Optimal Surface Chemistry for Peptide Immobilization in On-Chip

Chemistry Faculty of Engineering, Kyushu University, 744 Moto-oka Nishi-ku, Fukuoka 819-0395, Japan, and Institute of. Scientific and Industrial Resea...
0 downloads 0 Views 320KB Size
Download PDF
Anal. Chem. 2008, 80, 643-650

Optimal Surface Chemistry for Peptide Immobilization in On-Chip Phosphorylation Analysis Kazuki Inamori,† Motoki Kyo,† Kazuki Matsukawa,† Yusuke Inoue,‡ Tatsuhiko Sonoda,‡ Kenji Tatematsu,§ Katsuyuki Tanizawa,§ Takeshi Mori,‡ and Yoshiki Katayama*,‡

Biotechnology Frontier Project, Toyobo Co., Ltd., 10-24 Toyo-cho, Tsuruga, Fukui 914-0047, Japan, Department of Applied Chemistry Faculty of Engineering, Kyushu University, 744 Moto-oka Nishi-ku, Fukuoka 819-0395, Japan, and Institute of Scientific and Industrial Research, Osaka University, 8-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan

We investigated the optimal surface chemistry of peptide immobilization for on-chip phosphorylation analysis. In our previous study, we used a heterobifunctional crosslinker sulfosuccinimidyl-4-(N-maleimidomethyl) cyclohexane-1-carboxalate (SSMCC) to immobilize cysteineterminated peptides on an amine-modified gold surface. The study revealed that the phosphorylation efficiency and rate were low (only 20% at 2 h) comparing with the reaction in solution. In this study, to improve the phosphorylation efficiency, the kinase substrates were immobilized via poly(ethylene glycol) (PEG), a flexible, hydrophilic polymer. An improvement in cSrc phosphorylation was achieved (60% at 1 h) from using a PEGinserted peptide and SSMCC. However, no phosphorylation could be detected when the peptide was immobilized with a PEG-containing cross-linker. Fluorescence-labeled peptide studies revealed that the use of longer crosslinkers resulted in lower immobilization density. We considered that the flexible PEG linker was preferable to secure high phosphorylation efficiency for the immobilized peptide, probably due to the improvement of cSrc accessibility and peptide mobility, but the immobilization protocol is critical for keeping high density of the peptide immobilization. In addition, such an accelerating effect of PEG linker against on-chip phosphorylation of an immobilized peptide may depend on kinase structures or the position of the active center, because no improvement of on-chip peptide phosphorylation was observed in protein kinase A. However, PEG linker also did not suppress the phosphorylation in protein kinase A. Thus, we concluded that SSMCC and PEGylated peptide will be a good combination for the surface chemistry of on-chip phosphorylation in peptide array. Recently, assays using peptide arrays, on which peptides are immobilized on a solid support, have become popular.1,2 Peptide * To whom correspondence should be addressed. Phone: +81-92-802-2850. Fax: +81-92-802-2850. E-mail: [email protected]. † Toyobo Co. Ltd. ‡ Kyushu University. § Osaka University. (1) Emili, A. Q.; Cagney, G. Nat. Biotechnol. 2000, 18, 393-397. 10.1021/ac701667g CCC: $40.75 Published on Web 01/08/2008

