6852
Langmuir 2008, 24, 6852-6857
Probing the Interaction between Peptides and Metal Oxides Using Point Mutants of a TiO2-Binding Peptide Haibin Chen,† Xiaodi Su,‡ Koon-Gee Neoh,† and Woo-Seok Choe*,†,§ Department of Chemical & Biomolecular Engineering, National UniVersity of Singapore, 10 Kent Ridge Crescent, Singapore 119260, Institute of Materials Research and Engineering, 3 Research Link, Singapore 117602, and Department of Chemical Engineering, Sungkyunkwan UniVersity, Suwon, Korea 440-746 ReceiVed January 29, 2008. ReVised Manuscript ReceiVed April 7, 2008 An increasing number of peptides with specific binding affinity to inorganic materials are being isolated using combinatorial peptide libraries without prior knowledge about the interaction between peptides and target materials. The lack of understanding of the mechanism and the contribution of constituent amino acids to the peptides’ inorganicbinding ability poses an obstacle to optimizing and tuning of the binding affinity of peptides to inorganic materials and thus hinders the practical application of these peptides. Using the phage surface display technique, we previously identified a disulfide-bond-constrained peptide (-CHKKPSKSC-, STB1) cognitive of TiO2. In the present study, the interaction of STB1 with TiO2 was probed using a series of point mutants of STB1 displayed on phage surfaces. Their binding affinity was measured using a quartz crystal microbalance with energy dissipation measurement and compared on the basis of the ∆f or ∆D values. The three K residues of STB1 were found to be essential and sufficient for phage particle binding to TiO2. One mutant with five K residues showed not stronger but weaker binding affinity than STB1 due to its conformational restriction, as illustrated by molecular dynamics simulation, to align five K residues in a way conducive to their simultaneous interaction with the TiO2 surface. The contextual influence of noncharged residues on STB1’s binding affinity was also investigated. Our results may provide insight into the electrostatic interaction between peptides and inorganic surfaces.
Introduction Proteins are able to recognize and even condense specific inorganic materials as evidenced by various biomineralization processes in nature.1–3 Despite the encouraging examples from nature, there has been no rational route to identify a protein molecule with specific binding affinity to desired synthetic inorganic materials due to a poor understanding of the complicated interaction between protein molecules and inorganic materials.4–6 Currently, this problem is circumvented by employing combinatorial peptide libraries, which enables one to isolate peptides with specific binding affinity to inorganic materials without prior knowledge about the interaction between peptides and target inorganic materials.7–10 The desired proteins can be genetically engineered with the selected peptides to gain inorganic-binding ability.10–13 * To whom correspondence should be addressed. Phone: +82 31 290 7344. Fax: +82 31 290 7272. E-mail:
[email protected]. † National University of Singapore. ‡ Institute of Materials Research and Engineering. § Sungkyunkwan University. (1) Lowenstam, H. A.; Weiner, S. On Biomineralization; Oxford University Press: New York, 1989. (2) Mann, S. Biomineralization; Oxford University Press: Oxford, New York, 2001. (3) Ba¨uerlein, E. Biomineralization; Wiley-VCH: Weinheim, Germany, 2004. (4) Nakanishi, K.; Sakiyama, T.; Imamura, K. J. Biosci. Bioeng. 2001, 91, 233–244. (5) Gray, J. J. Curr. Opin. Struct. Biol. 2004, 14, 110–115. (6) Patwardhan, S. V.; Patwardhan, G.; Perry, C. C. J. Mater. Chem. 2007, 17, 2875–2884. (7) Sarikaya, M.; Tamerler, C.; Schwartz, D. T.; Baneyx, F. O. Annu. ReV. Mater. Res. 2004, 34, 373–408. (8) Kriplani, U.; Kay, B. Curr. Opin. Biotechnol. 2005, 16, 470–475. (9) Baneyx, F.; Schwartz, D. T. Curr. Opin. Biotechnol. 2007, 18, 312–317. (10) Sarikaya, M.; Tamerler, C.; Jen, A. K. Y.; Schulten, K.; Baneyx, F. Nat. Mater. 2003, 2, 577–585. (11) Dai, H.; Choe, W.-S.; Thai, C. K.; Sarikaya, M.; Traxler, B. A.; Baneyx, F.; Schwartz, D. T. J. Am. Chem. Soc. 2005, 127, 15637–15643. (12) Sano, K. I.; Sasaki, H.; Shiba, K. J. Am. Chem. Soc. 2006, 128, 1717– 1722.
