Biotin Reagents for Antibody Pretargeting. 7. Investigation of

Love–Hate ligands for high resolution analysis of strain in ultra-stable protein/small molecule interaction. Michael Fairhead , Di Shen , Louis K.M...
0 downloads 0 Views 138KB Size
Bioconjugate Chem. 2006, 17, 1514−1522

1514

Biotin Reagents for Antibody Pretargeting. 7. Investigation of Chemically Inert Biotinidase Blocking Functionalities for Synthetic Utility D. Scott Wilbur,* Donald K. Hamlin, and Ming-Kuan Chyan Department of Radiation Oncology, University of Washington, Seattle, Washington 98195. Received April 6, 2006; Revised Manuscript Received August 10, 2006

An investigation was conducted to evaluate three biotin derivatives designed to block biotinidase cleavage of the biotinamide bond. Difficulties in multistep syntheses of molecules containing tert-butyl protected hydroxymethyl and carboxylate groups positioned R to a biotinamide bond led to the investigation of alternative biotinidaseblocking moieties that do not require protection and deprotection. The targeted biotin derivatives contained serineO-methyl ether, 2-aminobutyric acid, and valine moieties conjugated to the biotin carboxylate functionality. Those derivatives were further modified with a radioiodinated aryl ring to study their biotinidase stability. As a comparison to previously studied biotin derivatives, radioiodinated versions of biotin conjugates that contained (a) no biotinidase stabilizing group, (b) an N-methyl (sarcosine) stabilizing group, (c) an R-carboxylate (aspartate) stabilizing group and hydroxymethyl (serine) stabilizing group were also prepared and tested. When tested in human serum, all of the radioiodinated biotinidase-stabilized biotin derivatives had biotinvaline-CN-Cbl > biotin-serine-O-methyl ether-CN-Cbl > biotin-aminobutyric acid-CN-Cbl. Due to the high cost of serine-O-ethyl ether (and its N-Boc derivative) and difficulty in syntheses of its biotin derivatives, that adduct is not an attractive candidate for application to compounds used in vivo. The higher lipophilicity and diminished binding of the biotin-valine adduct also makes its use in vivo less attractive. Thus, the biotinaminobutyric acid adduct appears to be the best candidate for incorporation into biotin derivatives used in vivo, as it simplifies the synthetic procedures, has low cost, and provides effective blocking of biotinidase while retaining high binding affinity.

INTRODUCTION This laboratory has been involved in the development of radiohalogenated biotin derivatives for therapy of cancer using the monoclonal antibody-based “pretargeting” approach (13). As part of those studies, we have investigated structural requirements in biotin derivatives to block the activity of the enzyme biotinidase (4-7). Biotinidase functions in a biotin salvage pathway, where it cleaves the biotinamide bond in the biotin-lysine adduct biocytin, which is present from peptidase degradation of biotin-containing carboxylases and histones (8, 9). Biotinidase may also function as a biotin carrier (10) to transfer biotin to other proteins such as histones (11, 12). The biotinamide cleavage by biotinidase is not specific for biocytin, as it occurs with biotin conjugates of other alkyl and aromatic amino compounds. This nonspecificity has allowed the development of colorimetric assays with p-aminobenzoic acid and aminoquinoline adducts of biotin (13), and radiometric assays with biotin adducts of p-[14C]aminobenzoic acid (14) and [125I]iodotyramine (15), for measuring bioactivity of biotinidase in clinical assays. Indeed, biotinidase may not be highly specific for biotin adducts either, as data has been published which indicates that it can cleave structurally similar desthiobiotinylp-aminobenzoic acid (16) and lipoic acid-lysine adducts (8, 17, 18). * Address correspondence to: D. Scott Wilbur, Ph.D., Department of Radiation Oncology, University of Washington, Box 359658, 325 Ninth Ave., Seattle, WA 98104, Phone: 206-341-5437, FAX: 206341-5438, E-mail: [email protected].

The nonspecific nature of the cleavage of biotinamide bonds in biotin conjugates has made it imperative that biotin derivatives employed in vivo be designed in a manner that blocks the enzyme activity. For this reason, a number of investigators (1921), including ourselves (22, 23), have studied biotin conjugates for use in pretargeting applications which have structural features designed to block biotinidase action. Our studies initially followed the lead reported by other investigators in which it was found that the N-methylglycine (sarcosine) adduct of biotin fully blocked the action of biotinidase (24). However, we felt it was important that the high binding affinity of biotin with avidin or streptavidin be retained, so competitive binding studies were conducted with biotin. From those studies, some biotin adducts were identified which were stable to biotinidase and retained the high binding affinity of the biotin-(strept)avidin system (as measured by a slow dissociation rate). The biotin adducts identified contained either a carboxylate (e.g., biotinaspartate, biotin-lysine, biotin-cysteine) or hydroxymethyl (e.g., biotin-serine) on the carbon R to the biotinamide bond (25). Subsequently, we have conducted a number of synthetic studies incorporating aspartate, lysine, and serine in the targeted biotin-containing compounds. In most synthetic studies, biotinaspartate or biotin-lysine was used, and the biotinidase-blocking carboxylate was protected as a tert-butyl ester. Unfortunately, unwanted side reaction products were obtained with the t-Bu ester protecting group. The problems encountered in synthetic procedures led us to investigate alternate biotinidase blocking

10.1021/bc060084m CCC: $33.50 © 2006 American Chemical Society Published on Web 10/28/2006

Chemically Inert Biotinidase Resistant Derivatives

Bioconjugate Chem., Vol. 17, No. 6, 2006 1515

Figure 1. Radioiodinated biotin derivatives and precursors prepared for evaluation in the biotinidase assay.

Figure 2. Cyanocobalamin adducts of biotin derivatives prepared for evaluation in the biotin dissociation assay.

groups that are chemically inert or functionalized in a manner that a protecting group does not have to be removed after the synthesis. The studies focused on three biotin-amino acid conjugates that have steric, but nonreactive, functional groups R to the biotinamide bond. The biotin derivatives were synthesized and evaluated for biotinidase stability and binding with streptavidin. The derivatives studied included biotin conjugates of (a) L-serine-methyl ether, (b) D,L-2-aminobutyric acid and (c) L-valine, which contain methyl hydroxymethyl, ethyl and isopropyl functional groups (respectively) R to the biotinamide functionality. Stability of the biotin-amino acid conjugates toward biotinidase cleavage in human serum was evaluated by preparing radioiodinated biotin derivatives from arylstannyl-biotin derivatives (22, 23). The structures of the biotin derivatives prepared, radioiodinated, and evaluated are shown in Figure 1. Changes in the binding affinity of two of the three biotin-amino acid adducts with recombinant streptavidin (rSAv) was evaluated using a size-exclusion HPLC assay (26). In that assay, relative rates of dissociation of biotin derivatives from rSAv can be observed after challenge with a large excess of biotin. The assay required preparation of CN-Cbl conjugates shown in Figure 2. The preparation of biotin derivatives, radioiodination of biotinarylstannanes, and evaluation of radioiodinated biotin derivatives

in the biotinidase assay and biotin-CN-Cbl derivatives in the HPLC dissociation assay are described herein.

EXPERIMENTAL PROCEDURES General. Most chemicals were purchased from SigmaAldrich Corporation (St. Louis, MO, and Milwaukee, WI) as analytical grade and were used without further purification. Bocserine(Me)-OH DCHA salt was obtained from Bachem Bioscience Inc. (King of Prussia, PA). Biotin N-hydroxysuccinimide ester, 11, was obtained as a gift from Quanta BioDesign, Ltd (Powell, OH). Solvents for HPLC analysis were obtained as HPLC grade and were filtered (0.2 µm) prior to use. Silica gel chromatography was conducted with 70-230 mesh 60 Å silica gel (Aldrich Chemical Co.). Microcon-30 centrifugation concentrators were obtained from Amicon (Beverly, MA). Melting 1 Abbreviations: Av, avidin; BSA, bovine serum albumin; CN-Cbl, cyanocobalamin; cpm, counts per minute; DCHA, dicyclohexylammonium salt; EDC, 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide; HOHgBz, hydroxymercuribenzoic acid; NCS, N-chlorosuccinimide; NHS, N-hydroxysuccinimide; PBS, phosphate-buffered saline; rt, room temperature; rSAv, recombinant streptavidin; SAv, streptavidin; t-Boc, tert-butyloxycarbonyl; TFA, trifluoroacetic acid; TFP, tetrafluorophenyl; TFP-OTFA, tetrafluorophenyl trifluoroacetate.

