Conditionally Cleavable Radioimmunoconjugates - American

Aug 19, 2003 - Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington 98109,. Departments of Chemistry and Pediatrics...
0 downloads 0 Views 113KB Size
Download PDF
Bioconjugate Chem. 2003, 14, 927−933

927

Conditionally Cleavable Radioimmunoconjugates: A Novel Approach for the Release of Radioisotopes from Radioimmunoconjugates Craig Beeson,‡,⊥ James E. Butrynski,† Michael J. Hart,‡ Cynthia Nourigat,† Dana C. Matthews,†,§ Oliver W. Press,† Peter D. Senter,¶ and Irwin D. Bernstein*,†,§ Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington 98109, Departments of Chemistry and Pediatrics, University of Washington, Seattle, Washington 98195, and Seattle Genetics, 21823 30th Dr. SE, Bothell, Washington 98021. Received December 14, 2002; Revised Manuscript Received July 1, 2003

One of the limitations of therapy with radiolabeled monoclonal antibodies (mAbs) is that significant toxicities can arise from circulating non-tumor-bound radiolabeled conjugate. Here, we describe a new method to reduce systemic radiation exposure from radiolabeled mAbs involving the attachment of the radioisotope through a linker that can be cleaved by an administered enzyme. To demonstrate the feasibility of this approach, we prepared a conditionally cleavable radioimmunoconjugate (RIC) composed of 131I-labeled cephalosporin conjugated to Tositumomab, a mAb against the CD20 antigen. The cleavable RIC bound antigen identically to directly iodinated antibody, and in the presence of β-lactamase, about 80-85% of the radioisotope was released. In vivo studies in mice revealed that the cleavable RIC and the directly iodinated anti-CD20 antibody had similar biodistribution patterns. Systemically administered β-lactamase induced a 2-3-fold decrease in the percent injected dose (ID) of the cleavable RIC/g of blood, marrow, spleen, lung, and liver 1 h after enzyme treatment, and a 4-6-fold decrease 20 h after enzyme treatment. This was accompanied by a 20-fold increase in % ID/g in urine 1 h after enzyme treatment, indicating that the released radiolabel was rapidly excreted through the kidneys. In mice with human tumor xenografts, there was no decrease in the %ID/g in tumor 1 h after enzyme treatment, but by 4 h after enzyme injection, decreases in tumor radioactive content began to diminish the targeting advantage. These studies demonstrate that the cleavable RIC substrate is able to bind to tumor antigens and localize within human tumor xenografts and that accelerated systemic clearance can be induced with β-lactamase.

INTRODUCTION

Several studies have demonstrated that radioimmunotherapy (RIT) using radiolabeled monoclonal antibodies can lead to dose-limiting toxicities through radiation exposure to normal organs. To minimize toxicity, a number of approaches have been devised to clear radiolabeled material from the systemic circulation. The most widely investigated method is commonly referred to as “pretargeting”, which involves the infusion of nonradioactive mAb conjugated with a tag that can be recognized by a radiolabeled binding partner. Biotin is the most commonly used tag, and it is recognized by radiolabeled avidin or streptavidin (1, 2). Alternatively, mAb-streptavidin conjugates can be targeted, which are then recognized by subsequently administered radiolabeled biotin* To whom correspondence should be addressed. Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. North,,P.O. Box 19024, D2-373, Seattle, WA 98109. Telephone: 206-6674886. Fax: 206-667-6084. E-mail: [email protected]. † Fred Hutchinson Cancer Research Center. ‡ Department of Chemistry, University of Washington. § Department of Pediatrics, University of Washington. ¶ Seattle Genetics. ⊥ Current address: Department of Pharmaceutical Sciences, Medical University of South Carolina, Charleston, SC 29425. 1 Abbreviations: ADEPT, antibody-directed enzyme prodrug therapy; ID, injected dose; ECIA, extracorporeal immunoadsorption; mAb, monoclonal antibody; RIC, radioimmunoconjugate; RIT, radioimmunotherapy; SPE, solid-phase extraction.

containing compounds (3, 4). Although these approaches have yielded substantially improved tumor to normal tissue ratios of delivered radiation compared to conventional RIT, issues have arisen due to treatment complexity, endogenous biotin, immunogenicity of avidin and streptavidin, and the prolonged time that the mAb conjugate must be retained on the tumor cell surface before the radiolabel is administered. An alternative approach for minimizing normal tissue radiation exposure involves plasmapheresis and extracorporeal immunoadsorption (ECIA). Using this method, blood is circulated ex vivo through a matrix that is capable of sequestering the radiolabeled agent and is then reinfused back into the body. Pharmacokinetic modeling studies have suggested a potential therapeutic advantage for this approach, particularly if high antibody doses and an isotope with a relatively long half-life such as 131I is used (5-7). Experimentally, it has been demonstrated that the ECIA procedure removed much of the blood radioactivity due to radiolabeled mAb, and that many organs were spared from radiation exposure (8). However, ECIA is a technically difficult procedure and entails significant radiation safety risks for technicians performing the procedure. A related method for removal of circulating radiolabeled mAbs that bypasses the need for ECIA involves the formation of immune complexes with a second-step clearing mAb (5, 8-11). While some success for this approach has been reported in preclinical