© 2008 American Chemical Society

arrays are applied for peptide-protein3-6/ peptide-cell7,8 interaction studies, epitope mappings9 and on-chip phosphorylation assays.10-16 Cellulosic membrane,5,10 glass,6,11-15 and gold16 have been used as solid supports of peptide arrays, and various surface chemistries for peptide immobilization have been developed. As a cellulosic membrane matrix, SPOT technology, in which covalently bound peptides are synthesized on the membrane via spacers, is commercially available and commonly used.13,15-18 Its advantage is that peptide synthesis and array fabrication can be done simultaneously. However, a large sample volume is necessary for detection. To minimize the required sample volume, the use of a glass matrix as the solid support of peptide arrays has become more widespread. Synthesized peptides have reportedly been covalently (2) Panicker, R. C.; Xuan, H.; Yao, S. Q. Comb. Chem. High Throughput Screening 2004, 7, 547-556. (3) Beattie, J.; Shand, J. H.; Flint, D. J. Eur. J. Biochem. 1996, 239, 479-486. (4) Stone, J. D.; Demkowicz, Jr., W. E.; Stern, L. J. Proc. Natl. Acad. Sci. U.S.A., 2005, 102, 3744-3749. (5) Eichler, J. Comb. Chem. High Throughput Screening 2005, 8, 135-143. (6) Usui, K.; Tomizaki, K.; Ohyama, T.; Nokihara, K.; Mihara, H. Mol. Biosyst. 2006, 2, 113-121. (7) Gannot, G.; Tangrea, M. A.; Gillespie, J. W.; Erickson, H. S.; Wallis, B. S.; Leakan, R. A.; Knezevic, V.; Hartmann, D. P.; Chuaqui, R. F.; Emmert-Buck, M. R. J. Mol. Diagn. 2005, 7, 427-436. (8) Kato, R.; Kaga, C.; Kunimatsu, M.; Kobayashi, T.; Honda, H. J. Biosci. Bioeng. 2006, 101, 485-495. (9) Omer, B. P.; Derda, R.; Lewis, R. L.; Thomson, J. A.; Kiessling, L. L. J. Am. Chem. Soc. 2004, 126, 10808-10809. (10) Schutkowski, M.; Reimer, U.; Panse, S.; Dong, L.; Lizcano, J. M.; Alessi, D. R.; Schneider-Mergener, J. Angew. Chem., Int. Ed. 2004, 43, 2671-2674. (11) Williams, D. M.; Cole, P. A. Trends Biochem. Sci. 2001, 26, 271-273. (12) Falsey, J. R.; Renil, M.; Park, S.; Li, S.; Lam, K. S. Bioconjugate Chem. 2001, 12, 346-353. (13) Lizcano, J. M.; Deak, M.; Morrice, N.; Kieloch, A.; Hastie, C. J.; Dong, L.; Schutkowski, M.; Reimer, U.; Alessi, D. R. J. Biol. Chem. 2002, 277, 2783927849. (14) Reimer, U.; Reineke, U.; Schneider-Mergener, J. Curr. Opin. Biotechnol. 2002, 13, 315-320. (15) Pal, M.; Moffa, A.; Sreekumar, A.; Ethier, S. P.; Barder, T. J.; Chinnaiyan, A.; Lubman, D. M. Anal. Chem. 2006, 78, 702-710. (16) Wenschuh, H.; Volkmer-Engert, R.; Schmidt, M.; Schulz, M.; SchneiderMergener, J.; Reineke, U. Biopolymers 2000, 55, 188-206. (17) Reineke, U.; Volkmer-Engert, R.; Schneider-Mergener, J. Curr. Opin. Biotechnol. 2001, 12, 59-64. (18) Panse, S.; Dong, L.; Burian, A.; Carus, R.; Schutkowski, M.; Reimer, U.; Schneider-Mergener, J. Mol. Diversity 2004, 8, 291-299. (19) Su, J.; Bringer, M. B.; Ismagilov, R. F.; Mrksich, M. J. Am. Chem. Soc. 2005, 127, 7280-7281.