1516 Bioconjugate Chem., Vol. 17, No. 6, 2006

points were obtained in open capillary tubes on a Mel-Temp II apparatus with a Fluke 51 K/J electronic thermometer and are uncorrected. Spectral Analyses. 1H NMR were obtained on either a Bruker AV 300 (300 MHz) or Bruker AV 500 (500 MHz) instrument. Chemical shifts are expressed as ppm using tetramethylsilane as internal standard (δ ) 0.0 ppm). High-resolution mass spectral (HRMS) data were obtained on a Bruker APEX III 47e Fourier transform mass spectrometer using electrospray ionization. For analysis, the samples were dissolved in 50/50 MeOH/ H2O and were introduced by an integral syringe infusion pump (Cole Parmer Series 74900). Radioactive Materials. Standard methods for safely using radionuclides of iodine were employed (27). Na[125I]I was purchased from Perkin-Elmer Life Sciences (Billerica, MA) as high pH, high concentration solutions in 0.1 N NaOH. Measurement of 125I was accomplished on the Capintec CRC-15R or a Capintec CRC-6A Radioisotope Calibrator. Radioactive samples were counted in a Wallac 1480 Wizard Gamma Counter (PerkinElmer Life and Analytical Services, Wellesley, MA). Chromatography. HPLC separations of nonradioactive compounds were obtained on Hewlett-Packard quaternary 1050 gradient pumping system with a variable wavelength UV detector (254 nm) and a Varex ELSD MKIII evaporative lightscattering detector. Analysis of the HPLC data were conducted on Hewlett-Packard HPLC ChemStation software. All reactions were monitored by HPLC. Reversed-phase HPLC chromatography was carried out using an Alltech Altima C-18 column (5 µm, 250 mm × 4.6 mm) using a gradient solvent system at a flow rate of 1 mL/min. A gradient of MeOH and H2O/0.1% HOAc was used. Starting with 40% MeOH, the initial solvent mixture was held for 2 min, the gradient was increased to 100% MeOH over the next 10 min, and then held at 100% MeOH for 8 min. Retention times (tR) under these conditions for biotin conjugates are provided with the compound experimental. HPLC separations of radioiodinated biotin derivatives were conducted using a gradient system consisting of a HP 1050 quaternary pump, a Waters Model 481 UV detector (254 nm), and a Beckman model 170 radioisotope detector. The separations were obtained using HPLC conditions described for Gradient I, above, except water was used in the place of the H2O/0.1% HOAc. Size-exclusion HPLC analyses used in the rSAv binding (dissociation) experiments were conducted on a system consisting of a Hewlett-Packard 1050 Multiple Wavelength Detector (280 and 362 nm), isocratic pump, and a Waters Protein-Pak 300SW glass column. The mobile phase was an aqueous solution containing 50 mM potassium phosphate (pH 6.8), 300 mM NaCl, 1 mM EDTA, and 1 mM NaN3. A flow rate of 1.0 mL/ min was used. Retention time for rSAv and bound biotincyanocobalamin derivatives was 10.5 min and free biotincyanocobalamin derivatives was 14.0 min. The following compounds were prepared as previously described: 1a, 1b, [125I]1b, 2a, 2b, [125I]2b (22); 3a, 3b, [125I]3b, 4a, 4b [125I]4b (23); 8 (26); 18 (28); 25, 26 (22); 27 (26). General Procedure for Synthesis of Biotin-Amino AcidTrioxatridecanediaminebenzamide Adducts, 5a/b-7a/b. A quantity of 19, 20, or 21 was dissolved in neat TFA (3-5 mL) and stirred at room temperature for 10 min. Then the TFA was evaporated under a stream of argon. The amine 22, 23, or 24 (TFA salt) was washed with EtOAc (3 × 20 mL) and dried under vacuum for 2 h. The crude amine TFA salt was dissolved in a solution of anhydrous DMF (10 mL) and Et3N (0.25 mL, 1.79 mmol), and then 1.3 equiv of the p-tri-n-butylstannylbenzoate TFP ester 25 or p-iodobenzoate TFP ester 26 was added and stirred at room temperature for 10 min to 1 h. The volatile materials were removed under vacuum. The residue was purified

Wilbur et al.

by a silica gel column (1.5 cm × 25 cm), eluting with a mixture of MeOH/EtOAc to give the desired product 5a, 5b, 6a, 6b, 7a, or 7b. 2-(N-Biotinoyl)-1-(N′-(13′-p-(tri-n-butylstannyl)benzoylamino)-4′,7′,10′-trio xatridecanamino)-L-serine-O-methyl Ether, 5a. Following the general procedure (stirring for 30 min), 5a was obtained after silica gel chromatography (20% MeOH/ EtOAc) as a colorless solid in 87% yield, mp ) 137-139 °C. 1H NMR 300 MHz (CD OD): δ 0.91 (t, J ) 7.3 Hz, 9H), 1.13 3 (t, J ) 8.1 Hz, 4H), 1.32-1.51 (m, 10H), 1.55-1.81 (m, 14H), 1.88-1.96 (m, 2H), 2.32 (t, J ) 7.3 Hz, 2H), 2.73 (d, J ) 12.5 Hz, 1H), 2.92-2.99 (m, 1H), 3.19-3.35 (m, 4H), 3.37 (s, 3H), 3.49-3.54 (m, 4H), 3.56-3.71 (m, 9H), 4.33 (dd, J ) 4.4, 7.9 Hz, 1H), 4.49-4.54 (m, 2H), 7.59 (d, J ) 8.2 Hz, 2H), 7.78 (d, J ) 8.2 Hz, 2H). HRMS (ES+) calcd for C43H75N5NaO8SSn (M + Na)+: 964.4256. Found: 964.4274. HPLC: tR ) 16.7 min. 2-(N-Biotinoyl)-1-(N′-(13′-p-iodobenzoylamino)-4′,7′,10′trioxatridecaneam ino)- L-serine-O-methyl Ether, 5b. Following the general procedure (stirring for 1 h), 5b was obtained after silica gel chromatography (40% MeOH/EtOAc) as a colorless solid in 83% yield, mp ) 152-154 °C. 1H NMR 300 MHz (CD3OD): δ 1.34 (t, J ) 7.3 Hz, 3H), 1.41-1.52 (m, 2H), 1.56-1.82 (m, 5H), 1.86-1.95 (m, 2H), 2.32 (t, J ) 7.3 Hz, 2H), 2.73 (d, J ) 12.6 Hz, 1H), 2.94 (dd, J ) 5.0, 12.6 Hz, 1H), 3.24 (q, J ) 7.3 Hz, 3H), 3.33-3.35 (m, 2H), 3.37 (s, 3H), 3.47-3.54 (m, 3H), 3.56-3.70 (m, 9H), 4.33 (dd, J ) 4.4, 8.0 Hz, 1H), 4.49-4.54 (m, 2H), 7.60 (d, J ) 8.5 Hz, 2H), 7.87 (d, J ) 8.5 Hz, 2H). HRMS (ES+) calcd for C31H48IN5NaO8S (M + Na)+: 800.2166. Found: 800.2157. HPLC: tR ) 11.9 min. 2-(N-Biotinoyl)-1-(N′-(13′-p-(tri-n-butylstannyl)benzoylamino)-4′,7′,10′-trio xatridecanamino)aminobutyramide, 6a. Following the general procedure (stirring for 30 min), 6a was obtained after silica gel chromatography (15% MeOH/EtOAc) as a colorless solid in 91% yield, mp ) 130-132 °C. 1H NMR 300 MHz (CD3OD): δ 0.90-0.99 (m, 10H), 1.14 (t, J ) 8.1 Hz, 4H), 1.32-1.51 (m, 10H), 1.55-1.83 (m, 14H), 1.86-1.96 (m, 2H), 2.29 (t, J ) 7.3 Hz, 2H), 2.72 (d, J ) 12.4 Hz, 1H), 2.94 (dd, J ) 5.0, 12.9 Hz, 1H), 3.19-3.35 (m, 6H), 3.51 (t, J ) 6.1 Hz, 4H), 3.56-3.70 (m, 9H), 4.16-4.24 (m, 1H), 4.32 (dd, J ) 4.4, 7.9 Hz, 1H), 4.51 (dd, J ) 4.4, 7.9 Hz, 1H), 7.59 (d, J ) 8.3 Hz, 2H), 7.77 (d, J ) 8.3 Hz, 2H). HRMS (ES+) calcd for C43H75N5NaO7SSn (M + Na)+: 948.4307. Found: 948.4302. HPLC: tR ) 16.8 min. 2-(N-Biotinoyl)-1-(N′-(13′-p-iodobenzoylamino)-4′,7′,10′trioxatridecanam ino)aminobutyramide, 6b. Following the general procedure (stirring for 10 min), 6b was obtained after silica gel chromatography (30% MeOH/EtOAc) as a colorless solid in 88% yield, mp ) 80-82 °C. 1H NMR 300 MHz (CD3OD): δ 0.96 (t, J ) 7.5 Hz, 3H), 1.41-1.51 (m, 2H), 1.551.83 (m, 8H), 1.86-1.97 (m, 2H), 2.29 (t, J ) 7.3 Hz, 2H), 2.72 (d, J ) 12.5 Hz, 1H), 2.95 (dd, J ) 5.0, 12.7 Hz, 1H), 3.19-3.35 (m, 4H), 3.50 (q, J ) 6.4 Hz, 4H), 3.56-3.69 (m, 9H), 4.19 (dd, J ) 5.9, 8.3 Hz, 1H), 4.32 (dd, J ) 4.5, 7.9 Hz, 1H), 4.52 (dd, J ) 4.5, 7.9 Hz, 1H), 7.60 (d, J ) 8.6 Hz, 2H), 7.87 (d, J ) 8.6 Hz, 2H). HRMS (ES+) calcd for C31H48IN5NaO7S (M + Na)+: 784.2217. Found: 784.2205. HPLC: tR ) 12.2 min. 2-(N-Biotinoyl)-1-(N′-(13′-p-(tri-n-butylstannyl)benzoylamino)-4′,7′,10′-trio xatridecanamino)-L-valine, 7a. Following the general procedure (stirring for 30 min), 7a was obtained after silica gel chromatography (15% MeOH/EtOAc) as a colorless solid in 89% yield, mp ) 134-136 °C. 1H NMR 300 MHz (CD3OD): δ 0.92 (t, J ) 7.4 Hz, 6H), 0.97 (dd, J ) 2.6, 6.5 Hz, 6H), 1.14 (t, J ) 8.0 Hz, 3H), 1.34 (t, J ) 7.3 Hz, 6H), 1.39-1.51 (m, 4H), 1.55-1.82 (m, 10H), 1.87-1.96 (m, 2H),

