Thermodynamics and Mechanisms of the Interactions between

May 10, 2017 - Murphy , C. B.; Zhang , Y.; Troxler , T.; Ferry , V.; Martin , J. J.; Jones , W. E. Probing Forster and Dexter Energy-transfer Mechanis...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/Langmuir

Thermodynamics and Mechanisms of the Interactions between Ultrasmall Fluorescent Gold Nanoclusters and Human Serum Albumin, γ‑Globulins, and Transferrin: A Spectroscopic Approach Miao-Miao Yin,† Ping Dong,† Wen-Qi Chen,† Shi-Ping Xu,† Li-Yun Yang,†,‡ Feng-Lei Jiang,*,† and Yi Liu*,†,‡,§ †

State Key Laboratory of Virology & Key Laboratory of Analytical Chemistry for Biology and Medicine (MOE), College of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072, People’s Republic of China ‡ College of Chemistry and Material Science, Guangxi Teachers Education University, Nanning 530001, People’s Republic of China § College of Chemistry and Chemical Engineering, Wuhan University of Science and Technology, Wuhan 430081, People’s Republic of China S Supporting Information *

ABSTRACT: Noble metal nanoclusters (NCs) show great promise as nanoprobes for bioanalysis and cellular imaging in biological applications due to ultrasmall size, good photophysical properties, and excellent biocompatibility. In order to achieve a comprehensive understanding of possible biological implications, a series of spectroscopic measurements were conducted under different temperatures to investigate the interactions of Au NCs (∼1.7 nm) with three model plasmatic proteins (human serum albumin (HSA), γ-globulins, and transferrin). It was found that the fluorescence quenching of HSA and γ-globulins triggered by Au NCs was due to dynamic quenching mechanism, while the fluorescence quenching of transferrin by Au NCs was a result of the formation of a Au NC−transferrin complex. The apparent association constants of the Au NCs bound to HSA, γglobulins, and transferrin demonstrated no obvious difference. Thermodynamic studies demonstrated that the interaction between Au NCs and HSA (or γ-globulins) was driven by hydrophobic forces, while the electrostatic interactions played predominant roles in the adsorption process for transferrin. Furthermore, it was proven that Au NCs had no obvious interference in the secondary structures of these three kinds of proteins. In turn, these three proteins had a minor effect on the fluorescence intensity of Au NCs, which made fluorescent Au NCs promising in biological applications owing to their chemical and photophysical stability. In addition, by comparing the interactions of small molecules, Au NCs, and large nanomaterials with serum albumin, it was found that the binding constants were gradually increased with the increase of particle size. This work has elucidated the interaction mechanisms between nanoclusters and proteins, and shed light on a new interaction mode different from the protein corona on the surface of nanoparticles, which will highly contribute to the better design and applications of fluorescent nanoclusters.

1. INTRODUCTION Noble metal nanoclusters (NCs), a newly emerging type of photoluminescent nanomaterials, provide the bridge between atomic and nanoparticles behavior in noble metals.1 The nanoclusters, in particular Au NCs, show great promise as nanoprobes for bioanalysis and cellular imaging in biological applications due to ultrasmall size, good photophysical properties, and low toxicity.2−7 Many of the previous work showed that when nanoparticles (NPs) (>3 nm) were exposed to an organism, proteins from body fluids bound to their surfaces, forming a so-called “protein corona”.8−10 Consequently, living systems usually interact with protein-coated rather than bare NPs, and the structure, dynamics, and stability of the corona can be decisive factors governing the biological response of an organism to NP exposure.11−13 Meanwhile, the © 2017 American Chemical Society

fastest binders will form the initial corona in blood serum, and subsequently be replaced by proteins with the highest affinity to the surface (Vroman effect). Thus, one may initially observe a “soft” corona forming in short time scales (seconds to minutes) that subsequently evolves into a “hard” corona over periods of hours.14 However, since the size of NCs is smaller than that of the NPs mentioned above, little is known about how NCs interact with proteins in living systems.15−17 On the other hand, it was reported that protein-stabilized Au NCs had been widely used in colorimetric probes18 and biosensing applications.19−21 Consequently, it is essential to achieve a comprehensive Received: January 19, 2017 Revised: March 15, 2017 Published: May 10, 2017 5108

