A QCM Study of the Immobilization of β-Galactosidase on

Nov 15, 2006 - Surface Activity. Rebecca E. Hamlin,‡ Talya L. Dayton,†,‡ Lewis E. Johnson, and Malkiat S. Johal*. Department of Chemistry, Pomon...
1 downloads 0 Views 231KB Size
Download PDF
4432

Langmuir 2007, 23, 4432-4437

A QCM Study of the Immobilization of β-Galactosidase on Polyelectrolyte Surfaces: Effect of the Terminal Polyion on Enzymatic Surface Activity Rebecca E. Hamlin,‡ Talya L. Dayton,†,‡ Lewis E. Johnson, and Malkiat S. Johal* Department of Chemistry, Pomona College, 645 North College AVenue, SeaVer North, Claremont, California 91711-6338 ReceiVed NoVember 15, 2006. In Final Form: January 13, 2007 This work describes the immobilization of β-galactosidase onto polyelectrolyte multilayer assemblies of the polyanion poly[1-[4-(3-carboxy-4-hydroxyphenylazo)benzenesulfonamido]-1,2-ethanediyl, sodium salt] (PAZO) and the polycation poly(ethylenimine) (PEI) constructed by electrostatic self-assembly (ESA). A single layer of β-galactosidase was deposited over a precursor film comprising up to five bilayers of the PEI/PAZO polyelectrolyte pair. The enzyme was deposited on both the polycationic (PEI) and the polyanionic (PAZO) surfaces. Quartz crystal microbalance with dissipation monitoring (QCM-D), single-wavelength ellipsometry, and UV-visible absorption spectroscopy revealed differences in both the amount of β-galactosidase incorporated in each of the multilayer assemblies and the resulting enzyme packing density in the films. The enzymatic films were immersed in a reaction solution containing o-nitrophenylβ-D-galactopyranoside (ONPG), and absorbance measurements were used to monitor the concentration of o-nitrophenyl (ONP), the product of the β-galactosidase catalyzed by hydrolysis of ONPG. Although our data indicate that comparable amounts of β-galactosidase are incorporated onto both surfaces, enzymatic activity is substantially inhibited when the β-galactosidase is immobilized on the polyanionic surface compared to the enzyme on the polycationic surface. The difference in catalytic activities reflects the different abilities of the two polyelectrolytes to screen the protein’s active site from the substrate environment. In both assemblies, the protein interpenetrated the PEI/PAZO multilayer, disrupting the J-aggregated state of the PAZO chromophores. This work demonstrates that the charge, conformation, and composition of underlying polyelectrolyte cushions have a significant effect on the structure and function of an immobilized protein within functional nanoassemblies.

Introduction There are many technological applications for biofunctionalized surfaces. In particular, the recent revival in studies of enzymatic surfaces has been accentuated by essential developments in enzyme immobilization and entrapment techniques. Compared to enzymes in solution, enzymatic surfaces are thermally stable, less susceptible to inhibition or protease attack, and are often reusable.1,2 Furthermore, the immobilization of enzymes allows for greater control of an enzyme’s catalytic activity. Traditionally, immobilization methods have included chemisorption,2 Langmuir-Blodgett deposition,3 intermolecular crosslinking,4 adsorption, and entrapment.5,6 More recently, layerby-layer (LbL) electrostatic self-assembly (ESA)7,8 has been employed to anchor a variety of enzymes on polyelectrolyte surfaces.9 Polyelectrolyte films offer a soft medium for enzyme entrapment in which charged protein segments (amino acid residues) and hydrophobic protein chains interact strongly with * Author to whom correspondence should be addressed. E-mail: [email protected]. Fax: (909) 607-7726. http://pages.pomona.edu/ ∼msj04747/. † Current address: Dipartimento di Chimica e Tecnologie Farmaceutiche, Universita di Palermo Via Archirafi 32, 90123 Palermo, Italy. ‡ These authors contributed equally to this work. (1) Ariga, K.; Nakanishi, T.; Michinobu, T. J. Nanosci. Nanotechnol. 2006, 6(8), 2278-2301. (2) Taylor, R. F., Ed. Protein Immobilization: Fundamentals and Applications; Marcel Dekker: New York, 1991. (3) Lvov, Y., Mo¨hwald, H., Eds. Protein Architecture: Interfacing Molecular Assemblies and Immobilization Biotechnology; Marcel Dekker: New York, 2000; p 99-124, 193-228, 229-250. (4) Siouffi, A.-M. J. Chromatogr., A 2003, 1000, 801-818. (5) Taylor, R. F., Ed. Protein Immobilization: Fundamentals and Applications; Marcel Dekker: New York, 1991. (6) Asanov, A. N.; DeLucas, L. J.; Oldham, P. B.; Wilson, W. W. J. Colloid Interface Sci. 1997, 62, 196. (7) Decher, G. Science 1997, 277, 1232-1237. (8) Decher, G.; Hong, J.-D.; Schmitt, J. Thin Solid Films 1992, 210-211, 831-835.

