Bioconjugate Chem. 2007, 18, 2197–2201
2197
Facile and Rapid Direct Gold Surface Immobilization with Controlled Orientation for Carbohydrates Jeong Hyun Seo,† Kyouichi Adachi,‡,§ Bong Kuk Lee,‡,§ Dong Gyun Kang,† Yeon Kyu Kim,† Kyoung Ro Kim,† Hea Yeon Lee,*,‡,§ Tomoji Kawai,‡,§ and Hyung Joon Cha*,† Department of Chemical Engineering, Pohang University of Science and Technology, Pohang 790-784, Korea, Institute for Scientific and Industrial Research, Osaka University, Osaka 567-0047, Japan, and Core Research for Evolutional Science and technology, Japan Science and Technology Agency, Saitama 332-0012, Japan. Received July 30, 2007; Revised Manuscript Received August 15, 2007
Effective surface immobilization is a prerequisite for numerous carbohydrate-related studies including carbohydrate–biomolecule interactions. In the present work, we report a simple and rapid modification technique for diverse carbohydrate types in which direct oriented immobilization onto a gold surface is accomplished by coupling the amine group of a thiol group-bearing aminophenyl disulfide as a new coupling reagent with an aldehyde group of the terminal reducing sugar in the carbohydrate. To demonstrate the generality of this proposed reductive amination method, we examined its use for three types of carbohydrates: glucose (monosaccharide), lactose (disaccharide), and GM1 pentasaccharide. Through successful mass identifications of the modified carbohydrates, direct binding assays on gold surface using surface plasmon resonance and electrochemical methods, and a terminal galactose-binding lectin assay using atomic force microscopy, we confirmed several advantages including direct and rapid one-step immobilization onto a gold surface and exposure of functional carbohydrate moieties through oriented modification of the terminal reducing sugar. Therefore, this facile modification and immobilization method can be successfully used for diverse biomimetic studies of carbohydrates, including carbohydrate–biomolecule interactions and carbohydrate sensor or array development for diagnosis and screening.
INTRODUCTION Carbohydrates encode information for specific molecular recognition, help determine protein folding, stability, and pharmacokinetics, play critical roles in determining biological functions, and affect diverse physiological processes (1–4). In addition, carbohydrate–protein interactions have been used to elucidate fundamental biochemical processes and identify new pharmaceutical substances in living cell systems (1–4). In order to fully study specific carbohydrate interactions with biomolecules such as proteins, DNA, and other carbohydrates in the cell, researchers need effective methods for carbohydrate immobilization (5–8). To date, modification for carbohydrate immobilization has been attempted using several techniques, including copolymerization, coupling with divinyl sulphone, coupling to CNBractivated substrates, and reductive amination (9, 10). However, although these modification methods have demonstrated promising results, general usefulness for diverse carbohydrate types has not been proven (11, 12). Previous studies have mainly focused on mono and/or disaccharides, with a few attempts made to immobilize oligo or polysaccharides, because of their structural complexity and problems with severe structural deformity during modification (13–15). Therefore, more advanced methods including alternative surface modification have been developed for diverse carbohydrate types (6, 16–20). However, these strategies also have tended to be multistep protocols aimed at indirect immobilization. Especially, for biosensor applications, it is desirable to directly and rapidly immobilize various carbohydrates on surfaces by one-step protocols.
Successful surface immobilization of carbohydrates has been further complicated by the need to retain the original structure and expose the functional sites so that the immobilized molecule can mimic the specific biomolecular interactions occurring on the cell surface. Because carbohydrates do not have functional groups for orientation, it has proven technically challenging to immobilize carbohydrates (especially oligo and polysaccharides) in an oriented fashionwhileretainingtheirinherentstructuresandfunctionalities(11,12). Therefore, we sought to develop a method for orienting and immobilizing various carbohydrate types directly and rapidly onto a gold surface that will allow specific carbohydrate–biomolecule recognition. Immobilization of carbohydrates on gold surface can be easily applicable for gold-based biosensors, such as electrochemical (EC) sensors, surface plasmon resonance (SPR), and quartz crystal microbalances (QCM). We herein report a simple and rapid reductive amination technique by using aminophenyl disulfide as a new coupling reagent for diverse carbohydrate types in which orientation is accomplished by coupling a thiol (-SH) group to the terminal reducing sugar of the carbohydrate, and the oriented carbohydrate is directly immobilized onto a bare gold surface (Scheme 1). To show the generality of our proposed modification/ immobilization method, we examined its use for three types of carbohydrates having terminal reducing sugar moieties, namely, glucose (monosaccharide), lactose (disaccharide), and GM1 pentasaccharide. We also demonstrated the interaction of a representative protein with carbohydrates surface-immobilized using our method (Scheme 1).