Chemically Inert Biotinidase Resistant Derivatives

1.99-2.11 (m, 1H), 2.30 (t, J ) 7.3 Hz, 2H), 2.72 (d, J ) 12.9 Hz, 1H), 2.94 (dd, J ) 5.0, 12.7 Hz, 1H), 3.24 (q, J ) 7.3 Hz, 6H), 3.29-3.35 (m, 4H), 3.51 (t, J ) 6.4 Hz, 4H), 3.55-3.70 (m, 9H), 4.11 (dd, J ) 1.8, 7.7 Hz, 1H), 4.32 (dd, J ) 4.4, 7.9 Hz, 1H), 4.51 (dd, J ) 4.4, 7.9 Hz, 1H), 7.58 (d, J ) 8.2 Hz, 2H), 7.77 (d, J ) 8.2 Hz, 2H). HRMS (ES+) calcd for C44H77N5NaO7SSn (M + Na)+: 962.4463. Found: 962.4453. HPLC: tR ) 16.8 min. 2-(N-Biotinoyl)-1-(N′-(13′-p-iodobenzoylamino)-4′,7′,10′trioxatridecanam ino)-L-valine, 7b. Following the general procedure (stirring for 10 min), 7b was obtained after silica gel chromatography (15% MeOH/EtOAc) as a colorless solid in 85% yield, mp ) 144-146 °C. 1H NMR 300 MHz (CD3OD): δ 0.97 (dd, J ) 2.9, 6.7 Hz, 6H), 1.41-1.51 (m, 2H), 1.561.82 (m, 6H), 1.87-1.95 (m, 2H), 1.97-2.11 (m, 1H), 2.31 (t, J ) 7.3 Hz, 2H), 2.73 (d, J ) 12.6 Hz, 1H), 2.95 (dd, J ) 5.0, 12.7 Hz, 1H), 3.19-3.35 (m, 4H), 3.51 (q, J ) 6.3 Hz, 4H), 3.56-3.70 (m, 9H), 4.11 (dd, J ) 1.8, 7.8 Hz, 1H), 4.32 (dd, J ) 4.5, 7.9 Hz, 1H), 4.52 (dd, J ) 4.5, 7.9 Hz, 1H), 7.60 (d, J ) 8.7 Hz, 2H), 7.87 (d, J ) 8.7 Hz, 2H). LRMS (ES+) calcd for C32H51IN5O7S (M + H)+: 776.2554. Found: 776.2563. HPLC: tR ) 12.6 min. General Procedure for Synthesis of Biotin-Cyanocobalamin Derivatives 9, 10, and 11. To a solution of 48 µmol biotin amino acid adduct 15, 16, or 17, and Et3N (9 µL, 64 µmol) in anhydrous DMF (2 mL), at room temperature was added 48 µmol trifluoroacetate tetrafluorophenyl ester (TFP-OTFA). The mixture was stirred at room temperature for 0.5 h, and then 32 µmol of the cyanocobalamin-trioxadiamine adduct 27 was added. That mixture was stirred at room temperature for 0.5 h, and EtOAc (15 mL) was added and stirred for an additional 5 min. The crude product was filtered, washed with EtOAc (3 × 10 mL), and then purified by TLC plate (RP-18 F254S, 20 × 20, Merck) eluting with 1:1 MeOH/H2O. Biotin-Serine-OMe-Cyanocobalamin Conjugate, 9. Following the general procedure, 9 was obtained as a red solid in 40% yield, mp 183-186 °C. 1H NMR 500 MHz (CD3OD): d -.49 (s, 3H), 1.23 (s, 6H), 1.28-1.42 (m, 6H), 1.49 (s, 3H), 1.63-1.86 (m, 10H), 1.91 (s, 6H), 1.95-2.26 (m, 6H), 2.29 (d, J ) 3.4 Hz, 6H), 2.33-2.42 (m, 6H), 2.50 (s, 3H), 2.59 (s, 6H), 2.63 (s, 3H), 2.85-2.97 (m, 2H), 3.10-3.30 (m, 6H), 3.37 (2xs, 3H), 3.41-3.45 (m, 3H), 3.52-3.71 (m, 14H), 3.79 (dd, J ) 3.7, 12.7 Hz, 1H), 3.93 (dd, J ) 3.6, 12.6 Hz, 1H), 4.094.22 (m, 2H), 4.32-4.38 (m, 1H), 4.50-4.54 (m, 1H), 6.08 (s, 1H), 6.31 (d, J ) 3.0 Hz, 1H), 6.54 (s, 1H), 7.17 (s, 1H), 7.28 (s, 1H). HRMS (ES+) calcd for C87H129CoN18Na2O22PS (MH+2Na)+: 1945.8115. Found: 1945.8079. HPLC: tR ) 7.87.9 min. Biotin-2-Aminobutyrate-Cyanocobalamin Conjugate, 10. Following the general procedure, 10 was obtained as a red solid in 90% yield, mp 196-199 °C. 1H NMR 300 MHz (CD3OD): δ 0.48 (s, 3H), 0.96-1.03 (m, 3H), 1.22 (s, 3H), 1.28 (d, J ) 5.5 Hz, 3H), 1.34 (t, J ) 7.3 Hz, 2H), 1.40 (d, J ) 5.7 Hz, 6H), 1.49 (s, 3H), 1.59-1.82 (m, 8H), 1.91 (s, 2H), 1.97-2.26 (m, 10H), 2.30 (s, 6H), 2.41 (s, 1H), 2.51 (s, 3H), 2.53-2.61 (m, 2H), 2.64 (s, 3H), 2.69 (s, 6H), 2.75-3.19 (m, 5H), 3.24 (q, J ) 7.3 Hz, 2H), 3.30-3.38 (m, 2H), 3.41-3.46 (m, 2H), 3.52-3.73 (m, 9H), 3.79 (dd, J ) 4.5, 12.8 Hz, 1H), 3.94 (dd, J ) 3.1, 12.5 Hz, 1H), 3.91-3.97 (m, 1H), 4.08-4.22 (m, 3H), 4.31-4.35 (m, 1H), 4.51 (dd, J ) 4.5, 8.0 Hz, 1H), 4.65-4.74 (m, 2H), 6.07 (s, 1H), 6.31 (d, J ) 2.8 Hz, 1H), 6.54 (s, 1H), 7.17 (s, 1H), 7.28 (s, 1H), 7.85 (s, 1H); LRMS (ES+) calcd for C87H130N18O21PSCoNa (M + Na)+: 1907.8. Found: 1907.9. HPLC: tR ) 8.5 min. Biotin-Valine-Cyanocobalamin Conjugate, 11. Following the general procedure, 11 was obtained as a red solid in 86% yield, mp 200-203 °C. 1H NMR 300 MHz (CD3OD): δ 0.48