DOI: 10.1021/acs.langmuir.7b00196 Langmuir 2017, 33, 5108−5116

Article

Langmuir

in the transportation of numerous endogenous substances and drugs or even toxicants and nutrition, providing biomarkers for disease diagnostic treatment.12 The investigation was carried out using fluorescence, UV−vis absorption, and circular dichroism (CD) spectroscopy. This work built a bridge between the interactions of small molecules (3 nm) with proteins, which would inspire future design and enable better application of fluorescent nanoclusters.

understanding of NC−protein interactions and their potential biological implications. Recently, some studies about the interactions between proteins and nanoparticles, such as silver nanoparticles,22 quantum dots,23−28 copolymer nanoparticles,29 graphene quantum dots,30 carbon-based nanoparticles31,32 and magnetic iron oxide nanoparticles33 have been reported. They either assume that proteins and nanoparticles form complexes and interaction parameters are given from the perspective of physical chemistry (binding constants, protein conformation changes, sites, numbers of binding protein, and hydrodynamic size alteration), then deduce the interaction patterns and strength, or assume that proteins and nanoparticles generate different protein coronas, which reflect the strength of the affinity of nanoparticles binding to the proteins.31 The theoretical models are beneficial to deepen quantitative understanding of the protein coronas. The surface ligands of nanoparticles have a great influence on the interactions between nanoparticles and proteins. Our group used MPA, NAC, GSH and CA as stabilizers to synthesize CdTe quantum dots (QDs); among them, CA-CdTe QDs were positively charged and others were negatively charged. It was found that the interactions between negatively charged QDs and human serum albumin (HSA) were mainly based on the formation of a complex, whereas the interaction mechanism between the positive CA-CdTe QDs and HSA was significantly different. The interaction between negatively charged QDs and HSA occurred due to adsorption behavior, which depended on the nanoparticle itself rather than on the coating molecules, and the adsorption of HSA onto the surface of positively charged QDs would result in the aggregation of nanoparticles.26 Nyokong et al. also investigated the interactions between different thiol capped CdTe QDs and protein. They used MPA, Cys, and TGA as ligands to synthesize different CdTe QDs, and found that the binding constants varied greatly because of different surface chemical properties of CdTe QDs.34 Yu et al. synthesized CdS QDs with different capping agents (Cys, GSH, and MPA) in aqueous solutions and investigated the interactions between these CdS QDs and proteins (BSA and LZY). The interactions between these CdS QDs and proteins showed different affinities. GSH-CdS QDs were the weakest quencher of the fluorescence of BSA and LZY, while MPA-CdS QDs was the strongest.35 Sasidharan et al. gave a systematic investigation on the kinetics of time-dependent adsorption and the formation of individual protein corona with citrate and lipoic acid-coated, 40 nm-sized Au NPs and Ag NPs. They found that different ligand-modified Au NPs and Ag NPs can form protein corona of different sizes during the same period, indicating that the surface chemical properties of nanomaterials had a certain effect on the formation of protein corona.36 In short, the surface ligands of nanoparticles will directly affect their interactions with protein. Apart from physical and chemical properties of nanoparticles, the kinds of proteins and corresponding surface charge will also influence the type of their interactions.32 Thus far, however, the interaction mechanisms between serum proteins and Au NCs (1−2 nm) have rarely been reported. Herein, we presented a systematic study on the interactions between fluorescent Au NCs modified with glutathione and serum proteins including HSA, γ-globulins, and transferrin. These serum proteins, abundantly exist in plasma, and the primary physiological function of serum proteins reflects in regulating osmotic pressure and pH, aiding