EXPERIMENTAL PROCEDURES * Corresponding author. E-mail:
[email protected] (H.J.C.). Email:
[email protected] (H.Y.L.). † Pohang University of Science and Technology. ‡ Osaka University. § Japan Science and Technology Agency.
Carbohydrate Modification. Glucose (Sigma), lactose (Sigma), and GM1 pentasaccharide (Alexis Biochemicals) at 100 mM each were separately dissolved in water, and 50 mM aminophenyl disulfide (Aldrich) was dissolved in acetic acid.
10.1021/bc700288z CCC: $37.00 2007 American Chemical Society Published on Web 10/05/2007
2198 Bioconjugate Chem., Vol. 18, No. 6, 2007
Seo et al.
Scheme 1. Strategy for the Modification and Direct Immobilization of Various Carbohydrate Typesa
a Modified SH-saccharides can be directly immobilized on gold and interact specifically with lectin. R1, glucose (monosaccharide); R2, lactose (disaccharide); R5, GM1 pentasaccharide.
The carbohydrate and aminophenyl disulfide solutions were mixed well and incubated in sealed tubes for 1 h at 20 °C (glucose and lactose) or 30 °C (GM1 pentasaccharide) for the amination reaction step. Then, freshly prepared reducing reagent, 100 mM dimethylamine borane (Fluka), was added to each reaction, and the tubes were incubated unsealed for 1 h at 20 °C for the reduction step. The tubes were then resealed and heated for 1 h at 50 °C under nitrogen gas steaming for the condensation step. All modified carbohydrates were dissolved in 5% acetic acid solution and preserved at -20 °C under light-blocking conditions. Without further purification steps, modified samples were used for MALDITOF MS analysis and immobilization. Mass Identification Using MALDI-TOF MS. To improve the ionization efficiency of MALDI-TOF MS, the samples were desalted using Zip-tip C18 (Millipore) and then eluted onto MALDI target plates using matrix solution (10 mg/mL; R-cyano4-hydroxycinnamic acid dissolved in a solution consisting of 50% acetonitrile and 0.5% trifluoroacetic acid). All mass spectra were acquired in reflection mode using a 4700 Proteomic Analyzer (Applied Biosystems) at Korea Basic Science Institute. Spectra were obtained in the mass range between 100 to 2000 Da with ∼200 laser shots. Internal calibration was performed with 4700 Cal Mix (Applied Biosystems). SPR-Based Immobilization Analysis. All SPR experiments were performed using a Biacore 2000 instrument (Biacore AB) with a flow system that loaded the solutions as discrete channels. The SIA Kit Au (Biacore AB) was used as the bare gold surface for the direct immobilization of modified SH-containing carbohydrates. The flow rate for binding onto the gold surface was 5 µL/min, and a noninjected channel was used as the baseline. Immobilization was performed by injecting 60 µL of each sample onto the gold surface, and immobilized carbohydrate amounts were measured as RU. All measurements were performed in HEPES-EP buffer (Biacore AB; 10 mM HEPES at pH 7.4, 150 mM NaCl, 3 mM EDTA, and 0.005% surfactant F20). EC-Based Immobilization Analysis. EC experiments were performed using a BAS 100 B/W potentiostat (Bioanalytical
System) at room temperature. The biosensor consisted of an 8-multielectode chip comprising an array of 200-nm diameter gold electrodes on glass (21–23). Modified GM1 pentasaccharide (1 µL at 5 mM in HEPES-buffered saline) was directly immobilized onto the multigold electrodes for 1 h at room temperature, and unmodified GM1 pentasaccharide (1 µL at 5 mM) was loaded as a negative control. Electrode pretreatment was carried out for 3 min at 1.8 V in 10 mM H2SO4. All CV measurements were performed in solutions containing 5 mM K3Fe(CN)6 at a scan rate of 100 mV/s. A solid bar of Ag/AgCl in 100 mM KCl and a 5-mm diameter platinum wire were used as the reference and the counter electrode, respectively. The SWV curves were recorded at a 10 mV/s scan rate with a pulse height of 25 mV and a step time of 0.2 s. Lectin Interaction Analysis Using AFM. We deposited a layer of 99.5% pure Au(111) on a cleaved mica surface using a high speed vacuum (base pressure, 5 × 10-7 Pa) at 450 °C. After the immobilization of SH-GM1 pentasaccharide onto the gold surface, we added 20 mg/mL of galectin (Biovision) and incubated the samples for 1 h. Topology was imaged using Multimode DI AFM (Veeco Instruments).
