13473
2008, 112, 13473–13475 Published on Web 10/07/2008
New Insight into the Surface Denaturation of Proteins: Electronic Sum Frequency Generation Study of Cytochrome c at Water Interfaces Pratik Sen, Shoichi Yamaguchi, and Tahei Tahara* Molecular Spectroscopy Laboratory, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan ReceiVed: July 11, 2008; ReVised Manuscript ReceiVed: September 3, 2008
In situ characterization of surface denaturation of a protein was realized by newly developed interface-selective multiplex electronic sum frequency generation spectroscopy. The observed electronic spectra of cytochrome c at the air/water interface exhibited a broad feature, which demonstrated coexistence of the nativelike and denatured protein at the interface. This situation of the mixed conformation at the air/water interface did not change in the acidic condition of pH ) 2 where the protein was completely denatured in the bulk water. In sharp contrast, only native spectrum was observed at the silica/water interface. Introduction The biologically active three-dimensional structure of a protein is realized by delicate balance of electrostatic and hydrophobic interactions among the residues.1 Slight imbalance, caused by certain agents (e.g., urea, surfactant, pH, heat, and so forth), induces denaturation of the protein and impairs its biological activity. Water soluble proteins are so constructed that the hydrocarbon parts are buried in the interior, leaving the surface covered by the polar parts. When a protein having such a structure reaches an interface, there is a strong tendency for the hydrocarbon parts of the protein to go to the hydrophobic phase of the interface. Consequently, the protein undergoes a conformational rearrangement to maximize the number of favorable interactions, leading to unfolding of the protein (see Figure 1). This phenomenon, that is, surface denaturation of proteins, is a well-known phenomenon from very early days.2 In addition, recently, the development of the engineering of twodimensional protein arrays becomes very important for their potential applications in biotechnology, in particular as biosensors.3 The main component of a biosensor is a thin native protein film, and the problem to be avoided is the surface denaturation of a protein. Thus, the study of surface denaturation is also important in this context. The surface denaturation of proteins has been studied by a variety of methods such as X-ray and neutron reflections,4 STM,5 conventional spectroscopic techniques with the foam formation method,6 and the stability of proteins at the air/water interface has been discussed. For instance, using X-ray reflection, Rice and co-workers4a reported the complete denaturation of glucose oxidase, alcohol dehydrogenase, and urease at the air/water interface. However, Lu and co-workers4b claimed that lysozyme remains native at the air/water interface. Although the X-ray and neutron reflectivity measurements provide unique information about the molecular dimension at the interface, the analysis is made on the assumption that the protein forms a uniform sheet at the interface. Some other techniques, such as STM, need the transfer of the protein monolayer from the air/water * To whom correspondence should be addressed. E-mail:
[email protected].
10.1021/jp8061288 CCC: $40.75
interface to the substrate before analysis. The conformation of the protein may be changed during the transfer process. In the case of spectroscopic study with the foam formation method, the renaturation of the protein during extraction of foam leaves some ambiguity. Therefore, it has been desired to perform in situ measurements of the protein conformation at interfaces using interface-selective spectroscopy. Second order nonlinear spectroscopy is a powerful tool to study molecules at interfaces.7-9 Several attempts to study proteins at different interfaces have been reported in the past few years using vibrational sum frequency generation (VSFG) or second harmonic generation (SHG) spectroscopy.8,9 Recently, we developed multiplex electronic sum frequency generation (ESFG) spectroscopy (see Figure 2) that provides electronic spectra of interfacial molecules with very high quality.7a,b In this work, we applied ESFG to in situ spectroscopic characterization of protein conformation at interfaces, using cytochrome c as a model protein. Experimental Section The experimental setup of the multiplex ESFG spectroscopy (Figure 2) has already been reported in detail elsewhere.7a-d A Ti:sapphire regenerative amplifier (Spitfire, Spectra Physics) seeded by a mode-locked oscillator (Tsunami, Spectra Physics) generated 795 nm, 0.9 mJ pulses at repetition rate of 1 kHz (fwhm ) 130 fs). A part of the output was attenuated by a variable-ND filter and used as the narrow-band ω1 pulse (bandwidth 160 cm-1). Another part of the output was focused into water to generate white light continuum that was used as the ω2 pulse. The spectrum of the ω2 pulse extended from 540 nm to 1.2 µm. Typical pulse energy of the narrow-band ω1 and broadband ω2 were 16 and 7 µJ, respectively, at the sample position. The ω1 and ω2 pulses were noncollinearly focused onto the same spot (φ ∼ 0.1 mm) at the interface of a sample solution. The incident angles of the ω1 and ω2 laser pulses from the interface normal were 48 and 38°, respectively, in the same plane of incidence. The linear polarization of the ω1 and ω2 beams was set in the plane of reflection (i.e., p-polarization). When the ω1 and ω2 pulses were temporally overlapped, sum 2008 American Chemical Society
13474 J. Phys. Chem. B, Vol. 112, No. 43, 2008
Letters
Figure 1. Sketch of a protein (a) in bulk water and (b) at the air/water interface. Hydrophobic and hydrophilic parts of the protein are represented by red and blue color, respectively.
