Modulation of Plasmon-Enhanced Resonance Energy Transfer to

Sep 21, 2015 - Qi-Qi Gai , Dong-Mei Wang , Rong-Fu Huang , Xia-Xia Liang , Hong-Lin ... aptasensor for ultrasensitive detection of aflatoxin B 1 ... g...
6 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCB

Modulation of Plasmon-Enhanced Resonance Energy Transfer to Gold Nanoparticles by Protein Survivin Channeled-Shell Gating Magdalena Stobiecka*,† and Agata Chalupa‡ †

Department of Biophysics, Warsaw University of Life Sciences (SGGW), 02776 Warsaw, Poland Institute of Nanoparticle Nanocarriers, 11010 Barczewo, Poland

‡

S Supporting Information *

ABSTRACT: The resonance energy transfer (RET) from excited fluorescent probe molecules to plasmonic gold nanoparticles (AuNPs) can be gated by modulating the width of channels (gates) in submonolayer protein shells surrounding AuNPs. We have explored the gated-RET (gRET) processes using an antiapoptotic protein survivin (Sur) as the gating material, citrate-capped gold nanoparticles (AuNP@Cit), and fluorescein isothiocyanate as the fluorescent probe. Despite the electrostatic repulsive forces between these components, a strong modulation of RET efficiency by Sur down to 240 pM (S/N = 3) is possible. Using piezometric measurements, we have confirmed the Sur adsorbability on Cit-coated Au surfaces with monolayer coverage: γSur = 5.4 pmol/cm2 and Langmuirian adsorption constant KL,Sur = 1.09 × 109 M−1. The AuNP@Cit/Sur stability has been corroborated using resonance elastic light scattering. The quantum mechanical calculations indicate that multiple hydrogen bonding between Cit ligands and −NH3+, =NH2+, and −NH2 groups of lysines and arginines of Sur have likely facilitated Sur bonding to nanoparticles. A theoretical model of gated-RET has been developed, enabling predictions of the system behavior. In contrast to the positive slope of the Stern−Volmer quenching dependence (F0/F) = f(QA), a negative slope has been obtained for gRET relationship (F0/F) = f(cP), attributed to the dequenching.

■

that the plasmon interacting with a fluorophore becomes an integral part of the excitation process. In addition to that, the FRET distance dependence is modified in plasmonic fields,18,34,35 leading to the nanomaterial surface energy transfer (NSET) which extends the dipole−dipole interactions to include multipoles. We have found that the resonance energy transfer (RET) from a fluorescent probe to a citrate-capped plasmonic nanoparticle (AuNP@Cit) can be modulated by a submonolayer film of cytochrome c (Cyt c) surrounding [email protected] The modulation of RET processes by a gating dielectric film has been further explored in the present work to identify basic parameters influencing the RET efficiency and gating sensitivity that would enable predictions of the system behavior. In the experiments performed, AuNP@Cit have been used with fluorescein isothiocyanate (FITC) as the fluorescent probe and the protein survivin as the gating material. The protein survivin (Sur) is an inhibitor of apoptosis which plays a crucial role in the regulation of cell division and cell cycle control. Sur is strongly expressed in embryos and malignant tumors but not in the fully differentiated cells.37−39 For this reason, Sur has been studied extensively as a cancer biomarker for therapeutic40 and diagnostic purposes.41 The use

INTRODUCTION Fluorescence resonance energy transfer (FRET) in the absence of strong electromagnetic fields has been extensively utilized in studies of biomolecules and their reactivity.1−5 In particular, FRET measurements have enabled evaluating biomolecule conformation and intramolecular distances based on Förster theory,6 derived for dipole−dipole interactions. However, recent studies indicate that FRET and optical transitions in biomolecules are strongly influenced by plasmonic fields emanating from colloidal gold and silver resulting in plasmonic field-tunability of absorption spectra of chromophores,7−10 plasmonic fluorescence enhancement,11−18 strong fluorescence quenching by gold and silver nanoparticles (AuNP, AgNP) due to plasmon-enhanced FRET,19 formation of “hot spots”,20−24 plasmon-enhanced Raman scattering,25 hot electron injection processes,26−28 catalytic enhancements,29 and other effects. Recently, Lakowicz and co-workers22 have found that Ag and Au islands on glass increased the fluorescent emission of a solution fluorophore up to ∼7 fold and in the hotspots up to ∼50 fold, while the gold nanorod created hotspots-generated fluorescence enhancement of ∼22 fold for upconversion excitation.30 Extensive studies conducted in this field have led to the proposed metal-enhanced fluorescence (MEF)11,12,31 which enables new insights into the nature of near-field induced electronic transitions in molecules, including the increased absorption and fast emission rate. It has been proposed22,25,32,33 © 2015 American Chemical Society

Received: August 10, 2015 Revised: September 18, 2015 Published: September 21, 2015 13227

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235

Article

The Journal of Physical Chemistry B

ence of gated-RET efficiency on specific component properties has been one of the goals of this work. The model mechanism proposed in this work may serve to determine adsorption constants for proteins on functionalized plasmonic metal NPs, as well as to design interface properties to optimize binding characteristics for nanoparticle nanocarriers42 for theranostic applications.

of oligonucleotide molecular beacons encoded for recognition of Sur mRNA has become one of the leading cancer diagnostic tools. Unlike Cyt c, Sur is negatively charged at neutral pH (pI = 5.66) and its interactions with AuNP@Cit (also negatively charged) have not been investigated. Despite the expected electrostatic repulsions, our preliminary measurements indicated that Sur may adsorb on AuNP@Cit. Therefore, in this work, using a quartz crystal nanogravimetry (EQCN) technique, adsorbability of Sur on a citrate-coated Au surface has been corroborated and then utilized in Sur-modulation testing of gated-RET. The operational principle of the gated-RET is presented in Figure 1. A fluorescent dye probe, labeled “dye” in panel A,

