Biomacromolecules 2005, 6, 3334-3344
3334
Calcium-Induced Changes to the Molecular Conformation and Aggregate Structure of β-Casein at the Air-Water Interface Christina R. Vessely,† John F. Carpenter,† and Daniel K. Schwartz*,‡ Department of Chemical and Biological Engineering, University of Colorado, Boulder, Colorado 80309, and Department of Pharmaceutical Sciences, University of Colorado Health Sciences Center, Denver, Colorado 80262 Received May 24, 2005; Revised Manuscript Received July 19, 2005
The influence of calcium on interactions of β-casein at the air-water interface has been studied by several techniques, including interfacial rheology, atomic force microscopy (AFM), infrared reflectance-absorbance spectroscopy (IRRAS), and ζ potential measurements. In the absence of calcium, a weak interfacial gel forms after about 2.5 h. Also in the absence of calcium, the adsorbed β-casein film exhibits some degree of both intra- and intermolecular structural organization. For example, IRRAS spectra show a measurable amount of R-helix content, and AFM images indicate the presence of interfacial aggregates with a characteristic lateral length scale of 20-30 nm, which we interpret as hemimicelles. Upon the addition of calcium, particularly at Ca:β-casein molar ratios above ∼5:1, a stronger interfacial gel forms more quickly; for example, the interfacial shear moduli increase twice as rapidly. Also under these conditions (5:1 Ca:β-casein ratio) there is little evidence of structural organization; i.e., the R-helix peaks are very weak, and AFM images show a disordered, but continuous film, without distinct hemimicelles. On the basis of these findings, we hypothesize that calcium binding destabilizes the coupled intra- and intermolecular structural organization, and that the loss of organization permits more rapid interfacial gelation. These phenomena are characteristic of the air-water interface; they are not accompanied by analogous structural changes in bulk solution. Introduction Proteins, because of their amphiphilic nature, adsorb readily at interfaces, especially those between hydrophobic and hydrophilic phases.1-3 The adsorption of proteins at the air-water interface can often result in significant changes to the protein structure and/or conformation.4-6 These conformational changes can enable intermolecular interactions including aggregation and the formation of an interfacial gel. These processes may be favorable, e.g., leading to greater stability of colloids, or unfavorable, leading to loss of activity in pharmaceutical proteins. In either case, a mechanistic understanding of adsorption-driven processes is clearly desirable. Rheology measurements are widely used to characterize viscoelastic bulk materials such as polymers and gels. By analogy, interfacial rheology can be used to study the mechanical properties of quasi-two-dimensional surface layers. While interfacial dilatational rheology has been used extensively to study the interactions of proteins at interfaces,7,8 interfacial shear rheology measurements are not as common. However, shear rheology measurements are particularly useful for the characterization of two-dimensional gels at the air-water interface;9-11 in fact, a comparison of the interfacial shear storage and loss moduli provides the characteristic signature of the gelation transition.9 * To whom correspondence should be addressed.
[email protected]. † University of Colorado Health Sciences Center. ‡ University of Colorado.
E-mail:
This study investigates the interactions of calcium with β-casein, and the influence on gel formation at the air-water interface. β-Casein is a 24 kDa protein which is considered to have little native secondary structure, although there is disagreement in the literature with respect to the exact amount of secondary structure present.12,13 β-Casein is known to interact readily at the air-water interface, and has previously been reported to form gel-like layers at the interface despite the absence of cysteine residues for the formation of disulfide cross-links between individual molecules.9 The majority of hydrophilic amino acid residues occur within the first 50 residues, largely contributing to the net charge of the protein of -12 at neutral pH.14 The remainder of the protein is predominantly hydrophobic in nature15 and proline rich. Because of the separation between hydrophilic and hydrophobic regions, the protein is often referred to as a natural diblock copolymer.16 The surfactantlike structure of the molecule makes it prone to the formation of certain types of aggregates in solution, such as micelles. The critical micelle concentration for β-casein has previously been reported as 500 µg/mL at 40 °C.17-19 The primary biological function of the casein molecules is the transport of calcium and phosphate in an easily absorbable form from mother to young through the mother’s milk. Each molecule has a number of potential binding sites for Ca2+, including ionic interactions at the negatively charged amino acid residues (of which there are approximately 22 per molecule) and more specific binding at the five phosphorylated serine residues.14,20-22 In milk
10.1021/bm050353w CCC: $30.25 © 2005 American Chemical Society Published on Web 08/25/2005
Ca2+ Effects on β-Casein at the Air-Water Interface
