Site-Specific Attachment of a Protein to a Carbon Nanotube End

---

Article pubs.acs.org/bc

Site-Specific Attachment of a Protein to a Carbon Nanotube End without Loss of Protein Function Shige H. Yoshimura,*,†,⊥ Shahbaz Khan,‡,⊥ Satoshi Ohno,§ Takashi Yokogawa,§ Kazuya Nishikawa,§ Takamitsu Hosoya,|| Hiroyuki Maruyama,‡ Yoshikazu Nakayama,‡,⊥ and Kunio Takeyasu†,⊥ †

Graduate School of Biostudies, Kyoto University, Yoshida-konoe-cho, Sakyo-ku, Kyoto, 606-8501, Japan Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, 565-0871, Japan § Faculty of Engineering, Gifu University, 1-1 Yanagido, Gifu, 501-1193, Japan || Graduate School of Biomedical Science, Tokyo Medical and Dental University, 2-3-10 Kanda-surugadai, Chiyoda-ku, Tokyo, 101-0062, Japan ⊥ CREST, Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, 332-0012, Japan ‡

S Supporting Information *

ABSTRACT: Establishing a nanobiohybrid device largely relies on the availability of various bioconjugation procedures which allow coupling of biomolecules and inorganic materials. Especially, site-specific coupling of a protein to nanomaterials is highly useful and significant, since it can avoid adversely affecting the protein’s function. In this study, we demonstrated a covalent coupling of a protein of interest to the end of carbon nanotubes without affecting protein’s function. A modified Staudinger-Bertozzi ligation was utilized to couple a carbon nanotube end with an azide group which is site-specifically incorporated into a protein of interest. We demonstrated that Ca2+-sensor protein, calmodulin, can be attached to the end of the nanotubes without affecting the ability to bind to the substrate in a calciumdependent manner. This procedure can be applied not only to nanotubes, but also to other nanomaterials, and therefore provides a fundamental technique for well-controlled protein conjugation.

■

INTRODUCTION A number of single-molecule-based measurement and manipulation techniques have made enormous contributions to biological and chemical sciences over the last 20 years.1−3 The information derived from a single-molecule event is useful in elucidating the molecular mechanisms of biological reactions. In these techniques, biomolecules such as DNA or proteins are attached to larger substrates such as micrometer beads or silicon surfaces. In a single-molecule force measurement using optical tweezers, proteins and DNAs are attached to micrometer beads made of latex or other materials,4−6 and in the case of atomic force microscopy, molecules are attached to a large probe made of silicon or silicon nitride.7,8 In most cases, a single substrate can carry a vast number of molecules on the surface, and the device measures or captures only a single event that happens to only one of the molecules bound to the substrate surface. Therefore, the result obtained from each single measurement may vary if the molecular species attached to the substrate are heterogeneous, or if the manner of attachment is not uniform. These effects of heterogeneous events can be accounted for by statistical analysis; however, when the number of molecules in the measurement system is very small, molecular heterogeneity has a significant impact on the accuracy of the measurement. © XXXX American Chemical Society

When examining the function of a protein using a singlemolecule system, it is important to take into consideration how the protein molecules are to be attached to the substrate. In cases in which the protein is directly adsorbed onto the substrate surface, the portion of the protein in contact with the surface is enzymatically inactive, since access to the active site is spatially hindered. Alternatively, proteins can be anchored to the substrate via a flexible polymer chain such as poly(ethylene glycol) or DNA.8 Even in such cases, however, the point of attachment (amino acid) with the polymer may affect access of the protein’s substrate. Thus, the orientation of a protein molecule on the substrate surface is a critical consideration, especially when protein−protein or protein−substrate interactions are to be analyzed. Carbon nanotubes (CNTs) have revolutionized material sciences and the development of biosensors because of their biocompatibility and chemical inertness. A variety of biological molecules (proteins, carbohydrates, DNA, etc.) have been conjugated to carbon nanotubes for use in biosensors and other devices.9−11 In many cases, biomolecules were attached to the Received: March 15, 2012 Revised: May 31, 2012

