA and C-DOTA as immunoconjugates with Herceptin™ with 177Lu. The results indicate that the 1B4M-DOTA possessed equivalent stability characteristics with an increase in complex formation rate as compared to the PA-DOTA and C-DOTA despite this chelating agent being comprised of a mixture of isomers. The CHX-DOTA, which was prepared as a specific isomer to add both additional rigidity and macrocycle pre-organization exhibited accelerated formation radiolabeling rates, yet also demonstrated the least stability of the four evaluated immunoconjugates. As such, in this preliminary investigation into the effects of macrocyclic ring substituents on both formation rates and stability, resolution of the impact of the non-stereospecifically added methyl group, 1B4M-DOTA, versus the stereospecific setting of three chiral centers in the CHX-DOTA seems to be in conflict to the fundamental premise that the fluxional ring geometry assumes a preferred geometry for complex formation. Additional studies to systematically investigate the chemistry of ring substituents and the effects of their chirality will be necessary to gain an understanding of how one might be able to control or optimize radio-metal complex formation of DOTA derivatives. Acknowledgments We would like to thank Dr. Bunsen Honeydew for his salient insights, discussions, and a thorough review of this manuscript. References [1] Brechbiel MW, Gansow OA, Atcher RW, Schlom J, Esteban J, Simpson DE, Colcher D, Synthesis of 1-(p-Isothiocyanatobenzyl) Derivatives of DTPA and EDTA. Antibody Labeling and TumorImaging Studies, Inorg Chem 1986;1986:2772–2781. [2] Brechbiel MW, Gansow OA, Backbone Substituted DTPA Ligands for 90Y Radioimmunotherapy, Bioconjugate Chem 1991;1991:187– 194.

[3] Brechbiel MW, Beitzel PM, Gansow OA, Purification of p-Nitrobenzyl C-Functionalized Diethylenetriamine Pentaacetic Acids for Clinical Applications using Anion-Exchange Chromatography, J Chromatog A 1997;1997:63– 69. [4] Chappell LL, Rogers BE, Khazaeli MB, Mayo MS, Buchsbaum DJ, Brechbiel MW, Improved synthesis of the bifunctional chelating agent 1,4,7,10-tetraaza-N-(1-carboxy-3-(4-nitrophenyl)propyl)-N⬘,N⬙,N⵮tris(acetic acid)cyclododecane (PA-DOTA), Bioorganic Med Chem 1999;1999:2313–2320. [5] Chappell LL, Deal KA, Dadachova E, Brechbiel MW, Synthesis, Conjugation and Radiolabeling of a Novel Bifunctional Chelating agent for 225Ac Radioimmunotherapy Applications, Bioconjugate Chem 2000;2000:510 –519. [6] Colcher D, Keenan AM, Larson SM, Schlom J, Prolonged binding of a radiolabeled monoclonal antibody (B72.3) used for the in situ radioimmunodetection of human colon carcinoma xenografts, Cancer Res 1984;1984:5744 –5751. [7] Dadachova E, Chappell LL, Brechbiel MW, Spectrophotometric method for determination of bifunctional macrocyclic ligands in macrocyclic ligand-protein conjugates, Nucl Med Biol 1999;1999:977– 982. [8] Durbin PW, Metabolic characteristics within a chemical family, Health Phys 1960;1960:225–238. [9] Forsberg JH, Delaney RM, Zhao Q, Harakas G, Chandran R, Analyzing Lanthanide_Induces Shifts in the NMR Spectra of Lanthanide(III) Complexes Deerives from 1,4,7,10-Tetrakis(N,N-diethylacetamido)-1,4,7,10-tetraazacyclododecane, Inorg Chem 1995;1995: 3705–3715. [10] Fraker PJ, Speck JC, Protein and cell membrane iodinations with a sparingly soluble chloramide 1,3,4,6-tetrachloro-3a, 6a-diphenyl-glycouril, Biochem Biophys Res Commun 1978;1978:849 – 857. [11] Gansow OA, Newer approaches to the radiolabeling of monoclonal antibodies by the use of metal chelates, Nucl Med Biol 1991;1991: 369 –381. [12] Garmestani K, Milenic DE, Plascjak PS, Brechbiel MW, A New and Convenient Method for Purification of 86Y using a Sr(II) Selective Resin and Comparison of Biodistribution of 86Y and 111In labeled Herceptin™, Nucl Med Biol 2002;2002:599 – 606. [13] Howard JAK, Kenwright AM, Moloney JM, Parker D, Port M, Navet M, Rousseau O, Woods M, Structure and dynamics of all of the stereoisomers of europium complexes of tetra(carboxyethyl) derivatives of dota: ring inversion is decoupled from cooperative arm rotation in the RRRR and RRRS isomers, Chem Commun (1998) 1381–1382. [14] Keire DA, Jang YH, Li L, Dasgupta S, Godard WA, Shively JE, Chelators for Radioimmunotherapy: I. NMR and ab initio calculation Studies on 1,4,7,10-Tetra(carboxyethyl)-1,4,7,10-tetraazacyclododecane (DO4Pr) and 1,4,7-Tris(carboxymethyl)-10-(carboxyethyl)1,4,7,10-tetraazacyclododecane (DO3A1Pr), Inorg Chem 2001;2001: 4310 – 4318. [15] Krasner C, Joyce RM, Zevalin™: 90Yttrium Labeled Anti-CD20 (Ibritumomab Rituxan), a New Treatment for Non-Hodgkin’s Lymphoma, Current Pharm Biotech 2001;2001:341–349. [16] Lewis MR, Raubitschek A, Shively JE, Facile A, Water-Soluble Method for Modification of Proteins with DOTA. Use of Elevated Temperature and Optimized pH To Achieve High Specific Activity and High Chelate Stability in Radiolabeled Immunoconjugates, Bioconjugate Chem 1994;1994:565–576. [17] Lewis MR, Kao JY, Anderson AL, Shively JE, Raubitschek A, An improved method for conjugating monoclonal antibodies with Nhydroxysulfosuccinimidyl DOTA, Bioconjugate Chem 2001;2001: 320 –324. [18] Lowry OH, Rosebrough NJ, Farr AL, J.Randall A, Protein measurement with the folin phenol reagent, J Biol Chem 1951;1951:265–275. [19] Ma D, Lu F, Overstreet T, Milenic D, Brechbiel MW, Novel Chelating Agents for Potential Clinical Applications of Copper, Nucl Med Biol 2002;2002:91–105.

