Bioconjugate Chem. 2007, 18, 175−182
175
Development of Tissue Plasminogen Activator Specific “On Demand Cleavable” (ODC) Linkers for Radioimmunotherapy by Screening One-Bead-One-Compound Combinatorial Peptide Libraries Pappanaicken R. Kumaresan,† Arutselvan Natarajan,‡ Aimin Song,† Xiaobing Wang,† Ruiwu Liu,† Gerald DeNardo,‡ Sally DeNardo,‡ and Kit S. Lam*,† Division of Hematology and Oncology, and Radiodiagnosis and Therapy, Department of Internal Medicine, UC Davis Cancer Center, University of California Davis, School of Medicine, Sacramento, California 95817. Received August 31, 2006; Revised Manuscript Received November 7, 2006
New strategies are needed to protect normal organs from radiation in cancer radioimmunotherapy (RIT). This can be achieved by rapid clearance of radiometal in the circulation after accumulation of radioimmunoconjugates (RIC) in the tumor. Our strategy is to place highly efficient and specific cleavable linkers between radiometal chelates and the tumor targeting agents. Such linkers must be resistant to cleavage by enzymes present in the plasma and the tumor. After radiotargeting agents have accumulated in the tumor, a cleaving agent can be administered “on demand” to cleave a specific linker, resulting in the release of radiometal from the circulating RIC in a form that will have rapid renal clearance. We have selected TNKase, a thrombolytic agent approved for patient use, as our model on-demand cleaving agent. To identify TNKase-specific linkers, we screened fluorescentquenched random “one-bead-one-compound” (OBOC) combinatorial peptide libraries. D-Amino acid containing peptides that were specific for TNKase but were resistant to cleavage by plasma and tumor-associated proteases were identified. One of these peptide substrates (rqYKYkf) was used to link the DOTA chelate to ChL6, a monoclonal antibody known to target breast cancer. This antibody conjugate was stable in plasma for 7 days while preserving the immunoreactivity to intact tumor cells. The addition of TNKase at clinical achievable plasma level (10 µg/mL) resulted in the release of 28% of the radiometal from the radioimmunoconjugate within 72 h. This lead linker, after further optimization to increase its response to TNKase, may be useful in the development of more effective radioimmunotherapeutic and radioimaging agents.
INTRODUCTION (RIT)1,
Radioimmunotherapy in which a targeting agent is conjugated to a radiometal, is a promising approach to treat cancer (1-3). However, its potential is limited by radiation exposure to normal organs, particularly bone marrow (4-6). One approach to clear nontargeted radioactivity is to increase renal excretion of circulating radiometal from the body after tumor targeting (7-9). Immunoadsorption has been utilized to remove circulating radioimmunoconjugates (RIC) and reduce radiation dose to normal tissues (10). Removal of free circulating RIC after adequate tumor uptake has been tried preclinically and clinically with modest success (11, 12). DeNardo et al. used a similar approach by treating B-cell lymphoma patients using 131I Lym-1, an anti-lymphoma monoclonal antibody, with and without immunoadsorption (13, 14). Results for two patients receiving RIT with immunoadsorption on four occasions demonstrated that RIC was reduced by 75% in blood and by 50% in the bone marrow when their plasma was pumped * Corresponding author. Kit S. Lam, M.D., Ph.D., Professor of Medicine, Chief, Division of Hematology & Oncology, UC Davis Cancer Center, 4501 X Street, Sacramento, CA 95817. Tel: (916)734-8012. Fax: (916)734-7946. E-mail:
[email protected]. † Division of Hematology and Oncology. ‡ Radiodiagnosis and Therapy, Division of Hematology and Oncology. 1 Abbreviations: RIT, radioimmunotherapy; RIC, radioimmunoconjugates; ODC, on-demand cleavable; OBOC, one-bead-onecompound; t-PA, tissue plasminogen activator; ChL6, chimeric mAb; DOTA, 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid; CAE, cellulose acetate electrophoresis. Standard single-letter abbreviation for amino acids is used. Upper case letters represent L-amino acids; lower case letters represent D-amino acids.
through an anti-gammaglobulin column for removing the circulating 131I Lym-1 antibodies. It is clear that a large portion of the RIC can be removed with this approach. However, the procedure is cumbersome and expensive, and there is need for a simpler strategy. Various forms of pretargeting RIT delivery have been developed to overcome this problem and used in clinical trials, for example, by administering tumor targeting molecules that bind subsequently injected radiochelates. The small radiochelates that do not bind to the tumor are cleared quickly from body. This results in a high therapeutic index (tumor dose vs normal tissue dose) (15-18). However, continued circulating pretargeting molecules have presented interference in tumor targeting of the radiochelates; clearing agents, such as biotinN-acetyl-galactosamine for the antibody-strepavidin conjugates, have resulted in effective radiochelate tumor targeting (1517). Although pretargeting approaches have great promise, these systems require multiple injections and administration of reagents at specific time points. Some of the reagents used in such systems, e.g., streptavidin, can be immunogenic. Another approach is to introduce a cleavable linker between the targeting agent and the radiochelate. For example, RIC with linkers cleavable by endogenous proteases such as liver-specific cathepsin-D or by an exogenous hydrolytic enzyme β-lactamase have been reported (19-21). Repeated usage of exogenous enzymes of nonhuman source, such as β-lactamase, however, may elicit an immune response. An ideal on-demand cleavable (ODC) enzyme should be of human origin and be primarily restricted to the intravascular space so that only the circulating radioconjugates will be cleaved. Recombinant tissue plasminogen activator (t-PA),
10.1021/bc0602681 CCC: $37.00 © 2007 American Chemical Society Published on Web 12/23/2006
176 Bioconjugate Chem., Vol. 18, No. 1, 2007
currently used as a thrombolytic agent in patients (TNKase) provides these characteristics (22). Thus, the aim of this study is to develop TNKase substrates with high specificity and efficiency that can be used as an ODC linker between radiometal chelate and tumor targeting agents such as mAbs. In this study, we have shown that currently known TNKase peptide substrates are not stable in plasma and cancer cell supernatants. We have utilized the one-bead-one-compound (OBOC) combinatorial library method (23, 24) to develop an ODC linker that is highly susceptible to TNKase but highly resistant to other proteases in plasma or in a tumor. In this study, we have selected one such TNKase peptide substrate as a lead ODC linker and have evaluated its potential to reduce radiation to normal tissues.
