Synthesis and nca-Radioiodination of Arylstannyl− Cobalamin

Synthesis and nca-Radioiodination of Arylstannyl−Cobalamin Conjugates. Evaluation of Aryliodo−Cobalamin Conjugate Binding to Transcobalamin II and...
0 downloads 0 Views 330KB Size
Download PDF
Bioconjugate Chem. 1996, 7, 461−474

461

Synthesis and nca-Radioiodination of Arylstannyl-Cobalamin Conjugates. Evaluation of Aryliodo-Cobalamin Conjugate Binding to Transcobalamin II and Biodistribution in Mice D. Scott Wilbur,*,† Donald K. Hamlin,† Pradip M. Pathare,† Shannon Heusser,† Robert L. Vessella,‡ Kent R. Buhler,‡ James E. Stray,‡ Janna Daniel,‡ Edward V. Quadros,§ Patricia McLoughlin,§ and A. Charles Morgan†,| Departments of Radiation Oncology and Urology, University of Washington, Seattle, Washington 98195, VA Medical Center and the SUNY Health Science Center, Brooklyn, New York 11209, and Receptagen Corporation, Edmonds, Washington 98020. Received January 18, 1996X

A new method of preparing radiolabeled cobalamin derivatives has been developed. The method involves the use of cobalamin-tri-n-butylstannyl hippurate conjugates as intermediates to obtain radioiodinated cobalamin-iodohippurate conjugates. The arylstannyl functionality was used as an exchangeable group to obtain high specific activity radioiodinations and to circumvent some deleterious side reactions common to cobalamins under electrophilic iodination conditions. The first step in the synthesis of tri-n-butylstannyl hippurate conjugates was to obtain free carboxylate groups on the cobalamin moiety. This was accomplished by mild acid hydrolysis of the b-, d-, or e-propionamide side chains on the corrin ring, followed by careful separation of the isomeric products. The second step was to couple a linking molecule (diaminododecane) to the carboxylate. The final step was to conjugate p-tri-n-butylstannyl hippurate to the cobalamin-diaminododecane adduct. All three isomeric cobalamin-p-tri-n-butylstannyl hippurate conjugates were prepared, as were the corresponding cobalamin-p-iodohippurate conjugates (HPLC standards). Radioiodination reactions were conducted with N-chlorosuccinimide and Na[*I]I in MeOH using conditions previously developed for arylstannylations. However, unlike the previous reactions, a key factor in obtaining the desired radioiodinated cobalamins was that the reaction be conducted under neutral conditions. Isolated yields of 40-65% were obtained for all three cobalamin isomers. Specific activities of 10-33% theoretical were obtained for the radioiodinated cobalamins. Evaluation of competitive binding of (nonradioactive) cobalaminiodohippurate conjugates with recombinant human transcobalamin II showed that the e-isomer bound nearly as well as [57Co]cyanocobalamin (50%), whereas the b-isomer had decreased binding (6%) and the d-isomer was significantly decreased in its binding (0.7%). Two biodistributions of the radioiodinated e-isomer were conducted in athymic mice. One biodistribution investigated tissue localization in mice bearing a renal cell carcinoma xenograft, and the other biodistribution investigated tissue localization when the radioiodinated cyanocobalamin was mixed with 1% BSA prior to injection. A comparison of the results of the two biodistributions and a discussion of how they relate to previous [57/60Co]cyanocobalamin biodistributions are provided.

INTRODUCTION

Cobalamin conjugates are being synthesized for evaluation as cellular growth blockers for the treatment of AIDS1 -related lymphoma. This therapeutic approach is * Address correspondence to this author at the Department of Radiation Oncology, University of Washington, 2121 N. 35th St., Seattle, WA 98103-9103 [telephone (206) 685-3085; fax (206) 685-9630; e-mail [email protected]]. † Department of Radiation Oncology. ‡ Department of Urology. § VA Medical Center and SUNY Health Science Center. | Receptagen Corp. X Abstract published in Advance ACS Abstracts, June 1, 1996. 1 Abreviations: AIDS, acquired immunodeficiency syndrome; BSA, bovine serum albumin; DCC, dicyclohexylcarbodiimide; DMF, dimethylformamide; DMSO, dimethyl sulfoxide; EDC, 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide; 2HEDS, 2-hydroxyethl sulfide; HOAc, acetic acid; HSA, human serum albumin; IF, intrinsic factor; LC, liquid chromatography; nca, no-carrier-added; NaOAc, sodium acetate; 3NBA, 3-nitrobenzyl alcohol; NCS, N-chlorosuccinimide; NHS, N-hydroxysuccinimide; PBS, phosphate-buffered saline; RCC, renal cell carcinoma; TCII, transcobalamin II; rhTCII, recombinant human transcobalamin II; TFA, trifluoroacetic acid; TFP, tetrafluorophenyl; TFP-OH, tetrafluorophenol.

S1043-1802(96)00033-X CCC: $12.00

based on the concept that cobalamin derivatives can be used to block entry of cyanocobalamin, 1 (vitamin B12), and related cobalamins into cells, thus depleting cells of this essential vitamin. Previous studies using the general anesthetic nitrous oxide (N2O) have shown that inactivation of cobalamin in rapidly proliferating cells, such as leukemias and lymphomas, can result in a therapeutic effect in vitro (1, 2) and in patients (3, 4). Depletion of cobalamin in cells is expected to provide results similar to those found when cobalamin was inactivated with N2O. While oxidation of cobalamin with N2O does not require specific biochemical/cellular mechanisms to obtain its effect, development of cobalamin derivatives that are designed to alter the entry of cobalamin into cells will require certain constraints. For instance, to be effective, cobalamin derivatives must bind avidly with transcobalamin II, the cobalamin-binding protein in plasma (5). This is due to the fact that at physiological concentrations cyanocobalamin does not enter cells by itself, nor does it bind with the cell surface receptor molecule (6). Further, design of optimal cobalamin conjugates will require that the cobalamin derivatives be evaluated with regard to interaction of the cobalamin/TCII complex with the cell-surface receptor and with regard to their in vivo pharmacokinetics. © 1996 American Chemical Society

462 Bioconjugate Chem., Vol. 7, No. 4, 1996

Wilbur et al.