Analytical Chemistry, Vol. 80, No. 3, February 1, 2008 643

immobilized on a glyoxylyl-modified glass slide,11 a PEGylated thioester-containing glass slide,2,20-23 a formyl-modified glass slide,24 and a semicarbazide glass slide.25 Gold surface matrixes have also been well-studied for peptide immobilization. A Diels-Alder reaction with a cyclopentadienemodified peptide,18,26-28 thiol-maleimide coupling with a cysteineterminated peptide,19 and a thiol-disulfide exchange reaction29 have been reported on gold as a support. However, the optimal surface chemistry for peptide immobilization on various surfaces has not been sufficiently discussed. Remarkably, Wegner et al. reported that SG or SGSG insertion improved antibody binding on a FLAG peptide.29 Most recently, Andresen et al. showed that a poly(ethylene glycol) (PEG) spacer between biotin and peptide increased antibody binding to a peptide-neutravidin conjugate, which was physically adsorbed onto amino-modified glass.30 These results might suggest that a spacer insertion could be widely applied for peptide array fabrications to improve on-chip reactions. We have already reported the detection and quantification of on-chip phosphorylation of immobilized peptides by a surface plasmon resonance (SPR) imaging technique31-33 using zinc(II) chelate compound,34 which can bind to the phosphate group.35-42 In the research, cysteine-terminated peptides were immobilized onto SAMs of amine-terminated alkanethiol by using small molecular cross-linkers, which have N-hydroxysuccinimide (NHS) (20) Uttamchandani, M.; Chan, E. W. S.; Chen, G. Y. J.; Yao, S. Q. Bioorg. Med. Chem. Lett. 2003, 13, 2997-3000. (21) Lesaicherre, M. L.; Uttamchandani, M.; Chen, G. Y. J.; Yao, S. Q. Bioorg. Med. Chem. Lett. 2002, 12, 2079-2083. (22) Lesaicherre, M. L.; Uttamchandani, M.; Chen, G. Y. J.; Yao, S. Q. Bioorg. Med. Chem. Lett. 2002, 12, 2085-2088. (23) Chen, G. Y. J.; Uttamchandani, M.; Lue, R. Y. P.; Lesaicherre, M. L.; Yao, S. Q. Curr. Top. Med. Chem. 2003, 3, 705-724. (24) Shigaki, S.; Yamaji, T.; Han, X.; Yamanouchi, G.; Sonoda, T.; Okitsu, O.; Mori, T.; Niidome, T.; Katayama, Y. Anal. Sci. 2007, 23, 271-275. (25) Duburcq, X.; Olivier, C.; Malingue, F.; Desmet, R.; Bouzidi, A.; Zhou, F.; Auriault, C.; Gras-Masse, H.; Melnyk, O. Bioconjugate Chem. 2004, 15, 307316. (26) Yousaf, M. N.; Mrksich, M. J. Am. Chem. Soc. 1999, 121, 4286-4287. (27) Kwon, Y.; Mrksich, M. J. Am. Chem. Soc. 2002, 124, 806-812. (28) Houseman, B. T.; Huh, J. H.; Kron, S. J.; Mrksich, M. Nat. Biotechnol. 2002, 20, 270-274. (29) Wegner, G. J.; Lee, H. J.; Corn, R. M. Anal. Chem. 2002, 74, 5161-5168. (30) Andresen, H.; Grotzinger, C.; Zarse, K.; Kreuzer, O. J.; Ehrentreich-Forster, E.; Bier, F. F. Proteomics 2006, 6, 1376-1384. (31) Rothenha¨usler, B.; Knoll, W. Nature 1988, 332, 615-617. (32) Nelson, B. P.; Frutos, A. G.; Brockman, J. M.; Corn, R. M. Anal. Chem. 1999, 71, 3928-3934. (33) Brockman, J. M.; Nelson, B. P.; Corn, R. M. Annu. Rev. Phys. Chem. 2000, 51, 41-63. (34) Inamori, K.; Kyo, M.; Nishiya, Y.; Inoue, Y., Sonoda, T.; Kinoshita, E.; Koike, T.; Katayama, Y. Anal. Chem. 2005, 77, 3979-3985. (35) Koike, T.; Inoue, M.; Kimura, E.; Shiro, M. J. Am. Chem. Soc. 1996, 118, 3091-3099. (36) Kimura, E.; Aoki, S.; Koike, T.; Shiro, M. J. Am. Chem. Soc. 1997, 119, 3068-3076. (37) Fujioka, H.; Koike, T.; Yamada, N.; Kimura, E. Heterocycles 1996, 42, 775787. (38) Kinoshita, E.; Takahashi, M.; Takeda, H.; Shiro, M.; Koike, T. Dalton Trans. 2004, 8, 1189-1193. (39) Takeda, H.; Kawasaki, A.; Takahashi, M.; Yamada, A.; Koike, T. Rapid Commun. Mass Spectrom. 2003, 17, 2075-2081. (40) Tanaka, T.; Tsutsui, H.; Hirano, K.; Koike, T.; Tokumura, A.; Satouchi, K. J. Lipid Res. 2004, 45, 2145-2150. (41) Kinoshita, E.; Yamada, A.; Takeda, H.; Kinoshita-Kikuta, E.; Koike, T. J. Sep. Sci. 2005, 28, 155-162. (42) Kinoshita, E.; Kinoshita-Kikuta, E.; Takiyama, K.; Koike, T. Mol. Cell. Proteomics 2006, 5, 749-757.

644 Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

Table 1. Peptide Sequences of Surface-Immobilized Probes for cSrc Reaction probe 1 2 3 4 5

peptide sequence CGIYGEFKKK-NH2 CGIFGEFKKK-NH2 CGIY(PO3)GEFKKK-NH2 CG(PEG12)IYGEFKKK-NH2 CG(PEG12)IY(PO3)GEFKKK-NH2

specifications Phe-substituted Tyr-phosphorylated PEG-inserted PEG-inserted and Tyr-phosphorylated