the underlying polycation or polyanion. ESA is a process that involves the interlaced deposition of a polycation and a polyanion to yield charge-alternating multilayers. The success of this alternating deposition technique is based upon the electrostatic adsorption of a polyion from aqueous solution onto a substrate primed with a polyion of opposite charge. Charge overcompensation and entropic gain due to counterion expulsion are generally cited as the driving mechanisms behind layer formation.10 The variety of biomolecules and enzymes that can immobilize onto polyelectrolyte multilayer surfaces underscores the importance of ESA as an important method for biofunctionalization. This array of molecule types includes: DNA,6 charged viral capsids,11 collagen,12 concanavalin A,13 myoglobin,14 lactate dehydrogenase,15 human serum albumin (HSA),16 and bovine serum albumin (BSA).6 The most important characteristic of each of these films is the marked retention of the biological activity of the biomacromolecule of interest. Caruso et al. were able to immobilize Immunoglobulin G (IgG) and anti-immunoglobulin G (anti-IgG) onto a film consisting of two bilayers (9) Tripathy, S. K., Jayant, K., Singh, N. H., MacDiarmin, A. G., Eds. Polyelectrolyte-Based Multilayers, Self-Assemblies and Nanostructures. In Handbook of Polyelectrolytes and Their Applications; American Scientific Publishers: Stevenson Ranch, CA, 2002; Vol. 1. (10) Schlenoff, J. B.; Dubas, S. T. Macromolecules 2001, 34, 592. (11) Lvov, Y.; Haas, H.; Decher, G.; Mohwald, H.; Mikhailov, A.; Mtchedlishvily, B.; Morgunova, E.; Vainshtein, B. Langmuir 1994, 10, 4232-4236. (12) Zhang, J.; Senger, B.; Vautier, D.; Picart, C.; Schaaf, P.; Voegel, J.-C.; Lavalle, P. Biomaterials 2005, 26, 3353-3361. (13) Lvov, Y.; Ariga, K.; Ichinose, I.; Kunitake, T. J. Am. Chem. Soc. 1995, 117, 6117-6123. (14) Lvov, Y. M.; Lu, Z.; Schenkman, J. B.; Zu, X.; Rusling, J. F. J. Am. Chem. Soc. 1998, 120, 4073-4080. (15) Martin, R. In Incorporation of ActiVe Lactate Dehydrogenase into Polyelectrolyte Multilayer Assemblies; Thesis submitted to the Division of Natural Science, New College of Florida, 2005. (16) Szyk, L.; Schwinte, P.; Voegel, J.-C.; Schaaf, P.; Tinland, B. J. Phys. Chem. B 2002, 106, 6049-6055.

10.1021/la063339t CCC: $37.00 © 2007 American Chemical Society Published on Web 03/13/2007

