SEPTEMBER/OCTOBER 2003 Volume 14, Number 5 © Copyright 2003 by the American Chemical Society
COMMUNICATIONS A Novel Approach for Affinity-Based Screening of Target Specific Ligands: Application of Photoreactive D-Glyceraldehyde-3-phosphate Dehydrogenase Masaki Kaneda, Yutaka Sadakane, and Yasumaru Hatanaka* Faculty of Pharmaceutical Sciences, Toyama Medical and Pharmaceutical University, 2630 Sugitani, Toyama 930-0194, Japan. Received April 1, 2003
A novel application of the photoaffinity technique has been developed for the efficient discovery of small ligand and macromolecule interaction. The approach, photoaffinity capture, uses a photoreactive protein together with immobilized ligand for the rapid screening of competitive inhibitors. The set of photoreactive glyceraldehyde-3-phosphate dehydrogenase (photo-GAPDH) and immobilized dye ligand was prepared and examined as a model system. The photo-GAPDH was shown to efficiently capture the immobilized ligand. When nonimmobilized competitive ligands were included in the system, the capture was prevented in accordance with the affinity of the ligands. The present approach would provide an efficient tool for affinity-based screening of ligand libraries.
Photoaffinity labeling is a useful method for the elucidation of molecular interaction processes (1). The method usually utilizes a photoreactive small ligand as a probe, which introduces cross-links on the target macromolecule (2-7). Photoreactive macromolecules such as proteins (8, 9) or oligonucleotides (10) have also been developed for probing the interactions between biomacromolecules. However, this macromolecular probe approach has never been applied for capturing ligands that specifically enter into the internal ligand binding sphere of photoreactive macromolecules. Here, we report a novel application of a photoreactive protein for the efficient discovery of its partner ligands. The approach uses a photoreactive protein and an im* To whom correspondence should be addressed. Phone: +81-76-434-7515. Fax: +81-76-434-5063. E-mail: yasu@ ms.toyama-mpu.ac.jp.
mobilized ligand for achieving efficient inhibitor screening. The protein has a photoreactive group, a diazirine, within the active site for enabling a covalent capture of a ligand (Figure 1A). The capture process can then be prevented with an inhibitor that has enough affinity for the active site (Figure 1B). Thus, the photoaffinity capture array composed of a known set of protein and ligand would provide a simple and efficient device for finding new ligands. We have examined this approach using GAPDH as a protein mold, and Cibacron Blue 3GA as an affinity ligand to be immobilized. GAPDH is a homo-tetramer protein that catalyzes the oxidation of glyceraldehyde-3-phosphate with concomitant reduction of β-nicotinamide adenine dinucleotide (NAD+) (11). Each subunit of GAPDH has an active site composed with a catalytic free sulfhydyl group Cys149 (12). The method for the selective modification of the active site Cys149 is already established (13, 14). Thus, a
10.1021/bc0340520 CCC: $25.00 © 2003 American Chemical Society Published on Web 07/26/2003
850 Bioconjugate Chem., Vol. 14, No. 5, 2003
Figure 1. General approach of photoaffinity capture. (A) The covalent capture of target-specific ligand by photoreactive protein. (B) Application to the screening of inhibitory ligands.
Kaneda et al.
Figure 2. UV-vis spectra of photo-GAPDH. The UV-vis spectrum (220-400 nm) of modified GAPDH was recorded in pyrophosphate buffer pH 7.5 containing 0.1 mM EDTA. The inset shows the spectral region of 320-400 nm, which was obtained with a concentrated solution of photo-GAPDH.
sulfhydryl group selective reagent 2 was designed and synthesized for the introduction of a useful carbene precursor, a diazirine (1), within the active site of GAPDH. The thiosulfonate group of the reagent is known to form a disulfide linkage with the sulfhydryl group of protein (15).