10.1021/bc025655z CCC: $25.00 © 2003 American Chemical Society Published on Web 08/19/2003

928 Bioconjugate Chem., Vol. 14, No. 5, 2003

models, immune complexes may pose significant clinical risks including induction of nephropathies or arthropathies. An alternative method for reducing normal organ radiation exposure from RIT has been reported, and it involves the use of linkers between the mAb carrier and the drug that can be hydrolyzed by endogenous enzymes in the liver. The rationale for this approach is that the liver, being the major organ for mAb and RIT clearance pathways, would be spared considerable radiation damage by rapid cleavage of the isotope off the mAb. This has been demonstrated using a linker that was hydrolyzed by cathepsin D, an abundant liver enzyme (1215). The beneficial effects of this radioisotope linkage strategy were mainly restricted to the liver. We wished to extend this methodology by developing radioisotope linkers that could be cleaved in a controlled manner in the blood and in any well-perfused organs. Here, we describe the synthesis of a cephalosporin-containing linker for the attachment of radioisotopes to mAbs. The approach was designed so that the radiolabel would rapidly be released from the mAb upon the systemic administration of β-lactamase, where it would then be rapidly excreted through the kidneys. We also report the in vitro and in vivo properties of the first such radioimmunoconjugate. MATERIALS AND METHODS

General. The human Ramos B lymphoma cell line (American Type Culture Collection, Bethesda, MD) was maintained in log-phase growth in RPMI media supplemented with 10% heat-inactivated bovine calf serum in a 5% CO2 incubator. NOD/SCID mice were bred at the Fred Hutchinson Cancer Research Center. The antiCD20 murine mAb Tositumomab (IgG2a) was a gift of Coulter Corporation. The G3G6 antibody, a murine IgG2a isotype control mAb, was produced by standard hybridoma techniques. All chemical reagents were obtained from Sigma/Fluka unless otherwise noted. HPLC chromatography was done with a Waters radial compression 10 µm C18 22 mm × 30 cm column. All mass spectra were obtained on a Hewlett-Packard Quanta ion trap spectrometer using standard electrospray ionization. Synthetic procedures given below are representative of multiple preparations, and the cited yields are for the specific example. Synthesis: Diphenylmethyl N-(Boc-4-hydroxyphenylacetyl)-7-aminocephalosporanate (3). Boc-4-hydroxyphenylacetic acid (2.83 g, 11.2 mmol), dicyclohexylcarbodiimide (2.54 g, 12.3 mmol), and triethylamine (4.69 mL, 33.6 mmol) were added to a suspension of 7-aminocephalosporanic acid 2 (3.05 g, 11.2 mmol) in 75 mL of dry dichloromethane. The resulting suspension was stirred under N2 at room temperature for 18 h. The solvent was removed in vacuo, and the remaining residue was dissolved in a solution of 10 mL of dimethoxyethane and 5 mL of water at 0 °C. A solution of NaOH (4.4 mL at 19 wt %/vol, 25 mmol) was added, and the reaction was stirred at 0 °C for 20 min. A solution of diphenyldiazomethane (16) (10.9 g, 56 mmol) in 30 mL of ethyl acetate was added to the flask, and the solution was brought to pH 4.0 with aqueous HCl. The solution was stirred at room temperature for 1 h, brought to pH 4.0 again, and stirred for a further 18 h. After extraction with ethyl acetate and subsequent evaporation, a red oil (11.9 g) was obtained. HPLC chromatography (C18, acetonitrile:water with 0.1% TFA, ret. time ) 74%) gave 1.9 g of product (25% yield) as a white fluffy powder. Analytical

Beeson et al.