EXPERIMENTAL PROCEDURES Materials and Instruments. PEGA resin (0.40 mmol/g) was purchased from Polymer Laboratories (Amherst, MA). Boc-2Abz-OH and Boc-Tyr(3-NO2)-OH were purchased from ChemImpex (Wood Dale, IL). All other amino acid derivatives and HOBt (1-hydroxybenzotriazole) were purchased from SynPep (Dublin, CA). DOTA-NHS-ester (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid mono(N-hydroxysuccinimide ester) and DOTA-mono-NHS-tris(t-Bu)-ester(1,4,7,10-tetraazacyclododecane-1,4,7-tris(t-butyl acetate)-10acetic acid mono(N-hydroxysuccinimide ester)) were purchased from Macrocyclics (Dallas, TX). DIC (N,N′ diisopropylcarbodiimide) was purchased from Advanced ChemTech (Louisville, KY). DMF (N,N′-dimethylformamide), DCM (dichloromethane), MeOH (methanol), diethyl ether, and acetonitrile were purchased from Fisher (Houston, TX). All other chemical reagents were purchased from Aldrich (Milwaukee, WI). Analytical HPLC analyses (Vydac column, 4.6 mm × 250 mm, 5 µm, 300 Å, C18, 1.0 mL/min) and preparative HPLC purification (Vydac column, 20 mm × 250 mm, 5 µm, 300 Å, C18, 7.0 mL/min, 45 min gradient) were performed on a Beckman System Gold HPLC system (Fullerton, CA). MALDI-TOF MS analysis was performed on a Bruker BIFLEX III mass spectrometer (Billerica, MA). All of the experiments are carried out at room temperature unless otherwise noted. ChL6 (Bristol-Myers Squibb Pharmaceutical Research Institute, Seattle, WA) was a chimera consisting of human IgG1-constant regions and the murine L6 MAb antigen-variable region. Carrier-free 111In(MDS Nordion, Vancouver, Canada) was purchased as chloride in 0.05 M HCl solution. Sephadex G50-80 (Pharmacia, Uppsala, Sweden), molecular sieving high-performance liquid chromatography (HPLC) (Beckman Coulter System Gold 127; Beckman, San Ramon, CA; SEC-3000 molecular sieving column), cellulose acetate electrophoresis (CAE) (Gelman Sciences, Inc., Ann Arbor, MI) were used for purification and characterization of RIC. TNKase was purchased from Genentech (San Francisco, CA) and reconstituted with distilled water to get 10 mg/mL concentration. The reconstituted samples were stored in -80 °C in aliquots. Synthesis of “Fluorescent-Quenched” Peptide Library. An OBOC combinatorial library, containing random Y(NO2)peptide-K(Abz), was synthesized on PEGA resin with a standard “split-mix” approach (25). The following fluorescent-quenched libraries were prepared on PEGA beads: Y(NO2)-xxxxxxxK(Abz), Y(NO2)-xxXxx-K(Abz), Y(NO2)-xxxXXxxx-K(Abz), and Y(NO2)-xxXXXxx-K(Abz). For these random peptides, Y(NO2) represents 3-nitrotyrosine, K(Abz) represents lysine bearing a 2-aminobenzoyl group on its -amino group, X represents one of 19 natural L-amino acids excluding cysteine, and x represents one of the D-isomers of 18 natural L-amino acids plus glycine. Theoretically, there are 197 ) 8.9 × 108 possible permutations of peptide sequences in this library.
Kumaresan et al.
Screening and Sequencing. Four different combinatorial libraries were screened against TNKase enzyme. Approximately 1 mL of the beads (about 325 000 beads) was used in each screening experiment, and a total of 1 million beads from each library were screened. Fluorescent-quenched combinatorial libraries were first incubated with pooled plasma, isolated from citrate-treated human blood obtained from normal volunteers. The beads were incubated with 50% plasma containing PBS overnight at 37 °C with occasional shaking. Beads with peptides that were labile in plasma and therefore fluoresced and those beads that autofluoresced were removed by a fluorescence bead sorter (COPAS, Union Biometrica). The remaining nonfluorescent beads were washed several times with PBS to remove plasma proteins. 10 µg/mL TNKase in PBS was added and incubated for 1 h at 37 °C with occasional shaking. After 1 h, beads were incubated with 8 M guanidine HCl to terminate the enzyme activity and washed with water several times to remove all the proteins attached on the beads. Positive fluorescent beads were picked under a fluorescent microscope at 360/410 excitation/emission spectrum and sequenced with Edman chemistry using an automatic microsequencer (ABI, Procise 494). On the basis of the plasma stability and TNKase susceptibility, 12 peptides were selected as candidate substrates for TNKase. For in vitro stability assay, the selected candidates were synthesized as soluble fluorescent-quenched substrates without tethering to beads as described below. Synthesis of Individual “Fluorescent-Quenched” ODC Linkers. Rink amide MBHA resin (70 mg, 0.09 mmol) was swollen in DMF overnight. The supernatant was removed, and 20% piperidine solution in DMF (2 mL) was added to the resin. The mixture was agitated for 15 min, and the supernatant was drained. This process was repeated. The resin was thoroughly washed with DMF, MeOH, and DMF. Fmoc-Lys(Alloc)-OH (122.2 mg, 0.27 mmol), HOBt (36.5 mg, 0.27 mmol), and DIC (42.3 µL, 0.27 mmol) in DMF (2 mL) were added to the resin. The resulting mixture was agitated until the ninhydrin test was negative. The resin was washed with DMF, MeOH, and DCM. In the presence of argon, a solution of PhSiH3 (266.5 µL, 2.16 mmol) in DCM (1 mL) was added to the resin, followed by a solution of Pd(PPh3)4 (26.0 mg, 0.0225 mmol) in DCM (2.2 mL). The mixture was shaken in an argon atmosphere for 30 min. This process was repeated. The supernatant was removed, and the resin was washed thoroughly with DCM, DMF, MeOH, and DMF. To the resin was added a solution of Boc-2-AbzOH (64.1 mg, 0.27 mmol), HOBt (36.5 mg, 0.27 mmol), and DIC (42.3 µL, 0.27 mmol) in DMF (2 mL). The resulting mixture was agitated until the ninhydrin test was negative. The supernatant was removed. The resin was washed with DMF, and the R-Fmoc protecting group of the lysine was removed by piperidine treatment as described above. The peptide was then synthesized on the R-amino group of the lysine using standard Fmoc chemistry. Boc-Tyr(3-NO2)-OH was used as the last amino acid. After the peptide sequence was finished, the resin was washed with DMF, MeOH, and DCM, and then dried under vacuum. 4 mL of cleavage mixture (TFA (trifluoroacetic acid)/phenol/thioanisole/water/TIS (triisopropylsilane), 82.5:7.5: 5:5:2.5, v/w/v/v/v) was added to the dried resin at 0 °C. The mixture was slowly warmed to room temperature and allowed to mix for 6 h. The supernatant was then collected, and the resin was washed with TFA (3 × 1 mL). The combined supernatants were concentrated to 1 mL under a stream of nitrogen. The peptide was precipitated by dilution with ethyl ether (10 mL), separated by centrifugation, washed with ethyl ether, and dried under vacuum. The crude peptide was analyzed and purified by HPLC and characterized by MALDI-TOF MS. Synthesis of DOTA-ODC Linker with an Aminooxyacetyl (Aoa) Group. The DOTA-ODC linker (DOTA-spacer-rqYKYkf-
