- Email: [email protected]
Adsorption of proteins on gold nanoparticles: One or more layers?
Accepted Manuscript Title: Adsorption of proteins on gold nanoparticles: One or more layers? Authors: Dmitriy V. Sotnikov, Anna N. Berlina, Vladislav S. Ivanov, Anatoly V. Zherdev, Boris B. Dzantiev PII: DOI: Reference:
S0927-7765(18)30722-7 https://doi.org/10.1016/j.colsurfb.2018.10.025 COLSUB 9709
To appear in:
Colloids and Surfaces B: Biointerfaces
Received date: Revised date: Accepted date:
19-6-2018 4-10-2018 9-10-2018
Please cite this article as: Sotnikov DV, Berlina AN, Ivanov VS, Zherdev AV, Dzantiev BB, Adsorption of proteins on gold nanoparticles: One or more layers?, Colloids and Surfaces B: Biointerfaces (2018), https://doi.org/10.1016/j.colsurfb.2018.10.025 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 of 16
Adsorption of proteins on gold nanoparticles: One or more layers?
Dmitriy V. Sotnikov, Anna N. Berlina, Vladislav S. Ivanov, Anatoly V. Zherdev, Boris B. Dzantiev *
[email protected], [email protected]. Tel.: +7-495-9543142
SC R
*Correspondence:
IP T
A.N. Bach Institute of Biochemistry, Research Center of Biotechnology of the Russian Academy of Sciences, Leninsky prospect 33, Moscow 119071, Russia;
The composition of conjugates of gold nanoparticles with proteins was analysed pH influences significantly on the adsorption of proteins on gold nanoparticles Both monolayer and multilayer structures from protein molecules are formed Explanation for the transition from monolayer to multilayer structures was proposed
A
CC E
PT
Highlights
ED
M
A
N
U
Graphical Abstract
2 of 16
SC R
IP T
Abstract: Adsorption of proteins on nanoparticles is a complex and poorly studied process. The mechanisms of protein layer formation can fundamentally differ depending on the composition of the medium, the nanoparticles’ structure, the protein’s nature, and other factors. In particular, monolayer or multilayer immobilization may occur. In the present work, the composition of conjugates of bovine serum albumin and immunoglobulin G with gold nanoparticles obtained by the Turkevich-Frens method are analyzed. The composition was studied by protein fluorescence measurement for particles ranging in size from 20–48 nm, depending on the pH of the immobilization medium (from 4–5 to 8–10). It was found that a pH shift of the immobilization medium from acidic to alkaline values is accompanied by a change in the mechanism of protein adsorption on the gold surface. In acidic pH conditions (4–5), effective binding of bovine serum albumin and gold nanoparticles occurs throughout the entire range of studied protein concentrations. In alkaline pH conditions (8-10), however, effective binding occurs only at concentrations of >10 μg/mL. This effect is not observed for immunoglobulin G, which is efficiently adsorbed onto nanoparticles throughout the entire range of studied concentrations and pH values. For acidic pH values, the surface of the particles is saturated with the amount of bound proteins, which approximately corresponds to the amount the monolayer is filled. For neutral and alkaline pH values, saturation is not observed and the amount of adsorbed protein certainly exceeds the monolayer filling, leading to multilayer immobilization.
N
U
Keywords: gold nanoparticles; protein conjugates; tryptophan fluorescence; bovine serum albumin; immunoglobulin G
M
A
1. Introduction
A
CC E
PT
ED
Gold nanoparticle (GNP) and their complexes with proteins are used for many analytical and therapeutic tasks [1-3]. The composition, structure and functional properties of protein nanoconjugates were characterized in a number of studies using various physical, chemical and biochemical methods [415]. Despite the advances in this field, the interaction of proteins and GNPs still raises many unknown questions. Many works contain data that, at first glance, may seem contradictory. In [11, 12, 16-18], it is claimed that proteins form a monolayer on the surface of GNP, and in [19, 20], it was observed that protein adsorption on GNPs continues at protein concentrations beyond those that fill the monolayer. The formation of multilayer structures from adsorbed proteins on the surface of nanoparticles is described in the literature by the model of so-called “hard” and “soft” protein coronas [21-27]. De Roe et al. [28] showed that protein A interacts with GNPs obtained by the citrate method with an equilibrium dissociation constant of 343.9 nM. However, Ghitescu et al. [29] stated that the dissociation constant at low protein A concentrations is 2–3 nM and at high concentrations increases to 500–900 nM, although both groups used the same radioactive labeling method. Differences in the interaction constants for proteins and GNPs can reach five orders of magnitude [16, 28, 30-32]. For example, for bovine serum albumin (BSA) conjugate and spherical GNP, the dissociation constant values, determined by quenching of tryptophan fluorescence, range from 2.7·10-5 M [33] to 1·10-9 M [30]. These differences can be explained by some methods’ insufficient accuracy, possible changes in the structure of conjugates during the experiments, and use of un-unified nanoparticles obtained by different techniques or under different conditions [10, 20, 34]. The interaction of macromolecules with GNP depends on the chemical nature of the adsorbed material, the composition of the medium, and other factors. Electrostatic, hydrophobic, van der Waals, and donor–acceptor interactions can contribute to this process [35, 36]. Further, gold is known to have an affinity for sulfur-containing chemical groups [37]. Due to the presence of many chemical groups with different natures in proteins, the interaction of proteins with the surface of nanoparticles occurs through several binding centers and mechanisms. However, in most studies, GNP–protein complexes are compared by varying a single
3 of 16
parameter, such as the size of nanoparticles. The search for more general patterns of protein coat formation on GNP surfaces requires multifactor analysis of conjugates that differ in terms of particle size, the nature of the protein to be adsorbed and its concentration, and the composition of the reaction medium. This study examined the GNPs that are most commonly used in practice: those synthesized by the citrate method with a diameter from 20 to 50 nm. Two proteins in the GNP conjugates—BSA and immunoglobulin G (IgG)—were examined as they are used for the functionalization and stabilization of nanoparticles in most studies. Conjugate composition was determined by protein concentration during synthesis and the pH of the immobilization medium. The study was carried out using a previously developed technique based on fluorescence measurements [38].
IP T
The obtained data enable identification of immobilization patterns of proteins on GNPs and explain some differences in the results of previous studies.
2. Materials and Methods
SC R
2.1. Reagents
Tetrachloroauric acid and sodium citrate were obtained from Sigma (USA); BSA was obtained from MP Biomedicals (USA); and goat anti-mouse IgG was obtained from Arista Biologicals (USA). All salts were analytically or chemically pure. Water was purified using a MilliQ system (Millipore, USA).
U
2.2. Preparation of gold nanoparticles
A
N
First, 0.2 mL of a 5% solution of tetrachloroauric acid was added to 97.5 ml of distilled water. The mixture was then brought to a boil. After that, 1, 1.5, or 2 ml of a 1% solution of sodium citrate was added. The mixture was boiled for 30 min, and then cooled to room temperature [39]. The resulting preparation was stored at 4 °C.
M
2.3. Determination of GNP size by transmission electron microscopy
PT
ED
GNP preparations were applied to 200-mesh hexagonal copper meshes coated with a pharmacopolymer film. A CX-100 transmission electron microscope (Jeol, Japan) with an accelerating voltage of 80 kV was used. Micrographs were taken at a magnification of 33,000X. Film negatives were scanned in grayscale on a scanner to obtain 1,200 dpi images. These images were saved as TIFFs and transferred to an image-processing program (Image Tool, University of Texas Health Science Center at San Antonio, USA). Particle size and the size distribution were then calculated according to the method used by Safenkova et al. [40]. 2.4. Preparation of protein and GNP conjugates at different pH values
A
CC E
The GNP solution was adjusted to the desired pH with 0.2 M Na2CО3, which was poured into 8 tubes of 1 mL and centrifuged at 13.4 rpm for 10 min. Then, the supernatant was separated from the sediment. The gold nanoparticle sediment was shaken and adjusted to 125 μL. Proteins were diluted to concentrations from 2 to 250 μg/mL using the GNP preparation supernatant. Only freshly prepared solutions were used in the experiments. Then, 125 μL of the resulting protein solutions were added to 125 μL of the centrifuged GNP solutions. The remaining protein solutions were used as calibration solutions. The protein was incubated with GNP for 1 h at room temperature and centrifuged at 10,000 g (20 min), and then 200 μL of supernatant was taken and transferred to microplate wells. The calibration solutions (200 μL aliquots) were also transferred to the microplate wells. Since microplates adsorb proteins, affecting fluorescence, the measurements were taken immediately after transferring the samples to the wells. 2.5. Fluorescence measurements The fluorescence spectra of proteins were measured by a Perkin Elmer En Spire 2300 microplate spectrophotometer (Waltham, MA, USA) in Nunc MaxiSorp white microplates (Roskilde, Denmark) at an excitation light wavelength of 280 nm and an emitted light wavelength of 350 nm.
4 of 16
2.6. Characterization of the proteins solutions by asymmetrical flow field-flow fractionation (AF4) The fractionations were performed by a Wyatt Eclipse 3+ Separation System (Wyatt Technology, USA) with an autosampler and pump (Agilent Technologies, USA). A 5-kDa MW cut off regenerated cellulose membrane (Microdyn-Nadir, Germany) was used for separation in the 275-mm channel with a 350-μm thick spacer. The channel was sequentially connected with a UV/VIS detector (Agilent Technologies, USA), Dawn HELEOS II multi-angle light scattering detector and Optilab T-Rex refractometer (both from Wyatt Technology, USA).
IP T
Solutions of BSA (1 mg/mL) and IgG (1 mg/mL) in 10 mM Na-citrate buffer, pH = 4.0, 50 mM Kphosphate buffer, pH = 7.3 (PBS), and 10 mM Na-carbonate buffer, pH = 9.0 were prepared. 80 μL of the solutions were loaded in the AF4 system at a rate of 0.2 mL/min. The focusing time during the loading was 2 min. The carrier fluids were the same as the buffers for the preparation of of BSA and IgG solutions. The main flow rate in the channel was 1 mL/min, and the focus flow rate was 1 mL/min. Separation was performed at a linear rate gradient of the cross flow from 5 to 0.1 mL/min (10 min).
SC R
The registered data were analyzed using ChemStation v.B.04.03 (Agilent Technologies, USA) and Astra v.6.1.1.17 (Wyatt Technology, USA) software.
3. Results and Discussion
N
3.1. Synthesis and determination of the size of GNPs
U
The described above AF4 measurements were conducted using the Shared-Access Equipment Centre “Industrial Biotechnology” of the Federal Research Centre “Fundamentals of Biotechnology” Russian Academy of Sciences (A.N. Bach Institute of Biochemistry RAS), Moscow, Russia.
M
A
To synthesize GNPs of different diameters, three concentrations of a reducing agent were used. Each synthesis was reproduced twice, resulting in six GNP preparations. Their dimensional characteristics were evaluated by transmission electron microscopy. According to the obtained data, the synthesized GNP preparations (Nos. 1–6) had average particle diameters of about 20, 22, 24, 27, 48, and 40 nm, respectively. Table 1 summarizes the micrograph analysis of the GNP preparations. The average deviation in particle diameter for each preparation did not exceed 25%.
