1-1. Introduction
Nowadays, proteases not only catalyze protein degradation by hydrolysis of peptide bonds but also are recognized as exceptionally important molecules that are engaged in
numerous vital life processes. Proteinase activity is considered an important biological marker in various pathologies, because the expression and activity of proteases are significantly different in several pathologies, including inflammatory, cardiovascular diseases, neurological disorders cancer, arthritis, and atherosclerosis [Thobhani et al., 2010]. Thus, the development of proteinase assay has been explored, which usually based on substrate zymography,
radioisotopes, on chromogenic, or fluorogenic substrate. However, these techniques are often time-consuming, expensive, discontinuous, or require specific instruments. More sensitive and convenient proteinase assay is needed; especially methods allow detecting and imaging protease activities in living organisms using different imaging modalities.
Gold nanoparticles (AuNPs) are a potential nanomaterial in biosensor field, account of their unique size- and distance-dependent optical properties and superior performances of being energy acceptors and quenchers [Guarise et al., 2006]. AuNPs interested many biological researchers for their feasible of surface coating and great biocompatibility; hence the applications of AuNPs in proteinase detection have arisen. For the purpose of high sensitivity sensing/detecting, fluorophore conjugated AuNPs protease activatable probes are popular for researchers.
2
1-2. Gold nanoparticles
AuNPs are some of the most widely investigated nanomaterial nowadays, because of their unique characteristic such as: optical, electrical, chemical and catalytic properties. Jans and Huo documented that since 2005, there are more than 5,000 literatures (including journal articles and reviews) were searched with key words “gold nanoparticles” and “detection”
using SciFinder Scholar [Jans et al., 2012]. The fact indicated that AuNPs application in detecting field is an arising trend.
1-2-1. Citrate reduction method
The most common chemical route of synthesizing AuNPs starts from Au (III) salts, which are then reduced to Au (0). The gold precursor is usually chloroauric acid (HAuCl4) dissolved in aqueous and the reducing agents could be sodium citrate, ascorbic acid, sodium boron hydride, or blockcopolymers [Polte et al., 2010]. Since 1981, more than 230 published studies have employed citrate reduction method to generate AuNPs. These researches showed scarce data on non-AuNPs components in the reaction system although some byproducts (such as ketoglutaric acid) in the synthesized AuNP solution have been reported [Kumar et al., 2007; Turkevich et al., 1951].
The sodium citrate reduction technique pioneered by Turkevich et al. in 1951 and refined by Frens in 1973. In brief, an aqueous HAuCl4 reduced by trisodium citrate as the reducing and stabilizing agent at the boiling point of solution. The stoichiometry of the citrate reduction method is confirmed as follows [Balasubramanian et al., 2010]:
2HAuCl4 + 3Na3C6H5O7 + 1.5 H2O
→ 2Au0 + Na2C5H6O5 + 3CH3COONa + 4NaCl + 4Cl- + 2H+ + 0.5 C3H6O + 5.5CO2
3
The resulting AuNPs acquires a citrate layer on surface, which confers negative charge and stability (Fig. 1-1). The citrate ligand can be easily displaced by several species that form stronger interactions with Au. The size of AuNPs range from 10 to 150 nm in diameter can be easily controlled by the ratio gold precursor (HAuCl4) to reduction agent (Na3C6H5O7) [Fanun, 2010; Frens, 1973]. However, the Turkevich-Frens citrate reduction method produce modestly monodisperse spherical AuNPs suspended in water of only around 10 to 15 nm in diameter;
and the larger particles may produce at the loss of monodispersity and shape [Kimling et al., 2006].
The recently proposed mechanism of gold nanoparticle formation could interpret as a four-step nucleation and growth process (Fig. 1-2). The initial stage is a rapid formation of nuclei and the second stage is that the nuclei coalesce into bigger particles. The third stage is low diffusion growth of AuNPs which comprised by ongoing process of gold reduction and a further coalescence. Finally, AuNPs completely consume their precursor and grow rapidly to their terminal size. This mechanism indicates that the coalescence of small nuclei plays a vital role throughout the AuNPs formation and determines the polydispersity of colloid AuNPs [Polte et al., 2010].
