Driving Force Behind Adsorption-Induced Protein Unfolding: A Time

Dec 11, 2008 - The time dependence of the density profile at the air/water interface ... adsorbed at interfaces is determined by a balance of various ...
0 downloads 0 Views 723KB Size
32

Langmuir 2009, 25, 32-35

Driving Force Behind Adsorption-Induced Protein Unfolding: A Time-Resolved X-ray Reflectivity Study on Lysozyme Adsorbed at an Air/Water Interface Yohko F. Yano,*,† Tomoya Uruga,‡ Hajime Tanida,‡ Hidenori Toyokawa,‡ Yasuko Terada,‡ Masafumi Takagaki,‡ and Hironari Yamada† Synchrotron Light Life Science Center, Ritsumeikan UniVersity, 1-1-1 Noji-Higashi, Kusatsu-shi, Shiga 525-8577, Japan, and Japan Synchrotron Radiation Research Institute (JASRI), 1-1-1 Kouto, Sayo, Hyogo 679-5198, Japan ReceiVed October 2, 2008. ReVised Manuscript ReceiVed NoVember 11, 2008 Time-resolved X-ray reflectivity measurements for lysozyme (LSZ) adsorbed at an air/water interface were performed to study the mechanism of adsorption-induced protein unfolding. The time dependence of the density profile at the air/water interface revealed that the molecular conformation changed significantly during adsorption. Taking into account previous work using Fourier transform infrared (FTIR) spectroscopy, we propose that the LSZ molecules initially adsorbed on the air/water interface have a flat unfolded structure, forming antiparallel β-sheets as a result of hydrophobic interactions with the gas phase. In contrast, as adsorption continues, a second layer forms in which the molecules have a very loose structure having random coils as a result of hydrophilic interactions with the hydrophilic groups that protrude from the first layer.

Introduction Since proteins fold their hydrophobic regions within hydrophilic regions in water, conformational changes are expected when they adsorb at water interfaces.1 The conformation of protein molecules adsorbed at interfaces is determined by a balance of various forces, including van der Waals attraction, electrostatic repulsion, hydrogen bonding, and more complex interactions resulting from the hydrophobic effect and changes in conformational entropy associated with adsorption.2 One approach toward understanding the mechanisms of protein folding is to study the procedures and conditions that lead to protein unfolding. Adsorption kinetics has been studied by a number of techniques, including dynamic surface tension,3-7 radiotracer technique,6 and ellipsometry7 to determine adsorption isotherms, and surfacesensitive spectroscopies8,9 to determine secondary structures of the adsorbed protein. X-ray or neutron reflectivity techniques are powerful tools for exploring surface phenomena and allow one to obtain a density profile along the surface normal.10,11 Since the conventional * Corresponding author. E-amil: [email protected]; phone: +8177-561-3798; fax: +81-77-561-2680. † Ritsumeikan University. ‡ JASRI.

(1) Cheesman, D. F.; Davies, J. T. AdV. Protein Chem. 1954, 9, 439–501. (2) Sengupta, T.; Razumovsky, L; Damodaran, S. Langmuir 1999, 15, 6991– 7001. (3) Graham, D. E.; Phillips, M. C. J. Colloid Interface Sci. 1979, 70, 403–414. (4) Tripp, B. C.; Magda, J. J.; Andrade, J. D. J. Colloid Interface Sci. 1995, 173, 16–27. (5) Yampolskaya, G.; Platikanov, D. AdV. Colloid & Interface Sci. 2006, 128, 159–183. (6) Hunter, J.; Kilpatrick, P. K.; Carbonell, R. G. J. Colloid Interface Sci. 1990, 137, 462–482. (7) Wierenga, P. A.; Egmond, M. R.; Voragen, A. G. J.; de Jongh, H. H. J. Colloid Interface Sci. 2006, 299, 850–857. (8) Kim, G.; Gurau, M.; Kim, J.; Cremer, P. S. Langmuir 2002, 18, 2807– 2811. (9) Lad, M. D.; Birembaut, F.; Matthew, J. M.; Frazier, R. A.; Green, R. J. Phys. Chem. Chem. Phys. 2006, 8, 2179–2186. (10) Daillant, J.; Gibaud, A. X-Ray and Neutron ReflectiVity: Principles and Applications; Springer: Berlin, 1999. (11) Tolan, M. X-Ray Scattering from Soft Matter Thin Films; Springer: Berlin, 1999.

