J. Phys. Chem. B 2007, 111, 11603-11610
11603
Structure and Dynamics of Lysozyme Encapsulated in a Silica Sol-Gel Matrix Isabel Pastor,† Maria L. Ferrer,‡ M. Pilar Lillo,£ Javier Go´ mez,† and C. Reyes Mateo*,† Instituto de Biologı´a Molecular y Celular, UniVersidad Miguel Herna´ ndez, 03202-Elche, Spain, Instituto de Ciencia de Materiales de Madrid, CSIC, Campus de Cantoblanco, 28049-Madrid, Spain, and Instituto de Quı´mica Fı´sica “Rocasolano”, CSIC, 28003-Madrid, Spain ReceiVed: June 20, 2007; In Final Form: July 30, 2007
Proteins entrapped in sol-gel matrices have been extensively studied during the last 15 years, showing that most of them can be encapsulated with retention of their native structure and functionality and with enhanced stability. However, relatively little is known about the structural and dynamical details of the biomoleculematrix interactions. To achieve this goal, the model protein hen egg white lysozyme (HEWL) has been entrapped in sol-gel matrices prepared from tetraethyl orthosilicate through an alcohol-free sol-gel route, and the photophysical properties of its fluorescent tryptophans have been determined using both steady-state and time-resolved fluorescence techniques. By combining fluorescence spectra, quenching experiments, lifetimes, and time-resolved fluorescence anisotropy measurements, we have obtained information on the structure, dynamics, and solvation properties of the entrapped protein. Our results show that the environment of HEWL within the silica pore as well as its internal dynamics is similar to that in aqueous solution, except that the protein showed no or, depending on conditions, very much slower global motion but retained its internal angularly restricted (hindered) segmental rotation upon entrapment. The experiments carried out at different experimental conditions indicate that, below the isoelectric point of the protein, a strong electrostatic interaction is established between the protein molecule and the negatively charged sol-gel walls, which is ultimately responsible for the total arrest of the overall rotation of the protein, but without significant effect upon its segmental rotational relaxation. The electrostatic nature of the interaction is clearly established since either reducing the positive charge of the protein (by increasing the pH toward its isoelectric point) or increasing the ionic strength of the solution (shielding against the attractive interaction) leads to a situation in which the protein freely rotates within the matrix pore, albeit an order of magnitude more slowly than that in free solution under similar macroscopic solution conditions, and still retains its segmental rotational properties.
Introduction Silica sol-gel materials have been shown to form excellent media for immobilization of proteins, among them, active enzymes, and other biologicals macromolecules.1-3 The past decade has seen a great deal of progress in the encapsulation of these biomolecules in sol-gel matrices by modifying older conventional procedures that were designed to afford more biocompatible environments in such ways as to optimize functionality in the resulting bioceramic.4-7 The entrapped biomacromolecule usually retains its structural integrity and functionality and, being accessible to small molecules diffusing intothegel,maybeusedinanalytical2,8-13 andbiotechnological1,14-17 applications. Bioencapsulation in silica matrices has also been investigated as an experimental system for testing the effects of molecular confinement on the structure and stability of proteins18,19 and for controlling the kinetics of the unfolding and refolding processes, as well as for trapping and characterizing unstable forms that are likely candidates for transitionstate species.20-22 Quite apart from extending the applicability of these materials to new applications, there is still a need for some fundamental questions to be addressed, for example, about the nature of the biomolecule-matrix interactions and how they * To whom correspondence should be addressed. Fax: +34 966 658 758. E-mail:
[email protected]. † Universidad Miguel Herna ´ ndez. ‡ Instituto de Ciencia de Materiales de Madrid. £ Instituto de Quı´mica Fı´sica “Rocasolano”.