The reagent 2 was readily synthesized from a known diazirine bromide 1 (16) by an established derivatization method (15). Since the bound NAD+ of holo-GAPDH is known to inhibit the modification of the active site sulfhydryl group (17), the cofactor was removed before the modification reaction (18). The apo-GAPDH was then reacted with compound 2 for 1 h at room temperature to give photoreactive GAPDH (photo-GAPDH). After a gel filtration chromatography, the GAPDH concentration of the pooled protein fraction was determined from UV absorption at 278 nm (19) and also by Bradford protein assay (20). The diazirine concentration was independently obtained from its characteristic n to π* absorption (λmax 360, ) 310) as shown in Figure 2 (inset). The extent of modification was calculated as 0.72 mol of diazirine per subunit of GAPDH. For the preparation of ligand immobilized matrix, a commercially available preactivated ligand, Cibacron Blue 3GA, was used. Cibacron Blue is an affinity ligand that specifically binds to the nucleoside phosphate binding sites of dehydrogenases and kinases (21). Since GAPDH has a NAD+ binding site near to the Cys149 in active site (17), Cibacron Blue should locate close to the diazirine of photo-GAPDH. Cibacron Blue 3GA was immobilized on a surface aminated tube that could be easily prepared by the graft polymerization of methacrylic acid on the inner wall of polypropylene tube (2224) followed by the amino scaffold coupling (Supporting Information). The Cibacron Blue immobilized tube was first used to evaluate the ligand-capturing ability of photo-GAPDH. The photo-GAPDH was incubated in the
Figure 3. Detection of cross-linked GAPDH onto immobilized Cibacron Blue ligand. (A) The photo-GAPDH was incubated in the ligand immobilized tube and examined. The GAPDH trapped on the solid matrix was visualized after biotinylation. Irradiated (spot a). Not irradiated (spot b). Irradiated then treated with 2-mercaptoethanol (spot c). (B) SDS-PAGE analysis of GAPDH without (lanes 1, 2, and 4) or with irradiation (lane 3). The affinity bound GAPDH on the immobilized Cibacron Blue 3GA was recovered with 1% SDS (lane 1). Experimental conditions were the same as the sample in lane 1 except that the Cibacron ligand was not attached on the matrix (lane 2). The photochemically cross-linked GAPDH on to the immobilized Cibacron Blue 3GA was recovered after reducing disulfide linkage with 2-mercaptoethanol (lane 3). Experimental conditions were the same as for the sample in lane 3 except that irradiation was not applied (lane 4).
Cibacron tube for 1 h at 37 °C and irradiated at 0 °C for 10 min. After the photolysis, the tube was completely washed with 1% sodium dodecyl sulfate (SDS) to remove untrapped GAPDH. The tube was then reacted with biotin N-hydroxysuccinimide and subjected for the chemiluminescent detection of trapped protein as shown in Figure 1. The chemiluminescence generated with a commercial kit (Wako Pure Chemical Industries, Ltd., Japan) was detected on an image analyzer (ChemiDoc, Bio-Rad Laboratries, Inc.). The trapped GAPDH was clearly visualized as a spot when the photoaffinity capture was applied (Figure 3A, spot a). The protein was not detected when the irradiation was not applied (spot b) or the irradiated tube was treated with 2-mercaptoethanol (spot c). The latter control experiment suggests that the trapped GAPDH could be released from the matrix by cleaving the S-S linkage. To confirm the result, the GAPDH was analyzed
Communications
Bioconjugate Chem., Vol. 14, No. 5, 2003 851
Figure 4. Photoaffinity capture in the presence of competitive small ligands. The GAPDH photoaffinity capture system was examined in the presence of Cibacron Blue 3GA (column a), ATP (column b). The concentration of additives was 100 µM (row 1) and 600 µM (row 2), respectively. No competitor was added to the control. Table 1. The Inhibition of Photoaffinity Capture in the Presence of Competitive Small Ligands. inhibition (% ( S. D.)a competitor
100 µM
600 µM
Cibacron Blue NAD+ ATP D-glyceraldehyde-3-phosphate L-tyrosine-O-phosphate NADH NADP+ NADPH
94.5 ( 1.5 4.5 ( 7.8 51.7 ( 6.2 13.3 ( 5.3 3.8 ( 5.5 4.3 ( 8.2 4.1 ( 11.3 3.8 ( 2.6
95.9 ( 1.2 14.1 ( 5.4 92.9 ( 1.9 62.7 ( 6.8 6.3 ( 5.3 12.8 ( 12.2 13.4 ( 5.3 12.9 ( 7.6
a The value is the average of 15 assays, and the errors are given in SD values.
