Bioconjugate Chem. 1990, 7, 65-71
(9) Brechbiel, M. W., Gansow, 0. A., Atcher, R. W., Schlom, J., Esteban, J., Simpson, D. E., and Colcher, D. (1986) Synthesis of 1-(p-isothiocyanatobenzyl)derivatives of DTPA and EDTA. Antibody labeling and tumor-imaging studies. Inorg. Chem. 25, 2772, 2781. (10) Brechbiel, M. W. (1988) New Bifunctional Ligands f o r Radioimmunoimaging and Radioimmunotherapy Ph. D. Thesis, The American University, Washington, D. C. (11) Meares, C. F., McCall, M. J., Reardon, D. T., Goodwin, D. A,, Diamanti, C. I., and McTigue, M. (1984) Conjugation of antibodies with bifunctional chelating agents: Isothiocyanate and bromoacetamide reagents, methods of analysis,and subsequent addition of metal ions. Anal. Biochem. 142,6878. (12) Kozak, R. W.; Raubitschek, A., Mirzadeh, S., Brechbiel, M. W., Junghaus, R., Gansow, 0. A., and Waldmann, T. A. (1989) Nature of the bifunctional chelating agent used for radioimmunotherapy with yttrium-90 monoclonal antibodies: Critical factors in determining in vivo survival and organ toxicity. Cancer Res. 49, 2639-2644. (13) Roselli, M., Schlom, J., Gansow, 0. A., Raubitschek, A., Mirzadeh, S., Brechbiel, M. W., and Colcher, D. (1989) Comparative biodistributions of yttrium- and indium-labeled monoclonal antibody B72.3 in athymic mice bearing human colon carcinoma xenografts. J. Nucl. Med. 30,672-682. (14) Paik, C. H., Murphy, P. R., Eckelman, W. C., Volkert, W. A., and Reba, R. C. (1983) Optimization of DTPA mixed-anhydride reaction with antibodies at low concentration. J. Nucl. Med. 24, 932-936. (15) Paik, C. H., Herman, E., Eckelman, W. C., and Reba, R. C. (1980) Synthesis, plasma clearance, and in vitro stability of protein containing a conjugated indium-111 chelate. J. Radioanal. Chem. 57,553-564. (16) Kawamura, A. (1969) in Fluorescent Antibody Tech-
65
niques and Their Applications (Kawamura, A., Ed.) Chap-
ter 3, University of Tokoyo, Tokyo. (17) Mizusawa, E. A,, Thompson, M. R., and Hawthorne, M. F. (1985) Synthesis and antibody-labeling studies with the p-isothiocyanatobenzene derivatives of 1,2-dicarba-closododecaborane and the dodecahydro-7,8-dicarba-nido-undecaborate(I-) ion for neutron-capture therapy of human cancer. Znorg. Chem. 24, 1911-1916. (18) Leonard, W. J., Depper, J. M., Robb, R. J., Waldmann, T. A., and Green, W. C. (1983) Characterization of the human receptor for T-cell growth factor. Proc. Natl. Acad. Sci. U.S.A. 80, 6957-6961. (19) Colcher,D., Horan Hand, P., Nuti, M., and Schlom,J. (1981) A spectrum of monoclonal antibodies reactive with mammary tumor cells. Proc. Natl. Acad. Sci. U.S.A. 78, 31993203. (20) Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fugimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C. (1985) Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 7685. (21) Nelson, F., Murase, F., and Kraus, K. A. (1964) I. Cation exchange in concentrated HC1 and HClO,. J. Chromatog. 13, 503. (22) Assony, S. J. (1961) The Chemistry of Isothiocyanates. In The Chemistry of Organic Sulfur Compounds (N. Kharasch and C. Y. Meyers, Eds.) pp 326-338, Pergamon Press, New York.
(23) Satchell, D. P. N. and Satchell, R. S. (1975) Acylation by ketenes and isothiocyanates. A mechanistic comparison. Chem. SOC.Rev. 4, 231-250. Registry No. lM3B-DTPA, 121806-84-6;l"In, 15750-15-9.