ESA of β-Galactosidase on Polyelectrolyte Surfaces

of polyallylamine hydrochloride (PAH) and polystyrene sulfonate (PSS). By showing that IgG would bind specifically to the immobilized anti-IgG, they demonstrated that the bound antibody retained its biological activity.17 Their results indicated that the specific binding increased as the amount of immobilized antiIgG was increased. When BSA was used to block nonspecific binding sites, less than 10% reduction in binding of IgG was observed.17 Constructing bioactive films on colloidal surfaces also allows one to counteract the problem encountered with planar substrates where diffusion of the reactant (substrate) in the bulk phase into the enzymatic film becomes limiting with increasing film thickness.18 Colloidal or spherical supports are particularly useful for the immobilization of various enzymes.18 After loading the porous spheres with enzymes, encapsulation using conventional polyelectrolytes has been shown to prevent desorption of the enzyme into the substrate solution. We use quartz crystal microbalance with dissipation measurements (QCM-D) to monitor the in situ deposition of β-galactosidase on pre-adsorbed polycation (poly(ethylenimine), henceforth denoted PEI) and polyanion (poly[1-[4-(3-carboxy-4hydroxyphenylazo)benzenesulfonamido]-1,2-ethanediyl, sodium salt], henceforth denoted PAZO) surfaces. Changes in adsorbed mass as a function of time allows the determination of binding constants that characterize the interaction between the polyelectrolyte surface and the enzyme. QCM measurements are noninvasive and have nanogram sensitivity with respect to measured masses. Although ESA has established itself as a viable method of immobilizing enzymes on polyelectrolyte surfaces, there has been little investigation into how the nature of the polyelectrolyte adsorbent influences the retention and catalytic properties of the immobilized enzyme. In this work, we explore the structure and activity of immobilized β-galactosidase on a pre-adsorbed polycation and polyanion surface and observe differences in deposition amounts on the two surfaces. For both polycationic and polyanionic surfaces, the J-aggregated state of the polyelectrolyte multilayer is disrupted by the presence of the enzyme, likely through extensive interpenetration. More importantly, when β-galactosidase is adsorbed on the anionic PAZO surface, the catalytic ability of the resulting assembly is strongly inhibited. This work demonstrates that an anionic enzyme can be immobilized on cationic or anionic polyelectrolyte surfaces and that the catalytic activity of the resulting assembly depends on the interaction of the immobilized enzyme with the underlying polyelectrolyte layers. Experimental Section The sequential adsorption of PEI and PAZO and the preparation of the glass and silicon substrates for conventional LbL ESA are described elsewhere.19-22 PEI (MW ) 25,000 g mol-1, mixture of linear and branched chains) and PAZO (average MW ≈ 65,000100,000 g mol-1, CAS 24615-84-7) were obtained from Aldrich and were used as received. All polyelectrolyte solutions were prepared at 1 mM (based on the molecular weight of the polymer repeat unit) using ultrapure water (resistivity >18 MΩ cm) at pH 7. Partially purified, lyophilized β-galactosidase was obtained from Worthington (17) Caruso, F.; Niikura, K.; Furlong, D. N.; Okahata, Y. Langmuir 1997, 13, 3427. (18) Yu, A.; Liang, Z.; Caruso, F. Chem. Mater. 2005, 17, 171. (19) Casson, J. L.; Wang, H.-L.; Roberts, J. B.; Parikh, A. N.; Robinson, J. M.; Johal, M. S. J. Phys. Chem. B 2002, 106, 1697. (20) Casson, J. L.; Johal, M. S.; Roberts, J. B.; Wang, H.-L.; Robinson, J. M. J. Phys. Chem. B 2000, 104, 11996. (21) Chiarelli, P. A.; Johal, M. S.; Holmes, D. J.; Casson, J. L.; Roberts, J. B.; Robinson, J. M.; Wang, H.-L. Langmuir 2002, 18, 168. (22) Johal, M. S.; Casson, J. L.; Chairelli, P.; Lui, D.-G.; Shaw, J. A.; Robinson, J. M.; Wang, H.-L. Langmuir 2003, 19, 8876.

Langmuir, Vol. 23, No. 8, 2007 4433 Biochemical Corporation and was used as received. Enzyme solutions were prepared at 1 mg/mL using standard Z-buffer (5 mM Na3PO4 buffer pH 7.0, 250 mM KCl, 1 mM MgSO4, ultrapure water). Adsorbed masses during LbL build-up, binding constants, and viscoelastic properties of the multilayer films were obtained using a quartz crystal microbalance (QCM) with dissipation measurement (E4, Q-Sense, Gothenburg, Sweden). The QCM sensor consisted of a disk-shaped, AT-cut piezoelectric quartz crystal, coated with metallic electrodes on both sides. The QCM sensor crystal (14 mm × 0.3 mm) operates at a frequency of 4.95 MHz ( 50 kHz. We used quartz crystals coated with an Au electrode (100 nm thick) on the back and an active surface layer of SiO2 (∼50 nm thick). The crystals were optically polished with a surface roughness of less than 3 nm (rms). The active side was in contact with the aqueous polyelectrolyte solution. The crystal was mounted in a flow cell with a total volume of 40 µL. The QCM crystal can be induced to oscillate at its fundamental resonant frequency by applying an oscillating potential across the electrodes. If a small amount of mass adsorbs to the electrode surface, a reduction in the resonance frequency is observed. If the adsorbed mass is rigidly fixed to the crystal surface and much smaller than the mass of the QCM crystal, the frequency reduction is proportional to adsorbed mass. The resonant frequency (∆F) of the crystal depends on the total oscillating mass (∆m), including water coupled to the oscillation. In this way, the QCM operates as a very sensitive balance. The mass of the adhering layer is calculated by using the Saurbrey relation,1 where C ) 17.7 ng cm-2 Hz-1 for a 5 MHz quartz crystal, and n represents the overtone number. ∆F ) -