Scheme 1. c-Lactone Formation in the Presence of Halogenation Reagents and Acid

To assist in the development of therapeutic cobalamin conjugates, a method for radiolabeling the synthesized cobalamin derivatives was sought. Early investigators demonstrated that cyanocobalamin could be labeled with cobalt radionuclides (Co-57, -58, -60) (7-14) and carbon14 (15). Tritium has also been used to radiolabel cyanocobalamin (16). While these labeling methods are adequate for obtaining information on cyanocobalamin, they were unattractive for our applications due to the potential difficulties with multistep derivatization of radioactive molecules and (in the case of C-14) the low specific activity that would be obtained. Also, the cost of radiolabeled starting materials in a multistep synthesis was considered a problem. Therefore, we chose to develop a new method of radiolabeling cyanocobalamin derivatives. An important consideration in developing a radiolabeling method is the need to incorporate the radionuclide in the last step of the synthesis. Additionally, we were interested in utilizing radionuclides of iodine, as they are readily available and relatively inexpensive. A further attraction of radioiodination was the fact that dual label experiments (e.g. I-125, -131) could be performed. Previous studies have shown that direct iodination of cyanocobalamin under neutral or acidic conditions results in an unstable adduct (17), and reaction with halogens or N-haloamides under acidic conditions results in formation of the c-lactone (Scheme 1) or c-lactam (18-21). Thus, we sought to develop a method for high specific activity radioiodination of cobalamin derivatives which would be facile and could be accomplished in a manner that did not alter the cobalamin portion of the conjugate. The approach chosen to prepare radioiodinated cobalamin derivatives was to use arylstannane derivatives as intermediates. These intermediates have been shown to provide rapid incorporation of no-carrier-added (nca) radioiodine into a large variety of compounds under mild reaction conditions (22, 23). The radioiodination of arylstannanes proceeds through an electrophilic substitution reaction, resulting in high radiochemical yields of compounds with high specific activity. On the basis of previous studies with cobalamin derivatives, we chose to conjugate the requisite arylstannane to the cobalamin moiety through a diaminododecane linking molecule attached to the b-, d-, and e-propionamide side chains of the corrin ring (24-27). Those studies had shown that conjugation of the diaminododecane with the three monocarboxylate derivatives, formed from mild acid hydrolysis of the corrin ring propionamide side groups, provided

cobalamin derivatives with various binding affinities for the cobalamin binding protein transcobalamin II (TCII) (28). Our approach to prepare the arylstannane derivatives involved three chemical steps: (1) hydrolysis of the corrin ring b-, d-, and e-propionamide side chains; (2) conjugation of the separate monocarboxylates with diaminododecane; and (3) conjugation of each cobalamin-diaminododecane adduct with an activated arylstannane. Once synthesized, radioiodination of the cobalamin-arylstannane conjugates was conducted by modification of conditions previously used for radioiodination of other arylstannane-containing molecules. Details of the synthesis and radioiodination of three isomeric stannylhippurylcobalamin conjugates, evaluation of TCII binding with three isomeric iodohippuryl-cobalamin conjugates, and two biodistributions of one radioiodinated isomer in athymic mice are reported herein. EXPERIMENTAL PROCEDURES

General. All chemicals purchased from commercial sources were of analytical grade or better and were used without further purification. Cyanocobalamin (vitamin B12) was obtained from Sigma Chemical Co. (St. Louis, MO). N-Hydroxysuccinimide was purchased from Lancaster Synthesis Inc. (Windham, NH). All other reagents were obtained from Aldrich Chemical Co. (Milwaukee, WI), except where noted. Solvents for HPLC analysis were obtained as HPLC grade and were filtered (0.2 µm) prior to use. Ion exchange chromatography was conducted with 200-400 mesh strongly basic anion, 2% cross-linked Dowex 1 chloride (Aldrich). Amberlite XAD-2 nonionic polymeric adsorbent and octadecyl functionalized silica gel for column chromatography were obtained from Aldrich. Bio-Sil NH2 (aminopropyl bonded silica) (40-63 µm) for column packing was purchased from BioRad Laboratories (Hercules, CA). Phosphate-buffered saline (PBS), pH 7.4, was prepared as a solution containing 8.1 mM Na2PO4, 1.2 mM KH2PO4, and 138 mM NaCl. Human serum albumin (HSA) was obtained from Miles, Inc. (Elkhart, IN). Bovine serum albumin (BSA) was obtained from ICN Biomedicals, Inc. (Costa Mesa, CA), and was diluted to 2% in PBS (10 mM sodium phosphate/ 150 mM NaCl), pH 7.4. Elemental analyses were obtained from Desert Analytics (Tucson, AZ). Cobalamin derivatives retain various quantities of water even after drying under high vacuum. Data obtained from elemental analysis were evaluated

Radioiodinated Cobalamin Conjugates

by calculating the elemental abundance for various quantities of water. The calculated carbon abundance closest to that obtained in the analysis was used to determine the number of water molecules associated with the compound. Even though all analytical samples were >98% pure by HPLC analysis, on the basis of experience with other stannyl compounds, we anticipated difficulties in obtaining satisfactory elemental analyses of stannyl derivatives 11-13. Therefore, HPLC chromatograms and NMR spectra have been submitted in support of the identity and purity of these compounds (see Supporting Information). Na[125I]I and Na[131I]I were purchased from NEN/ DuPont (Billerica, MA) as high concentration/high specific activity radioiodide in 0.1 N NaOH. Measurement of 125I and 131I was accomplished on the Capintec CRC15R or a Capintec CRC-6A radioisotope calibrator. Tissue samples were counted in a LKB 1282 gamma counter. [57Co]Cyanocobalamin was obtained from Amersham Corp. (Amersham, U.K.) at a specific activity of 210 µCi/ µg (7.78 MBq/µg). Counting of [57Co]cyanocobalamincontaining samples was conducted with a Beckman 300 gamma counter. Handling of Radioactivity. Reactions involving radioactive materials were conducted in a charcoalfiltered Plexiglas enclosure within a fume hood. All radioactive materials were opened, handled, and stored in the Plexiglas enclosure until used in studies or removed as radiation waste. Radioiodination reactions were conducted in 1 mL vials with Teflon-coated septa. The reaction mixtures were vented through a 10 mL charcoal-filled syringe, and additions of reagents or removal of materials was conducted by passing a syringe needle through the septum. Standard radiation safety procedures, including double gloves, were used. Radiation monitoring followed approved procedures. Spectroscopic Data. 1H NMR spectra were obtained on a Bruker AC-500 (500 MHz) instrument. The chemical shifts are expressed as parts per million using tetramethylsilane as an internal standard (δ ) 0.0 ppm). IR data were obtained on a Perkin-Elmer 1420 infrared spectrophotometer. UV data were obtained on a PerkinElmer Lambda 2 UV-vis spectrophotometer or a Shimadzu UV 160U spectrophotometer. UV absorbances were obtained by dissolving a weighed sample (approximately 1.00 mg) into MeOH or H2O in a 10 mL volumetric flask. A 1 mL aliquot of this solution was diluted to 2 mL in a volumetric flask. An aliquot of the diluted solution was placed in a quartz cuvette and the absorption was recorded over the wavelength range of 200-1100 nm. Extinction coefficients () were calculated from the cobalamin’s absorbance at the 360 nm peak using the Beer-Lambert equation. Mass spectral data were obtained on a VG 70SEQ mass spectrometer with 11250J data system. Fast atom bombardment (FAB+) mass spectral data were obtained at 8 kV using a matrix of 3-nitrobenzyl alcohol (3NBA) or 90% thioglycerol, 9% DMSO, and 1% TFA (DMIX). Obtaining mass spectral data for the stannyl derivatives 11-13 proved to be difficult using standard mass spectra instrumental methods. However, fully opening the mass spectrometer collector slit resulted in the obtainment of spectra consistent with the proposed structures. Analytical Chromatography. HPLC separations of (nonradioactive) compounds were obtained on a HewlettPackard 1050 quaternary gradient pumping system with a multiple-wavelength UV detector (360 nm) and a Varex ELSD MKIII evaporative light-scattering detector. Analysis of the HPLC data were conducted on Hewlett-Packard HPLC Chemstation software. All reactions were moni-