ED
3.2. Determination of the GNP concentration
CC E
PT
GNP concentration was determined based on the following method. Assuming that the particles are spherical, the volume of one particle (Vp) is found using the data on particle sizes obtained from electron micrographs. Taking into account the density of gold (ρ = 19.3 g/cm3), we obtain the mass of one particle (mp = ρ ∙ Vp). During synthesis, 0.2 mL of 5% HAuCl4 was added to 100 mL of water. Accordingly, the concentration of gold is 0.05 mg/mL, and the number of particles in one milliliter is n = C / m p. The resulting particle concentrations are given in Table 1. Before conjugation, GNPs were precipitated and the precipitate was re-dissolved in a smaller volume, increasing the GNP concentration four-fold. 3.3. Preparation and characterization of GNP–protein conjugates
A
BSA and IgG were selected for analysis in this study because IgG is the most common receptor molecule in bioanalytical systems and BSA is most often used as a carrier protein for haptens and as a stabilizer for colloidal particles. In addition, BSA and IgG are the main protein components of blood, and when nanoparticles enter the body, they interact mainly with these proteins. Thus, studying these interactions is of great importance for the fields of medicine and toxicology. GNP preparations 1, 3, and 5 were conjugated with BSA, whereas preparations 2, 4, and 6 were conjugated with IgG. The method used to determine the composition of conjugates is based on measurement of the intrinsic fluorescence of proteins. It does not require any labels, which simplifies the analysis, reduces
5 of 16
errors, and enables study of the conjugation of GNPs with native proteins. Most proteins have intrinsic fluorescence, which is mainly due to tryptophanic residues. Tryptophan features maximum light absorption at 280 nm, and maximum emission at 340–360 nm [41]. The principle of this technique is described in detail in [38]. Briefly, it involves comparing the initial fluorescence of the calibration solutions (F0) and the residual fluorescence of the reaction solutions after separation of the synthesized conjugates (F). The difference between these two quantities (ΔF) is proportional to the amount of protein in the conjugate:
Ck = C0·ΔF/F0,
(1)
IP T
where C0 is the added protein concentration and Ck is the concentration of proteins bound to GNP in the solution.
SC R
This technique excludes the influence of GNP on tryptophan fluorescence. Since GNP is a strong modulator of fluorophore properties [42], its presence in the solution changes the value of fluorescence and does not allow for comparison of fluorescence between samples and calibration solutions. For this reason, the accuracy of measurements largely depends on the completeness of removal of gold particles from the solution. 3.4. Determination of the amount of protein adsorbed on GNP
N
U
GNP solutions synthesized by the citrate method have an acidic or slightly acidic pH (3–6) depending on particle size. For larger-sized GNP, pH is decreased due to the lower sodium citrate content, which shifts the pH towards neutral values (the measured pH value of a 1% sodium citrate solution is 7.3). The pH of GNP solutions was adjusted by adding an alkaline 0.2 M Na2CO3 solution.
ED
M
A
After synthesis and separation of conjugates by centrifugation, 200 μL aliquots of supernatant were placed in microtiter plate wells. Calibration solutions (also 200 μL) were placed in wells of the same plate, and fluorescence spectra were measured. Based on the measured spectra, the dependences of fluorescence intensity on the concentration of the added protein at 350 nm were plotted for the calibration solutions and samples. At protein concentrations of 0–250 μg/mL, the calibration dependences are practically linear (R linearization factor is >0.99 for both proteins).
PT
Based on the difference in fluorescence of the samples and calibration solutions and the data concerning particles’ content during conjugation (see Table 1), the number of protein molecules on one particle was calculated according to the following formula:
[(F0 – F)/F0] × ([L0]/[R0]) = RL,
(2)
CC E
where RL is the number of protein molecules bound to GNPs, F0 is the fluorescence of the calibration solution, F is the fluorescence of the supernatant after centrifugation of the synthesized conjugate, [L0] is the molar concentration of the added protein, and [R0] is the molar concentration of GNPs.
A
In some cases, during the synthesis of conjugates, aggregation was observed, which was accompanied by a change in the solution color from red to blue: • Preparations of GNPs sized 48 nm aggregate when alkalinized to pH = 8 (data not shown); • A slight discoloration after alkalization of GNPs sized 22, 27, and 40 nm at pH = 10 (see Section
3.6). • Significant aggregation of conjugates with a low content of adsorbed protein after centrifugation. However, because no residual protein was observed in the supernatant, the protein was adsorbed completely on the particles, and aggregation under these conditions did not affect our conclusions. For all other cases, the conjugates were stable, and discoloration of the particles after protein addition was not observed. Therefore, we state that the processes of GNP aggregation were nonsignificant for the article’s conclusions. This is in agreement with the results of experiments on the
6 of 16
adsorption of proteins at different GNP concentrations; as well, in the case of aggregation, it was enforced by the increase in the GNP concentration and it caused non-linearity in the dependence of the amount of adsorbed protein on the concentration of nanoparticles (See section ‘Studying on the dependence of adsorbed protein quantity from GNP concentration’ in the Supplementary material). 3.5. Examination of the composition of GNP–BSA conjugates
SC R
IP T
Data concerning BSA fluorescence in calibration solutions and in the supernatant after centrifugation of the conjugates with average GNP diameters of 20 are shown in Fig. 1 (the same data for 24, and 48 nm GNPs are shown in Supplementary material: Fig. S1–S2, respectively). Measurements were carried out in triplicate for each type of conjugate. As shown in the figures, the dependence on BSA concentration differs significantly for different pH values. At an acidic pH (4-5) with a low concentration of added BSA, the amount of unbound protein molecules does not differ from the background; that is, almost all the added protein was adsorbed on the GNP surface. The parallel course of the fluorescence curves of the calibration solutions and samples at high protein concentrations indicates saturation of all sorption centers on the surface of nanoparticles. In neutral (pH 7) and alkaline (pH 8) media, BSA fluorescence in the supernatant liquid (protein concentrations of less than 20 μg/mL) is much higher than in acidic medium and the gold surface is not saturated. The obtained dependences are similar for all three investigated GNP preparations (average diameters of 20, 24, and 48 nm). The 48 nm preparation was not stable after alkalization and centrifugation of the solution, which made it impossible to determine the composition of its conjugate with BSA at a pH of 8.
CC E
PT
ED
M
A
N
U
Based on the fluorescence differences between the calibration solutions and supernatant, the amount of BSA adsorbed on one particle was calculated using formula (2). The obtained data concerning conjugates’ composition (Fig. 2 and Fig. S3 in Supplementary material) demonstrate that in acidic pH conditions effective binding of BSA to particles occurs at all studied protein concentrations, but in alkaline pH conditions effective binding occurs only at concentrations greater than 10 μg/mL. At BSA concentrations of ~1–80 μg/mL (for GNPs with an average diameter of 20 nm (Fig. 2)), ~1–60 μg/mL (for GNPs with an average diameter of 24 nm (Fig. S3A)), and ~1–25 μg/mL (for GNPs with an average diameter of 48 nm (Fig. S3B)), more BSA is absorbed by GNPs in acidic solutions (pH 4–5) than in alkaline ones and at higher protein concentrations – more BSA is absorbed in alkaline solutions. These effects can be explained by the fact that, in an alkaline environment, the negative charge of BSA molecules increases (pI in water at 25 °C is 4.7) and, accordingly, their repulsion from GNP particles with negatively charged surfaces is enhanced [2]. This creates an electrostatic barrier that can be overcome only by protein molecules with sufficient kinetic energy, the number of which increases with increasing concentration. For larger particles (diameter of 48 nm), in neutral pH conditions, this barrier practically stops BSA adsorption on GNPs at concentrations below 8 μg/mL (Fig. S3B). However, for particles with diameters of 20 and 24 nm, this effect takes place only at a pH of 8. In alkaline solutions, the GNP surface is not saturated, even at BSA concentrations of 250 μg/mL, whereas in acidic solutions, saturation occurs at concentrations of about 60 μg/mL (for GNPs with diameters of 20 and 24 nm) and 16 μg/mL (for GNPs with a diameter of 48 nm). Thus, at BSA concentrations <60 μg/mL more BSA is adsorbed in an acidic medium than in an alkaline medium and at higher BSA concentrations – more BSA is adsorbed in an alkaline medium. Additional studies to understand these effects and their mechanism further will be necessary.
A
The dependence of the maximum amount of adsorbed BSA from the pH of the immobilization medium (Fig. S4) demonstrates that the least amount of protein is adsorbed in an acidic medium. 3.6. Examination of the composition of GNP–IgG conjugates IgG adsorption on GNP was studied in a similar manner to the described above study of BSA adsorption. The fluorescence of IgG in calibration solutions and the supernatant after centrifugation of the conjugates is shown in Fig. S4–7 (Supplementary material). The obtained dependences are somewhat different from those for BSA, but a lack of saturation in alkaline pH conditions is also observed.
7 of 16
The obtained dependences of the number of IgG molecules adsorbed on one particle from the initial protein concentration in solutions with different pH values are shown in Fig. 3 and Fig S8 (Supplementary material). Unlike for BSA, there is no significant decrease in binding at the initial sections of the curves in alkaline solutions. This may be due to the much higher values of the isoelectric points of IgG molecules compared to BSA (the pI of IgG ranges from 6.4–9.0 depending on amino acid composition and degree of glycosylation [43]). For this reason, it is difficult for IgG to reach a sufficient negative charge to prevent contact with the GNP surface. At an IgG concentration of ~60 μg/mL and a pH of 4–5, the amount of adsorbed protein reaches a plateau, whereas in neutral and alkaline solutions, adsorption continues until a concentration of 250 μg/mL.
SC R
IP T
The maximum amount of adsorbed IgG on GNPs with average diameters of 22, 27, and 40 nm, depending on the pH of the immobilization solution, are shown in Fig. S9 (Supplementary material). The highest number of protein molecules on one GNP is reached at a pH of 8. In addition, the observed decrease in adsorption at higher pH values may be due to an increase in the concentration of sodium carbonate, which leads to a decrease in the total surface area of nanoparticles and their aggregation. 3.7. Calculating the surface area occupied by proteins on GNP surface
U
The maximum quantities of protein molecules adsorbed on a single particle are shown in Table 2. By dividing the surface area of a particle by the amount of adsorbed proteins, we can determine the area occupied by one protein molecule.
M
A
N
Calculation of the area occupied by one protein molecule on the surface of the particle demonstrates that, at the maximum load, the occupied area corresponds to a circle with a diameter of about 3 nm for BSA and 3.5 nm for IgG. This is considerably smaller than the dimensions of BSA and IgG molecules (Fig. S10) and the linear size of one Fc or Fab fragment of IgG, leading to the conclusion that multilayer immobilization occurs for IgG on the GNP surface in alkaline pH conditions. That is, depending on the pH of the immobilization and protein concentrations in a solution, both monolayer and multilayer adsorption can be observed. This observation explains differences in the mechanisms of adsorption and structures of conjugates (see Introduction).