Bioconjugate Chem., Vol. 18, No. 1, 2007 177
On Demand Cleavable Linkers for Radioimmunotherapy
spacer-K(Aoa)-NH2) was synthesized using a procedure similar to the synthesis of fluorescent-quenched peptides. After the Alloc deprotection, instead of Boc-2-Abz-OH, Boc-Aoa-OH was coupled to the -amino group of the lysine. The R-Fmoc protecting group was then removed by piperidine treatment as described above. To increase ODC accessibility for TNKase digestion, a short hydrophilic spacer was introduced on both sides of the ODC linker. This linear hydrophilic spacer has a diamine component at the N-terminal end derived from 2,2′(ethylenedioxy)-bis(ethylamine) and an anhydride component at the C-terminal end derived from succinic anhydride, respectively (26). A Fmoc-protected spacer was coupled to the R-amino group of the lysine using our previously published procedure (26). After the coupling of the last amino acid, the N-terminus Fmoc protection group was removed by piperidine treatment. An additional spacer was introduced between the DOTA and the peptide. After Fmoc deprotection, the resin was incubated with a solution of DOTA-NHS-ester (112.0 mg, 0.135 mmol) and DIEA (156.8 µL, 0.9 mmol) until the ninhydrin test was negative. The peptide was released from the resin and purified using the procedure described above. A total of 26 mg of DOTA-spacer-rqYKYkf-spacer-K(Aoa)-NH2 was obtained from 200 mg of resin (yield 14%), and mass (2088.1) was calculated by MALDI-TOF MS (M + H)+. All the substrates were purified by preparative HPLC, and purity was verified by both analytical and MALDI-TOF mass spectrometry. The products that showed g95% purity were taken for the TNKase susceptibility and stability assay. Linker Stability Analysis. 100 nmol/mL of fluorescentquenched ODC linkers were incubated in PBS containing 10% human plasma at 37 °C. The susceptibility was monitored every hour in a fluorescence plate reader (Tecan) at 360/410 nm excitation/emission spectrum. The ODC linker that was stable in plasma overnight was taken for further stability analysis. For this, the cell culture supernatant was isolated from several cancer cell lines (approximately 1-5 million cells/mL) that were split 1 day before the collection. 100 nmol/mL of the ODC linkers was incubated with the cancer cell culture supernatant, and its susceptibility was monitored in a fluorescence plate reader at 360/410 nm excitation/emission spectrum. ODC linkers that were stable in both plasma and various cancer cell culture supernatants were taken for the TNKase susceptibility assay. TNKase Susceptibility Assay. Potential ODC linkers (100 nmol/mL) were incubated with TNKase (10 µg/mL) in PBS at 37 °C with occasional shaking. To simulate the in vivo condition, plasma was added (10%) in the assay medium. The susceptibility was monitored at various time intervals at 360/ 410 nm excitation/emission spectrum using a Tecan fluorescence plate reader. The entire assay was done in triplicate, and two independent experiments were performed. Antibody-Linker Conjugation and MALDI-MS Analysis. For on-demand cleavable studies, ChL6, an antibody targeted against tumor antigen expressed in colon, lung, and breast cancer cells, was used (27, 28). DOTA-ODC-aminoxyamine was conjugated with ChL6 by the ketone-oxime method developed in our lab (29). In brief, primary amines (mainly lysine) present ChL6 mAb was coupled with levulinic succinimidyl esters by incubating a 1:30 molar ratio in 40 mM TEA, pH 8.0, for 5 h in a cold room. The uncoupled levulinic succinimidyl esters were removed by dialysis with 40 mM TEA, pH 5.5 buffer. ChL6 was equilibrated with 40 mM TEA, pH 5.5, and incubated with 20 molar excess of DOTA-ODC-ONH2 linkers for 5-12 h. To verify that ChL6 Ab was conjugated, a small aliquot was removed at various time intervals, and the degree of conjugation was determined by the expected mass shift in MALDI-TOF MS. Estimation of Metal-Chelatable DOTA (c) per MAb (a) in RIC. The DOTA-ODC-ChL6 conjugate was purified by
Sephadex G50 column chromatography (30) and was tested for its pyrogen-free status by standard procedures. DOTA-ODCChL6 was assayed in triplicate by cobalt metal binding assay to determine the ratio of the metal chelatable DOTA-ODC (chelate) per ChL6 (c/a) (30-32). Generally, an aliquot of DOTA-ODC-ChL6 (480 µg, 3 nmol) dissolved in 0.1 M tetramethylammonium phosphate buffer at pH 7.0 was treated with excess standardized aqueous solution of 57CoCl2 (ICN Radiochemicals, Irvine, CA) and incubated for 30 min at room temperature. To this reaction mixture, aqueous EDTA solution (Fisher Scientific, Fair Lawn, NJ) was added to a final concentration of 10 mM and incubated for 15 min to complex with excess unbound cobalt ions. This EDTA chased mixture was purified by a Penefsky column. The concentration of the 57Co/CoCl -chelated DOTA-ODC-ChL6 was determined by 2 radiation counting using a γ counter against a known quantity of the standardized 57CoCl2 solution. The metal-chelatable DOTA was calculated as the ratio of the molar concentration of the 57Co-DOTA-ODC-ChL6 (chelate) and ChL6 (MAb). Radiolabeling of DOTA-ODC-ChL6 and Quality Control (QA). To prepare RIC, DOTA-ODC-ChL6 and 111In were combined in 0.1 M ammonium acetate (pH 5.3) and incubated for 30 min at 37 °C. To scavenge unbound 111In, EDTA was added to a final concentration of 10 mM for 30 min at room temperature. The final RIC was purified by molecular sieving chromatography. QA of the RIC was established by SEC-3000 HPLC, cellulose acetate electrophoresis (CAE), and immunoreactivity (32, 33). Immunoreactivity, Plasma Stability, and TNKase Cleavage Assay. 111In-DOTA-ODC-ChL6 was incubated with ChL6 antibody-binding breast cancer cells (HBT 3477) and ChL6 nonbinding lymphoma cancer cells (Raji). Immune reactivity was assayed by performing a competitive assay against lightly iodinated 125I-ChL6. Approximately 1 million cells were incubated for 1 h and washed with PBS to remove any unbound antibody. This experiment was carried out in triplicate and repeated twice. The cells were counted using a β counter. The in vitro plasma stability of RIC was evaluated by mixing 10 µg of 111In-DOTA-ODC-ChL6 incubated in freshly prepared human plasma (1:1 v/v dilution; total volume, 1 mL) at 37 °C in a humidified incubator maintained at 5% CO2 and 95% air. At various intervals (1, 3, 5, 7, 9, 13, and 14 days), aliquots were taken and analyzed by CAE and high-performance liquid chromatography (32, 33). For TNKase susceptibility assay, RIC (0.39 µCi/µg)) was incubated in human plasma with and without TNKase at the clinical dosage level (10 µg/mL). Plasma samples were fractioned by HPLC. Radioactivity and UV absorption (280 nm) from each fraction were monitored. The degree of susceptibility was calculated by either decreased radioactivity to the peak corresponding to the ChL6 mAb or increased radioactivity to the fractions corresponding to the peptide peak.