Bioconjugate Chem., Vol. 7, No. 4, 1996 463

tored by HPLC. Separations of the hippurates 16-18 were accomplished on a reversed-phase C18 column (4.6 mm × 250 mm, 5 µm, Alltima) using a gradient elution. The solvents in the gradient were acetonitrile (A) and aqueous 1% HOAc (B). The gradient began with 70% A/30% B for 2 min, and then it was increased to 100% A over the next 13 min, where it was held for an additional 10 min. Retention times for the hippurates were as follows: 16 ) 3.1 min; 17 ) 7.2 min; and 18 ) 21.7 min. Separations of the cyanocobalamin, 1, and cobalamin derivatives 2-13 were conducted on two different columns, an aminopropyl-silica column and a C18 reversedphase column at a flow rate of 1 mL/min. The methods employed for each of these columns are described separately below. Method 1: The HPLC separations of the precursor compounds 1-7 were conducted on a 5 µm, 4.6 mm × 250 mm aminopropyl column (Rainin microsorbMV amino column) eluting with 58 mM pyridine acetate, pH 4.4, in H2O/THF (96:4) solution (28). Retention times for the cobalamins evaluated with this system were as follows: 1 ) 4.3 min; 2 ) 6.5 min; 3 ) 8.0 min; 4 ) 8.8 min; 5 ) 3.0 min; 6 ) 2.9 min; and 7 ) 2.9 min. Method 2: The iodohippuryl- and stannylhippuryl-cobalamin conjugates 8-13 were evaluated by reversed-phase chromatography. Reversed-phase HPLC chromatography was carried out on a Hewlett-Packard LiChrospher 100 RP-18 (5 µm; 4.6 mm × 125 mm) C18 column using a gradient solvent system. Solvent A in the gradient was MeOH. Solvent B was H2O. Starting from 70% A, the initial solvent mixture was held for 2 min, then the gradient was increased to 100% A over the next 10 min, and 100% A was held for 5 min. Retention times under these conditions were as follows: 8 ) 8.1 min; 9 ) 7.5 min; 10 ) 7.7 min; 11 ) 14.7 min; 12 ) 14.7 min; and 13 ) 14.7 min. RadioHPLC. Chromatographic separations involving radiolabeled cobalamins were conducted on a system consisting of two Beckman Model 110B pumps, a Beckman 420 controller, a Beckman Model 153 UV detector (254 nm), and a Beckman Model 170 radioisotope detector. Separations were accomplished using two different columns at a flow rate of 1 mL/min. Method A: Initial studies used an Alltima C18 column (5 µm; 4.6 mm × 250 mm) with a gradient separation. Solvent A in the gradient was MeOH. Solvent B was H2O. Starting from 70% A, the initial solvent mixture was held for 2 min, then the gradient was increased to 100% A over the next 10 min, and then 100% A was held for 5 min. Retention times under these conditions were as follows: 8 ) 9.2 min; 9 ) 9.3 min; 10 ) 9.2 min; 11 ) 18.4 min; 12 ) 18.3 min; and 13 ) 18.3 min. Method B: This method was used to obtain purified samples of [*I]-10 from the radioiodination of e-isomer 13. For the separation, a Hamilton PRP-1 column (10 µm; 4.6 mm × 250 mm; Alltech) was used with a gradient of MeOH/H2O as in method A. Under these conditions the retention times were as follows: 10 ) 12.7 min and 13 ) 18.4 min. Preparative LC. A preparative LC system containing a Rainin Rabbit-plus peristaltic pumping system, a Dynamax (Model UV-1) UV-visible absorbance detector, and a Dynamax Model FC1 fraction collector was used to obtain pure samples of 8-13. Preparative purifications of cobalamin-stannylhippuryl conjugates 11-13 for use in radioiodination reactions were conducted by reversed-phase chromatography on an analytical system. The system was as described under Analytical Chromatography (above), except a 100 µL injection loop was used. The purification was conducted on an Alltima C18 column (5 µm; 4.6 mm × 250 mm; Alltech, Deerfield, IL) employing a gradient of MeOH/H2O. The gradient was as