■

MATERIALS AND METHODS Chemicals. Human survivin full length protein (Apoptosis Inhibitor Protein 4, AIP4), was obtained as a recombinant protein using Escherichia coli. It was tagged with calmodulin (CaM), 17 kDa protein, purchased from Abcam (Cambridge, UK). Survivin was purified by chromatography and tested by Western blotting (purity > 90% using 15% SDS-PAGE, 33 kDa). It was stored in a solution of 0.1 M NaCl, 20 mM Tris, pH 7.5. Fluorescein isothiocyanate and other agents were purchased from Sigma-Aldrich and used as received. All chemicals used for investigations were of analytical grade purity. Aqueous solutions were prepared with freshly deionized water with 18.2 MΩ cm resistivity (Hydrolab, Wiślina, Poland). The RET gating measurements were performed using a 50 mM Na3Cit solution with pH 7.45. The stock solution of FITC (20 μM) was prepared in dimethyl sulfoxide ((CH3)2SO, DMSO). All concentrations of added reagents cited in this paper are final concentrations obtained after mixing. Curve fitting was performed using the Simplex algorithm. Instrumentation. The fluorescence spectra were recorded using Spectrometer model LS55 (PerkinElmer, Waltham, MA), with 20 kW Xenon light source and a photomultiplier tube detector. The excitation and emission slit widths were set to 7.5 nm. The excitation and emission wavelengths were set to λex = 495 nm and λem = 517 nm, respectively. The nanogravimetric measurements were performed using an Electrochemical Quartz Crystal Nanobalance, Model EQCN-940 (Elchema, US), and a Data Logger and Experiment Control System, Model DAQ-716v, operating under Voltscan 5.0 data acquisition and processing software. The resonant frequency, f 0, of the Au-piezoelectrodes with 5 mm diameter Au disk was 9.975 MHz. The interfacial mass changes were determined from the changes in oscillation frequency of the EQCN according to the Sauerbrey relationship fulfilled for thin rigid films.43−45 For quartz crystal resonators with 10 MHz fundamental resonance frequency, the relationship between film mass change, Δm, and the measured fundamental frequency shift, Δf, is given by Δm = −0.8673 Δf, where Δm is in nanograms and Δf is in hertz.46 The transmission electron microscopy (TEM) images were recorded in the Electron Microscopy Platform of the Mossakowski Medical Research Centre of the Polish Academy of Sciences in Warsaw, Poland. Procedures. The citrate-capped gold nanoparticles (AuNP@Cit) were synthesized using the borohydride-citrate method as reported earlier.47 Briefly, tetrachloroauric acid (HAuCl4) solution (2.56 mL, 10 mM) was mixed with trisodium citrate (Na3C6H5O7) solution (9.6 mL, 10 mM) and poured into distilled water (88 mL). Next, sodium borohydride (NaBH4) solution (8.9 mL, 5 mM) was added dropwise. The solution slowly turned light gray and then ruby red. Stirring was maintained for 30 min. The obtained citratecapped gold nanoparticles were stored at 4 °C. The concentrations of gold nanoparticle solutions were determined from exact amounts of reagents used in synthesis. The size of AuNP determined by TEM image analysis was 4.9 ± 0.1 nm.

Figure 1. Principle of gated fluorescence resonance energy transfer (RET). (A) Plasmon-enhanced RET due to the gold nanoparticle (AuNP) plasmonic field leads to a low fluorescence. (B) Hindered RET caused by an adsorbed protein shell, fully covering AuNPs (θ = 1), results in high fluorescence. (C) Gated RET (gRET) in which a protein film on a AuNP (with partial coverage, θ < 1) is gating the plasmon-enhanced RET and makes the system very sensitive to protein concentration.

undergoes a plasmon-enhanced RET when in close proximity to a AuNP@Cit. This means that emission of an excited probe is efficiently quenched by AuNP@Cit. However, when AuNP@ Cit is surrounded by an impenetrable shell of protein molecules (panel B), with surface coverage θ = 1, the approach of the probe to AuNP@Cit surface is blocked. Therefore, the efficiency of RET decreases significantly because of the large distance and low probability of protein-mediated energy migration. There may be still some quenching of the probe by the protein itself, but usually the RET efficiency is low. In the intermediate case (panel C), when protein coverage is not full, θ < 1, there is some quenching of the excited probes able to enter the channels (voids) in the protein film and interact directly with plasmonic fields of AuNP. Therefore, by modulating the width of these channels, the RET efficiency can be modulated. The high sensitivity of the gated-RET to low protein levels in the solution seems to arise from the ease of protein selfassembled monolayer (SAM) formation, the large thickness of protein shell (due to the size of protein molecules), and the intermolecular channel width modulation in protein SAM. The identification of physical parameters influencing the depend13228

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235

Article

The Journal of Physical Chemistry B The inner-filter effect in fluorescence measurements due to the high absorption of AuNPs was corrected when necessary according to the procedure published earlier.19 For the Stern− Volmer constant determination, the initial slope of the experimental curve was utilized instead of the entire curve which shows deviations from linearity at higher AuNP concentrations. By using the initial slope, the measurements are extrapolated to [AuNP] = 0. Quantum mechanical calculations of electronic structures for the interactions of model amino acids, lysine and arginine, with citrate molecules of AuNP@Cit shell were performed using modified Hartree− Fock methods, semiempirical PM3 method, and density functional theory (DFT) method at B3LYP functional level with 6-31G* basis set and pseudopotentials. The molecular dynamics simulations and quantum mechanical calculations were carried out using procedures embedded in Wavefunction (Irvine, CA) Spartan ’14 software.