Preparation and Characterization of Paramagnetic Polychelates and Their Protein Conjugates Paul F. Sieving, Alan D. Watson, and Scott M. Rocklage* Salutar, Incorporated, 428 Oakmead Parkway, Sunnyvale, California 94086. Received July 31, 1989
The gadolinium complexes of poly-L-lysine-poly(diethylenetriamine-N,N~,N",N"-pentaacetic acid) (Gd-PL-DTPA) and poly-L-lysine-poly(1 ,4,7,1O-tetraazacyclododecane-~,N'JV",N"'-tetraaceticacid) (Gd-PL-DOTA) and their conjugates with human serum albumin (HSA) have been prepared and characterized. Poly-L-lysine (PL, degree of polymerization = 100) was N-acylated with a mixed anhydride of the chelating ligand (DTPA or DOTA). Sixty to ninety chelating groups per molecule of PL could be attached in this way. Following purification of the polychelate by size-exclusion chromatography, the gadolinium complexes were prepared by standard methods and conjugated to HSA with heterobifunctional cross-linking reagents. The molar relaxivities of these macromolecular species were 2-3-fold higher than those of the corresponding monomeric metal complexes ([Gd(DTPA)] and [Gd(DOTA)]). The conjugation conditions were optimized to produce conjugates containing 60-90 metal centers per molecule of HSA (ca. one polychelate per protein).
We are conducting a research program to investigate the feasibility of preparing derivatives of biologically active macromolecules containing a large number of covalently bound metal chelates. The utility of such species for producing target-specific contrast enhancement in mag1043-1802/90/2901-0065$02.50/0
netic resonance imaging (MRI), as well as for applications in nuclear medicine, is anticipated. The present system was proposed as a model of the physical and chemical properties of such systems in general. The polychelate approach was developed as a means of preserving 0 1990 American Chemical Society
66
Sieving et at.
Bioconjugate Chem., Vol. 1, No. 1, 1990
the biological activity of the targeting protein, consistent with the proposed high degree of loading with metal chelates. Earlier attempts focused on direct attachment of chelates to proteins of interest (I,2). These employed DTPA (diethylenetriamine-N,N,N',N'',N"-pentaaceticacid) as the chelating group and utilized two different methods for attachment of DTPA to macromolecules. The mixed anhydride method (3) reduced the likelihood of crosslinking or polymerization of the macromolecular substrate, while the bicyclic anhydride method ( 4 ) offered the advantage of greater ease of use. Both methods, however, suffered from a common drawback when applied to direct modification of proteins. It was observed ( 5 , 6 ) that increased levels of functional group modification of a protein, such as an antibody, resulted in a decrease in the biological activity of the protein. This compromise in biological activity was a serious limitation in view of the requirement for high metal loading of proteins when used as target specific MRI contrast agents (7-9). It is believed that a 50% reduction in T , relaxation time of water protons in the target tissue is the minimum requirement for an effective MRI contrast agent. Considering the affinity of antibodies for their antigens and the concentration of these antigens in the target tissues, each antibody molecule must carry many paramagnetic tenters to bring about these levels of T , reduction. More recent efforts in this area (10, 11) have focused on the use of intermediary carriers of chelates as a means of overcoming this limitation. Although Manabe and coworkers prepared a polylysine-DTPA (PL-DTPA) species, the cyclic anhydride method was used and resulted in only ea. 40 DTPA groups per polylysine (PL) (n = 109). Torchilin e t al. prepared PL-DTPA and PLEDTA as well as their respective MAb conjugates. Although they used the mixed anhydride method, their method of conjugation was nonspecific and the end result was a conjugate that lacked structural definition. Another drawback common to both methods, with or without the use of an intermediary carrier, arose from the chemical transformations employed in linking the chelate to the