Article pubs.acs.org/Langmuir
Real-Time Analysis of Porphyrin J‑Aggregation on a Plant-Esterase-Functionalized Surface Using Quartz Crystal Microbalance with Dissipation Monitoring Limin Yang,† Lei Jiang,*,† Weijing Yao,‡,§ Junling Liu,† and Juan Han† †
State Key Laboratory of Heavy Oil Processing and Center for Bioengineering and Biotechnology, China University of Petroleum (East China), Qingdao, Shandong 266555, P. R. China ‡ Evidence Identification Center, Public Security Bureau, Chongqing 400021, P. R. China § Engineering Research Center for Criminal Investigation, Chongqing 400021, P. R. China S Supporting Information *
ABSTRACT: The J-aggregation of meso-tetra (4-sulfonatophenyl) porphine (TPPS4) on a plant-esterase-functionalized surface in a 1:1 v/v mixture of 0.05 M HCl/ethanol (pH ∼1.38) was analyzed in real time using a quartz crystal microbalance with dissipation monitoring (QCM-D). Simultaneous changes in frequency (Δf) and energy dissipation (ΔD) correlated well with mass and structural changes during the sequential phases of slow nucleation, rapid aggregation, and equilibration in J-aggregation. The time-dependent mass adsorption could be quantitatively analyzed with a model, which integrated two simple equations obtained when the surface concentration of TPPS4 (ΓTPPS4) was below and above the critical aggregation surface concentration (CASC). This study provides a new view for the proteininduced J-aggregation process, which may be helpful for understanding the interactions of self-assembled nanostructures with biomolecules.
■
and nonlinear optical properties.19−21 Spectroscopic features and excitonic interactions in TPPS4 J-aggregates have been interpreted to a limited extent, for example the sharp and red-shifted absorption band (J-band)22 and nonextended exciton coupling between chromophores.23 Protein scaffolds are essential for the biological activities of biological pigments.24 In the natural light-harvesting systems, the pigments are generally noncovalently bound to proteins, forming the pigment−protein complexes.25 This binding ensures the energy transfer efficiency by regulating both the orientation and distance among pigments.26 Study of J-aggregates interactions with proteins can provide direct information on their physiological properties and on the relationship between their structures and biological activities. Although several proteins, such as human serum albumin (HSA),27 bovine serum albumin (BSA),7,28 lysozyme,28 and myoglobin,29 have been reported to promote J-aggregation of TPPS4 in acidic medium, knowledge of the mechanisms of protein-induced J-aggregation is still very limited. Based on the distinct spectral features of the monomeric and aggregated species, UV/vis absorption and fluorescence spectroscopies have been used to study the kinetics of the aggregation process induced by proteins in solution. Andrade and
INTRODUCTION Molecular J-aggregates, well-known for their unique optical properties and biological and technological significance, have been extensively studied since first described by Jelley and Scheibe.1 Porphyrin aggregates have also attracted much attention because they are considered as a good model system for studying the energy transfer in the light-harvesting antennas in photosynthetic organisms.2,3 Porphyrin aggregates have been classified as H- or J-type, defined by the relative orientations of induced transition dipoles of the constituent molecules, either “face-to-face” or “head-to-tail”, respectively.4 The natural chlorophyll aggregates in the light-harvesting proteins or chlorosomes have a strong transition dipole moment aligned in the “head-totail” direction.5 Therefore, porphyrin J-aggregates are particularly interesting as model systems for the study of photosynthesis and biological pigment systems. Water-soluble synthetic porphyrins are expected to simulate the functions of native porphyrins in the physiological environment and their structural simplicity makes the interpretation of the structure−function relationship easier.6 Mesotetra (4-sulfonatophenyl) porphine (TPPS4) is one of the most commonly studied water-soluble porphyrins. It tends to selfassemble into stable J-aggregates at low pH and/or high ionic strength via electrostatic interactions between positively charged macrocycle inner core and negatively charged peripheral sulfonate groups of adjacent molecules.7 Most studies on TPPS4 J-aggregates have focused on their formation with and without additives or templates,8−15 or on their unique ordered structures16−18 © XXXX American Chemical Society
k
Costa30 proposed a simple kinetic model based on the X → Y reaction to fit the aggregation process. Both J-aggregate formation Received: February 14, 2014 Revised: July 11, 2014