Δf = −

2Δmnf0 2 A μq ρq

(1)

where f 0 is the fundamental-mode oscillation frequency of unperturbed QC resonator, n the overtone number, A the piezoactive surface area of the resonator, μq the shear modulus of quartz (μq = 2.947 × 1011 g cm−1 s−2), and ρq the quartz density (ρq = 2.648 g cm−3). The total mass of the citrate SAM was Δm = 22.3 ng (Δf = 25.7 Hz), corresponding to the monolayer mass mmono,Cit = 87.4 ng/cm2 and surface coverage γCit = 4.55 × 10−10 mol/cm2. This value is in good agreement with the surface excess of citrate species, γCit = 4.6 × 10−10 mol/cm2, determined for Au(111) from double-layer capacitance measurements.48 During the time transient recorded after the injection of Sur (Figure 2), the total mass change Δm = 45.8 ng (Δf = 52.8 Hz) was observed, corresponding to the monolayer mass mmono,Sur = 179.6 ng/cm2 and surface coverage γSur = 5.4 pmol/cm2. This surface excess γSur is equivalent to the area of 30.5 nm2 per adsorbed Sur molecule. Based on crystallographic data of Chantalat and coworkers,49 the dimensions of one Sur molecule are 5.475 × 4.36 × 5.63 nm3. Therefore, the Sur cross-sectional areas for horizontal, side-on, and vertical orientations are 30.8, 23.9, and 24.5 nm2, respectively. The experimental value of 30.5 nm2 is close to the theoretical cross section of the Sur molecule for horizontal orientation. Survivin-Gated RET. Modulating the probe fluorescence by adsorptive protein gating of excited-state energy transfer, previously demonstrated for positively charged Cyt c protein, has been further explored in this work by employing Sur protein molecules to form gating submonolayer films on AuNPs. In contrast to Cyt c, the protein Sur is charged negatively. At the same time, the probe is also charged negatively and the nanoparticles are negatively charged as well. Building on the results of piezogravimetric measurement, described in a previous section and which clearly indicated adsorbability of Sur on a citrate-coated Au surface, we have investigated the gating properties of Sur in a gRET system: AuNP@Cit/Sur−FITC. The solution-soluble fluorophore probe (FITC) exhibits strong fluorescence (λex = 495 nm, λem = 517 nm) which is quenched by AuNPs. The excellent quenching of FITC fluorescence by AuNPs is due to the surface plasmon absorption band (with maximum at λmax = 516 nm) which perfectly overlaps with the FITC emission spectrum.36 Any approach of a FITC molecule to a AuNP at a distance closer than ca. 10 nm will result, to a high probability, in an NSET from the excited-FITC donor to the AuNP’s surface plasmon as the acceptor. In the gated-RET, the quenching process is tamed by inserting a protein film around AuNP so that a direct approach of the dye probe to the surface of AuNP is inhibited. At sufficiently low added Sur concentration, only a submonolayer film of Sur is formed, enabling the dye probes to pass through the channels between Sur molecules and transfer their excited-state energy to AuNP. Note that after Sur adsorption, the gates are not modulated any further: they are static. Figure 3 illustrates all these situations for FITC probe, Sur, and the citrate-capped AuNPs (AuNP@Cit). The TEM image of AuNPs is presented in panel A, and their Gaussian size distribution is shown in panel B (nanoparticle mean diameter, d = 4.9 ± 0.2 nm). Panel C shows the normalized absorbance spectrum of spherical gold nanoparticles (AuNP@Cit, 5 nm dia.) and the fluorescence spectrum of a FITC dye. FITC

■

RESULTS AND DISCUSSION Piezometric EQCN Testing of Sur Adsorbability on Au Surfaces. To corroborate the interactions of Sur molecules with a citrate-coated gold surface, in support of the preliminary observations of possible Sur adsorbability on plasmonic gold nanoparticles and suitability of the multishell AuNP@Cit/Sur model for gated-RET devices, nanogravimetric experiments have been performed using the EQCN technique. The mass transients obtained for a bare AuQC electrode after injection of 20 mM Cit and then 0.99 μM Sur (final concentrations) are presented in Figure 2. The temporal evolution of resonance

Figure 2. Nanogravimetric transients observed during citrate selfassembled monolayer adsorption on a Au piezoelectrode in 0.1 M NaClO 4 after injection of a 20 mM Na3Cit solution (final concentration), followed by a monolayer film formation by a protein survivin (Sur) after the injection of a 0.99 μM Sur solution (final concentration). Arrows indicate the injections of Cit and Sur; smooth solid lines represent fitting to the experimental points.

frequency shift, −Δf in Figure 2, illustrates the apparent mass gain, Δm, due to the formation of a citrate self-assembled monolayer on the surface of a gold-coated quartz crystal resonator wafer (QC/Ti/Au). The relation between the resonant frequency shift, Δf, and the film mass change, Δm, is given by the Sauerbrey equation:43−45 13229

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235

Article

The Journal of Physical Chemistry B

Figure 3. (A) TEM image of citrate-capped gold nanoparticles (AuNP@Cit). (B) Size distribution of AuNP@Cit: mean diameter d = 4.9 ± 0.2 nm (number of particles, 142). (C) Normalized absorbance spectrum for AuNP@Cit and fluorescence spectrum for FITC dye. (D) Dependence of FITC fluorescence on AuNP@Cit quencher concentration. (E) Fluorescence emission spectra for (1) FITC; (2) AuNP@Cit; (3) FITC and survivin; (4) AuNP and FITC; and (5) AuNP, survivin, and FITC. Solution: 10 mM PB, pH 7.4. Dye: FITC, cFITC = 66.7 nM, cAuNP@Cit = 2.03 nM, cSur = 3.03 nM, λex = 495 nm.