CC E
PT
ED
For a fully filled monolayer, the calculated surface area of a nanoparticle per one BSA molecule is 13, 17, and 35 nm2 for GNPs with diameters of 20, 24, and 48 nm, respectively. The surface areas occupied by one IgG molecule are 15, 18, and 35 nm2 for GNPs with diameters of 22, 27, and 40 nm, respectively. The increase in the surface area occupied by one protein molecule can be due primarily to two factors. First, on the more flattened surface, the impact of steric hindrances created by neighboring molecules increases. Second, a large deformation of macromolecules occurs, and, consequently, a larger area of the occupied surface is usually observed on larger particles [44]. This statement is confirmed as well by an increase in the binding constant of the molecule–nanoparticle interaction due to more contact of the chemical groups of the protein with the surface [29].
A
3.8. Study of protein aggregation at different pH values The shift from monolayer to multilayer protein immobilization can be either a consequence of a change in the sorption properties of nanoparticles or spontaneous aggregation of proteins with an increase in the pH of the medium. To investigate this issue, we analyzed the preparations of BSA and IgG in solutions with different pH values by the AF4 method. BSA and IgG solutions with a protein concentration of 1 mg/ml were investigated in three buffer solutions with pH 4.0, 7.3 and 9.0. Fig. S11 presents the outputs of protein fractions from the capillary for AF4. Monomeric BSA and IgG pass through the capillary for 8-10 and 13-14 min, respectively. The presence of aggregates in the preparations leads to the appearance of additional peaks with a release time of 10-12 min for BSA and 20-22 min for IgG, respectively.
8 of 16
IP T
For BSA, an increase in the pH leads to a decrease in the time of release of the protein fraction and a decrease in the aggregates peaks intensities (see Fig. S11A). Thus, increased pH reduces aggregation of BSA. For IgG, the release time slightly increases with decreasing pH, whereas the value of the aggregates remains practically unchanged (Fig. S11B). Thus, aggregation of IgG preparations varies insignificantly. This facts mean that multilayer sorption of the proteins on gold at high pH values is due to changes in the properties of the nanoparticles, and not to the state of the protein molecules in the solution. Note that the AF4 studies were implemented for the concentration of proteins equal to 1 mg/ml for better accuracy of measurements. The concentration of proteins used in the preparation of nanoconjugates was an order of magnitude lower than the concentration for AF4 studies. So the degree of aggregation of the preparations used for nanoconjugates was much lower than that of the preparations used in the AF4 experiments. 3.9. Modeling of GNP-protein conjugate formation
SC R
By summarizing the obtained data and common notions about the structure of GNPs in solution [45], we can suggest a model for the formation of conjugates under different conditions.
U
The surface of a gold particle is charged negatively (the surface potential is about –50 mV) (Fig. S13). Cations create the next layer around the anionic layer and form a double ionic shell around the GNP, which generates repulsive forces between different GNPs. This shell determines the stability of the colloidal solution.
ED
M
A
N
At a pH below the isoelectric point of the adsorbed protein, the positively charged protein molecule is attracted to the surface of the particle and is easily conjugated to it until a monolayer is formed. At a pH above the isoelectric point, on the contrary, the negatively charged protein molecule repels from the surface of the particle. In this case, protein binding to the surface is possible due to donor–acceptor, van der Waals, and other interactions, if the kinetic energy of the protein molecule is sufficient to overcome the electrostatic barrier. Therefore, electrostatic repulsion only leads to a deceleration of adsorption and a decrease in the binding strength, but it does not prevent adsorption. According to experimental data, this effect is more pronounced for BSA with an isoelectric point in the acidic region (pI 4.7–6.0) than for IgG (pI 6.4–9.0). According to Fig. 3, the effect is also observed for IgG, but the influence of the alkaline medium is weakened, as IgG acquires a smaller negative charge than BSA.
CC E
PT
With an increasing pH, the negative charge on the surface of the particles increases. Respectively, the layer of counter ions near the particle increases as well. In addition, the adsorption of the first layer of protein leads to an even greater increase in the negative charge near the surface and an additional expansion of the counter ion layer. As a result, negatively charged protein molecules can electrostatically attach to an increased cationic “coat” (Fig. S14). The interaction with the cationic “coat” is the likely reason for the appearance of the protein poly-layers, as according to AF4, an increase in pH does not enhance protein–protein interactions.
4. Conclusions
A
It is important to develop a general theory about protein and nanoparticle conjugation. To do so, researchers need data on conjugates’ composition when several parameters are varied. However, most studies on conjugates have varied one or two parameters [16, 29, 31, 46-50]. This study performs multifactorial analysis of GNP conjugates’ composition, varying particle size, the nature of the protein, protein concentration, and the pH of the immobilization medium. The obtained results contribute new data about the formation of protein layers on the surface of GNPs, which are undoubtedly of interest to researchers. In particular, this study showed the effect of the pH of the immobilization medium on the adsorption of proteins on GNPs: when pH varies from 4–5 to 8–10, the maximum amount of adsorbed protein molecules on a GNP increases, possibly due to the transition from monolayer to multilayer protein immobilization. The AF4 study of proteins solutions has shown that pH-dependent changes of the amount of adsorbed proteins are not caused by aggregation of the proteins.
9 of 16
Differences in the adsorption of BSA and IgG on GNPs were found. The shift in the pH of the immobilization medium from acidic to alkaline values is accompanied by a change in the adsorption mechanism. In acidic pH conditions, effective binding of BSA to nanoparticles occurs at all protein concentrations, but in alkaline pH conditions, effective binding occurs only at concentrations of >10 μg/mL. IgG is adsorbed effectively at all studied concentrations and pH values.
IP T
Significant differences in the adsorption of proteins with different isoelectric points were considered by Geoghegan and Ackerman [51]. They studied the stabilization of gold particles by different proteins and found that the pH dependence of the stabilization of conjugates in solutions with high ionic strength is U-shaped for proteins with an isoelectric point in the acidic region and L-shaped for proteins with an isoelectric point in the alkaline region. However, this study was implemented at a fixed protein concentration and needs additional data. Our results (see Fig. S3) indicate that the pH dependence of BSA adsorption on GNP 24 nm is U-shaped for concentrations in the range of ~50–150 μg/mL. However, the form of the dependence changes for other concentration ranges. This example confirms the need for a multifactor analysis of protein adsorption on nanoparticles to describe this process fully.
SC R
These observations explain the differences in the results obtained by different groups of researchers regarding GNP–protein conjugates.
U
Acknowledgments:
N
The authors are grateful to Irina V. Safenkova (A.N. Bach Institute of Biochemistry) for her assistance in AF4 studies and interpretation of the obtained data.
ED
M
A
This study was financially supported by the Ministry of Science and Education of the Russian Federation (grant agreement No. 14.613.21.0080 on 22.11.2017, unique identifier RFMEFI61317X0080).
PT
References
1. S. Zeng, K.T. Yong, I. Roy, X.Q. Dinh, X. Yu, F. Luan. A review on functionalized gold nanoparticles for biosensing applications. Plasmonics, 6 (2011), 491-506, DOI: 10.1007/s11468-011-9228-1.
CC E
2. K.E. Sapsford, W.R. Algar, L. Berti, K.B. Gemmill, B.J. Casey, E. Oh, M.H. Stewart, I.L. Medintz. Functionalizing nanoparticles with biological molecules: developing chemistries that facilitate nanotechnology. Chem. Rev., 113 (2013), 1904-2074, DOI: 10.1021/cr300143v.
A
3. O.V. Salata. Applications of nanoparticles in biology and medicine. J. Nanobiotechnol. 2, 3 (2004), DOI: 10.1186/1477-3155-2-3. 4. K.E. Sapsford, K.M. Tyner, B.J. Dair, J.R. Deschamps, I.L. Medintz. Analyzing nanomaterial bioconjugates: A review of current and emerging purification and characterization techniques. Anal. Chem., 83 (2011), 4453-4488, DOI: 10.1021/ac200853a. 5. N.C. Bell, C. Minelli, A.G. Shard. Quantitation of IgG protein adsorption to gold nanoparticles using particle size measurement. Anal. Methods, 5 (2013), 4591-4601, DOI: 10.1039/C3AY40771C.
10 of 16
6. N.G. Khlebtsov, V.A. Bogatyrev, B.N. Khlebtsov, L.A. Dykman, P. Englebienne. A multilayer model for gold nanoparticle bioconjugates: application to study of gelatin and human IgG adsorption using extinction and light scattering spectra and the dynamic light scattering method. Colloid J., 65 (2003), 622-635, DOI: 10.1023/A:1026140310601. 7. S. Neupane, Y. Pan; S. Takalkar, K. Bentz, J. Farmakes, Y. Xu, B. Chen, G. Liu; S.Y. Qian, Z. Yang. Probing the aggregation mechanism of gold nanoparticles triggered by a globular protein. J. Phys. Chem. C, 121 (2017), 1377-1386, DOI: 10.1021/acs.jpcc.6b11963.
IP T
8. M.J. Pollitt, G. Buckton, R. Piper, S. Brocchini. Measuring antibody coatings on gold nanoparticles by optical spectroscopy. RSC Adv., 5 (2015), 24521-24527, DOI: 10.1039/C4RA15661G. 9. S. Liu, J. Horak, M. Höldrich, M. Lämmerhofer. Accurate and reliable quantification of the protein surface coverage on protein-functionalized nanoparticles. Anal. Chim. Acta, 989 (2017), 29-37; DOI: 10.1016/j.aca.2017.08.004.
SC R
10. A.M. Davidson, M. Brust, D.L. Cooper, M. Volk. Sensitive analysis of protein adsorption to colloidal gold by differential centrifugal sedimentation. Anal. Chem., 89 (2017), 6807–6814, DOI: 10.1021/acs.analchem.7b01229.
U
11. S. Zhang, Y. Moustafa, Q. Huo. Different interaction modes of biomolecules with citrate-capped gold nanoparticles. ACS Appl. Mater. Interfaces, 6 (2014), 21184-21192, DOI: 10.1021/am506112u.
A
N
12. L. Shang, G.U. Nienhaus. In situ characterization of protein adsorption onto nanoparticles by fluorescence correlation spectroscopy. Acc. Chem. Res., 50 (2017), 387-395, DOI: 10.1021/acs.accounts.6b00579.
ED
M
13. K.E. Woods, Y.R. Perera, M.B. Davidson, C.A. Wilks, D.K. Yadav, N.C. Fitzkee. Understanding protein structure deformation on the surface of gold nanoparticles of varying size. J. Phys. Chem. C, 120 (2016), 27944-27953, DOI: 10.1021/acs.jpcc.6b08089. 14. A. Wang, K. Vangala, T. Vo, D. Zhang, N.C. Fitzkee. A three-step model for protein–gold nanoparticle adsorption. J. Phys. Chem. C, 118 (2014), 8134-8142, DOI: 10.1021/jp411543y.
PT
15. Q. Shao, C.K. Hall. Protein adsorption on nanoparticles: model development using computer simulation. J. Phys. Condens. Matter, 28 (2016), 414019, DOI: 10.1088/0953-8984/28/41/414019.
CC E
16. S. Dominguez-Medina, S. McDonough, P. Swanglap, C.F. Landes, S. Link. In situ measurement of bovine serum albumin interaction with gold nanospheres. Langmuir 28 (2012) 9131-9139, DOI: 10.1021/la3005213.