ester and maleimide groups.43,44 The research revealed that both the reaction rate and the efficiency of on-chip phosphorylation (20% in 2 h) were much lower than those in solution reaction,45 even when a GG spacer was inserted in the peptide sequence. We have also already reported that a transcription factor could bind to its recognition sequence, only when the sequence was immobilized via PEG.46 We speculated that the accessibility of the transcription factor and the mobility of the immobilized DNA were secured by the PEG cross-linker. In this paper, an optimal surface chemistry for peptide immobilization on gold using PEG spacers was investigated to improve the phosphorylation rate and efficiency. Kinase substrates were immobilized via PEG using a PEG cross-linker46 and PEG-inserted peptide.47 The on-chip phosphorylation by cSrc and PKA were investigated with autoradiography using [γ-32P/33P]-ATP,10,11,15 and an SPR imaging technique with a zinc(II) chelate compound. EXPERIMENTAL SECTION Materials. Sulfosuccinimidyl-4-(N-maleimidomethyl) cyclohexane-1-carboxalate (SSMCC, Pierce), N-hydroxysuccinimidePEG maleimide MW 3400 (NHS-PEG3400-MAL, Nektar), MALdPEG12 NHS ester (NHS-PEG875-MAL, Quanta Biodesign), thiolterminated methoxypoly(ethylene glycol) MW 5000 (PEG-thiol, NOF), 8-amino-1-octanethiol hydrochloride (8-AOT, Dojindo Laboratories), cAMP-dependent protein kinase catalytic subunit (PKA, Promega), Src, active (cSrc, Upstate), adenosine 5′-triphosphete disodium salt hydrate (ATP, Sigma-Aldrich), [γ-33P]-ATP (GE Healthcare), streptavidin (SA, Invitrogen), anti-streptavidin antibody (anti-SA, Vector), and N-(5-(2-(+)-biotin aminoethylcarbamoyl)pyridin-2-ylmethyl)-N,N′,N′-tris(pyridin-2-ylmethyl)-1,3-diaminopropan-2-ol (Phos-tag biotin, Nard Institute, Japan) were used without further purification. All other reagents and solvents used were of analytical quality. All aqueous solutions were prepared using deionized and distilled water. Peptide Preparation. All peptides used in this study and shown in Tables 1 and 2 were synthesized by a peptide synthesizer (Pioneer, Applied Biosystems) and purified by HPLC (BioCAD perfusion chromatography system, Applied Biosystems) using an (43) Nelson, B. P.; Grimsrud, T. E.; Liles, M. R.; Goodman, R. M.; Corn, R. M. Anal. Chem. 2001, 73, 1-7. (44) Brockman, J. M.; Frutos, A. G.; Corn, R. M. J. Am. Chem. Soc. 1999, 121, 8044-8051. (45) Sonoda, T.; Shigaki, S.; Nagashima, T.; Okitsu, O.; Kita, Y.; Murata, M.; Katayama, Y. Bioorg. Med. Chem. Lett. 2004, 14, 847-850. (46) Kyo, M.; Yamamoto, T.; Motohashi, H.; Kamiya, T.; Kuroita, T.; Tanaka, T.; Engel, J. D.; Kawakami, B.; Yamamoto, M. Genes Cells 2004, 9, 153-164. (47) Song, A.; Wang, X.; Zhang, J.; Marik, J.; Lebrilla, C. B.; Lam, K. S. Bioorg. Med. Chem. Lett. 2004, 14, 161-165.

Table 2. Peptide Sequences of Surface-Immobilized Probes for PKA Reaction probe

peptide sequence

6 7 8 9 10

CGGLRRASLG-NH2 CGGLRRAALG-NH2 CGGLRRAS(PO3)LG-NH2 CGG(PEG12)LRRASLG-NH2 CGG(PEG12)LRRAS(PO3)LG-NH

specifications Ala-substituted Ser-phosphorylated PEG-inserted PEG-inserted and Ser-phosphorylated

ODS-A column with a 12-nm pore size (YMC). Kemptide peptide,48 to which cysteine and two glycine residues were attached at the C-terminus, was used as a PKA substrate (probe 1). We prepared a negative control of Kemptide (probe 2), in which serine was substituted with alanine, and a positive control (probe 3), which had a phosphoserine residue. The substrate of cSrc kinase28 (probe 4) and its positive control (probe 5) were also synthesized to determine the reaction specificity of PKA. PEG-inserted peptides were synthesized by using N-Fmoc-amindo-dPEG12 acid (Quanta Biodesign). Cy3-labeled peptides were obtained using Cy3 NHS ester (Amersham). Cy3-labeled lysine residue was added to the N-terminus of probe 1. Fabrication of Peptide Arrays. The covalently immobilized peptide arrays were obtained on the basis of the scheme described in our previous paper.34 Three heterobifunctional cross-linkers, SSMCC (MW 436), NHS-PEG875-MAL, and NHS-PEG3400-MAL, were used to immobilize peptides in the array fabrications. A patterned amino-modified Au-coated chip (Toyobo), which has amino groups in 96 areas with a 500 µm square and PEG background, was reacted for 30 min with 300 µL of 1 mM SSMCC or for 90 min with 300 µL of 1 mM NHS-PEG875-MAL or NHSPEG3400-MAL to create a maleimide-modified surface. Then, 10nL drops of 1 mg/mL cysteine-terminated peptide substrates were delivered automatically on the maleimide surface using an automated spotter (Toyobo), and the maleimide-thiol reaction was carried out overnight. PEG-thiol solution was reacted on the peptide array to block the unreacted maleimide group. The surface was then rinsed with phosphate buffer and water. On-Chip Phosphorylation by Protein Kinases. The peptide arrays were reacted at 30 °C with 400 µL of 0.2 unit/µL PKA solution in reacting buffer (50 mM Tris-HCl, 50 mM MgCl2, pH 7.4) containing 10 µM ATP, or with 300 µL of 0.1 unit/µL cSrc solution in reacting buffer (50 mM MES-NaOH buffer, 50 mM MgCl2, pH 6.8) containing 10 µM ATP, respectively. For the autoradiography assay in both protein kinases, 1.85 × 1017 Bq/ mol [γ-33P]-ATP was used. After the reactions, the arrays were rinsed with PBS and MilliQ water and then dried with airflow. SPR Imaging Analysis. A 300-µL aliquot of 2 µg/mL Phostag biotin solution, which was dissolved in the buffer (10 mM HEPES-NaOH, 0.2 M sodium nitrate, 1 mM zinc nitrate, 0.005% (w/v) Tween 20, 10% ethanol, pH 7.4), was poured on the peptide array and incubated for 30 min at room temperature. The array was then immediately placed on the SPR instrument and exposed by 1 µg/mL SA solution in the running buffer (10 mM HEPESNaOH, pH 7.4) for 10 min and then rinsed with the running buffer (48) Kemp, B. E.; Graves, D. J.; Benjamini, E.; Krebs, E. G. J. Biol. Chem. 1977, 252, 4888-4894.