HPLC (C18 Toso Haas ODS-Tm, 4.6 mm × 20 cm) using acetonitrile-water gave a single peak (g95% pure) at 83% acetonitrile. Analytical HPLC (C8 Toso Haas ODSTm, 4.6 mm × 20 cm) using methanol-water gave a single peak (g95% pure) at 47% methanol. 1H NMR (CDCl3): δ (ppm) 7.44 (m, 10H, CHPh2); 7.29 (d, J ) 6 Hz, 2H, ArH), 7.13 (d, J ) 6 Hz, 2H, ArH); 6.98 (s, 1H, CHPh2); 6.78 (d, J ) 7 Hz, 1H, NH); 5.93 (dd, JNH ) 7 Hz, J7,6 ) 5 Hz, 1H, C(7)H); 4.95 (d, J6,7 ) 5 Hz, 1H, C(6)H); 4.75 (br. s, 1H, OH); 4.47, 4.03 (AB dd, JAB ) 11 Hz, 2H, SCH2); 3.63 (dd, J ) 10, 4 Hz, 2H, CH2O); 3.49 (br. s, 2H, ArCH2); 1.54 (s, 9H C(CH3)3). 13C NMR (CDCl3): δ (ppm) 174.4, 166.6, 162.2, 156.9, 141.3, 135.8, 130.4, 129.7, 129.2, 129.0, 128.6, 128.4, 127.3, 125.8, 97.5, 80.1, 61.4, 59.9, 57.6, 42.6, 27.8, 26.4. C34H34N2O8S (630.71 Da) calcd %: C 64.75, H 5.43, N 4.44, S 5.03, found %: C 64.59, H 5.21, N 4.64, S 5.02. 3-[(1,6-Diaminohexanamidoxy)methyl]-4-hydroxyphenylacetyl-5-oxo-7-aminocephalosporanic Acid (4). 1,1,1,2-Tetrachloroethyl orthochloroformate (32 µmol) was added to a solution of 3 (20 mg, 32 µmol) in dry dichloromethane at 0 °C, stirring was continued for 20 min, and then m-chloroperoxybenzoic acid (7.8 mg of 70 wt %/wt, 32 µmol) was added. The reaction solution was stirred at 0 °C for 15 min under N2 at which point 0.5 mL of 10% Na2S2O3 was added to quench any remaining m-chloroperoxybenzoic acid, followed by addition of 5 mL of CH2Cl2 and then washing with H2O. The solution was dried with saturated NaCl and evaporated in vacuo. The resulting oil was dissolved in THF (500 µL at 0 °C), and to this was added 32 µmol of mono-Boc-1,6-diaminohexane in 100 µL of THF followed by 1 equiv of diisopropylethylamine. The crude oil obtained after evaporation was dissolved in 500 µL of 1:1 TFA:anisole, and after 15 min this reaction mixture was evaporated to give an oily gum that was dissolved in DMSO for chromatography. After HPLC purification (C18, acetonitrile-water with 0.1% TFA, ret. time ) 44% acetonitrile), 4 was obtained as a white fluffy powder (5.1 mg, 25% yield). Analytical HPLC (C18 Toso Haas ODS-Tm, 4.6 mm × 20 cm) using acetonitrile-water gave a single peak (g95% pure) at 49% acetonitrile. Analytical HPLC (C8 Toso Haas ODSTm, 4.6 mm × 20 cm) using methanol-water gave a single peak (g95% pure) at 32% methanol. 1H NMR (d6DMSO): δ (ppm) 11.10 (br. s, 1H); 9.12 (br. s, 1H); 7.71 (br. s, 3H); 7.38 (d, J ) 6 Hz, 2H, ArH), 7.24 (d, J ) 6 Hz, 2H, ArH); 6.85 (t, 6 Hz, 1H, NHCH2); 6.73 (d, J ) 7 Hz, 1H, NHCH); 5.78 (dd, JNH ) 7 Hz, J7,6 ) 5 Hz, IH, C(7)H); 5.13 (d, J6,7 ) 5 Hz, 1H, C(6)H); 4.28 (dd, J ) 10, 6 Hz, 2H, CH2O); 3.44 (dd, J ) 14, 7 Hz, 2H, SCH2); 3.55 (br. s, 2H, ArCH2); 3.35 (m, 2H, CH2NH3+); 2.97 (m, 2H, NHCH2); 2.18-2.03 (m, 4H, CH2); 1.51 (m, 2H, CH2); 1.40 (m, 2H, CH2). C23H3ON4O8S (522.57 Da); MS (ES) [MH]+ calc. 523.186, found 523.191. 3-[(6-(2-Sulfo-N-hydroxysuccinimidosuberyl)1,6-diaminohexanamidoxy)methyl]-4-hydroxyphenylacetyl-5-oxo-7-aminocephalosporanic Acid (1). A solution of 5.3 mg of 4 (8.4 µmol) and 24 mg (42 µmol) of disulfo-N-succinimidyl suberate (Pierce, Chicago, IL) in 1 mL of dry DMSO containing 7.3 µL of diisopropylethylamine (42 µmol) was stirred for 1 h at room temperature. The reaction mixture was chromatographed by HPLC (C18, acetonitrile-water with 0.1% TFA, ret. time ) 48% acetonitrile) to give 1 as a white powder (6.7 mg, 92% yield). Analytical HPLC (C18 Toso Haas ODS-Tm, 4.6 mm × 20 cm) using acetonitrile-water gave a single peak (g95% pure) at 43% acetonitrile. Analytical HPLC (C8 Toso Haas ODS-Tm, 4.6 mm × 20 cm) using methanolwater gave a single peak (g95% pure) at 27% methanol.