macromolecular substrate. Both of the methods described relied on the chemical modification of one of the carboxylate groups normally involved in metal binding and resulted in decreased chelate stability with respect to the monomeric chelates (12). Methods were developed to address this problem, particularly when applied to the radiolabeling of proteins by direct attachment of chelates (13). Meares et al. prepared and characterized a series of bifunctional chelators, based on macrocyclic ligands such as DOTA (1,4,7,10-tetraazacyclododecaneN,N',N",N"'-tetraacetic acid), which featured the linking group attached to the backbone of the ligand. None of the coordination sites on these ligands were compromised, and full denticity was retained. The methods employed in the present work have not addressed this issue specifically; however, the use of DOTA in place of DTPA is expected to result in a lower rate of metal release in vivo. This improvement may well be sufficient to minimize metal loss for applications in which the agent is cleared from the body in under 48 h. The present work is an extension of efforts to prepare intermediary carriers and incorporates some practical improvements as described below. EXPERIMENTAL PROCEDURES Reagents. DOTA was synthesized by published methods (14) and isolated as a zwitterion by precipitation from concentrated aqueous solution at pH 2.5 following ion-
exchange chromatography. All other reagents were purchased from commercial suppliers and used as received unless otherwise specified. NMR Measurements. Proton nuclear magnetic resonance spectra were recorded on a Bruker AM250 spectrometer at an ambient temperature of 21 f 1 "C. In the case of DOTA, the spectra were used to establish identity and estimate purity. In the case of PL-DTPA and PL-DOTA, the spectra were used to estimate the extent of acylation of PL by the mixed anhydride of the chelator via integration of the appropriate peaks. Relaxivity Measurements. Relaxation times (T,) in water of the title compounds were measured on a RADX Model 530 proton spin analyzer a t 10 MHz, 37 "C, for gadolinium concentrations of 0.04-2.0 mM. The data was fitted by nonlinear regression to the equation T , = 1/ A[Gd] B to determine relaxivities, where A is the relaxivity and B is 1/T, for pure water. Titrimetric Measurements. The titration data were acquired with an automatic titrator system as previously described (15). In a typical measurement, a sample of one of the polychelate complexes (Gd-PL-DTPA or Gd-PL-DOTA), 20-50 mg/mL in water, was adjusted to pH 1, degassed to remove CO,, and titrated to pH 11.5 with 0.100 M KOH. The data were plotted as pH vs volume of titrant added, the first derivative (dpH/ dvol) was taken, and the volume of titrant added between the two maxima in the pH range 9-11 was measured. Directly Coupled Plasma Atomic Absorption (DCPAA) Measurements. DCP-AA determinations of gadolinium concentration were performed on a Beckman SpectraSpan IV instrument. Synthesis of DTPA Mixed Anhydride. A 25-mL round-bottom flask was fitted with a water condenser and charged with 1.622 g (4 mmol) of DTPA and 7.0 mL of acetonitrile (dried over 4-A molecular sieves). Triethylamine (2.79 mL, 20 mmol) was added and the mixture was stirred magnetically under an atmosphere of nitrogen at 60 "C for 1 h until homogeneous. This solution was cooled to -30 "C under an atmosphere of nitrogen and stirred while adding 0.520 mL (4 mmol) of isobutyl chloroformate (IBCF) slowly over 5 min. The resultant slurry was stirred for 30 min a t -30 "C. Synthesis of PL-DTPA. The mixed anhydride slurry was added slowly over 5 min to a solution of 0.250 g of poly-L-lysine hydrobromide (Sigma Chemical Company, degree of polymerization = 105, MW = 22 000) in 12.5 mL of 0.1 N sodium bicarbonate, pH 9.0, which was cooled in an ice bath. The resulting mixture was allowed to warm to room temperature and stirred for 6 h. The majority of the acetonitrile was removed by rotary evaporation a t 60 "C and the resulting aqueous solution was dialyzed in 1 2 500 MW cutoff tubing against 3.5 L of 0.02 M oxalic acid, pH 2.0, for 6 