might be modulated in a rational way. To help toward achieving this goal, the rotational motions of different model proteins entrapped in sol-gel glasses have been analyzed through different spectroscopic techniques, and the results have been interpreted in terms of electrostatic and/or hydrogen bond interactions between the protein and the silica matrix7,23-25 (which has an isoelectric point, pI, of ∼2) or in terms of changes in the effective viscosity sensed by the protein.26-28 It should be noted that the influence of immobilization on the dynamic properties may vary among different proteins, depending on constraints caused by heterogeneity of the surface charge density of the protein and its pH-dependent interaction with the negatively charged silica matrix. In the case of the green fluorescent protein mutant GFPmut2, negatively charged at physiological pH, time-resolved fluorescence depolarization experiments indicated that the protein environment inside of the silica pores is similar to the one sensed by the protein in the aqueous solution and that unhindered molecular rotations occur.24 On the contrary, for the human serum albumin, significant restriction of its global rotational motion was observed, despite its overall negative charge at physiological pH.7 This behavior was ascribed to differences in the electrostatic interactions between the sol-gel matrix and the individual domains of the protein. These results are also consistent with the significantly restricted rotation observed, at neutral pH, for myoglobin (pI ∼ 7.2) by fluorescence depolarization27 and for highly basic horse heart cytochrome c (pI ) 9.45) by NMR.23
10.1021/jp074790b CCC: $37.00 © 2007 American Chemical Society Published on Web 09/13/2007
11604 J. Phys. Chem. B, Vol. 111, No. 39, 2007 It has been suggested that these proteins interact with, or are adsorbed to, the porous walls of the host matrix through hydrogen bonds or electrostatic interactions which effectively hinder the free rotation of the protein.29 With the aim of casting some light on the role of proteinmatrix interactions on the structural and dynamic properties of the entrapped biomolecules, we have encapsulated the protein hen egg white lysozyme (HEWL), a small globular, wellcharacterized enzyme (molecular mass 14.4 kDa), which is highly positively charged at neutral pH (isoelectric point ∼11) and therefore likely to be attracted to the silica surface of the sol-gel glass. HEWL contains six tryptophan residues at locations 28, 62, 63, 108, 111, and 123, two of which (Trp 62 and Trp 108), both partially exposed to the solvent, are responsible for most of its intrinsic fluorescence.30-32 In the present work, HEWL was entrapped in sol-gel matrices prepared from tetraethyl orthosilicate through an alcohol-free sol-gel route, and the photophysical properties of its fluorescent tryptophans were determined using steady-state and timeresolved fluorescence experiments. By combining the results for fluorescence spectra, quenching experiments, lifetimes, and time-resolved fluorescence anisotropy measurements, we have obtained information on the structure, dynamics, and average environment sensed by the entrapped protein. Our results indicate that, at neutral pH and low ionic strength, there is a qualitative change in the rotational behavior of HEWL entrapped in the pores as compared with that in the bulk aqueous solution, whereas in the latter, global rotation is free and unrestricted, and entrapment in the pores under these conditions completely inhibits the global rotation, though it does leave the segmental rotation essentially unaffected. The origin of the complete ablation of free global rotation of the protein when in the pores under these conditions is shown to be mainly electrostatic in nature since either reduction of the total positive charge carried by the protein (by increasing the pH toward its pI) or an increase in the ionic strength of the solution leads to complete removal of this constraint, though the global rotation ensuing is very much slower than that in the bulk solution. Materials and Methods Chemicals. HEWL (EC 3.2.1.17; 50200 U mg-1) and tetraethyl orthosilicate (TEOS) were purchased from SigmaAldrich Chemical Co. (Milwaukee, WI). Other chemicals were of analytical or spectroscopic reagent grade. HEWL was dissolved in a 10 mM, pH 7.4, sodium phosphate buffer prepared with deionized doubly distilled water. Other buffers were sodium phosphate buffer (10 mM, pH 7.4, 3 M NaCl) and borate buffer (10 mM, pH 9). The protein concentration was determined from UV absorbance at 280 nm using an absorption coefficient of ) 2.65 mg-1 mL cm-1.33 Immobilization of HEWL in Sol-Gel Monoliths. HEWL was encapsulated in pure silica matrices prepared through the alcohol-free sol-gel route described in Ferrer et al.5 (2002). Briefly, 4.46 mL of TEOS, 1.44 mL of H2O, and 0.04 mL HCl (0.62 M) were mixed with vigorous stirring at 22 °C in a closed vessel. After 1 h, 1 mL of the resulting sol was mixed with 1 mL of deionized water and submitted to rotaevaporation for a weight loss of 0.62 g (i.e., 0.62 g is approximately the alcohol mass resulting from alkoxyde hydrolysis). The aqueous sol was mixed with 1 mL of buffered HEWL (20 µM) in a disposable poly(methyl methacrylate) cuvette. Gelation occurred readily after mixing. Following gelation, monoliths were wet-aged in the phosphate buffer solution at 4 °C for 48 h. After aging, the monoliths, of size ∼9 × 9 × 19 mm, were removed from the