Conditionally Cleavable Radioimmunoconjugates 1H

NMR (d6-DMSO): δ (ppm) 11.67 (s, 1H), 11.42 (br. s, 1H); 9.43 (br. s, 1H); 7.31 (d, J ) 6 Hz, 2H, ArH), 7.15 (d, J ) 6 Hz, 2H, ArH); 6.87 (t, 7 Hz, 1H, NHCH2); 6.87 (t, 7 Hz, 1H, NHCH2); 6.65 (d, J ) 6 Hz, 1H, NHCH); 5.76 (dd, JNH ) 6 Hz, J7,6 ) 5 Hz, IH, C(7)H); 5.33 (d, J6,7 ) 5 Hz, 1H, C(6)H); 4.33 (dd, J ) 9, 6 Hz, 2H, CH2O); 3.76 (dd, J ) 9, 5 Hz, 1H, CH(H)CHS); 3.58 (br. s, 2H, ArCH2); 3.49 (dd, J ) 12, 6 Hz, 2H, SCH2); 3.34 (m, 2H, CH2NH); 3.31 (m, 2H, NHCH2); 3.12 (dd, J ) 18, 9 Hz, 1H, CH(H)CHS); 2.99 (m, 2H, CH2), 2.70 (dd, J ) 18, 5 Hz, 1H, CH(H)CHS); 2.32 (br. t, J ∼ 7 hz, 2H, COCH2); 2.20-2.07 (m, 4H, CH2); 1.60-1.52 (m, 6H, CH2); 1.401.25 (m, 6H, CH2). C35H45N5O16S2 (855.89 Da) calcd %: C 49.12, H 5.30, N 8.18, S 7.49, found %: C 48.85, H 5.10, N 7.82, S 7.52. Preparation of the Standard and Cleavable Radioimmunoconjugate. Chloramine T was freshly prepared in deionized water to 60 mM. Carrier-free I-125 or I-131 in 0.1 M NaOH (NEN Life Sciences, Boston, MA) was neutralized with 5 volumes of 0.5 M phosphate/0.150 M NaCl buffer pH 7.4. Standard labeling of Tositumomab antibody with I-125 or I-131 was performed using the chloramine T method as previously published (3). Labeling of Tositumomab antibody with radiolabeled cleavable linker was done as follows: 8.6 µg (10 nmol) of cleavable linker in 10 µL of DMSO was added to 10 nmol of chloramine T and 3-5 mCi I-131, and the solution was agitated at room temperature for 5 min. The reaction was quenched with a 5 molar excess of sodium metabisulfite, and the reaction contents were added to a small reverse phase column (C18 SPE, 50 mg; Alltech, Deerfield, IL) previously equilibrated with 4% acetonitrile in water solution. Multiple aqueous washes removed unreacted iodide and aqueous soluble reagents. Acetonitrile was used to elute the radiolabeled cleavable linker. The acetonitrile solution containing radiolabeled cleavable linker was evaporated with a gentle stream of air, and B1 antibody (6 nmol in 1 mL of phosphate-buffered saline pH 7.4) was added to the dry radiolabeled cleavable linker and allowed to react for 1 h at room temperature with agitation. The cleavable radioimmunoconjugate was isolated by size exclusion chromatography (Pharmacia PD10, phosphate-buffered saline pH 7.4). Purification of β-Lactamase. β-Lactamase was purified from crude penicillinase type IV, Enterobacter cloacae, (Sigma), according to the method of Cartwright (17). Enzymatic activities of β-lactamase were determined using nitrocefin as a substrate (18). Characterization of Cleavable Radioimmunoconjugates. Reducing and nonreducing SDS gels of the cleavable RIC and standard size markers were run on 10-12% polyacrylamide. The gels were then dried and exposed to Kodak film for autoradiography. Immunoreactivity of the cleavable radioimmunoconjugate and standard labeled Tositumomab antibody were performed according to the method of Lindmo (19). In vitro cleaving assay were performed using 5-10 µL (1-10 µg) of cleavable RIC added to an eppendorf tube with 1 µg of β-lactamase enzyme. The contents were mixed and allowed to incubate for 15-60 min at 37 °C. Following the incubation period, 900 µL of phosphate-buffered saline was added, the contents were mixed, the total volume was added to a PD10 (Pharmacia) size exclusion column, and 12 1 mL fractions were collected and counted on a scintillation counter. In Vivo Biodistribution in Murine Model. Tenweek old NOD/SCID mice were injected in the tail vein with 25-50 µg of standard radiolabeled antibodies or 2550 µg of cleavable RIC. Mice were coinjected with 400 µg

Bioconjugate Chem., Vol. 14, No. 5, 2003 929 Scheme 1. Synthesis of Cleavable Linker 1a

a Conditions: (a) BocOPhCH CO H, DCC, DIEA; (b) NaOH; 2 2 (c) Ph2CN2; (d) CCl3CHClOCOCl, DIEA; (e) mCPBA; (f) BocNH(CH2)6NH2; (g) TFA, anisole; (h) di-sulfoNHS suberate, DIEA.

of a nonspecific IgG2a irrelevant antibody (G3G6) to block nonspecific binding to Fc receptors in spleen, liver, and marrow, and, 5-20 h following injection of RIC, groups of mice were injected in the tail vein with 10 µg of β-lactamase or buffer. Blood, urine, and tissue samples were collected immediately before and at various times after injection of enzyme or buffer. Collected body fluids and excised tissue samples were weighed, and 125I and 131 I radioactivity was counted. Counts are corrected for radioactive decay using an aliquot of the injectate. The percent injected dose of radionuclide per gram (%ID/g) of fluid or tissue was calculated and is reported as the mean ( standard deviation for n ) 3-5 mice. The statistical significance between mean tissue uptake in enzyme-treated and untreated mice for a given experiment was calculated with the unpaired Student’s t test for p < 0.05. In Vivo Biodistribution in Murine Lymphoma Xenograft Model. NOD/SCID mice were irradiated with 375 cGy and injected with 2 × 107 Ramos cells subcutaneously in the flank. Mice were monitored until palpable (3-10 mm) tumor nodules appeared (7-14 days) at which time mice were used for biodistribution studies as described above. RESULTS