A
17. E. Casals, T. Pfaller, A. Duschl, G.J. Oostingh, V. Puntes. Time evolution of the nanoparticle protein corona. ACS Nano, 4 (2010), 3623-3632, DOI: 10.1021/nn901372t. 18. D.-H. Tsai, F.W. DelRio, A.M. Keene, K.M. Tyner, R.I. MacCuspie, T.J. Cho, M.R. Zachariah, V.A. Hackley. Adsorption and conformation of serum albumin protein on gold nanoparticles investigated using dimensional measurements and in situ spectroscopic methods. Langmuir 27 (2011) 24642477, DOI: 10.1021/la104124d. 19. S.H.D.P. Lacerda, J.J. Park, C. Meuse, D. Pristinski, M.L. Becker, A. Karim, J.F. Douglas. Interaction of gold nanoparticles with common human blood proteins. ACS Nano 4 (2010) 365-379; DOI: 10.1021/nn9011187.
11 of 16
20. S. Chakraborty, P. Joshi, V. Shanker, Z. Ansari, S.P. Singh, P. Chakrabarti. Contrasting effect of gold nanoparticles and nanorods with different surface modifications on the structure and activity of bovine serum albumin. Langmuir, 27 (2011), 7722-7731, DOI: 10.1021/la200787t. 21. M. Rahman, S. Laurent, N. Tawil, L. Yahia, M. Mahmoudi. Protein-nanoparticle interactions. In Springer Series in Biophysics, B. Martinac, Ed.; Springer: Heidelberg, Germany, 15 (2013), 21-25; ISBN 978-3-642-37554-5.
IP T
22. M.P. Monopoli, D. Walczyk, A. Campbell, G. Elia, I. Lynch, F.B. Bombelli, K.A. Dawson. Physicalchemical aspects of protein corona: relevance to in vitro and in vivo biological impacts of nanoparticles. J. Am. Chem. Soc., 113 (2011), 2525–2534, DOI: 10.1021/ja107583h. 23. D. Docter, D. Westmeier, M. Markiewicz, S. Stolte, S.K. Knauer, R.H. Stauber. The nanoparticle biomolecule corona: Lessons learned–challenge accepted? Chem. Soc. Rev., 44 (2015), 6094-6121, DOI: 10.1039/C5CS00217F.
SC R
24. R.M. Pearson, V.V. Juettner, S. Hong. Biomolecular corona on nanoparticles: a survey of recent literature and its implications in targeted drug delivery. Front. Chem., 2 (2014), 108, DOI: 10.3389/fchem.2014.00108.
N
U
25. W. Liu, J. Rose, S. Plantevin, M. Auffan, J.Y. Bottero, C. Vidaud. Protein corona formation for nanomaterials and proteins of a similar size: hard or soft corona? Nanoscale, 5 (2013), 1658-1668, DOI: 10.1039/C2NR33611A.
M
A
26. S. Winzen, S. Schoettler, G. Baier, C. Rosenauer, V. Mailaender, K. Landfester, K. Mohr. Complementary analysis of the hard and soft protein corona: sample preparation critically effects corona composition. Nanoscale, 7 (2015), 2992-3001, DOI: 10.1039/C4NR05982D.
ED
27. T. Miclaus, V.E. Bochenkov, R. Ogaki, K.A. Howard, D.S. Sutherland. Spatial mapping and quantification of soft and hard protein coronas at silver nanocubes. Nano Lett., 14 (2014), 20862093, DOI: 10.1021/nl500277c.
PT
28. C. De Roe, P.J. Courtoy, P. Baudhuin. A model of protein-colloidal gold interactions. J. Histochem. Cytochem., 35 (1987), 1191-1198, DOI: 10.1177/35.11.3655323.
CC E
29. L. Ghitescu, M. Bendayan. Immunolabeling efficiency of protein A-gold complexes. J. Histochem. Cytochem., 38 (1990), 1523-1530, DOI: 10.1177/38.11.2212613. 30. T. Sen, K.K. Haldar, A. Patra. Au nanoparticle-based surface energy transfer probe for conformational changes of BSA protein. J. Phys. Chem. C, 112 (2008), 17945-17951, DOI: 10.1021/jp806866r.
A
31. S. Naveenraj, S. Anandan, A. Kathiravan, R. Renganathan, M. Ashokkumar. The interaction of sonochemically synthesized gold nanoparticles with serum albumins. J. Pharm. Biomed. Anal., 53 (2010), 804-810, DOI: 10.1016/j.jpba.2010.03.039. 32. S.H. Brewer; W.R. Glomm; M.C. Johnson; M.K. Knag; S. Franzen. Probing BSA binding to citratecoated gold nanoparticles and surfaces. Langmuir 21 (2005) 9303-9307, DOI: 10.1021/la050588t. 33. S. Naveenraj, S. Anandan. Binding of serum albumins with bioactive substances–nanoparticles to drugs. J. Photochem. Photobiol. C., 14 (2013), 53-71, DOI: 10.1016/j.jphotochemrev.2012.09.001.
12 of 16
34. M. Cui, R. Liu, Z. Deng, G. Ge, Y. Liu, L. Xie. Quantitative study of protein coronas on gold nanoparticles with different surface modifications. Nano Res., 7 (2014), 345-352, DOI: 10.1007/s12274-013-0400-0. 35. A. Wang, Y.R. Perera, M.B. Davidson, N.C. Fitzkee. Electrostatic interactions and protein competition reveal a dynamic surface in gold nanoparticle–protein adsorption. J. Phys. Chem. C, 120 (2016), 24231-24239. 36. L. Dykman, N. Khlebtsov. Gold nanoparticles in biomedical applications: Recent advances and perspectives. Chem. Soc. Rev., 41 (2012), 2256-2282, DOI: 10.1039/C1CS15166E.
IP T
37. H. Sellers, A. Ulman, Y. Shnidman, J.E. Eilers. Structure and binding of alkanethiolates on gold and silver surfaces: implications for self-assembled monolayers. J. Am. Chem. Soc., 115 (1993), 93899401, DOI: 10.1021/ja00074a004.
SC R
38. D.V. Sotnikov, A.V. Zherdev, B.B. Dzantiev. Development and application of a label-free fluorescence method for determining the composition of gold nanoparticle–protein conjugates. Intern. J. Mol. Sci., 16 (2014), 907-923, DOI: 10.3390/ijms16010907.
U
39. G. Frens. Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions. Nature, 241 (1973), 20-22, DOI: 10.1038/physci241020a0.
A
N
40. I. Safenkova, A. Zherdev, B. Dzantiev. Factors influencing the detection limit of the lateral-flow sandwich immunoassay: A case study with potato virus X. Anal. Bioanal. Chem., 403 (2012), 15951605, DOI: 10.1007/s00216-012-5985-8.
M
41. J.R. Lakowicz. Principles of Fluorescence Spectroscopy. 3rd ed.; Springer: New York, USA, (2010), 859 pp., ISBN: 978-0-387-31278-1.
ED
42. K.A. Kang, J. Wang, J.B. Jasinski, S. Achilefu. Fluorescence manipulation by gold nanoparticles: from complete quenching to extensive enhancement. J. Nanobiotechnol., 9, 16 (2011), DOI: 10.1186/1477-3155-9-16.
PT
43. G. Li, R. Stewart, B. Conlan, A. Gilbert, P. Roeth, H. Nair. Purification of human immunoglobulin G: A new approach to plasma fractionation. Vox Sanguinis, 83 (2002), 332-338, DOI: 10.1046/j.14230410.2002.00241.x.
CC E
44. Aubin-Tam, M.E., Hamad-Schifferli, K. (2008). Structure and function of nanoparticle-protein conjugates. Biomedical Materials, 3 (3), 034001. 45. Mays D.C. (2007) Using the Quirk-Schofield diagram to explain environmental colloid dispersion phenomena. J. Natural Resources & Life Sci. Education, 36 (1), 45–52.
A
46. F. Benetti, M. Fedel, L. Minati, G. Speranza, C. Migliaresi. Gold nanoparticles: Role of size and surface chemistry on blood protein adsorption. J. Nanopart. Res., 15 (2013), 1694, DOI: 10.1007/s11051013-1694-2. 47. J. Deng, M. Sun, J. Zhu, C. Gao. Molecular interactions of different size AuNP–COOH nanoparticles with human fibrinogen. Nanoscale, 5 (2013), 8130-8137, DOI: 10.1039/C3NR02327C. 48. K. Kaur, J.A. Forrest. Influence of particle size on the binding activity of proteins adsorbed onto gold nanoparticles. Langmuir, 28 (2012), 2736-2744, DOI: 10.1021/la203528u.
13 of 16
49. M. Horisberger, M. Vauthey. Labelling of colloidal gold with protein. Histochem. Cell Biol., 80 (1984), 13-18, DOI: 10.1007/BF00492765. 50. H. Jans, X. Liu, L. Austin, G. Maes, Q. Huo. Dynamic light scattering as a powerful tool for gold nanoparticle bioconjugation and biomolecular binding studies. Anal. Chem., 81 (2009), 9425-9432, DOI: 10.1021/ac901822w.
A
CC E
PT
ED
M
A
N
U
SC R
IP T
51. Geoghegan, W. D., & Ackerman, G. A. (1977). Adsorption of horseradish peroxidase, ovomucoid and anti-immunoglobulin to colloidal gold for the indirect detection of concanavalin A, wheat germ agglutinin and goat anti-human immunoglobulin G on cell surfaces at the electron microscopic level: a new method, theory and application. Journal of Histochemistry & Cytochemistry, 25(11), 1187-1200.
M
A
N
U
SC R
IP T
14 of 16
A
CC E
PT
ED
Fig. 1. Dependences of fluorescence intensity at 350 nm from BSA concentration added to calibration solutions (1) and a supernatant after centrifugation of conjugates with 20 nm GNPs (2). A. – pH = 5; B. – pH = 7; C. – pH = 8.
Fig. 2. Dependence of the number of BSA molecules adsorbed on one GNP (20 nm) with different diameters and at different pH from a BSA concentration.
SC R
IP T
15 of 16
A
CC E
PT
ED
M
A
N
U
Fig. 3. Dependence of the number of IgG molecules adsorbed on one nanoparticle (22 nm) from the IgG concentration being synthesized.
16 of 16
Table 1. TEM characteristics of GNP preparations. Size of sample, pcs
Length of major axis, nm
Length of smaller axis, nm
Average diameter, nm
Ratio of axes
1
117
22±4
19±3
20±4
0.8
Concentration of particles, pcs/mL 6.2·1011
2
152
24±5
20±4
22±5
0.8
4.6·1011
3
112
26±3
22±3
24±3
0.8
3.6·1011
4
117
29±6
25±4
27±6
0.9
2.5·1011
5
57
54±7
40±6
48±7
0.7
6
46
44±5
35±5
40±5
0.8
IP T
Preparation
4.5·1010
SC R
7.7·1010
PT CC E A
24 107 154 231 -
M
20 96 138 185 -
ED
Diameter of GNPs, nm pH 4–5 pH 7 pH 8 pH 10
N
BSA
A
Protein
U
Table 2. The maximum number of protein molecules adsorbed on one particle.