1-2-2. Local surface plasmon resonance
Surface plasmon resonance (SPR) is a physical characteristic that metallic materials such as Ag, Cu, Au, and Al, process a negative real and small positive imaginary dielectric constant over a range of wavelengths [Henry et al., 2011]. When these metallic materials stimulated by electromagnetic radiation, these would form an electron gas that moves away from its
equilibrium position and be displaced by induced surface polarization charges that act as a restoring force. This positive imaginary arises from Coulomb attraction between electrons and nuclei to against the electron gas. The collective oscillation of the conducted electrons is
4
called the dipole plasmon resonance. The oscillation frequency is determined by four factors:
the density of electrons, the effective electron mass, and the shape and size of the charge distribution [Jans et al., 2012; Kelly et al., 2003]. The SPR could category to propagating or localized. Propagating surface plasmon is observed on thin metallic films, called surface plasmon polaritons (SPPs); whereas localized surface plasmon is observed on nanoscale structures, which nanoparticles (NPs) are much smaller than the incident wavelength (Fig.
1-3). That plasmon oscillates locally around the nanoparticle with a frequency called local surface plasmon resonance (LSPR). The LSPR also sensitive to changes in the local dielectric environment like SPR do [Willets et al., 2007].
The LSPR spectrum is strongly dependent on the nanoparticle’s size, shape, dielectric constant and the surrounding environment under dielectric constant as mentioned (Fig. 1-4).
AuNPs have an optical property which is bulk plasmon resonance in the visible region of the spectrum, while for most other metals this resonance only occurs in the ultraviolet (UV) region. Therefore, the strong absorption or scattering of AuNPs at the visible light region could make them could be observed by naked eyes or be easy to detect color change. Besides, the possibility of tuning the LSPR band of AuNPs (including nanorods, shells, stars, and other shapes) at the near IR region makes them promising materials for in vivo imaging and
analysis [Jans et al, 2012].
LSPR-based sensors are label-free techniques. By LSPR change, there are two types of LSPR-based sensors are conducted. One is based on aggregation of the colloid and results in apparent color change (from red to blue). Aggregation causes a coupling of the colloid plasma modes resulting in a red shift and broadening of the longitudinal plasma resonance in the UV-vis spectrum. Due to dipole—dipole interactions occur, the wavelength of absorption may be varied from 520 nm (effectively isolated particles) through 750 nm (particles that are separated by only 0.5 nm). The resulting spectrum consists of the conventional plasmon
5
resonance due to single spherical particles and the new peak due to particle—particle interactions [Ghosh et al., 2007].
The other type is based on LSPR extinction (or scattering) wavelength maximum, λmax, which is sensitive to the dielectric constant of the surrounding medium or adsorbents. Thus, extinction wavelength maximum changes in the local environment, for example it should cause a shift in λmax in the presence of an adsorbed species. This relationship is expressed by the following equation:
∆λ𝑚𝑎𝑥 = 𝑚∆𝑛 [1 − 𝑒𝑥𝑝 (−2𝑑𝑙
𝑑 )] (1) Here 𝑚 is the bulk refractive index response of the nanoparticle, ∆𝑛 is the change in the refractive index induced by the adsorbents, 𝑑 is the effective adsorbent layer thickness, and 𝑙𝑑 is the characteristic EM-field-decay length (approximated as an exponential decay).
This relationship is the basis of LSPR wavelength-shift sensing experiments. When molecules bind to a AuNPs the refractive index will change and the LSPR band occurs red-shift. It can be deduced from this equation that the sensitivity towards refractive index changes is distance dependent. Only at close proximity to the nanoparticle surface will give rise to a shift of the LSPR wavelength. This makes the refractive index based biosensor be specific for
interactions close to the NPs surface [Jans et al., 2012; Willets et al., 2007].
Depending on the optical properties of spherical AuNPs as mentioned, the size and concentration of the spherical AuNPs could be determined by UV-Vis spectra. The extinction coefficient is an important parameter that can be used to calculate the NPs concentration or estimating the NPs size. According to Lambert-Beers law, the molar concentration of the solution can be obtained.
A = εbC (2)
6
Liu et al. [2007] used high resolution transmission electron microscopy analysis and UV-vis absorption spectroscopic measurement to determine the extinction coefficients of AuNPs with the size ranging from 4 to 40 nm. An equation was provided that it is
independent to the capping ligands on the NPs surface and the solvent dissolve the NPs, ln 𝜀 = 𝑘𝑙𝑛𝐷 + 𝑎 (3)
Where 𝜀 is extinction coefficient in M-1cm-1, 𝐷 is the core diameter of the NPs, and 𝑘 = 3.32, 𝑎 = 10.8. The correlation coefficient is 0.99 and the standard deviation is 0.22.