Design and Synthesis of Cleavable Linker 1. The minimal requirements for a cleavable linker are a radionuclide-carrying moiety, an enzyme substrate that can be efficiently cleaved, and a protein-conjugation moiety. A cephalosporin nucleus was chosen as the cleavable enzyme substrate (compound 1, Scheme 1) since β-lactamase enzymes are highly specific and efficient for hydrolysis of the lactam, which then promotes spontaneous decomposition of the carbamate. In addition, there

930 Bioconjugate Chem., Vol. 14, No. 5, 2003

Beeson et al.

Figure 1. Characterization and in vitro lactamase treatment of cleavable Tositumomab RIC. Linker 1 was iodinated with I-131 and then conjugated to Tositumomab. Shown in panel A is a nonreducing SDS-PAGE gel of the antibody labeled with linker 1 (cleavable RIC, lane 2) and antibody conventionally labeled with I-131 and chloramine T (conv., lane 1). Shown in panel B is the cpm of fractions eluted off a size exclusion column for the cleavable RIC after 1 h treatment with β-lactamase or buffer (control).

is no endogenous β-lactamase activity in mammals. The aryl group of 4-hydroxyphenylacetic acid is activated for electrophilic iodination using traditional techniques such as chloramine T, and the NHS-ester in 1 readily reacts with the -nitrogen of lysine residues. The synthesis of 1 was readily achieved using conventional side-chain protection and peptide coupling chemistries (Scheme 1). Initially 7-aminocephalosporinic acid 2 was acylated with Boc-protected 4-hydroxyphenylacetic acid. The position-3 oxymethylene acetate was hydrolyzed with base, and the carboxylic acid was protected as the diphenylmethyl ester to give the alcohol 3 with overall yields of 20-40%. The hydroxy group of 3 was activated with 1,1,1,2-tetrachloroethyl orthochloroformate, the thioether was oxidized, and the tetrachloroethyl carbonate was then coupled to mono-Boc-protected 1,6-diaminohexane to give 4 after TFA deprotection (20-30% yield from 3). The linker 1 was then obtained from the reaction of 4 with the sulfo-NHS suberate diester (>90% yield from 4). Preparation and in Vitro Characterization of Cleavable Radioimmunoconjugates. The linker 1 was radioiodinated with chloramine T, and the iodinated linker was then incubated with Tositumomab to give the cleavable RIC. The efficiency of the cleavable RIC synthesis as assessed by percent incorporation of radioisotope into the final purified cleavable RIC was 5-10%. The cleavable RIC exhibited a molecular weight of approximately 175 kDa in nonreducing SDS PAGE analysis (Figure 1A) and comigrated with antibody radioiodinated by conventional techniques. Image analysis of gels showed about an 80% decrease of radioactivity in the region of the antibody after a 1 h treatment with β-lactamase (data not shown). Isolation of the cleavable radioimmunoconjugate-containing fraction by size exclusion chromatography also revealed an average 80-85% decrease in radioactive counts per minute in enzyme treated versus nontreated samples (Figure 1B). In antigen binding assays at antigen excess, the cleavable RIC, conventionally labeled RIC, and isotype control bound to 56%, 59%, and 0% of antigen-containing cells, respectively.

In Vivo Cleavage and Biodistribution in Mice. Nonspecific uptake of IgG2a antibodies in the reticuloendothelial system in tumor and non-tumor-bearing animals was minimized by the coinjection of control IgG2a antibody against an irrelevant specificity (G3G6). We found that coinjection of 400 µg of G3G6 was the most effective for inhibiting nonspecific uptake of radiolabeled B1 (data not shown). Various doses of radiolabeled Tositumomab were used in biodistribution experiments with the most optimal biodistributions observed after infusing 25-50 µg of radiolabeled Tositumomab per mouse; biodistribution of the cleavable RIC in mice was identical to that of standard radioiodinated Tositumomab (data not shown). The in vivo cleavage and clearance of the β-lactamase-sensitive anti-CD20 cleavable RIC was demonstrated with blood clearance and urinary excretion in mice. Two hundred micrograms of cleavable Tositumomab RIC trace-labeled with I-131 was injected iv into the tail vein of NOD/SCID mice. After 5 h, 10 µg of β-lactamase was administered iv to one-half of the mice, with remaining mice serving as controls. Blood clearance in enzyme-treated and control mice revealed a 3-fold decrease in %ID/g at 30 min after enzyme infusion, and a 4-fold decrease at the 20-h time point (Figure 2A). In the same mice, urine obtained 30 min following the β-lactamase injection showed a 300-fold increase in %ID/g compared to nontreated mice (Figure 2B). Representative data illustrating the effects of enzymatic cleavage on radioactive uptake in normal tissues 1 and 20 h after enzyme infusion are shown in Figure 3A and 3B. At 1 h we observed a statistically significant 3-fold decrease in spleen %ID/g and a 2-fold decrease in lung and liver %ID/g in the enzyme-treated group (Figure 3A). At 20 h after enzyme infusion, we observed a statistically significant 4-6-fold decrease in marrow and spleen %ID/g, a 3-5-fold decrease in lung and liver %ID/ g, and a 3-fold decrease in kidney %ID/g in the enzymetreated group (Figure 3B). Rapid renal clearance of the low molecular weight isotope-containing moiety led to the transient 2-fold increase in the kidney %ID/g 1 h following enzyme treatment. The β-lactamase-mediated relative decreases in radioactive uptake shown in Figure 3

