Monoclonal Antibody Radiopharmaceuticals - American Chemical

546

Bioconjugate Chem. 2003, 14, 546−553

Monoclonal Antibody Radiopharmaceuticals: Cationization, Pegylation, Radiometal Chelation, Pharmacokinetics, and Tumor Imaging Hwa Jeong Lee and William M. Pardridge* College of Pharmacy, Ewha Womans University, Seoul 120-750, South Korea and Department of Medicine, UCLA School of Medicine, Los Angeles, California 90024. Received December 31, 2002; Revised Manuscript Received February 20, 2003

The 528 murine monoclonal antibody (MAb) to the human epidermal growth factor receptor (EGFR) was sequentially cationized with hexamethylenediamine and conjugated with diethylenetriaminepentaacetic acid (DTPA) as a potential antibody radiopharmaceutical for imaging EGFR-expressing cancer. The cationized 528 MAb was characterized with isoelectric focusing and electrophoresis, and an immunoradiometric assay, which showed the affinity of the 528 MAb for the human EGFR was retained following cationization. The native or cationized 528 MAb, labeled with 111In, was injected intravenously in scid mice bearing human U87 flank tumors, which express the EGFR, and tumor imaging was performed with both external detection in live animals and with whole body autoradiography. However, the tumor signal was not increased with the cationized MAb, relative to the native MAb, and this was due to a serum inhibition phenomenon that was confirmed by a pharmacokinetics analysis in control mice. In an attempt to block the serum inhibition, the cationized 528 MAb was pegylated with 2000 Da poly(ethylene glycol), and the cationized/pegylated MAb was conjugated with DTPA and labeled with 111In. However, a pharmacokinetics analysis showed the pegylation did not reverse the serum inhibition of the cationic charge on the MAb. These studies describe methods for reformulating monoclonal antibodies to develop improved radiopharmaceuticals, but show that radiolabeling a cationized MAb with DTPA produces a serum neutralization of the initial cationization modification.

INTRODUCTION

Monoclonal antibody radiopharmaceuticals are potential diagnostic agents for the imaging of cancer and other regional disorders with standard imaging modalities such as single photon emission computed tomography. However, the initial promise of monoclonal antibody radiopharmaceuticals as tumor imaging agents has not been realized to date (1), because the intensity of the image over the tumor, relative to surrounding tissue, is generally not sufficiently high to allow for detection of small carcinomas. The single most important factor limiting imaging with monoclonal antibody radiopharmaceuticals is the poor transport of these large molecule drugs across the microvascular endothelial barrier of the capillaries perfusing the tumor or target end organ. The capillary wall of the tumor forms a blood-tumor barrier, and the poor transport of the antibody radiopharmaceutical across the microvascular barrier has two effects, both of which are deleterious to the tumor image. First, the slow exodus of the antibody from the vascular compartment within the tumor reduces the access of the antibody to the target antigen on the tumor cell, which reduces the imaging “signal”. Second, the slow exodus of the antibody from the vascular compartment results in the prolonged blood residence time, which increases the imaging “noise” from the tissue surrounding the tumor. * To whom correspondence should be addressed: Dr. William M. Pardridge, UCLA Warren Hall 13-164, 900 Veteran Ave., Los Angeles, CA 90024. Phone: (310) 825-8858. Fax: (310) 2065163. E-mail: [email protected].

Monoclonal antibody radiopharmaceuticals could be new tumor imaging agents given the development of a technology that enhances monoclonal antibody transport across microvascular barriers. The antibody that is enabled to cross microvascular barriers must be reformulated in such a way that the affinity of the antibody for the target antigen is not compromised. These objectives may be accomplished with antibody cationization (2). In this approach, surface acidic amino acids on the monoclonal antibody are conjugated with amino group ligands, resulting in the conversion of surface carboxyl groups into extended amino groups on the modified antibody. Cationized proteins are generally taken up by cells via absorptive-mediated endocytosis (3, 4). This uptake is caused by electrostatic interactions between the cationic groups on the protein and anionic moieties on the plasma membrane of the cell (5). Native monoclonal antibodies are general excluded from the cell and it is necessary to physically inject an antibody into the cell to enable access of the antibody to an intracellular antigen (6). However, a cationized antibody is rapidly endocytosed by cells, and cationized antibodies rapidly undergo absorbtive-mediated transcytosis across microvascular endothelial barriers in vivo (2). The monoclonal antibody must be cationized in such a way that the affinity of the antibody for the target antigen is largely retained. Prior work has shown that monoclonal antibodies directed against a variety of antigens, including amyloid peptides (7), oncogenes (8, 9), or viral proteins (10), can be cationized with retention of antigen affinity. Pharmacokinetic analyses of cationized monoclonal antibody clearance from blood have

10.1021/bc0256648 CCC: $25.00 © 2003 American Chemical Society Published on Web 03/25/2003