Haiss et al. [2007] provided equations according to SPR peak that can determine bare spherical AuNPs size and concentration. For particles ranging from 35 to 100 nm can be calculated from the peak position according to eq (4):
𝑑 = ln (
𝜆𝑠𝑝𝑟−𝜆0 𝐿1 )
𝐿2 (4) Where d is the diameter of the spherical AuNPs, 𝜆𝑠𝑝𝑟 is the wavelength at the peak of the SPR; 𝜆0 =512; L1 = 6.53; and L2 = 0.0216. Haiss et al. found the average of the absolute error is only 3%.
The particle diameter in the size ranging from 5 to 80 nm has a better agreement between theory and experiment as the particle diameter is found if the absorbance ratios are
determined in the wavelength region below 600 nm. Hence the ratio Aspr/A450 should be particularly suitable to calculate the particle diameter (in nanometers) from 5 to 80 nm, eq (5):
𝑑 = exp (𝐵1𝐴𝐴𝑠𝑝𝑟
450− 𝐵2) (5) Here, 𝐴𝑠𝑝𝑟 is the absorbance at the SPR peak, A450 is the absorbance at 450nm, B1 = 3.00 and B2 = 2.20. According to eq (5), the particle diameter average deviation is about 11%.
The number density of the particles (N) can be determined by following eq (6):
7
N = 𝐴450×1014
𝑑2[−0.295+1.36𝑒𝑥𝑝(−(𝑑−96.878.2 )2)] (6) Where A450 is the absorbance at 450nm, and d is the particle diameter (nm). Haiss et al. found this equation to be accurate to ~6% [Haiss et al., 2007]. Above equations are also used by NanopartzTM as a tool to determine size and concentration of AuNPs.
1-2-3. Fluorescence quenching mechanisms by AuNPs
Fluorescence activatable probes are comprised of at least two components: the
fluorophore (donor, D) and the quencher (acceptor, A). Any process that causes a decrease in intensity can be considered to be quenching [Lakowicz, 2006]. The fluorescence emission can be altered when the fluorophore is placed near an entity possessing an EM field. Typically metal entities are nano-sized, such as gold, silver, platinum, copper NPs, etc [Kang et al., 2011]. In the development of activatable probes, inorganic AuNPs owns very high sensitivity, which has the highest quenching efficiency (up to 99%) [Swierczewska et al., 2011].
There are two main factors considered to affect the changes on the fluorescence by metal NPs: (1) the plasmon field generated around the particle by the incident light, increasing the excitation decay rate of the fluorophore or enhancing the level of fluorescence emission; and (2) the dipole energy around the nanoparticle reduces the ratio of the radiative to
non-radiative decay rate and the quantum efficiency of the fluorophore, resulting in the fluorescence quenching [Kang et al., 2011]. Plasmon field on AuNPs effect fluorescence by quenching/enhancing was discussed as followings. Mie theory presented a solution to
Maxwell’s equations that describes the extinction spectra (extinction = scattering + absorption) of spherical particles of arbitrary size [Kelly et al., 2003]. According to Mie theory and the size and shape of the particle, the extinction of metal colloids can be due to either absorption or scattering [Yguerabide et al., 1998ab]. Incident energy is dissipated by absorption; and far-field radiation is created by scattering. The radiating plasmon (RP) model provided by
8
Lakowicz [2005] conducted with Mei theory that small colloids are expected to quench fluorescence because absorption is dominant over scattering; while larger colloids are expected to enhance fluorescence because the scattering component is dominant over absorption. The RP model for NPs explains that their induced plasmon can radiate when the scattering cross section rules over the absorption cross-section. The particle cross section for extinction (𝐶𝐸) with a dielectric constant ε1 is dependent on the cross section due to
absorption (𝐶𝐴) and scattering (𝐶𝑠) by:
𝐶𝐸 = 𝐶𝐴 + 𝐶𝑆 = 𝑘1𝐼𝑚(𝛼) +𝑘6𝜋14|𝛼|2 (7)
where 𝑘1 is the wavevector of the incident light in medium. Polarizability (𝛼) of a sphere with a radius r is:
α = 4πr3(𝜀𝜀𝑚−𝜀1
𝑚+2𝜀1) (8) where 𝜀𝑚 is the complex dielectric constant of the metal. The absorption term, 𝐶𝐴 is responsible for quenching, while the scattering term, 𝐶𝑆 can cause fluorescence enhancement.
As seen by this model, the NPs size plays a more significant role in 𝐶𝑆 (𝑟3) over 𝐶𝐴 (𝑟3).
Therefore, smaller NPs are preferred for quenching. In accordance with the RP model, AuNPs with diameter below 40 nm are more efficient fluorophore quenchers; while larger colloids above 40 nm are expected to enhance fluorescence, because scattering becomes the dominant mechanism [Swierczewska et al., 2011].