for 5 min. For the quantification of phosphorylation, 1 µg/mL antiSA solution in the running buffer was injected subsequently for 20 min into the SPR imaging instrument to enhance the signals. The signal data were collected with the SPR analysis program (Toyobo). All SPR experiments were performed at 30 °C. Autoradiography Experiment. The array was reacted with [γ-33P]-ATP and exposed to SG Imaging Plates (Fuji Photo Film) for 30 min. The plates were observed with an imaging analyzer (BAS-1800 II: Fuji Film). Using an analysis program (Fuji Film), the quantity of 33P uptake on each spot was analyzed. The obtained images were processed using PhotoShop (Adobe). The quantity of 33P uptake as the relative amount of radiation is shown by the photo-stimulated luminescence value, which was calculated by Image Gauge (Fuji Film). Fluorescent Analysis of Peptide Arrays. The immobilized Cy3-labeled peptide array was obtained by the following procedure. A glass slide with a thin chromium underlayer (2 nm) beneath a gold layer (45 nm) was immersed in 8-AOT solution (1 mM in ethanol) for 90 min to introduce the amine group. Then, 1 mM SSMCC solution, 1 mM NHS-PEG875-MAL, or 1, 5, 10, or 20 mM NHS-PEG3400-MAL solution was reacted for 15 or 90 min, respectively, with the amine group on the gold surface. Finally, a 2.5, 1.0, 0.5, 0.1, 0.05, or 0.01 mg/mL solution of Cy3-labeled peptides was spotted on the gold with an automated spotter (Toyobo) to react on the introduced MAL group. The prepared arrays were immersed in PBS and water for 15 min, respectively, and were dried with airflow. Then, they were analyzed by a fluorescent array scanner (GenePix 4200AL: Axon Instruments) at the wavelength of 532 nm. The obtained images were modified using PhotoShop (Adobe). The immobilized peptide density was quantified with a calibration curve, which was established with “unwashed” spots of Cy3labeled peptide. Drops (10 nL) of Cy3-labeled peptide solutions at 5.0, 6.3, 8.0, 10.0, 25.0, and 50.0 µg/mL were spotted on the maleimide-modified gold, then dried and analyzed by a fluorescent array scanner without washing. All procedures in fluorescent analysis were performed under a light-shielding condition. RESULTS AND DISCUSSION Surface Chemistries for Peptide Immobilization. Surface chemistries of peptide immobilization on gold surface are described in Figure 1. Heterobifunctional cross-linkers, which contain NHS ester and maleimide groups, were reacted with amino-modified chips to create maleimide surfaces (Figure 1A). Then, cysteine-terminated peptide probes were immobilized through thiol-maleimide reactions using one of three schemes (Figure 1B). In scheme i, the small-molecule cross-linker, SSMCC, and a Cys-terminated peptide were used. This surface chemistry is the same as that of our previous study, and showed a low phosphorylation efficiency and reaction rate.34 To improve the efficiency and rate, we examined two schemes in which peptides were immobilized via PEG to provide spacer effects. In scheme ii, we used flexible and longer cross-linkers; NHS-PEG875-MAL or NHSPEG3400-MAL, which were appropriate for protein-DNA interaction studies in our previous researches.46 In scheme iii, SSMCC and PEG-inserted Cys-terminated peptides were used to achieve higher immobilization efficiency with a PEG spacer. N-Fmocamindo-dPEG12-acid (Figure 1C) was used to obtain PEG inserted Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

