pH-Dependent Behavior of Surface-immobilized Artificial Leucine

pH-Dependent Behavior of Surface-immobilized Artificial. Leucine Zipper Proteins. Molly M. Stevens,†,‡ Stephanie Allen,*,† Jill K. Sakata,§ Mar...
0 downloads 0 Views 83KB Size
Langmuir 2004, 20, 7747-7752

7747

pH-Dependent Behavior of Surface-immobilized Artificial Leucine Zipper Proteins Molly M. Stevens,†,‡ Stephanie Allen,*,† Jill K. Sakata,§ Martyn C. Davies,† Clive J. Roberts,† Saul J. B. Tendler,† David A. Tirrell,| and Philip M. Williams† Laboratory of Biophysics and Surface Analysis, School of Pharmacy, The University of Nottingham, Nottingham, NG7 2RD, United Kingdom, Department of Polymer Science and Engineering, University of Massachusetts, Amherst, Massachusetts 01003, and Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125 Received December 29, 2003. In Final Form: June 10, 2004 The coiled-coil protein motif occurs in over 200 proteins and has generated interest for a range of applications requiring surface immobilization of the constituent peptides. This paper describes an investigation of the environment-responsive behavior of a monolayer of surface-immobilized artificial proteins, which are known to assemble to form coiled-coil structures in bulk solution. An extended version of the quartz crystal microbalance (QCM-D) and surface plasmon resonance (SPR) are independently employed to characterize the adsorption of the proteins to a gold surface. The data suggest that the molecules arrange in a closely packed layer orientated perpendicular to the surface. QCM-D measurements are also employed to measure pH-induced changes in the resonant frequency (f) and the energy dissipation factor (D) of a gold-coated quartz crystal functionalized with the formed monolayer. Exposure of the protein monolayer to a pH 4.5 solution results in a shift of 43 Hz in f and a shift of -0.7 × 10-6 in D as compared to pH 7.4. In contrast, increasing the pH to 11.2, results in f and D shifts of -17 Hz and 0.6 × 10-6, respectively. The magnitude of the observed shifts suggests that the proteins form a rigid layer at low pH that can be hydrated to a fluid layer as the pH is increased. These observations correlate with spectroscopic changes that indicate a reduction in the helical content of the protein in bulk solutions of high pH.

Introduction The coiled-coil protein motif is found in a diverse group of over 200 proteins,1 and de novo designed peptides that adopt this structure have become established as model systems to use in understanding the fundamentals of protein folding and stability.2,3 Furthermore, this “universal” dimerization domain has generated interest in a plethora of applications where there is a requirement for the constituent peptides to be immobilized at a surface, including heterodimer technology for the detection and purification of recombinant peptides and proteins;4 coiledcoils as templates for combinatorial helical libraries for drug discovery and basic research;5 and as a dimerization domain for biosensors.6 The studies presented here concern the investigation of a synthetic protein, termed Acys, the sequence of which is displayed in Figure 1. The protein was engineered7 with a terminal cysteine residue, which facilitates its covalent * To whom correspondence should be addressed (e-mail: [email protected]). † The University of Nottingham. ‡ Currently at Imperial College London. § University of Massachusetts. | California Institute of Technology. (1) Lupas, A.; Van Kyke, M.; Stock, J. Science 1991, 252, 11621164. (2) Adamson, J. G.; Zhou, N. E.; Hodges, R. S. Curr. Opin. Biotechnol. 1993, 4, 428-437. (3) Hodges, R. S. Biochem. Cell Biol. 1996, 74, 133-154. (4) Tripet, B.; Yu, L.; Bautista, D. L.; Wong, W. Y.; Irvin, R. T.; Hodges, R. S. Protein Eng. 1996, 9, 1029-1042. (5) Houston, M. E., Jr.; Wallace, A.; Bianchi, E.; Pessi, A.; Hodges, R. S. J. Mol. Bio. 1996, 262, 270-282. (6) Chao, H.; Houston, M. E., Jr.; Grothe, S.; Kay, C. M.; O’ConnorMcCourt, M.; Irvin, R. T.; Hodges, R. S. Biochemistry 1996, 35, 1217512185. (7) Petka, W. A.; Hardin, J. L.; McGrath, K. P.; Wirtz, D.; Tirrell, D. A. Science 1998, 281, 389-392.