with SDS-PAGE (Figure 3B). First, the photo-GAPDH was examined without the irradiation. The lane 1 demonstrate that the modification at the sulfhydryl group does not significantly reduce the recognition of Cibacron Blue. The affinity bound GAPDH was recovered from the matrix with 1% SDS and was resolved at 37 kD with the expected subunit molecular weight (lane 1), whereas no GAPDH was detected from the matrix without Cibacron Blue ligand (lane 2). The conditions for lanes 3 and 4 were the same as those for spots c and b of Figure 3A, respectively. When the irradiation was applied, the covalently trapped GAPDH on the matrix was detected only after cleaving the disulfide bond between the crosslinked moiety and Cys149 (lane 3). In contrast to this, the control lane 4 showed that no GAPDH was trapped without irradiation. From the densitometric analysis, the amount of captured protein in lane 3 was determined as 36% of the bound GAPDH (lane 1). The photoaffinity capture system of photo-GAPDH and immobilized Cibacron Blue was shown to have desired nature for progression to the competitive inhibitor screening. The GAPDH photoaffinity capture system was examined in the presence of known inhibitors, Cibacron Blue 3GA and ATP, at the concentration of 100 µM (row 1) and 600 µM (row 2), respectively. The addition of 100 µM Cibacron Blue 3GA was enough for preventing the capture of GAPDH (column a row 1). ATP, a well-known inhibitor of GAPDH (25), inhibited the capture of GAPDH in a dose dependent manner as shown in the column b of Figure 4. The inhibition with various small ligands is summerized in Table 1. In the presence of NAD+, the trap of GAPDH on the matrix was moderately inhibited compared to the result of Cibacron Blue 3GA. From the study of lactate dehydrogenase and phosphoglycerate kinase, these enzymes were reported to have 10 to 100 times higher affinity for Cibacron Blue 3GA than for NAD+ (26). GAPDH only uses NAD+ and NADH for the enzyme reaction whereas the similar amount of the
Figure 5. The advantage of photoreactive protein. (A) Photoaffinity labeling of target protein with a conventional small photoreactive ligand. (B) Photoaffinity capture of small ligand by photoreactive protein.
capture inhibition was also observed with NADP+ and NADPH. The similar change in the cofactor specificity was reported for an active site modified GAPDH (13). The result clearly showed that the photoaffinity capture differentiated between ligands with different affinities. The addition of substrate D-glyceraldehyde-3-phosphate inhibited the capture whereas a more bulky phosphate, L-tyrosine-O-phosphate, was ineffective. Since D-glyceraldehyde-3-phosphate binds near to the active site sulfhydryl group (27), it is interesting that the enough space for the substrate binding is still remains after the introduction of phenyl diazirine moiety. Although bulky ligands may have a conflict with the photoreactive group, the result also suggests that a small phosphate may be a candidate for a lead structure of the GAPDH inhibitor. Taken together, the results showed that the photoaffinity capture method could be a rapid and efficient approach of ligand screening. The technique based on surface plasmon resonance spectroscopy was developed for the real-time analysis of ligand-receptor interaction (28, 29). The technique, however, requires special equipment for the analysis whereas the photoaffinity capture only uses a set of photoreactive protein and an immobilized ligand array. The advantage of the photoaffinity approach could be a simple affinity-based entry into the massive screening of ligands. Conventional methods based on the interaction between ligand and receptor usually have a problem of nonspecific binding. The advantage of current method is that the nonspecifically bound proteins can be removed from the matrix before the detection. By making the protein photoreactive, the