Bioconjugate Chem., Vol. 17, No. 6, 2006 1517

(s, 3H), 0.97 (dd, J ) 2.8, 6.9 Hz, 6H), 1.17-1.29 (m, 6H), 1.41 (d, J ) 8.7 Hz, 6H), 1.49 (s, 3H), 1.60-1.84 (m, 8H), 1.87-2.12 (m, 10H), 2.19 (s, 1H), 2.22-2.26 (m, 1H), 2.29 (s, 6H), 2.50 (s, 3H), 2.55-2.60 (m, 2H), 2.63 (s, 3H), 2.90 (s, 6H), 3.15 (t, J ) 6.4 Hz, 3H), 3.25 (t, J ) 7.2 Hz, 3H), 3.323.35 (m, 2H), 3.41 (s, 6H), 3.45 (s, 3H), 3.51-3.72 (m, 9H), 3.84-3.96 (m, 3H), 4.09-4.23 (m, 3H), 4.31-4.39 (m, 1H), 4.50-4.58 (m, 1H), 4.68-4.76 (m, 1H), 6.07 (s, 1H), 6.31 (d, J ) 2.8 Hz, 1H), 6.54 (s, 1H), 7.17 (s, 1H), 7.28 (s, 1H), 7.85 (s, 1H); LRMS (ES+) calcd for C88H133N18O21PSCo (M + H)+: 1899.9. Found: 1899.9. HPLC: tR ) 9.1 min. General Procedure for Synthesis of Biotin-Amino Acid Adducts, 15-17. A quantity of the amino acid 12 (see below), 13, or 14 dissolved in a solution of water (15 mL) and sodium hydrogen carbonate (0.50 g, 6.24 mmol), then biotin Nhydroxysuccinimide ester (0.71 g, 2.08 mmol) and acetone (15 mmol) were added respectively. The resultant solution was stirred at room temperature for 16 h, then the solution was evaporated to dryness under vacuum. The residue was redissolved in water (50 mL), and the aqueous solution was acidified to pH 2.0 by 1 N HCl. The white precipitate was filtered, washed with water (300 mL), and dried under vacuum to afford the compound 15, 16 or 17. N-Biotinoyl-L-serine-O-methyl Ether, 15. N-(tert-Butoxycarbonyl)-L-serine-O-methyl ether DCHA salt (1.0 g, 2.50 mmol) was used rather than 12. To prepare 12, the tBoc protected amino acid was dissolved in neat trifluoroacetic acid (3 mL) and stirred at rt for 10 min. After volatile materials were removed under a stream of argon, the residue was washed with ethyl acetate (3 × 20 mL), then dried under vacuum for 1 h. The dried residue was used directly in the synthesis. Following the general procedure, 15 was obtained as a colorless solid in 81% yield, mp ) 245-247 °C. 1H NMR 300 MHz (D2O/ NaOD): d 1.39-1.50 (m, 2H), 1.56-1.81 (m, 4H), 2.35 (t, J ) 7.4 Hz, 2H), 2.79 (d, J ) 13.4 Hz, 1H), 3.01 (dd, J ) 5.0, 13.1 Hz, 1H), 3.33-3.39 (m, 1 H), 3.37 (s, 3H), 3.73-3.75 (m, 2H), 4.40 (2xd, J ) 4.1, 4.3 Hz, 1 H), 4.44 (dd, 4.3, 8.0 Hz, 1H), 4.62 (dd, J ) 4.3, 8.0 Hz, 1H). HRMS (ES+) calcd for C14H23N3NaO5S (M + Na)+: 368.1256. Found: 368.1259. HPLC 4.4 min. N-Biotinoyl-2-aminobutyric Acid, 16. Following the general procedure, 16 was obtained as a colorless solid in 95% yield, mp 216-218 °C dec. 1H NMR 500 MHz (DMSO-d6): δ 0.87 (t, J ) 7.2 Hz, 3H), 1.27-1.37 (m, 2H), 1.41-1.65 (m, 4H), 1.66-1.74 (m, 1H), 2.10-2.14 (m, 2H), 2.56, 2.59 (2xs, 2H), 2.80, 2.83 (2xd, J ) 5.0 Hz, 1H), 3.07-3.12 (m, 1H), 4.074.14 (m, 2H), 4.31 (dd, J ) 4.3, 8.0 Hz, 1H), 6.35 (s, 1H), 6.40 (s, 1H), 8.02 (s, 1H). HRMS (ES+) calcd for C14H24N3O4S (M + H)+: 330.1488. Found: 330.1493. HPLC: tR ) 6.5 min. N-Biotinoyl-L-valine, 17. Following the general procedure, 17 was obtained as a colorless solid in 86% yield, mp 207209 °C dec [lit. 215-216 °C] (29). 1H NMR 300 MHz (DMSOd6): δ 0.87 (d, J ) 6.8 Hz, 6H), 1.24-1.38 (m, 2H), 1.401.68 (m, 4H), 1.97-2.08 (m, 1H), 2.16 (t, J ) 7.2 Hz, 2H), 2.57 (d, J ) 12.5 Hz, 1H), 2.82 (dd, J ) 5.0, 12.4 Hz, 1H), 3.06-3.12 (m, 2H), 4.10-4.15 (m, 2H), 4.30 (dd, J ) 5.1, 7.8 Hz, 1H), 6.38 (d, J ) 15.9 Hz, 2H), 7.91 (d, J ) 8.5 Hz, 1H). HRMS (ES+) calcd for C15H26N3O4S (M + H)+: 344.1644. Found: 344.1645. HPLC: tR ) 8.1 min. General Procedure for Syntheses of Biotin-Amino AcidTrioxadiamine Adducts, 19-21. To a solution containing 0.7 mmol of 15, 16 or 17 and 1-[3-(dimethylamino)propyl]-3ethylcarbodiimide hydrochloride (0.9 mmol) in anhydrous DMF (8 mL) was added 0.9 mmol of 18. The reaction was stirred at room temperature for 4 h to 5 days. After that period, DMF was removed under vacuum and the crude residue was purified