RESULTS Identification of TNKase (t-PA) Specific Substrates. To be effective, ODC linkers need to be (i) stable to proteases present in the plasma, (ii) stable to proteases present in normal tissues and tumor, and (iii) highly susceptible to TNKase. Prior to screening with TNKase, the four OBOC fluorescent-quenched combinatorial libraries were prescreened with human plasma so that those beads with peptides susceptible to plasma could be identified and removed. A library with more L-amino acids in their structure yielded more positive beads. In fact, about 10-15% of the bead library (xxXXXxx) became fluorescent after mixing with plasma for 1 h at room temperature. Fluorescent beads from the plasma-treated peptide libraries were physically removed either manually or by passing through a fluorescence bead sorter (COPAS); the residual beads were
178 Bioconjugate Chem., Vol. 18, No. 1, 2007
Kumaresan et al.
Table 1. TNKase Substrates Identified from OBOC Libraries P4
P3
P2
P1
C
P P x v d k r l r r p q q r
G F X R A G N H Q F G Y Y Y
R R X K R K K K K K R K K K
x y p p l f f G k r G
V
P 1′
P 2′
P3′
P4′ a
V G X Q K K K S W M R Y Y G
V S x w f G G r G a H k k k
G Ac xd k n e G i k t k le fe se
Gb
p
V; a depicts the TNKase cleaving site on peptides and P and P′ amino acids positions are numbered accordingly. b Peptide substrate derived from t-PA cleaving site of its native substrate plasminogen. c Peptide substrate derived from phage display library. d Structure of OBOC combinatorial library. e Motifs identified by aligning 3 peptide substrate homologues based on the cleavage site.
screened with TNKase. Similar to the plasma experiment, the xxXXXxx peptide library yielded more positive beads than the xxxXXxxx library. xxXxx and xxxxxxx libraries, on the other hand, were totally resistant to digestion by TNKase. Since only a portion of the peptides on each positive bead were cleaved by the enzyme, there were enough intact peptides left on each bead for sequence determination. On the basis of the microsequencing result (Edman degradation), we determined the precise TNKase cleavage site. The amino acid sequences of the 12 TNKase-susceptible peptides isolated from the xxXXXxx bead library are shown in Table 1. All twelve peptides except one were found to be cleaved at the peptide bond between the fourth and fifth residues. One peptide (pGRRHkp) was cleaved between the third and fourth residues. In both cases, the peptides were all cleaved between two L-amino acids. This is not unexpected, as peptide bonds between two D-amino acids or one D- and one L-amino acid are generally resistant to proteolysis. Table 1 also includes the peptide sequence (PGRVVGG) of the cleavage site of plasminogen, the native substrate for TNKase, and a peptide substrate, PFRGSA, identified by the phage-display peptide library method. The P1 position of all fourteen peptides was found to contain either an Arg or a Lys. Tyr and Phe are the
preferred amino acids in the P2 position. Eight of the twelve substrates have one basic residue (D-Arg or D-Lys) at either the P3 or P4 position. Similarly, there is at least one basic residue in position P1′, P2′, or P3′. Peptides rqYKYkf and kqYKYkl share the same structural homology. Stability in Biological Fluids. For in vitro stability and susceptibility studies, the identified candidates were synthesized in a free form that is not tethered to beads to rule out steric effects of beads on TNKase digestion. As shown in Figure 1, all the peptides listed in Table 1 were stable in plasma over a 12 h period except for PFRGSA (phage display peptide) and GrFKMat. Peptides yvRKQwk, kqYKYkl, flHKSri, frQKWGk, and rqYKYkf were susceptible to TNKase digestion. Of these, rqYKYkf, frQKWGk, and yvRKQwk were the most susceptible and were completely cleaved within 1 h of TNKase addition (10 µg/mL). Since an effective ODC linker needs to be stable to proteolytic enzymes at the tumor site, we evaluated the susceptibility of these peptide substrates against tissue culture supernatants from various cancer cell lines: pancreatic cancer (Panc1), prostate cancer (PC3), B-cell lymphomas (Raji, Ramos), T-cell lymphomas (Jurkat), and normal immortalized hepatocytes (PHH) (Figure 2). Even though most of the peptides showed stability up to 5 h in the supernatant, peptides flHKSri and rqYKYkf were stable in cancer cell supernatants and plasma for at least 10 h. Out of these two peptides, rqYKYkf was more susceptible to TNKase (Figure 1), and this peptide (Table 1) was also identified twice from library screening. Therefore, we elect rqYKYkf as the first prototype ODC linker for further characterization. Conjugation and Radiolabeling. Peptide rqYKYkf was derivatized with DOTA and conjugated to a chimeric ChL6 antibody for radiolabeling and in vitro susceptibility studies with TNKase. Prior to conjugation, ChL6 was analyzed by HPLC, CAE, and PAGE, and its purity was found to be over 95% (data not shown). After conjugation, as revealed by MALDI-TOF mass spectrometry, the number of DOTA peptides conjugated to ChL6 was determined to be 1-3 (Figure 3). 57Co/CoCl2 was utilized, which demonstrated that 0.5 c/a was available to chelate radiometal. The 111In-DOTA-ODC-ChL6 was purified by Sephadex-G50 column chromatography, and the purity was confirmed by the HPLC and CAE radioactive profiles. Linker Stability and Immunoreactivity Studies. The plasma stability and TNKase susceptibility of 111In-DOTAODC-ChL6 was assayed in vitro prior to in vivo study. 111In-
Figure 1. The fluorescence intensity of the ODC-peptides (100 nmol/mL) with plasma (10%) alone is shown on the left. On the right, the fluorescence intensity of the peptides (100 nmol/mL) with 10% of plasma containing 10 µg/mL of TNKase. Peptide “rqYKYkf” is stable in plasma and highly susceptible to TNKase digestion.