exhibits a large spectral overlap of its fluorescence spectrum with the absorbance spectrum of AuNP@Cit, providing the basis for efficient fluorescence resonance energy transfer (FRET) from FITC acting as a donor to AuNP@Cit acting as the acceptor. Panel D shows the FITC emission intensity decrease observed upon its interaction with AuNP@Cit. Gold nanoparticles strongly quench the fluorescence of FITC due to the efficient NSET. In the fluorescence emission spectra presented in panel E, curve 1 shows the emission of FITC alone while curve 2 (flat on the bottom) shows no emission of AuNPs. When Sur is added to a solution of FITC probe, almost no change in fluorescence is observed (curve 3), which means that Sur does not quench appreciably the probe at the concentration level of the experiments. However, the addition of AuNP@Cit to the probe solution quenches the probes considerably (curve 4). When Sur is injected to the solution of AuNP@Cit and FITC probe, a reduction in quenching is clearly seen (curve 5 rising above curve 4). All the observations described above are consistent with the gated-RET, as defined previously.36 In the above experiments, the concentrations of components were selected with the following goals in mind: for FITC, to obtain strong fluorescence (cFITC = 66.7 nM); for AuNP@Cit, to achieve considerable quenching of 40−50% (cAuNP = 2.03 nM); and for Sur, to form submonolayer films on AuNP@Cit (cSur = 3.03 nM). In Figure 4, the RET efficiency, E, is plotted versus survivin concentration cSur for constant concentrations of FITC and AuNP@Cit. The RET efficiency, E, is defined as E = (F0 − F )/F0

that the RET efficiency increases linearly with cSur at low Sur levels and then tends to saturate at cSur > 5 nM. The saturation of E is due to the formation of a full monolayer of Sur molecules on AuNPs. The limiting value of E is then due to the probe traveling through the residual channels between the adsorbed Sur molecules (i.e., those which cannot be closed in a compressed full-coverage film) and also due to the RET through the protein film and FRET by the protein itself. The latter effect is likely to be very small, as shown in Figure 3C by comparing spectra 1 and 3. Because the saturation of E at cSur > 5 nM corresponds to θSur = 1, the Langmuir adsorption constant KL,Sur can be determined from the decrease of E to the midpoint. The obtained KL,Sur = 1.09 × 109 M−1. The expanded view of the E−cSur plot at the low-concentration range of Sur shows that the limit of detection (LOD) for Sur, determined using 3σ method, is very low: LOD = 240 pM (Figure 4B). Therefore, the modulation of RET efficiency is very sensitive to the Sur level in solution and the Sur surface coverage on AuNP@Cit. Interparticle Interactions of AuNP@Cit/Sur Probed Using RELS Spectroscopy. The assessment of interparticle interactions between AuNP@Cit/Sur nanoparticles is important to be able to prevent the assembly of Sur-capped AuNPs which could undermine the gRET effect. Hence, the control experiments testing interparticle interactions have been performed using RELS measurements. The RELS spectroscopy has been chosen for its remarkable sensitivity and diagnostic features which enabled us in our previous studies to distinguish between the effects of dielectric function changes in nanoparticle shells and the actual nanoparticle assembly.50,51 In Figure 5, the dependence of RELS intensity on total Sur concentration is presented. It is seen that the RELS intensity saturates for cSur > 1 nM. The maximum increase of the RELS

(2)

where F is the fluorescence intensity of the dye probe and F0 is its fluorescence intensity in the absence of a quencher. It is seen 13230

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235

Article

The Journal of Physical Chemistry B

where a0 and asat are the particle diameters before and after aggregation, respectively. It follows from this formula that for the smallest aggregate of 2 nanoparticles with the resulting diameter ratio asat/a0 ≈ 2, the RELS intensity increase should be 8-fold, much more than the increase observed experimentally (59%). Also, taking into account the thickness of the Sur shell forming on AuNP@Cit, the diameter of a fully coated AuNP@Cit/Sur should be ca. 14.8 nm up from 6.0 nm for AuNP@Cit (the thickness of the Cit shell has been assumed to be 0.5 nm on the basis of atomic distances of a DFT citrate ion). When the concentration of AuNPs is unchanged and any variation in the dielectric function could be neglected, the increase in RELS intensity due to the particle size increase follows the Rayleigh dependence: IRELS,sat /IRELS,0 = (asat /a0)6

(4)

Therefore, in the absence of dielectric function changes, the expected increase in the scattering should be (14.8/6)6 = 225 fold. Because the observed increase in RELS intensity is much smaller, it is most likely due to the dielectric function changes occurring in the nanoparticle shells due to the adsorption of Sur molecules rather than due to the particle size or assembly. Theoretical Model of Protein-Gated RET. To analyze the relationships governing the resonance energy transfer in gRET systems, a simple model presented in Figure 6 has been designed. The fluorescence emission, F0, of a free fluorophore probe (D) in the absence of AuNP quencher (A) and protein molecules (P) is given by

Figure 4. (A) Dependence of the gRET efficiency for energy transfer from FITC to AuNP on survivin concentration showing saturation at higher Sur concentrations. (B) Details of the limit of detection (LOD) determination using 3σ method (LOD = 240 pM). FITC concentration: 66.7 nM, cAuNP@Cit = 2.03 nM, λex = 495 nm, λem = 517 nm.

F0 = ε0c D,0

(5)

where ε0 is the fluorescence coefficient (ε0 = 1.366 × 1010 M−1) and cD,0 is the probe concentration. In the presence of AuNPs, the quenched fluorescence emission F1 is given by F1 = ε1(Q A )c D,0

(6)

where the coefficient ε1(QA) is a function of AuNP@Cit concentration, QA. This function can be derived from the Stern−Volmer equation for static quenching F0 = 1 + KSV,AQ A F1

(7)

as follows: F1 =

ε0c D,0 F0 = 1 + KSV,AQ A 1 + KSV,AQ A

ε1(Q A ) = Figure 5. Light scattering intensity for a solution of AuNP@Cit (3 nM) containing protein survivin; 20 mM citrate buffer, pH 7.46.