1518 Bioconjugate Chem., Vol. 17, No. 6, 2006

by silica column (1.5 cm × 25 cm) eluted with a MeOH/EtOAc mixture to obtain 19, 20, or 21. 2-(N-Biotinoyl)-1-(N′-(13′-tert-butoxycarbonylamino)-4′,7′,10′-trioxatrideca neamino)-L-serine-O-methyl Ether, 19. Following the general procedure (stirring for 5 d), 19 was obtained after silica gel chromatography (40% MeOH/EtOAc) as a colorless solid product in 77% yield, mp 142-144 °C. 1H NMR 300 MHz (CD3OD): δ 1.49 (s, 9H), 1.62-1.92 (m, 8H), 1.962.04 (m, 1H), 2.33-2.45 (m, 2H), 2.77 (d, J ) 12.8 Hz, 1H), 2.99 (dd, J ) 4.8, 12.8 Hz, 1H), 3.14-3.20 (m, 4H), 3.243.38 (m, 3H), 3.41 (s, 3H), 3.55-3.61 (m, 4H), 3.63-3.75 (m, 9H), 4.38 (dd, J ) 4.5, 7.7 Hz, 1H), 4.52-4.59 (m, 2H). HRMS (ES+) calcd for C29H53N5NaO9S (M + Na)+: 670.3462. Found: 670.3449. HPLC: tR ) 11.0 min. 2-(N-Biotinoyl)-1-(N′-(13′-tert-butoxycarbonylamino)benzoylamino)-4′,7′,10′-trioxatri-decaneamino)aminobutyramide, 20. Following the general procedure (stirring for 4 h), 20 was obtained after silica gel chromatography (20% MeOH/ EtOAc) as a colorless solid in 83% yield, mp 125-127 °C. 1H NMR 300 MHz (CD3OD): δ 0.99 (t, J ) 7.3 Hz, 3H), 1.47 (s, 9H), 1.60-1.87 (m, 10H), 2.31 (t, J ) 7.2 Hz, 2H), 2.74 (d, J ) 12.6 Hz, 1H), 2.96 (dd, J ) 4.9, 12.8 Hz, 1H), 3.16 (t, J ) 7.9 Hz, 2H), 3.21-3.35 (m, 4H), 3.52-3.57 (m, 4H), 3.603.69 (m, 9H), 4.17-4.25 (m, 1H), 4.34 (dd, J ) 5.0, 7.9 Hz, 1H), 4.53 (dd, J ) 5.0, 7.9 Hz, 1H). HRMS (ES+) calcd for C29H53N5NaO8S (M + Na)+: 654.3513. Found: 654.3505. HPLC: tR ) 11.4 min. 2-(N-Biotinoyl)-1-(N′-(13′-tert-butoxycarbonylamino)benzoylamino)-4′,7′,10′-trioxatri-decaneamino)-L-valine, 21. Following the general procedure (stirring for 16 h), 20 was obtained after silica gel chromatography (15% MeOH/EtOAc) as a colorless solid in 88% yield, mp 133-135 °C. 1H NMR 300 MHz (CD3OD): δ 1.08 (dd, J ) 2.3, 6.7 Hz, 6H), 1.56 (s, 9H), 1.69-1.94 (m, 8H), 2.10-2.22 (m, 1H), 2.41 (t, J ) 7.3 Hz, 2H), 2.83 (d, J ) 12.8 Hz, 1H), 3.05 (dd, J ) 4.9, 12.8 Hz, 1H), 3.25 (t, J ) 6.8 Hz, 2H), 3.30-3.46 (m, 4H), 3.62-3.67 (m, 4H), 3.70-3.78 (m, 9H), 4.21 (dd, J ) 1.8, 7.8 Hz, 1H), 4.43 (dd, J ) 4.9, 7.8 Hz, 1H), 4.62 (dd, J ) 4.9, 7.8 Hz, 1H). HRMS (ES+) calcd for C30H56N5O8S (M + H)+: 646.3850. Found: 646.3856. HPLC: tR ) 11.9 min. Radioiodination. To 100 µL of a 1 mg/mL solution of 5a, 6a, or 7a was added 1-5 µL of Na[125I]I in 0.1 N NaOH, followed by 15 µL of a 1 mg/mL solution of N-chlorosuccinimide in MeOH/5% HOAc. After 5 min, 15 µL of a 1 mg/mL aqueous sodium metabisulfite solution was added. The reaction mixture was then drawn into a syringe for purification by HPLC. The radioiodinated [125I]5b, [125I]6b, or [125I]7b was obtained from the HPLC effluent, collecting the peak that had a retention time 0.2 min later than the nonradioactive 5b, 6b, or 7b (this compensation is made due to the difference in UV and radioactivity detectors). 2-(N-Biotinoyl)-1-(N′-(13′-p-[125I]iodobenzoylamino)-4′,7′,10′-trioxatrideca neamino)-L-serine-O-methyl Ether, [125I]5b. This compound was obtained in 13% isolated radiochemical yield. 2-(N-Biotinoyl)-1-(N′-(13′-p-iodobenzoylamino)-4′,7′,10′trioxatridecaneam ino)amino-butyramide, 6b. This compound was obtained in 39% isolated radiochemical yield. 2-(N-Biotinoyl)-1-(N′-(13′-p-iodobenzoylamino)-4′,7′,10′trioxatridecaneam ino)-L-valine, 7b. This compound was obtained in 19% isolated radiochemical yield. Biotinidase Cleavage Assay. The same experimental conditions were used as previously reported (22). Briefly; the radiolabeled biotin derivative (isolated from HPLC effluent) was diluted to obtain approximately 500 000 cpm per 10 µL in a well counter. To each set of three 0.5 mL aliquots of diluted serum (1:10 with 20 mM phosphate buffer, pH 6.8) was added

Wilbur et al.

10 µL of the HPLC purified radiolabeled biotin derivative in MeOH. Each tube was lightly vortexed and placed in a 37 °C heating block. After the desired time (e.g., 2 h), the vials were removed and 40 µL of 2 mM 4-(hydroxymercuri)benzoic acid (HOHgBz), sodium salt in H2O, was added to inhibit biotinidase. To that solution was added an excess of rSAv (40 µL; 0.24 mg), and it was incubated at 37 °C for another 30 min. A 200 µL aliquot of each triplicate sample was transferred to a Microcon-30 filter and centrifuged at 3000g for 10 min. The material remaining on the filter was washed with 4 × 100 µL phosphate buffer, concentrating by centrifugation each time. The top and bottom (containing filtrate and washes) of the Microcon30 were counted in a gamma counter. The ratio of nonbound counts (bottom) to the total counts (top and bottom) was calculated to determine the percent of the biotin derivative that had been cleaved. As a control, this procedure was repeated as above except that the 4-(hydroxymercuri)benzoic acid was added prior to the radiolabeled biotin derivative to inhibit biotinidase activity. HPLC Biotin Derivative Dissociation Measurements. The same experimental conditions were used as previously reported (26). Briefly, to 400 µL of a 0.5 mg/mL solution of rSAv (200 µg; 3.8 nmol) in a microcentrifuge vial was added ∼4 equiv (16 nmol) of a biotin conjugate (8, 9, or 10) dissolved in 20 µL of a 10% aqueous DMSO solution. The rSAv/biotin derivative mixture was incubated for 1 h at room temperature, and 100 µL of the solution was injected onto the SE-HPLC column and eluted to assay binding. To the remaining avidin/biotin derivative solution was added 60 µL of a 0.575 mg/mL solution of biotin (34.5 µg; 141 nmol) in 10% aqueous DMSO. That solution was incubated for 1 h, and 100 µL was removed for analysis by HPLC. The remaining solution was incubated for an additional 2-3 h, and then another 100 µL was removed and analyzed by HPLC.

RESULTS Synthesis of Biotin Derivatives. Biotin derivatives that contain an unprotected biotinamide bond (1a/b), a N-methylsubstituted biotinamide bond (2a/b), a carboxylate R to the biotinamide bond (3a/b), and a hydroxymethyl group R to the biotinamide bond (4a/b) were prepared as previously described (23). The synthetic route to obtain biotin derivatives that contain a methoxymethyl (5a/b), an ethyl (6a/b), or an isopropyl (7a/ b) group R to the biotinamide bond is shown in Scheme 1. 2-Aminobutyric acid, 13, and valine, 14, are commercially available, but the serine-O-methyl ether, 12, was not readily obtained, so an alternate, the N-t-Boc protected derivative, was used instead. The t-Boc protecting group was readily removed to provide 12, and all three amino acids (12, 13, 14) were conjugated with biotin N-hydroxysuccinimidyl ester to give the desired adducts in 81-95% yields. The resulting biotin derivatives 15, 16, and 17 were coupled with mono-t-Boc protected trioxadiamine, 18, to provide adducts 19, 20, and 21 in 7788% yields. Deprotection of 19, 20, and 21 was essentially quantitative in neat trifluoroacetic acid (TFA) to provide the free amine derivatives 22, 23, and 24. The biotin-amine derivatives were used without isolation in reactions with benzoate esters 25 and 26. Iodobenzamide derivatives 5b, 6b, and 7b, used as HPLC standards, were prepared in 83-88% yield by reaction of p-iodobenzoate tetrafluorophenyl (TFP) ester 26 with 22, 23, or 24 (respectively). Similarly, the stannylbenzoate-containing biotin derivatives 5a, 6a, and 7a, used as intermediates for radioiodination, were prepared in 87-91% yields by reaction of the tri-n-butylstannylbenzoate TFP ester, 25, with 22, 23, or 24. Four biotin-cyanocobalamin derivatives (8, 9, 10, 11) were prepared for evaluation of their relative binding with rSAv.