∆m nC

(1)

A film that is “soft” (viscoelastic) will not fully couple to the oscillation of the crystal, which dampens the crystal’s oscillation. The dissipation (D) of the crystal’s oscillation is a measure of the film’s softness (viscoelasticity). The dissipation of the crystal is measured by recording the response of a freely oscillating crystal that has been vibrating at its resonant frequency. Thus viscosity, elasticity, and the correct thickness may be extracted even for soft films. Prior to use, the SiO2-coated quartz crystals were washed in 10% SDS solution and then cleaned in a UV/ozone chamber for 30 min. UV/ozone cleaning was followed by a 1-h continuous rinse with ultrapure water (resistivity >18 MΩ cm). All polyelectrolyte and β-galactosidase solutions were introduced into the flow cell at a rate of 300 µL/min. In each deposition step, ∆F decreased with time, and D increased with time. The QCM cell was rinsed with ultrapure water between polyelectrolyte/enzyme depositions, which were considered to be complete when ∆F and D stopped changing with time. For the QCM experiments, β-galactosidase was constructed on a precursor polyelectrolyte layer comprising three bilayers of PEI/PAZO. For thickness and absorbance measurements, the substrates for film deposition were glass microscope slides for UV-visible spectroscopy measurements and 1-in. square polished silicon wafers for ellipsometry. The same native oxide surface layers for all substrate types provided reproducible surface conditions for film deposition. Substrates were prepared by immersion in a 30:70 H2O2(30% weight)/ H2SO4 mixture for 1 h at 80 °C (piranha etch treatment). This treatment exposes free silanol (-Si-OH) groups on the substrate surface, which subsequently deprotonate in higher pH solutions (pH > 3), thus resulting in an overall negatively charged surface. Following this treatment, substrates were rinsed thoroughly in water, sonicated for 15 min to remove any remaining etch solution, and then stored in water. Prior to film deposition, substrates were rinsed thoroughly in water and dried under a stream of nitrogen gas. For thickness and absorbance measurements, β-galactosidase was constructed on a precursor polyelectrolyte layer comprising five bilayers of PEI/ PAZO. UV-visible absorbance measurements of the multilayered films built on glass substrates were recorded between 190 and 800 nm on

4434 Langmuir, Vol. 23, No. 8, 2007

Hamlin et al.

a Varian Cary 300 spectrophotometer. Spectra were obtained for each of the five bilayers and for each terminal β-galactosidase layer. Thickness measurements were collected by single-wavelength null ellipsometry using a Rudolph Instruments model 439L633P ellipsometer. Data were collected at a beam incidence angle of 70° and a wavelength of 632.8 nm. A refractive index of 1.5 ( 0i was used to manually calculate ellipsometric film thickness from ∆ and Ψ parameters. The thickness of the native oxide layer on Si was determined for every substrate used. This value was then subtracted from the film measurements to determine actual ellipsometric film thickness. The reported (2 Å error represents the instrument’s accuracy. The actual roughness of the substrate is much greater. All thickness measurements were taken at exactly the same spot after each deposition. The presence of active β-galactosidase was monitored by its reaction with the lactose analogue o-nitrophenyl-β-D-galactopyranoside (ONPG), which generates a colored product, o-nitrophenol (ONP). The concentration of ONP can be quantified by measuring the absorbance of the reaction solution at 420 nm, the λmax for ONP.23 To study the catalytic activity of the immobilized β-galactosidase, the films were immersed in 45 mL of solution 1, which consisted of 36.6 mL of Z-buffer, 100 µL of β-mercaptoethanol, and 8.4 mL of ONPG (4 mg/mL). At 10 min intervals, 1 mL of solution 1 was removed into a quartz cuvette, 0.4 mL of 1 M Na2CO3 was added, and the absorbance at 420 nm was measured using a Spectronic 20 (Genesys) spectrophotometer. After 270 min, 0.8 mL of 1 M Na2CO3 was added to 2 mL of the remaining reaction solution, and the UV-visible absorbance spectrum between 190 nm and 800 nm was recorded. The baseline sample for the spectra consisted of a mixture of 1 mL of Z-buffer and β-mercaptoethanol and 0.4 mL of 1 M Na2CO3.