(9)

where KSV,A is the Stern−Volmer quenching constant (KSV,A = 1.32 × 109 M−1).36 The probe fluorescence emission F2 in the absence of AuNP may be weakly quenched by a protein, with a quenching constant KSV,P, so the fluorescence can be described by the equations

intensity from the case of Sur absence (IRELS,0 = 5.73 au) to the saturation level (IRELS,sat = 9.12 au) is 59.2%. This increase is smaller than that expected for AuNP assembly, based on the formula derived earlier51 which takes into account the decrease in particle concentration and increase of the particle size while neglecting variability in the dielectric function of the shell and medium: IRELS,sat /IRELS,0 = (asat /a0)3

ε0 1 + KSV,AQ A

(8)

F2 = ε2c D,0

ε2 =

ε0 −η 1 + KSV,Pc P

(10)

(11)

where cP is the free protein concentration and the function η represents contributions that may come from the energy

(3) 13231

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235

Article

The Journal of Physical Chemistry B one obtains F = F0,A +

KLc P (Ffin,P − F0,A ) 1 + KLc P

(15)

where F0,A is the new reference (after addition of AuNP quencher at QA,fin concentration) for subsequent changes due to the additions of protein. The value of the Langmuir adsorption constant, KL, can be determined using eq 15 by reading the cP concentration corresponding to F at the midpoint between F0,A and Ffin,P, for which KL = 1/cP. In this way, one obtains KL,Sur = 1.09 × 109 M−1. It can also be shown that the gated-RET relationship (F0,A/F) = f(cP) is more complex than the Stern−Volmer expression for static quenching and has an opposite slope to that of the Stern− Volmer plot because of the “dequenching” observed with increasing protein concentration: F0,A F

(

KLc P =1−

1+

ε2(c P) ε1(Q A )

)

−1

ε (c ) KLc P ε 2(QP ) 1 A

(16)

The intercept of the plots (F0,A/F) = f(cP) is 1, and the initial slope is ⎛ ∂(F0,A /F ) ⎞ ⎟ ⎜ ⎝ ∂c P ⎠c

migration and low-probability energy transfer from the probe to AuNP through the protein molecules. Note that F2 is experimentally observed as Ffin on the plot of F versus cP, when a saturation sets in at high cP values because of the formation of the full protein monolayer film (θ = 1). The net fluorescence in a gRET system is dependent on the protein coverage, θ, on AuNPs: (12)

The expression for θ can be obtained assuming that the protein obeys the Langmuirian adsorption isotherm for noninteracting adsorbates: KLc P θ= 1 + KLc P (13)

pKFITC = 6.4, pK Cit,1 = 3.14, pK Cit,2 = 4.77, pK Cit,3 = 6.39, pISur = 5.66

While these constants have been determined for species in solution and they may somewhat shift under surface confinement conditions, the general behavior is expected to remain unaffected. It is seen that at pH 7.4, 90% of FITC molecules in solution are deprotonated and most of the citrate ions are deprotonated as well. The Sur molecules are also negatively charged because the solution pH > pISur. In this situation, all major system components should electrostatically repel each other provided that there is no screening by counterions and the surface pK values are not shifted dramatically by surface confinement to reverse the system behavior. On the other hand, there is a strong experimental evidence, shown in the section on EQCN experiments (vide supra), that

where KL is the adsorption equilibrium constant for a protein adsorption on NPs and cP is the free protein concentration in solution. (This equation holds also if cP is replaced with the total protein concentration, cP,0, when the amount of adsorbed protein is small. If that is not the case, an exact solution of the quadratic equation can be used.) By combining eqs 12 and 13 and taking into account that Ffin,P ε2 = ε1(Q A,fin) F0,A

(17)

Thus, the value of the Langmuir adsorption constant KL can be determined from the initial slope of the (F0,A/F) = f(cP) characteristic if the ratio ε2/ε1 is known. It follows from the analysis of eqs 16 and 17 that with increasing value of KL, the gated-RET efficiency decreases steeply in the lower protein concentration range, making the gating increasingly more sensitive. Role of Electrostatic Forces and Hydrogen Bonding on Sur Interactions with AuNP@Cit and FITC. The net charges carried by Sur, FITC, and AuNP@Cit play an important role in binding Sur molecules to form a gating film on a AuNP, and in the case of the carrier probes, they contribute to the control of probe permeation through the gate channels and transfer the excited-state energy to surface plasmons in AuNP. The net charges on Sur, FITC, and AuNP@Cit can be readily evaluated taking into account the respective pK values and the isoelectric point of Sur, pISur, as follows:

Figure 6. Dependence of F0/F on (1) AuNP@Cit concentration, cAuNP, for Stern−Volmer quenching by static RET and (2) on Sur protein concentration, cP, for dequenching by protein film in gatedRET. KSV,AuNP, Stern−Volmer static quenching constant; KL,Sur, Langmuir adsorption constant for Sur protein on AuNP@Cit. Insets: model schematics illustrating the processes involved.

F = (1 − θ )ε1(Q A )c D,0 + θε2c D,0

P= 0

⎛ ε (0) ⎞ = −KL⎜⎜ 2 − 1⎟⎟ ⎝ ε1(Q A ) ⎠

(14) 13232

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235

Article

The Journal of Physical Chemistry B

(N3) and −NH2 group (N4) of Arg and two COO− groups of Cit. The strength of the H-bonds between the positively charged protonated amine groups and negatively charged deprotonated COO− groups are augmented by the local electrostatic attractions between these groups. Thus, the formation of strong H-bonds and the local electrostatic attraction are likely to facilitate the adsorption of Sur on Citcoated Au surfaces. Effect of Resonance Energy Migration on Gated-RET. The sensitive detection of biomolecules is often challenging because of the resonance energy migration and background fluorescence. In fluorescence analysis of proteins by staining with fluorescent dyes, the resonance energy migration (REM) is encountered when an excess of fluorophore in relation to protein is used, resulting in sample overlabeling. REM is similar to FRET, except that resonance energy transfer takes place between the same kinds of molecules and can proceed through a chain of transfers52 when a dense film of fluorophores is encountered (as in the case of overlabeling). A gradual loss in emission intensity is usually observed. Under these conditions, the interpretation of fluorescence data becomes problematic. From the analysis of the gated-RET model presented in Figure 1, it becomes clear that the effect of energy migration in gatedRET is marginalized because the main analytical signal comes from the dye passing through the voids in the protein film rather than from the dye molecules attached to the protein where the REM originates. Therefore, in applications of gRET for the analysis of dilute protein solutions where gRET can replace ultraviolet−visible (UV−vis) and direct fluorescence methods, the exact ratio of dye to protein concentrations does not need to be observed.