Figure 1. Amino acid sequence of Acys. The protein consists of 76 amino acids, 42 of which are of the leucine zipper, Helix, which is shown in the helical wheel representation as viewed from the NH2-terminus. Charge patterns in the e and g positions of the Helix heptad repeat, (a-b-c-d-e-f-g)6 are underlined.

immobilization onto gold substrates via the formation of a gold-thiolate bond. Acys consists of 76 amino acids of which 42 residues form putative leucine zipper domains (Helix), which can associate with the leucine zipper domain of another Acys to form a dimeric coiled-coil structure as illustrated by the helical wheel diagram in Figure 1. The rationale for the design of the Acys protein has been described previously.7 The Helix region comprises six of the heptad repeats (a-b-c-d-e-f-g), which characterize the sequence of coiled-coil proteins,3 and is separated from the surface by a 15-amino-acid tether. Although hydrophobic packing at the dimer interface is the dominant force of stabilization in coiled-coils, the stability can also be modulated by electrostatic interactions.3 These electrostatic effects originate from the close interhelical approach of the side

10.1021/la030440e CCC: $27.50 © 2004 American Chemical Society Published on Web 08/06/2004

7748

Langmuir, Vol. 20, No. 18, 2004

chains of residues in the e and g positions, of which 9 out of the 12 in Acys are occupied by glutamic acid residues (Figure 1). This promotes the destabilization of homodimeric Acys coiled-coils in basic solutions through glutamate to glutamate ionic interactions across the dimer interface as confirmed by previous thermal denaturation studies of the protein in bulk solution.7 The quartz crystal microbalance (QCM) is well established as a sensitive mass detection technique, in which changes in the resonant frequency (f) of a quartz crystal may be monitored as nanograms of material bind to it.8,9 A useful extension of the QCM, termed “QCM-D”, is the ability to simultaneously measure changes in the damping of the crystal oscillation in liquid.10 The measured energy dissipation factor (D), can be related to changes in the viscoelastic properties of surface-immobilized molecules.11 This additional facility has led to various applications for QCM-D, including monitoring of protein12,13 and lipid vesicle adsorption,14 interactions of specific antibodies and antigens,15 interactions of peptides and DNA,16 monitoring of cell adhesion and spreading,17-19 and investigation of changes in the rheological properties of polymer films.20 This study uses QCM-D and the complementary technique of surface plasmon resonance (SPR) to characterize the adsorption of protein Acys to a gold surface, and to reveal details of the likely packing arrangement of the proteins at this surface. QCM-D is then employed to investigate the response of the surface-immobilized proteins to changes in the environmental pH. Experimental Section Materials. Acys was prepared by bacterial expression of the corresponding artificial gene followed by purification by metal affinity column chromatography.7 Unless otherwise indicated, materials were obtained from Sigma-Aldrich Co. Ltd (Gillingham, Dorset, United Kingdom) and were of analytical grade quality. General Methods. For all measurements, the pH-adjusted solutions were 10 mM sodium phosphate and 150 mM NaCl which were adjusted to pH 4.5, 7.4, and 11.2 with 1 M HCl or 1 M NaOH unless specified. All solutions were prepared using high-purity deionized water (purified on an ELGA water system, Maxima HPLC, Elga Ltd, Bucks., United Kingdom, resistivity approximately 15 MΩ cm), and were degassed and filtered using an 0.2 µm filter prior to use (Sartorius A. G., Go¨ttingen, Germany). CD Spectroscopy. CD spectra were recorded on an Aviv 60DS spectropolarimeter (Lakewood, NJ). Acys was dissolved in phosphate-buffered saline (10 mM sodium phosphate, 150 mM NaCl, pH adjusted with 0.5 N HCl or NaOH) to a concentration of 15 µM. Protein concentrations were determined by Trp absorbance in 6 M guanidinium hydrochloride at 25 °C using an extinction coefficient at 280 nm of 5690 M-1 cm-1.21 The protein (8) Sauerbrey, G. Z. Phys. 1959, 155, 206-222. (9) O’Sullivan, C. K.; Guilbault, G. G. Biosens. Bioelectron. 1999, 14, 663-670. (10) Marx, K. A. Biomacromolecules 2003; 4, 1099-1120. (11) Rodahl, M.; Hook, F.; Krozer, A.; Brzezinski, P.; Kasemo, B. Rev. Sci. Instrum. 1995, 66, 3924-3930. (12) Rodahl, M.; Hook, F.; Fredriksson, C.; Keller, C. A.; Krozer, A.; Brzezinski, P.; Voinova, M.; Kasemo, B. Faraday Discuss. 1997, 229246. (13) Hook, F.; Rodahl, M.; Kasemo, B.; Brzezinski, P. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 12271-12276. (14) Keller, C. A.; Kasemo, B. Biophys. J. 1998, 75, 1397-1402. (15) Hook, F.; Rodahl, M.; Brzezinski, P.; Kasemo, B. Langmuir 1998, 14, 729-734. (16) Hook, F.; Ray, A.; Norden, B.; Kasemo, B. Langmuir 2001, 17, 8305-8312. (17) Fredriksson, C.; Khilman, S.; Kasemo, B.; Steel, D. M. J. Mater. Sci.: Mater. Med. 1998, 9, 785-788. (18) Fredriksson, C.; Kihlman, S.; Rodahl, M.; Kasemo, B. Langmuir 1998, 14, 248-251. (19) Nimeri, G.; Fredriksson, C.; Elwing, H.; Liu, L.; Rodahl, M.; Kasemo, B. Colloids Surf., B 1998, 11, 255-264. (20) Kim, J.-M.; Chang, S.-M.; Muramatsu, H. Polymer 1999, 40, 3291-3299.