48 280 720 -
IgG 22 102 103 139 132
27 127 178 208 128
40 145 495 446 388
S0927-7765(18)30722-7 https://doi.org/10.1016/j.colsurfb.2018.10.025 COLSUB 9709
To appear in:
Colloids and Surfaces B: Biointerfaces
Received date: Revised date: Accepted date:
19-6-2018 4-10-2018 9-10-2018
Please cite this article as: Sotnikov DV, Berlina AN, Ivanov VS, Zherdev AV, Dzantiev BB, Adsorption of proteins on gold nanoparticles: One or more layers?, Colloids and Surfaces B: Biointerfaces (2018), https://doi.org/10.1016/j.colsurfb.2018.10.025 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 of 16
Adsorption of proteins on gold nanoparticles: One or more layers?
Dmitriy V. Sotnikov, Anna N. Berlina, Vladislav S. Ivanov, Anatoly V. Zherdev, Boris B. Dzantiev *
[email protected], [email protected]. Tel.: +7-495-9543142
SC R
*Correspondence:
IP T
A.N. Bach Institute of Biochemistry, Research Center of Biotechnology of the Russian Academy of Sciences, Leninsky prospect 33, Moscow 119071, Russia;
The composition of conjugates of gold nanoparticles with proteins was analysed pH influences significantly on the adsorption of proteins on gold nanoparticles Both monolayer and multilayer structures from protein molecules are formed Explanation for the transition from monolayer to multilayer structures was proposed
A
CC E
PT
Highlights
ED
M
A
N
U
Graphical Abstract
2 of 16
SC R
IP T
Abstract: Adsorption of proteins on nanoparticles is a complex and poorly studied process. The mechanisms of protein layer formation can fundamentally differ depending on the composition of the medium, the nanoparticles’ structure, the protein’s nature, and other factors. In particular, monolayer or multilayer immobilization may occur. In the present work, the composition of conjugates of bovine serum albumin and immunoglobulin G with gold nanoparticles obtained by the Turkevich-Frens method are analyzed. The composition was studied by protein fluorescence measurement for particles ranging in size from 20–48 nm, depending on the pH of the immobilization medium (from 4–5 to 8–10). It was found that a pH shift of the immobilization medium from acidic to alkaline values is accompanied by a change in the mechanism of protein adsorption on the gold surface. In acidic pH conditions (4–5), effective binding of bovine serum albumin and gold nanoparticles occurs throughout the entire range of studied protein concentrations. In alkaline pH conditions (8-10), however, effective binding occurs only at concentrations of >10 μg/mL. This effect is not observed for immunoglobulin G, which is efficiently adsorbed onto nanoparticles throughout the entire range of studied concentrations and pH values. For acidic pH values, the surface of the particles is saturated with the amount of bound proteins, which approximately corresponds to the amount the monolayer is filled. For neutral and alkaline pH values, saturation is not observed and the amount of adsorbed protein certainly exceeds the monolayer filling, leading to multilayer immobilization.
N
U
Keywords: gold nanoparticles; protein conjugates; tryptophan fluorescence; bovine serum albumin; immunoglobulin G
M
A
1. Introduction
A
CC E
PT
ED
Gold nanoparticle (GNP) and their complexes with proteins are used for many analytical and therapeutic tasks [1-3]. The composition, structure and functional properties of protein nanoconjugates were characterized in a number of studies using various physical, chemical and biochemical methods [415]. Despite the advances in this field, the interaction of proteins and GNPs still raises many unknown questions. Many works contain data that, at first glance, may seem contradictory. In [11, 12, 16-18], it is claimed that proteins form a monolayer on the surface of GNP, and in [19, 20], it was observed that protein adsorption on GNPs continues at protein concentrations beyond those that fill the monolayer. The formation of multilayer structures from adsorbed proteins on the surface of nanoparticles is described in the literature by the model of so-called “hard” and “soft” protein coronas [21-27]. De Roe et al. [28] showed that protein A interacts with GNPs obtained by the citrate method with an equilibrium dissociation constant of 343.9 nM. However, Ghitescu et al. [29] stated that the dissociation constant at low protein A concentrations is 2–3 nM and at high concentrations increases to 500–900 nM, although both groups used the same radioactive labeling method. Differences in the interaction constants for proteins and GNPs can reach five orders of magnitude [16, 28, 30-32]. For example, for bovine serum albumin (BSA) conjugate and spherical GNP, the dissociation constant values, determined by quenching of tryptophan fluorescence, range from 2.7·10-5 M [33] to 1·10-9 M [30]. These differences can be explained by some methods’ insufficient accuracy, possible changes in the structure of conjugates during the experiments, and use of un-unified nanoparticles obtained by different techniques or under different conditions [10, 20, 34]. The interaction of macromolecules with GNP depends on the chemical nature of the adsorbed material, the composition of the medium, and other factors. Electrostatic, hydrophobic, van der Waals, and donor–acceptor interactions can contribute to this process [35, 36]. Further, gold is known to have an affinity for sulfur-containing chemical groups [37]. Due to the presence of many chemical groups with different natures in proteins, the interaction of proteins with the surface of nanoparticles occurs through several binding centers and mechanisms. However, in most studies, GNP–protein complexes are compared by varying a single
3 of 16
parameter, such as the size of nanoparticles. The search for more general patterns of protein coat formation on GNP surfaces requires multifactor analysis of conjugates that differ in terms of particle size, the nature of the protein to be adsorbed and its concentration, and the composition of the reaction medium. This study examined the GNPs that are most commonly used in practice: those synthesized by the citrate method with a diameter from 20 to 50 nm. Two proteins in the GNP conjugates—BSA and immunoglobulin G (IgG)—were examined as they are used for the functionalization and stabilization of nanoparticles in most studies. Conjugate composition was determined by protein concentration during synthesis and the pH of the immobilization medium. The study was carried out using a previously developed technique based on fluorescence measurements [38].
IP T
The obtained data enable identification of immobilization patterns of proteins on GNPs and explain some differences in the results of previous studies.
2. Materials and Methods
SC R
2.1. Reagents
Tetrachloroauric acid and sodium citrate were obtained from Sigma (USA); BSA was obtained from MP Biomedicals (USA); and goat anti-mouse IgG was obtained from Arista Biologicals (USA). All salts were analytically or chemically pure. Water was purified using a MilliQ system (Millipore, USA).
U
2.2. Preparation of gold nanoparticles
A
N
First, 0.2 mL of a 5% solution of tetrachloroauric acid was added to 97.5 ml of distilled water. The mixture was then brought to a boil. After that, 1, 1.5, or 2 ml of a 1% solution of sodium citrate was added. The mixture was boiled for 30 min, and then cooled to room temperature [39]. The resulting preparation was stored at 4 °C.
M
2.3. Determination of GNP size by transmission electron microscopy
PT
ED
GNP preparations were applied to 200-mesh hexagonal copper meshes coated with a pharmacopolymer film. A CX-100 transmission electron microscope (Jeol, Japan) with an accelerating voltage of 80 kV was used. Micrographs were taken at a magnification of 33,000X. Film negatives were scanned in grayscale on a scanner to obtain 1,200 dpi images. These images were saved as TIFFs and transferred to an image-processing program (Image Tool, University of Texas Health Science Center at San Antonio, USA). Particle size and the size distribution were then calculated according to the method used by Safenkova et al. [40]. 2.4. Preparation of protein and GNP conjugates at different pH values
A
CC E
The GNP solution was adjusted to the desired pH with 0.2 M Na2CО3, which was poured into 8 tubes of 1 mL and centrifuged at 13.4 rpm for 10 min. Then, the supernatant was separated from the sediment. The gold nanoparticle sediment was shaken and adjusted to 125 μL. Proteins were diluted to concentrations from 2 to 250 μg/mL using the GNP preparation supernatant. Only freshly prepared solutions were used in the experiments. Then, 125 μL of the resulting protein solutions were added to 125 μL of the centrifuged GNP solutions. The remaining protein solutions were used as calibration solutions. The protein was incubated with GNP for 1 h at room temperature and centrifuged at 10,000 g (20 min), and then 200 μL of supernatant was taken and transferred to microplate wells. The calibration solutions (200 μL aliquots) were also transferred to the microplate wells. Since microplates adsorb proteins, affecting fluorescence, the measurements were taken immediately after transferring the samples to the wells. 2.5. Fluorescence measurements The fluorescence spectra of proteins were measured by a Perkin Elmer En Spire 2300 microplate spectrophotometer (Waltham, MA, USA) in Nunc MaxiSorp white microplates (Roskilde, Denmark) at an excitation light wavelength of 280 nm and an emitted light wavelength of 350 nm.
4 of 16
2.6. Characterization of the proteins solutions by asymmetrical flow field-flow fractionation (AF4) The fractionations were performed by a Wyatt Eclipse 3+ Separation System (Wyatt Technology, USA) with an autosampler and pump (Agilent Technologies, USA). A 5-kDa MW cut off regenerated cellulose membrane (Microdyn-Nadir, Germany) was used for separation in the 275-mm channel with a 350-μm thick spacer. The channel was sequentially connected with a UV/VIS detector (Agilent Technologies, USA), Dawn HELEOS II multi-angle light scattering detector and Optilab T-Rex refractometer (both from Wyatt Technology, USA).
IP T
Solutions of BSA (1 mg/mL) and IgG (1 mg/mL) in 10 mM Na-citrate buffer, pH = 4.0, 50 mM Kphosphate buffer, pH = 7.3 (PBS), and 10 mM Na-carbonate buffer, pH = 9.0 were prepared. 80 μL of the solutions were loaded in the AF4 system at a rate of 0.2 mL/min. The focusing time during the loading was 2 min. The carrier fluids were the same as the buffers for the preparation of of BSA and IgG solutions. The main flow rate in the channel was 1 mL/min, and the focus flow rate was 1 mL/min. Separation was performed at a linear rate gradient of the cross flow from 5 to 0.1 mL/min (10 min).
SC R
The registered data were analyzed using ChemStation v.B.04.03 (Agilent Technologies, USA) and Astra v.6.1.1.17 (Wyatt Technology, USA) software.
3. Results and Discussion
N
3.1. Synthesis and determination of the size of GNPs
U
The described above AF4 measurements were conducted using the Shared-Access Equipment Centre “Industrial Biotechnology” of the Federal Research Centre “Fundamentals of Biotechnology” Russian Academy of Sciences (A.N. Bach Institute of Biochemistry RAS), Moscow, Russia.
M
A
To synthesize GNPs of different diameters, three concentrations of a reducing agent were used. Each synthesis was reproduced twice, resulting in six GNP preparations. Their dimensional characteristics were evaluated by transmission electron microscopy. According to the obtained data, the synthesized GNP preparations (Nos. 1–6) had average particle diameters of about 20, 22, 24, 27, 48, and 40 nm, respectively. Table 1 summarizes the micrograph analysis of the GNP preparations. The average deviation in particle diameter for each preparation did not exceed 25%.