Bioconjugate Chem., Vol. 17, No. 6, 2006 1519

Chemically Inert Biotinidase Resistant Derivatives

Scheme 1. Synthetic Route for Preparation of Iodinated and Radioiodinated Biotin Derivatives 5b, 6b, and 7ba

a (a) H2O, Na2CO3, 12, 13, or 14, acetone, rt, 16 h; (b) DMF, EDC, 15, 16, or 17, 4 h to 5 d; (c) TFA, rt, 10 min; (d) DMF, Et3N, 25 or 26, rt, 10 min to 1 h; (e) MeOH/5% HOAc, 5a, 6a, or 7a, NCS, 2 min.

Scheme 2. Reaction for Preparation of Biotin-Cyanocobalamin Derivatives 9, 10, and 11a

a

(a) DMF, 15, 16, or 17, TFP-OTFA, rt, 0.5 h; (b) 27, rt, 30 min.

Biotin cyanocobalamin 8 was prepared as previously described (26). Biotin-cyanocobalamin derivatives 9, 10, and 11 were prepared using the same synthetic route, as shown in Scheme 2. Reaction of biotin adducts 15, 16, and 17 with trifluoroacetate tetrafluorophenyl ester (TFA-OTFP) to form the corresponding biotin TFP esters in situ, followed by reaction with the trioxadiamine adduct of cyanocobalamin 27 provided 9, 10, and 11 in 40%, 90%, and 86% yield (respectively). Cyanocobalamin derivative 27 was prepared by reaction of 1,13-diamino-4,7,10-trioxatridecane as previously described (26). Radioiodination of Biotin Derivatives. The radioiodinated derivatives [125I]1b, [125I]2b, [125I]3b, [125I]4b, [125I]5b, [125I]6b, and [125I]7b were prepared in 13-80% yield from the corresponding tri-n-butylstannylbenzamide derivatives 1a, 2a,

3a, 4a 5a, 6a, or 7a in MeOH/1% HOAc using N-chlorosuccinimide (NCS) as the oxidant as previously described (22, 23). The radioiodinated derivatives were purified by collection from the HPLC effluent. Identification was accomplished by coinjection of the isolated radioiodinated derivative with the HPLC standards. The radiochemical yields were not optimized, as only a small amount of material was required for the biotinidase evaluations. Assessment of Biotinidase Activity. In this study, as in our previous studies, incubation of the radioiodinated samples in a 1 to 10 dilution of fresh human serum was used to assess biotinidase activity. Cleavage of the biotinamide bond is demonstrated as a lack of binding of radioactivity with rSAv. In the experiment, parallel sets of triplicate samples (for most

1520 Bioconjugate Chem., Vol. 17, No. 6, 2006

Wilbur et al.

Table 1. Estimates of Biotinamide Cleavage in Biotin Derivativesa compound no. [blocking group]

% boundb (serum)

% boundc (serum+HOHgBz)

% cleavedd by biotinidase

1b [H] 2b [N-CH3] 3b [CO2H] 4b [CH2OH] 5b [CH2OCH3] 6b [CH2CH3] 7b [CH(CH3)2]

8.34 ( 0.50 96.93 ( 0.07 92.47 ( 0.04* 86.92 ( 0.07 90.57 ( 0.21 92.92 ( 0.10 91.84 ( 0.31

98.52 ( 0.41 97.76 ( 0.23 92.57 ( 0.14 87.13 ( 0.07 91.43 ( 0.05* 93.42 ( 0.10 91.66 ( 0.16

91.5 0.8 0.1 0.2 0.9 0.5 0.0

a Values were obtained from percent [125I]iodophenyl-biotin derivative bound to rSAv after incubation with human serum or human serum treated with 4-(hydroxymercuri)benzoic acid (HOHgBz). All % bound values represent the percentage of total radioactivity remaining in a Microcon-30 centrifugation unit (bound with rSAv or nonspecifically bound) after repeated (3×) centrifugations/washes at 2000 rpm for 5 min. Values are an average of three analyses ( SD, except those denoted by * which are from two analyses. Human serum was obtained on the day of the experiments. b Radiolabeled biotin was placed in 10% human serum for 2 h at 37 °C, rSAv was added, and an aliquot was spun in Microcon-30 centrifugation unit. c Radiolabeled biotin was placed in HOHgBz-treated 10% human serum for 2 h at 37 °C, rSAv was added, and an aliquot was spun in a Microcon-30 centrifugation unit. d Percentages are derived from the formula (100 - [% bound for noninhibited/% bound after HOHgBz inhibition] × 100]). Values for percent biotinamide cleaved that are less than 1% may be reflective of experimental error and not actual biotinidase cleavage.

compounds) either were treated with the sulfhydryl-reactive HOHgBz to block biotinidase activity (as a control) or were untreated to measure biotinidase cleavage. All samples were incubated at 37 °C for 2 h, then samples previously untreated with HOHgBz were treated with that reagent to quench the enzyme activity. Following that, rSAv was added to all samples and they were incubated at 37 °C for 30 min. Ultrafiltration was then conducted on each sample using Microcon-30 filters to separate the rSAv-bound radioactivity from the nonbound radioactivity. The results obtained are shown as percent bound radioactivity in Table 1. The control group (serum + HOHgBz) provides a value for the amount of radiolabeled biotin that will bind with rSAv under the conditions of the study. In theory, this value should be 100%, but other factors may cause a decrease in the binding of the radiolabeled biotin derivatives with rSAv. One such factor is production of biotin sulfoxide under the radioiodination conditions (30). To assess the percent biotinamide bond cleaved, the percent bound values obtained after incubation with serum is subtracted from the control binding values (serum + HOHgBz). At least 90% of the unprotected biotinamide bond in 1b was cleaved by biotinidase, demonstrating the activity of the enzyme. All of the other samples evaluated had less than 1% cleaved, indicating that they are very stable to biotinidase activity. It is likely that the 4 equiv) of the biotin-cyanocobalamin reagent prior to the competitive reaction so that both protein-bound activity and free biotin derivative could be seen on the chromatogram (UV trace). Thus, the percent bound values observed at t ) 0 are less than 100%. It should be noted that the variance in free biotin-CN-Cbl will potentially askew the values at the later time points, so caution must be used in comparing data between studies. However, the data in Table 2 clearly show the differences in dissociation of the biotin-CNCbl derivatives tested. The N-methylbiotinamide derivative 8 was included in the assay as a control to compare with the rates of dissociation of 9, 10, and 11. The data indicate that the release of the R-isopropyl (biotin-valine) derivative 11 is slower than

Table 2. Percent Biotin-Cyanocobalamin Derivatives 8, 9, 10, and 11 Bound with rSAv before and after Challenged by a Large Excess of Biotina compound no. [blocking group]

t)0 (%)b

t)1h (%)

t)3h (%)

t)5h (%)

8 [N-CH3] 8 [N-CH3]c 9 [CH2OCH3]c 10 [CH2CH3] 11 [CH(CH3)2]

86 100 89 95 93

42 53 77 86 65

11 15 56 74 42

4 4 47 62 32

a Biotin derivative (4+ equiv) was incubated for 1 h with rSAv (t ) 0), 50 equiv of biotin was added, and aliquots of the mixture were taken at 1, 3, and 5 h for analysis by size exclusion HPLC. Samples were covered from light and kept at room temperature during the experiment. Values from t ) 0 are not 100% due to an excess of the reagent being present. b Percent values were obtained from analysis of peak areas from size exclusion HPLC chromatograms. UV traces were obtained for 280 and 362 nm, where the protein and vitamin B12, respectively, have strong absorbances. The % bound values were obtained from the 362 nm traces and represent the [area of protein-bound peak divided by the area of protein bound peak + free biotin derivative] × 100. c Analysis of binding run at a separate time from cyanocobalamin derivatives 8, 9, and 10. Cyanocobalamin derivative 8 was run a second time as control.

the N-Me (biotin-sarcosine) derivative 8, but is faster than the R-methylmethyl ether and the smaller R-ethyl derivative 10 (i.e., dissociation rates are: 8 > 11 > 9 > 10).