464 Bioconjugate Chem., Vol. 7, No. 4, 1996

described for method 2 (above). Retention times were as follows: 11 ) 8.1 min; 12 ) 7.7 min; and 13 ) 8.1 min. Preparation of Cyanocobalamin Monocarboxylic Acids 2, 3, or 4. The b-, d-, and e-cobalamin monocarboxylates (2, 3, and 4, respectively) were prepared as previously described (28). Briefly, cyanocobalamin was hydrolyzed in 0.1 N HCl over 10 days at room temperature. Following the hydrolysis reaction, the isomeric monocarboxylates were separated from cyanocobalamin and from di- and triacids by ion exchange chromatography. Separation of the individual carboxylate isomers was accomplished by preparative liquid chromatography on an aminopropyl-silica column (25 mm × 1000 mm) at a flow rate of 0.15 mL/min. General Procedure for Conjugation of 2, 3, or 4 with 1,12-Diaminododecane; Preparation of 5-7. The conjugation of the cobalamin monocarboxylates with diaminododecane was accomplished as previously described (28). Briefly, reaction of 2, 3, or 4 with diaminododecane, EDC, and NHS in a 1:1 mixture of DMF and H2O for 4 days yielded the desired compounds after purification by ion exchange chromatography. Preparation of p-Iodohippuric Acid-Cobalamin Conjugates 8-10. To a solution containing 0.30 g (0.192 mmol) of 5, 6, or 7 and 18 µL of triethylamine in 40 mL of DMF was added 0.105 g (0.231 mmol) of p-iodohippurate TFP ester, 17, over a 5-10 min period. The reaction mixture was stirred at room temperature for 2 h (HPLC monitored) and then evaporated to dryness. The resultant solid residue was dissolved in 20 mL of 80% methanol/ H2O and loaded on a preparative liquid chromatography system. The product was eluted on a reversed-phase column (500 mm × 25 mm, Alltech, 150 psi) (octadecyl) with the same solvent mixture as used for loading the column. The fractions containing the final product were evaporated to dryness to yield 8, 9, or 10. b-Isomer (8): yield, 94 mg (80%); mp 198-204 °C (dec); 1H NMR (MeOH) δ 0.43 (s, 3H), 1.18 (s, 4H), 1.24 (d, 5H), 1.28 (m, 22H), 1.36 (m, 16H), 1.42 (s, 6H), 1.86 (s, 3H), 2.25 (d, 6H), 2.5 (d, 10H), 2.8 (s, 4H), 2.97 (s, 5H), 3.15 (m, 3H), 3.2 (s, 4H), 3.6 (m, 1H), 3.68 (d, 1H), 3.75 (m, 1H), 3.9 (d, 1H), 3.97 (d, 1H), 4.04 (m, 1H), 4.1 (d, 1H), 4.19 (m, 1H), 4.3 (m, 1H), 4.47 (m, 1H), 4.65 (m, 1H), 6.0 (s, 1H), 6.3 (d,1H), 6.5 (s,1H), 7.1 (s, 1H), 7.2 (s, 1H), 7.6 (d, 2H), 7.8 (d, 2H); MS (FAB+), mass calcd for C84H119N16O16CoPI 1825, found 1825; IR (KBr) 3400, 3200, 2950, 2060, 1660, 1570, 1490, 1060 cm-1; UV (MeOH) λ 360 nm ( ) 20 400). Anal. Calcd for C84H119N16O16CoPI‚9H2O: C, 50.72; H, 6.89; N, 11.27. Found: C, 50.92; H, 6.86; N, 11.55. d-Isomer (9): yield, 100 mg (85%); mp 236-238 °C (dec); 1H NMR (MeOH) δ 0.42 (s, 3H), 1.18 (s, 4H), 1.3 (m, 22H), 1.36 (m, 16H), 1.45 (s, 6H), 1.9 (s, 6H), 2.14 (s, 3H), 2.25 (d, 8H), 2.56 (d, 10H), 2.8-3.0 (m, 6H), 3.3 (s, 6H), 3.6 (m, 1H), 3.68 (d, 1H), 3.75 (m, 1H), 3.9 (d, 1H), 4.0 (d, 1H), 4.07 (m, 1H), 4.12 (d, 1H), 4.17 (d, 1H), 4.3 (m, 1H), 4.47 (m, 1H), 4.65 (m, 1H), 6.0 (s, 1H), 6.2 (d,1H), 6.5 (s,1H), 7.1 (s, 1H), 7.2 (s, 1H), 7.6 (d, 2H), 7.8 (d, 2H); MS (FAB+), mass calcd for C84H119N16O16CoPI 1825, found 1825; IR (KBr) 3400, 3200, 2950, 2060, 1660, 1570, 1490, 1060 cm-1; UV (MeOH) λ 360 nm ( ) 17 000). Anal. Calcd for C84H119N16O16CoPI‚8H2O: C, 51.19; H, 6.85; N, 11.37. Found: C, 51.27; H, 6.84; N, 11.33. e-Isomer (10): yield, 79 mg (67%); mp 229-232 °C (dec); 1H NMR (MeOH) δ 0.45 (s, 3H), 1.18 (s, 4H), 1.25 (d, 5H), 1.29 (m, 22H), 1.36 (m, 16H), 1.4 (s, 6H), 1.87 (s, 3H), 2.25 (d, 6H), 2.36 (d, 4H), 2.55 (d, 10H), 2.8 (m, 4H), 2.97 (s, 5H), 3.15 (m, 3H), 3.3 (s, 4H), 3.6 (m, 1H), 3.68 (d, 1H), 3.75 (m, 1H), 3.9 (d, 1H), 4.0 (d, 1H), 4.07

Wilbur et al.