From a material engineering point of view, metal oxides are frequently encountered at interfaces, and many peptides have been isolated for metal oxides, such as MnO2,14 Cr2O3,14 CoO,14 PbO2,14 Fe2O3,15 SiO2,16–18 TiO2,18 Cu2O,19 ZnO,19,20 and Al2O3,13 using combinatorial peptide libraries. Among the selected peptides, a high percentage of basic residues (His, Lys, Arg) are generally observed, and electrostatic interaction is usually postulated as the major force to bind peptides to metal oxides. This postulation is quite plausible because metal oxide surfaces are generally negatively charged21 and the basic residues of the selected peptides are positively charged in an aqueous environment at neutral pH. Besides basic residues, nonbasic residues are also found in affinity peptides selected for metal oxides. However, the contribution of each constituent residue of peptide binders to their metal oxide-binding affinity is not clear in most studies because the interaction between proteins or peptides with inorganic surfaces is so complicated that it is quite challenging to reveal the mechanism at the atomic level using either experimental methods or theoretical simulations.4–6 Therefore, many questions remain unanswered regarding the interaction between selected peptides with target metal oxides: e.g., whether all the constituent basic residues are involved in the electrostatic interaction with metal oxide surfaces, which nonbasic residues (13) Krauland, E. M.; Peelle, B. R.; Wittrup, K. D.; Belcher, A. M. Biotechnol. Bioeng. 2007, 97, 1009–1020. (14) Schembri, M. A.; Kjaergaard, K.; Klemm, P. FEMS Microbiol. Lett. 1999, 170, 363–371. (15) Brown, S. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 8651. (16) Naik, R. R.; Brott, L.; Carlson, S. J.; Stone, M. O. J. Nanosci. Nanotechnol. 2002, 2, 95–100. (17) Eteshola, E.; Brillson, L. J.; Lee, S. C. Biomol. Eng. 2005, 22, 201–204. (18) Chen, H.; Su, X.; Neoh, K.-G.; Choe, W.-S. Anal. Chem. 2006, 78, 4872– 4879. (19) Thai, C. K.; Dai, H.; Sastry, M. S. R.; Sarikaya, M.; Schwartz, D. T.; Baneyx, F. Biotechnol. Bioeng. 2004, 87, 129–137. (20) Kjærgaard, K.; Sorensen, J. K.; Schembri, M. A.; Klemm, P. Appl. EnViron. Microbiol. 2000, 66, 10–14. (21) Hunter, R. J. Foundations of Colloid Science; Oxford University Press: New York, 1989.
10.1021/la800314p CCC: $40.75 2008 American Chemical Society Published on Web 06/06/2008
Interaction between Peptides and Metal Oxides
are dispensable, and how the binding affinity or selectivity of a peptide binder will be affected by changing some of its residues. The answers to these questions may help us optimize and tune the binding affinity or selectivity of peptides toward the target. In a previous study, we used peptide libraries displayed on phage surfaces and isolated disulfide-bond-constrained peptides with specific binding affinity to TiO2 nanoparticles.18 A total of 24 phage clones bound to TiO2 were finally analyzed by DNA sequencing, and 22 clones displayed the same peptide sequence. Such a consensus peptide binder provides an excellent model to study the influence of amino acid composition and sequence on the peptide’s binding affinity to TiO2. In the present study, 16 point mutants of STB1 (-CHKKPSKSC-) were created by mutating the STB1 peptide displayed on the phage surface. As demonstrated in the previous study,18 unique ∆f and ∆D signatures of quartz crystal microbalance with energy dissipation (QCMD) measurements provide useful tools to differentiate the STB1mediated binding of phage particles to TiO2 from the nonspecific interaction between phage bodies and TiO2. Therefore, the QCM-D technique was used again to measure and qualitatively compare the binding affinity among various STB1 mutants. Phage particles were employed as a contextual scaffold necessary for binding affinity study of STB1 and its point mutants because of the following three merits: (1) Only ∼5 copies of peptides are displayed on the pIII protein of each phage particle; therefore, the nonspecific interaction between phage bodies and metal oxides can compete with and even screen weak interactions between the displayed peptides and metal oxides, which makes weaker binders easily differentiated. (2) For applications, inorganicbinding peptides are usually fused to some host proteins and behave in the context of host proteins; therefore, the pIII protein on phage surfaces can serve as a host protein. (3) If free peptides are used, they can form large clusters when absorbed on solid surfaces, which complicates binding affinity assessment,22 and hence, displaying peptides on phage surfaces may prevent such aggregation. To help illustrate the changes in the peptides’ binding affinity, the conformations of STB1 and its point mutants were demonstrated by molecular dynamics simulations.