DISCUSSION Animals and humans have significant amounts of biotinidase in serum and tissues (8), raising the possibility that biotinylated compounds used in vivo will be degraded (biotin released) before the intended binding with avidin or streptavidin can occur. For this reason, all of the biotin derivatives prepared for in vivo application in our laboratory incorporate biotinidasestabilizing functionalities as a precautionary measure to assure that no cleavage (release) of the biotin occurs. It should be noted that there may be in vivo applications where biotinidase stabilization is not required, or even desirable. For example, in some applications, modification for biotinidase stabilization may impart undesirable properties to biotinylated molecules, such as increasing the lipophilicity. Additionally, there may be instances where biotinidase cleavage offers an advantage for blood and tissue clearance, particularly if an avidin or streptavidin binding of the biotinylated compound postinjection is not required. Another example where biotinidase stabilization may not be required is in vivo use of biotinylated proteins such as monoclonal antibodies. In studies of biotinylated antibodies in this laboratory no difference was observed in avidin column capture of monoclonal antibodies biotinylated with biotinidase stabilized or nonstabilized biotin reagents after subjecting to human serum (unpublished results). These results do not necessarily indicate that biotinidase is incapable of acting on biotinylated antibodies, but rather that the rate of cleavage from the (usual) multi-biotinylated protein is slow enough that the biotinylated protein can be captured on an avidin column after incubation in serum. In fact, published data suggests that intact proteins can be debiotinylated by biotinidase (9), and biotinidase stabilizing groups can block the cleavage (31). It became apparent in our early studies that biotinidaseblocking functional groups were essential for optimal pretargeting of therapeutic radionuclides. In those studies, we identified biotin conjugates that contain carboxylate and hydroxymethyl functionalities on the carbon R to the biotinamide bond as derivatives of choice for in vivo application (25). However, later synthetic studies to prepare dendrimers and dendrons with multiple biotin conjugates containing these functionalities were fraught with difficulties. Importantly, in some examples, unexpected side reactions prevailed, and in other

Bioconjugate Chem., Vol. 17, No. 6, 2006 1521

Chemically Inert Biotinidase Resistant Derivatives

examples, removal of the t-Bu protecting group was problematic. Further, the cost of the protected amino acids (e.g., aspartate t-Bu ester and serine-O-t-Bu ether) used in the syntheses was perceived as a potential roadblock to their widespread application. These shortcomings led us to consider alternate biotinidaseblocking functionalities that were inert to most chemical transformations and could be obtained at a lower cost. A primary goal of the studies was to identify biotinidaseblocking moieties that were inert to most reaction conditions, such that they did not have to be protected and deprotected in the syntheses of targeted biotin derivatives. We reasoned that it was best to incorporate functional groups just large enough to block the biotinidase without dramatically affecting avidin/ streptavidin binding or other properties of the conjugates. A nonreactive derivative previously tested had a methyl group R to the biotinamide bond. Unfortunately, the methyl group did not fully block biotinidase cleavage, only decreased its rate of cleavage. Since it appeared that the R-methyl group was not large enough to fully block biotinidase activity, slightly larger nonreactive functional groups adjacent to the biotinamide were sought. We chose to incorporate 2-aminobutyric acid and valine as biotin adducts as those amino acid adducts have either an ethyl or isopropyl group R to the biotinamide bond. The ethyl and isopropyl group are inert to most reaction conditions and are similar in steric size of a hydroxymethyl and carboxylate functionality (respectively), which have been previously shown to be stable to biotinidase activity. We also chose to include the biotin adduct of serine-O-methyl ether in the study, as it is nonreactive under most reaction conditions and has a similar steric size but would be expected to be less lipophilic. Radioiodinated derivatives of the serine-O-methyl ether, 2-aminobutyric acid, and valine adducts of biotin, ([125I]5b, [125I]6b, and [125I]7b) were prepared and evaluated in a biotinidase assay. As controls, radioiodinated derivatives of biotin without an amino acid blocking group ([125I]1b) or containing the amino acids sarcosine, aspartate, and serine ([125I]2b, [125I]3b, [125I]4b) were also prepared and tested. All of the biotin derivatives containing an amino acid conjugated with the biotin carboxylate were found to be very stable toward biotinidase cleavage (Table 1). On the basis of the theory that blockage of the enzyme is controlled by steric bulk, one might expect the steric isopropyl group in valine to block biotinidase more effectively. However, under the conditions studied there was essentially no difference between the compounds. More concentrated serum samples or longer incubation times may be required to determine if there are small differences in the biotinidase-blocking capacity of the various derivatives. It should be noted that biotin derivatives of interest for our studies have a very short half-life in blood (i.e., 4 min in R-phase and 1 h in β-phase (33)], making the 2 h incubation time adequate to assess their in vivo use. Biotin derivatives of molecules that have long biological half-lives in blood may require further studies of biotinidase stability. In addition to biotinidase stability, we believe it is very important to retain the high binding affinity in the biotin derivatives used in vivo to alleviate the potential for displacement from streptavidin by endogenous biotin (34, 35). While there are methods for decreasing biotin concentrations in blood, it appears that some biotin will always be present to compete for the biotin-binding pockets on SAv. To assess the relative retention of biotin derivatives bound with rSAv, a simple sizeexclusion HPLC assay has been developed (26). In that assay, 4 equiv of a biotin-CN-Cbl derivative are bound with rSAv and then challenged with a large excess (50 equiv) of biotin. Under these conditions, any biotin-CN-Cbl derivative that is dissociated will be replaced with biotin. Thus, the relative rate of dissociation of a biotin derivative can be estimated by the quantity of free biotin-CN-Cbl derivative versus rSAv-bound

biotin-CN-Cbl derivative. The use of biotin-CN-Cbl derivatives makes HPLC analysis possible, as UV detection of the protein can be measured at 280 nm and the biotin-cyanocobalamin derivative at 362 nm, where there is very little absorption by the protein. Significant differences in the rate of dissociation were measured between the biotin derivatives (Table 2). As in previous studies, nearly all biotin-sarcosine-CN-Cbl, 8, is displaced within 5 h at room temperature. The biotinvaline-CN-Cbl, 11, has the second fastest rate of dissociation, with the biotin-serine-O-methyl ether-CN-Cbl, 9, a little slower, and the biotin-2-aminobutyric acid-CN-Cbl, 10, having the slowest dissociation. It is interesting to note that the rate of dissociation correlates with the steric bulk of the substituent on the carbon R to the biotinamide bond. While the differences in amount of biotin-amino acid adduct dissociated by 5 h is fairly large, this may not be important for in vivo application. Indeed, it is likely that all three new biotin derivatives would provide adequate binding to be useful in most in vivo applications. This statement is made based on the fact that highly successful pretargeting of rSAv conjugates on tumors in vivo has been accomplished with a biotin-sarcosine derivative (36), which has the fastest dissociation from SAv out of the compounds studied.

CONCLUSION Problems encountered in prior studies employing t-Bu protected hydroxyl and carboxylate biotinidase stabilizing groups led to this investigation to develop alternate biotinidase stabilizing functionalities. In the investigation, new functionalized biotin derivatives were targeted that retained criteria set previously, i.e., that they effectively block biotinidase while retaining high binding affinity with rSAv but had the additional criteria that they (1) are chemically inert, (2) do not require protection/ deprotection in the syntheses, and (3) have a low cost such that they can be applied widely. While all three of the biotin-amino acid derivatives tested were found to be stable to biotinidase, the cost of serine-O-methyl ether and difficulties encountered in syntheses using it makes that derivative less attractive than the 2-aminobutyric acid and valine adducts. The data obtained indicate that biotin-valine adduct could be used for in vivo studies, but higher lipophilicity and faster dissociation than that of the 2-aminobutyric acid adduct make it a second choice. The low cost and high binding affinity of the biotin-2-aminobutyric acid derivative make it particularly attractive for development of reagents to be used in vivo. While there is little doubt that other untested functional groups can also provide biotinidase blockage, it seems likely that a relatively small number would meet all of the criteria set. It is anticipated that a majority of our subsequent synthetic studies with biotin derivatives will incorporate the 2-aminobutyric acid moiety to block biotinidase.

ACKNOWLEDGMENT We are grateful for the financial support provided by the Department of Energy, Medical Applications and Biophysical Research Division, OHER (DE-FG03-98ER62572), and National Cancer Institute, NIH (5 RO1 CA113431). We thank Quanta BioDesign (Powell, OH) for providing the biotin N-hydroxysuccinimide ester as a gift.