Antibody Radiopharmaceuticals

shown that the rate of exodus of the antibody from the blood is increased manyfold following cationization (79, 11). In these previous investigations of cationized antibodies, the cationized monoclonal antibody was radiolabeled with 125-iodine. However, the use of the 125iodine radionuclide is not optimal for in vivo imaging, because the blood rapidly accumulates with low molecular weight, highly diffusable radioactive metabolites. These 125I-labeled metabolites are generated by the rapid cellular metabolism of the 125I-labeled cationized antibody, which occurs following the organ uptake of the antibody. The accumulation of the low molecular weight, diffusable 125I-labeled metabolites in the organ diminishes the quality of the image. An alternative radionuclide for labeling cationized monoclonal antibodies is 111-indium. The principal advantage of this radionuclide over 125I is that 111In labeled metabolites are retained in the peripheral tissues and do not return to the blood circulation (12). In contrast, the 125I-labeled metabolites produced in peripheral tissues are exported back to blood to accumulate in the other organs. Labeling an antibody with 111In requires the conjugation of the cationized antibody with a radiometal chelating moiety, such as diethylenetriaminepentaacetic acid (DTPA).1 The present studies report on the sequential cationization and DTPA conjugation of the 528 murine monoclonal antibody (MAb) to the human epidermal growth factor receptor (EGFR). The EGFR is overexpressed in the majority of solid cancers (13), and a radiolabeled, cationized anti-EGFR MAb is a potential new radiopharmaceutical for the imaging of cancer. In the course of the present studies, a serum inhibition phenomenon was observed in which the beneficial effects of the antibody cationization were neutralized in vivo in the circulation. Therefore, the present studies evaluated the triple conjugated 528 MAb, in which the antibody is sequentially cationized, conjugated with poly(ethylene glycol), and conjugated with DTPA, followed by labeling with 111In and in vivo pharmacokinetic analysis. EXPERIMENTAL PROCEDURES

Materials. U87 human brain glioma cells (ATCC HTB 14) and the 528 hybridoma (ATCC HB 8509) were supplied by the American Type Culture Collection (Rockville, MD). [111In]Cl3 was purchased from NEN Life Science Products Inc. (Boston, MA). [125I]Na was obtained from Amersham (Arlington Heights, IL). Diethylenetriaminepentaacetic (DTPA) dianhydride and hexamethylenediamine (HMD) were supplied by Aldrich Chemical Co., Inc. (Milwaukee, WI). NHS-PEG2000 was obtained from Shearwater Polymers (Huntsville, AL), where NHS ) N-hydroxysuccinimide and PEG2000 ) poly(ethylene glycol) of 2000 Da molecular weight. Superose 12 HR 10/ 30 FPLC columns, ampholine polyacrylamide (PAG) plates (pH ) 3.5-9.5) for isoelectric focusing (IEF), and 1 Abbreviations: MAb, monoclonal antibody; IEF isoelectric focusing; pI, isoelectric point; ID, injected dose; DTPA, diethylenetriaminepentaacetic acid; NHS, N-hydroxysuccinimide; PEG, poly(ethylene glycol); EGFR, epidermal growth factor receptor; HMD, hexamethylenediamine; EDC, N-ethyl-N′-(3(dimethylamino)propyl)carbodiimide; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; SCID, severe combined immunodeficiency; NHS, N-hydroxysuccinimide; Bmax, maximal receptor binding; KD, receptor binding dissociation constant; KI, receptor binding inhibitory constant; NSB, nonspecific binding; AUC, area under the plasma concentration curve; FPLC, fast protein liquid chromatography; IRMA, immunoradiometric assay; mIgG, mouse immunoglobulin G.; CMC, carboxymethyl cellulose.