solutions, the casein molecules exist in a complex conformation consisting of soaplike micelles that form large conglomerate structures. Individual micelles include about 30 β-casein monomers and are on the order of 10-20 nm in diameter. These individual micelles aggregate to form large conglomerates that generally average about 150 nm in diameter, although sizes as large as 600 nm have been observed.14 These conglomerates are unfortunately known as “casein micelles” in the food science literature, causing confusion with the more traditional colloid terminology. However, in this paper we will use the term “micelle” to refer to the soaplike micelles. The formation of the large conglomerates may occur through interactions with calcium phosphate nanoclusters, with the phosphorylated serine residues grouped around the nanoclusters while the hydrophobic blocks of the molecules associate with one another to form larger structures.23,24 Despite disagreement with respect to the specific structure of the conglomerates, calcium-induced aggregation of β-casein has been shown repeatedly in the literature.20,25-27 The goal of this study was to evaluate the aggregation behavior of β-casein at the air-water interface through the use of rheological studies and atomic force microscopy (AFM). β-Casein concentrations were held constant for the experiments to focus the investigations on the effects of both calcium and aging time. The β-casein concentration chosen was ∼50 µg/mL, which is an order of magnitude below the critical micelle concentration for β-casein. The data obtained using these analytical methods led to further investigation of β-casein at the molecular level through the use of spectroscopic techniques. Experimental Section Materials. The β-casein (C-6905, minimum 90% purity by electrophoresis, essentially salt free) used in all of the experiments was obtained from Sigma-Aldrich. The certificate of analysis for the material (lot no. 082K7405) showed the actual purity to be >99%. Surfactin (approximately 98%, Bacillus subtilis) and poly-γ-benzyl-L-glutamate (molecular weight 30000-70000) used in infrared reflectance-absorbance spectroscopy (IRRAS) studies were obtained from Sigma-Aldrich. Sodium azide (NaN3), paraffin, hexane (C6H14), hydrochloric acid (HCl; 37%), and deuterium oxide (D2O; 100 atom % D) were also obtained from SigmaAldrich. Sodium citrate (C6H5O7Na3‚2H2O), sulfuric acid (H2SO4; ACS reagent), hydrogen peroxide (H2O2; 30%), hydrofluoric acid (HF; 47-51%), sodium chloride (NaCl; USP grade), calcium chloride (CaCl2; >99%), and chloroform (CHCl3; g99.9%) were all purchased from Fisher Scientific. Microscopic single-well culture slides were also purchased from Fisher Scientific. Silicon wafers (Si(100) type P/boron/ Cz resistivity 0.01-0.04 Ω cm) were purchased from Electrooptic Materials. The silica substrates were cut into small pieces (∼1 × 1 cm), cleaned with ethanol, and then boiled in a solution of ∼33% H2O2 and ∼66% H2SO4. The silicon substrates were then rinsed with Millipore water, and kept in a solution of Millipore water until use (>24 h). The resulting surface was hydrophilic in nature. Initial experi-
Biomacromolecules, Vol. 6, No. 6, 2005 3335
ments had been performed using a hydrophobic surface prepared by etching the silica surface with HF. However, this resulted in a beading up of the solution upon sampling, and a disturbance to the interfacial protein layer. Therefore, all AFM results reported here are for samples prepared using hydrophilic silica substrates. The cage plates, extension rods, mounting post, cage clamp, and optics mirror with protected gold coating used in the IRRAS system were purchased from Thorlabs. The Spectra-Tech IR polarizer used with the IR system was originally purchased from Thermo Electron Corp. For the purpose of these experiments, a citrate buffer system was chosen over a phosphate buffer system, which is more commonly used to study the behavior of β-casein in solution. Because a goal of this study was to evaluate the interactions between calcium and β-casein, phosphate buffer, which would result in the precipitation of calcium as insoluble calcium phosphate, was not appropriate. The citrate buffers with varying calcium concentrations were prepared by first making a stock solution of 100 mM sodium citrate solution and adjusting to pH 7.4 using HCl. The individual solutions were then prepared by diluting the stock solution to a concentration of 10 mM, and adding calcium chloride at the desired concentration (based on the intended calcium: casein ratio for the sample). The ionic strength of the solutions was kept constant by adding NaCl to the diluted buffer solutions as necessary. Casein was weighed out and reconstituted in buffer to obtain a final concentration of 50 µg/mL (∼2.08 µM). This concentration is approximately 1/10 of the reported value for the critical micelle concentration of β-casein.17 Rheology Measurements. Rheological experiments were performed using a magnetic rod interfacial rheometer similar to those developed by Shahin and others,28-30 specifically described by Bantchev and Schwartz.9 The instrument consists of a hydrophobic, ferromagnetic rod that floats at the air-solution interface as the result of capillary forces. Electromagnetic coils, positioned