RESULTS AND DISCUSSION For the efficient immobilization of diverse carbohydrate types onto a bare gold surface, we used an amination reaction to couple the aldehyde (-CHO) group of the carbohydrate’s terminal reducing sugar with the amine (-NH2) group of a thiolbearing aminophenyl disulfide (Scheme 1). A reductive amination reaction, namely, coupling the aldehyde group of a carbohydrate’s terminal reducing sugar with the amine group of 2-aminopyridine, is commonly used for pyridylamination tagging of glycans with 2-aminopyridine in order to facilitate their purification and molecular weight analysis (13, 14, 24, 25). In the present work, we found that mono and disaccharides could be modified at 20 °C reaction temperature, whereas modification of GM1 pentasaccharide was successful at 30 °C. This may suggest that a comparatively higher reaction temper-
Bioconjugate Chem., Vol. 18, No. 6, 2007 2199
Figure 2. SPR binding assays of SH-modified monosaccharide glucose (A), disaccharide lactose (B), and GM1 pentasaccharide (C).
Figure 1. MALDI-TOF MS analysis of SH-modified monosaccharide glucose (A), disaccharide lactose (B), and GM1 pentasaccharide (C).
ature is required to give a higher ratio of an acyclic form of the terminal reducing sugar of complex carbohydrates. We further found that the modification efficiency depended on the concentration of reactants, with experiments containing 100 mM GM1 pentasaccharide showing a higher efficiency than those containing 10 mM of the same carbohydrate at the same concentration (50 mM) of aminophenyl disulfide (data not shown). To check the feasibility of the modification of target carbohydrates using the proposed reductive amination reaction, we first performed nuclear magnetic resonance spectroscopy (NMR). The NMR spectrum contained features consistent with a -CH2-NH- moiety (for glucose, 1H NMR [600 MHz, D2O] δ 3.42–3.48 [3H, m], 3.58–3.62 [5H, m], 6.48 [2H, d, J)8.0 Hz], 7.09 [2H, d, J)8.0 Hz]). We then used matrix-assisted laser desorption–ionization mass spectrometry with time-of-flight (MALDI-TOF MS) analysis to investigate the molecular weights of the modified carbohydrates (Figures 1A–C). Although we used aminophenyl disulfide as the coupling reagent, the modification primarily occurred through the aminophenyl sulfide
that might be formed from disulfide bond breakage under acidic (pH 2–3) conditions or reduction of its bond by dimethylamine borane. Therefore, the molecular weights of the modified glucose, lactose, and GM1 pentasaccharide were 290, 452, and 1108 (1130: +Na), respectively. Interestingly, all of the examined samples yielded a double mass value, representing a small peak of modified carbohydrates that had reacted directly with the aminophenyl disulfide. In addition, we also performed the modification of four other types of carbohydrates and confirmed their modifications using mass analyses (data not shown). Overall, these mass analyses confirmed that all of the target carbohydrates were successfully modified. Without further purification, we then directly immobilized the SH-modified carbohydrates onto a gold surface. All of the SH-modified carbohydrates showed successful binding on the gold surface, as assessed by SPR binding assays (Figure 2A–C). Since the mass of lactose is somewhat higher than that of glucose at the same mole concentration, the resonance unit (RU) value of SH-lactose (950 RUs) was slightly higher than that of SH-glucose (900 RUs). In the case of GM1 pentasaccharide, although we used a carbohydrate concentration of 100 mM for the modification step, we diluted the SH-GM1 pentasaccharide 10-times for immobilization because of its limited available amount. Although GM1 pentasaccharide has a much higher
2200 Bioconjugate Chem., Vol. 18, No. 6, 2007
Seo et al.