Figure 2. (a) Ray diagram of the multiplex ESFG setup. ω1 is 795 nm (bandwidth: 160 cm-1) and ω2 is white light continuum (540–1200 nm). (ω1 + ω2) is the ESFG signal. (b) Energy diagram of the ESFG process.
Figure 3. ESFG spectra of cytochrome c at (a) silica/water interface, bulk pH ) 7, (b) air/water interface, bulk pH ) 7, (c) air/water interface, bulk pH ) 2. Red lines represent ESFG spectra. Orange and blue lines represent the UV-visible absorption spectra of native and denatured protein in the bulk water, respectively.
frequency (SF, ω1 + ω2) was generated at the interface. The SF light was collected and focused onto the entrance slit of a single polychromator (HR-320, Jobin Yvon). The spectrally dispersed SF light was detected by liquid nitrogen cooled CCD (Spec-10:2KBUV, Roper Scientific). Unwanted second harmonic generation (SHG) signals (i.e., 2ω1 and 2ω2) were rejected by a spatial filter and by a background subtraction on a computer. A cylindrical glass cell was used to contain the sample solution that was stirred by a 2 mm magnetic stirring flea.
Cytochrome c was purchased from Sigma and dissolved in HPLC-grade distilled water (Wako) without further purification. The bulk concentration of cytochrome c was 65 µM in this study. All the experiments were performed at 299 K. Result and Discussion Horse heart cytochrome c is a heme protein, exhibiting its characteristic Soret absorption band around 400 nm.10 At neutral pH, cytochrome c has the native structure and two strong field
Letters ligands (His18 and Met80) are coordinated in the axial position of the heme iron.11 The axial ligation to the iron is sensitive to the structural change of the protein, which can be induced by the changes of pH, temperature, ionic strength, and solvent composition, etc. The difference in the axial ligation reflects on the spin state of the iron, which causes the shift of the Soret band. Thus, the peak wavelength of the Soret band indicates the condition of ligand binding, oxidation state of iron, and conformation of the polypeptide in the vicinity of the heme group. In the native state, the Soret band is centered at 410 nm.11 In the fully denatured state (pH ) 2), neither His18 nor Met80 is coordinated with the heme,11 and the Soret band is blue-shifted to 394 nm. In the intermediate five coordinated state, Met80 is absent and the maxima of the Soret band appear in between 397 and 406 nm, depending on the nature of the intermediate state. Thus, we can directly obtain information about conformation of cytochrome c at the interface by measuring interface-selective electronic spectra in the Soret band region. We first measured the ESFG spectrum of cytochrome c at the silica/water interface (Figure 3a). It is readily seen that, the multiplex ESFG technique reliably provides the electronic spectrum of the protein at the interface. In the present experimental condition, the ESFG spectrum was measured under the two-photon resonant and one-photon nonresonant condition, so that it can be directly compared with the linear absorption spectrum of cytochome c. The spectrum at the silica/water interface is very similar to the absorption spectrum of the native cytochrome c, indicating the retention of the native structure of the protein at this interface. In the case of the silica/water interface, the two phases have similar polarity and the perturbation at this interface does not induce denaturation. Recently, Chemg et al. also obtained a similar result using total internal reflection absorption spectroscopy.11c Figure 3b shows the ESFG spectrum of cytochrome c at the air/water interface at pH ) 7, where the protein has the native structure in the bulk water. It is readily seen that the Soret band of cytochrome c at the air/water interface is quite different from that of native cytochrome c in the bulk water. This reveals that the conformation of cytochrome c at the air/water interface is certainly different from that in the bulk environment. It should be noted that the interfacial Soret spectrum is also very different from the Soret spectrum of the denatured protein in the bulk (pH ) 2). The observed ESFG spectrum at the air/water interface is much broader and exhibits a plateaulike feature around the intensity maximum in the region between the absorption maxima of the native and denatured states. We consider two possibilities as the origin of this different interfacial Soret spectrum. First, an intermediate (e.g., molten globule state, and so forth) is realized at the air/water interface, which exhibits the absorption maximum between the maxima of the native and denatured state. The second possibility is the presence of mixed conformation of the protein at the air/water interface. It was reported that the widths of the Soret band of different intermediate states of cytochrome c are very similar to those of the native and denatured state.10 Thus, the broad feature of the Soret band at the air/water interface is not assignable to a single conformation of cytochrome c. In addition, the plateaulike feature around the intensity maximum is readily rationalized as the overlap of two closely located absorption bands. Therefore, the broad ESFG spectrum observed at the air/water interface is attributable to the mixed conformation of the protein at the air/water interface, that is, simultaneous presence of nativelike and denatured proteins at the interface. We note that the Soret band is sensitive only to the structural