Conditionally Cleavable Radioimmunoconjugates

Bioconjugate Chem., Vol. 14, No. 5, 2003 931

Figure 2. Representative clearance curves for blood (panel A) and urine (Panel B), expressed as %ID/g. Cleavable radiolabeled Tositumomab antibody was injected at time 0, and β-lactamase enzyme was injected 5 h later. Three mice were evaluated at each time point. Solid lines indicate enzyme-treated mice, and broken lines indicate control mice not injected with β-lactamase. Error bars are the standard deviation, and asterisks denote the concentrations in enzyme-treated mice that are significantly (p < 0.05) different from the controls.

were not significantly different than those observed in other experiments where the cleavable RIC was allowed to circulate for up to 24 h before administration of β-lactamase (data not shown). In Vivo Cleavage and Biodistribution in Lymphoma Xenograft Mice. Biodistribution studies in mice bearing lymphoma xenografts were undertaken to better determine the effects of the β-lactamase enzyme on the binding of radioisotope by cleavable RIC to tumor compared to normal tissues. We determined that maximum tumor localization of iodinated Tositumomab antibody occurs at 20-24 h (data not shown). Thus, the cleavable RIC was allowed to circulate for 20 h before enzyme was introduced. Immunodeficient mice with subcutaneous Ramos B lymphoma cell tumors received 25 µg cleavable B1-RIC labeled with 18 µCi of I-131. Mice also received 400 µg of irrelevant antibody to decrease nonspecific binding activity. After 20 h, one-half of the mice were treated iv into the tail vein with 6.4 µg of β-lactamase, and groups of 3-4 mice each were sacrificed at later time points. Representative biodistributions for 1 and 4 h after enzyme infusion are listed in Table 1. Evaluation of organ and tumor distribution revealed a statistically greater than 2-fold decrease in blood and marrow radioisotope concentrations 1 and 4 h after enzyme infusion. Significant decreases in radioactivity contents in lung and liver were also seen, with about a

Figure 3. Concentration of radioactivity in tissues expressed as %ID/g. Mice were injected with cleavable radiolabeled Tositumomab antibody at time 0, and half received β-lactamase enzyme at 5 h. Necropsy was performed in groups of three mice each at 1 h (panel A) and 20 h (panel B) after enzyme infusion. Error bars are the standard deviation, and asterisks denote the concentrations in enzyme-treated mice that are significantly (p < 0.05) different from the controls.

2-fold difference 1 and 4 h after enzyme infusion. An increase in kidney radioactivity was again seen at the 1 h time point. In contrast to blood and the normal organs, tumor radioactive content was not significantly decreased 1 h after enzyme infusion. However, by 4 h, the tumor %ID/g had decreased by 40% in mice receiving enzyme. We also observed a greater than 2-fold increase in tumor/ blood ratio of %ID/g at 1 h (1.3 vs 0.6) and a 50% increase at 4 h (0.9 vs 0.6) for enzyme-treated versus control. The gain in tumor/blood ratio is lost at later time points. By 20 h the biodistributions in enzyme-treated and untreated normal tissues are comparable to those illustrated in Figure 3, but the tumor radioactive content is typically reduced by about 70% compared to the control (data not shown). Thus, at these later time points there

932 Bioconjugate Chem., Vol. 14, No. 5, 2003

Beeson et al.

Table 1. I-131 RIC Biodistribution in Mice Bearing Lymphoma Xenografts at 1 h and 4 h %ID/g at 1 ha tissue blood marrow lung liver kidney tumor

%ID/g at 4 h

β-lactamse

control

β-lactamase

control

5.1d ( 0.2b, 1.3c,d 3.6d ( 0.2, 1.9d 2.0d ( 0.2, 3.5d 1.4d ( 0.1, 4.9d 4.0d ( 0.2, 1.7 6.9 ( 1.0

11.9 ( 1.0, 0.6 8.4 ( 0.9, 0.8 3.7 ( 0.2, 1.9 2.0 ( 0.1, 3.6 2.1 ( 1.22, 3.4 7.1 ( 0.6

4.5d ( 0.4, 0.9d 2.7d ( 0.8, 1.5d 1.6d ( 0.2 2.5 1.0d ( 0.1 4.0 2.3 ( 0.2 1.7 4.0d ( 0.2