h a t room temperature then against 3.5 L of 0.05 M sodium bicarbonate, pH 8.0, for 12 h at room temperature. Residual low molecular weight components were removed by gel filtration on Sephadex G25 with UV detection a t 254 nm, eluting with 10 mM sodium bicarbonate, 15 mM NaC1, pH 8.0. Lyophilization (10 pm, 24 h) afforded 0.532 g of white, amorphous solid. Analysis by 'H NMR demonstrated 0.88 DTPA group per lysine residue, indicating that 88% of the lysine t-amines were acylated. Synthesis of Gd-PL-DTPA. PL-DTPA (0.100 g) was dissolved in 3.0 mL of standardized 50.1 mM GdCl, in 0.1 N HC1. The solution was adjusted to pH 7.0 with 7.0 N NaOH and stirred for 30 min at room temperature. A small aliquot of the solution tested positive for free Gd3+ when added to 1.0 mL of 10 pM arsenazo I11 in
+
Paramagnetic Polychelates
acetate buffer, pH 4.0. A further 2 mg of PL-DTPA was added to the solution, and another aliquot tested negative for free Gd3+. The relaxivity of the complex was determined to be 10.80 mM-' s-l, on the basis of a DCPAA determination of Gd concentration. Titrimetric determination of free t-amines, in conjunction with the DCP data, confirmed 88% acylation of t-amines with DTPA (% acylation = ([Gd]/[Gd] + [RNH,]) X 100 and assuming that [Gd] i= [DTPA]). HPLC analysis (column, Supelco C18 deactivated; mobile phase, 5 mM triethylammonium acetate, pH 6.8) demonstrated the absence ( e l % ) of free [Gd(DTPA)]. The samples from nondestructive analyses were recombined with the stock solution and desalted by gel filtration on Sephadex G-25. Lyophilization afforded 0.135 g of white solid. Synthesis of DOTA Mixed Anhydride. A 25-mL round-bottom flask was fitted with a water condenser and charged with 0.808 g (2 mmol) of DOTA and 5.0 mL of acetonitrile. Tetramethylguanidine (1.0 mL, 8 mmol) was added and the mixture was stirred until homogeneous (5 min). Stirring was discontinued and the solution was dried overnight over 4-A molecular sieves. The solution was decanted from the sieves, placed under an atmosphere of nitrogen, cooled to -30 "C, and stirred while adding 0.260 mL (2 mmol) of IBCF slowly over 5 min. The resultant slurry was stirred for 1 h at -30 "C. Synthesis of PL-DOTA. The slurry from above was added slowly over 5 min to a solution of 0.100 g of PL-HBr in 6.0 mL of 0.1 N sodium bicarbonate, pH 9.0, which was cooled in an ice bath. The resulting solution was allowed to warm to room temperature and was stirred for 6 h. The reaction mixture was worked up and purified as described for PL-DTPA, affording 0.183 g of white solid. Analysis by 'H NMR demonstrated 0.68 DOTA group per lysine residue, indicating that 68% of the lysine t-amines were acylated. Synthesis of Gd-PL-DOTA. PL-DOTA (0.300 g) was dissolved in 5.0 mL of 50.1 mM GdC1, in 0.1 N HC1 and adjusted to pH 7.0 with 7.0 N NaOH. After stirring for 1 h a t room temperature, the solution tested negative for free Gd3+. The mixture was maintained between pH 6.0 and 7.0 with 7.0 N NaOH, while additional 0.5-mL aliquots of 50.1 mM GdC1, were added a t 1-h intervals until the solution tested positive for free Gd3+. The solution was stirred overnight and 1-mg aliquots of PLDOTA were added until the solution tested negative for free Gd3+. The relaxivity of the complex was 13.03 mM-l s-l. Data from titrimetry and DCP-AA confirmed 68% acylation of t-amines with DOTA. HPLC analysis demonstrated the absence (e170)of free Gd-DOTA. Desalting and lyophilization afforded 0.360 g of white solid. Activation of HSA for Conjugation. HSA contains one native sulfhydryl residue, which was blocked by alkylation. HSA (1 g, 15 pmol) was dissolved in 50 mL of 0.05 M Tris-HC1, pH 8.0, in a 100-mL round-bottom flask. The flask was purged with dry nitrogen, sealed with a septum and wrapped with aluminum foil to exclude light. A solution of 15 mg (80 pmol) of iodoacetamide in 4.0 mL of 1.0 N NaOH was added by syringe through the septum and the mixture was stirred for 45 min a t room temperature in the dark. The reaction mixture was dialyzed against 3.5 L of 0.05 M sodium bicarbonate, pH 8.0, for 1 2 h with a buffer change at 6 h. Lyophilization of the dialysate afforded 0.903 g of white fibers. The absence of free sulfhydryls in the preparation was demonstrated by the method of Ellman (16). Thiol-blocked HSA (0.100 g) was dissolved in 50 mL of 50 mM triethanolamine, 7 mM monopotassium phos-