645

Figure 1. Design of array surface chemistry for the immobilization of peptide probes. (A) Introduction of maleimide group using heterobifunctional cross-linker to the amino-modified Au-coated chip. (B) Three strategies for immobilization of the cysteine-terminated peptide probes on gold modified by a maleimide group: (i) a small-molecule cross-linker, SSMCC; (ii) a longer-sized cross-linker, NHS-PEG-MAL; (iii) cysteine-terminated and PEG-inserted peptide probes immobilized on gold using SSMCC as a cross-linker. (C) The chemical structure of the PEG derivative compound that was used in PEG-inserted peptide synthesis.

peptides by the solid-phase peptide synthesis method. The PEG length of N-Fmoc-amindo-dPEG12-acid is identical to that of NHSPEG875-MAL. On-Chip Phosphorylation Analysis by cSrc with Autoradiography and SPR Imaging. We compared the three surface chemistries for on-chip phosphorylation shown in schemes i-iii. Peptide arrays were fabricated using three heterobifunctional cross-linkers (SSMCC, NHS-PEG875-MAL, and NHS-PEG3400-MAL) and five peptide probes shown in Table 1. On-chip phosphorylations by tyrosine kinase, cSrc, were investigated with the fabricated arrays. The amino acid sequence IYGEFKKK in probe 1 is known as a cSrc substrate.28 The liquid scintillation counting experiment showed that probe 1 was efficiently phosphorylated by cSrc (data not shown). Probe 2 is a Phe-substituted probe 1 as a negative control so that it should not be phosphorylated at all. Probe 3 is “100% phosphorylated” probe 1 and cannot be phosphorylated further. In probe 4, PEG12 is inserted between G and I in probe 1. Probe 5 is a “100% phosphorylated” probe 4. Figure 2A indicates the 33P uptakes by the five probes immobilized using SSMCC in autoradiography after the incubation with [γ-33P]-ATP for 30 or 120 min. The array pattern is shown in Figure 2B. As expected, 33P uptakes were observed in probes 1 and 4. However, when the PEG cross-linker (NHS-PEG875-MAL 646

Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

or NHS-PEG3400-MAL) was used for immobilization, no 33P uptake could be detected (data not shown). Even when the PEG crosslinkers were used at higher concentrations (20 mM), no signal was obtained. 33P uptakes at 30, 60, and 120 min with the three surface chemistries were assayed and are summarized in Figure 2C. The surface chemistry using “SSMCC + probe 1”, “PEG cross-linker + probe 1”, and “SSMCC + probe 4” corresponded to schemes i, ii, and iii, respectively. As described, no 33P uptake could be observed in scheme ii. On the other hand, the 33P uptake in scheme iii was significantly larger than that in scheme i. These results suggest that PEG insertion may improve phosphorylation efficiency; but the PEG cross-linkers cannot be used to insert a PEG moiety. We used the SPR imaging technique to investigate the phosphorylation rate and efficiency quantitatively. The phosphorylation ratios were calculated by our established method.34 Briefly, SPR signals for mixtures of probes 1 and 3 or of probes 4 and 5 at various mixture ratios were obtained, and then the SPR signal ratios against probe 3 or probe 5 were calculated respectively. Subsequently, calibration curves between SPR signal intensities and mixture ratios, which were hypothetically identical to phosphorylation efficiencies, were established.

Figure 3. On-chip phosphorylation kinetic study of the immobilized peptide probe by SPR imaging analysis. The phosphorylation ratios were calculated using a calibration array on which various mixture ratios of probes 1 and 3 or of probes 4 and 5 (Table 1) were immobilized.

Figure 2. (A) Autoradiograms for the peptide array by on-chip cSrc reaction. (B) The array pattern of the immobilized peptide probes for autoradiography. (C) 33P uptake with phosphorylation time in autoradiography by cSrc comparing three surface chemistries for peptide immobilization described in Figure 1. In scheme ii, two PEG crosslinkers were compared.