12.1 ( 2.1, 0.6 6.8 ( 1.3, 3.3 3.3 ( 0.9 2.2 2.0 ( 0.6 3.6 2.7 ( 0.4 2.6 7.1 ( 1.2

a Time after iv injection of β-lactamase. b Standard deviation, n ) 4. c Ratio of tumor/normal tissue %ID/g. d Statistically different from relevant control at p < 0.05.

is no statistically significant difference in the tumor/blood ratios for treated and untreated mice. DISCUSSION

This report demonstrates that a conditionally cleavable RIC that utilizes a cephalosporin-based linker can be prepared and, when exposed to β-lactamase, radioactivity can be released and rapidly cleared by the kidneys. In addition, in vitro antigen binding studies and in vivo studies in murine xenograft models demonstrated tumor cell localization by the conditionally cleavable RIC. The ability to target antigen was not altered by the addition of the cleavable linker to the antibody since in vitro and in vivo antigen localization of the cleavable RIC and a standard labeled RIC was nearly identical. Biodistribution experiments demonstrated in vivo cleavage of the RIC with rapid clearance of radioisotope by kidney and decreased radioactivity contents in blood and organs known to be dose limiting for RIT, namely liver, lung, and marrow. Biodistribution studies of the cleavable RIC in a tumor xenograft model demonstrated that clearance of radioactivity from the blood and dose-limiting normal organs was enhanced relative to loss at the tumor immediately following β-lactamase infusion. However, this advantage was transient, because β-lactamase treatment also caused loss of radioactive content from the tumor. Since tumor-bound conjugate is in equilibrium with nonbound cleaved and noncleaved conjugate, this decrease in tumor content could be due to exchange of bound, labeled RIC with unlabeled conjugate. Alternatively, the decrease could be due to enzyme cleavage of tumor-bound RIC. Regardless, at early time points after enzyme infusion (i.e., 1 h to 4 h), use of the cleavable RIC significantly enhanced the tumor/blood %ID/g ratio although this gain is lost at later time points (i.e., >20 h). This study, to our knowledge, represents the initial synthesis of a conditionally cleavable radioimmunconjugates that, upon exposure to cleaving enzyme, will release the radioactive moiety in vitro or in vivo. The readily achieved release of the radioactive moiety in vivo following infusion of β-lactamase enzyme was successful in causing rapid clearance of radiation from blood and normal organs. However, the systemic administration of an exogenous enzyme can cause toxicity and immune responses, which are problems also encountered in antibody-directed enzyme prodrug therapies (ADEPT; 20, 21). Fortunately, the amounts of β-lactamase needed to clear radionuclide are considerably lower than the amounts used for ADEPT strategies, and this advantage

could be greatly increased with the use of, for example, dextran-modified β-lactamase, which has a very slow clearance rate and is not immunogenic (22). Although the β-lactamase used here also induced a loss of the radiolabel from tumor cells, the loss was delayed compared to the loss of radioactivity from normal organs. These observations suggest that effective use of a cleavable RIC for improving antibody-mediated delivery of radiolabel to tumor might be achieved if the antibodies used would not be available for cleavage following enzyme infusion. It should be possible to achieve this objective by employing antibodies that are rapidly internalized by the cell after membrane binding and subsequently retained within the cell, rather than antibodies such as Tositumomab, which are retained on the surface of cells for prolonged periods of time. In this way, enzyme-induced clearance of the radiolabel would only occur from conjugates not specifically bound to antigen and thus exposed to the enzyme. The data presented in this study also suggest that the current approach may be useful for improving tumor imaging with radiolabeled antibodies since radiolabel was cleared more rapidly from the blood than tumor or normal organs, which should allow greater discrimination between conjugate within the blood pool and at the tumor site. ACKNOWLEDGMENT

J.E.B. and M.J.H contributed equally to this work. This work is supported by NIH grants CA44991 and CA86963. I.D.B. is supported by an F. M. Kirby/American Cancer Society Clinical Research Professorship. LITERATURE CITED (1) Goodwin, D. A., and Meares, C. F. (1999) Pretargeted peptide imaging and therapy. Cancer Biother. Radiopharm. 14, 145-152. (2) Goodwin, D. A., and Meares, C. F. (1997) Pretargeting: general principles. Cancer 80 (Suppl.), 2675-2680. (3) Press, O. W., Corcoran, M., Subbiah, K., Hamlin, D. K., Wilbur, D. S., Johnson, T., Theodore, L., Yau, E., Mallett, R., Meyer, D. L., and Axworthy, D. A. (2001) Comparative evaluation of conventional and pretargeted radioimmunotherapy of CD20-expressing lymphoma xenografts. Blood 98, 2535-2543. (4) Hamblett, K. J., Kegley, B., Hamlin, D., Chyan, M., Hyre, D., Press, O., Wilbur, D., and Stayton, P. (2002) A streptavidin-biotin binding system that minimizes blocking by endogenous biotin. Bioconjugate Chem. 13, 588-598. (5) DeNardo, G. L., Maddock, S. W., Sgouros, G., Scheibe, P. O., and DeNardo, S. J. (1993) Immunoadsorption: an enhancement strategy for radioimmunotherapy. J. Nucl. Med. 34, 1020-1027. (6) Sgouros, G. (1992) Plasmapheresis in radioimmunotherapy of micro-metastases: a mathematical modeling and dosimetrical analysis. J. Nucl. Med. 33, 2167-2179. (7) Norrgren, K., Strand, S. E., and Ingvar, C. (1992) Contrast enhancement in RII and modification of the therapeutic ratio in RIT: A theoretical evaluation of simulated extracorporeal immunoadsorption. Antibody Immunoconjugate Radiopharm. 3, 61-73. (8) Chen, J. Q., Strand, S. E., Tennvall, J., Lindgren, L., Hindorf, C., and Sjogren, H. O. (1997) Extracorporeal immunoadsorption compared to avidin chase: enhancement of tumor-to-normal tissue ratio for biotinylated Rhenium-188chimeric BR96. J. Nucl. Med. 38, 1934-1939. (9) Dienhart, D. G., Kasliwal, R., and Lear, J. L. (1994) Extracorporeal immunoadsorption of radiolabeled monoclonal antibody: a method for reduction of background radioactivity and its potential role during the radioimmunotherapy of cancer. Antibody Immunoconj. Radiopharm. 7, 225-252. (10) Marshall, D., Pedley, R. B., Boden, J. A., Boden, R., and Begent, R. H. (1994) Clearance of circulating radio-antibodies