A
dx.doi.org/10.1021/la501986e | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
and monomer transformation fitted the equation well. However, no allowance was made for the initial induction period, although the kinetics for porphyrin aggregation should begin with an induction period and it may be the rate-limiting step.31,32 The kinetic model is therefore incomplete and the actual aggregation process remains not well understood. Valanciunaite and coworkers7 also presented a hypothetical model demonstrating the different stages of TPPS4 interaction with BSA in acid. Unfortunately, the analysis of the early events in the aggregation process was not mentioned. In addition, these spectrophotometric techniques are “static” in that they study the system at a fixed time and rarely provide any information about porphyrin aggregation dynamics. The lack of in situ methods to monitor the aggregation kinetics hinders further development of proteininduced J-aggregates. A final problem is that the aggregation of TPPS4 has been studied in solution, in which porphyrins sometimes aggregate as ordered precipitates.33 Quartz crystal microbalance with dissipation monitoring (QCM-D) technique is ideally suited for in situ characterization of self-assembly because it is a label-free, in-line technique that directly reports the mass and viscoelastic properties of aggregates as they grow on the surface in real time.34 QCM-D is known to be a promising tool for kinetic evaluations.35−38 Interfacial adsorbates behave in different manners at solid/ liquid interfaces from those in bulk phases. It is therefore interesting to study the in situ fabrication of nanostructured TPPS4 J-aggregate on protein surfaces. In this study, a plantesterase-functionalized QCM-D surface was prepared and employed to monitor the in situ growth of TPPS4 J-aggregates. Plant-esterase (EC 3.1.1.X) was used as the model protein since it is widely distributed in plants and plays key roles in activation of signal molecule, regulation of the bioactivity of endogenous products, and other biological process.39 It also meets the requirement of being positively charged at low pH (< pI = 4.3−4.6).40
■
the nonadsorbed material, and blow-dried in a stream of nitrogen. The cystamine-coated crystal was subsequently immersed in a 25% (v/v) aqueous glutaraldehyde solution for 1 h to change amine groups to aldehyde on the gold surface. After being rinsed with water and dried, the crystal was mounted in the flow chamber and exposed to the plantesterase solution for 1 h in flow mode at a flow rate of 50 μL min−1. All the signal frequency and dissipation variations during plant-esterase immobilization process were monitored by QCM-D E1 equipment (Q-SENSE, Västra Frölunda, Sweden) (see SI Figure S3). Finally, the plant-esterase-functionalized QCM-D surface was prepared. All these procedures were performed at 37 °C. It is worth noting that the polymerization of glutaraldehyde might occur in this process, considering the high concentration and temperature used.43,44 As a result, the glutaraldehyde layer formed might be acoustically rigid.45 Protein-Induced TPPS4 Aggregation Monitored with QCM-D. The assembled plant-esterase−QCM-D chip was mounted in the thermostatic flow chamber at (25 ± 0.5) °C. Before measurements, an HCl− ethanol mixed solution (0.05 M HCl in 50% v/v ethanol, pH 1.38) was employed in the flow until the baseline stabilized to ±1 Hz. The quartz crystal was then exposed to the TPPS4 solution at concentrations of 0.01, 0.08, 0.12, 0.16, 0.20, 0.28, 0.36, and 0.44 mM. Measurements were performed in a flow−static−flow mode. Each measuring cycle consisted of 3 min sample injection (flow on), 60 min interaction step (flow off), and 20 min washing step (flow on). Water was used in the washing step to eliminate all the porphyrin bound to the enzyme surface. Changes in resonance frequency (Δf) and energy dissipation (ΔD) as a function of time were obtained at a step time of 0.36 s. Time evolutions were obtained by plotting values Δf (hertz) or values ΔD (10−6) versus time (minutes). Analysis and Modeling of QCM-D Data. Values of Δf and ΔD were recorded at the third, fifth, seventh, and ninth harmonics as a function of time. The original data were processed in Q-tools software (Q-SENSE) before being exported for further analysis in OriginPro 8 (OriginLab). The Voigt-based viscoelastic model was applied to Δf and ΔD data in order to extract information on the adsorbed mass and viscoelastic properties of the aggregate layer using Q-tools. The input parameters, including the viscosity and density of the solvent and the density of the film, were