Figure 3. EC response of gold electrodes bound with unmodified GM1 pentasaccharide (red line) and modified SH-GM1 pentasaccharide (green line) compared to that of the bare gold electrode (black line). Inset, CV; main figure, SWV. The schematic illustration represents immobilization of modified SH-GM1 pentasaccharide on the gold surface.
molecular weight than glucose or lactose, the RU value was lower (800 RUs) because the utilized concentration was onetenth that used in the other experiments. Collectively, these findings indicate that all of the SH-modified carbohydrates were successfully immobilized onto the gold surface without requiring additional treatment steps. To further evaluate the direct immobilization of SH-modified carbohydrates onto the gold surface, we used EC analysis, which has the advantages of easy detection and high sensitivity at low concentrations (19–21) and which might be a useful tool for practical application as a carbohydrate sensor that needs high sensitivity due to very low portion of carbohydrates in the cellular materials. EC analysis was used to obtain the cyclic voltammetry (CV) and square wave voltammetry (SWV) curves from SH-modified and unmodified GM1 pentasaccharide exposed to 200-µm diameter gold electrodes (Figure 3). The CV measurements showed that the EC responses were significantly enhanced for immobilized SH-GM1 pentasaccharide (Figure 3, inset). The peaks and dips in the redox current of the CV signal showed the typical oxidation and reduction responses in the electrolyte of K3Fe(CN)6. The SWV curves (Figure 3, main figure) revealed that the limiting redox current decreased from 1.05 on the bare gold electrode to 0.2 µA in the presence of immobilized SH-GM1 pentasaccharide. The change ratios of the redox currents of bare gold electrodes in the presence or absence of unmodified or SH-modified GM1 pentasaccharide were ∼45% and ∼80%, respectively. This clearly showed that the EC signal ratio changed in response to carbohydrate immobilization, indicating that EC could have potential carbohydrate sensor applications. The reduced redox current is due to the decrease in reactive surface area exposed to the electrolyte. Notably, while the current ratio decreased to some extent in the presence of unmodified GM1 pentasaccharide, likely due to physical adsorption from the relatively small carbohydrate size, the redox current was almost completely blocked by the binding of SH-GM1 pentasaccharide to the gold electrode. Future systematic studies may be warranted to examine additional levels of control that may be exerted over this physical adhesion. Therefore, these results collectively indicate that our proposed method is a highly efficient strategy for immobilizing carbohydrates to the tested substrate. Although the structure of the terminal reducing sugar is changed to the acyclic form by coupling of the aldehyde and amine groups (see Scheme 1), this should not be a problem in long chain carbohydrates such as GM1 pentasaccharide because
the generation of the acyclic form only occurred at the surfacelinked terminal sugar. A previous study showed that galactose and sialic acid, two terminal sugars of the branched GM1 pentasaccharide, exhibited substantial specific binding interactions (∼39% from galactose and ∼43% from sialic acid), with a smaller contribution (∼17%) arising from the N-acetyl galactosamine residue (26). However, the central galactose and terminal reducing glucose residues do not appear to make direct interactions with the receptor protein (26), indicating that the terminal reducing and central sugars are more likely to act as bridges or linkers. Accordingly, we examined the binding of GM1 pentasaccharide to its terminal galactose-binding lectin, galectin, by topology analysis using atomic force microscopy (AFM) (Figure 4). Our results revealed that galectin did not bind in the case of unmodified, nonimmobilized GM1 pentasaccharide (Figure 4A), whereas galectin was clearly detected as a white spot on SH-GM1 pentasaccharide-immobilized gold
Figure 4. AFM analysis of the GM1 pentasaccharide–galectin interaction on Au(111). (A) Unmodified pentasaccharide, (B) modified SHpentasaccharide, (C) magnified image of box in B, and (D) line profile of the SH-modified GM1 pentasaccharide–galectin interaction.