measurement typically requires at least 30 min even using a synchrotron source, this technique can only be used for samples that are sufficiently stable. Here, we investigate the adsorption process of a globular protein lysozyme (LSZ) on an air/water interface using time-resolved measurements at a time resolution of 1 min performed on the recently developed X-ray liquid interface reflectometer at SPring-8.12 LSZ has an elliptical shape with approximate dimensions of 30 × 30 × 45 Å3, and is regarded as a rigid molecule because of the presence of four disulphide bridges. In the present study, LSZ was chosen because its three-dimensional conformation is very stable in solution, and its adsorption behavior has already been studied widely.2-6,8,9,13-15 However, the adsorption of LSZ on an air/water interface has been subject to debate, with conflicting interpretations on the structure of the LSZ molecules.6,9,13,14 On the basis of measurements of the surface excess and thickness in the equilibrium state, it was concluded that LSZ molecules at an air/water interface remain in their native structure, but the molecular orientation switches from side-on to end-on with increasing bulk concentration.6,13 In contrast, surfacesensitive IR spectroscopy measurements have revealed that some R-helices change to β-sheets upon adsorption.9,14 Here, we show that the LSZ molecules at the air/water interface have a very flat shape soon after adsorption and gradually change their conformation as a result of interaction with other LSZ molecules.

Experimental Methods Materials. Three times crystallized and lyophilized hen egg LSZ was purchased from Sigma (prod. no. L6876) and used as supplied. Protein solutions were made using a phosphate buffer solution (0.02 M KH2PO4/K2HPO4) of pH 7 (ionic strength 0.035 M) using UHQ(12) Yano, Y. F.; Uruga, T.; Tanida, H.; Toyokawa, H.; Terada, Y.; Takagaki, M. J. Physics: Conf. Ser. 2007, 83, 012024. (13) Lu, J. R.; Su, T. J.; Thomas, R. K.; Penfold, J.; Webster, J. J. Chem. Soc. Faraday Trans. 1998, 94, 3279–3287. (14) Postel, C.; Abillon, O.; Desbat, B. J. Colloid Interface Sci. 2003, 266, 74–81. (15) Erickson, J. S.; Sundaram, S.; Stebe, K. J. Langmuir 2000, 16, 5072– 5078.

10.1021/la803235x CCC: $40.75  2009 American Chemical Society Published on Web 12/11/2008

Letters

Langmuir, Vol. 25, No. 1, 2009 33 Table 1. Structural Parameters Obtained by Refining the Reflectivity Shown in Figure 1

Figure 1. (a) X-ray reflectivity profiles measured 1 min after LSZ injection. The data were divided by the Fresnel reflectivity of the air/ buffer interface. The continuous lines are the fit to the data. (b) Electron density profiles corresponding the fits to the data.

grade water. Protein solutions were made to concentrations of 43 and 0.43 mg/mL, from which 1 cm3 portions were added to a 42 cm3 buffer subphase in a Langmuir trough to give final concentrations of 1 and 0.01 mg/mL. Methods. The X-ray reflectivity measurements were performed using the recently developed liquid-interface reflectometer at the undulator beamline BL37XU of SPring-8.12 The reflectometer, equipped with the two-dimensional pixel array detector PILATUS,16 is able to achieve X-ray reflectivity values toward 10-8 in only 150 s with an integration time at each angle of 1 s. The time resolution of the reflectometer was limited by the time needed for position adjustment of the seven motorized stages at each reflection angle. The measurements were made at angles of incidence in the range of 0.005-2.0° with an X-ray energy of 15 keV, which corresponds to qz of 0.001-0.5 Å-1. In the experiment, a LSZ solution was injected using a liquid dispenser (Ultra Dispensing Stations 30 PSI, San-Ei Tech) into a phosphate buffer solution of pH 7 contained in a Langmuir trough. The Langmuir trough, covered with an acrylic hood to suppress liquid evaporation, was mounted on a heat sink, and the temperature was controlled at 25 ( 1 °C. Measurements were started 4 s after injection of the sample. The position of the X-ray on the sample was changed horizontally at intervals of 1 s to avoid radiation damage. The surface pressure was monitored throughout using a Wilhelmy plate made of glass mounted on the Langmuir trough. Data Analysis. The X-ray reflectivity data were fitted using a three-slab model with the Parrat32 software17 taking parameters of the film thickness, the electron densities, and roughness of the three slabs. A surface excess Γ of LSZ was calculated by