fixed to 0.001 kg m−1s−1, 1000 kg m−3, and 1000 kg m−3, respectively. The boundary conditions used for all the analyses are shown in SI Table S1. The best fit between the Voigt model and experimental data for all overtones was obtained considering minimum value of χ2 parameter. Surface Characterization with Spectroscopic and Microscopic Methods. Spectroscopic Methods. For spectroscopic investigations, plant-esterase was immobilized on the surface of an optically transparent quartz slide (20 × 10 × 4 mm) instead of a QCM-D gold chip following almost the same procedures except that cystamine was replaced by APTES (see SI Figure S2b). The quartz slides were first treated in piranha solution for 15 min, followed by extensive rinsing in water and drying in nitrogen. The cleaned and oxidized slides were then silanized with a 10% (v/v) solution of APTES in toluene overnight at room temperature, followed by rinsing in toluene to remove noncovalently adsorbed silane compounds and drying in an oven at 110 °C for 1 h to achieve complete formation of Si−O−Si bonds via condensation. As a result, reactive amino groups were introduced at the slide surface just as in the cystamine treatment. A glutaraldehyde linker was then coupled to the amino-silanized slide surface by covering it with 25% (v/v) aqueous glutaraldehyde solution for 1 h at 37 °C. Enzyme coupling was achieved through Schiff base formation by incubating the activated slides in plant-esterase solution for 1 h at 37 °C. After being thoroughly washed with water and dried, the plantesterase-functionalized quartz surface was prepared. Before measurements, the slides with plant-esterase surface were reacted with the 0.44 mM pH 1.38 TPPS4 solution for 1 h at 25 °C, and then thoroughly washed with the HCl−ethanol mixed solvent to remove any physically adsorbed material. Finally, the slides were placed inside a quartz cuvette (optical path 10 mm) filled with the HCl−ethanol mixed solvent. The porphyrin aggregates already formed stably existed in the solvent. In particular, the slides were maintained in vertical position so that the beam of light could pass through the porphyrin aggregates
EXPERIMENTAL SECTION
Reagents and Materials. Plant-esterase was extracted from wheat flour and further purified to homogeneity as described previously.40 The enzyme preparation was stored frozen at −20 °C in small aliquots. Meso-tetra (4-sulfonatophenyl) porphine (TPPS4) was obtained as tetrasodium salt from Sigma-Aldrich. Aqueous stock solutions of TPPS4 (2 mM) were prepared in water before use. The porphyrin solutions for kinetic analysis were prepared according to the “porphyrin first (PF)” mixing order protocol reported by Romeo et al.41 The proper amount of porphyrin stock was first diluted with water and mixed with ethanol. Then, concentrated hydrochloride acid (HCl) solution (1 M) was added. The final concentration of ethanol and HCl was 50% (v/v) and 0.05 M, respectively. The pH value was around 1.38. A series of kinetic experiments was conducted over a porphyrin concentration range of 0.01−0.44 mM. All the porphyrin solutions were prepared immediately before injection and the molecules exist in the protonated form (see Supporting Information (SI) Figure S1). Cystamine, glutaraldehyde (grade I, 50% aqueous solution), and 3-aminopropyltriethoxysilane (APTES) were obtained from Sigma-Aldrich and used as received. Glutaraldehyde was stored at −20 °C to ensure the lowest possible rate of polymerization.42,43 Deionized water from a Millipore Direct-Q Water system (resistivity, 18.2 MΩ cm) was used in all the aqueous solutions. QCM-D Measurements. Preparation of Protein-Functionalized QCM-D Surface. Plant-esterase was immobilized on the gold surface of QSX 301 AT-cut quartz crystals (Q-SENSE, Västra Frölunda, Sweden) using the cystamine−glutaraldehyde method (see SI Figure S2a). Prior to the experiments, the gold surface was cleaned in piranha solution (three portions of concentrated H2SO4 to one portion 30% H2O2 in volume) for 15 min. The quartz crystal was first incubated in an aqueous solution of cystamine for 1 h to deposit a self-assembled monolayer of 2-aminoethanethiolate, then thoroughly rinsed in water to remove all B
dx.doi.org/10.1021/la501986e | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