Bioconjugate Chem., Vol. 1. No. 1. 1990 67
phate, 100 mM NaC1, 1 mM EDTA, pH 8.0. The solution was degassed for 10 min by stirring under vacuum and was covered with an atmosphere of nitrogen, and the flask was sealed with a septum and cooled to 4 "C. A solution of 8.6 mg of 2-iminothiolane in 100 pL of 1 M triethanolamine hydrochloride, pH 8.0, was added by syringe through the septum and the mixture was stirred for 90 minutes a t 4 "C. The reaction mixture was dialyzed overnight against 3.5 L of 0.08 M sodium phosphate, 0.5 mg/mL EDTA, pH 8.0, with frequent buffer changes for the first few hours. Assay by the method of Ellman demonstrated 1.32 sulfhydryl residues per mole of HSA. Activation of Gd-PL-DTPA or Gd-PL-DOTA for Conjugation to HSA. A 0.200-g sample of either GdPL-DTPA or Gd-PL-DOTA was dissolved in 18 mL of 8 mM sodium phosphate, pH 8.0, and the mixture was treated dropwise with a solution of 10.7 mg of N-succinimidyl 4-(N-maleimidomethyl)cyclohexane-l-carboxylate (SMCC) in 2.0 mL of DMSO, and stirred for 30 min. The reaction mixture was dialyzed overnight against 3.5 L of deionized water with frequent changes for the first few hours. Reaction of an aliquot of the dialysate with a known amount of 2-mercaptoethanol and measurement of the residual sulfhydryls by the method of Ellman, indicated 1.30 maleimide residues per mole of polychelate. Conjugation of Polychelates to HSA. The solutions containing activated polychelate and activated HSA were combined and the mixture was stirred for 4 h a t room temperature. Analysis of an aliquot by the method of Ellman indicated that over 90% of the thiols were consumed. The mixture was lyophilized, the residue was dissolved in 10 mL of deionized water and dialyzed for 6 h against 3.5 L of deionized water. The dialysate was fractionated by size-exclusion chromatography (SEC) on Sephacryl S-300 eluting with 10 mM NaH,PO,, 15 mM NaCl, pH 8.0, and the fractions with a significant absorbance a t 280 nm were pooled and lyophilized to yield 0.30 g of fibrous solid. The relaxivities of the conjugates were essentially the same as those of the polychelates from which they were prepared ( 10 mM-' s-l for GdPL-DTPA-HSA, 13 mM-l s-l for Gd-PL-DOTAHSA). Absorbance (280 nm) and DCP-AA data indicated that the metal content of the conjugates was 6080 mol of Gd/mol of HSA.
-
N
RESULTS AND DISCUSSION
Synthesis of PL-DTPA and PL-DOTA. The syntheses of PL-DTPA and PL-DOTA are shown in Schemes I and 11,respectively. The published procedure for preparation of DTPA mixed anhydride was revised to improve efficiency and reproducibility. Anhydrous grades of reagents (DTPA, triethylamine, acetonitrile) were used and the reaction was carried out under an atmosphere of nitrogen a t low temperature to minimize hydrolytic decomposition of the mixed anhydride prior to reaction with polylysine. A series of experiments was carried out to determine a ratio of mixed anhydride to t-amines required to achieve 80-90% acylation which left some amines available for subsequent conjugation reactions. A 4-fold molar excess of mixed anhydride with respect to t-amines was satisfactory. Purification of the polychelate was effected by a combination of size-exclusion techniques. After evaporative removal of acetonitrile, the major impurities were excess DTPA, triethylamine, and isobutyl alcohol. It was found that dialysis at low pH (6) was required to remove excess DTPA,
Sieving et al.
Bioconjugate Chem., Vol. 1, No. 1, 1990
68
Scheme I. Synthesis of PL-DTPA
Scheme 11. Synthesis of PL-DOTA H02C7
N - N P 0 2 H
,>
\OzH N' HO,C
\O,-
-0
2 C ~ N n N n N r C 0 2 - . 5Et3NH+ k 0 , -
I
-0,c-
+
IBCF
+
N
N
- 0C , -J
ACN 7
\--CO,H
ACN
4TMGH'
-30 "C
\co,-
- 0 , c ~/I
4TMG
- 0,c /-C(0)OC02CH&H(CH3), N -02'4Et3NH+
+ Et3NHCI
Lo,-
-
-
- O z C 7
(O)OC02CH2CH(CH,), o'c7NANP
N.
- 0,c
r" (O)OCO~CH,CH(CHJ),
'N
3TMGH'
A W L-GO2-