The time dependencies of phosphorylation ratios in cSrc exposure are shown in Figure 3. The phosphorylation efficiency in schemes i and iii were saturated at 40% in 120 min and at 65% in 60 min, respectively. The SPR signal on probe 2, the negative control peptide probe, was negligible. Additionally, the SPR signal in scheme ii using the PEG cross-linkers was also negligible (data not shown). Although the phosphorylation profiles of SPR imaging were different from those of autoradiography, improvement of the phosphorylation rate and efficiency was observed in scheme iii by both SPR and autoradiography. We considered that the reason for the disadvantage of scheme ii may be caused by the steric hindrance of the maleimide group due to flexibility of PEG chains. We imagined that the terminatory maleimide groups were buried in the random coil of PEG molecules, and the collision possibility between the maleimide groups and the peptide probes might be drastically reduced in consequence. On the other hand, improvement of the phosphorylation rate and efficiency in scheme iii was probably due to the increase of molecular mobility of the substrate peptide and kinase accessibility to the peptide due to the presence of the inserted PEG linker. We speculated that the difference of the phosphorylation efficiency

between schemes ii and iii may be due to the difference of immobilized density of the peptide. Qualification of Immobilized Peptide Density. We hypothesized that the use of the PEG cross-linkers (NHS-PEG3400-MAL and NHS-PEG875-MAL) decreased the immobilization density of peptides and, consequently, disabled detection of phosphate groups. To verify the difference in the immobilized peptide density depending on the molecular weight of the cross-linker, we examined the fluorescence analysis of a peptide and a PEGinserted peptide, which were both Cy3-labeled and immobilized using three heterobifunctional cross-linkers, SSMCC, NHS-PEG875MAL, and NHS-PEG3400-MAL. The results of fluorescence imaging are shown in Figure 4A. The array pattern is shown in Figure 4B. The fluorescence intensities were drastically lower using NHS-PEG875-MAL and NHS-PEG3400-MAL than using SSMCC. This suggests that PEG insertion with the PEG cross-linker decreased the peptide density on the gold surface. In contrast, the immobilization density of a PEG-inserted peptide using SSMCC was almost the same as that of the regular peptide. It is interesting, because the PEG lengths of NHS-PEG875-MAL and PEG-inserted peptide were identical. The PEG insertion was achieved without decrease of immobilization density using SSMCC and the PEG-preinserted peptide. Figure 4C indicates the quantification of immobilized peptide density using the Cy3-labeled peptide. When SSMCC was used with 0.5 mg/mL peptide solutions, the immobilized peptide density was more than 300 pg/mm2 in both the PEG-inserted peptide probe and the regular peptide probe. On the other hand, when NHS-PEG3400-MAL was used, the immobilized peptide density was less than 30 pg/mm2. The PEG surface is known to have high surface free energy. Probably, its high mobility tends to suppress the reaction between the maleimide group in the PEG terminus and the N-terminal cysteine in the peptide probe. The abovementioned immobilized peptide density, 30-300 pg/mm2 corresponds to approximately 1.8 × 1012-1.8 × 1013 molecules/cm2. Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

647

Figure 4. Evaluation of peptide immobilization quantity using Cy3-labeled peptide probe. Three heterobifunctional cross-linkers were compared. (A) The fluorescence images of Cy3-labeled peptide array. The fluorescence intensity is considered to indicate the efficiency of peptide immobilization. (B) The array pattern of the immobilized peptide probes in autoradiography. Probe 1 was used in schemes i and ii), and probe 4 in scheme iii. (C) Comparison of immobilized peptide density depending on the cross-linker. (D) Effect of immobilization conditions using NHS-PEG3400-MAL at various peptide concentrations on the immobilized peptide density.

The distances between immobilized peptide probes were calculated as 23-74 Å. The obtained peptide densities were slightly higher than the DNA densities, which were measured by “washoff” of the fluorescent complements in previous studies for a DNAimmobilizing gold plate.49-51 Thus, this estimation may include unavailable probes, which adsorb nonspecifically on the surface, (49) Okumura, A.; Sato, Y.; Kyo, M.; Kawaguchi, H. Anal. Biochem. 2005, 339, 328-337. (50) Peterson, A. W.; Wolf, L. K.; Georgiadis, R. M. J. Am. Chem. Soc. 2002, 124, 14601-14607.

648 Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

because the wash-off measurement could estimate only the available probes. We investigated peptide immobilization densities in scheme ii with NHS-PEG3400-MAL in various reaction conditions. However, we could not improve the immobilized peptide density, which was ∼30 pg/mm2 for all concentrations of NHS-PEG3400-MAL or peptide (Figure 4D). When NHS-PEG3400-MAL was used, no (51) Smith, E. A.; Wanat, M. J.; Cheng, Y.; Barreira, S. V. P.; Frutos, A. G.; Corn, R. M. Langmuir 2001, 17, 2502-2507.