on both sides of the apparatus, are used to apply an oscillatory force to the rod. The rod moves lengthwise within a glass channel, while a CCD camera centered over the rod tracks its motion. The knowledge of the applied force, which is proportional to the difference in the current between the two electromagnetic coils, allows for the determination of the applied stress. The rod displacement as measured by the CCD camera allows for the determination of the resulting strain. The stress and strain waves are then used to calculate the rheological parameters, including the complex dynamic interfacial shear modulus (G*), represented as a combination of an elastic (storage) modulus (G′) and a viscous (loss) modulus (G′′), G* ) G′ + iG′′. G′ is related to the in-phase, or real, component of the oscillatory flow (G′ ) (|σ|/|γ|) cos φ), while G′′ is related to the out-of-phase, or imaginary, component of the oscillatory flow (G′′ ) (|σ|/|γ|) sin φ). See Figure 1 for a diagram of the data collected in a rheological experiment. In our experiments, 100 mL of buffer solution was added to the sample holder to calibrate the apparatus. The sample holder was a rectangular dish with a glass channel in the center. The inner channel had an open-ended configuration
3336
Biomacromolecules, Vol. 6, No. 6, 2005
Figure 1. Schematic diagram of the data collected in a rheological experiment. Dotted lines represent the stress or force applied to the rod by the magnetic coils, while the solid lines represent the resulting motion of the rod as recorded by the CCD camera.
to reduce the constriction of the interfacial layer due to end effects. The top of the channel was open, but the dish was covered with an acrylic lid during the experiments to reduce solution evaporation. The channel width was 1.64 cm and the channel length 15 cm. The channel and dish were precleaned with a strong oxidizing solution (∼1/3 hydrogen peroxide, ∼2/3 sulfuric acid) to remove any protein residue or other surface contaminants from the channel. This procedure also acted to maximize the hydrophilicity of the channel, which aided in the proper centering of the rod between the channel walls due to capillary forces. The channel was centered between the two electromagnetic coils for both the calibration and the experiment, its length parallel to the direction of the magnetic field, so that the motion of the rod was along the axis of the coils. The rod was made of straightened, hardened steel music wire, 0.6 mm in diameter and ∼3.0 cm in length, with alternating dark and light stripes produced through an anodization procedure. The stripes on the wire allowed for more accurate detection of the rod’s motion by the CCD camera. The music wire was magnetized prior to performance of the calibration by placing it in the center of a coil of magnet wire and oscillating power through the coil at 5 × 10-2 A and a frequency of 0.25 Hz to create a magnetic field of approximately 6 × 10-2 T. Prior to the experiments, the rod was dipped in a warm solution of paraffin wax in hexanes. As the solution on the rod cooled, the hexanes evaporated, leaving a hydrophobic paraffin coating on the rod surface. The calibration was performed following the methods described by Brooks.28 Once the buffer was added to the channel, the rod was placed at the air-solution interface and the calibration was performed through a series of oscillation experiments performed at varying frequencies. When the calibration was complete, the rod was removed from the interface and the buffer solution removed from the channel using a vacuum system. This method prevented the need for repositioning the channel, so it was unnecessary to realign and focus the CCD camera. As a result, data were collected almost immediately upon pouring of the protein solution into the channel. A 100 mL portion of the protein solution was added to the channel for the experiment. The rod was again placed on the air-solution interface, and the sample analysis initiated. The oscillation experiments were repeated in a cyclic manner until the analysis was terminated. The analysis was run for between 12 and 72 h, with the majority of the experiments run in the 12-16 h time frame. At very long
Vessely et al.
times, movement of the rod was prohibited by the adsorbed protein layers, resulting in large inconsistencies in the data. A minimum of three replicates of the experiment were performed for each condition tested. Surface Pressure Measurements. Surface pressure was measured using a Wilhelmy plate tensiometer made by Riegler & Kirstein, with a 3 mm wide filter paper plate. The instrument was first calibrated in Millipore water. In these experiments, the measurements were intended to give the difference in surface pressure between a buffer solution and a protein solution. The measurements were performed by placing the filter paper plate into a solution containing buffer only and zeroing the instrument. The solution was then removed through a vacuum system and replaced by an equal volume of protein solution. The experiments were performed in this manner to mimic the procedures used in the rheological experiments. Because there was some motion to the plate upon pouring in the protein solution, there was some noise in the measurements at very early time points (