indicates clearly the adsorption of Sur on Cit-coated Au piezoelectrodes (Figure 2). Because of this and due to the fact that the citrate shell on AuNP has no pronounced hydrophobic properties, the plausible mechanism of Sur binding to AuNP@ Cit is likely to be based on the involvement of hydrogen bonding and counterions that are screening the excessive negative charges. We have previously indicated50 that citrate SAMs on AuNPs may have an extensive lateral intermolecular hydrogen bonding developed to strengthen the SAM shell. The hydrogen bonding has also been indicated in the initial interactions of AuNP@Cit with various biogenic ligands.51,42 The most likely sites on the Sur surface to form stable attachment to AuNP@Cit are primary amine-bearing amino acids, lysine (Lys) and arginine (Arg). There are 24 lysines and 12 arginines present in the Sur structure. Most of them are located on the Sur surface and thus are accessible for Hbonding to AuNP@Cit. To identify the possible H-bonds that could be formed between citrate ligands of AuNP shell and lysine/arginine moieties on the surface of Sur, we have performed quantum mechanical calculations of the model structures formed in the interactions of Cit ligands with Lys and Arg amino acids. These interactions are illustrated in Figure 7, showing representative ensembles formed in direct interactions of Cit ligands with Lys and Arg. It is seen that three H-bonds are readily formed between ε NH3+ group of Lys and two COO− groups of Cit. They are shown with dashed green lines. Similarly, three H-bonds are formed between =NH2+ group

■

CONCLUSIONS We have demonstrated that the protein survivin can be utilized in forms of submonolayer films on AuNP@Cit to modulate resonance energy transfer between fluorophore probe in solution and nanoparticles’ surface plasmons. In this gatedRET phenomenon, the plasmonic field-enhanced energy transfer is precisely controlled by the width of the channels (voids) in the protein film, with very high sensitivity down to 240 pM Sur concentrations. The protein ability to shield plasmonic fields arises from the dielectric loss in the protein being higher than in water, which is due to the value of the imaginary part of the dielectric function of protein being higher than that of water at the plasmonic frequency. Using piezometric measurements, we have determined that Sur adsorbs on citrate-modified gold surfaces (γSur = 5.4 pmol/ cm2) despite repulsive electrostatic forces. The strong binding is likely due to the multiple hydrogen bonding originating from −NH3+, =NH2+, and −NH2 groups of lysines and arginines of Sur and electrolyte ion screening. While the immobilization of Sur on AuNP@Cit causes a decrease in the efficiency of RET from FITC to surface plasmons of nanoparticle cores, no appreciable quenching of the probe fluorescence by Sur in solution has been observed. Using the proposed theoretical treatment of the gated-RET model, based on Langmuirian adsorption, one can identify main parameters determining the RET efficiency dependence on gating-material surface coverage. This model can be utilized for determination of effective binding constants for proteins. As an example, the Langmuirian constant, KL,Sur, for Sur adsorption on AuNP@Cit has been determined from the gated-RET measurements (KL,Sur = 1.09 × 109 M−1). The high sensitivity of the gated-RET method can

Figure 7. Electron density surfaces with electrostatic potential mapping for a model adsorbed citrate (Cit) interacting with (A) lysine (Lys) and (B) arginine (Arg) amino acids of the Sur molecule with the formation of hydrogen bonds between two COO− groups of Cit and amino groups of Lys or Arg, marked with green dashed lines. (C) Structural details of the H-bonding between COO- groups of Cit and amine groups of N3 and N4 atoms of Arg enhanced by electrostatic interactions between local interfacial charges indicated with color circles. ρ = 0.1 au−3; color-coded electrostatic potential: from high (blue) to low (red). 13233