ED
3.2. Determination of the GNP concentration
CC E
PT
GNP concentration was determined based on the following method. Assuming that the particles are spherical, the volume of one particle (Vp) is found using the data on particle sizes obtained from electron micrographs. Taking into account the density of gold (ρ = 19.3 g/cm3), we obtain the mass of one particle (mp = ρ ∙ Vp). During synthesis, 0.2 mL of 5% HAuCl4 was added to 100 mL of water. Accordingly, the concentration of gold is 0.05 mg/mL, and the number of particles in one milliliter is n = C / m p. The resulting particle concentrations are given in Table 1. Before conjugation, GNPs were precipitated and the precipitate was re-dissolved in a smaller volume, increasing the GNP concentration four-fold. 3.3. Preparation and characterization of GNP–protein conjugates
A
BSA and IgG were selected for analysis in this study because IgG is the most common receptor molecule in bioanalytical systems and BSA is most often used as a carrier protein for haptens and as a stabilizer for colloidal particles. In addition, BSA and IgG are the main protein components of blood, and when nanoparticles enter the body, they interact mainly with these proteins. Thus, studying these interactions is of great importance for the fields of medicine and toxicology. GNP preparations 1, 3, and 5 were conjugated with BSA, whereas preparations 2, 4, and 6 were conjugated with IgG. The method used to determine the composition of conjugates is based on measurement of the intrinsic fluorescence of proteins. It does not require any labels, which simplifies the analysis, reduces
5 of 16
errors, and enables study of the conjugation of GNPs with native proteins. Most proteins have intrinsic fluorescence, which is mainly due to tryptophanic residues. Tryptophan features maximum light absorption at 280 nm, and maximum emission at 340–360 nm [41]. The principle of this technique is described in detail in [38]. Briefly, it involves comparing the initial fluorescence of the calibration solutions (F0) and the residual fluorescence of the reaction solutions after separation of the synthesized conjugates (F). The difference between these two quantities (ΔF) is proportional to the amount of protein in the conjugate:
Ck = C0·ΔF/F0,
(1)
IP T
where C0 is the added protein concentration and Ck is the concentration of proteins bound to GNP in the solution.
SC R
This technique excludes the influence of GNP on tryptophan fluorescence. Since GNP is a strong modulator of fluorophore properties [42], its presence in the solution changes the value of fluorescence and does not allow for comparison of fluorescence between samples and calibration solutions. For this reason, the accuracy of measurements largely depends on the completeness of removal of gold particles from the solution. 3.4. Determination of the amount of protein adsorbed on GNP
N
U
GNP solutions synthesized by the citrate method have an acidic or slightly acidic pH (3–6) depending on particle size. For larger-sized GNP, pH is decreased due to the lower sodium citrate content, which shifts the pH towards neutral values (the measured pH value of a 1% sodium citrate solution is 7.3). The pH of GNP solutions was adjusted by adding an alkaline 0.2 M Na2CO3 solution.
ED
M
A
After synthesis and separation of conjugates by centrifugation, 200 μL aliquots of supernatant were placed in microtiter plate wells. Calibration solutions (also 200 μL) were placed in wells of the same plate, and fluorescence spectra were measured. Based on the measured spectra, the dependences of fluorescence intensity on the concentration of the added protein at 350 nm were plotted for the calibration solutions and samples. At protein concentrations of 0–250 μg/mL, the calibration dependences are practically linear (R linearization factor is >0.99 for both proteins).
PT
Based on the difference in fluorescence of the samples and calibration solutions and the data concerning particles’ content during conjugation (see Table 1), the number of protein molecules on one particle was calculated according to the following formula:
[(F0 – F)/F0] × ([L0]/[R0]) = RL,
(2)
CC E
where RL is the number of protein molecules bound to GNPs, F0 is the fluorescence of the calibration solution, F is the fluorescence of the supernatant after centrifugation of the synthesized conjugate, [L0] is the molar concentration of the added protein, and [R0] is the molar concentration of GNPs.
A
In some cases, during the synthesis of conjugates, aggregation was observed, which was accompanied by a change in the solution color from red to blue: • Preparations of GNPs sized 48 nm aggregate when alkalinized to pH = 8 (data not shown); • A slight discoloration after alkalization of GNPs sized 22, 27, and 40 nm at pH = 10 (see Section
3.6). • Significant aggregation of conjugates with a low content of adsorbed protein after centrifugation. However, because no residual protein was observed in the supernatant, the protein was adsorbed completely on the particles, and aggregation under these conditions did not affect our conclusions. For all other cases, the conjugates were stable, and discoloration of the particles after protein addition was not observed. Therefore, we state that the processes of GNP aggregation were nonsignificant for the article’s conclusions. This is in agreement with the results of experiments on the
6 of 16
adsorption of proteins at different GNP concentrations; as well, in the case of aggregation, it was enforced by the increase in the GNP concentration and it caused non-linearity in the dependence of the amount of adsorbed protein on the concentration of nanoparticles (See section ‘Studying on the dependence of adsorbed protein quantity from GNP concentration’ in the Supplementary material). 3.5. Examination of the composition of GNP–BSA conjugates
SC R
IP T
Data concerning BSA fluorescence in calibration solutions and in the supernatant after centrifugation of the conjugates with average GNP diameters of 20 are shown in Fig. 1 (the same data for 24, and 48 nm GNPs are shown in Supplementary material: Fig. S1–S2, respectively). Measurements were carried out in triplicate for each type of conjugate. As shown in the figures, the dependence on BSA concentration differs significantly for different pH values. At an acidic pH (4-5) with a low concentration of added BSA, the amount of unbound protein molecules does not differ from the background; that is, almost all the added protein was adsorbed on the GNP surface. The parallel course of the fluorescence curves of the calibration solutions and samples at high protein concentrations indicates saturation of all sorption centers on the surface of nanoparticles. In neutral (pH 7) and alkaline (pH 8) media, BSA fluorescence in the supernatant liquid (protein concentrations of less than 20 μg/mL) is much higher than in acidic medium and the gold surface is not saturated. The obtained dependences are similar for all three investigated GNP preparations (average diameters of 20, 24, and 48 nm). The 48 nm preparation was not stable after alkalization and centrifugation of the solution, which made it impossible to determine the composition of its conjugate with BSA at a pH of 8.
CC E
PT
ED
M
A
N
U
Based on the fluorescence differences between the calibration solutions and supernatant, the amount of BSA adsorbed on one particle was calculated using formula (2). The obtained data concerning conjugates’ composition (Fig. 2 and Fig. S3 in Supplementary material) demonstrate that in acidic pH conditions effective binding of BSA to particles occurs at all studied protein concentrations, but in alkaline pH conditions effective binding occurs only at concentrations greater than 10 μg/mL. At BSA concentrations of ~1–80 μg/mL (for GNPs with an average diameter of 20 nm (Fig. 2)), ~1–60 μg/mL (for GNPs with an average diameter of 24 nm (Fig. S3A)), and ~1–25 μg/mL (for GNPs with an average diameter of 48 nm (Fig. S3B)), more BSA is absorbed by GNPs in acidic solutions (pH 4–5) than in alkaline ones and at higher protein concentrations – more BSA is absorbed in alkaline solutions. These effects can be explained by the fact that, in an alkaline environment, the negative charge of BSA molecules increases (pI in water at 25 °C is 4.7) and, accordingly, their repulsion from GNP particles with negatively charged surfaces is enhanced [2]. This creates an electrostatic barrier that can be overcome only by protein molecules with sufficient kinetic energy, the number of which increases with increasing concentration. For larger particles (diameter of 48 nm), in neutral pH conditions, this barrier practically stops BSA adsorption on GNPs at concentrations below 8 μg/mL (Fig. S3B). However, for particles with diameters of 20 and 24 nm, this effect takes place only at a pH of 8. In alkaline solutions, the GNP surface is not saturated, even at BSA concentrations of 250 μg/mL, whereas in acidic solutions, saturation occurs at concentrations of about 60 μg/mL (for GNPs with diameters of 20 and 24 nm) and 16 μg/mL (for GNPs with a diameter of 48 nm). Thus, at BSA concentrations <60 μg/mL more BSA is adsorbed in an acidic medium than in an alkaline medium and at higher BSA concentrations – more BSA is adsorbed in an alkaline medium. Additional studies to understand these effects and their mechanism further will be necessary.
A
The dependence of the maximum amount of adsorbed BSA from the pH of the immobilization medium (Fig. S4) demonstrates that the least amount of protein is adsorbed in an acidic medium. 3.6. Examination of the composition of GNP–IgG conjugates IgG adsorption on GNP was studied in a similar manner to the described above study of BSA adsorption. The fluorescence of IgG in calibration solutions and the supernatant after centrifugation of the conjugates is shown in Fig. S4–7 (Supplementary material). The obtained dependences are somewhat different from those for BSA, but a lack of saturation in alkaline pH conditions is also observed.
7 of 16
The obtained dependences of the number of IgG molecules adsorbed on one particle from the initial protein concentration in solutions with different pH values are shown in Fig. 3 and Fig S8 (Supplementary material). Unlike for BSA, there is no significant decrease in binding at the initial sections of the curves in alkaline solutions. This may be due to the much higher values of the isoelectric points of IgG molecules compared to BSA (the pI of IgG ranges from 6.4–9.0 depending on amino acid composition and degree of glycosylation [43]). For this reason, it is difficult for IgG to reach a sufficient negative charge to prevent contact with the GNP surface. At an IgG concentration of ~60 μg/mL and a pH of 4–5, the amount of adsorbed protein reaches a plateau, whereas in neutral and alkaline solutions, adsorption continues until a concentration of 250 μg/mL.
SC R
IP T
The maximum amount of adsorbed IgG on GNPs with average diameters of 22, 27, and 40 nm, depending on the pH of the immobilization solution, are shown in Fig. S9 (Supplementary material). The highest number of protein molecules on one GNP is reached at a pH of 8. In addition, the observed decrease in adsorption at higher pH values may be due to an increase in the concentration of sodium carbonate, which leads to a decrease in the total surface area of nanoparticles and their aggregation. 3.7. Calculating the surface area occupied by proteins on GNP surface
U
The maximum quantities of protein molecules adsorbed on a single particle are shown in Table 2. By dividing the surface area of a particle by the amount of adsorbed proteins, we can determine the area occupied by one protein molecule.
M
A
N
Calculation of the area occupied by one protein molecule on the surface of the particle demonstrates that, at the maximum load, the occupied area corresponds to a circle with a diameter of about 3 nm for BSA and 3.5 nm for IgG. This is considerably smaller than the dimensions of BSA and IgG molecules (Fig. S10) and the linear size of one Fc or Fab fragment of IgG, leading to the conclusion that multilayer immobilization occurs for IgG on the GNP surface in alkaline pH conditions. That is, depending on the pH of the immobilization and protein concentrations in a solution, both monolayer and multilayer adsorption can be observed. This observation explains differences in the mechanisms of adsorption and structures of conjugates (see Introduction).