Results and Discussion β-galactosidase was deposited on a precursor film composed of PEI and PAZO bilayers. The enzyme was deposited on both the PAZO- and the PEI-ending films. These films are henceforth denoted as [PEI/PAZO]3-β-gal and [PEI/PAZO]3-PEI-β-gal, respectively, where 3 refers to the three precursor bilayers composed of PEI and PAZO (including coupled water). Figure 1 shows the final mass adsorbed for all layers, including the terminal β-galactosidase layer. These masses were determined from the QCM frequency shifts using eq 1. In both systems, the precursor polyelectrolytes PEI and PAZO built linearly with respect to adsorbed mass, with PAZO adding slightly more mass per bilayer. The addition of β-galactosidase, which included coupled water, resulted in a relatively large mass uptake in both systems. The amount of enzyme adsorbed on a terminal polycationic PEI layer (i.e., the [PEI/PAZO]3-PEI-β-gal system) is 990 ( 2 ng/cm2. The amount of enzyme adsorbed on the negatively charged PAZO layer (i.e., the [PEI/PAZO]3-β-gal) is significantly less (620 ( 2 ng/cm2). This difference likely reflects differences in noncovalent interactions between amino acid residues on the exterior of the enzyme and the polyelectrolyte surface. Displacement of the underlying polyelectrolyte by β-galactosidase can also influence the frequency shifts; however, UV-visible absorption spectroscopy did not indicate any significant polyelectrolyte displacement (see below). In both cases the adsorption of the enzyme was measured as a function of time (Figure 2). Fitting a simple first-order rate expression allowed the determination of the rate constant (Figure 3). Despite the different charges of the underlying polyelectrolytes, the rate constants were comparable (k ) 0.1099 ( 0.0005 s-1 for both the [PEI/PAZO]3-PEI-β-gal and the [PEI/PAZO]3-β-gal systems). It should be noted that the starting mass of the [PEI/ PAZO]3 film is greater than that of the [PEI/PAZO]3-PEI film. (23) Langley, K. E.; Villarejo, M. R.; Fowler, A. V.; Zamnhof, P. J.; Zabin, I. Proc. Natl. Acad. Sci. U.S.A. 1975, 72(4), 1254-1257.

Figure 1. Frequency shift response (Saurbrey mass) as a function of polyelectrolyte layer number. The first point (layer 1) corresponds to the deposition of PEI. The terminal layer (solid circle) corresponds to the deposition of β-galactosidase. (a) Changes in mass when the enzyme is added to a terminal layer of PEI and (b) changes in mass when the enzyme is added to a terminal layer of PAZO. Total buildup of mass (∆m) due to enzyme deposition is indicated in both systems. The uncertainly in ∆m is about 2 ng cm-2.

The overall thickness (or mass) of the polyelectrolyte multilayer depends on the charge density of the substrate, which itself is a somewhat difficult variable to control. Note that in Figure 1, the mass differences are largest near the substrate (layers 1-3). This was the main reason for depositing the enzyme on a relatively thick polyelectrolyte precursor layer. Despite the substrate effect, our laboratory has consistently observed greater amounts of enzyme deposition on PEI compared to that on PAZO. Figure 2 shows the mass of adsorbed β-galactosidase on both surfaces and the corresponding energy dissipation as a function of time. The time W corresponds to the water rinse following enzyme deposition. In both cases, the dissipation increases concurrently with mass, indicating the decreasing rigidity of the resulting film. At time W, there is a significant decrease in mass and dissipation. This drop in mass is followed by a slight rise in mass, which then levels off. We interpret the initial drop as slight enzyme desorption, and the following rise and leveling off as swelling of the film due to slow incorporation of water molecules into the film. This interpretation is consistent with the energy dissipation measurements where desorption leads to an increase in film rigidity and swelling leads to a decrease in rigidity. In both cases, it takes about 15 min to reach equilibrium. Although the decrease in mass at time W and the subsequent increase are comparable in both cases, the dissipation drop is larger in the [PEI/PAZO]3-PEI-β-gal system. At equilibrium,

ESA of β-Galactosidase on Polyelectrolyte Surfaces

Figure 2. Frequency shift response (Saurbrey mass) and dissipation changes due to the adsorption of β-galactosidase on the polyelectrolyte film. (a) Mass and dissipation as a function of time for deposition of the enzyme on the PEI surface, and (b) mass and dissipation as a function of time for deposition of the enzyme on the PAZO surface. Total build up of mass (∆m) due to enzyme deposition is indicated in both systems.