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235

Article

The Journal of Physical Chemistry B

(11) Lakowicz, J. R. Plasmonics in Biology and Plasmon-Controlled Fluorescence. Plasmonics 2006, 1, 5−33. (12) Lakowicz, J. R.; Ray, K.; Chowdhury, M.; Szmacinski, H.; Fu, Y.; Zhang, J.; Nowaczyk, K. Plasmon-controlled fluorescence: a new paradigm in fluorescence spectroscopy. Analyst 2008, 133, 1308− 1346. (13) Anger, P.; Bharadwaj, P.; Novotny, L. Enhancement and quenching of single molecule fluorescence. Phys. Rev. Lett. 2006, 96, 113002. (14) Ray, K.; Chowdhury, M. H.; Zhang, J.; Fu, Y.; Szmacinski, H.; Nowaczyk, K.; Lakowicz, J. R. Plasmon-controlled fluorescence towards high-sensitivity optical sensing. Adv. Biochem. Eng./Biotechnol. 2009, 116, 29−72. (15) Volpati, D.; Spada, E. R.; Cid, C. C. P.; Sartorelli, M. L.; Aroca, R. F.; Constantino, C. J. L. Exploring copper nanostructures as highly uniform and reproducible substrates for plasmon-enhanced fluorescence. Analyst 2015, 140, 476−482. (16) Zhang, J.; Fu, Y.; Lakowicz, J. R. Enhanced forster resonance energy transfer (FRET) on a single metal particle. J. Phys. Chem. C 2007, 111, 50−56. (17) Zhang, J.; Fu, Y.; Chowdhury, M. H.; Lakowicz, J. R. Plasmoncoupled fluorescence probes: Effect of emission wavelength on fluorophore-labeled silver particles. J. Phys. Chem. C 2008, 112, 9172−9180. (18) Akbay, N.; Lakowicz, J. R.; Ray, K. Distance-dependent metalenhanced intrinsic fluorescence of proteins using polyelectrolyte layerby-layer assembly and aluminum nanoparticles. J. Phys. Chem. C 2012, 116, 10766−10773. (19) Stobiecka, M.; Hepel, M. Multimodal coupling of optical transitions and plasmonic oscillations in rhodamine B modified gold nanoparticles. Phys. Chem. Chem. Phys. 2011, 13, 1131−1139. (20) Kinkhabwala, A.; et al. Large single-molecule fluorescence enhancements produced by a bowtie nanoantenna. Nat. Photonics 2009, 3, 654−657. (21) Hao, E.; Schatz, G. C. Electromagnetic fields around silver nanoparticles and dimmers. J. Chem. Phys. 2004, 120, 357. (22) Choudhury, S. D.; Badugu, R.; Ray, K.; Lakowicz, J. R. Silver− gold nanocomposite substrates for metal-enhanced fluorescence: Ensemble and single-molecule spectroscopic studies. J. Phys. Chem. C 2012, 116, 5042−5048. (23) Galarreta, B. C.; Rupar, I.; Young, A.; Lagugne-Labarthet, F. Mapping hot-spots in hexagonal arrays of metallic nanotriangles with azobenzene polymer thin films. J. Phys. Chem. C 2011, 115, 15318− 15323. (24) Fruhnert, M.; Kretschmer, F.; Geiss, R.; Perevyazko, I.; CiallaMay, D.; Steinert, M.; Janunts, N.; Sivun, D.; Hoeppener, S.; Hager, M. D.; et al. Synthesis, separation, and hypermethod characterization of gold nanoparticle dimers connected by a rigid rod linker. J. Phys. Chem. C 2015, 119, 17809−17817. (25) Kusch, P.; Heeg, S.; Lehmann, C.; Müller, N. S.; Wasserroth, S.; Oikonomou, A.; Clark, N.; Vijayaraghavan, A.; Reich, S. Quantum nature of plasmon-enhanced raman scattering. 2015, http://arxiv.org/ abs/1503.03835v3. (26) Cushing, S. K.; Li, J.; Bright, J.; Yost, B. T.; Zheng, P.; Bristow, A. D.; Wu, N. Controlling plasmon-induced resonance energy transfer and hot electron injection processes in metal@TiO2 core−shell nanoparticles. J. Phys. Chem. C 2015, 119, 16239−16244. (27) Qiu, J.; Wei, W. D. Surface plasmon-mediated photothermal chemistry. J. Phys. Chem. C 2014, 118, 20735−20749. (28) Brongersma, M. L.; Halas, N. J.; Nordlander, P. Plasmoninduced hot carrier science and technology. Nat. Nanotechnol. 2015, 10, 25−34. (29) Zhong, Y.; Ueno, K.; Mori, Y.; Oshikiri, T.; Misawa, H. Cocatalyst effects on hydrogen evolution in a plasmon-induced watersplitting system. J. Phys. Chem. C 2015, 119, 8889−8897. (30) Feng, A. L.; You, M. L.; Tian, L.; Singamaneni, S.; Liu, M.; Duan, Z.; Lu, T. J.; Xu, F.; Lin, M. Distance-dependent plasmonenhanced fluorescence of upconversion nanoparticles using polyelectrolyte multilayers as tunable spacers. Sci. Rep. 2015, 5, 7779.

also be used to develop new procedures for standardization of dilute protein solutions for which UV−vis spectroscopy is not sufficiently sensitive. Further studies are being conducted to develop gated-RET techniques for microfluidic devices.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcb.5b07778. Additional figures and information about Sur (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +48 22 593 8621. Fax: +48 22 593 8619. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported by funding provided by the Ministry of Science and Higher Education Grant Iuventus Plus, No. IP2012058072.

■

ABBREVIATIONS RET, resonance energy transfer; FRET, fluorescence RET; NSET, nanomaterial surface RET; gRET, gated-RET; EQCN, electrochemical quartz crystal nanogravimetry; RELS, resonance elastic light scattering

■

REFERENCES

(1) Stobiecka, M.; Molinero, A. A.; Chalupa, A.; Hepel, M. Mercury/ homocysteine ligation-induced ON/OFF-switching of a T-T mismatch-based oligonucleotide molecular beacon. Anal. Chem. 2012, 84, 4970−4978. (2) Stobiecka, M.; Chalupa, A. Biosensors based on molecular beacons. Chem. Pap. 2015, 69, 62−76. (3) Alam, R.; Karam, L.; Doane, T.; Zylstra, J.; Fontaine, D.; Branchini, B.; Maye, M. Near infrared bioluminescence resonance energy transfer from firefly luciferase quantum dot bionanoconjugates. Nanotechnology 2014, 25, 1−7. (4) Naczynski, D. J.; Tan, M. C.; Zevon, M.; Wall, B.; Kohl, J.; Kulesa, A.; Chen, S.; Roth, C. M.; Riman, R. E.; Moghe, P. V. Rareearth-doped biological composites as in vivo shortwave infrared reporters. Nat. Commun. 2013, 4, 1−10. (5) Larson, D. R.; Zipfel, W. R.; Williams, R. M.; Clark, S. W.; Bruchez, M. P.; Wise, F. W.; Webb, W. W. Water-soluble quantum dots for multiphoton fluorescence imaging in vivo. Science 2003, 300, 1434−1436. (6) Michalet, X.; Weiss, S.; Jager, M. Single-molecule fluorescence studies of protein folding and conformational dynamics. Chem. Rev. 2006, 106, 1785−1813. (7) Taminiau, T. H.; Stefani, F. D.; Segerink, F. B.; van Hulst, N. F. v. Optical antennas direct single-molecule emission. Nat. Photonics 2008, 2, 234−237. (8) Grubisic, A.; Schweikhard, V.; Baker, T. A.; Nesbitt, D. J. Coherent multiphoton photoelectron emission from single Au nanorods: The critical role of plasmonic electric near-field enhancement. ACS Nano 2013, 7, 87−99. (9) Zhao, L.; Ming, T.; Chen, H. J.; Liang, Y.; Wang, J. F. Plasmoninduced modulation of the emission spectra of the fluorescent molecules near gold nanorods. Nanoscale 2011, 3, 3849−3859. (10) Narband, N.; Uppal, M.; Dunnill, C. W.; Hyett, G.; Wilson, M.; Parkin, I. P. The interaction between gold nanoparticles and cationic and anionic dyes: enhanced UV-visible absorption. Phys. Chem. Chem. Phys. 2009, 11, 10513−10518. 13234