surface vertically as shown in SI Figure S4. The cuvette was then immediately placed into the spectrometer and the spectroscopic characteristics were measured. The UV/vis absorption spectra were collected on a UV-2450 spectrophotometer (Shimadzu, Kyoto, Japan). Wavelength scans were recorded from 350 to 750 nm using a wavelength step size of 0.5 nm. Fluorescence spectra were acquired using an F-2500 spectrophotometer (Hitachi, Tokyo, Japan). Scans were performed at room temperature using excitation wavelengths of 434 and 490 nm, respectively. Emission spectra were collected in the range of 600−800 nm using increments of 1 nm. Resonance light scattering (RLS) measurements were also conducted on the F-2500 spectrophotometer in the wavelength region 350−750 nm, adopting for RLS experiments a synchronous scan protocol with a right angle geometry.46 Both the excitation and emission slit widths were kept at 10 nm. Circular dichroism (CD) spectral measurements were carried out on a MOS-450/AF-CD spectropolarimeter (Bio-Logic, Claix, France) under constant nitrogen flow. Wavelength scans were performed in the range of 380−600 nm using a wavelength step size of 1.0 nm. Atomic Force Microscope (AFM) Imaging. AFM measurements were performed in liquid using a Nanoscope VIII-MultiMode (Bruker, CA). Before use, the AFM fluid cell was washed by sonication in successive baths of ethanol and water. The V-shaped silicon nitride cantilevers with a spring constant of 0.06 N/m (DNP-S10, Bruker) were cleaned with a PDC-32G plasma cleaner (Harrick, NY) before mounting in the AFM cell. After incubation of TPPS4 in the QCM-D chamber, the chip was installed on the AFM scanner. Images were collected in contact mode. Images presented were height images and zero-order flattened using a standard algorithm within the Nanoscope Analysis software.
frequency, which corresponds to the increase in mass density over the adsorption of TPPS4.15 However, Δf increases with increasing overtone, and ΔD is significantly high. Thus, Δf can not only be related to mass changes, but also to changes in the viscoelastic properties.47 The significance of this viscoelastic change will be discussed later. The frequency dropped immediately after introducing TPPS4, showing that TPPS4 molecules were adsorbed to the plantesterase surface. The time evolution of adsorption was complex. At first, the frequency slowly decreased to about −84 Hz over a period of 3 min. Then there was a sharp decrease in frequency from −84 to −1162 Hz before the frequency leveled off. The overall profile therefore showed a sigmoid curve, which could be divided into three phases: (i) an initial “induction” period, followed by (ii) rapid adsorption up to 80%−85%, and (iii) a slow approach to the equilibrium state. The induction period is different from the previous reports.7,30 To further investigate the TPPS4 adsorption process, different concentrations of TPPS4 monomers were incubated on the plant-esterase surface. Figure 2 shows the frequency responses of the plant-esterase surface to different concentrations of TPPS4 ranging from 0.01 to 0.44 mM at the third overtone (responses at three other overtones are available in SI Figure S5). It is observed that the frequency shift is related to the concentration of injected TPPS4. Higher bulk solution concentrations of TPPS4 lead to more adsorption, as expected. What is interesting is that lower concentrations can slow down the adsorption of TPPS4, and this allows more detailed kinetics analysis. First, small concentrations of TPPS4 (0.01−0.08 mM) caused only a slow variation in frequency, as shown in Figure 2a. The frequency decreased slowly right up to the equilibrium and there was no lag phase or steep decrease. At intermediate concentrations of TPPS4 (0.12−0.20 mM), two kinds of changes took place consecutively. The frequency decreased slowly for the first 4 min (phase i) and then a sharp decrease occurred and there was no equilibration, i.e. phase ii, as illustrated in Figure 2b. At the highest TPPS4 concentrations (0.28−0.44 mM), all three kinetic phases were observed (Figure 2c). The adsorption behavior of TPPS4 on the plant-esterase surface in neutral medium was also investigated as a negative control. As shown in the inset in SI Figure S6, the frequency was changed a little at pH 7.0, and the maximum Δf value was only about −13 Hz. However, at pH 1.38, the adsorption of TPPS4 on the plant-esterase surface showed a comparatively larger frequency shift (Δf = −1740 Hz), and therefore larger overall adsorbed mass, compared with the adsorption in neutral medium. Viscoelasticity of the Adsorbed TPPS 4 Layer. The dissipation measured by QCM-D gives a measure of the damping of the oscillation, which is related to the viscoelastic properties of the adsorbed layer.48 Figure 1 shows that the loss of frequency is accompanied by an increase in ΔD, indicating the formation of a viscoelastic TPPS4 layer.49 The mechanical properties of the TPPS4 adsorption layers can be analyzed by examining both the f and D data using Voigt-based viscoelastic model, which allows for the response to be translated into three effective physical parameters: mass (mVoigt = thickness (dVoigt) × density (ρeffective)), shear viscosity (ηVoigt), and shear elasticity (μVoigt).36 As shown in SI Figure S7, this model fits the measurements well (deviation of CASC n ⎩ k + t n′