Experimental Section Oligonucleotide-Directed Mutagenesis of M13 Phage DNA. Phage clones displaying each point mutant of STB1 were created by oligonucleotide-directed mutagenesis of M13 phage DNA. All the protocols for phage manipulation and mutagenesis were adopted from ref 23. Briefly, the template DNA for site-directed mutagenesis is uracil-containing single-stranded DNA of an appropriate phage clone (e.g., STB1-P) following its amplification using Escherichia coli CJ236 (New England Biolabs, catalog no. E4141S). The purified template DNA was annealed with a primer, phosphorylated mutagenic oligonucleotide containing the desired base changes (synthesized by Research Biolabs, Singapore), and mutated DNA was synthesized in the presence of T4 DNA polymerase, T4 DNA ligase, and dNTP in T4 DNA ligase reaction buffer (New England Biolabs, catalog nos. M0203S, M0202S, N0447S, and B0202S). Subsequently, both template DNA and mutated DNA were transformed by electroporation into competent E. coli ER2738 cells capable of destructing the uracil-containing template DNA strand while leaving the mutated DNA strand to duplicate and direct the production of phage particles. The transformed cells were mixed with melted 2× YT top agar and plated on LB agar plates. The plates were incubated overnight at 37 °C to allow phage plaques to form, and then individual phage plaques (22) Goede, K.; Grundmann, M.; Holland-Nell, K.; Beck-Sickinger, A. G. Langmuir 2006, 22, 8104–8108. (23) Sambrook, J.; Russell, D. W. Molecular Cloning: A Laboratory Manual, 3rd ed.; Cold Spring Harbor Laboratory Press: New York, 2001; Vol. 2, Chapter 13.
Langmuir, Vol. 24, No. 13, 2008 6853 were sampled (6-12 plaques) and amplified separately for DNA extraction. Finally, the desired mutants were identified by DNA sequencing, amplified at large scale, and stored at -20 °C. Multiple rounds of mutagenesis were performed when the target clone had two or more residues different from those of the template clone. QCM-D Measurement. The simultaneous measurements of f and D were undertaken using the QCM-D system from Q-Sense AB (Gothenburg, Sweden). Five megahertz, AT-cut quartz crystal sensors coated at one side with Ti were purchased from Q-Sense AB (catalog no. QSX 310). Immediately prior to use, the Ti surface was treated in a standard UV/ozone chamber for 30 min and subsequently washed with DI water. Using XPS analysis, it was confirmed that a layer of TiO2 was formed on the Ti-coated quartz crystal following the UV/ozone treatment (XPS spectrum not shown). The measurement of phage particle binding was conducted in a stagnant liquid cell where the solution contacts the TiO2 side of the quartz crystal sensor. TBS buffer (50 mM Tris-HCl, 150 mM NaCl, pH 7.5) with 0.5% (v/v) Tween 20 was used as the binding buffer. The phage particle solutions for QCM-D measurements were prepared by diluting the phage stock in the binding buffer to the final concentration of 7 × 1010 pfu/mL. For each measurement, the binding buffer was first loaded into the liquid cell. The solution and the liquid cell were temperature stabilized at 24.7 ( 0.1 °C to avoid drifts in f and D. Following the achievement of stable baselines for f and D, the phage solution was added into the liquid cell to replace the binding buffer. After ∼30 min of interaction, the liquid cell was rinsed with the binding buffer. Each measurement was repeated twice, and the data deviation is within 5%. The f and D values measured at the third overtone were presented for analysis. The noise values of f and D with the liquid load are 1 Hz and 0.2 × 10-6, respectively. Molecular Dynamics Simulation. The conformations of the free STB1 peptide and its point mutants in solutions were determined using commercial molecular simulation software, Discovery Studio 1.7 (Accelrys Inc., San Diego, CA, http://www.accelrys.com/). The peptide chains were constructed using the peptide builder module in Discovery Studio. The disulfide bond between two flanking cysteine residues was manually connected, and a cyclic structure was obtained. Then the cyclic peptide was solvated in a orthorhombic water box extending at least 7 Å from any peptide atom (see an example in Figure 3a). To make the simulation box neutral, salt ions (Na+ and Cl-) at a concentration of 145 mM were also added. We performed a molecular dynamics study in the NVT ensemble using the Charmm27 force field. First, the configurational energy of the system was minimized by performing 200 steps of an adopted basis NR algorithm followed by 200 steps of a conjugate gradient algorithm. Then the system was heated to the target temperature (298 K) gradually and the dynamics run for 80 ps to reach equilibrium. Subsequently, a 200 ps production stage was carried out at a time step of 1 fs with the nonbond cutoff being 12 Å. To save computation time, the SHAKE algorithm and particle mesh Ewald method were applied in both the equilibrium stage and the production stage. The simulated conformations of the peptides were visualized and their surface electrostatic potentials were obtained using Discovery Studio 1.7.
Results and Discussion Table 1 lists the 17 phage clones with the peptide sequences displayed on their surfaces that were used in this study. A total of 16 mutants of STB1-P phage particles (single or multiple point(s)) were created by oligonucleotide-directed mutagenesis. We observed that the efficiencies of mutagenesis (the percentage of correct mutants in the pools of sampled phage clones) were very low (