Figure 3. Frequency shift response (Saurbrey mass) as a function of time due to the adsorption of β-galactosidase on the polyelectrolyte film. The solid dark line represents a fit to the data using the functional form A(1 - exp(-kt)) + B. The fit shown represents the deposition of β-galactosidase on the PAZO surface of the [PEI/PAZO]3 precursor film.

the dissipation factor for the [PEI/PAZO]3-PEI-β-gal system is approximately one-half of the corresponding value for the

Langmuir, Vol. 23, No. 8, 2007 4435

[PEI/PAZO]3-β-gal system. These differences in dissipation demonstrate that the [PEI/PAZO]3-PEI-β-gal film is more rigid than the [PEI/PAZO]3-β-gal film, despite the fact that the former contains an “extra” polyelectrolyte (PEI) layer. It should be noted that these differences in rigidity may also be influenced by the high ionic strength of the enzyme solution, since salt is known to have a dramatic effect on the multilayer morphology. Figures 1 and 2 provide strong evidence that β-galactosidase adsorbs to a greater extent on PEI than on PAZO, and adsorption on PEI results in a film that is more dense and rigid. It is likely that this is indicative of strong intermolecular interactions between the PEI and the exterior of the enzyme. In addition to precise values for adsorbed masses, QCM provides valuable insights into the dynamics of adsorption and the mechanical properties of the resulting film. Furthermore, measurements of thickness and UV-visible absorbance can be used in conjunction with QCM data to provide film packing density and information about interlayer interpenetration. Previous work has shown that PAZO within PEI/PAZO bilayers shows a prominent red-shift in the UV-visible absorbance spectrum compared to the aqueous bulk phase (λmax(bulk) ) 350 nm, λmax(film) ) 360 nm).24 This shift is due to the presence of noncovalent interactions, or chromophore-coupling effects, between the azo-benzene chromophores in aggregated systems that are absent in dilute solutions.25 For a perfect one-dimensional aggregate with transition dipoles aligned parallel to one another, the two idealized arrangements of chromophores are called H-aggregates for dipoles arranged in a “head-to-head” and “tailto-tail” arrangement, and J-aggregates for dipoles that are arranged in a “head-to-tail” arrangement. Aggregation leading to a prominent spectral blue-shift is attributed to H-type aggregation, while aggregation leading to a red-shift is attributed to J-type aggregation.26 In order to obtain ellipsometric thickness and absorption measurements, we used chemically etched silicon substrates with roughnesses considerably greater than the surface of the QCM crystals used for the deposition experiment. Thus, precursors comprised of five PEI/PAZO layers were constructed before enzyme deposition. Figure 4a shows the thickness of the polyelectrolyte film as a function of the PEI/PAZO bilayers. The thickness of the [PEI/PAZO]5 precursor is 171 ( 2 Å, and the average increase in thickness per bilayers is 34 ( 2 Å. With the addition of β-galactosidase, the thickness increases to 196 ( 2 Å. This increase in 25 Å is attributed to the deposition of the protein. Figure 4b shows the absorbance of the film at λmax ) 360 nm as a function of bilayer number. A linear increase in absorbance is observed for the precursor bilayers. Although the addition of β-galactosidase shows no additional increase in absorbance at this wavelength, a significant blue-shift (∆ λmax ) 10 nm) is observed in the absorbance spectrum (Figure 4c). Furthermore, as shown in Figure 4c, the absorbances at 190 and 230 nm increase after the adsorption of the protein (∆A(190 nm) ) 0.035, ∆A(230 nm) ) 0.020). A similar pattern is observed for the [PEI/PAZO]5-PEI-β-gal system (Figure 5). In this case, the precursor film grows by only 15 Å after the addition of the protein, and a similar blue-shift of 10 nm in the absorbance spectrum of the precursor film is observed. The absorbance increases at 190 and 230 nm following protein deposition are similar for the [PEI/PAZO]5-PEI-β-gal system (∆A(190 nm) (24) Campbell, V. E.; Chiarelli, P. A.; Kaur, S.; Johal, M. S. Chem. Mater. 2005, 17, 186-190. (25) Kuhn, H.; Kuhn, C. In J-Aggregates; Kobayashi, T., Ed.; World Scientific: Singapore, 1996; pp 1-40. (26) Dante, S.; Advincula, R.; Frank, C. W.; Stroeve, P. Langmuir 1999, 15, 193.

