One-step homogeneous non-stripping chemiluminescence metal immunoassay based on catalytic activity of gold nanoparticles

Analytical Biochemistry 449 (2014) 1–8

Contents lists available at ScienceDirect

Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio

One-step homogeneous non-stripping chemiluminescence metal immunoassay based on catalytic activity of gold nanoparticles Yingying Qi a,b, Fu-Rong Xiu b, Baoxin Li a,⇑ a Key Laboratory of Analytical Chemistry for Life Science of Shaanxi Province, School of Chemistry and Materials Science, Shaanxi Normal University, Xi’an 710062, People’s Republic of China b Department of Environment and Equipment Engineering, Fujian University of Technology, Fuzhou 350108, People’s Republic of China

a r t i c l e

i n f o

Article history: Received 16 September 2013 Received in revised form 26 November 2013 Accepted 4 December 2013 Available online 9 December 2013 Keywords: Homogeneous Non-stripping Chemiluminescence Immunoassay Gold nanoparticles

a b s t r a c t The catalytic activity of gold nanoparticles (AuNPs) on a luminol–H2O2 chemiluminescence (CL) system is found to be greatly enhanced after its crosslinking aggregation induced by immunoreaction. Based on this observation, a one-step homogeneous non-stripping CL metalloimmunoassay was designed. In the presence of corresponding antigen (Ag), the immunoreaction caused the aggregation of antibody (Ab)modified AuNPs, and these crosslinking aggregated AuNPs could catalyze luminol–H2O2 CL reaction to produce a much stronger CL signal than dispersed Ab-modified AuNPs. The assay, including immunoreaction and detection, can be accomplished in homogeneous solution. In the assay, no tedious and strict stripping of metal nanoparticles, difficult synthesis of labels, multiple steps of immunoreactions and washings, and complicated magnetic separation process were required. The detection limit of human immunoglobulin G (IgG, 3r) was estimated to be as low as 3.2  10 11 g ml 1. The sensitivity was increased by two orders of magnitude over that of other AuNP-based CL immunoassay. The current CL metalloimmunoassay offers the advantages of being simple, cheap, rapid, and sensitive. Ó 2013 Elsevier Inc. All rights reserved.

Sensitive, rapid, and selective immunoassays are applied widely in various biomolecule detections, which are critical for clinical applications and biochemical studies [1–3]. After the radioimmunoassay was introduced [4], other immunoassays employing enzyme, chelate complex of metal ions, chemiluminophore, and fluorescent dye labels were successively developed [5–9]. However, these methods have some drawbacks such as being a health hazard, waste disposal problems, bad stability, and poor sensitivity [5,10]. Gold nanoparticles (AuNPs)1 are used widely as labels because of several advantages such as easy preparation, correspondingly large surface-to-volume ratio, special physicochemical properties, and good biocompatibility [11–13]. On the other hand, chemiluminescence (CL) analysis has become an important detection method because of the high sensitivity, wide linear response range, low consumption of inexpensive reagents, and use of a simple instrument [14]. Therefore, the CL immunoassay (CLIA) using AuNPs as

⇑ Corresponding author. Fax: +86 29 85307774. E-mail address: [email protected] (B. Li). Abbreviations used: AuNP, gold nanoparticle; CL, chemiluminescence; CLIA, CL immunoassay; Ab, antibody; Ag, antigen; IgG, immunoglobulin G; UV–Vis, ultraviolet–visible; TEM, transmission electron microscopy; BSA, bovine serum albumin; HRP, horseradish peroxidase; OPD, O-phenylenediamine; PBS–BSA, phosphate buffer containing BSA; ELISA, enzyme-linked immunosorbent assay; PBS-T, phosphate buffer containing Tween 20; SPR, surface plasmon resonance. 1

0003-2697/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ab.2013.12.007

the biological tags attracted great interest [15–22]. In 2005, Lu’s group [18] and Li’s group [15] separately developed two CL metalloimmunoassays exploiting AuNPs as labels in which, after immunoreactions, AuNPs bound to the antibody (Ab) or antigen (Ag) were dissolved and stripped to produce Au(III) or AuCl4 so as to catalyze luminol CL reaction; thus, the indirect measurement of Ab or Ag was realized. However, the dissolution of AuNPs required extremely strict conditions (highly concentrated HNO3–HCl or poisonous HBr–Br2), which resulted in high CL background so as to reduce sensitivity. In 2006, the other CL metalloimmunoassay was developed using silver precipitation on colloidal AuNP tags for the determination of human immunoglobulin G (IgG) [19]. Although this work avoided the dissolution of gold and improved the sensitivity, a stripping process was still inescapable. Among the above CL metalloimmunoassays, the dissolution of gold or silver was required to be conducted under extremely strict conditions for more than 12 h to ensure complete dissolution. Hence, the stripping CL metalloimmunoassays were tedious and time-consuming. To address the issue, the development of non-stripping CLIA attracted tremendous interest [20,23]. Li and coworkers [20] first established a non-stripping CLIA based on the phenomenon that the irregular AuNPs could greatly enhance the CL intensity of a luminol–H2O2 system. Although this assay avoided the strict stripping procedure, it still had an important disadvantage in that the synthesis of irregular nanoparticles was hard to control, which was required to react under 40 °C for