(m, 1H), 4.1 (d, 1H), 4.18 (m, 1H), 4.3 (m, 1H), 4.5 (m, 1H), 4.65 (m, 1H), 6.0 (s, 1H), 6.3 (d,1H), 6.5 (s,1H), 7.1 (s, 1H), 7.2 (s, 1H), 7.6 (d, 2H), 7.8 (d, 2H); MS (FAB+), mass calcd for C84H119N16O16CoPI 1825, found 1825; IR (KBr) 3400, 3200, 2950, 2060, 1660, 1570, 1490, 1060 cm-1; UV (MeOH) λ 360 nm ( ) 17 900). Anal. Calcd for C84H119N16O16CoPI‚9H2O: C, 50.72; H, 6.89; N, 11.27. Found: C, 50.80; H, 6.48; N, 11.10. Preparation of p-Tri-n-butylstannylhippuric Acid-Cobalamin Conjugates 11-13. The same procedure was used as described for the synthesis of 8, 9, or 10 (above), except 0.141 g (0.231 mmol) of p-tributylstannyl hippurate TFP ester, 18, was added in the place of 17. b-Isomer (11): yield, 89 mg (80%); mp 182-185 °C (dec); 1H NMR (MeOH) δ 0.43 (s, 3 H), 0.88 (t, 15 H), 1.15 (t, 12 H), 1.17 (s, 4 H), 1.22 (d, 4 H), 1.29 (s, 24 H), 1.36 (br s, 6 H), 1.4 (s, 6 H), 1.6 (m, 3 H), 1.87 (s, 8 H), 1.9 (s, 1 H), 2.05 (m, 2 H), 2.25 (s, 6 H), 2.36 (m, 3 H), 2.55 (d, 10 H), 2.8 (s, 4 H), 2.97 (s, 5 H), 3.06 (t, 2 H), 3.1 (m, 3 H), 3.3 (s, 9 H), 3.34 (m, 1 H), 3.4 (m, 1 H), 3.58 (m, 1 H), 3.65 (m, 1 H), 3.75 (d, 1 H), 3.9 (d, 1 H), 4.00 (s, 2 H), 4.07 (m, 1 H), 4.1 (d, 1 H), 4.16 (m, 1 H), 4.3 (m, 2 H), 4.48 (m, 2 H), 4.6 (m, 1 H), 6.0 (s, 1 H), 6.3 (d, 1 H), 6.5 (s, 1 H), 7.0 (s, 1 H), 7.2 (s, 1 H), 7.6 (d, 2 H), 7.8 (d, 2 H); MS (FAB+), mass calcd for C96H146N16O16CoPSn 1988, found 1988, 2010 (M + Na); IR (KBr) 3400, 3200, 2950, 2060, 1660, 1570, 1490, 1060 cm-1; UV (MeOH) λ 360 nm ( ) 17 000). d-Isomer (12): yield, 95 mg (82%); mp 188-200 °C (dec); 1H NMR (MeOH) δ 0.43 (s, 3 H), 0.88 (t, 15 H), 1.15 (t, 12 H), 1.18 (s, 4 H), 1.3 (m, 13 H), 1.39 (m, 13 H), 1.45 (s, 5 H), 1.6 (m, 4 H), 1.72 (m, 2 H), 1.9 (s, 6 H), 2.25 (d, 6 H), 2.35 (m, 5 H), 2.56 (m, 5 H), 2.8-3.0 (m, 8 H), 3.15 (m, 4 H), 3.3 (m, 2 H), 3.4 (m, 2 H), 3.6 (m, 1 H), 3.68 (m, 1 H), 3.75 (m, 1 H), 3.9 (d, 1 H), 4.00 (d, 2 H), 4.07 (m, 1 H), 4.12 (d, 1 H), 4.2 (br s, 1 H), 4.3 (m, 1 H), 4.47 (m, 1 H), 4.7 (m, 1 H), 6.0 (s, 1 H), 6.2 (d,1 H), 6.5 (s,1 H), 7.1 (s, 1 H), 7.2 (s, 1 H), 7.6 (d, 2 H), 7.8 (d, 2 H); MS (FAB+), mass calcd for C96H146N16O16CoPSn 1988, found 1988, 2010 (M + Na); IR (KBr) 3400, 3200, 2950, 2060, 1660, 1570, 1490, 1060 cm-1; UV (MeOH) λ 360 nm ( ) 16 600). e-Isomer (13): yield, 78 mg (70%); mp 178-182 °C (dec); 1H NMR (MeOH) δ 0.43 (s, 3 H), 0.88 (t, 15 H), 1.15 (t, 12 H), 1.17 (m, 3H), 1.2 (d, 4 H), 1.27 (m, 15 H), 1.35 (br s, 9 H), 1.42 (s, 3 H), 1.53 (m, 2 H), 1.6 (m, 4 H), 1.86 (s, 4 H), 1.9 (s, 1 H), 2.25 (d, 6 H), 2.5 (d, 10 H), 2.8 (s, 3 H), 2.9 (m, 6 H), 2.97 (s, 5 H), 3.15 (m, 3 H), 3.2 (m, 4 H), 3.3 (s, 9 H), 3.4 (m, 3 H), 3.6 (d, 1 H), 3.75 (d, 1 H), 3.96 (d, 1 H), 4.0 (s, 2 H), 4.08 (m, 2 H), 4.19 (m, 1 H), 4.3 (m, 2 H), 4.65 (m, 1 H), 6.0 (s, 1 H), 6.3 (d, 1 H), 6.5 (s, 1 H), 7.1 (s, 1 H), 7.2 (s, 1 H), 7.6 (d, 2 H), 7.8 (d, 2 H); MS (FAB+), mass calcd for C96H146N16O16CoPSn 1988, found 1988, 2010 (M + Na); IR (KBr) 3400, 3200, 2950, 2060, 1660, 1570, 1490, 1060 cm-1; UV (MeOH) λ 360 nm ( ) 17 700). Synthesis of p-Iodohippuric Acid TFP Ester, 17. A 5.3 g (7.1 mmol) quantity of glycine was dissolved in 100 mL of 10% NaOH. To this solution was added 19.4 g (7.3 mmol) of p-iodobenzoyl chloride in several portions. The reaction mixture was allowed to stir for 10 min and then cooled in an ice-H2O bath. To facilitate stirring, 100 mL of water was added. The solution was then acidified to pH 5 with dropwise addition of 6 N HCl. The thick faint yellow precipitate was collected by vacuum filtration and dried. Recrystallization from boiling MeOH yielded 15.3 g (70%) of off-white crystals of 16, mp 228230 °C, with noticeable decomposition beginning at 202 °C: 1H NMR (DMSO-d6) δ 3.77 (d, 2H, J ) 2.7 Hz), 7.65

Radioiodinated Cobalamin Conjugates

(d, 2H, J ) 4.4 Hz), 7.85 (d, 2H, J ) 4.4 Hz), 8.57 (t, 1H, J ) 5.4 Hz); IR (nujol) 3260, 1740, 1635, 1570, 1535, 995, 830 cm-1; HRMS, calcd for C9H9NO3I (M + H) 305.9627, found 305.9615. A solution containing 1.0 g (3.3 mmol) of 16 suspended in 30 mL of anhydrous EtOAc was cooled in an ice-H2O bath. To this solution was added 1.6 g (9.6 mmol) of 2,3,5,6-tetrafluorophenol (TFP-OH) in 5 mL of anhydrous EtOAc, followed by 0.74 g (3.6 mmol) of 1,3-dicyclohexylcarbodiimide (DCC). The reaction was stirred for 1 h at 0 °C and then allowed to warm to room temperature while stirring overnight. The EtOAc was removed by rotary evaporation to yield a tacky white solid. This material was dissolved in 10 mL of hexanes, filtered, and dried to yield 0.89 g of a light yellow solid. The solid was eluted on a 2.5 cm × 40 cm silica gel 60 column with a 25% EtOAc/1% HOAc/74% hexanes mixture. Fractions of 100 mL were collected. The desired product eluted in fractions 6-11. Those fractions were combined, and the solvent was removed to yield 0.79 g (53%) of 17 as a white solid: mp 152-154 °C; 1H NMR (CDCl3) δ 4.63 (d, 2H, J ) 2.7 Hz), 6.72 (t, 1H, J ) 4.4 Hz), 7.05 (m, 1H), 7.55 (d, 2H, J ) 4.3 Hz), 7.82 (d, 2H, J ) 4.3 Hz); IR (nujol) 3270, 1775, 1630, 1525, 1190, 1180, 1150, 1060, 950, 840, 835, 775 cm-1; HRMS, calcd for C15H9NO3F4I (M + H) 453.9563, found 453.9564. Synthesis of p-Stannylhippuric Acid TFP Ester, 18. A 2.0 g (4.4 mmol) quantity of 17 was dissolved in 40 mL of anhydrous toluene under argon. To this solution was added 8.0 mL (15.8 mmol) of hexabutylditin and 100 mg (0.09 mmol) of tetrakis(triphenylphosphine)palladium(0). After stirring at room temperature for 30 min, the mixture was heated to reflux for 2 h and another 100 mg of the catalyst was added. Within 30 min, the mixture turned black. The reaction mixture was allowed to cool and toluene was removed by rotary evaporation, yielding a black oil. This oil was applied to a 250 g (4 cm × 45 cm) silica gel 60 column and eluted with 100% hexanes; 100 mL fractions were collected. At fraction 11, the eluting solvent was changed to 10% EtOAc in hexanes. The fractions were examined by HPLC, and fractions 9-15 were combined. The solvent was removed from the combined fractions to yield 1.82 g (68% yield) of a colorless oil: 1H NMR (CDCl3) δ 0.88 (t, 9H, J ) 7.3 Hz), 1.08 (m, 6H), 1.32 (m, 6H), 1.54 (m, 6H), 4.65 (d, 2H, J ) 2.6 Hz), 6.78 (t, 1H, J ) 5.4 Hz), 7.03 (m, 1H), 7.56 (d, 2H, J ) 4.0 Hz), 7.76 (d, 2H, J ) 4.0 Hz); IR (neat) 3300, 1800, 1785, 1635, 1520, 1480, 1175, 1105, 950 cm-1; HRMS, calcd for C27H36NO3F4Sn 618.1653 (M + H, Sn-120), found 618.1655. Preparation of Cyanocobalamin-Diaminododecane-p-[*I]Iodohippurate Derivatives [*I]11-[*I]13. Purification of Cyanocobalamin-Stannyl Hippurate Conjugates 8-10 by HPLC. A saturated solution of the conjugate was prepared in methanol. This solution was passed through a 45 µm filter, and three to five 20 µL aliquots were injected on the HPLC (C18 column). The peak at 8-9.2 min was collected from each injection, and the collected fractions were combined. The concentration of cobalamin derivative in the combined fractions was estimated by UV. The cobalamin-containing solution was then divided into aliquots containing 0.5 mg, and a stream of argon was passed over each aliquot to remove the solvent. The samples were stored in a refrigerator, covered from light with aluminum foil. Method A (Example for e-Isomer). This procedure was followed to obtain [131I]-10 for the first biodistribution study. In a small vial was placed 50 µL of a 1 mg/mL solution of 13 (HPLC purified) in MeOH. To this solution was added 1.5 µL (1.5 mCi) of Na[131I]I in 0.1 N NaOH,