Bioconjugate Chem., Vol. 18, No. 1, 2007 179
On Demand Cleavable Linkers for Radioimmunotherapy
Figure 2. The ODC-linker (rqYKYkf) demonstrated notable stability, while others demonstrated varying degrees of susceptibility to enzymes present in these cell supernatants.
Figure 5. Cellulose acetate electophoresis (CAE): 45 min analysis was carried out at pH 8.6, with CAE strip. After 45 min, the strips were cut at 0.5 cm of each and counted on a γ counter. The radioactivity traces represented are migration patterns of 111In-DOTA-ODC-ChL6 on CAE strips of days 1 and 7 with human plasma. The traces are identical. No radioactivity was observed in the lower molecular weight region for any breakdown product. Figure 3. ChL6 antibody mass spectrum is designated as IgG and ChL6-ODC-DOTA mass spectrum designated as IgG + peptide. On the basis of the mass spectrum, we estimate that ∼1 to 3 peptides were conjugated per antibody (calculated by shift in mass).
ODC-ChL6 on HBT cells was greater than 50%, whereas it was only 6% to Raji cells. In the competitive assay, the RIC competed with 125I-labeled ChL6. This further suggests that conjugating the ODC linker to the antibody did not alter its antigen binding site (data not shown). 111In-DOTA-ODCChL6 was incubated in plasma with and without TNKase at 37 °C, and we found that 10-15% and 28% of the radiometal were released from ChL6-ODC-DOTA conjugates at the clinical dosage level in 2 and 72 h, respectively (Figure 6).
DISCUSSION
Figure 4. HPLC analysis was performed at UV 280 nm, and each 0.5 mL fraction was collected and radioactivity was counted. The radioactive profiles of 111In-DOTA-ODC-ChL6 with human plasma at days 1 and 7 are identical. Thus, the RIC was essentially stable for a minimum of 6 days.
DOTA-ODC-ChL6 was monitored for 14 days in plasma. Stability of the conjugated ODC-linker in RIC was assessed by HPLC for radioactive and protein peak. Figure 4 showed that the RIC was intact for 7 days (Figure 4). The appearance of small MW species at days 1 and 7 was due to a small amount (e5%) of 111In-EDTA present in the original preparation. CAE and HPLC analysis indicated that g95% of the radiometal was associated with RIC, and g98% of RIC was in monomeric form (Figure 5). The appearance of a small percentage of radioactivity on small molecular peaks was noticed after 7 days of incubation with plasma. 111In-DOTA-ODC-ChL6 selectively bound to specific breast cancer cells (HBT) and did not bind to negative control cells (Raji). Immunoreactivity of the 111In-DOTA-
To protect vital organs in myeloablative RIT, an equal amount of unlabeled mAb is preinfused before the radiolabeled mAb infusion. The unlabeled mAb will saturate nonspecific antibodybinding sites and Fc receptors in the host body, that will aid labeled mAb in targeting the tumor and reduce nonspecific binding during targeting. In spite of that, radioactive uptake in normal organs is still very high, resulting in bone marrow abalation (34, 35), and nonhematological toxicity such as lung, gastrointestinal, liver, and kidney toxicity (36-39). It has been proven that complete removal of radioimmunoconjugates from circulation after tumor targeting is the best way to protect the vital organs. Various methods have been developed, but none has reached the clinic because of practical difficulties and cost associated with the techniques. Thus, to date, no optimal strategy has been developed to limit the radiation toxicity to normal organs. Introduction of cathepsin-D degradable linkers between Ab and radiometal-DOTA-chelate has shown partial protection for the liver that is dose-limiting for myeloablative RIT using radiometals (20). Pretargeting approaches are promising but have limitations. The administration of exogenous protease from a nonhuman source, such as β-lactamase, has been shown to
180 Bioconjugate Chem., Vol. 18, No. 1, 2007
Kumaresan et al.