Pastor et al. disposable cuvettes and placed in 10 × 10 mm quartz cuvettes in the presence of 500 µL of the desired buffer. In most of experiments, this buffer was sodium phosphate buffer (10 mM, pH 7.4). For quenching experiments, the same buffer containing different concentrations of acrylamide was used, while for high ionic strength and alkaline pH experiments, the buffer was replaced by phosphate buffer, 10 mM (3 M NaCl, pH 7.4) and borate buffer (10 mM, pH 9), respectively, and was allowed to equilibrate with the monoliths for 48 h before measurements were made. Since the pKa of phosphate is known to change dramatically with ionic strength,34 the pH of the NaCl-containing buffer was always adjusted in the presence of the salt. Blank monoliths were prepared as above but with the protein solution replaced by the working buffer. Absorption and Steady-State Fluorescence Measurements. Absorption measurements were carried out at 22 °C using a Shimadzu UV-1603 spectrophotometer (Shimadzu, Tokyo, Japan). Fluorescence measurements were performed in an SLM8000C spectrofluorimeter (SLM Instruments Inc., Urbana, IL) fitted with Glan-Thompson polarizers. The experimental samples (sol-gel monoliths and protein aqueous solution) were placed in 10 × 10 mm path length quartz cuvettes. The steadystate anisotropy 〈r〉, defined by
〈r〉 )
IVV - GIVH IVV + 2GIVH
(1)
was obtained as a function of temperature by measuring the vertically (parallel) and horizontally (perpendicular) polarized components of the fluorescence emission with the excitation polarized vertically. The G factor (G ) IHV/IHH) corrects for bias in transmissivity between vertically and horizontally polarized components of the emission introduced by the detection system. Samples were excited at 300 nm with a bandwidth of 4 nm, and the polarized emission was detected at 340 nm with a bandwidth of 8 nm. Background intensities arising from the sol-gel matrix, which contribute less than 5% to the total signal, were always taken into account and subtracted from the measured sample intensities. Time-Resolved Fluorescence Measurements. The decay of the total fluorescence intensity, and those of the parallel and perpendicularly polarized components, were recorded at 20 °C in a single-photon timing system using an experimental setup similar to that described previously by Lillo et al.35 In brief, the tryptophan fluorophores of HEWL were excited by means of vertically polarized light pulses from a Ti:sapphire picosecond laser (Tsunami, Spectra Physics) pumped with a 5W Nd:YVO4 laser (Millennia, Spectra Physics) and associated with a third harmonic generator tuned to 297 nm. Pulses of 1-2 µs width were generated at a repetition rate of 4 MHz, giving ∼20 µW of average power at the samples. The experimental samples (sol-gel monoliths and aqueous protein solution) were placed in 10 × 10 mm path length cuvettes. Data were stored in 4K channels, at a resolution of 11 or 6.1 ps/channel, up to ∼5 × 106 total counts. The decay of the total fluorescence intensity Im(t) was recorded at 340 nm with the emission polarizer set at the magic angle (54.7°) relative to the vertically polarized excitation beam. The parallel and perpendicular components of emission were sequentially recorded at 340 nm by alternating the orientation of the emission polarizer every 2 min. The kinetic parameters of the impulse response fluorescence intensity decay, im(t) (lifetimes τi and normalized amplitudes Ri), were determined by reconvolution and fitting, using nonlinear least-squares regression methods. The amplitude-
Lysozyme Encapsulated in a Silica Sol-Gel Matrix
J. Phys. Chem. B, Vol. 111, No. 39, 2007 11605
weighted lifetime τj, proportional to the quantum yield and defined by36
τj )
∑ Riτi
(2)
and the average fluorescence lifetime defined by36
)
2 ∫0∞ t × i(t)dt ∑i Riτi
∫0∞ i(t)dt
)
∑i
(3)
Riτi
were calculated. The anisotropy decay function, r(t), was determined by simultaneous reconvolution and fitting to the two experimentally determined polarized components of the fluorescence intensity, using nonlinear least-squares global analysis implemented by GLOBALS Unlimited (Urbana, IL). In this case, the analysis consisted of finding the numerical parameters for r(t) that best fit the two polarized decay functions, iVV(t) ) [im(t)/3][1 + 2r(t)] and iVH(t) ) [1/G][im(t)/3][1 - r(t)], to the experimental traces IVV(t) and GIVH(t) upon introducing the lifetime obtained from the magic angle data as a fixed parameter. The general expression used for the anisotropy decay fitting was a sum of n exponentials and a constant term37 n
βi exp(-t/φi)] + r∞ ∑ i)1
r(t) ) (r(0) - r∞)[
(4)
where n
∑β )1 i
i)1
and φi are the rotational correlation times, βi are normalized amplitudes, and r∞ is the residual anisotropy, containing information about the restriction of the depolarizing processes. To perform this analysis, the first channels of the experimental decays were not included in the fit in order to avoid different scattering contributions from the sol-gel monolith among different samples. The fits tabulated here both for fluorescence intensity and anisotropy decay represent the minimum set of adjustable parameters which satisfy the usual statistical criteria, namely, a reduced χ2 value of