2. EXPERIMENTAL SECTION 2.1. Materials. All chemicals were commercially available and used without further purification. L-Glutathione in reduced form (GSH), hydrogen tetrachloroaurate trihydrate (HAuCl4·3H2O), human serum protein (HSA), γ-globulins, and transferrin were obtained from SigmaAldrich. Sodium chloride (NaCl), potassium chloride (KCl), potassium phosphate dibasic (K2HPO4), and potassium dihydrogen phosphate (KH2PO4) were purchased from Sinopharm Chemical Reagent Co. (China). All chemicals used in this study were of analytical reagent grade. All aqueous solutions were prepared with ultrapure water (18.2 MΩ cm−1, Millipore). 2.2. The Synthesis of Au NCs. The synthesis procedure of Au NCs was according to the literature method with minor modifications.2 The freshly prepared aqueous solutions of HAuCl4 (20 mM, 0.50 mL) and GSH (100 mM, 0.15 mL) were mixed with 4.35 mL of ultrapure water at 25 °C. The reaction mixture was heated to 75 °C under gentle stirring (500 rpm) for 24 h. An aqueous solution of strongly orange-emitting Au NCs was formed. The orangeemitting Au NCs were precipitated with 3-fold amount of 2-propanol, and then resuspended in ultrapure water and precipitated with 2propanol three times. The purified Au NCs were dried overnight at 37 °C in vacuum, and the final product in the powder form could be dispersed in phosphate buffered saline (PBS) and further purified by dialysis against 0.01 mol·L−1 PBS buffer for 4 h. 2.3. Characterization. High-resolution transmission electron microscopy (TEM) images were recorded on a JEOL JEM-2100 (HR) electron microscope. The elemental composition was analyzed by X-ray photoelectron spectroscopy (XPS, Thermo Fisher, ESCALAB 250Xi) with a monochromatic Al Kα X-ray source. The the concentration of gold element was determined by inductively coupled plasma atomic energy spectrometry (ICP-AES, IRIS Intrepid II XSP). 2.4. Fluorescence Spectroscopy. Fluorescence analyses were performed on an LS-55 fluorophotometer (PerkinElmer, USA) in the ratio mode with temperature maintained by circulating bath. The excitation wavelengths of HSA, γ-globulins, and transferrin were 280 nm, and the emission wavelengths of HSA, γ-globulins, and transferrin were respectively 343, 339, and 336 nm. Each spectrum was the average of two scans. Titrations were performed by trace syringes. Synchronous fluorescence spectra were obtained at a fixed excitation wavelength (Δλ) (15 or 60) nm. The concentration of the three proteins was kept at 2 μM, and Au NCs of different concentrations were added into the solution, resulting in fluorescence quenching. 2.5. UV−Visible Absorption Spectroscopy. A UNICO 4802 UV−vis double-beam spectrophotometer was used to measure absorption spectra of Au NCs. The absorption spectra were measured with a 1 cm quartz cell. The wavelength of the spectra was measured between 190 and 500 nm. 2.6. Circular Dichroism (CD) Spectroscopy. CD measurements were performed on a circular dichroism photomultiplier (Applied Photophysics Limited, UK) at 25 °C. The CD spectra of HSA, γglobulins, and transferrin were recorded in the range of 260−190 nm. The instrument was controlled by Chirascan software. Quartz cells with path length of 0.1 cm were used, and the scanning speed was set at 200 nm min−1. Appropriate buffer solutions, measured under the same experimental conditions, were taken as blanks and subtracted from the sample spectra. Each spectrum was the average of three scans. 5109

DOI: 10.1021/acs.langmuir.7b00196 Langmuir 2017, 33, 5108−5116

Article

Langmuir

3. RESULTS AND DISCUSSION 3.1. Characterization of Au NCs. Figure 1 shows the characterization of the synthesized Au NCs. The TEM image

Au36SG32, Au39SG35, and Au43SG37 by means of gel electrophoresis. Combined with the results of XPS experiments in our work, Au36(SG)32 was used as the average particle composition. Therefore, the concentration of Au NCs is calculated by dividing the gold concentration obtained by ICP-AES by 36. As shown in Figure 1c, Au NCs exhibit a narrow and almost symmetrical emission spectrum with a strong emission peak at 588 nm; however, they do not demonstrate a distinct peak in its excitation spectrum, corresponding to its lack of characteristic absorption peaks in UV−vis absorption spectra. The fluorescence quantum yield (QY) of the Au NCs was calculated to be 1.2% with a reference of Rhodamine B (with a standard QY of 89%). 3.2. Fluorescence Quenching Mechanism and Binding Constants. Apart from the metal core quantum confinement effect, surface states may have a nonignorable effect on the photoluminescent property of metal NCs.16 Given this, we studied the photoluminescent changes of Au NCs when serum proteins were adsorbed on the surface of Au NCs. In Figure 2, the shift of the emission peak and partial photoluminescence quenching of Au NCs suggested that the Au NCs may interact