Kang et al. [2011] theoretically studied the plasmon field on AuNPs effect to the fluorescence and provided five important factors for designing the quenching and enhancement effect by metal NPs. The normalized enhancement of excitation decay rate (𝛾𝛾𝑒𝑥𝑐
𝑒𝑥𝑐0 ) , which is the main cause for fluorescence enhancement, shows more significant differences with the size, because it has a relationship with the square of the field strength
9 Where the superscript ‘0’ is the value of the system without AuNPs, the plasmon field strength at a distance (𝛾) to AuNP core is 𝐸𝑝, and an incident light field to AuNP is 𝐸0.
The quantum yield (q) indirectly influenced by the plasmon field 𝐸𝑝 can be described as: resulted from the radiated energy absorbed by the particle, and 𝑞0 is the intrinsic quantum yield of the fluorophore. For a spherical particle with a quasi-static polarizability, 𝛾𝛾𝑟
𝑟0= 𝛾𝛾𝑒𝑥𝑐
𝑒𝑥𝑐0 . The fluorescence enhancement rate (φ) is, therefore, the combined effect of the
enhancement of the excitation decay rate and the change in the quantum yield, both influenced by the plasmon field.
φ =𝛾𝛾𝑒𝑥𝑐
𝑒𝑥𝑐0 𝑞
𝑞0 (11) Kang et al. [2011] provided five important factors for designing the quenching and enhancement effect by metal NPs: (1) The metal type of the particle for example: AuNP, which the dielectric permittivity of the metal determines the plasmon field distribution. (2) The NPs size, i.e., field strength and the enhancement of the excitation decay rate depends on the particle size, (eq (9)). (3) The fluorophore to be used, which determines the wavelength.
The field strength depends upon the excitation wavelength, and the level of absorption of the emission light by the NPs varies depending upon the emission wavelength, (eq (10)). (4) The
10
intrinsic quantum yield of the flurophore: It is one of the major factors that determine the quantum yield of the fluorophore placed near the nanoparticle, (eq (10)). (5) The placement of a shell on the surface. The plasmon field distribution may change significantly depending on the material properties of the shell on the particle [Kang et al., 2011].
(2) The radiative and non-radiative energy transfer of fluorophore to AuNPs
Quenching efficiency depends on the measure of the fluorescence decay rate (Rfluo), radiative decay rate (Rrad), nonradiative decay rate (Rnonrad), and the fluorescence quantum efficiency (η). The fluorescence decay rate is the inverse of the fluorescence lifetime (t), Rfluo
= 1/τ and can be expressed as the sum of the radiative and nonradiative decay rates [Dulkeith et al., 2002]:
Rfluo = Rrad + Rnonrad (12) η = Rrad/ Rfluo (13)
The process of the dye molecule releases a photon returning to the ground state is called radiative decay; while the excited photon cannot return to its ground state due to various processes (such as intersystem crossings or heat dissipation) is called non-radiative decay.
Radiative and nonradiative rates depend on the size and shape of the NPs, the distance between the fluorophores and NPs, the orientation of the dye molecule binding onto the AuNPs, and also on the overlap of the fluorophore’s emission and NP absorption [Swierczewska et al., 2011].
At those small distances, the large fluorescence quenching efficiency of 99.8% is due to two effects of equal importance: first, the AuNPs increase the non-radiative rate (Rnonrad ) of the molecules due to energy transfer, and second, the radiative rate (Rrad) of the molecules is decreased because the molecular dipole and the dipole induced on the AuNP radiate out of phase if the molecules are oriented tangentially to the AuNPs surface [Dulkeith et al., 2005].
11
Three mechanisms used to explain energy transfer between a donor and AuNP of fluorescence quenching are discussed here, such as Gersten–Nitzan (GN), Fluorescence Resonance Energy Transfer (FRET) and Nanometal Surface Energy Transfer (NSET) models.
In general, the quantum efficiency (𝜑𝐸𝑛) of energy transfer efficiency can be written as:
𝜑𝐸𝑛 = 1
1+(𝑅0𝑅)𝑛 (14) Where 𝜑𝐸𝑛 is dependent on the distance between the donor and acceptor (𝑅), and the 50% quenching distance (𝑅0) [Ray et al., 2007].
The Gersten-Nitzan model (GN Model) describes a coupling of the fluorophore induced plasmon and strong electric field of AuNPs. Both radiative (fluorescence enhancement) and non-radiative (fluorescence quenching) rates are taken into account under this model. In eq (14), n = 6, and 𝑅0 is: dielectric constant of the metal, respectively, and 𝑐 is the speed of light. The GN model is able to show that a small dipole from the fluorophore can induce a large dipole in the NPs.