Stevens et al.

Figure 2. Circular dichroism for 15 µM Acys in PBS (10 mM phosphate, 150 mM NaCl), pH adjusted to 4.5(O), 7.4 ([), and 11.2 (0). solution was analyzed in a quartz cuvette (1 mm path length) from 200 to 250 nm in 1 nm steps with 1 nm bandwidth at 25 °C, where each spectrum was the average of three scans. QCM-D Measurements. The QCM-D measurements were performed using a Q-Sense D300 measurement system (Q-Sense AB, Gothenburg, Sweden). The QCM-D sensor crystal consisted of a 14 mm diameter, 5 MHz AT-cut quartz crystal that was gold coated. In liquid environments, the limit for the mass sensitivity was on the order of 5 ng.cm-2, and the dissipation factor (D) was approximately 3 × 10-7 for the unloaded 5 MHz crystal. Changes in the resonant frequency (∆f) and the dissipation factor (∆D) of the oscillator were measured simultaneously at the fundamental resonant frequency (5 MHz) and at the third overtone (15 MHz). The shifts from the measurements performed at the 15 MHz third overtone are presented in this study due to the increased sensitivity of the signal at this frequency. A frequency shift of 1 Hz corresponds to 5.9 ng.cm-2 when the third overtone of 15 Hz is used, compared to 17.7 ng.cm-2 when the fundamental frequency of 5 MHz is employed22 (see later discussion on the limitations of this mass-response model). All samples were introduced into an axial flow chamber (QAFC 301) that comprised a T-loop to thermally equilibrate the sample (0.5-0.6 mL) at 23 ( 0.1 °C for 2 min before it was introduced into the measurement chamber. This resulted in small pressure changes observable in the D and f traces as the samples were introduced into the T-loop. A steady baseline was acquired prior to starting all measurements. SPR Measurements. The SPR measurements were performed using a BIAcore 3000 biosensor system (Pharmacia Biosensor AB, Uppsala, Sweden). The SPR slides were gold coated (Pioneer Chip J1, Pharmacia Biosensor AB). The mass uptake of proteins at the surface was determined in realtime by a measured change in response units (∆RU). This is a dimensionless quantity that is proportional to the change in refractive index at the interfacial region.23 A shift of 1 RU corresponds to 77 ng.cm-2 of protein mass adsorbed, as calibrated for the Pioneer Chip J1.24 In the detailed studies, the temperature was set to 25 °C, and the experimental system was left to equilibrate for a few minutes to avoid any drift due to thermal effects. The running buffer for the immobilization process was pH 7.4 HBS buffer (containing 0.1 M HEPES and 0.15 M NaCl, BIAcore) which flowed at a rate of 2 µL/min. A steady baseline was always obtained prior to starting the SPR measurements.