CC E
PT
ED
For a fully filled monolayer, the calculated surface area of a nanoparticle per one BSA molecule is 13, 17, and 35 nm2 for GNPs with diameters of 20, 24, and 48 nm, respectively. The surface areas occupied by one IgG molecule are 15, 18, and 35 nm2 for GNPs with diameters of 22, 27, and 40 nm, respectively. The increase in the surface area occupied by one protein molecule can be due primarily to two factors. First, on the more flattened surface, the impact of steric hindrances created by neighboring molecules increases. Second, a large deformation of macromolecules occurs, and, consequently, a larger area of the occupied surface is usually observed on larger particles [44]. This statement is confirmed as well by an increase in the binding constant of the molecule–nanoparticle interaction due to more contact of the chemical groups of the protein with the surface [29].
A
3.8. Study of protein aggregation at different pH values The shift from monolayer to multilayer protein immobilization can be either a consequence of a change in the sorption properties of nanoparticles or spontaneous aggregation of proteins with an increase in the pH of the medium. To investigate this issue, we analyzed the preparations of BSA and IgG in solutions with different pH values by the AF4 method. BSA and IgG solutions with a protein concentration of 1 mg/ml were investigated in three buffer solutions with pH 4.0, 7.3 and 9.0. Fig. S11 presents the outputs of protein fractions from the capillary for AF4. Monomeric BSA and IgG pass through the capillary for 8-10 and 13-14 min, respectively. The presence of aggregates in the preparations leads to the appearance of additional peaks with a release time of 10-12 min for BSA and 20-22 min for IgG, respectively.
8 of 16
IP T
For BSA, an increase in the pH leads to a decrease in the time of release of the protein fraction and a decrease in the aggregates peaks intensities (see Fig. S11A). Thus, increased pH reduces aggregation of BSA. For IgG, the release time slightly increases with decreasing pH, whereas the value of the aggregates remains practically unchanged (Fig. S11B). Thus, aggregation of IgG preparations varies insignificantly. This facts mean that multilayer sorption of the proteins on gold at high pH values is due to changes in the properties of the nanoparticles, and not to the state of the protein molecules in the solution. Note that the AF4 studies were implemented for the concentration of proteins equal to 1 mg/ml for better accuracy of measurements. The concentration of proteins used in the preparation of nanoconjugates was an order of magnitude lower than the concentration for AF4 studies. So the degree of aggregation of the preparations used for nanoconjugates was much lower than that of the preparations used in the AF4 experiments. 3.9. Modeling of GNP-protein conjugate formation
SC R
By summarizing the obtained data and common notions about the structure of GNPs in solution [45], we can suggest a model for the formation of conjugates under different conditions.
U
The surface of a gold particle is charged negatively (the surface potential is about –50 mV) (Fig. S13). Cations create the next layer around the anionic layer and form a double ionic shell around the GNP, which generates repulsive forces between different GNPs. This shell determines the stability of the colloidal solution.
ED
M
A
N
At a pH below the isoelectric point of the adsorbed protein, the positively charged protein molecule is attracted to the surface of the particle and is easily conjugated to it until a monolayer is formed. At a pH above the isoelectric point, on the contrary, the negatively charged protein molecule repels from the surface of the particle. In this case, protein binding to the surface is possible due to donor–acceptor, van der Waals, and other interactions, if the kinetic energy of the protein molecule is sufficient to overcome the electrostatic barrier. Therefore, electrostatic repulsion only leads to a deceleration of adsorption and a decrease in the binding strength, but it does not prevent adsorption. According to experimental data, this effect is more pronounced for BSA with an isoelectric point in the acidic region (pI 4.7–6.0) than for IgG (pI 6.4–9.0). According to Fig. 3, the effect is also observed for IgG, but the influence of the alkaline medium is weakened, as IgG acquires a smaller negative charge than BSA.
CC E
PT
With an increasing pH, the negative charge on the surface of the particles increases. Respectively, the layer of counter ions near the particle increases as well. In addition, the adsorption of the first layer of protein leads to an even greater increase in the negative charge near the surface and an additional expansion of the counter ion layer. As a result, negatively charged protein molecules can electrostatically attach to an increased cationic “coat” (Fig. S14). The interaction with the cationic “coat” is the likely reason for the appearance of the protein poly-layers, as according to AF4, an increase in pH does not enhance protein–protein interactions.
4. Conclusions
A
It is important to develop a general theory about protein and nanoparticle conjugation. To do so, researchers need data on conjugates’ composition when several parameters are varied. However, most studies on conjugates have varied one or two parameters [16, 29, 31, 46-50]. This study performs multifactorial analysis of GNP conjugates’ composition, varying particle size, the nature of the protein, protein concentration, and the pH of the immobilization medium. The obtained results contribute new data about the formation of protein layers on the surface of GNPs, which are undoubtedly of interest to researchers. In particular, this study showed the effect of the pH of the immobilization medium on the adsorption of proteins on GNPs: when pH varies from 4–5 to 8–10, the maximum amount of adsorbed protein molecules on a GNP increases, possibly due to the transition from monolayer to multilayer protein immobilization. The AF4 study of proteins solutions has shown that pH-dependent changes of the amount of adsorbed proteins are not caused by aggregation of the proteins.
9 of 16
Differences in the adsorption of BSA and IgG on GNPs were found. The shift in the pH of the immobilization medium from acidic to alkaline values is accompanied by a change in the adsorption mechanism. In acidic pH conditions, effective binding of BSA to nanoparticles occurs at all protein concentrations, but in alkaline pH conditions, effective binding occurs only at concentrations of >10 μg/mL. IgG is adsorbed effectively at all studied concentrations and pH values.
IP T
Significant differences in the adsorption of proteins with different isoelectric points were considered by Geoghegan and Ackerman [51]. They studied the stabilization of gold particles by different proteins and found that the pH dependence of the stabilization of conjugates in solutions with high ionic strength is U-shaped for proteins with an isoelectric point in the acidic region and L-shaped for proteins with an isoelectric point in the alkaline region. However, this study was implemented at a fixed protein concentration and needs additional data. Our results (see Fig. S3) indicate that the pH dependence of BSA adsorption on GNP 24 nm is U-shaped for concentrations in the range of ~50–150 μg/mL. However, the form of the dependence changes for other concentration ranges. This example confirms the need for a multifactor analysis of protein adsorption on nanoparticles to describe this process fully.
SC R
These observations explain the differences in the results obtained by different groups of researchers regarding GNP–protein conjugates.
U
Acknowledgments:
N
The authors are grateful to Irina V. Safenkova (A.N. Bach Institute of Biochemistry) for her assistance in AF4 studies and interpretation of the obtained data.
ED
M
A
This study was financially supported by the Ministry of Science and Education of the Russian Federation (grant agreement No. 14.613.21.0080 on 22.11.2017, unique identifier RFMEFI61317X0080).
PT
References
1. S. Zeng, K.T. Yong, I. Roy, X.Q. Dinh, X. Yu, F. Luan. A review on functionalized gold nanoparticles for biosensing applications. Plasmonics, 6 (2011), 491-506, DOI: 10.1007/s11468-011-9228-1.
CC E
2. K.E. Sapsford, W.R. Algar, L. Berti, K.B. Gemmill, B.J. Casey, E. Oh, M.H. Stewart, I.L. Medintz. Functionalizing nanoparticles with biological molecules: developing chemistries that facilitate nanotechnology. Chem. Rev., 113 (2013), 1904-2074, DOI: 10.1021/cr300143v.
A
3. O.V. Salata. Applications of nanoparticles in biology and medicine. J. Nanobiotechnol. 2, 3 (2004), DOI: 10.1186/1477-3155-2-3. 4. K.E. Sapsford, K.M. Tyner, B.J. Dair, J.R. Deschamps, I.L. Medintz. Analyzing nanomaterial bioconjugates: A review of current and emerging purification and characterization techniques. Anal. Chem., 83 (2011), 4453-4488, DOI: 10.1021/ac200853a. 5. N.C. Bell, C. Minelli, A.G. Shard. Quantitation of IgG protein adsorption to gold nanoparticles using particle size measurement. Anal. Methods, 5 (2013), 4591-4601, DOI: 10.1039/C3AY40771C.
10 of 16
6. N.G. Khlebtsov, V.A. Bogatyrev, B.N. Khlebtsov, L.A. Dykman, P. Englebienne. A multilayer model for gold nanoparticle bioconjugates: application to study of gelatin and human IgG adsorption using extinction and light scattering spectra and the dynamic light scattering method. Colloid J., 65 (2003), 622-635, DOI: 10.1023/A:1026140310601. 7. S. Neupane, Y. Pan; S. Takalkar, K. Bentz, J. Farmakes, Y. Xu, B. Chen, G. Liu; S.Y. Qian, Z. Yang. Probing the aggregation mechanism of gold nanoparticles triggered by a globular protein. J. Phys. Chem. C, 121 (2017), 1377-1386, DOI: 10.1021/acs.jpcc.6b11963.
IP T
8. M.J. Pollitt, G. Buckton, R. Piper, S. Brocchini. Measuring antibody coatings on gold nanoparticles by optical spectroscopy. RSC Adv., 5 (2015), 24521-24527, DOI: 10.1039/C4RA15661G. 9. S. Liu, J. Horak, M. Höldrich, M. Lämmerhofer. Accurate and reliable quantification of the protein surface coverage on protein-functionalized nanoparticles. Anal. Chim. Acta, 989 (2017), 29-37; DOI: 10.1016/j.aca.2017.08.004.
SC R
10. A.M. Davidson, M. Brust, D.L. Cooper, M. Volk. Sensitive analysis of protein adsorption to colloidal gold by differential centrifugal sedimentation. Anal. Chem., 89 (2017), 6807–6814, DOI: 10.1021/acs.analchem.7b01229.
U
11. S. Zhang, Y. Moustafa, Q. Huo. Different interaction modes of biomolecules with citrate-capped gold nanoparticles. ACS Appl. Mater. Interfaces, 6 (2014), 21184-21192, DOI: 10.1021/am506112u.
A
N
12. L. Shang, G.U. Nienhaus. In situ characterization of protein adsorption onto nanoparticles by fluorescence correlation spectroscopy. Acc. Chem. Res., 50 (2017), 387-395, DOI: 10.1021/acs.accounts.6b00579.
ED
M
13. K.E. Woods, Y.R. Perera, M.B. Davidson, C.A. Wilks, D.K. Yadav, N.C. Fitzkee. Understanding protein structure deformation on the surface of gold nanoparticles of varying size. J. Phys. Chem. C, 120 (2016), 27944-27953, DOI: 10.1021/acs.jpcc.6b08089. 14. A. Wang, K. Vangala, T. Vo, D. Zhang, N.C. Fitzkee. A three-step model for protein–gold nanoparticle adsorption. J. Phys. Chem. C, 118 (2014), 8134-8142, DOI: 10.1021/jp411543y.
PT
15. Q. Shao, C.K. Hall. Protein adsorption on nanoparticles: model development using computer simulation. J. Phys. Condens. Matter, 28 (2016), 414019, DOI: 10.1088/0953-8984/28/41/414019.
CC E
16. S. Dominguez-Medina, S. McDonough, P. Swanglap, C.F. Landes, S. Link. In situ measurement of bovine serum albumin interaction with gold nanospheres. Langmuir 28 (2012) 9131-9139, DOI: 10.1021/la3005213.