nonspecific cross-linking is also minimized. Photoreactive small ligands often label the protein surface in a nonspecific manner (Figure 5A). The photoreactive protein limits the photochemical reaction to only the active site (Figure 5B). This directs the crosslink to the specific capture of ligands. In summary, the application of the present method may be limited to these protein with a sulfhydryl group near the active site. The proteins have to be also stable during the modification and need to be readily available. The development of protein expression and site-specific mutagenesis, however, could provide a tailor-made (engineered) protein in which the desired sulfhydryl group is incorporated. The photoaffiniy capture would be an efficient tool for affinity-based parallel screening of
852 Bioconjugate Chem., Vol. 14, No. 5, 2003
ligands in the early step of the drug discovery and developments. ACKNOWLEDGMENT
We are very grateful to Professor G. D. Holman (University of Bath, U. K.) for valuable advice throughout the manuscript. This work was supported by Grants-inAid for Scientific Research (12470504), for Encouragement of Young Scientists (13771410), and for Exploratory Research (14658185), from the Ministry of Education, Culture, Sports, Science and Technology of Japan. Supporting Information Available: Synthetic procedures, NMR, MS, and elemental analysis of compound 1. Procedures for the preparation of photo-GAPDH and Cibacron immobilized tube. Conditions for photoaffinity capture experiments. These materials are available free of charge via the Internet at http://pubs.acs.org/BC. LITERATURE CITED (1) Hatanaka, Y., and Sadakane, Y. (2002) Photoaffinity Labeling in Drug Discovery and Developments: Chemical Gateway for Entering Proteomic Frontier. Curr. Top. Med. Chem. 2, 271-288. (2) Prestwith, G. D., and Dorma´n G. (2000) Using Photolabile Ligands in Drug Discovery and Developments. Trends Biotechnol. 18, 64-77. (3) Dadhish, M., Hennig, L., Findeisen, M., Sabine, G., Schumer, F., Hennig, H., Beck-Sickinger, A. G., and Welzel, P. (2002) Tetrafunctional Photoaffinity Labels Based on Nakanishi’s m-Nitroalkoxy-Substituted Phenyltrifluoromethyldiazirine. Angew. Chem., Int. Ed. 41, 2293-2297. (4) Strømgaard, K., Saito, D. R., Shindou, H., Ishii, S., Shimizu, T., and Nakanishi, K. (2002) Ginkgolide Derivatives for Photolabeling Studies: Preparation and Pharmacological Evaluation. J. Med. Chem. 45, 4038-4046. (5) Me´sange, F., Sebbar, M., Capdevielle, J., Guillemot, J., Ferrara, P., Bayard, F., Poirot, M., and Faye, J. (2002) Identification of Two Tamoxifen Target Proteins by Photolabeling with 4-(2-Morpholinoethoxy)benzophenone. Bioconjugate Chem. 13, 766-772. (6) Hatanaka, Y., Hashimoto, H., and Kanaoka, Y. (1998) A Rapid and Efficient Method for Identifying Photoaffinity Biotinylated Sites within Proteins. J. Am. Chem. Soc. 120, 453-454. (7) Jahn, O., Eckart, K., Braun, O, Tezval, H, and Spiess, J. (2002) The Binding Protein of Corticotropin-Releasing Factor: Ligand-Binding Site and Subunit Structure. Biochemistry 99, 12055-12060. (8) Wahlstrom, J. L., Randall, Jr, M. A., Lawson, J. D., Lyons, D. E., Siems, W. F., Crouch, G. J., Barr, R., and Facemyer, K. C. (2003) Structural Model of the Regulatory Domain of Smooth Muscle Heavy Meromysin. J. Biol. Chem. 278, 51235131. (9) Brown, C. E., Howe, L., Sousa, K., Alley, S. C., Carrozza, M. J., Tan, S., and Workman, J. L. (2001) Recruitment of HAT Complexes by Direct Activator Interactions with the ATMRelated Tra1 Subunit. Science 292, 2333-2337. (10) Kovach, M. J., Tirumalai, R., and Landy, A. (2002) Sitespecific Photocross-linking between λ Integrase and Its DNA Recombination Target. J. Biol. Chem. 277, 14530-14538. (11) Buehner, M., Ford, G. C., Moras, D., Olsen, K. W., and Rossmann, M. G. (1974) Three-dimensional Structure of D-Glyceraldehyde-3-phosphate Dehydrogenase. J. Mol. Biol. 90, 25-49.