Conditionally Cleavable Radioimmunoconjugates using streptavidin or second antibodies in a xenograft model. Br. J. Cancer 69, 502-507. (11) Kobayashi, H., Sakahara, H., Hosono, M., Yao, Z. S., Toyama, S., Endo, K., and Konishi, J. (1994) Improved clearance of radiolabeled biotinylated monoclonal antibody following the infusion of avidin as a “chase” without decreased accumulation in the target tumor. J. Nucl. Med. 35, 16771684. (12) Studer, M., Kroger, L. A., DeNardo, S. J., Kukis, D. L., and Meares, C. F. (1992) Influence of a peptide linker on biodistribution and metabolism of antibody-conjugated benzylEDTA: Comparison of enzymatic digestion in vitro and in vivo. Bioconjugate Chem. 3, 424-429. (13) Stein, R., Goldenberg, D. M., Ong, G. L., Thorpe, S. R., and Mattes, M. J. (1997) Manipulation of blood clearance to optimize delivery of residualizing label-antibody conjugates to tumor cells in vivo. J. Nucl. Med. 38, 1392-1400. (14) DeNardo, G. L., Kroger, L. A., Meares, C. F., Richman, C. M., Salako, Q., Shen, S., Lamborn, K. R., Peterson, J. J., Miers, L. A., Zhong, G. R., and DeNardo, S. J. (1998) Comparison of 1,4,7,10-tetraazacyclododecane-n,n′,n′′,n′′′-tetraacetic acid (DOTA)-peptide-CHL6, a novel immunoconjugate with catabolizable linker, to 2-iminothiolane-2-[p(bromoacetamido)benzyl]-DOTA-CHL6 in breast cancer xenografts. Clin. Cancer Res. 4, 2483-2490. (15) Arano, Y., Wakisaka, K., Akizawa, H., Ono, M., Kawai, K., Nakayama, M., Sakahara, H., Konishi, J., and Saji, H. (1998) Assessment of the radiochemical design of antibodies with a metabolizable linkage for target-selective radioactivity de-

Bioconjugate Chem., Vol. 14, No. 5, 2003 933 livery. Bioconjugate Chem. 9, 497-506. (16) Smith, L. I., and Howard, K. L. (1955) Preparation of diphenyldiazomethane. Organic Syntheses; Collect. Vol. III, p 351, Wiley, New York. (17) Cartwright, S. J., and Waley, S. G. (1984) Purification of beta-lactamases by affinity chromatography on phenylboronic acid-agarose. Biochem. J. 221, 505-512. (18) Shannon, K., and Phillips, I. (1980) Beta-lactamase detection by three simple methods: intralactam, nitrocefin and acidimetric. J. Antimicrob. Chemother. 6, 617-621. (19) Lindmo, T., Boven, E., Cuttitta, F., Fedorko, J., and Bunn. P. A., Jr. (1984) Determination of the immunoreactive fraction of labeled monoclonal antibodies by linear extrapolation to binding at infinite antigen excess. J. Immunol. Methods 72, 77-89. (20) Melton, R. G., and Sherwood, R. F. (1996) Antibody-enzyme conjugates for cancer therapy. J. Natl. Cancer Inst. 88, 153165. (21) Syrigos, K. N., and Epenetos, A. A. (1999) Antibody directed enzyme prodrug therapy (ADEPT): a review of the experimental and clinical considerations. Anticancer Res. 19, 605613. (22) Mikolajcyzk, S. D., Meyer, D. L., Fagnani, R., Hagan, M. S., Law, K. L., and Starling, J. J. (1996) Dextran modification of a Fab′-β-lactamase conjugate modulated by variable pretreatment of Fab′ with amine-blocking reagents. Bioconjugate Chem. 7, 150-158.

BC025655Z