A
17. E. Casals, T. Pfaller, A. Duschl, G.J. Oostingh, V. Puntes. Time evolution of the nanoparticle protein corona. ACS Nano, 4 (2010), 3623-3632, DOI: 10.1021/nn901372t. 18. D.-H. Tsai, F.W. DelRio, A.M. Keene, K.M. Tyner, R.I. MacCuspie, T.J. Cho, M.R. Zachariah, V.A. Hackley. Adsorption and conformation of serum albumin protein on gold nanoparticles investigated using dimensional measurements and in situ spectroscopic methods. Langmuir 27 (2011) 24642477, DOI: 10.1021/la104124d. 19. S.H.D.P. Lacerda, J.J. Park, C. Meuse, D. Pristinski, M.L. Becker, A. Karim, J.F. Douglas. Interaction of gold nanoparticles with common human blood proteins. ACS Nano 4 (2010) 365-379; DOI: 10.1021/nn9011187.
11 of 16
20. S. Chakraborty, P. Joshi, V. Shanker, Z. Ansari, S.P. Singh, P. Chakrabarti. Contrasting effect of gold nanoparticles and nanorods with different surface modifications on the structure and activity of bovine serum albumin. Langmuir, 27 (2011), 7722-7731, DOI: 10.1021/la200787t. 21. M. Rahman, S. Laurent, N. Tawil, L. Yahia, M. Mahmoudi. Protein-nanoparticle interactions. In Springer Series in Biophysics, B. Martinac, Ed.; Springer: Heidelberg, Germany, 15 (2013), 21-25; ISBN 978-3-642-37554-5.
IP T
22. M.P. Monopoli, D. Walczyk, A. Campbell, G. Elia, I. Lynch, F.B. Bombelli, K.A. Dawson. Physicalchemical aspects of protein corona: relevance to in vitro and in vivo biological impacts of nanoparticles. J. Am. Chem. Soc., 113 (2011), 2525–2534, DOI: 10.1021/ja107583h. 23. D. Docter, D. Westmeier, M. Markiewicz, S. Stolte, S.K. Knauer, R.H. Stauber. The nanoparticle biomolecule corona: Lessons learned–challenge accepted? Chem. Soc. Rev., 44 (2015), 6094-6121, DOI: 10.1039/C5CS00217F.
SC R
24. R.M. Pearson, V.V. Juettner, S. Hong. Biomolecular corona on nanoparticles: a survey of recent literature and its implications in targeted drug delivery. Front. Chem., 2 (2014), 108, DOI: 10.3389/fchem.2014.00108.
N
U
25. W. Liu, J. Rose, S. Plantevin, M. Auffan, J.Y. Bottero, C. Vidaud. Protein corona formation for nanomaterials and proteins of a similar size: hard or soft corona? Nanoscale, 5 (2013), 1658-1668, DOI: 10.1039/C2NR33611A.
M
A
26. S. Winzen, S. Schoettler, G. Baier, C. Rosenauer, V. Mailaender, K. Landfester, K. Mohr. Complementary analysis of the hard and soft protein corona: sample preparation critically effects corona composition. Nanoscale, 7 (2015), 2992-3001, DOI: 10.1039/C4NR05982D.
ED
27. T. Miclaus, V.E. Bochenkov, R. Ogaki, K.A. Howard, D.S. Sutherland. Spatial mapping and quantification of soft and hard protein coronas at silver nanocubes. Nano Lett., 14 (2014), 20862093, DOI: 10.1021/nl500277c.
PT
28. C. De Roe, P.J. Courtoy, P. Baudhuin. A model of protein-colloidal gold interactions. J. Histochem. Cytochem., 35 (1987), 1191-1198, DOI: 10.1177/35.11.3655323.
CC E
29. L. Ghitescu, M. Bendayan. Immunolabeling efficiency of protein A-gold complexes. J. Histochem. Cytochem., 38 (1990), 1523-1530, DOI: 10.1177/38.11.2212613. 30. T. Sen, K.K. Haldar, A. Patra. Au nanoparticle-based surface energy transfer probe for conformational changes of BSA protein. J. Phys. Chem. C, 112 (2008), 17945-17951, DOI: 10.1021/jp806866r.
A
31. S. Naveenraj, S. Anandan, A. Kathiravan, R. Renganathan, M. Ashokkumar. The interaction of sonochemically synthesized gold nanoparticles with serum albumins. J. Pharm. Biomed. Anal., 53 (2010), 804-810, DOI: 10.1016/j.jpba.2010.03.039. 32. S.H. Brewer; W.R. Glomm; M.C. Johnson; M.K. Knag; S. Franzen. Probing BSA binding to citratecoated gold nanoparticles and surfaces. Langmuir 21 (2005) 9303-9307, DOI: 10.1021/la050588t. 33. S. Naveenraj, S. Anandan. Binding of serum albumins with bioactive substances–nanoparticles to drugs. J. Photochem. Photobiol. C., 14 (2013), 53-71, DOI: 10.1016/j.jphotochemrev.2012.09.001.
12 of 16
34. M. Cui, R. Liu, Z. Deng, G. Ge, Y. Liu, L. Xie. Quantitative study of protein coronas on gold nanoparticles with different surface modifications. Nano Res., 7 (2014), 345-352, DOI: 10.1007/s12274-013-0400-0. 35. A. Wang, Y.R. Perera, M.B. Davidson, N.C. Fitzkee. Electrostatic interactions and protein competition reveal a dynamic surface in gold nanoparticle–protein adsorption. J. Phys. Chem. C, 120 (2016), 24231-24239. 36. L. Dykman, N. Khlebtsov. Gold nanoparticles in biomedical applications: Recent advances and perspectives. Chem. Soc. Rev., 41 (2012), 2256-2282, DOI: 10.1039/C1CS15166E.
IP T
37. H. Sellers, A. Ulman, Y. Shnidman, J.E. Eilers. Structure and binding of alkanethiolates on gold and silver surfaces: implications for self-assembled monolayers. J. Am. Chem. Soc., 115 (1993), 93899401, DOI: 10.1021/ja00074a004.
SC R
38. D.V. Sotnikov, A.V. Zherdev, B.B. Dzantiev. Development and application of a label-free fluorescence method for determining the composition of gold nanoparticle–protein conjugates. Intern. J. Mol. Sci., 16 (2014), 907-923, DOI: 10.3390/ijms16010907.
U
39. G. Frens. Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions. Nature, 241 (1973), 20-22, DOI: 10.1038/physci241020a0.
A
N
40. I. Safenkova, A. Zherdev, B. Dzantiev. Factors influencing the detection limit of the lateral-flow sandwich immunoassay: A case study with potato virus X. Anal. Bioanal. Chem., 403 (2012), 15951605, DOI: 10.1007/s00216-012-5985-8.
M
41. J.R. Lakowicz. Principles of Fluorescence Spectroscopy. 3rd ed.; Springer: New York, USA, (2010), 859 pp., ISBN: 978-0-387-31278-1.
ED
42. K.A. Kang, J. Wang, J.B. Jasinski, S. Achilefu. Fluorescence manipulation by gold nanoparticles: from complete quenching to extensive enhancement. J. Nanobiotechnol., 9, 16 (2011), DOI: 10.1186/1477-3155-9-16.
PT
43. G. Li, R. Stewart, B. Conlan, A. Gilbert, P. Roeth, H. Nair. Purification of human immunoglobulin G: A new approach to plasma fractionation. Vox Sanguinis, 83 (2002), 332-338, DOI: 10.1046/j.14230410.2002.00241.x.
CC E
44. Aubin-Tam, M.E., Hamad-Schifferli, K. (2008). Structure and function of nanoparticle-protein conjugates. Biomedical Materials, 3 (3), 034001. 45. Mays D.C. (2007) Using the Quirk-Schofield diagram to explain environmental colloid dispersion phenomena. J. Natural Resources & Life Sci. Education, 36 (1), 45–52.
A
46. F. Benetti, M. Fedel, L. Minati, G. Speranza, C. Migliaresi. Gold nanoparticles: Role of size and surface chemistry on blood protein adsorption. J. Nanopart. Res., 15 (2013), 1694, DOI: 10.1007/s11051013-1694-2. 47. J. Deng, M. Sun, J. Zhu, C. Gao. Molecular interactions of different size AuNP–COOH nanoparticles with human fibrinogen. Nanoscale, 5 (2013), 8130-8137, DOI: 10.1039/C3NR02327C. 48. K. Kaur, J.A. Forrest. Influence of particle size on the binding activity of proteins adsorbed onto gold nanoparticles. Langmuir, 28 (2012), 2736-2744, DOI: 10.1021/la203528u.
13 of 16
49. M. Horisberger, M. Vauthey. Labelling of colloidal gold with protein. Histochem. Cell Biol., 80 (1984), 13-18, DOI: 10.1007/BF00492765. 50. H. Jans, X. Liu, L. Austin, G. Maes, Q. Huo. Dynamic light scattering as a powerful tool for gold nanoparticle bioconjugation and biomolecular binding studies. Anal. Chem., 81 (2009), 9425-9432, DOI: 10.1021/ac901822w.
A
CC E
PT
ED
M
A
N
U
SC R
IP T
51. Geoghegan, W. D., & Ackerman, G. A. (1977). Adsorption of horseradish peroxidase, ovomucoid and anti-immunoglobulin to colloidal gold for the indirect detection of concanavalin A, wheat germ agglutinin and goat anti-human immunoglobulin G on cell surfaces at the electron microscopic level: a new method, theory and application. Journal of Histochemistry & Cytochemistry, 25(11), 1187-1200.
M
A
N
U
SC R
IP T
14 of 16
A
CC E
PT
ED
Fig. 1. Dependences of fluorescence intensity at 350 nm from BSA concentration added to calibration solutions (1) and a supernatant after centrifugation of conjugates with 20 nm GNPs (2). A. – pH = 5; B. – pH = 7; C. – pH = 8.
Fig. 2. Dependence of the number of BSA molecules adsorbed on one GNP (20 nm) with different diameters and at different pH from a BSA concentration.
SC R
IP T
15 of 16
A
CC E
PT
ED
M
A
N
U
Fig. 3. Dependence of the number of IgG molecules adsorbed on one nanoparticle (22 nm) from the IgG concentration being synthesized.
16 of 16
Table 1. TEM characteristics of GNP preparations. Size of sample, pcs
Length of major axis, nm
Length of smaller axis, nm
Average diameter, nm
Ratio of axes
1
117
22±4
19±3
20±4
0.8
Concentration of particles, pcs/mL 6.2·1011
2
152
24±5
20±4
22±5
0.8
4.6·1011
3
112
26±3
22±3
24±3
0.8
3.6·1011
4
117
29±6
25±4
27±6
0.9
2.5·1011
5
57
54±7
40±6
48±7
0.7
6
46
44±5
35±5
40±5
0.8
IP T
Preparation
4.5·1010
SC R
7.7·1010
PT CC E A
24 107 154 231 -
M
20 96 138 185 -
ED
Diameter of GNPs, nm pH 4–5 pH 7 pH 8 pH 10
N
BSA
A
Protein
U
Table 2. The maximum number of protein molecules adsorbed on one particle.
48 280 720 -
IgG 22 102 103 139 132
27 127 178 208 128
40 145 495 446 388