Figure 1. Characterization of the synthesized Au NCs. (a) HRTEM images of Au NCs (the scale bar in the inset is 5 nm); (b) XPS spectra of Au NCs; (c) absorption (black solid line), fluorescence excitation (red solid line), and emission spectra (blue solid line) of the synthesized Au NCs.

(Figure 1a) demonstrates that the nanoclusters are well dispersed and have uniform and spherical shapes. The corresponding histograms reveal that the particle size distributions are rather narrow, and its mean size is 1.7 ± 0.6 nm in statistical method (Figure S1). Figure 1b shows the XPS spectra of Au NCs. The XPS full spectrum clearly indicates that Au NCs contain oxygen, nitrogen, carbon, sulfur, and gold. Because Au NCs were synthesized with chloroauric acid (HAuCl4) and glutathione (GSH), so the expected cluster composition is Aux(SG)y, that is, Aux(C9H17N3O6S)y, where x and y are integers. XPS further gives the content of each element (see Table S1), the calculated content of each element is also well in line with the expected composition, namely, N:S (atomic number) = 3:1, O:S = 6:1, and C:S ≈ 10:1. The ratio of Au:S = 1.154:1 indicates that the average composition of Au NCs is Aux(SG)0.866x. In the literature,2 it is determined that the Au NCs contained five categories: Au29SG27, Au30SG28,

Figure 2. Effects of serum proteins on Au NCs fluorescence emission spectra. For panels a, b, and c, [Au NCs] = 21.6 μM. (a) [HSA] = 0, 5.4, 7.2, 10.8, 21.6, 43.2 μM; (b) [γ-globulins] = 0, 5.4, 7.2, 10.8, 21.6, 43.2 μM; (c) [transferrin] = 0, 5.4, 7.2, 10.8, 21.6, 43.2 μM. 5110

DOI: 10.1021/acs.langmuir.7b00196 Langmuir 2017, 33, 5108−5116

Article

Langmuir with these three kinds proteins. By comparing the amount of fluorescence quenching of Au NCs with the addition of these three kinds of proteins, it was found that HSA and γ-globulins had little effect on the photophysical properties of Au NCs, whereas transferrin had relatively larger effect on the fluorescence intensity of Au NCs (∼25% decrease). Next, from the perspective of physical chemistry, we quantitatively studied the strength of the interactions between Au NCs and three proteins with steady-state fluorescence techniques.29 As shown in Figure S2 (a, b, and c), both fluorescence intensity of HSA, γ-globulins, and transferrin were quenched obviously by Au NCs (λex = 280 nm, 298 K). The decreased intensity indicated that Au NCs can bind to three serum proteins. In a typical theory, the quenching mechanism is classified into dynamic quenching (collisional encounters) and static quenching (complex formation). They can be distinguished by different dependence on temperature and viscosity, or preferably by fluorescence lifetime measurements.37 In general, higher temperature results in larger diffusion coefficients and hence increased dynamic quenching; on the contrary, higher temperature usually results in the dissociation of weakly bound complexes, and decreased static quenching. The fluorescence quenching can be described by the well-known Stern−Volmer (SV) equation38 expressed as F0 = 1 + KSV[Q ] F