Figure 6. 111In-DOTA-ODC-ChL6 was incubated with human plasma containing 10 µg/mL of TNKase. HPLC radiochromatograms are shown from samples, before (b) and after [2 h (red 4) and 72 h (green O)] the addition of TNKase. Peak A: 111In-DOTA-ODCChL6, 150 kD. Peak B: 111In-DOTA-ODC, 5 kD. At 10 µg/mL, TNKase digested 28% of RIC by 72 h. Scheme 1. Hypothesis
Schematic
Representation
of
ODC
Linker
cleave the radioconjugates in circulation; however, radioconjugates at tumor sites are also cleaved, suggesting that such agents should be active only in circulation in order to be effective (21). In this study, we provide an alternative novel approach to lower radiation exposure to all normal organs by producing radioimmunoconjugates with TNKase-susceptible “on-demand cleavable” (ODC) linkers. After adequate uptake of radiometal in the tumor, TNKase can be administered intravenously to cleave the ODC linker, resulting in the release of radiometalDOTA-chelate from the circulating RIC for renal excretion (Scheme 1). Consequently, the unbound radiometal-DOTAchelate can be rapidly eliminated from the circulation, resulting in diminished toxicity to normal organs. TNKase was chosen over other proteases as an on-demand cleavage agent because (i) it has already been used safely for several years as a thrombolytic agent (22, 40), (ii) it has one known physiological substrate, plasminogen (41), (iii) it is already commercially available in both the short-acting (Ativase) and longer-acting
forms (TNKase) (42), and (iv) its volume of distribution is very similar to the blood volume, indicating that the enzyme resides predominantly within the intravascular space. Bleeding is the main side effect of the t-PA therapy at clinical dosage (10 µg/ mL), so we intend to use only one-tenth of the clinical dosage (1 µg/mL) or less, by developing highly susceptible ODC linkers. A linear peptide substrate derived from the primary sequence of plasminogen (t-PA’s physiological substrate) is not cleaved efficiently as intact plasminogen, suggesting that it is not a good candidate for an ODC linker (43, 44). Ding et al. described the use of phage-display combinatorial peptide libraries to identify PFGRSA peptide as an efficient substrate for t-PA (44). However, this peptide is also highly susceptible to cleavage by proteases present in plasma (Figure 1) and cancer cell supernatants (Figure 2). Thus, it is unsuitable for use as an ODC linker. One main reason that PFGRSA is rapidly degraded in human plasma and cancer cell supernatant is that it is an all L-amino acid peptide. Every peptide bond within this hexapeptide is a potential cleavage site for the numerous proteases present in plasma or tissue proteases. In order to eliminate many of the unwanted cleavage sites by other proteases but still retain appropriate amino acid side chains at the P2, P3, P4, P2′, P3′, and P4′ sites for specific interaction with TNKase, we needed to develop peptide libraries with a combination of D-amino acids and L-amino acids. This is impossible with phage-display libraries but can be easily accomplished with synthetic peptide libraries such as the OBOC combinatorial peptide library method. Screening an OBOC library is one of the best strategies to identify potential enzyme substrates and novel cancer-specific targeting ligands. We have recently reported the use of the focused OBOC library strategy to develop a lymphoma-targeting ligand with binding affinity of approximately 2 pM (45, 46). By screening the OBOC libraries with a fluorescence-quench assay, we identified unique peptide linkers that were susceptible to TNKase cleavage but stable in human plasma and cancer cell supernatants. One of these peptides rqYKYkf, containing both D- and L-amino acids, is particularly promising, as it can be completely degraded by TNKase within 2 h, resulting in an enzyme level that is clinically achievable (10 µg/mL), and yet it is stable in plasma and cancer cell supernatants. When rqYKYkf was used as an ODC linker to ligate DOTA to ChL6 antibody and DOTA was labeled with 111In, only 28% of the radiometal was released from the RIC after incubation with TNKase (10 µg/mL) for 72 h. The less-efficient cleavage of rqYKYkf when linked to an antibody when compared to a solution peptide may in part be explained by steric hindrance caused by the antibody molecule or by the DOTA.
CONCLUSION AND FUTURE PLAN We have demonstrated that highly specific and efficient ODCpeptide linkers can be discovered through screening OBOC combinatorial peptide libraries containing both D- and L-amino acids. To improve the efficiency of on-demand cleavage by eliminating possible steric hindrance caused by DOTA and/or the antibody molecule, we plan to insert short PEG linkers between DOTA and rqYKYkf and between rqYKYkf and the antibody. We also plan to link two rqYKYkf peptides serially so that cleavage of any of the two peptides will lead to radiometal-DOTA-chelate release. Furthermore, we plan to design focused OBOC libraries based on the rqYKYkf sequence and use a more stringent screening method, e.g., a much lower dose of TNKase (1µg/mL), to further optimize a highly specific and efficient substrate for TNKase.
On Demand Cleavable Linkers for Radioimmunotherapy
ACKNOWLEDGMENT This work was supported by PO1 grant (CA47829) from NIH. The authors thank Jan Marik for help in mass spectrometry analysis, Gary Mirick for performing immunoreactivity assay, and Amanda Enstrom for help in manuscript preparation.