2

Chemiluminescence metal immunoassay using AuNPs / Y. Qi et al. / Anal. Biochem. 449 (2014) 1–8

24 h during purging of oxygen, limiting the practical application of this method. Recently, a non-stripping microplate-compatible CLIA was developed for the determination of human IgG based on a luminol–AgNO3–AuNP CL system [23]. Normal spherical AuNPs can be used in this method, but it still needs extraordinary apparatus such as a microplate luminometer and several cycles of consecutive binding and washing steps. So, it was sophisticated and had a long dwell time. In addition, some researchers [21,24,25], using a magnetic separation/mixing process and the amplification feature of AuNP labels, also developed the non-stripping CL immunoassays. These protocols involved a sandwich format in which an extraordinary CL enhancer and several labeling and washing steps were indispensable, and the preparation of antibody-immobilized magnetic beads and a magnetic separation process made these methods complicated. In addition, the analytic time was very long, with even dozens of hours being needed. Moreover, the use of a CL enhancer and magnetic beads greatly increased the analytic cost. Very recently, a capillary electrophoresis-based CLIA using AuNPs as a label also was developed [22]. Our previous works [26–28] found that the non-crosslinking aggregated AuNPs could display stronger catalytic activity on luminol CL reaction than the dispersed AuNPs. Based on this, label-free homogeneous DNA hybridization detection [26] and a label-free aptamer-based CL biosensor [27] were established. They were based on the fact that different configuration oligonucleotides had different propensities to adsorb on AuNPs in colloidal solution, and DNA hybridization and aptamers’ conformational changes could result in non-crosslinking aggregation of AuNPs, producing a strong CL signal. On the other hand, the crosslinking aggregation of AuNPs induced by the immunoreaction between Ab-modified AuNPs and Ag was reported frequently and has been applied for various immunoassays, including the hyper-Rayleigh scattering (HRS) technique, light scattering analysis technique, colorimetric assay, and single AuNP counter (SGNPC) [29–32]. Therefore, the crosslinking aggregation of AuNPs induced by the immunoreaction offers the possibility of applying this phenomenon to direct CLIA. The applicability of crosslinking aggregated AuNPs used in the direct CLIA depends on whether or not the crosslinking aggregated AuNPs conjugated with protein induced by immunoreaction show different catalytic activity on the CL reaction. In this work, the catalytic ability of crosslinking aggregated AuNPs induced by immunoreaction on the luminol CL system was first studied. It was found that crosslinking aggregated AuNPs surrounded by protein could also enhance the luminol CL signal in comparison with dispersed AuNPs. Based on this finding, the luminol–H2O2–crosslinking aggregated AuNP CL system was used to develop a new one-step and non-stripping CLIA protocol for the determination of human IgG. A schematic diagram of this method is shown in Scheme 1. In the absence of Ag, Ab-modified AuNPs were monodispersed and could induce a weak CL signal of the luminol–H2O2 system. After the immunoreaction between Ab and Ag, the binding of Ag could cause AuNPs to form dimers, induce their aggregation, and initiate a strong CL signal.