A

dx.doi.org/10.1021/bc300131w | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

was described previously.20 The fusion protein was expressed in bacteria and purified using Ni-NTA resin (Qiagen) according to the manufacturer’s protocol. Coupling of Azide Proteins to CNTs and Observation Using Fluorescence Microscopy. SWNTs synthesized using the chemical vapor deposition (CVD) method (Nanocyl s.a.) were oxidized as described in the previous study22 and dispersed in ultrapure water using ultrasonication. CDIactivated Tween 20 was then added to a final concentration of 10%. The nanotubes were washed with 50 mM sodium borate (pH 7.5) to remove excess Tween 20 and then incubated with a fluorophore carrying an amino group (Texas Red hydrazide or BODIPY FL-hydrazide, Invitrogen) at 25 °C for 4 h. Excess fluorophore was removed either by passing the solution through a gel filtration column (Micro Bio-Spin 6, Bio Rad) or by repeated centrifugation and redissolution. The fluorescent-labeled oxidized nanotubes were reacted with ethylenediamine (2.5 mg/mL) and EDC (2.5 mg/mL) in 50 mM borate buffer (pH 7.5) for 20 min at 20 °C. After removal of free cross-linkers, the nanotubes were incubated with a NHScoupled triarylphosphine derivative (3-diphenylphosphino-4(methoxycarbonyl) benzoic acid N-hydroxysuccimide ester) (47 μM) at 20 °C for 60 min, and subsequently with purified calmodulin with or without azidotyrosine at position 80 (56 nM). The nanotubes were then incubated with purified ECFP carrying calmodulin peptide (1 μM) in the presence or absence of 5 μM CaCl2. The sample was observed on a conventional fluorescence microscope (IX71, Olympus) equipped with an EM-CCD camera (Hamamatsu Photonics) and fluorescence filter sets. Expression and Purification of VLPs. Expression vectors for rotavirus GFP-ΔVP2 and VP6 were kind gifts from Dr. Didier Poncet (Laboratoire de Virologie Moléculaire et Structurale, France). The DNA region encoding GFP was replaced by a DNA fragment encoding ECFP, which was amplified from ECFP-C1 vector using PCR. The plasmid (ECFP-ΔVP2 and VP6 in pFastBac1) was introduced into E. coli DH10Bac (Invitrogen) to produce a bacmid. Sf9 cells were transfected with the purified bacmid (ECFP-ΔVP2 or VP6) and the virus was harvested. The Sf9 cells were simultaneously infected with these two different viruses (ECFP-ΔVP2 and VP6) and VLPs were purified as described previously.23 Briefly, the cells were lysed with Triton X-100 and the cell debris was removed by passage through a 0.2 μm filter. Virus-like particles were concentrated using an ultrafiltration membrane (300 000 MWCO, Amicon). Purified VLPs were observed using the same microscopic settings as were used for the observation of CNTs (see above). The fluorescence intensity of VLPs was quantified using image-analysis software (MetaMorph, Molecular Imaging).

sidewall of the CNT via either noncovalent or covalent bonding. Noncovalent couplings are typically based on hydrophobic interactions between a CNT and a biomolecule (or hydrophobic moieties, such as pyrene, that are attached to the biomolecule).11,12 This type of method, due to its simplicity, has been used extensively in field-effect transistor (FET)-based sensor devices.10,13,14 Nonspecific adsorption on CNTs is an interesting phenomenon, but represents a relatively less controllable mode of attaching molecules to CNTs. Covalent couplings include chemical cross-linking between an oxidized CNT (carboxylic acid) and a biomolecule. Oxidation of a CNT (either by acid treatment or heating in air) produces carboxylic groups at the nanotube ends (as well as defect sites on the sidewalls),15−17 and has been used to attach molecules of interest to the ends of CNTs.16,18 Because of its small diameter, the end of a CNT is a good target site for trapping a small number of protein molecules, and therefore can be used for single-molecule measurements. However, even in such cases, it is difficult to control the attachment site within the protein because of the lack of suitable protein engineering techniques. Therefore, a procedure for covalently attaching a protein of interest to CNTs in a desired position and orientation is important and necessary for establishing new bioconjugation systems. In the previous study, we reported that a modified Staudinger-Bertozzi reaction can be used to covalently couple CNTs and azide groups in a protein.19 In this report, we utilized this reaction scheme to site-specifically attach a protein of interest to the end of a CNT without affecting the protein’s enzymatic function.