L.L. Chappell et al. / Nuclear Medicine and Biology 30 (2003) 581–595 [20] McMurry TJ, Brechbiel MW, Kumar K, Gansow OA, Convenient Synthesis of Bifunctional Tetraaza Macrocycles, Bioconjugate Chem 1992;1992:108 –117. [21] Milenic DE, Garmestani K, Chappell LL, Dadachova E, Yordanov A, Ma D, Schlom J, Brechbiel MW, In Vivo Comparison of Macrocyclic and Acyclic Ligands for Radiolabeling of Monoclonal Antibodies with 177Lu for Radioimmunotherapeutic Applications, Nucl Med Biol 2002;2002:431– 442. [22] Mirzadeh S, Brechbiel MW, Atcher R, Gansow OA, Radiometal Labeling of Immunoproteins: Covalent Linkage of 2-(4-Isothiocyanatobenzyl) diethylenetriaminepentaacetic Acid Ligand to Immunoglobulin, Bioconjugate Chem 1990;1990:59 – 65. [23] Moi MK, Meares CF, DeNardo SJ, The Peptide Way to Macrocyclic Bifunctional Chelating Agents: Synthesis of 2-(p-Nitrobenzyl)1,4,7,10-tetraazacyclododecane-N,N⬘,N⬙,N⵮,-tetraacetic Acid and Study of Its Yttrium(III) Complex, J Am Chem Soc 1988;1988: 6266 – 6267. [24] O’Donoghue JA, Bardies M, Wheldon TE, Relationships between Tumor Size and Curability for Uniformly Targeted Therapy with Beta-Emitting Radionuclides, J Nucl Med 1995;1995:1902–1909. [25] Packard AB, Kronauge JF, Brechbiel MW, Metalloradiopharmaceuticals, in M.J. Clarke, P.J. Sadler (Eds), Metalloradiopharmaceuticals II, Diagnosis and Therapy, Springer-Verlag: NY, 2000, pp. 45–116 . [26] Park JG, Frucht H, LaRocca RV, Bliss DP, Kurita Y, Chen TR, Henslee JG, Trepel JB, Jensen RT, Johnson BE, Characteristics of cell lines established from human gastric carcinoma, Cancer Res 1990;1990:2773–2780.

595

[27] Parker D, Tumour targeting with radiolabelled macrocycle-antibody conjugates, Chem Soc Rev 1990;1990:271–291. [28] Peterson JJ, Pak RH, Meares CF, Total Solid-Phase Synthesis of 1,4,7,10-Tetraazacyclododecane-N,N⬘,N⬙,N⵮-tetraacetic Acid-Functionalized Peptides for Radioimmunotherapy, Bioconjugate Chem 1999;1999:316 –320. [29] Ruegg CL, Anderson-Berg WT, Brechbiel MW, Mirzadeh S, Gansow OA, Strand M, (1990) Improved in vivo stability and tumor targeting of bismuth-labeled antibody, Cancer Res 1990;1990:4221– 4226. [30] Srivastava SC, Criteria for the selection of radionuclides for targeting nuclear antigens for cancer radioimmunotherapy, Cancer Biother Radiopharm 1996;1996:43–50. [31] Takenouchi M, Tabe K, Watanabe A, Hazato Y, Kato M, Shionoya T, Koike T, Kimura E, Novel Pendant-Type Macrocyclic Bifunctional Chelating Agents; (Carboxymethy) amino Derivatives of 2 - (4 - Nitrobenzyl) - 1,4,7,10 - tetraazacyclo - dodecane - N,N⬘,N⬙,N⵮tetraacetic Acid and Their Complex Formation with Yttrium (III), J Org Chem 1993;1993:6895– 6899. [32] Tom RH, Rutzky LH, Jakstys MH, Human colonic adenocarcinoma cells. Establishment and description of a new cell line, In Vitro 1976;1976:180 –191. [33] Wu C, Kobayashi H, Sun B, Yoo TM, Paik CH, Gansow OA, Carrasquillo JA, Pastan I, Brechbiel MW, Stereochemical Influence on the Stability of Radio-Metal Complexes In Vivo. Synthesis and Evaluation of the Four Stereoisomers of 2-(p-nitrobenzyl)-trans-CyDTPA, Bioorg Med Chem 1997;1997:1925–1934.