J. Phys. Chem. B, Vol. 112, No. 43, 2008 13475 change around the heme. Therefore, the species that exhibits the nativelike spectrum might be somewhat different from the real native cytochrome c, and the structure of its outer part might be changed. In this sense, we cannot completely rule out the possibility of the presence of multiple conformations of denatured cytochrome c at the air/water interface. To examine the effect of bulk denaturation on the interfacial situation of cytochrome c, we also measured the ESFG spectra of cytochrome c at the air/water interface under an acidic condition (pH ) 2) (Figure 3c). Surprisingly, the ESFG spectrum is very similar to the spectrum measured under the neutral condition. This indicates that the nativelike and denatured protein simultaneously exists at the air/water interface also at pH ) 2, although the protein is completely denatured in the bulk. The interfacial perturbation at the air/water interface is so strong that the existence of multiple conformation of cytochrome c is not noticeably affected by the change of pH in the bulk. In conclusion, with newly developed ESFG spectroscopy, we have successfully measured electronic spectra of a protein, cytochrome c, at the silica/water and air/water interfaces and realized in situ characterization of its conformation. We observed the native conformation of cytochrome c at the silica/water interface, whereas a broad Soret spectrum, which was interpreted as the sum of the spectra of the native and denatured protein, was observed at the air/water interface. This demonstrates that cytochrome c exists as a mixture of native and denatured states at the air/water interface. The situation of the protein at the air/ water interface did not noticeably change under neutral and acidic conditions, because the effect of the strong interfacial perturbation was predominant. Acknowledgment. P.S. thanks JSPS for a postdoctoral fellowship. This work is supported by Grant-in-Aid for Scientific Research (A) (No. 19205005) from JSPS and Grant-in-Aid for Science Research on Priority Area (No. 19056009) from MEXT. References and Notes (1) (a) Karplus, M. Fold Design 1997, 2, 569. (b) Arai, M.; Kuwajima, K. AdV. Protein Chem. 2000, 53, 209. (2) (a) Bull, H. B.; Neurath, H. J. Biol. Chem. 1937, 118, 163. (b) Macritchie, F. AdV. Protein Chem. 1978, 32, 283. (3) (a) Stelzle, M.; Weissmuller, G.; Sackmann, E. J. Phys. Chem. 1993, 97, 2974. (b) Miyasaka, T.; Koyama, K.; Itoh, I. Science 1992, 255, 342. (4) (a) Gidalevitz, D.; Huang, Z.; Rice, S. A Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 2608. (b) Lu, J. R.; Su, T. J.; Thomas, R. K.; Penfold, J.; Webster, J. J. Chem. Soc., Faraday Trans. 1998, 94, 3279. (5) (a) Liu, Z.-F.; Manivannan, A.; Inokuchi, H.; Yanagi, H. J. Vac. Sci. Technol., B 1993, 11, 1766. (b) Liu, Z.-F.; Manivannan, A.; Yanagi, H.; Ashida, M.; Fujishima, A.; Inokuchi, H. Surf. Sci. Lett. 1993, 284, L411. (6) (a) Clarkson, J. R.; Cui, Z. F.; Darton, R. C. J. Colloid Interface Sci. 1999, 215, 323. (b) Lechevalier, V.; Croguennec, T.; Pezennec, S.; GuerinDubiard, C.; Pasco, M.; Nau, F. J. Agric. Food Chem. 2003, 51, 6354. (7) (a) Yamaguchi, S.; Tahara, T. J. Phys. Chem. B 2004, 108, 19079. (b) Yamaguchi, S.; Tahara, T. J. Chem. Phys. 2006, 125, 194711. (c) Sekiguchi, K.; Yamaguchi, S.; Tahara, T. J. Chem. Phys. 2008, 128, 114715. (d) Yamaguchi, S.; Tahara, T. Laser Photon. ReV. 2008, 2, 74. (e) Shen, Y. R.; Ostroverkhov, V. Chem. ReV. 2006, 106, 1140. (8) (a) Petralli-Mallow, T. P.; Plant, A. L.; Lewis, M. L.; Hicks, J. M. Langmuir 2000, 16, 5960. (b) Chen, X.; Wang, J.; Boughton, A. P.; Kristalyn, C. B.; Chen, Z. J. Am. Chem. Soc. 2007, 129, 1420. (9) (a) Salafsky, J. S. Chem. Phys. Lett. 2003, 381, 705. (b) Salafsky, J. S. J. Chem. Phys. 2006, 125, 74701. (10) (a) Goto, Y.; Hagihara, Y.; Hamada, D.; Hoshino, M.; Nishii, I. Biochemistry 1993, 32, 11878. (b) Sanghera, N.; Pinheiro, T. J. Protein Sci. 2000, 9, 1194. (c) Moosavi-Movahedi, A. A.; Chamani, J.; Goto, Y.; Hakimelahi, G. H. J. Biochem. 2003, 133, 93. (11) (a) Babul, J.; Stellwagen, E. Biochemistry 1972, 11, 1195. (b) Bren, K. L.; Gray, H. B. J. Am. Chem. Soc. 1993, 115, 10382. (c) Chemg, Y.-Y.; Lin, S. H.; Chang, H.-C.; Su, M.-C. J. Phys. Chem. A 2003, 107, 10687.
JP8061288