(1)

where F0 and F are fluorescence intensities in the absence and presence of the Au NCs, and Q is the concentration of Au NCs. Stern−Volmer constant (Ksv) is conventionally taken as a magnitude to estimate the quenching efficiency.39 In Figure 3a,b, an excellent agreement was achieved by fitting the fluorescence intensity of serum proteins (F) as a function of Au NCs concentration ([Au NCs]) with eq 1, yielding a series of quenching constants (Ksv) at different given temperatures. The increased Ksv with the elevated temperatures demonstrated that the fluorescence quenching of HSA and γ-globulins triggered by Au NCs was initiated by the dynamic collision rather than the complex formation. In Figure 3c, it could be seen that the quenching constant decreased accordingly with the temperature rising, demonstrating the moderate influence of temperature on the fluorescence quenching and the probable static quenching mode of transferrin by Au NCs. In order to further confirm the quenching mechanisms of HSA, γ-globulins, and transferrin by Au NCs, UV−vis absorption spectroscopy was employed (Figure 4). According to the theoretical models, dynamic quenching affects the excited state of the fluorophore, but does not change its UV− vis absorption spectra, so the corresponding differential absorption spectra (the difference between the absorption spectra of Au NCs together with protein and that of Au NCs alone) and UV−vis absorption spectra of protein may overlap. By contrast, ground-state complex formation will frequently result in perturbation of the absorption spectra of the fluorophore, and the corresponding differential absorption spectra may not overlap with the UV−vis absorption spectra of the protein.40 The UV−vis absorption spectra of HSA (Figure 4a, blue line) and γ-globulins (Figure 4b, blue line) and the corresponding differential absorption spectra (green line) demonstrated no obvious difference in neither the peak shape nor position in the experimental error. Herein, it can be deduced that the fluorescence quenching of HSA and γ-

Figure 3. Stern−Volmer plots for the (a) Au NCs + HSA, (b) Au NCs + γ-globulins, and (c) Au NCs + transferrin systems at different temperatures at pH 7.4.

globulins triggered by Au NCs was due to dynamic quenching mechanism. In Figure 4c, the corresponding differential absorption spectra (green line) demonstrated obvious differences in the peak intensity, which confirmed the fluorescence quenching of transferrin by Au NCs was due to the static quenching mechanism. This was consistent with the results from fluorescence quenching. The different interactions of Au NCs with these three proteins may be related to the different structures of proteins. Because γ-globulins contains five species of immunoglobulin (Ig): IgG, IgA, IgM, IgD, and IgE, and among them the main component is IgG (75%), so the structure of IgG can be approximately considered the same as that of γ-globulins. On one hand, HSA has a smaller and relatively spherical structure, and transferrin consists of two similar globular structures and has two binding sites of iron, which are located in the N and C terminus, respectively. However, γ-globulins is a relatively large globular protein (molecular mass: 150 kDa), which can provide a larger surface for the adsorption of Au NCs when increasing the concentration of Au NCs. On the other hand, there is an order for these three proteins in terms of their hydrophobic surface residues: 37 (IgG) > 29 (transferrin) > 20 (HSA), which can contribute to the adsorption of Au NCs to these three proteins.41 5111

DOI: 10.1021/acs.langmuir.7b00196 Langmuir 2017, 33, 5108−5116

Article

Langmuir

system.43,44 As shown in Figure 5a, the dependence of F0/ΔF on the reciprocal value of the quencher concentration [Q]−1

Figure 4. UV−vis absorption spectra of Au NCs, proteins, proteins/ Au NCs, (proteins/Au NCs)−Au NCs. [Au NCs] = 2 μM; (a) [HSA] = 2 μM; (b) [γ-globulins] = 2 μM; (c) [transferrin] = 2 μM.

Figure 5. (a) Modified Stern−Volmer plots of transferrin: doublelogarithm plots of (b) HSA and (c) γ-globulins.