LITERATURE CITED (1) Denardo, S. J., Richman, C. M., Goldstein, D. S., Shen, S., Salako, Q., Kukis, D. L., Meares, C. F., Yuan, A., Welborn, J. L., and Denardo, G. L. (1997) Yttrium-90/indium-111-DOTA-peptidechimeric L6: pharmacokinetics, dosimetry and initial results in patients with incurable breast cancer. Anticancer Res. 17, 17351744. (2) DeNardo, G. L., Kukis, D. L., Shen, S., DeNardo, D. A., Meares, C. F., and DeNardo, S. J. (1999) 67Cu-versus 131I-labeled Lym-1 antibody: comparative pharmacokinetics and dosimetry in patients with non-Hodgkin’s lymphoma. Clin. Cancer Res. 5, 533-541. (3) Press, O. W., Shan, D., Howell-Clark, J., Eary, J., Appelbaum, F., Matthew, D., Hinman, L., Shochat, D., and Bernstein, I. D. (1996) Comparative metabolism and retension of iodine-125, yttrium-90, and indium-111 radioimmunoconjugates by cancer cells. Cancer Res. 56, 2123-2129. (4) Breitz, H. (2002) Dosimetry in a myeloablative setting. Cancer Biother. Radiopharm. 17, 119-128. (5) Bouchet, L. G., Bolch, W. E., Blanco, H. P., Wessels, B. W., Siegel, J. A., Rajon, D. A., Clairand, I., and Sgouros, G. (2003) MIRD pamphlet no. 19: Absorbed fractions and radionuclide values for six age-dependent multiregion models of the kidney. J. Nucl. Med. 44, 1113-1147. (6) Wessels, B. W., Bolch, W. E., Bouchet, L. G., Breitz, H. B., DeNardo, G. L., Meredith, R. F., Stabin, M. G., and Sgouros, G. (2004) Bone marrow dosimetry using blood-based models for radiolabeled antibody therapy: A multiinstitutional comparison. J. Nucl. Med. 45, 1725-1733. (7) Goodwin, D. A., and Meares, C. F. (2001) Advances in pretargeting biotechnology. Biotechnol. AdV. 19, 435-450. (8) Paganelli, G. (2005) Pretargeted radioimmunotherapy. Ernst Schering Res. Found. Workshop 49, 73-84. (9) Breitz, H. B., Fisher, D. R., Goris, M. L., Knox, S., Ratliff, B., Murtha, A. D., and Weiden, P. L. (1999) Radiation absorbed dose estimation for 90Y-DOTA-biotin with pretargeted NR-LU-10/ streptavidin. Cancer Biother. Radiopharm. 14, 381-395. (10) Norrgren, K., Strand, S. E., Nilsson, R., Lindgren, L., and Sjogren, H. O. (1993) A general, extracorporeal immunoadsorption method to increase the tumor-to-normal tissue ratio in radioimmunoimaging and radioimmunotherapy. J. Nucl. Med. 34, 448-454. (11) Dienhart, D. G., Kasliwal, R., Lear, J. L., Johnson, T. K., Bloedow, D. C., Hartmann, C., Seligman, P. A., Miller, G. J., Glenn, S. D., Mcateer, M. J., Thickman, D., Feyerabend, A., Maddock, E. N., Maddock, S. W., and Bunn, P. A. (1994) Extracorporeal immunoadsorption of radiolabeled monoclonal-antibody - a method for reduction of background radioactivity and its potential role during the radioimmunotherapy of cancer. Antibody Immunoconjugates and Radiopharmaceuticals 7, 225-252. (12) Tennvall, J., Garkavij, M., Chen, J., Sjogren, H. O., and Strand, S. E. (1997) Improving tumor-to-normal-tissue ratios of antibodies by extracorporeal immunoadsorption based on the avidin-biotin concept: development of a new treatment strategy applied to monoclonal antibodies murine L6 and chimeric BR96. Cancer 80, 2411-2418. (13) Denardo, G. L., Denardo, S. J., Meares, C. F., Kukis, D., Diril, H., Mccall, M. J., Adams, G. P., Mausner, L. F., Moody, D. C., and Deshpande, S. V. (1991) Pharmacokinetics of copper-67 conjugated Lym-1, a potential therapeutic radioimmunoconjugate, in mice and in patients with lymphoma. Antibody Immunoconjugates and Radiopharmaceuticals 4, 777-785. (14) 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.
Bioconjugate Chem., Vol. 18, No. 1, 2007 181 (15) Chang, C. H., Sharkey, R. M., Rossi, E. A., Karacay, H., McBride, W., Hansen, H. J., Chatal, J. F., Barbet, J., and Goldenberg, D. M. (2002) Molecular advances in pretargeting radioimunotherapy with bispecific antibodies. Mol. Cancer Ther. 1, 553-563. (16) Boerman, O. C., van Schaijk, F. G., Oyen, W. J., and Corstens, F. H. (2003) Pretargeted radioimmunotherapy of cancer: progress step by step. J. Nucl. Med. 44, 400-411. (17) Gruaz-Guyon, A., Raguin, O., and Barbet, J. (2005) Recent advances in pretargeted radioimmunotherapy. Curr. Med. Chem. 12, 319-338. (18) Weiden, P. L., Breitz, H. B., Press: O., Appelbaum, J. W., Bryan, J. K., Gaffigan, S., Stone, D., Axworthy, D., Fisher, D., and Reno, J. (2000) Pretargeted radioimmunotherapy (PRIT (TM)) for treatment of non-Hodgkin’s lymphoma (NHL): Initial phase I/II study results. Cancer Biother. Radiopharm. 15, 15-29. (19) Meares, C. F., McCall, M. J., Deshpande, S. V., DeNardo, S. J., and Goodwin, D. A. (1988) Chelate radiochemistry: cleavable linkers lead to altered levels of radioactivity in the liver. Int. J. Cancer Suppl. 2, 99-102. (20) DeNardo, G. L., DeNardo, S. J., Peterson, J. J., Miers, L. A., Lam, K. S., Hartmann-Siantar, C., and Lamborn, K. R. (2003) Preclinical evaluation of cathepsin-degradable peptide linkers for radioimmunoconjugates. Clin. Cancer Res. 9, 3865S-72S. (21) Beeson, C., Butrynski, J. E., Hart, M. J., Nourigat, C., Matthews, D. C., Press, O. W., Senter, P. D., and Bernstein, I. D. (2003) Conditionally cleavable radioimmunoconjugates: a novel approach for the release of radioisotopes from radioimmunoconjugates. Bioconjugate Chem. 14, 927-933. (22) Davydov, L., and Cheng, J. W. (2001) Tenecteplase: a review. Clin. Ther. 23, 982-97; discussion 981. (23) Lam, K. S., Salmon, S. E., Hersh, E. M., Hruby, V. J., Kazmierski, W. M., and Knapp, R. J. (1991) A new type of synthetic peptide library for identifying ligand-binding activity. Nature (London) 354, 82-84. (24) Lam, K. S., Lebl, M., and Krchnak, V. (1997) The “one-beadone-compound’’ combinatorial library method. Chem. ReV. 97, 411448. (25) Meldal, M., Svendsen, I., Breddam, K., and Auzanneau, F. I. (1994) Portion-mixing peptide libraries of quenched fluorogenic substrates for complete subsite mapping of endoprotease specificity. Proc. Natl. Acad. Sci. U.S.A. 91, 3314-3318. (26) Song, A. M., Wang, X. B., Zhang, J. H., Marik, J., Lebrilla, C. B., and Lam, K. S. (2004) Synthesis of hydrophilic and flexible linkers for peptide derivatization in solid phase. Bioorg. Med. Chem. Lett. 14, 161-165. (27) Goodman, G. E., Hellstrom, I., Yelton, D. E., Murray, J. L., Ohara, S., Meaker, E., Zeigler, L., Palazollo, P., Nicaise, C., Usakewicz, J., and Hellstrom, K. E. (1993) Phase-I trial of chimeric (humanmouse) monoclonal-antibody L6 in patients with non-small-cell lung, colon, and breast-cancer. Cancer Immunol. Immunother. 36, 267273. (28) Denardo, S. J., Mirick, G. R., Kroger, L. A., Ogrady, L. F., Erickson, K. L., Yuan, A., Lamborn, K. R., Hellstrom, I., Hellstrom, K. E., and Denardo, G. L. (1994) The biologic window for chimeric L6 radioimmunotherapy. Cancer 73, 1023-1032. (29) Xu, Q., Miyamoto, S., and Lam, K. S. (2004) A novel approach to chemical microarray using ketone-modified macromolecular scaffolds: application in micro cell-adhesion assay. Mol. DiVersity 8, 301-310. (30) Meares, C. F., McCall, M. J., Reardan, 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, 68-78. (31) Goodwin, D. A., Meares, C. F., Watanabe, N., Mctigue, M., Chaovapong, W., Ransone, C. M., Renn, O., Greiner, D. P., Kukis, D. L., and Kronenberger, S. I. (1994) Pharmacokinetics of pretargeted monoclonal-antibody 2d12.5 and Y-88 Janus-2-(p-nitrobenzyl)1,4,7,10-tetraazacyclododecanetetraacetic acid (DOTA) in Balb/C mice with Khjj mouse adenocarcinoma - a model for Y-90 Cancer Res. 54, 5937-5946. (32) Kukis, D. L., DeNardo, G. L., DeNardo, S. J., Mirick, G. R., Miers, L. A., Greiner, D. P., and Meares, C. F. (1995) Effect of the extent of chelate substitution on the immunoreactivity and biodistribution of 2IT-BAT-Lym-1 immunoconjugates. Cancer Res. 55, 878-884.