LITERATURE CITED (1) Goodwin, D. A., and Meares, C. F. (1997) Pretargeting: General Principles. Cancer (Suppl.) 80, 2675-2680. (2) Zhu, H., Jain, R. K., and Baxter, L. T. (1998) Tumor Pretargeting for Radioimmunodetection and Radioimmunotherapy. J. Nucl. Med. 39, 65-76. (3) Stoldt, H. S., Aftab, F., Chinol, M., Paganelli, G., Luca, F., Testori, A., and Geraghty, J. G. (1997) Pretargeting Strategies for Radio-

1522 Bioconjugate Chem., Vol. 17, No. 6, 2006 immunoguided Tumour Localisation and Therapy. Eur. J. Cancer 33, 186-192. (4) Wright, L. D., Driscoll, C. A., and Boger, W. P. (1954) Biocytinase, an enzyme concerned with hydrolytic cleavage of biocytin. Proc. Soc. Exp. Biol. Med. 86, 335-7. (5) Wolf, B., Hymes, J., and Heard, G. S. (1990) Biotinidase Methods Enzymol. 184, 103-111. (6) Oizumi, J., and Hayakawa, K. (1993) Biocytin-specific 110-kDa biotinidase from human serum. Clin. Chim. Acta. 215, 63-71. (7) Hymes, J., Fleischhauer, K., and Wolf, B. (1997) Biotinidase in Serum and Tissues. Methods Enzymol. 279, 422-434. (8) Pispa, J. (1965) Animal Biotinidase. Ann. Med. Expl. Biol. Fenniae 43, 3-40. (9) Ballard, T. D., Wolff, J., Griffin, J. B., Stanley, J. S., van Calcar, S., and Zempleni, J. (2002) Biotinidase catalyzes debiotinylation of histones. Eur. J. Nutr. 41, 78-84. (10) Chauhan, J., and Dakshinamurti, K. (1988) Role of Human Serum Biotinidase as Biotin-Binding Protein. Biochem. J. 256, 265-270. (11) Stanley, C. M., Hymes, J., and Wolf, B. (2004) Identification of alternatively spliced human biotinidase mRNAs and putative localization of endogenous biotinidase. Mol. Genet. Metab. 81, 300-12. (12) Zempleni, J. (2005) Uptake, localization, and noncarboxylase roles of biotin. Annu. ReV. Nutr. 25, 175-96. (13) Hayakawa, K., Yoshikawa, K., Oizumi, J., and Yamauchi, K. (1997) Determination of Biotinidase Activity with Biotinyl-6aminoquinoline as Substrate. Methods Enzymol. 279, 435-442. (14) Wolf, B., and McVoy, J. S. (1983) A sensitive radioassay for biotinidase activity: deficient activity in tissues of serum biotinidasedeficient individuals. Clin. Chim. Acta 135, 275-281. (15) Livaniou, E., Kakabakos, S. E., Evangelatos, S. A., Evangelatos, G. P., and Ithakissios, D. S. (1997) Determination of Serum Biotinidase Activity with Radioiodinated Biotinylamide Analogs. Methods Enzymol. 279, 442-451. (16) Chauhan, J., and Dakshinamurti, K. (1986) Purification and Characterization of Human Serum Biotinidase. J. Biol. Chem. 261, 4268-4275. (17) Nilsson, L., and Kagedal, B. (1994) Lipoamidase and biotinidase activities in the rat: tissue distribution and intracellular localization. Eur. J. Clin. Chem. Clin. Biochem. 32, 501-9. (18) Nilsson, L., and Kagedal, B. (1993) Co-purification of human serum lipoamidase and biotinidase: evidence that the two enzyme activities are due to the same enzyme protein. Biochem. J. 291 (Pt 2), 545-51. (19) Gustavson, L. M., Srinivasan, A., Fritzberg, A. R., Reno, J. M., and Axworthy, D. B. (1994) NeoRx Corporation, US Patent 5,283,342. (20) Rosebrough, S. F. (1993) Plasma Stability and Pharmacokinetics of Radiolabeled Deferoxamine-Biotin Derivatives. J. Pharmacol. Exp. Ther. 265, 408-415. (21) Foulon, C. F., Alston, K. L., and Zalutsky, M. R. (1998) Astatine211-Labeled Biotin Conjugates Resistant to Biotinidase for Use in Pretargeting Radioimmunotherapy. Nucl. Med. Biol. 25, 81-88. (22) Wilbur, D. S., Hamlin, D. K., Pathare, P. M., and Weerawarna, S. A. (1997) Biotin reagents for antibody pretargeting. Synthesis, radioiodination and in vitro evaluation of water soluble, biotinidase resistant biotin derivatives. Bioconjugate Chem. 8, 572-584.

Wilbur et al. (23) Wilbur, D. S., Hamlin, D. K., Chyan, M. K., Kegley, B. B., and Pathare, P. M. (2001) Biotin reagents for antibody pretargeting. 5. Additional studies of biotin conjugate design to provide biotinidase stability. Bioconjugate Chem. 12, 616-23. (24) Axworthy, D. B., Theodore, L. J., Gustavson, L. M., and Reno, J. M. (1997) NeoRx Corporation, US Patent 5,608,060. (25) Wilbur, D. S., Chyan, M.-K., Pathare, P. M., Hamlin, D. K., Frownfelter, M. B., and Kegley, B. B. (2000) Biotin Reagents for Antibody Pretargeting. 4. Selection of Biotin Conjugates for in Vivo Application Based on Their Dissociation Rate from Avidin and Streptavidin. Bioconjugate Chem. 11, 569-583. (26) Wilbur, D. S., Pathare, P. M., Hamlin, D. K., Frownfelter, M. B., Kegley, B. B., Leung, W. Y., and Gee, K. R. (2000) Evaluation of biotin-Dye conjugates for use in an HPLC assay to assess relative binding of biotin derivatives with avidin and streptavidin. Bioconjugate Chem. 11, 584-98. (27) Wilbur, D. S., Hamlin, D. K., Vessella, R. L., Stray, J. E., Buhler, K. R., Stayton, P. S., Klumb, L. A., Pathare, P. M., and Weerawarna, S. A. (1996) Antibody fragments in tumor pretargeting. Evaluation of biotinylated Fab′ colocalization with recombinant streptavidin and avidin. Bioconjugate Chem. 7, 689-702. (28) Wilbur, D. S., Hamlin, D. K., Buhler, K. R., Pathare, P. M., Vessella, R. L., Stayton, P. S., and To, R. (1998) Streptavidin in antibody pretargeting. 2. Evaluation of methods for decreasing localization of streptavidin to kidney while retaining its tumor binding capacity. Bioconjugate Chem. 9, 322-330. (29) Bhuniya, S., Park, S. M., and Kim, B. H. (2005) Biotin-amino acid conjugates: an approach toward self-assembled hydrogelation. Org. Lett. 7, 1741-4. (30) Wilbur, D. S., Hamlin, D. K., and Chyan, M.-K. (2005) Addition of Methionine to Radiohalogenation Reactions of Biotin Derivatives Decreases Formation of Biotin Sulfoxide by Oxidants. J. Labelled Compd. Radiopharm. 48, S305 (abstract). (31) Bogusiewicz, A., Mock, N. I., and Mock, D. M. (2005) A biotinprotein bond with stability in plasma. Anal. Biochem. 337, 98-102. (32) Visintin, C., Aliev, A. E., Riddall, D., Baker, D., Okuyama, M., Hoi, P. M., Hiley, R., and Selwood, D. L. (2005) Membrane receptor probes: solid-phase synthesis of biotin-Asp-PEG-arvanil derivatives. Org. Lett. 7, 1699-702. (33) Fritzberg, A. R., Beaumier, P. L., Bottino, B. J., and Reno, J. M. (1994) Approaches to improved antibody- and peptide-mediated targeting for imaging and therapy of cancer. J. Controlled Release 28, 167-173. (34) Rusckowski, M., Fogarasi, M., Fritz, B., and Hnatowich, D. J. (1997) Effect of Endogenous Biotin on the Applications of Streptavidin and Biotin in Mice. Nucl. Med. Biol. 24, 263-268. (35) Wang, H., and Pevsner, J. (1999) Detection of endogenous biotin in various tissues: novel functions in the hippocampus and implications for its use in avidin-biotin technology. Cell Tissue Res. 296, 511-6. (36) Theodore, L. J., Fritzberg, A. R., Schultz, J. E., and Axworthy, D. B. (2000) Evolution of a Pretarget Radioimmunotherapeutic Regimen. In Radioimmunotherapy of Cancer (Abrams, P. G., and Fritzberg, A. R., Eds.) pp 195-221, Marcel Dekker, New York. BC060084M