phosphorylation could be detected on the chip in SPR analysis at any immobilization conditions (data not shown). We think that the immobilized peptide density was not sufficient for the detection of on-chip phosphorylation at any conditions when NHS-PEG3400MAL was used. In the case of NHS-PEG875-MAL, the immobilized peptide density was slightly higher than that with NHS-PEG3400-MAL. This is probably due to the less hindered maleimide group attached at the PEG875 terminus than that at longer PEG3400 terminus. Immobilized density of the PEG linkers may also be another factor contributing this effect, because the longer PEG3400 should form PEG brush with lower density.52,53 However, the density was still much lower than that for PEG preinserted peptide in scheme iii. It is considered that the reactivity of NHS-PEG875-MAL against amino groups on the gold surface is very similar to that of PEG preinserted peptide against maleimide groups on the surface, because the molecular length and flexibility of both compounds should not be so different. Thus, these results suggest that the difference of the detection sensitivity between schemes ii and iii is caused by the difference of the reactivity of maleimide group, which is attached at the PEG terminus and cyclohexylmethyl group in SSMCC. On-Chip Phosphorylation Analysis by PKA. We also investigated the three surface chemistries in the PKA reaction to expand our strategy. The peptide arrays were fabricated using two heterobifunctional cross-linkers, SSMCC and NHS-PEG3400MAL, and the five peptide probes shown in Table 2, and were analyzed with autoradiography and SPR measurement. The peptide sequence LRRASLG in probe 6 is a well-known PKA substrate, which is generally called Kemptide.48 Probe 7, which is Ala-substituted probe 6, is a negative control. Probe 8 is “100% phosphorylated” probe 6 as a positive control. Probe 9 is a PEGinserted probe 6. Probe 10 is a “100% phosphorylated” probe 9. 33P uptakes by PKA exposure on the array were measured by autoradiography and described in Figure 5A. The 33P uptakes in SSMCC + probe 6 (scheme i) and SSMCC + probe 9 (scheme iii) were almost identical. This result means that the phosphorylation rate and efficiency were not improved but also not suppressed by PEG-insertion chemistry in the PKA reaction. When NHS-PEG3400-MAL was used as cross-linker (including scheme ii, NHS-PEG3400-MAL + probe 6), 33P uptake was not observed as in the cSrc study. Additionally, 33P uptakes were not observed in the negative control (probe 7) and “100% phosphorylated” probes (probe 8 and 10) (data not shown). The phosphorylation efficiencies by PKA were analyzed quantitatively by SPR imaging (Figure 4B). No improvements and suppression of phosphorylation efficiency were observed for the PEG-inserted peptide of Kemptide. The phosphorylation efficiencies were saturated at less than 20%. We considered that PEG insertion in an immobilized peptide improved phosphorylation efficiency in cSrc, but not in PKA. The reason for the difference was unclear; however, protein structures, or the positions of the active center, may be the source of the difference. (52) Uchida, K.; Otsuka, H.; Kaneko, M.; Kataoka, K.; Nagasaki, Y. Anal. Chem. 2005, 77, 1075-1080. (53) Otsuka, H.; Nagasaki, Y.; Kataoka, K. Biomacromolecules 2000, 1, 39-48.

Figure 5. (A) 33P uptake in the immobilized peptide probes with PKA reaction time by autoradiography. (B) Comparison of kinetics for on-chip phosphorylation using the immobilized peptide probes with PKA reaction by SPR imaging analysis.

CONCLUSIONS In this paper, we investigated the optimal surface chemistry for peptide immobilization in the detection of on-chip phosphorylation. We expected that the phosphorylation rate and efficiency would be improved when peptide probes were immobilized via PEG. Such improvement was actually achieved using the smallmolecule cross-linker, SSMCC, and a PEG-inserted peptide in cSrc phosphorylation. However, no phosphorylation signal was observed by autoradiography or SPR imaging, when NHS-PEG875MAL and NHS-PEG3400-MAL were used as cross-linker to insert a PEG moiety between gold and the kinase substrate sequence. We evaluated the immobilization density using fluorescent-labeled peptides and found that the immobilization density became lower when using higher molecule weight cross-linkers. The immobilized peptide density could be kept higher by using SSMCC and a PEG-inserted peptide. PEG linker improved the detection sensitivity, but protocol for PEG insertion was critical for the expression of this effect. We concluded that the optimal surface chemistry for on-chip phosphorylation is to use a small molecular cross-linker and a PEG-inserted peptide. In the PKA reaction, however, we could not observe any improvements in the phosphorylation efficiency and rate by PEG insertion. We speculated that the kinase structures, or the positions of the active center, may cause the difference in improvement by intrapeptide PEG insertion. However, PEG insertion did not decrease the detection efficiency for PKA phosphorylation. Thus, the immobilization protocol using SSMCC and PEG-inserted peptide is offered as the preferred Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

649

surface chemistry for the detection of on-chip phosphorylation.

Koike and Prof. Kinoshita at Hiroshima University for useful information on the biotinylated zinc(II) complex.

ACKNOWLEDGMENT

Received for review August 6, 2007. Accepted October 26, 2007.

This work was supported by the New Energy and Industrial Technology Development Organization (NEDO). We thank Prof.

650

Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

AC701667G