Kaneda et al. (12) Walker, J. E., Carne, A. F., Runswick, M. J., Bridgen, J., and Harris, J. I. (1980) D-Glyceraldehyde-3-Phosphate Dehydrogenase. Eur. J. Biochem. 108, 549-565. (13) Hilvert, D., Hatanaka, Y., and Kaiser, E. T. (1988) A High Active Thermophilic Semisynthetic Flavoenzyme. J. Am. Chem. Soc. 110, 682-689. (14) Hilvert, D., and Kaiser, E. T. (1985) New Semisynthetic Flavoenzymes Based on a Tetrameric Protein Template, Glyceraldehyde-3-phosphate Dehydrogenase. J. Am. Chem. Soc. 107, 5805-5806. (15) Kenyon, G. L., and Bruice, T. W. (1977) Novel Sulfhydryl Reagents. Methods Enzymol. 47, 407-430. (16) Nassal, M. (1984) 4′-(1-Azi-2,2,2-trifluoroethyl)phenylalanine, a Photolabile Carbene-Generating Analogue of Phenylalanine. J. Am. Chem. Soc. 106, 7540-7545. (17) Trentham, D. R. (1968) Aspect of the Chemistry of D-Glyceraldehyde-3-phosphate Dehydrogenase. Biochem. J. 109, 603-612. (18) Kreimsky, I., and Racker, E. (1962) Separation of Oxidative from Phosphorylative Activity by Proteolysis of Glyceraldehyde-3-Phosphate Dehydrogenase. Biochemistry 2, 512-518. (19) Murdock, A., and Koppe, O. (1964) The Content and Action of Diphosphopyridine Nucleotide in Triosephosphate Dehydrogenase. J. Biol. Chem. 239, 1983-1988. (20) Bradford, M. M. (1976) A Rapid and Sensitive Method for the Quantitation of Microgram Quantities of Protein Utilizing the Principle of Protein-Dye Binding. Anal. Biochem. 72, 248254. (21) Hanggi, D. and Carr, P. (1985) Analytical Evaluation of the Purity of Commercial Preparations of Cibacron Blue F3GA and Related Dyes. Anal. Biochem. 149, 91-104. (22) Tazuke, S., and Kimura, H. (1978) Surface Photografting. I. Graft Polymerization of Hydrophilic Monomers onto Various Polymer Films. J. Polym. Sci. 16, 497-500. (23) Tazuke, S., and Kimura, H. (1978) Surface Photografting, 2. Modification of Polypropylene Firm Surface by Graft Polymerization of Acrylamide. Makromol. Chem. 179, 26032612. (24) Wenschuh, H., Volkmer-Engert, R., Schmidt, M., Schulz, M., Schneider-Mergener, J., and Reineke, U. (2000) Coherent Membrane Supports for Parallel Microsynthesis and Screening of Bioactive Peptides. Biopolymers (Pept. Sci.) 55, 188206. (25) Oguti, M., Gerth, E., Fitzgerald, B., and Park, J. (1973) Regulation of Glyceraldehyde-3-Phosphate Dehydrogenase by Phosphocreatine and Adenosine Triphosphate. IV. Factor affecting in vivo Control of Enzymatic Activity. J. Biol. Chem. 248, 5571-5576. (26) Thompson, S. T., and Stellwagen, E. (1976) Binding of Cibacron Blue F3GA to Proteins Containing Dinucleotide Fold. Proc. Natl. Acad. Sci. U.S.A. 73, 361-365. (27) Cardon, J. W., and Boyer, P. D. (1982) Subunit Interaction in Catalysis: Some Experimental and Theoretical Approaches with Glyceraldehyde-3-Phosphate Dehydrogenase. J. Biol. Chem. 257, 7615-7622. (28) Salamon, Z., Macleod, A. H., and Tollin, G. (1997) Surface Plasmon Resonance Spectroscopy as a Tool for Investigating the Biochemical and Biophysical Properties of Membrane Protein Systems. I: Theoretical Principles. Biochim. Biophys. Acta 1331, 117-129. (29) Salamon, Z., Macleod, A. H., and Tollin, G. (1997) Surface Plasmon Resonance Spectroscopy as a Tool for Investigating the Biochemical and Biophysical Properties of Membrane Protein Systems. II: Application to biological systems. Biochim. Biophys. Acta 1331, 131-152.
BC0340520