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235

Article

The Journal of Physical Chemistry B (31) Zhang, J.; Fu, Y.; Lakowicz, J. R. Fluorescent Metal Nanoshells: Lifetime-tunable molecular probes in fluorescent cell imaging. J. Phys. Chem. C 2011, 115, 7255−7260. (32) Lakowicz, J. R. Radiative decay engineering: Biophysical and biomedical applications. Anal. Biochem. 2001, 298, 1−24. (33) Lakowicz, J. R.; Shen, Y.; D’Auria, S.; Malicka, J.; Fang, J.; Gryczynski, Z.; Gryczynski, I. Radiative decay engineering. 2. Effects of silver island films on fluorescence intensity, lifetimes, and resonance energy transfer. Anal. Biochem. 2002, 301, 261−277. (34) Yun, C. S.; Javier, A.; Jennings, T.; Fisher, M.; Hira, S.; Peterson, S.; Hopkins, B.; Reich, N. O.; Strouse, G. F. Nanometal surface energy transfer in optical rulers, breaking the FRET barrier. J. Am. Chem. Soc. 2005, 127, 3115−3119. (35) Zhang, X.; Marocico, C. A.; Lunz, M.; Gerard, V. A.; Gun’ko, Y. K.; Lesnyak, V.; Gaponik, N.; Susha, A. S.; Rogach, A. L.; Bradley, A. L. Experimental and theoretical investigation of the distance dependenceof localized surface plasmon coupled Forster rsonance energy transfer. ACS Nano 2014, 8, 1273−1283. (36) Stobiecka, M. Novel plasmonic field-enhanced nanoassay for trace detection of proteins. Biosens. Bioelectron. 2014, 55, 379−385. (37) Altieri, D. C. The molecular basis and potential role of survivin in cancer diagnosis and therapy. Trends Mol. Med. 2001, 7, 542−547. (38) Altieri, D. C. Survivin, versatile modulation of cell division and apoptosis in cancer. Oncogene 2003, 22, 8581−8589. (39) Blanc-Brude, O. P.; Mesri, M.; Wall, N. R.; Plescia, J.; Dohi, T.; Altieri, D. C. Therapeutic targeting of the survivin pathway in cancer: Initiation of mitochondrial apoptosis and suppression of tumorassociated angiogenesis. Clin. Cancer Res. 2003, 9, 2683−2692. (40) Mita, A. C.; Mita, M. M.; Nawrocki, S. T.; Giles, F. J. Survivin: Key regulator of mitosis and apoptosis and novel target for cancer therapeutics. Clin. Cancer Res. 2008, 14, 5000−5005. (41) Altieri, D. C. Survivin, cancer networks and pathway-directed drug discovery. Nat. Rev. Cancer 2008, 8, 61−70. (42) Stobiecka, M.; Hepel, M. Double-shell gold nanoparticle-based DNA-carriers with poly-L-lysine binding surface. Biomaterials 2011, 32, 3312−3321. (43) Sauerbrey, G. Verwendung von Schwingquarzen zur Wägung dünner Schichten und zur Mikrowägung. Eur. Phys. J. A 1959, 155, 206−222. (44) Hepel, M. Electrode-solution interface studied with electrochemical quartz crystal nanobalance. In Interfacial Electrochemistry. Theory, Experiment, and Applications; Wieckowski, A., Ed.; M. Dekker: New York, 1999; pp 599−631. (45) Hepel, M.; Cateforis, E. Studies of copper corrosion inhibition using electrochemical quartz crystal nanobalance and quartz crystal immittance techniques. Electrochim. Acta 2001, 46, 3801−3815. (46) Hepel, M.; Stobiecka, M. Interactions of adsorbed albumin with underpotentially deposited copper on gold piezoelectrodes. Bioelectrochemistry 2007, 70, 155−164. (47) Turkevich, J.; Stevenson, P. C.; Hillier, J. A study of the nucleation and growth processes in the synthesis of colloidal gold. Discuss. Faraday Soc. 1951, 11, 55−75. (48) Kunze, J.; Burgess, I.; Nichols, R.; Buess-Herman, C.; Lipkowski, J. Electrochemical evaluation of citrate adsorption on Au(1 1 1) and the stability of citrate-reduced gold colloids. J. Electroanal. Chem. 2007, 599, 147−159. (49) Chantalat, L.; Skoufias, D.; Kleman, J.; Jung, B.; Dideberg, O.; Margolis, R. Crystal structure of human survivin reveals a bow tieshaped dimer with two unusual alpha-helical extensions. Mol. Cell 2000, 6, 183−189. (50) Stobiecka, M.; Hepel, M. Rapid functionalization of metal nanoparticles by moderator-tunable ligand-exchange process for biosensor designs. Sensor. Sens. Actuators, B 2010, 149, 373−380. (51) Stobiecka, M.; Deeb, J.; Hepel, M. Ligand exchange effects in gold nanoparticle assembly induced by oxidative stress biomarkers: Homocysteine and cysteine. Biophys. Chem. 2010, 146, 98−107. (52) Wang, F.; Deng, R.; Wang, J.; Wang, Q.; Han, Y.; Zhu, H.; Chen, X.; Liu, X. Tuning upconversion through energy migration in core-shell nanoparticles. Nat. Mater. 2011, 10, 968−973. 13235

DOI: 10.1021/acs.jpcb.5b07778 J. Phys. Chem. B 2015, 119, 13227−13235