182 Bioconjugate Chem., Vol. 18, No. 1, 2007 (33) Moi, M. K., Meares, C. F., DeNardo, S. J. (1988) The peptide way to macrocyclic bifunctional chelating agents: synthesis of 2-(pnitrobenzyl)-1,4,7,10-tetraazacyclododecane-N,N′,N′’,N′′′-tetraacetic acid and study of its yttrium (III) complex. J. Am. Chem. Soc. 110, 6266-6267. (34) Lim, S. M., DeNardo, G. L., DeNardo, D. A., Shen, S., Yuan, A., O’Donnell, R. T., and DeNardo, S. J. (1997) Prediction of myelotoxicity using radiation doses to marrow from body, blood and marrow sources. J. Nucl. Med. 38, 1374-1378. (35) Breitz, H. B., Fisher, D. R., and Wessels, B. W. (1998) Marrow toxicity and radiation absorbed dose estimates from rhenium-186labeled monoclonal antibody. J. Nucl. Med. 39, 1746-1751. (36) Divgi, C. R., Bander, N. H., Scott, A. M., O’Donoghue, J. A., Sgouros, G., Welt, S., Finn, R. D., Morrissey, F., Capitelli, P., Williams, J. M., Deland, D., Nakhre, A., Oosterwijk, E., Gulec, S., Graham, M. C., Larson, S. M., and Old, L. J. (1998) Phase I/II radioimmunotherapy trial with iodine-131-labeled monoclonal antibody G250 in metastatic renal cell carcinoma. Clin. Cancer Res. 4, 2729-2739. (37) Behr, T. M., Sharkey, R. M., Sgouros, G., Blumenthal, R. D., Dunn, R. M., Kolbert, K., Griffiths, G. L., Siegel, J. A., Becker, W. S., and Goldenberg, D. M. (1997) Overcoming the nephrotoxicity of radiometal-labeled immunoconjugates - Improved cancer therapy administered to a nude mouse model in relation to the internal radiation dosimetry. Cancer 80, 2591-2610. (38) Press, O. W., Appelbaum, F. R., Eary, J. F., and Bernstein, I. D. (1995) Radiolabeled antibody therapy of lymphomas. Important AdV. Oncol. 157-171. (39) Press, O. W., Eary, J. F., Appelbaum, F. R., Martin, P. J., Nelp, W. B., Glenn, S., Fisher, D. R., Porter, B., Matthews, D. C., Gooley, T., and Bernstein, I. D. (1995) Phase II trial of 131I-B1 (anti-CD20)
Kumaresan et al. antibody therapy with autologous stem cell transplantation for relapsed B cell lymphomas. Lancet 346, 336-340. (40) Guerra, D. R., Karha, J., and Gibson, C. M. (2003) Safety and efficacy of tenecteplase in acute myocardial infarction. Expert Opin. Pharmacother. 4, 791-798. (41) Madison, E. L., Coombs, G. S., and Corey, D. R. (1995) Substratespecificity of tissue-type plasminogen-activator - characterization of the fibrin independent specificity of T-Pa for plasminogen. J. Biol. Chem. 270, 7558-7562. (42) Modi, N. B., Fox, N. L., Clow, F. W., Tanswell, P., Cannon, C. P., Van, de Werf, F., and Braunwald, E. (2000) Pharmacokinetics and pharmacodynamics of tenecteplase: results from a phase II study in patients with acute myocardial infarction. J. Clin. Pharmacol. 40, 508-515. (43) Madison, E. L. (1994) Probing structure-function-relationships of tissue-type plasminogen-activator by site-specific mutagenesis. Fibrinolysis 8, 221-236. (44) Ding, L., Coombs, G. S., Strandberg, L., Navre, M., Corey, D. R., and Madison, E. L. (1995) Origins of the specificity of tissuetype plasminogen activator. Proc. Natl. Acad. Sci. U.S.A. 92, 76277631. (45) Peng, L., Liu, R. W., Marik, J., Wang, X. B., Takada, Y., and Lam, K. S. (2006) Combinatorial chemistry identifies high-affinity peptidomimetics against R(4)β(1) integrin for in vivo tumor imaging. Nat. Chem. Biol. 2, 381-389. (46) Peng, L., Liu, R., Wang, X., Marik, J., Takada, Y., and Lam, K. S. (2005) High affinity high specificity R4 β1 integrin targeting peptides for lymphoid cancers. Biopolymers 80, 559-559. BC0602681