Results and Discussion Adsorption of Acys to a Gold Surface. The secondary structure of Acys in bulk solution at pH 4.5, 7.4, and 11.2 was characterized using circular dichroism (CD). Figure 2 shows the stepwise decrease in molar ellipticity (increase in [θ]222), from pH 4.5 to 7.4 to 11.2, indicating that the helicity of the Acys coiled-coil decreases with increasing pH. (21) Edelhoch, H. Biochemistry 1967, 6, 1948-1954. (22) Personal communication with Q-sense AB, Gothenburg, Sweden. (23) Lofas, S.; Malmqvist, M.; Roennberg, I.; Stenberg, E.; Liedberg, B.; Lundstroem, I. Sens. Actuators, B 1991, B5, 79-84. (24) Personal communication with Pharmacia Biosensor AB, Gothenburg, Sweden.

Artificial Leucine Zipper Proteins

Langmuir, Vol. 20, No. 18, 2004 7749

provided that the mass is small compared to the mass of the crystal, the mass is evenly distributed, the mass does not slip on the electrode, and the mass is sufficiently rigid and/or thin to have negligible internal friction. In this case the Sauerbrey relation8 in eq 1 holds

∆f ) -n∆m/C

Figure 3. QCM-D measurement of the frequency, ∆f (bottom trace), and dissipation shifts, ∆D (top trace) observed upon formation of a monolayer of Acys on a gold-coated quartz sensor at pH 7.4. The dotted lines indicate the solution exchange from the Acys solution at pH 7.4 to a 10 mM NaOH solution, which was used to remove noncovalently bound protein from the surface.

Figure 3 displays the changes in frequency and dissipation signals observed during the adsorption of protein Acys onto the gold-coated sensor surface. Prior to introducing protein Acys into the measurement chamber, a steady baseline was acquired with pH 7.4 solution. At t ) 1 min, a solution of Acys at pH 7.4 was introduced into the measurement chamber. Following this time, an initial rapid frequency decrease (∆f ≈ -145 Hz), corresponding to a mass increase at the surface due to protein binding, can be observed in Figure 3. For the same time period a small increase in the dissipation signal (∆D ) 0.8 × 10-6), can also be observed. A second sample of Acys (in pH 7.4 buffer) was introduced into the measurement chamber at t ) 5 min, with the aim of attaining maximum surface coverage. No marked shifts in the frequency or dissipation signals accompanied the second injection, and the plateau in both traces before t ) 11 min indicates that the process of protein binding to the gold sensor surface had reached saturation. Exposure of the surface to protein was then interrupted by exchanging the solution in the measurement chamber for a 10 mM NaOH solution at t ) 11 min (Figure 3). At this point, both traces moved out of the range of the displayed scale. The NaOH solution was introduced to remove any noncovalently bound protein by promoting dissociation of any formed coiled-coil structures that may be loosely bound and/or participating in multilayer formation. At such high pH values, dissociation of the coiled-coils is promoted through increased interhelical repulsive interactions due the presence of ionized glutamic acid residues, consistent with the decrease in helicity at high pH as shown in Figure 2. A pH 7.4 solution was then reintroduced into the measurement chamber at t ) 18 min and t ) 21 min to return the sample to the original pH value. The corresponding f and D-shifts suggest that NaOH treatment resulted in loss of only a small amount of protein. Additional NaOH washes (data not shown) resulted in no further changes in the frequency or dissipation signals on returning to pH 7.4. From the data presented in Figure 3, a total ∆f of -120 Hz can be observed for the whole experiment. ∆f can be converted to a change in mass (∆m) due to bound protein,