Bioconjugate Chem., Vol. 18, No. 6, 2007 2201
(Figure 4B and C). Line profile analysis revealed that the protein spots were 3–4 nm in size (Figure 4D). Importantly, since galectin should interact with only the terminal galactose, these findings confirmed that the modified carbohydrate had been successfully immobilized with the proper orientation. In the present work, we kept our method simple by using aminophenyl disulfide as both the coupling reagent and the linker. In the future, the use of a longer coupling reagent might help increase the flexibility and receptor binding of the target carbohydrate. However, since increasing the length of the alkyl chain group can change the hydrophobicity of the molecule, the use of an overly long chain could inhibit carbohydrate modification. One of the advantages of our immobilization method is that only the terminal reducing sugar is specifically coupled with the linker reagent, so that the functional sites (e.g., biomolecule recognition or binding sites) of the carbohydrate can be preserved. In summary, we herein demonstrated that the amine group of thiol-bearing aminophenyl disulfide was successfully coupled with the aldehyde group of the terminal reducing sugar in three types of carbohydrates, namely, glucose (monosaccharide), lactose (disaccharide), and GM1 pentasaccharide. The proposed modification method has several advantages over previously reported methods, including direct and rapid one-step immobilization onto a gold surface without surface pretreatment(s) by thiol group coupling in a mild reaction environment and exposure of functional carbohydrate moieties through oriented immobilization of the terminal reducing sugar. This modification and immobilization method should prove useful for diverse biomimetic studies in carbohydrates, including carbohydrate– biomolecule interaction and carbohydrate sensor or array development for diagnosis and screening.
ACKNOWLEDGMENT This work was supported by National R&D Project for Nano Science and Technology and Advanced Environmental Biotechnology Research Center at POSTECH from the Korea Science and Engineering Foundation and the Brain Korea 21 Program from the Ministry of Education, Korea.
LITERATURE CITED (1) Jelinek, R., and Kolusheva, S. (2004) Carbohydrate biosensors. Chem. ReV. 104, 5987–6015. (2) Bertozzi, C. R., and Kiessling, L. L. (2001) Chemical glycobiology. Science 291, 2357–2364. (3) Zachara, N. E., and Hart, G. W. (2002) The emerging significance of O-GlcNAc in cellular regulation. Chem. ReV. 102, 431–438. (4) Ritchie, G. E., Moffatt, B. E., Sim, R. B., Morgan, B. P., Dwek, R. A., and Rudd, P. M. (2002) Glycosylation and the complement system. Chem. ReV. 102, 305–320. (5) Nyquist, R. M., Eberhardt, A. S., Silks, L. A., Li, Z., Yang, X., and Swanson, B. I. (2000) Characterization of self-assembled monolayers for biosensor applications. Langmuir 16, 1793–1800. (6) Park, S. J., and Shin, I. J. (2002) Fabrication of carbohydrate chips for studying protein-carbohydrate interactions. Angew. Chem., Int. Ed. 41, 3180–3182. (7) Lazcka, O., Campo, F. J. D., and Munoz, F. X. (2007) Pathogen detection: A perspective of traditional methods and biosensors. Biosens. Bioelectron. 22, 1205–1217. (8) Ni, J. H., Singh, S., and Wang, L. X. (2003) Synthesis of maleimide-activated carbohydrates as chemoselective tags for site-specific glycosylation of peptides and proteins. Bioconjugate Chem. 14, 232–238.