Γ)

m

∫ [F(z) - Fbuffer]dz Ne

with the obtained electron density profile F(z), the electron density of the buffer solution Fbuffer of 0.334 e/ Å,3 the molecular mass and the number of electrons of an LSZ molecule, m ) 2.38 × 10-17 mg and Ne ) 7626 e-, respectively. This equation gives an underestimated value of Γ, since it assumes the adsorbed layer consists of the buffer solution. From the previous work by Lu et al., however, a part of the LSZ molecule protrudes out of the buffer solution.13

Results and Discussion Figure 1 shows X-ray reflectivity profiles measured 1 min (at the midpoint of the scan) after LSZ injection, along with electron density profiles obtained by fitting the data. All fitting parameters are listed in Table 1. The adsorbed LSZ has a dense upper slab (16) Broennimann, C.; Eikenberry, E. F.; Henrich, B.; Horisberger, R.; Huelsen, G.; Pohl, E.; Schmitt, B.; Schulze-Briese, C.; Suzuki, M.; Tomizaki, T.; Toyokawa, H.; Wagner, A. J. Synchrotron Rad. 2006, 13, 120–130. (17) http://www.hmi.de/bensc/instrumentation/instrumente/v6/refl/parratt_en.htm.

slab number

slab thickness [Å]

electron density [e-/ Å3]

interface roughness [Å]

1.0 mg/mL 1 2 3 bulk

6.0 10.7 11.2

0.52 0.39 0.345 0.334a

3.0 2.9 3.0 2.8

0.01 mg/mL 1 2 3 bulk

4.1 8.9 1.8

0.56 0.37 0.35 0.334a

3.0 3.0 3.0 3.2

a

Fixed parameters.

Figure 2. Time dependence of R/RFbuffer vs qz X-ray reflectivity profiles with bulk concentrations of (a) 1 mg/mL and (b) 0.01 mg/mL. The solid lines are the fits to the data. The curves are displaced vertically for clarity. The adsorption times are shown in minutes.

10 Å thick and a diffuse lower slab 10 Å thick, suggesting that the protein is highly denatured in comparison to the globular structure that is observed in the bulk. Furthermore, the observed thickness was less for the 0.01 mg/mL sample than for the 1 mg/mL sample. The time dependence of the X-ray reflectivity profiles is shown in Figure 2. The adsorption of LSZ led to a slight shift in the broad maximum to the lower qz for the 0.01 mg/mL sample. This shift can be explained by increasing the total thickness of the adsorbed LSZ layer. The profile at adsorption time of 175 min coincides with that observed at 3 min for the 1 mg/mL sample, suggesting that the early stage of the adsorption process for the higher concentration slows down in the case of lower concentration. For the higher concentration, an additional peak appeared in qz < 0.1 Å-1 after 6 min. The whole adsorption process is easier to understand from the electron density profiles as shown in Figure 3. For a LSZ concentration of 0.01 mg/mL, the density of all slabs increases during adsorption, while for a concentration of 1 mg/mL, the density and thickness of the third slab drastically increases. The time dependence of the surface pressure and surface excess calculated by integration of the electron density profile are shown in Figure 4. For the 1 mg/mL sample, the changes in surface pressure and surface excess do not exactly correspond: a continuous increase is observed in the surface pressure, while two plateaus are observed in the surface excess. The first plateau in surface excess occurs at a value about half-that of the second plateau, indicating that a monolayer formed at the first plateau, and a second layer formed after that. In contrast, for the 0.01 mg/mL sample, the two properties for 0.01 mg/mL are more

34 Langmuir, Vol. 25, No. 1, 2009

Figure 3. Time dependence of electron density profiles with bulk concentrations of (a) 1 mg/mL and (b) 0.01 mg/mL. The curves are the fits to the data shown in Figure 2.