Bioconjugate Chem., Vol. 14, No. 3, 2003 547

Broad pI standard kit (pH ) 3-10) were purchased from Pharmacia Biotech (Piscataway, NJ). Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) supplies were obtained from Bio-Rad Inc. (Richmond, CA). Chloramine-T was purchased from MCB Reagents (Cincinnati, OH). N-Ethyl-N′-(3-(dimethylamino)propyl)carbodiimide (EDC) and all other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO). Female BALB/c mice and female severe combined immunodeficient (SCID) mice were supplied by Harlan/Sprague Dawley (San Diego, CA). Antibody Cationization and Pegylation. The 528 MAb was cationized with 2 different reaction conditions designed to give a protein with increasing degrees of cationization. In the first reaction, 3.2 mL of 2 M HMD was added to 1.0 mL of native 528 MAb (8.3 mg/mL in 0.01 M PBS/7.4) and the pH was adjusted to 6.8. After mixing, 44 mg of fresh EDC was added to the mixture, and the pH was readjusted to 6.8. In the second reaction, the pH was changed to 7.4, and the amounts of HMD and EDC were reduced to 1.6 mL of 2 M HMD and 21 mg EDC, respectively. The mixtures were gently rocked end over end for 3 h at room temperature, and the reaction was stopped by the addition of 1 mL of 1 M glycine, followed by incubation for 30 min at room temperature. The mixtures were dialyzed against 10 Ls of 0.01 M PBS/7.4 overnight at 4 °C using a 12 kDa MW cut off dialysis tubing. The cationized 528 MAbs were stored at -20 °C. The cationized 528 MAb prepared at pH 7.4 was pegylated with a 100-fold or 1000-fold molar excess of NHS-PEG2000 as follows: 0.1 mL of 0.5 M NaHCO3/pH 8.5 was added to 0.9 mL of the cationized 528 MAb (1.1 mg/mL). Either 1.2 mg or 12 mg of NHS-PEG2000 was directly added to the mixture. The mixture was gently rocked for 60 min at room temperature, and the reaction was quenched by the addition of 50 µL of 1 M glycine, followed by incubation for 60 min at room temperature. The pI of the native or cationized 528 MAbs was determined with polyacrylamide gel IEF. Approximately 10-20 µg of either native 528 MAb, cationized 528 MAb/ pH ) 6.8 or cationized 528 MAb/pH ) 7.4, was solubilized in 2% NP-40 and applied to the IEF gel in parallel with pI standards. IEF was performed as described previously (8), followed by staining of the gel with Coomassie blue. Antibody Radioiodination or DTPA Conjugation and Indium Chelation. The native 528 MAb (83 µg, 0.55 nmol) was iodinated with [125I]Na (1 mCi, 0.53 nmol) and chloramine T (8.4 nmol) in 0.05 M PBW/pH 7.4 at room temperature for 2 min. After the addition of sodium metabisulfite (12.5 nmol) to quench the reaction, [125I]native 528 MAb was purified by Sephadex G25 gel filtration chromatography (0.7 × 28 cm column) with an elution buffer of 0.01 M PBS/pH 7.4 containing 0.1% BSA (PBSB). The final specific activity and the TCA precipitability of [125I]-native 528 MAb were 4.9 µCi/µg and > 99%, respectively. The [125I]-native 528 MAb was used in the immunoradiometric assay described below. For [111In] labeling of the native 528 MAb, DTPA dianhydride (6.72 nmol) was added in a 12:1 molar ratio to 83 µg of the native 528 MAb (0.55 nmol) in 0.1 M NaHCO3/pH 8.3, followed by incubation at room temperature for 90 min. The DTPA-native 528 MAb was radiolabeled with 0.75 mCi of [111In]Cl3 in 0.05 M Na acetate/0.15 M NaCl (ABS)/pH 5.5, followed by incubation at room temperature for 30 min. [111In]-native 528 MAb was separated from free [111In] by Sephadex G25 gel filtration chromatography (0.7 × 28 cm column) in 0.01

548 Bioconjugate Chem., Vol. 14, No. 3, 2003

M PBS/pH ) 7.4. [111In]-native 528 MAb had a specific activity of 6.4 µCi/µg. For [111In] labeling of the cationized 528 MAb, DTPA dianhydride (21 nmol) was added in a 30:1 molar ratio to 105 µg of the cationized 528 MAb (0.7 nmol) in 0.1 M NaHCO3/pH 8.3, followed by incubation at room temperature for 90 min. The 528 MAb cationized at pH ) 7.4 was used for DTPA conjugation. The DTPA-cationized 528 MAb was radiolabeled with 0.75 mCi [111In]Cl3 in 0.05 M ABS/pH 5.5, followed by incubation at room temperature for 30 min. [111In]-cationized 528 MAb was purified by Sephadex G25 size-exclusion chromatography (0.7 × 28 cm column) in 0.05 M ABS/pH 5.5 containing 0.05% Tween 20 (ABST). The final specific activity of [111In]-cationized 528 MAb was 6.5 µCi/µg. For [111In] labeling of the PEG2000-cationized 528 MAb, DTPA dianhydride (50.4 nmol) was added in a 100:1 molar ratio to 100 µg of the PEG2000-cationized 528 MAb (0.5 nmol) in 0.1 M NaHCO3/pH 8.3, followed by incubation at room temperature for 90 min. The PEG2000-DTPAcationized 528 MAb was radiolabeled with 1 mCi [111In]Cl3 in 0.05 M ABS/pH 5.5, followed by incubation at room temperature for 30 min. PEG2000-[111In]-cationized 528 MAb was separated from free [111In] by Sephadex G25 gel filtration chromatography (0.7 × 28 cm column) with an elution buffer of 0.01 M PBS/pH 7.4 containing 0.05% Tween 20 (PBST). The PEG2000-[111In]-cationized 528 MAb had a specific activity of 9.3 µCi/µg. Immunoradiometric Assay (IRMA). U87 human brain glioma cells (3 × 104 cells/well) were plated on sterile 96-well plate and grown to confluency. These cells express the EGFR on the plasma membrane in cell culture (14). The cells were fixed by ice-cold methanol at -20 °C for 10 min. Each well (100 µL) contained 0.04 µCi of [125I]-native 528 MAb and 1-300 µg/mL of either unlabeled native 528 MAb, cationized 528 MAb (pI ) 8.7 or pI ) 9.2), native mouse IgG (mIgG), or cationized mIgG in PBSB. After incubation for 3 h at room temperature, the wells were washed three times with 180 µL of cold 0.01 M PBS/pH 7.4. Then, 100 µL of 1 N NaOH was added to each well to solubilize the cells at 60 °C for 30 min, followed by rocking at room temperature for 30 min. The solubilized cells (80 µL) from each well were counted for [125I] radioactivity using a γ-counter (Beckman Instruments, Inc., Fullerton, CA). The binding constants for native or cationized 528 MAb binding to the human EGFR expressed on the fixed U87 cells were determined by fitting the binding data to equations derived for a single saturable binding site plus a nonsaturable binding site. The saturation of [125I]native 528 MAb binding by unlabeled native 528 was fit to a Scatchard plot:

B/F ) [(Bmax)/(KD + F)] + NSB were B ) bound MAb, F ) free MAb, Bmax ) maximal binding at the saturable site, KD ) the saturable binding site dissociation constant, and NSB ) nonsaturable binding. The saturation of [125I]-native 528 MAb binding by unlabeled cationized 528 MAb was fit to a modified Scatchard plot:

B/F ) [(Bmax/KD)/(1 + I/KI)] + NSB where I ) the concentration of cationized 528 MAb, and KI ) the concentration of 50% inhibition of native 528 MAb binding. All binding data were fit with a weighted least squares regression analysis described below.

Lee and Pardridge

Pharmacokinetics. The BALB/c mice were anesthetized with ketamine (100 mg/kg) and xylazine (2 mg/kg) intraperitoneally. Either 5 µCi of [111In]-native 528 MAb in 0.01 M PBS/pH 7.4, 6.7 µCi of [111In]-cationized 528 MAb in 0.05 M ABST/pH 7.4, or 6 µCi of PEG2000-[111In]cationized 528 MAb in 0.01 M PBST/pH 7.4 was injected into the jugular vein of each mouse. Arterial blood was removed from the aorta at 0.25, 2, 5, 15, and 60 min after intravenous injection of the isotope solution. The blood samples were centrifuged and serum was counted for [111In]-radioactivity using a γ-counter. At 60 min after isotope administration, the mice were sacrificed and liver, kidney, heart, and lung were removed for counting of [111In]-radioactivity. Pharmacokinetic parameters were estimated by fitting the serum concentration-time data to a monoexponential equation with a derivative-free nonlinear regression analysis (PARBMDP, Biomedical Computer P-Series, developed at UCLA Health Science Computing Facilities), using a weighting factor of 1/(concentration)2, where A(t), % ID (injected dose)/mL serum, and k, min-1, are the intercept and slope, respectively, of the monoexponential equation. The area under the serum concentration-time curve at 60 min [AUC(t)] and at steady state [AUCss] were calculated from the injected dose (ID), the slope, and the intercept of the monoexponential equation. The AUCss is considered only an approximation because the pharmacokinetics analysis was extended to only 60 min. The organ uptake of each compound was determined as follows:

% ID/g organ ) [Vd - V0]Cp(60 min) where Cp(60 min) is the terminal serum concentration (%ID/µL), Vd is the organ volume of distribution (µL/g), and V0 is the organ plasma volume, which was reported previously for mice (15). Gel Filtration Fast Protein Liquid Chromatography (FPLC). The metabolic stability of [111In]-native 528 MAb, [111In]-cationized 528 MAb, and PEG2000-[111In]-cationized 528 MAb was examined by injecting the pooled serum taken from 3 mice at 60 min on to a Superose 12 HR 10/30 FPLC column with elution at 0.5 mL/min in PBST for 60 min. The fractions (0.5 mL per each tube) were counted for [111In]-radioactivity using a γ-counter. External Detection Imaging and Whole Body Autoradiography of Flank Tumors. U87 human glioma cells (106 cells/mouse) were injected into right flank of female scid mice subcutaneously. U87 experimental tumors overexpress the human EGFR on the tumor cell membrane (14). Tumor sizes were measured twice a week and tumor volume was calculated by following equation:

tumor volume (mm3) ) (4/3)πab2 where a ) long radius and b ) short radius. Tumor sizes ranged from 524 to 1600 mm2 at 2 weeks after the injection of U87 cells. The tumors of the scid mice were imaged at 15 days after implantation. The scid mice with flank tumors were divided into two groups and 50 µCi of either [111In]-native 528 MAb or [111In]-cationized 528 MAb was injected into the tail vein of each animal. At 24 h after the isotope injection, the mice were anesthetized with ketamine (100 mg/kg) and xylazine (2 mg/kg). The mice were placed on the plate in the prone position and were scanned with a stationary LFVO camera (Orbiter ZLC-7500, Siemens) at 247 keV or 172



Figure 1. Reaction schemes for reformulation of 528 MAb. The native MAb with surface carboxyl groups is shown as I. The native 528 MAb is cationized with hexamethylenediamine, NH2(CH2)6NH2, and EDC (also called carbodiimide) to form the cationized MAb, shown as II, and the surface carboxyl groups have now been converted to extended primary amine groups. The latter are pegylated with NHS-PEG2000 to form the cationized, pegylated MAb as shown in III. Either II or III are reacted with DTPA dianhydride as shown in IV and V, respectively.