Bioconjugate Chem., Vol. 7, No. 4, 1996 465

followed by the addition of 10 µL of a 1 mg/mL solution of N-chlorosuccinimide (NCS) in MeOH containing 5% HOAc. After 5 min, 5 µL of a 0.72 mg/mL solution of sodium metabisulfite in H2O was added to quench the reaction. A total of 1.21 mCi in an approximately 65 µL volume was removed from the reaction vessel and injected on the HPLC (C18 column). Collection of the desired peak (6-7 min) yielded 390 µCi (32%). Method B (e-Isomer). In a small vial was placed 50 µL of a 1 mg/mL solution of 13 (HPLC purified) in MeOH. To this solution was added 4.5 µL (977 µCi) of Na[125I]I in 0.1 N NaOH, followed by the addition of 4.5 µL of 0.1 N HOAc. To the resulting solution was added 10 µL of a 1 mg/mL solution of NCS in MeOH. After 15 min, 10 µL of a 1.0 mg/mL solution of sodium metabisulfite in H2O was added to quench the reaction. A total of 951 µCi in approximately 75 µL was removed from the reaction vessel by syringe and injected on the HPLC (PRP-1 column). The iodinated product was collected from the HPLC effluent (12.4-13.6 min), yielding 581 µCi (61%) of [125I]-10. The radioiodinated product was evaluated by HPLC and set in the refrigerator for storage in the HPLC solvent (Figure 4). Procedure Followed To Obtain [125I]-10 for the Second Biodistribution Study. This reaction was conducted as method B above with the following changes. A 6 µL aliquot of Na[125I]I in 0.1 N NaOH was added to the cobalamin solution. To that solution was added 6 µL of 0.1 N HOAc in MeOH. The entire reaction mixture (1.93 mCi) was drawn into a syringe and injected onto an HPLC (PRP column). The desired peak (12 min) was collected to yield 0.74 mCi (38%).2 Binding of Cobalamin-Iodohippurate Derivatives with Recombinant Human Transcobalamin II (rhTCII). rhTCII was obtained as previously reported (29). Samples (2.0 mg) of 8-10 were dissolved in 100200 µL of DMSO and diluted to 1.0 mL with H2O. Quantification of the cobalamin derivatives was accomplished by UV absorbance based on the extinction coefficients at 360 nm (data provided with individual compounds). Thus, a 20 µL aliquot was diluted to 1 mL with H2O, and the absorbance at 360 nm was determined. An aliquot of the original solution was diluted to obtain a 1 µM solution, which was further diluted to 0.1 µM for use in the assay. Apo-rhTCII was partially purified on a cation exchange column as previously described (30). The rhTCII was diluted in PBS containing 0.025% HSA to bind approximately 10 pmol of cyanocobalamin (1)/mL. The assay tubes contained 0.01 pmol of [57Co]-1, 0.1-30 pmol of unmodified 1 or cobalamin derivatives 8-10, and 100 µL of the diluted rhTCII solution. The solution volume in each tube was adjusted to 1.0 mL with PBS, and the samples were incubated at room temperature for 1 h. The free [57Co]-1 and protein-bound [57Co]-1 were separated by adsorption of free cobalamin to hemoglobin-coated charcoal (31), and the amount of radioactivity in each fraction was determined in a gamma counter. The decrease in binding of [57Co]-1 in the presence of various amounts of a cobalamin derivative was calculated and graphed (Figure 5). The quantity of each derivative required to inhibit binding of the [57Co]-1 to rhTCII by 50% was estimated, and the binding affinity of each compound relative to cyanocobalamin was determined. 2 These radioiodination conditions generally result in 45-60% isolated radiochemical yields. The low yield obtained in this experiment appeared to be due to having more base present than anticipated, resulting in more free iodide being observed than in other similar experiments.