As described above, HSA has a smaller and relatively spherical structure and can move fast in the solution, which may give opportunities to the collision of HSA and Au NCs. Although γ-globulins has a larger spherical structure, the movement is slower, and it has a large surface, which can make the collision between γ-globulins and Au NCs easier. Therefore, the interaction between Au NCs and HSA (or γ-globulins) is based on a dynamic mode. However, transferrin consists of two similar globular structures and has more tyrosine and phenylalanine residues, which may lead to π−π interactions between transferrin and Au NCs. So, the interaction between Au NCs and transferrin is based on a static mode. For the static quenching process of transferrin and Au NCs, the quenching data can be further analyzed through the modified Stern−Volmer equation: F0 1 1 1 = + ΔF fa K a [Q ] fa

was linear with the slope equal to the value of (fa Ka)−1 and the intercept equal to the value of fa−1. The corresponding associative binding constants Ka were calculated from the quotient of the intercept fa−1 and the slope (fa Ka)−1 through eq 2. As shown in Table S2, the associative binding constant also decreased with the increasing temperature, which was in accordance with the variation trend of Ksv. Since the quenching mechanism between Au NCs and HSA and γ-globulins are dynamic rather than static, a doublelogarithm equation45 was applied to determine the binding constants of Au NCs and these two proteins. F0 − F = lg K a + n lg[Q ] (3) F where Ka is the binding constant, and n is the number of binding sites per protein. Figure 5b,c show the typical doublelogarithm curves of the fluorescence quenching of HSA and γglobulins by Au NCs. 3.3. Thermodynamic Parameters of Interactions of Au NCs and Proteins. The binding forces between drugs and biomolecules usually include electrostatic interactions, hydrogen bonding, van der Waals interactions and hydrophobic lg

(2)

Herein, fa represents the mole fraction of solvent-accessible fluorophore. Ka, which represents the effective quenching constant for the accessible fluorophore,42 is analogous to the associative binding constant for the quencher−acceptor 5112

DOI: 10.1021/acs.langmuir.7b00196 Langmuir 2017, 33, 5108−5116

Article

Langmuir

and HSA and γ-globulins were driven by hydrophobic forces, and the negative ΔH and positive ΔS indicated that the interaction between Au NCs and transferrin was driven by specific electrostatic binding.47 3.4. Conformation Investigations. The interactions between Au NCs and serum proteins might cause conformation change of proteins. First, the synchronous fluorescence spectroscopy was introduced to study the microenvironment of amino acid residues by measuring the shift of emission wavelength. When the Δ-values (Δλ) between emission and excitation wavelength are stabilized at 15 or 60 nm, the synchronous fluorescence gives the characteristic information on tyrosine or tryptophan residues, respectively.48 Furthermore, tryptophan accounts for the biggest contribution to the fluorescence and is most sensitive to the ambient environment. Thus, the fluorescence peak position of synchronous fluorescence spectrum was usually used as the basis of a qualitative judgment for the changing microenvironment surrounding tryptophan residues of serum proteins. As shown in Figure 7, the position of emission peak wavelength for

forces,46 which can be elucidated by the thermodynamic parameters. The enthalpy change (ΔH) can be evaluated from the van’t Hoff equation: ⎛ ⎞ ⎜ ∂InK a ⎟ = − ΔH ⎜ ∂1 ⎟ R ⎝ T ⎠P

(4)

where Ka is analogous to the associative binding constants at the corresponding temperature, and R is the gas constant. As shown in Figure 6, the enthalpy change (ΔH) was calculated from the slope of the van’t Hoff relationship. The free energy change (ΔG) is then calculated from the following relationship: ΔG = −RT InK a

(5)

Figure 7. Synchronous fluorescence spectra of HSA in the absence and presence of Au NCs. [HSA] = 2 μM, [Au NCs] = 0−2.5 μM. (a) Δλ = 15 nm and (b) Δλ = 60 nm.

tyrosine remained constant in spite of a gradual quenching for the synchronous fluorescence spectra of HSA (Figure 7a), γglobulins (Figure S3a) and transferrin (Figure S3c) in the absence and presence of the Au NCs when Δλ was set as 15 nm. It was induced that the microenvironment of the tyrosine residue was not perturbed upon the addition of Au NCs.49 When Δλ was set as 60 nm, the maximum emission wavelength of HSA (Figure 7b), γ-globulins (Figure S3b) and transferrin (Figure S3d) had no shift, proving that the microenvironment of the tryptophan residues was also not perturbed upon the addition of Au NCs.32 Circular dichroism (CD) spectra was used to investigate the influence of Au NCs on the secondary structure of serum

Figure 6. van’t Hoff plots of Au NCs-protein system: (a) Au NCstransferrin system; (b) Au NCs-HSA system; (c) Au NCs-γ-globulins system.