Figure 2. Isoelectric focusing of the native 528 MAb (lane 1) and the cationized 528 MAb (lanes 2 and 3); pI standards are shown in lane 4. The 528 MAb was cationized at either pH 7.4 or 6.8 to form the cationized MAb with a pI of 8.7 (lane 2) or 9.2 (lane 3), respectively. The gel was stained with Coomasie blue.

keV with 15% window width for 10 min. The scan data were saved on a Siemens ICON computer prior to generation of X-ray film images. After γ-counter external detection, the mice were euthanized and placed on a Styrofoam board in the prone position and immersed in carboxymethyl cellulose (CMC) and dipped in liquid N2 for 10 min and stored at -20 °C overnight. The frozen CMC-mouse specimen was placed in a cryostat and 50 µm thick sections were obtained. The frozen sections were placed in a Fuji-BAS casette and exposed at -20 °C for 24 h. The plate was analyzed with a Fuji-BAS reader, saved as a MacBAS file with 100 micron resolution and 65 bit color, and converted to a pict file in Adobe Photoshop. RESULTS

The native 528 MAb is shown as intermediate I in Figure 1 and the cationized 528 MAb is shown as II in

Figure 3. Sodium dodecyl sulfate polyacrylamide gel electrophoresis with Coomasie blue staining of molecular weight standards (lanes 1 and 6), native 528 MAb (lane 2), cationized 528 MAb/pI ) 8.7 (lane 3), or PEG-cationized 528 MAb (lanes 4 and 5). The cationized 528 MAb pegylated with a 100-fold or 1000-fold molar excess of NHS-PEG2000:antibody is shown in lanes 4 and 5, respectively.

Figure 1. The 528 MAb that undergoes sequential cationization and DTPA conjugation followed by chelation with 111In is shown as IV in Figure 1. The 528 MAb was cationized at two different reaction pH’s and the pI of the MAb following cationization at pH 7.4 is 8.7 and the pI of the antibody following cationization at pH 6.8 is 9.2 as shown by the IEF gel in Figure 2. The pI of the native 528 MAb ranges from 8.1 to 8.3 (lane 1, Figure 2). There was minimal formation of intramolecular bonds between antibody light or heavy chains following cationization as demonstrated by SDS-PAGE (Figure 3). The size of the light chain and heavy chain of the cationized 528 MAb (lane 3, Figure 3) is identical to the size of the light and heavy chains of the native 528 MAb (lane 2, Figure 3). The affinity of the native or cationized 528 MAb for the target human EGFR was determined with the immunoradiometric assay. The binding of [125I] native 528 MAb to human U87 glioma cells was inhibited by increasing concentrations of unlabeled native 528 MAb


Lee and Pardridge

Figure 4. Immunoradiometric assay (IRMA) with methanol-fixed human U87 glioma cells as the solid-phase source of the human EGFR and [125I]-native 528 MAb as the tracer ligand. The competitive displacement of binding of the [125I]-native 528 MAb to the U87 cells was measured in the presence of 1-30 µg/mL concentrations of native 528 MAb, cationized 528 MAb/pI ) 9.2, cationized 528 MAb/pI ) 8.7, native mouse IgG (mIgG), and cationized mIgG. Data are mean ( SE (n ) 3). The affinity of the PEG2000cationized 528 MAb for the EGFR was not measured.

Figure 5. Whole body autoradiography [left] or gamma counter external detection [right] of female scid mice implanted with U87 subcutaneous tumors in the right flank (arrows). Mice were injected with 50 µCi of either [111In]-native 528 MAb or [111In]-cationized 528 MAb. At 24 h after injection, the animals were anesthetized for external detection and then sacrificed for whole body autoradiography. The tumors were imaged 15 days after implantation. Table 1. EGF Receptor Binding Constantsa monoclonal antibody

KD (nM)

KI (nM)

Bmax NSB (fmol/dish) (fmol/dish)

native 528 1.4 ( 0.5 21.0 ( 4.2 cationized 528/pI ) 8.7 5.6 ( 3.6 cationized 528/pI ) 9.2 27 ( 6 -

1.7 ( 0.2 0.8 ( 0.4 2.4 ( 0.4

a Parameters determined by nonlinear regression analysis of immunoradiometric assay (IRMA) shown in Figure 4. Bmax ) maximal binding; NSB ) nonspecific binding.

but not by increasing concentrations of native or cationized preimmune mouse IgG (mIgG), shown in Figure 4. The binding curve for the native 528 MAb was analyzed by nonlinear regression analysis, which indicated the binding KD was 1.4 ( 0.5 nM for the native 528 MAb (Table 1). Increasing concentrations of both the cationized 528 MAb/pI ) 8.7 and the cationized 528 MAb/pI ) 9.2 decreased binding of [125I] native 528 MAb to the U87 cells (Figure 4). Nonlinear regression analysis demonstrated the KI of the cationized 528/pI ) 8.7 and the cationized 528/pI ) 9.2 was 5.6 ( 2.6 nM and 27 ( 6 nM, respectively (Table 1). These findings indicated the affinity of the cationized 528/pI ) 8.7 was comparable to that of the native 528 MAb, but that the affinity of the

cationized 528/pI ) 9.2 was too low owing to the higher degree of cationization. Therefore, all subsequent work was performed with the cationized 528/pI ) 8.7 MAb, which was cationized at pH ) 7.4 (Methods). The native or cationized 528 MAb was conjugated with DTPA anhydride (Figure 1) and radiolabeled with 111In and injected into tumor bearing SCID mice with unilateral U87 glioma flank tumors. The flank tumors could be detected with either the native or the cationized 528 MAb using either an external detection gamma camera or whole body autoradiography (Figure 5). Although there was increased hepatic uptake with the cationized 528 MAb, the uptake of the native or cationized 528 MAb over the tumor was comparable (Figure 5). The lack of an increased tumor signal with the cationized 528 suggested a possible serum inhibition phenomenon that neutralized the effects of the cationization modification of the antibody. This was confirmed with a series of pharmacokinetic studies in BALB/c mice. The [111In] native 528 MAb or the [111In] cationized 528 MAb was injected intravenously into mice, and serum radioactivity was measured over a 60 min period (Figure 6). The serum data were analyzed by nonlinear regres-