(1)
where (mVoigt)t is the effective mass obtained through Voigt model at time t, and (mVoigt)max is the maximum binding value (corresponding to the minimum frequency measured, or plateau value, as seen in Figure 1). The relaxation time (τ), the reciprocal of the porphyrin binding rate, can be calculated from curve fittings of the adsorbed mass increase during the nucleation. SI Table S2 lists the kinetic parameters for nucleation obtained after fitting the experimental values presented in phase i. As expected, the kinetic parameters indicate that the larger the adsorbed mass, the faster the binding rates. When ΓTPPS4 is above the CASC, two kinds of interactions take place consecutively. First, the critical aggregation nucleus is formed through the protein-induced porphyrin binding (phase i). Then, the nucleus already formed acts as a catalyst, catalyzing the aggregation of more and more porphyrin monomers (phase ii). At the turning point in the frequency plot, there is a sharply decreased oscillation frequency (Figure 2b). This rapid growth step (phase ii) is illustrated in Figure 7. H4TPPS42− molecules stack in a J-aggregation manner on the plant-esterase surface by electrostatic forces and hydrophobic interactions. Increase of the concentration of TPPS4 leads to the frequency variation becoming smaller and smaller until a plateau is reached (Figure 2c). This indicates that there is no further change in the aggregate, and that the equilibrium in H4TPPS42− aggregation has been reached (Figure 7, phase iii). Cooperativity is often considered as one of the features of the noncovalent biological self-assembly.66 This effect takes place when the binding of one ligand influences the binding strength of a macromolecule toward a subsequent ligand. As the model system of natural chlorophyll aggregates, porphyrin J-aggregates should display cooperativity. Furthermore, cooperativity is also
(3)
The surface concentration ΓTPPS4 was estimated to be around 4031.3 and 5085.6 ng cm −2 when the bulk solution concentration of TPPS4 was 0.08 and 0.12 mM, respectively. Thus, the CASC is probably within the range of 4000.3− 5141.2 ng cm−2. However, it is worth noting that the value may be overestimated due to the fact that the mass measured by QCM-D includes the extra mass of water coupled in the TPPS4 layer.70 In neutral aqueous solution, the global charge of the enzyme is negative and the two pyrrolic nitrogen atoms in porphyrin macrocycle remain unprotonated. The aggregation will be hindered by the electrostatic anionic repulsion.17 Consequently, only a small amount of porphyrins were bound to the enzyme surface via hydrophobic interactions, resulting in a small change in frequency (SI Figure S5). Structural Changes in TPPS4 Aggregate Layer. Besides the changes in mass, the dynamic structural changes in aggregation process are also discussed. In the early nucleation (phase i), a large quantity of H4TPPS42− monomers are attracted to the plant-esterase surface, inducing a sharp increase of the effective surface concentration of the TPPS4 in the layer, which are associated with steep increases in elasticity (Figure 3b) and thickness (Figure 3c). At the same time, some loosely bound H4TPPS42− molecules desorb from the surface. As a result, the elasticity, as well as thickness, increases slowly in the later nucleation until small aggregate nuclei are formed. The formation of J-aggregates can be considered to be a noncovalent polymerization process.71 For polymerization process, it is expected that the viscosity of the film should be similar to that of the solvent in the beginning. G
dx.doi.org/10.1021/la501986e | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 8. Schematic illustration of structure changes in the three different kinetic phases: (i) nucleation, (ii) aggregation, and (iii) equilibrium.
in Figure 4b. There might also be a critical surface concentration that can trigger distinct conformational transition, as indicated by the sigmoid curves in Figure 4. This critical surface concentration probably exists in the inflection point of the elasticity curve. The TPPS4 layer shows a flexible conformation of the aggregated TPPS4 molecules with different extent of water coupled below the critical surface concentration, and shows a compact and dehydrated conformation above the critical surface concentration.
Indeed, it is found that the effective viscosity of the TPPS4 aggregate layer is