Figure 4. (a) Variation of surface pressure. (b) Variation of surface excess calculated by integration of the electron density profiles shown in Figure 3.

consistent: both sets of data remain at almost zero for the first 100 min. This duration, the so-called induction time, is usually observed in a dilute solution with a rigid protein.4 During this period, LSZ molecules are found to form a surface gaseous phase, which can be observed by fluorescence microscopy.15 The surface excess for the 0.01 mg/mL sample is less than the value of the first plateau for the 1 mg/mL sample, suggesting that monolayer formation is not complete after 175 min. From these results, we can build up a picture of the adsorption process of LSZ molecules at the air/water interface, which is considered to be a four-step process as follows: 1 The LSZ molecules initially adsorb at the air/water interface as a monomolecular structure, with each molecule adopting a flat unfolded conformation that does not change up to full monolayer surface coverage. Such a flat structure supports the result that some R-helices change to antiparallel β-sheets upon adsorption, as observed in previous work.9 The surface pressure does not change during this period (induction time). When the concentration of the solution becomes higher, the

Letters

Figure 5. Schematic diagram of the LSZ adsorption process. (1) LSZ molecules initially adsorb as a monomolecular structure with a flat unfolded conformation, forming antiparallel β-sheets as a result of hydrophobic interactions with the gas phase. (2) LSZ molecules form a monolayer and begin to protrude their hydrophilic groups toward the water phase. (3) A second layer forms in which the molecules have a very loose structure having random coils as a result of hydrophilic interactions with the hydrophilic groups that protrude from the first layer. (4) LSZ molecules form a multilayer in which the molecules might have a very loose structure having random coils.

adsorption rate increases, and hence the induction time becomes significantly shorter. 2 Near the full monolayer surface coverage, the LSZ molecules at the surface interact with each other, and hence begin to protrude their hydrophilic groups toward the water phase. As a result, the surface pressure starts to increase. 3 The LSZ molecules then start to adsorb underneath the monolayer to form a second layer. The conformation of the LSZ in the second layer is more loosely packed than that in the first layer. A random coil structure is highly possible taking into account previous work that shows a gradual shift in the amide I band away from the strongly antiparallel β-sheet structure of the initially formed protein layer.9 4 The conformation of the LSZ molecules in the second layer becomes more loosely packed as a multilayer forms. The surface pressure gradually increases up to an equilibrium value of 18 mN/m for the 1 mg/mL sample.9 A schematic diagram of the four-step process is shown in Figure 5.

Conclusions We investigated the adsorption kinetics of LSZ at an air/water interface to study the mechanism of adsorption-induced protein unfolding. The LSZ molecules that initially adsorb at the air/ water interface have an unfolded structure, such as a flat structure with antiparallel β-sheets. Interaction with other molecules upon further adsorption leads to changes in the conformation. Considering that LSZ molecules unfold at solid/water interfaces that are hydrophobic,18 and do not unfold on a hydrophilic interface,19 the crucial driving force for unfolding at the air/ water interface is thought to be hydrophobic interactions with the gas phase. In the second layer, the molecular conformation has a very loose structure having random coils, which results from hydrophilic interactions with the hydrophilic groups that protrude from the first layer. The present results, obtained using time-resolved X-ray reflectivity, offer valuable insight into the (18) Lu, J. R.; Su, T. J.; Thirtle, P. N.; Thomas, R. K.; Rennie, A. R.; Cubitt, R. J. Colloid Interface Sci. 1998, 206, 212–223.

Letters

mechanism of LSZ adsorption at an interface, and the mechanism we consider here is consistent with previous results. Acknowledgment. The synchrotron radiation experiments were performed at the BL37XU in the SPring-8 facility with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (Proposal No. 2007A1197; 2007B1091).

Langmuir, Vol. 25, No. 1, 2009 35

Supporting Information Available: Computational details and all fitting parameters. This material is available free of charge via the Internet at http://pubs.acs.org. LA803235X (19) Evers, F.; Shokuie, K.; Paulus, M.; Sternemann, C.; Czeslik, C.; Tolan, M. Langmuir 2008, 24, 10216–10221.