Such an enhancement in the dipole increases energy transfer efficiencies by 104 ~ 105.
However, such strong interactions may underestimate the quenching abilities of AuNPs due to the rapid damping of the electric field on their surface [Swierczewska et al., 2011].
Dulkeith et al. [2005] applied GN model to find out how the influence of AuNPs on the quantum yield of fluorophores ceases when the separation between the two species is
gradually increased. They proved that the energy transfer rate is 2 orders of magnitude smaller than that calculated using the GN model and that the distance independent (2.2 ~ 16.2 nm) is
12
opposite expected from the Förster theory. Implying energy transfer does not play an important role in quenching with longer distance. In other words, the reduced quantum efficiency may be due to the reduced radiative rate and not energy transfer. This discrepancy is attributed to the fact that the GN model does not take into account nonlocal effects but the Förster theory does [Dulkeith et al., 2005].
FRET is an electro-dynamic phenomenon, occurring between a donor (D) molecule in the excited state and an acceptor (A) molecule in the ground state. The donor molecules typically emit at shorter wavelengths that overlap with the absorption spectrum of the acceptor (Fig. 1-5A). The term resonance energy transfer (RET) is preferred because the process of long range dipole— dipole interactions between the donor and acceptor does not involve emission and reabsorption photons. The theory of energy transfer is based on a
fluorophore acts as an oscillating dipole, which can exchange energy with another dipole with a similar resonance frequency. Hence RET is similar to the behavior of coupled oscillators and is a non-radiative energy transfer.
The rate of transfer for a donor and acceptor separated by a distance r is given by 𝑘𝜏(𝑟) =𝜑𝜏𝐷𝜅2
𝐷𝑟6(9000(𝑙𝑛10)128𝜋5𝑁𝑛4) ∫ 𝐹0∞ 𝐷(𝜆)𝜀𝐴(𝜆)𝜆4𝑑𝜆 (16) where 𝜑𝐷 is the quantum yield of the donor in the absence of acceptor, n is the
refractive index of the medium, N is Avogadro's number, 𝑟 is the distance between the donor and acceptor, and 𝜏𝐷 is the lifetime of the donor in the absence of acceptor. The refractive index (n) is typically assumed to be 1.4 for biomolecules in aqueous solution. 𝐹𝐷(𝜆) is the corrected fluorescence intensity of the donor in the wavelength range λ to λ + Δλ with the total intensity (area under the curve) normalized to unity. 𝜀𝐴(𝜆) is the extinction coefficient of the acceptor at λ, which is typically in units of M–1 cm–1. The term 𝜅2 is a factor
describing the relative orientation in space of the transition dipoles of the donor and acceptor.
13
𝜅2 is usually assumed to be equal to 2/3, which is appropriate for dynamic random averaging of the donor and acceptor.
The overlap integral (𝐽(𝜆)) expresses the degree of spectral overlap between the donor emission and the acceptor absorption:
𝐽(𝜆) = ∫ 𝐹𝐷(𝜆)𝜀𝐴(𝜆)𝜆4𝑑𝜆 =∫ 𝐹𝐷(𝜆)𝜀𝐴(𝜆)𝜆4𝑑𝜆 distance, and 𝑟 is the donor-to-acceptor distance. Hence, the rate of transfer is equal to the decay rate of the donor (1/𝜏𝐷) when the D-to-A distance (𝑟) is equal to the Förster distance (𝑅0), and the transfer efficiency is 50%. At this distance (𝑟 = 𝑅0) the donor emission would be decreased to half its intensity in the absence of acceptors. The rate of RET depends strongly on distance, and is proportional to 𝑟 = 6 [Lakowicz, 2006].
The distance at which RET is 50% efficient is called the Förster distance (𝑅0) , which is typically in the range of 20 to 60 Å . At this distance, half the donor molecules decay by energy transfer and half decay by the usual radiative and non-radiative rates. With 𝑘𝜏(𝑟) = 1/𝜏𝐷 obtains:
𝑅0𝐹𝑅𝐸𝑇 = [9000(𝑙𝑛10)𝜅2𝜑𝐷𝐽(𝜆) 128𝜋5𝑁𝑛4 ]
1⁄6
(19) In FRET, the acceptor, AuNPs, are estimated to be molecular with little disruption placed on it by the donor. Therefore, this energy transfer model does not describe the strong effect of
(19) In FRET, the acceptor, AuNPs, are estimated to be molecular with little disruption placed on it by the donor. Therefore, this energy transfer model does not describe the strong effect of