■

EXPERIMENTAL PROCEDURES Proteins. The cDNA encoding calmodulin carrying an amber mutation at residue 80 was prepared as described previously.20 The cDNA fragment encoding mouse importin α (a.a. 1−133) was amplified using PCR and subcloned into the pIVEX vector (Roche Applied Science). The codon corresponding to Phe54 or Phe124 was substituted with an amber codon (TAG) using PCR with mutated oligonucleotides. In vitro transcription and translation were carried out as described previously.20,21 Briefly, the suppressor tRNATyr(CUA) was synthesized in E. coli cells from the plasmid carrying the yeast amber suppressor tRNA gene, and then purified using ion exchange chromatography (DEAE-sepharose, GE Healthcare). The transcription/translation-coupled reaction (50 μL) contained 50 mM HEPES-KOH (pH 7.5), 7.7 mM Mg(OAc)2, 27.5 mM NH4OAc, 200 mM KOAc, 1.7 mM DTT, 1.25 mM ATP, 0.83 mM GTP, UTP, and CTP, 80 mM creatine phosphate, 0.21 mg/mL creatine kinase, 0.1 mg/mL T7 RNA polymerase, 4% poly(ethylene glycol) (average MW 8000 Da), 3.4 A260 units/mL E. coli tRNA mix, 30% E. coli cell extract, 80 μM tyrosine, 200 μM each of 19 amino acids (excluding tyrosine) (Peptide Institute Inc.), 500 μM 3-azidotyrosine (Watanabe Chemical Industries Ltd.), 0.17 mg/mL tyrosyltRNA synthetase mutant (TyrRS Y43G), 0.2 A260 units/mL yeast suppressor tRNATyr(CUA), and 0.02 mg/mL template DNA. Synthesized protein was purified using Ni-NTA resin (Qiagen) for importin α fragment or phenyl sepharose CL-4B (GE Healthcare) for calmodulin. To confirm the incorporation of azidotyrosine, synthesized protein was reacted with a rhodamine-conjugated triarylphosphine derivative (3-diphenylphosphino-4-(methoxycarbonyl) benzamide), separated by SDS-PAGE, and the gel was imaged using a fluoroimager. The construction of ECFP-fused calmodulin-binding peptide

■

RESULTS AND DISCUSSION Amino Acid-Specific Incorporation of an Azide Group into a Protein of Interest. The azide group is a chemically useful reactive group, but it does not exist in natural proteins.24−26 We employed an azide group to site-specifically attach a protein of interest to the end of a CNT. Amino acidspecific incorporation of azidotyrosine (Figure 1a) into the protein of interest was achieved using a previously reported procedure.20 Briefly, a cDNA encoding the protein of interest was mutated such that the codon for the target amino acid was replaced by the amber codon (TAG) (Figure 1b). Using this cDNA as a template, transcription and translation were carried B

dx.doi.org/10.1021/bc300131w | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