The entropy change (ΔS) is calculated from eq 6: ΔG = ΔH − T ΔS

(6)

The calculated ΔH, ΔS, and ΔG values were all summarized in Table S2. Due to the negative values of ΔG, the binding interaction between Au NCs and three kinds of proteins occurred spontaneously. In addition, both the positive values of ΔH and ΔS indicated that the interactions between Au NCs 5113

DOI: 10.1021/acs.langmuir.7b00196 Langmuir 2017, 33, 5108−5116

Article

Langmuir

characteristic peptide CD bands of HSA (Figure 8a), γglobulins (Figure 8b) or transferrin (Figure 8c), which are sensitive to secondary structure, remain essentially unchanged when the concentration of Au NCs is less than 12.8 μM, indicating that Au NCs does not induce any obvious perturbations of protein fold.46 When the concentration of Au NCs is increased to 25.6 μM, there are some different degrees of changes in the CD spectra. However, compared with the reported literatures,15,31,36,41 the changes in all these CD spectra are not very obvious upon addition of Au NCs. Although Au NCs interact with these three proteins with different mechanisms, Au NCs have weak effects on the secondary structures of all these three proteins, as evidenced by the CD spectra, indicating the excellent biocompatibility of Au NCs for biomedical applications. 3.5. Comparison of the Binding Constants of Small Molecules, NCs, and Nanoparticles Interacting with Serum Albumin. There are a lot of research work on the interactions between serum albumin and small molecules and large-sized quantum dots. However, as a bridge between atoms and nanoparticles, noble metal nanoclusters have relatively less studied on their interactions with serum albumin. To sum up, as shown in Table 1, it is found that when small molecules (berberine, palmatine, and rotenone) interact with HSA, the magnitude of Ka is 104. With the particle size is gradually increased, the magnitude of Ka is also gradually increased to 106, showing a stronger binding, which may be due to the different binding ways of particles to serum albumin. The hydrated particle size of serum albumin is about 10 nm, and when interacting with large-sized particles, serum albumin may be adsorbed on the particle surface to form the so-called protein corona, but when the particle size is less than 2 nm, the particles may bind to serum albumin with a totally different mode. Particles entering into the protein cavity is more difficult than adsorption on the protein surface, so the larger the particle size, the stronger the interaction. Compared with the small molecule berberine, the noble metal nanoclusters have a stronger combination with HSA because nanoclusters contain much more units (∼tens to hundreds), which can all act as quenching units. Similarly, the binding of noble metal nanoclusters to HSA is weaker than that of larger quantum dots. HSA is the most abundant protein in the human body. When the nanoparticles are strongly bound to HSA, it is likely that the metabolism of the organism will be disordered and even disease will be caused. Therefore, the relatively weak interaction of Au NCs and HSA may make it very useful in medical treatment and biological research. In short, nanoclusters have the advantages of high fluorescence stability, strong biocompatibility, and large Stokes shift, so they can be well applied in biosensing, labeling, imaging, etc.

protein. Figure 8 showed typical CD spectra of these three kinds of proteins in the presence and absence of Au NCs.

Figure 8. CD spectra of (a) HSA, (b) γ-globulins and (c) transferrin in the presence of different concentrations of Au NCs. [Au NCs] = 0, 0.8, 1.6, 3.2, 6.4, 12.8, 25.6 μM. [HSA] = [γ-globulins] = [transferrin] = 5 μM.

Consistent with the literature, the CD spectra of HSA exhibited two negative bands in the UV region at 210 and 220 nm, γglobulins showed two negative bands at 218 and 231 nm, and transferrin also had two negative bands at 208 and 217 nm. The

Table 1. Stern−Volmer Quenching Constants Ksv and Association Constants Ka of Interactions of Small Molecules, NCs, and Nanoparticles with HSA (T = 298 K) quenchers

size (nm)

ligand

berberine palmatine rotenone Au NCs CdTe QDs