466 Bioconjugate Chem., Vol. 7, No. 4, 1996

Animals and Tumor Model. Male athymic mice (nu/nu), obtained from Simonson Laboratories (Gilroy, CA), were housed for 1 week in the isolator facility prior to the beginning of the study. For those mice bearing tumor xenografts, tumor pieces (5-10 mg) of TK-82 renal cell carcinoma (RCC) (32) were implanted (subcutaneously) above the right shoulder using a protocol approved by the Animal Care Committee at the University of Washington. Mice in the first biodistribution had an average weight of 26.1 ( 1.9 g, and the TK-82 tumor xenografts had an average weight of 160 ( 116 mg. Mice in the second biodistribution had an average weight of 24.94 ( 1.80 g. Biodistribution Studies. Biodistribution 1. A 500 µL aliquot of sterile saline was added to the [131I]-10 fraction collected from the HPLC, and the solution was reduced to 500 µL under argon flow. The solution was further diluted with sterile saline to prepare an injectate of 5 µCi/100 µL. Twenty-four athymic mice with TK-82 RCC xenografts were injected with 5.1 µCi each of [131I]10 (specific activity of 190 µCi/µg) via the lateral tail vein. The actual amount of injectate administered to each animal was determined by weighing the syringe before and after injection (range 96-103 µL). The amount of radioactivity administered to each animal was determined by counting quadruplicate standards of 1 µL volume of the injectate and multiplying the average standard by the actual volume of injectate. Groups of eight mice were sacrificed by cervical dislocation at 1, 4, and 24 h postinjection. The biodistribution data obtained are given in Table 2, group 1. In the biodistribution, tissues were excised, blotted free of blood, weighed, and counted. The muscle in front of the neck was excised to obtain the thyroid gland. This was done to evaluate whether free radioiodide was present. The stomach was evaluated for the same reason. Urine samples were collected by a bladder tap (using a syringe) after sacrifice. The urine samples were placed in preweighed tubes and weighed after collection to quantify each sample. It is not uncommon in this procedure to obtain urine samples from some, but not all, animals. At the 1 h sacrifice time, urine samples were collected from all eight animals in the group; at 4 h, urine samples were obtained from only two animals; and at 24 h, urine samples were collected from five animals. Calculation of the percent injected dose per gram (% ID/g) and standard deviation (( SD) in the tissues was accomplished with an Excel spreadsheet. All calculations involving iodine-131 were corrected for radioactive decay. Biodistribution 2. This biodistribution was conducted as described for biodistribution 1, with the following differences. The [125I]-10 fraction isolated from the HPLC was diluted with 300 µL of DMSO, and the mixture was reduced in volume to approximately 300 µL under a stream of argon. Half of the solution (355 µCi) was transferred to a new vial, and 1.5 mL of sterile saline was added. After mixing, 1 mL was drawn into a syringe and the amount of activity in the syringe was checked. Less than 10% of the expected dose was in the syringe. The contents of the syringe was returned to the vial, and 1.5 mL of a 2% BSA/PBS solution was added, making a 1% BSA solution. Upon checking the amount of activity in a 1 mL syringe, we found that the expected amount was obtained. The 1% BSA solution was then diluted with additional 1% BSA to obtain a concentration of 7 µCi/100 µL for injection. Eighteen athymic mice were injected with 7.1 µCi each of [125I]-10 (specific activity of 0.29 µCi/µg) in approximately 100 µL (range 98-105 µL) of 1% BSA. Groups of six mice were sacrificed at 1, 4, and 24 h postinjection. The mice did not have tumor xenografts, but additional tissues were evaluated. Those

Wilbur et al.

tissues were skin (not reported), heart, and brain. The brain was obtained by cutting the back of the skull for removal. The data obtained is provided in Table 2, group 2. RESULTS

Synthesis of Arylstannane Conjugates. The syntheses of stannyl hippurate conjugates 11-13, and the corresponding iodohippuryl conjugates 8-10, were accomplished by first preparing the cobalamin monocarboxylic acids 2-4 (Table 1) (28). Preparation of monocarboxyl cyanocobalamin derivatives was accomplished by hydrolysis of the corrin ring propionamides at room temperature in 0.1 N HCl over a 10 day period. After separation from nonhydrolyzed cyanocobalamin and the di- and tricarboxylates by ion chromatography, the b-, d-, and e-monocarboxylates (2, 3, and 4, respectively) were separated by preparative liquid chromatography using an aminopropyl-silica column. The individual monocarboxylates 2-4 were coupled with diaminododecane by reaction with a carbodiimide (EDC) in DMF/H2O containing NaCN/NHS to give compounds 5-7. The carboxylate-activated hippurate derivatives 17 and 18 were prepared as depicted in Scheme 2. Preparation of iodohippurate was readily accomplished by reaction of glycine in 10% NaOH with p-iodobenzoyl chloride. Formation of the TFP ester was accomplished by reaction of tetrafluorophenol (TFP-OH) and dicyclohexylcarbodiimide (DCC) in THF. Difficulties in purification resulted in a moderate yield of 17. Stannylation was accomplished on the TFP ester containing 17 by reaction of hexabutylditin using the palladium catalyst tetrakis(triphenylphosphine)palladium(0). Iodination of 18 with NaI and NCS in MeOH resulted in conversion to the iodohippurate 17. The iodohippuryl conjugates 8-10 and stannylhippuryl conjugates 11-13 were obtained by reaction of 5, 6, or 7 with the hippuric acid derivative 17 or 18. After purification, analytical HPLC indicated that single compounds were obtained (e.g. Figure 1; Supporting Information). Radioiodination Studies.3 Initial radioiodination reactions of 11-13 were carried out in MeOH/1% HOAc solution employing NCS/Na[125I]I. Reaction with each of the three isomers provided a radioiodinated compound in high yield, but the retention time of the radiolabeled species did not correspond to that of the co-injected iodo standards (8, 9, or 10). An example is shown in Figure 2, where the radioHPLC had a major radioactivity peak at 8.4 min and a minor peak at 9.5 min (Figure 2A). After co-injection of the radiolabeled material and the iodo standard (Figure 2B), it became apparent that the radioactivity peak at 8.4 min was not the desired product, as the iodo standard peak had a retention time of 9.2 min (with shoulder at 8.4 min) by UV detection, and the UV detector comes before the radioactivity detection in the effluent flow. Thus, it seemed likely that the radioactivity peak corresponding to the desired product had a retention time of 9.5 min. A concern about oxidative side products, particularly lactone formation (Scheme 1), produced under acidic conditions, led to an investigation of (nonradioactive) iodination reactions under acidic, basic, and neutral conditions. To be more reflective of radiolabeling conditions, the studies were conducted at 0.1 equiv of NCS and NaI. Under acidic conditions, a radioiodinated product that did not correspond to the iodo standard was 3 Note that three different HPLC systems and conditions were used to monitor radioactive and nonradioactive reactions. The equipment and conditions are described under Experimental Procedures.

Bioconjugate Chem., Vol. 7, No. 4, 1996 467

Radioiodinated Cobalamin Conjugates Table 1. Synthesized Derivatives of Cyanocobalamin for Radioiodination Studies

obtained. Under basic conditions, no reaction occurred. However, under neutral conditions, a product that coeluted with the iodo standard was obtained. Although the arylstannanes 11-13 appeared to be >95% pure by HPLC after the first purification (e.g. Figure 1), the starting material was repurified by preparative HPLC prior to further radioiodinations. A second purification step is often conducted to assure that reactive minor impurities will not compete in ncaradioiodinations. Similar to the iodination reactions, radioiodination of the HPLC-purified starting material under neutral conditions provided the desired radioiodi-

nated product (i.e. 9.5 min). Unfortunately, the yield was low (25-30%), and an additional radioiodinated species which was similar in lipophilicity to the starting arylstannane (i.e. 19 min) was also obtained. The HPLC retention time for the new radioiodinated species indicated that the arylstannane was not substituted by iodine, suggesting that the radioiodide may have displaced the cyano group on the cobalt. To reverse this, the reaction was quenched with the addition of NaCN. Indeed, the lipophilic radioactivity peak previously seen at 19 min disappeared and a new radioactivity peak at 3.7 min appeared. Addition of NaCN to the reaction

Scheme 2. Synthesis of Hippuryl TFP Esters

a

Glycine, 10% NaOH, 76%.

b

TFP-OH, DCC, 52%. c n-Bu6Sn2, [(C6H5)3P]4Pd(0), 67%.

d

NaI, NCS, MeOH/5% HOAc.