(9) Hermanson, G. T., Mallia, A. K., and Smith, P. K. (1992) Immobilized Affinity Ligand Techniques, Vol. 1, pp 137–200, Academic Press, London. (10) Hatanaka, K., Takeshige, H., and Akaike, T. (1994) Synthesis of a new polymer containing uridine and galactose as pendant groups. J. Carbohydr. Biochem. 13, 603–610. (11) Shin, I., Park, S., and Lee, M. R. (2005) Carbohydrate microarrays: An advanced technology for functional studies of glycans. Chem.–Eur. J. 11, 2894–2901. (12) de Paz, J. L., and Seeberger, P. H. (2006) Recent advances in carbohydrate microarrays. QSAR Comb. Sci. 25, 1027–1032. (13) Kondo, A., Suzuki, J., Kuraya, N., Hase, S., Kato, I., and Ikenaka, T. (1990) Improved method for fluorescence labeling of sugar chains with sialic acid residues. Agric. Biol. Chem. 54, 2169–2170. (14) Hase, S., Ibuki, T., and Ikenaka, T. (1984) Reexamination of the pyridylamination used for fluorescence labeling of oligosaccharides and its application to glycoproteins. J. Biochem. (Tokyo) 95, 197–203. (15) Hase, S., Koyama, S., Daiyasu, H., Takemoto, H., Hara, S., Kobayashi, Y., Kyogoku, Y., and Ikenaka, T. (1986) Structure of a sugar chain of a protease inhibitor isolated from barbados pride (Caesalpinia pulcherrima Sw.) seeds. J. Biochem. (Tokyo) 100, 1–10. (16) (a) Wang, D. N., Liu, S. Y., Trummer, B. J., Deng, C., and Wang, A. L. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20, 275–281. (17) Lee, M., and Shin, I. (2005) Facile preparation of carbohydrate microarrays by site-specific, covalent immobilization of unmodified carbohydrates on hydrazide-coated glass slides. Org. Lett. 7, 4269–4272. (18) Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. G. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions. Nat. Biotechnol. 20, 1011–1017. (19) Zhi, Z., Powell, A. K., and Turnbull, J. E. (2006) Fabrication of carbohydrate microarrays on gold surfaces: Direct attachment of nonderivatized oligosaccharides to hydrazide-modified selfassembled monolayers. Anal. Chem. 78, 4786–4793. (20) Zhou, X., and Zhou, J. (2006) Oligosaccharide microarrays fabricated on aminooxyacetyl functionalized glass surface for characterization of carbohydrate-protein interaction. Biosens. Bioelectron. 21, 1451–1458. (21) Lee, H. Y., Park, J. W., and Kawai, T. (2004) SNPs feasibility of nonlabeled oligonucletides on using electrochemical sensing. Electroanalysis 16, 1999–2002. (22) Kim, J. M., Jung, H. S., Park, J. W., Yukimasa, T., Oka, H., Lee, H. Y., and Kawai, T. (2005) Spontaneous immobilization of liposomes on electron-beam exposed resist surfaces. J. Am. Chem. Soc. 127, 2358–2362. (23) Lee, H. Y., Park, J. W., Kim, J., Jung, H., and Kawai, T. (2006) Well-oriented nanowell array metrics for integrated digital nanobiosensors. Appl. Phys. Lett. 89, 113901. (24) Takahashi, C., Nakakita, S., and Hase, S. (2003) Conversion of pyridylamino sugar chains to corresponding reducing sugar chains. J. Biochem. (Tokyo) 134, 51–55. (25) Kuraya, N., and Hase, S. (1992) Release of O-linked sugar chains from glycoproteins with anhydrous hydrazine and pyridylamination of the sugar chains with improved reaction conditions. J. Biochem. (Tokyo) 112, 122–126. (26) Merritt, E. A., Sarfaty, S., Vandenakker, F., Lhoir, C., Martial, J. A., and Hol, W. (1994) Crystal-structure of cholera-toxin B-pentamer bound to receptor G(M1) pentasaccharide. Protein Sci. 3, 166–175. BC700288Z