4436 Langmuir, Vol. 23, No. 8, 2007

Figure 4. (a) Ellipsometric thickness and (b) absorbance at 360 nm as a function of PEI/PAZO bilayer number. The solid data point represents addition of β-galactosidase. (c) UV-visible absorbance spectrum of [PEI/PAZO]5 (solid line) and [PEI/PAZO]5-β-gal (dashed line).

) 0.028, ∆A(230 nm) ) 0.022). The changes in absorbance and the thickness values imply that the [PEI/PAZO]5-PEI-β-gal assembly is more densely packed compared to the [PEI/PAZO]5β-gal system. At pH 7.0, β-galactosidase bears an overall negative charge; however, deposition of β-galactosidase occurs to a significant extent on both a polycationic and a polyanionic surface. It is important to note that, although the ESA process is dominated by electrostatic interactions, adsorption is observed even if the overall charges of the enzyme and the terminal polyelectrolyte layer are the same.27 The electrostatic argument for the adsorption of proteins onto like-charged surfaces emphasizes that the net charge of a protein is simply the net sum of the charged amino (27) Ladam, G.; Schaaf, P.; Cuisinier, F. J. G.; Decher, G.; Voegel, J.-C. Langmuir 2001, 17, 878-882.

Hamlin et al.

Figure 5. (a) Ellipsometric thickness and (b) absorbance at 360 nm as a function of PEI/PAZO bilayer number. The solid data point represents addition of β-galactosidase. (c) UV-visible absorbance spectrum of [PEI/PAZO]5-PEI (solid line) and [PEI/PAZO]5-PEIβ-gal (dashed line).

acids that comprise the protein. Thus, although the overall charge of β-galactosidase is negative, there exist domains of positively charged amino acid residues that can interact with the negatively charged PAZO surface. The observed blue-shift in the spectrum of both films results in a λmax value consistent with a dilute unaggregated solution of PAZO. We attribute this shift to the disruption of the J-aggregated state of the PAZO chromophores within the assembly due to extensive interpenetration of the protein into the multilayer. Furthermore, since we know the mass and thickness, we can calculate the density of the polymer and protein layers. The effective density for the protein (enzyme) film is about 6 g cm-3 for adsorption on PEI and 2.5 g cm-3 for adsorption on PAZO. The density of polymer (PEI/PAZO) films is about 0.7 g cm-3. This calculation shows that enzyme molecules can penetrate to all depths of the film and can, thus, disrupt the PAZO aggregation. The absorbance data suggest significant mixing between the enzyme and the polyelectrolyte layer. The QCM data indicate that more enzyme is added to the PEI layer than to the PAZO

ESA of β-Galactosidase on Polyelectrolyte Surfaces

Figure 6. Absorbance per ng at 420 nm as a function of time for the hydrolysis of ONPG catalyzed by the active β-galactosidase assemblies. [PEI/PAZO]5-PEI-β-gal (squares) and [PEI/PAZO]5β-gal (circles).

layer. The thickness data show a smaller increase in thickness when the enzyme is added to the PEI layer. The combined thickness and mass data suggest that the [PEI/PAZO]n-PEIβ-gal system yields a more dense and rigid film. The greater binding of β-galactosidase to PEI is further supported by a much smaller energy dissipation factor compared to that of the [PEI/ PAZO]n-β-gal system. Thus, even though the same polyelectrolytes were used in both systems, there are large differences in the resulting film structures. These differences also affect the catalytic activity of the enzyme. We tested the catalytic activity of both the [PEI/PAZO]5-β-gal and [PEI/PAZO]5-PEI-β-gal assemblies. Figure 6 shows the absorbance at 420 nm of each of the reaction solutions as a function of time. The observed increase in the absorbance at this wavelength is indicative of an increasing concentration of o-nitrophenyl (ONP) in the reaction solution, which can be attributed to the hydrolysis of o-nitrophenyl-β-D-galactopyranoside (ONPG) catalyzed by active β-galactosidase. For comparison purposes, absorbance values shown are per unit mass. The rate of the reaction catalyzed by the [PEI/PAZO]5-PEI-β-gal assembly is much greater than that of the reaction catalyzed by the [PEI/PAZO]5-β-gal assembly. At 270 min, the former reaches an absorbance value of ∼1.9 × 10-3 ng-1 at 420 nm, whereas the latter only reaches an absorbance value of ∼1.6 × 10-4 ng-1 at this wavelength. We suggest that the direct interaction of PEI and β-galactosidase in [PEI/PAZO]5-PEI-β-gal assemblies stabilizes the enzyme’s active conformation to a greater degree than the direct interaction of PAZO and β-galactosidase in [PEI/PAZO]5-β-gal assemblies. It must be noted that at 270 min, no further increase in absorbance was observed in the enzyme-free solutions. This demonstrates that very little or no β-galactosidase leaches from the multilayered assembly into the ONPG solution. It should be emphasized that electrostatic effects may not be the only factors influencing the enzymatic activity. Hydrophobic and steric factors may also play an important role. Previous circular dichroism and enzyme activity measurements of immobilized lactate dehydrogenase and β-galactosidase show that PEI stabilizes the structure of these enzymes in the bulk phase.28 (28) Andersson, M. M.; Hatti-Kaul, R. J. Biotechnol. 1999, 72, 21.