Table 2. Pharmacokinetic Parametersa parameter

[111In]-native 528 MAb

[111In]-cationized 528 MAb

[111In]-PEG2000-cationized 528 MAb

AUC (60 min) (%ID‚min/mL) AUCss (%ID‚min/mL) liver uptake (%ID/g) kidney uptake (%ID/g) heart uptake (%ID/g) lung uptake (%ID/g)

4180 ( 210 25887 ( 3380 11.0 ( 0.5 0 0 2.5 ( 1.6

3976 ( 192 17770 ( 2301 18.6 ( 0.8 4.8 ( 0.4 0 4.9 ( 0.5

4581 ( 45 43447 ( 9864 0 0 0 3.5 ( 0.5

a AUC was calculated from data in Figure 6; organ %ID/gram are mean ( SE (n ) 3 mice per group). An organ uptake (%ID/gram) of 0 indicates the organ volume of distribution was no greater than the organ blood volume in mice (15).

Figure 6. Serum concentration of [111In]-native 528 MAb, [111In]-cationized 528 MAb, and PEG2000-[111In]-cationized 528 MAb in mice at 0.25-60 min after a single intravenous injection. Data are mean ( SE (n ) 3). The PEG2000-[111In]-cationized 528 MAb used in this study was prepared with a 1000:1 molar ratio of PEG:antibody (Methods).

sion analysis to yield the pharmacokinetic parameters shown in Table 2. Although the steady-state area under the plasma concentration curve (AUCss) was reduced 31% following cationization, there was no difference in the 60 min AUC for cationized antibody as compared to the native antibody (Table 2). Cationization caused minor increases in the uptake of the MAb by organs such as liver, kidney, or lung, but had no effect on uptake by the heart. The 60 min serum from mice injected with either native or cationized 528 MAb was analyzed by gel filtration fast protein liquid chromatography (FPLC) as shown in Figure 7A and 7B, respectively. The data showed that the cationized 528 MAb migrated at the same elution volume as the native 528 MAb. In an attempt to retard the serum inhibition of the cationized 528 MAb, this antibody was triple conjugated with sequential cationization, pegylation, and DTPA conjugation as outlined in Figure 1. The cationized 528 MAb was pegylated at two different molar ratios of PEG2000:antibody. Pegylation of the cationized 528 MAb at 100:1 molar ratio resulted in only minor increases in the size of the heavy or light chain of the antibody. However, with the 1000:1 molar ratio of PEG2000: antibody, the molecular weight of the cationized MAb was increased from 55K to approximately 110K (lane 5, Figure 3). Similarly, the molecular weight of the light chain was increased from 25K to approximately 55K (lane 5, Figure 3). The 528 MAb that was sequentially cationized, pegylated, and DTPA conjugated was radiolabeled with 111In and injected into control mice for a pharmacokinetic analysis. The serum concentration profile over

Figure 7. Superose 12HR gel filtration fast protein liquid chromatography (FPLC) of pooled serum taken from 3 female BALB/c mice at 60 min after intravenous injection of [111In]native 528 MAb (A), [111In]-cationized 528 MAb (B), and PEG2000[111In]-cationized 528 MAb (C).

60 min is shown in Figure 6, and the pharmacokinetic parameters are shown in Table 2. Pegylation of the cationized 528 MAb resulted in a 10% increase in the plasma AUC at 60 min and a 68% increase in the steadystate plasma AUC (Table 2). Pegylation of the cationized 528 eliminated uptake of the antibody by liver or kidney and inhibited lung uptake of the cationized antibody (Table 2). DISCUSSION

These studies are consistent with the following conclusions. First, cationization of the murine 528 MAb to the human EGFR results in increased pI (Figure 2), stable heavy chain and light chain structure on SDS-PAGE (Figure 3), and high affinity binding to the EGFR in human U87 cells (Figure 4). Second, imaging EGFR expression in flank tumors in SCID mice in vivo does not result in increased tumor signal using the [111In]-cationized 528 MAb, as compared to the [111In] native 528 MAb (Figure 5). This correlates with a minimal increase in plasma clearance of the 528 MAb following cationization and radiolabeling with 111In (Figure 6, Table 2). Third, the clearance of the [111In]-cationized 528 MAb is further