(1)

where C is the mass sensitivity constant (C ) 17.7 ng.cm-2.Hz-1 for a crystal with a fundamental resonance frequency of 5 MHz), and n is the overtone number. This mass-response model has been successfully applied to characterize the adsorption from the liquid phase of rigid, evenly distributed thin layers of alkanethiols25 and lipid mono- or bilayers.26 It is however generally accepted that when this relation is applied to describe protein adsorption phenomena in liquids, significantly higher mass-uptake estimations are observed than for techniques such as SPR, ellipsometry, and optical waveguide mode spectroscopy.12,16,27,28 This discrepancy arises as the latter techniques are unable to detect water within and associated with the protein layer, and thus provide an estimate of its molar (“dry”) mass. In contrast, the f-shift in QCM is due to the change in the total coupled mass, including hydrodynamically coupled water and water associated with the hydration layer of proteins or trapped within the protein layer.27,29 It is therefore likely that water associated with or trapped within the immobilized layer of protein Acys will be a significant factor affecting the mass determination.12,16 The f-shift of -120 Hz due to binding of Acys to the gold substrate can be converted to a gain in mass of 708 ng.cm-2 using eq 1. To estimate the surface coverage of Acys, we note that it has a molecular mass of approximately 8.5 kDa and we assume that approximately 20% of the measured mass is due to H2O associated with the protein. Avogadro’s constant can then be used to calculate the surface density of adsorbed Acys dimers as 1.95 × 1013 per cm2 (which corresponds to an approximate area per dimer of 5 nm2). Using such figures, we are able to estimate the likely arrangement of Acys molecules on the surface. Upper and lower limits (depending on the orientation of the molecule) for the number of molecules present in a tightly packed monolayer of Acys can be determined using the dimensions of the Acys dimer, estimated using the crystal structure of the peptide GCN4-p130 which is closely related in sequence and structure. A cross-sectional area of 7.1 nm2 and a length of 10.4 nm were used here for calculating the projected area of an Acys dimer with the helical axis oriented either perpendicular or parallel to the surface. For a layer comprising molecules in an upright orientation, assuming hexagonal close packing, we calculate that a ∆f of -120 Hz corresponds to a mass nearly one and a half times larger than a theoretical monolayer (100% surface coverage). In contrast, to achieve such a frequency change with molecules orientated approximately parallel to the gold surface, the formed layer would have to (25) Schneider, T. W.; Buttry, D. A. J. Am. Chem. Soc. 1993, 115, 12391-12397. (26) Keller, C. A.; Glasmastar, K.; Zhdanov, V. P.; Kasemo, B. Phys. Rev. Lett. 2000, 84, 5443-5446. (27) Hook, F.; Voros, J.; Rodahl, M.; Kurrat, R.; Boni, P.; Ramsden, J. J.; Textor, M.; Spencer, N. D.; Tengvall, P.; Gold, J.; Kasemo, B. Colloids Surf., B 2002, 24, 155-170. (28) Hook, F.; Kasemo, B.; Nylander, T.; Fant, C.; Sott, K.; Elwing, H. Anal. Chem. 2001, 73, 5796-5804. (29) Caruso, F.; Furlong, D. N.; Kingshott, P. J. Colloid Interface Sci. 1997, 186, 129-140. (30) O’Shea, E. K.; Klemm, J. D.; Kim, P. S.; Alber, T. Science 1991, 254, 539-544.

7750

Langmuir, Vol. 20, No. 18, 2004

Figure 4. Example SPR trace of the shift in response units (∆RU) recorded as a function of time for the adsorption of Acys onto a gold surface. The points at which solutions of 10 µM Acys (points A, C, and E), HBS buffer (points B, D, and F) and 10 mM NaOH (point G) were injected into the SPR chamber are indicated.

comprise 16 molecular layers. However, the presence of a multilayer of Acys is unlikely following exposure of the surface to NaOH. It should also be noted that the total dissipation shift of 0.5 × 10-6 is relatively small and indicative of a tightly packed arrangement of molecules (D is a dimensionless quantity; changes in D are qualitatively assessed).13 For further interpretation of these results the reader is reminded of the assumptions made before using the Sauerbrey equation. For example, it is known that slip of the film at the surface, internal dissipative effects, and entrapped water can all influence mass determination by QCM.12-15 However, previous QCM-D studies12,15 have also indicated that for thin (