Langmuir, Vol. 23, No. 8, 2007 4437

Thus, it is possible that we are observing the stabilizing effect of PEI on the β-galactosidase adsorbed onto PEI/PAZO multilayer assemblies. Previously, Schwinte et al. found that the interactions of PSS and hen egg white lysozyme (HEL) adsorbed onto PSS/ PAH multilayer assemblies stabilized the structure of a protein.29 Furthermore, Asanov et al. found that protein secondary structure remains mostly unchanged by the electrostatic interactions between an enzyme and the terminal polyelectrolyte layer onto which it has been adsorbed.6 Greater structural changes have been observed when the protein is adsorbed onto an oppositely charged terminal polyelectrolyte layer.16 Such structural changes can render the protein inactive, as observed in the [PEI/PAZO]5β-gal assembly. Finally, for the [PEI/PAZO]5-PEI-β-gal assemblies, the observed increase in thickness upon deposition of the terminal layer of β-galactosidase was notably less than the observed increase in thickness upon deposition of the enzyme onto a terminal, like-charged layer of PAZO. Previous work, however, reports that the observed thickness of protein layers adsorbed onto an oppositely charged surface was larger than that observed for protein layers adsorbed onto a like-charged surfacesthe thicknesses of protein layers adsorbed onto oppositely charged surfaces were consistent with protein aggregation.30 It is clear that the observed thicknesses of the [PEI/PAZO]5-PEI-β-gal and [PEI/PAZO]5-β-gal films are not directly related to the amount of enzyme immobilized but indicative of the density of the β-galactosidase layers. Therefore, the composition of polyelectrolyte cushions has marked effects on the structure, aggregation, and catalytic function of terminally adsorbed proteins.

Conclusions We demonstrated that β-galactosidase can be successfully immobilized onto terminal layers of both PEI and PAZO. The layer of β-galactosidase adsorbed onto an oppositely charged layer of PEI is denser than the layer adsorbed onto a like-charged layer of PAZO. Despite this difference, the enzyme had a comparable binding constant for both surfaces. The catalytic assays performed indicate that the [PEI/PAZO]n-PEI-β-gal film shows a higher degree of catalytic activity than the [PEI/PAZO]nβ-gal film, and β-galactosidase was, therefore, better able to retain its active structure. We suggest that PEI is able to stabilize the structure of β-galactosidase when the two are in direct contact within a polyelectrolyte multilayer assembly. Despite the aforementioned differences between [PEI/PAZO]n-PEI-β-gal and [PEI/PAZO]n-β-gal films, the UV-visible absorbance spectra of both these films demonstrated a blue-shift of 10 nm at 360 nm, suggesting that in both assemblies, the enzyme interpenetrated the multilayer, disrupting the J-aggregated state of the PAZO chromophores. Acknowledgment. Ellipsometric thickness data were collected at New College of Florida (NCF) by Miss Talya Dayton. This work was supported by the New College Foundation and the Pomona College Department of Chemistry. We thank Dr. Katherine Walstrom (NCF), Dr. Paul Scudder (NCF), and Dr. E. J. Crane (Pomona College) for useful discussions. LA063339T (29) Schwinte, P.; Ball, V.; Szalontai, B.; Haikel, Y.; Voegel, J.-C.; Schaaf, P. Biomacromolecules 2002, 3, 1135-1143. (30) Ladam, G.; Gergely, C.; Senger, B.; Decher, G.; Voegel, J.-C.; Schaaf, P.; Cuisinier, F. J. G. Biomacromolecules 2000, 1, 674-687.