nanotube was first oxidized to produce carboxylic acid group(s),22 which then reacted with EDC and ethylenediamine to generate amino group(s) at the end. The coupling of the amino group with the azide group incorporated into the protein was achieved using a modified Staudinger-Bertozzi ligation,27,28 which utilizes an NHS-coupled triarylphosphine derivative (Figure 1c). The Staudinger reaction is highly specific for azide groups and therefore does not normally occur on CNT surfaces. In a previous study, an azido-carrying molecule (sugar) was specifically attached to acid-treated CNTs via a carbodiimide (DCC/HOBt).17 Our modified reaction scheme includes Staudinger-Bertozzi ligation as the final step of the conjugation, which catalyzes formation of the linkage between an amino group on the CNT and an azide group on the protein (Figure 1c). This reaction scheme was validated by a simple experimental system utilizing quantum dots (Supporting Information, Figure S1), as well as in our previous report.19 Calmodulin Functions at the End of a Carbon Nanotube. We applied the procedure described above to attach calmodulin to the end of a CNT and monitor enzymatic function. Calmodulin is a ubiquitous calcium-binding protein (KD = 0.5−5 (nM)) that regulates the activities of a variety of cellular enzymes. It undergoes a conformational change when bound to Ca2+, which then facilitates the interaction with the substrate (Figure 2a). We prepared full-length calmodulin carrying azidotyrosine at position 80 using the procedure described above. Tyr80 is not directly involved in the interaction with the substrate (Figure 2a). Biochemical analysis demonstrated that calmodulin containing azidotyrosine at position 80 retains Ca2+- and substrate-binding properties comparable to those of the wildtype protein (KD ∼10 nM).20 Azido-calmodulin was attached to the end of the CNT using the procedure described in Figure 1c (see also Experimental Procedures). CNTs (Figure 1d) were oxidized in the air to produce carboxyl groups at the ends.22 We employed mild oxidation condition in order to minimize damage to the sidewalls of the CNTs.22 The oxidized CNTs were mixed with fluorophore-coupled detergent (Tween 20Texas Red) to enable observation under fluorescence microscopy,29 and then reacted with EDC and ethylenediamine (EDA) to produce amino groups. Finally, the CNTs were incubated with the NHS-coupled phosphine derivative, after which they were ready for reaction with the azide group. The successful attachment of azido-calmodulin to the CNT was monitored by observing the binding and unbinding of ECFPfused calmodulin-binding peptide (ECFP-CBP). To obtain maximum binding efficiency, a saturating amount of the substrate (1 μM) was added to the nanotube/calmodulin hybrid. In the presence of 5 μM Ca2+ in the observation buffer, an ECFP signal was observed mainly at the ends of the CNTs (Figure 2b, arrowheads). Some of the CNTs carried the ECFP signal at both ends, but others had signal at only one end, because under our mild oxidation process, not all of the CNT ends were oxidized and some of them still remained as closed ends judging from electron micrographs. In total, approximately 17% of the CNTs in the observation fields had an ECFP signal on at least one end (Figure 2c). Additional images are shown in Figure S2, Supporting Information. The ECFP signal was also detected on the sidewall of the CNT, but with very low frequency (approximately 2% of the total nanotubes) (Figure 2b, asterisk). This was probably due to the production of carboxyl groups not only at the ends of the CNT during the

Figure 1. Reaction scheme for the site-specific coupling of a protein to the tip of a nanotube. (a) Structure of azidotyrosine. (b) Schematic illustration of amino acid specific incorporation of azidotyrosine into the protein of interest. The details of the procedure are described in the text and in the Experimental Procedures section. (c) Reaction scheme for coupling an azido-containing protein to the end of the CNT. The details of the reaction are described in the main text. (d) Transmission electron micrograph of CNTs utilized in this study. Most of the nanotubes (>95%) are single-walled. We utilized mild oxidation process to avoid the damage to the sidewall of the CNT. Because of this oxidation procedure, not all of the nanotube ends were oxidized and opened. Scale bar: 50 nm.

out in vitro. The translation reaction mixture contained a suppressor tRNA carrying a CUA anticodon, azidotyrosine (Figure 1b), and the corresponding aminoacyl-tRNA synthetase, so that the azidotyrosine could be incorporated specifically into the position of the TAG codon (see Experimental Procedures section for the detailed procedure). The reaction scheme for the covalent coupling of an azide group to the end of the nanotube is depicted in Figure 1c. The end of the C

dx.doi.org/10.1021/bc300131w | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

oxidation process, but also on its sidewall, even though our oxidation process was mild. In the absence of Ca2+, in contrast, the frequency of detecting ECFP signals on the CNTs was drastically reduced (