slowed by surface pegylation of the cationized MAb followed by radiolabeling with 111In (Figure 6, Table 2). The mechanism of antibody cationization used in these studies is the conversion of surface carboxyl groups on aspartic acid and glutamic acid residues (intermediate I, Figure 1) to extended primary amino groups using hexamethylenediamine (intermediate II, Figure 1). The degree of cationization is enhanced with an increased molar ratio of hexamethylenediamine to antibody or with a decreased reactant pH (Figure 2). When the 528 MAb is cationized at pH ) 7.4, there is minimal change in affinity of the cationized antibody for the target EGFR (Figure 4, Table 1). However, if the 528 MAb is cationized at pH 6.8, there is a 5-fold loss of affinity of the antibody for the target antigen (Figure 4, Table 1). By determining the optimal conditions of the cationization reaction, it is possible to cationize the monoclonal antibody without altering the structure of the heavy or light chains (Figure 3) and without substantially inhibiting the affinity of the antibody for the target antigen (Figure 4, Table 1). Cationized monoclonal antibodies have been radiolabeled in previous work with 125-iodine. There is a profound increase in plasma clearance of the [125I]antibody following cationization. Even if the native MAb has a net cationic charge, the cationization of the MAb has a marked effect on plasma clearance in vivo. For example, the native D146 MAb to the ras oncogene has an alkaline pI of 8.9, but this MAb is cleared slowly from plasma (8). Following cationization, the pI was raised to 9.5, and the plasma clearance in BALB/c mice was increased 58-fold compared to the plasma clearance of the [125I]-native D146 MAb (8). This change in pI of the native and cationed D146 MAb is comparable to that observed in the present study with the 528 MAb (Figure 2). The plasma clearance in scid mice of human immunoglobulins was increased >1000-fold following cationization and radiolabeling with 125-iodine (16). The plasma clearance in rats of the humanized HER2 antibody was increased 22-fold following cationization and radiolabeling with 125-iodine (9). The plasma clearance of the AMY33 murine monoclonal antibody to the Aβ amyloid peptide was increased 29-fold in BALB/c mice following cationization and radiolabeling with 125-iodine (7). However, the plasma clearance of the AMY33 antibody in mice was increased only 4-fold following cationization and radiolabeling with 111-indium (7). In the dog, the plasma clearance of the [111In]-cationized AMY33 MAb was actually slower than the plasma clearance of the [111In]-native AMY33 MAb (7). The findings of the present study corroborate previous work in mice and dogs with AMY33 MAb and indicate that radiolabeling of a cationized MAb with 111-indium eliminates the ability of the antibody to rapidly cross microvascular barriers and undergo rapid exodus from the blood compartment. Prior work with isolated microvessels and in vitro binding assays with or without serum showed that the inhibition of plasma clearance of the [111In]-cationized MAb could be traced to inhibition by factors in serum (7). The serum inhibitory factors are low molecular weight since the elution of the cationized MAb on a gel filtration FPLC column is identical to that of the native MAb labeled with 111In (Figure 7B). Pegylation of proteins inhibits the interaction of the protein with plasma constituents and thereby delays plasma clearance of the protein (17). The PEG moieties may be attached to either surface amino or carboxyl groups. In the present studies, the PEG was conjugated to amino groups on the cationized MAb (Figure 1), which would be expected to partially reverse the cationization,

Lee and Pardridge

which should reduce the plasma clearance of the cationized MAb. However, we hypothesized that pegylation of a cationized MAb may block the serum inhibition of the DTPA conjugated-cationized MAb and actually result in an increase in plasma clearance of the triply conjugated antibody following cationization, pegylation, and DTPA conjugation. The pharmacokinetic studies showed that pegylation of the cationized antibody that was subsequently DTPA conjugated was actually slower than the plasma clearance of the nonpegylated cationized antibody (Figure 6, Table 2). The cationized 528 MAb was effectively pegylated based on SDS-PAGE analysis (Figure 3). Pegylation at a 100:1 molar ratio of PEG:antibody resulted only in minor pegylation of the protein (lane 4, Figure 3). However, pegylation at 1000:1 molar ratio of PEG:antibody resulted in an increase in the size of the heavy chain from approximately 55 kDa to approximately 110 kDa (lane 5, Figure 3). This suggests that approximately 20-25 PEG2000 moieties were attached to each heavy chain. The size of the light chain was increased from approximately 25 kDa to approximately 50-55 kDa (lane 5, Figure 3) indicating approximately 12-15 PEG2000 moieties were conjugated to each light chain. However, pegylation of the cationized 528 antibody did not block the serum inhibition and actually caused a further slowing of the clearance of the antibody from the blood (Figure 6, Table 2). In summary, these studies describe the reformulation of the 528 murine MAb to the human EGFR that involves sequential cationization, pegylation, and DTPA conjugation (Figure 1). Similar to previous studies with the AMY33 MAb to the amyloid peptide (7), the combined cationization and DTPA conjugation followed by radiolabeling with 111-indium resulted in an abrogation of the beneficial effects of cationization on plasma clearance of the MAb. The exodus of an antibody following cationization and radiolabeling with 125-iodine is increased from 20-fold to >1000-fold in mice and rats. However, the increased rate of exodus from blood of a cationized antibody is not observed when the antibody is radiolabeled with DTPA conjugation and 111In chelation, as described in these studies. Cationization of an MAb is a simple approach to antibody reformulation that causes rapid removal of the antibody from the blood compartment in vivo. This rapid clearance can result in increased signal and decreased noise in a tumor imaging modality in vivo. However, future work will be necessary to find the optimal method of radiolabeling the cationized monoclonal antibody that avoids the serum inhibition observed with DTPA conjugated cationized antibodies. ACKNOWLEDGMENT