468 Bioconjugate Chem., Vol. 7, No. 4, 1996

Wilbur et al.

Figure 1. HPLC chromatograms (UV detection) for the isolated products from the reaction of aminododecylcobalamin, 7, with TFP ester of p-iodohippurate, 17, and p-tri-n-butylstannylhippurate, 18, to yield cobalamin conjugates 10 (A, left) and 13 (B, right). A reversed-phase C18 column was used for the separation. Conditions are described under Experimental Procedures, Analytical Chromatography, Method 2.

mixture was not desired, and the radiolabeling yields were not optimal, so additional radiolabeling studies were conducted. Subsequent radioiodination studies demonstrated that the reactions could be conducted under near-neutral conditions to give 38-65% yields (in nine syntheses) of the desired radioiodinated product. Furthermore, most reaction conditions did not produce the more lipophilic (later eluting) radioiodinated species. Isolation of the desired radioiodinated product was achieved by collection of the HPLC effluent at the appropriate time. However, initial attempts at isolation of the radioiodinated cobalamins from the HPLC effluent were inefficient, as it appeared that a large proportion of the radioactivity was being retained on the C18 HPLC column. While it was not apparent why this occurred, it appeared that the compound was binding nonspecifically to the column. This nonspecific binding was also observed with glass vials and the syringe needles used to administer the compound to mice.4 Use of a different HPLC column (Hamilton PRP-1) resulted in good recovery of the radiolabeled cobalamin. Thus, optimization of the radioiodination reactions included varying reaction conditions and using a column that did not retain the product. The in vitro stability of [125I]-10 was evaluated by re-examining the radioiodinated cobalamin after storage in a refrigerator for over 3 months. Radiochromatograms of the isolated product after storage for 1 h, 16 h, and 37 days are shown in Figure 4. Radiochromatograms taken at various times up to 99 days postisolation looked identical to those shown in Figure 4. TCII Binding. Binding of cobalamin derivatives with TCII is essential to the receptor binding and cellular uptake of these compounds by cells. To assess the binding of nonradioactive iodohippuryl conjugates of cyanocobalamin, competitive binding assays were performed using [57Co]cyanocobalamin, as the tracer, and partially purified recombinant human apo-rhTCII. The binding curves for compounds 8-10 are compared with that of cyanocobalamin in Figure 5. In comparison with cyanocobalamin, which binds rhTCII with high affinity (Ka ) 10-12 M), the iodohippurate derivatives

of the e-carboxylate isomer (10) and b-carboxylate isomer (8) bind rhTCII with 2- and 16-fold lower apparent affinities, respectively. The d-iodohippurylcobalamin derivative (9), on the other hand, binds with a 75-fold lower apparent affinity than [57Co]-1, requiring 152 pmol to inhibit binding of [57Co]-1 to rhTCII by 50% (value derived by extrapolation of the binding data). Biodistribution in Mice. Two biodistributions of radioiodinated 10 were carried out to determine the tissue distribution in mice and to assess the in vivo integrity of the radioiodine label. In one distribution, mice bearing human renal cell carcinoma (RCC) tumor xenografts, TK-82 (32), were studied, as it was of interest to determine whether this radiolabeled derivative would localize, or be retained in, that tumor model. We chose to investigate the TK-82 tumor model even though data on the numbers of cobalamin/TCII receptor sites on this tumor xenograft could not be readily obtained due to the fact that it is not propagated as a cell culture line. In the first biodistribution, radiolabeling of 13 with nca-[131I]iodide resulted in the obtainment of 390 µCi of [131I]-10 with a specific activity of 0.19 Ci/mg (0.19 µCi/ ng). A solution containing approximately 5 µCi/100 µL was prepared, and this amount was injected via the lateral tail vein into 24 athymic mice bearing TK-82 tumor xenografts. Animals were sacrificed at 1, 4, and 24 h postinjection. Data obtained for the distribution of radioactivity (% ID/g) in nine tissues, blood, and urine are provided in Table 2, group 1. All tissue concentrations decreased with time, except the stomach, which increased at 4 h over that seen at 1 h. Kidney concentrations were the highest at the 1 h time point and, with the exception of neck (thyroid), retained the highest level of radioactivity per gram. Urine collection indicated that a large portion of the radioactivity may have been excreted by this route.5 The results from the first biodistribution led to an uncertainty regarding the stability of the radioiodine label in vivo, as high neck (thyroid) and stomach concentrations were observed. The fact that the concentra-

4 There were difficulties with an apparent nonspecific binding of radioiodinated 10. Upon removal of the MeOH from the isolated HPLC solvent, the radioactivity began to adhere to glass surfaces and to syringe needle surfaces, making it difficult to obtain the quantities of radiolabeled material sought. This is thought to be related to the decreased water solubility of the iodohippuryldiaminododecane-linked cyanocobalamin from that of cyanocobalamin.

5 The total percentage of radioactivity being excreted in the urine was not determined. The feces were not collected, and therefore it is difficult to determine how much activity was excreted via that route. However, on the basis of the relatively low values (% ID/g) in the intestines versus the quantities seen in the kidneys and urine, it would seem that hepatobiliary was not the major route of excretion for the radioiodinated cyanocobalamin derivative.

Radioiodinated Cobalamin Conjugates

Bioconjugate Chem., Vol. 7, No. 4, 1996 469

Figure 2. HPLC radiochromatograms (left) and UV chromatograms (right) of (A, top) the reaction mixture from radioiodination of 13 using NCS and Na[125I]I under acidic conditions and (B, bottom) the same reaction mixture with added iodo standard, 10 (9.23 min on UV trace). A reversed-phase C18 column was used for the separation with conditions described under Experimental Procedures, RadioHPLC, Method A.

tion of radioiodine in the neck decreased at 4 and 24 h was suggestive of having a bolus of radioiodine injected with the preparation, rather than in vivo degradation to free iodide. To better assess this, a second biodistribution was conducted in athymic mice. Because previous investigators had reported that high concentrations of vitamin B12 (cyanocobalamin) were found in rat hearts and pituitary glands (33), we chose to include heart and brain tissues in the study. In the second biodistribution, 13 was radiolabeled with nca-[125I]iodide, resulting in the obtainment of 740 µCi of [125I]-10 having a specific activity of 0.29 Ci/mg. Due to difficulties with the radiolabeled cobalamin adhering to the glass vial in the previous biodistribution study, the fraction containing [125I]-10 collected from the HPLC

effluent was diluted with 300 µL of DMSO prior to removal of the MeOH/H2O under a stream of N2. This procedure worked well, but dilution to a concentration of DMSO of