Supported by a grant 3IB-006 from the California Breast Cancer Research Program and by the U.S. Department of Energy. Dr. Sanjiv Gambhir’s laboratory assisted with the whole body autoradiography. LITERATURE CITED (1) Vriesendorp, H. M., and Quadri, S. M. (2001) Radiolabeled immunoglobulin therapy: Old barriers and new opportunities. Expert Rev. Anticancer Ther. 1, 461-478. (2) Triguero, D., Buciak, J. B., Yang, J., and Pardridge, W. M. (1989) Blood-brain barrier transport of cationized immunoglobulin G: Enhanced delivery compared to native protein. Proc. Natl. Acad. Sci. 86, 4761-4765. (3) Basu, S. K., Goldstein, J. L., Anderson, R. G. W., and Brown, M. S., (1976) Degradation of cationized low-density lipoprotein and regulation of cholesterol metabolism in homozygous

Antibody Radiopharmaceuticals familial hypercholesterolemia fibroblasts. Proc. Natl. Acad. Sci. U.S.A. 73, 1378-3182. (4) Shen, W.-C. and Ryser, H. J.-P. (1978) Conjugation of polyL-lysine to albumin and horseradish peroxidase: A novel method of enhancing the cellular uptake of proteins. Proc. Natl. Acad. Sci. U.S.A. 75, 1872-1876. (5) Vorbrodt, A. W. (1989) Ultracytochemical characterization of anionic sites in the wall of brain capillaries. J. Neurocytol. 18, 359-368. (6) Riabowol, K. T., Vosatka, R. J., Ziff, E. B., Lamb, N. J., and Feramisco, J. R. (1998) Microinjection of fos-specific antiboides blocks DNA synthesis in fibroblast cells. Mol. Cell. Biol. 8, 1670-1676. (7) Bickel, U., Lee, V. M. Y., and Pardridge, W. M. (1995) Pharmacokinetic differences between 111In- and 125I-labeled cationized monoclonal antibody against β-amyloid in mouse and dog. Drug Deliv. 2, 128-135. (8) Pardridge, W. M., Kang, Y.-S., Yang, J., and Buciak, J. L. (1995) Enhanced cellular uptake and in vivo biodistribution of a monoclonal antibody following cationization. J. Pharm. Sci. 84, 943-948. (9) Pardridge, W. M., Buciak, J., Yang, J., and Wu, D. (1998) Enhanced endocytosis in cultured human breast carcinoma cells and in vivo biodistribution in rats of a humanized monoclonal antibody following cationization of the protein. J. Pharmacol. Exp. Ther. 286, 548-554. (10) Pardridge, W. M., Bickel, U., Buciak, J., Yang, J., Diagne, A., and Aepinus, C. (1994) Cationization of a monoclonal antibody to the human immunodeficiency virus rev protein

Bioconjugate Chem., Vol. 14, No. 3, 2003 553 enhances cellular uptake but does not impair antigen binding of the antibody. Immunol. Lett. 42, 191-195. (11) Hong, G., Bazin-Redureau, M. I., and Scherrmann, J. M. G. (1999) Pharmacokinetics and organ distribution of cationized colchicine-specific IgG and Fab fragments in rat. J. Pharm. Sci. 88, 147-153. (12) Kurihara, A., Deguchi, Y., and Pardridge, W. M. (1999) Epidermal growth factor radiopharmaceuticals: 111In chelation, conjugation to a blood-brain barrier delivery vector via a biotin-polyethylene linker, pharmacokinetics, and in vivo imaging of experimental brain tumors. Bioconjugate Chem. 10, 502-511. (13) Nicholson, R. I., Gee, J. M. W., and Harper, M. E. (2001) EGFR and cancer prognosis. Eur. J. Cancer 37, S9-S15. (14) Kurihara, A. and Pardridge, W. M. (1999) Imaging brain tumors by targeting peptide radiopharmaceuticals through the blood-brain barrier. Cancer Res. 54, 6159-6163. (15) Skarlatos, S. and Pardridge, W. M. (1995) Targeting of an anti-CR3 (CD11b/CD18) monoclonal antibody to spleen, but not brain, in vivo in mice. J. Drug Targeting 3, 9-14. (16) Pardridge, W. M., Kang, Y. S., Diagne, A., and Zack, J. A. (1996) Cationized hyperimmune immunoglobulins: Pharmacokinetics, toxicity evaluation, and treatment of HIV-infected scid-hu mice. J. Pharmacol. Exp. Ther. 276, 246-252. (17) Nucci, M. L., Shorr, R., and Abuchowski, A. (1991) The therapeutic value of poly(ethylene glycol)-modified proteins. Adv. Drug Deliv. Rev. 6, 133-151.

BC0256648

Monoclonal Antibody Radiopharmaceuticals - American Chemical

Recommend Documents