Antigen

AuNPs-Antibody

Without antigen

Materials and methods Apparatus The CL intensity was measured and recorded on an IFFL-D chemiluminescence analyzer (Xi’an Ruimai Electronic Technology, Xi’an, China). Ultraviolet–visible (UV–Vis) adsorption spectra were recorded on a Hitachi U-3900H UV–Vis spectrophotometer. The transmission electron microscopy (TEM) images of AuNPs were taken by using a JEM-2100 TEM device (Japan Electronics). The pH detections were carried out on a PHS-3E analyzer (Jiangsu, China). Reagents Polystyrene 96-well microtiter plates (FOLCON) were used to perform the immunoreactions. Human IgG, goat anti-human IgG, rabbit IgG, goat IgG, bovine serum albumin (BSA) and horseradish peroxidase (HRP)-labeled goat anti-human IgG were purchased from Beijing Dingguo Biotechnology (Beijing, China). Chloroauric acid (HAuCl4
4H2O) was purchased from Sinopharm Chemical Reagent (Shanghai, China). Sodium citrate and sodium chloride were purchased from Tianjing Chemical Reagent (Tianjing, China). Other reagents and chemicals were of analytical grade and used without further purification. Doubly distilled and deionized water was used throughout. Luminol stock solution (2.5  10 2 M) was prepared by dissolving 4.43 g of luminol in 20 ml of NaOH solution (0.10 M) and then diluting to 1 L with water. The luminol solution was stored in the dark for 1 week prior to use to ensure that the reagent property had stabilized. Working solutions of luminol were prepared by diluting the stock solution. Working solutions of H2O2 were prepared fresh daily from 30% (w/w) H2O2. O-Phenylenediamine (OPD) served as the substrate of HRP. The OPD solution (0.4 mg ml 1) was prepared by dissolving 10.0 mg of OPD in 6.1 ml of 0.1 M citric acid solution and then adding 6.4 ml of 0.2 M Na2HPO4 solution and 12.5 ml of deionized distilled water. Next, 40 ll of 30% H2O2 was added in the OPD solution immediately prior to use. The human serum (provided by Shanxi Normal University Hospital) was used as the sample to evaluate the reliability of the proposed immunoassay. Preparation of AuNPs and AuNP-labeled goat anti-human IgG AuNPs (13 nm) were prepared according to a literature method [33]. Briefly, a sodium citrate solution (0.1 M, 1.94 ml) was rapidly added to a boiled HAuCl4 solution (50 ml H2O and 0.167 ml 10% HAuCl4) under vigorous stirring. The mixed solution was boiled for 10 min and further stirred for 15 min. The resulting solution was cooled to room temperature (20 °C) and then stored in the refrigerator (4 °C) before use. The size and shape of the synthesized AuNPs were characterized by TEM. The images showed that their diameter was uniform and their dispersion was very good.

Luminol+H2O2

Strong CL

Luminol+H2O2 Weak CL

Scheme 1. Schematic representation of the proposed CL immunoassay.

3

Chemiluminescence metal immunoassay using AuNPs / Y. Qi et al. / Anal. Biochem. 449 (2014) 1–8

Procedure of immunoreaction and CL detection At room temperature, 200 ll of the above prepared Ab-modified AuNP solution was added to 200 ll of human IgG solutions with different concentrations or diluted serum samples. After that, 100 ll of NaCl solution (0.005 M) was added. The mixture was blended by a WH-861 vortex mixer and incubated for 30 min at room temperature. After that, 100 ll of the Ab-modified AuNP/human IgG mixture solution was introduced into a 40  14-mm quartz tube (used as CL reactor), and then 200 ll of luminol–H2O2 CL solution (the volume ratio of 5.0  10 4 M luminol and 5.0  10 2 M H2O2 was 2:1) was injected, and the CL signal was measured and recorded with the IFFL-D chemiluminescence analyzer. ELISA test The enzyme-linked immunosorbent assay (ELISA) is a classic method of clinical immunoassay and is widely used to detect specific Ag or Ab. In this study, ELISA was used as the standard method to test the reliability of the proposed CL metalloimmunoassay for the determination of human serum samples. The procedure of ELISA was as follows. First, a polystyrene 96well microplate was coated with 100 ll of 1.0  10 5 g ml 1 goat anti-human IgG per well overnight at 4 °C, and the unbound Ab was washed away with 0.01 M phosphate buffer containing 0.05% Tween 20 (PBS-T, pH 7.4) three times. Second, each well was blocked with 200 ll of 0.01 M PBS containing 1% BSA (PBS– BSA, pH 7.4) for 1 h at 37 °C to fill up the uncoated vacancies by Ab and avoid nonspecific binding. Third, after the wells were washed three times with PBS-T (pH 7.4), 100 ll of human serum samples or human IgG solutions with different dilutions in PBS– BSA were pipetted into the wells and incubated for 2 h at 37 °C. Fourth, after the wells were washed six times with PBS-T (pH 7.4), 100 ll of 1.0  10 5 g ml 1 HRP-labeled goat anti-human IgG in PBS-T, which acted as a secondary Ab, was added to each well and incubated for 2 h at 37 °C. Fifth, after a final washing cycle, 200 ll of OPD solution was added as the substrate to react with HRP. The reaction was stopped after 15 min by adding 100 ll of H2SO4 solution (2 M), and the absorbance at 492 nm was detected.

AuNPs-Antibody+Antigen AuNPs-Antibody AuNPs

Abs

The preparation of AuNP-labeled goat anti-human IgG was performed according to the literature [15]. Briefly, after the pH value of AuNP solution was adjusted to 9.0 with 0.1 M K2CO3 solution, 1.0 ml of goat anti-human IgG was added. The mixed solution was agitated and incubated at room temperature for 1 h to make AuNPs and Ab bind fully. The resulting bioconjugate was centrifuged at 12,000 rpm at 4 °C for 30 min to get rid of unbound goat anti-human IgG, and the supernatant was removed. The oily ruby sediment was washed by 0.01 M phosphate buffer containing 1% BSA (PBS–BSA, pH 7.4), and the sediment was resuspended in 5 ml of 0.01 M phosphate buffer (pH 7.4). The bioconjugate solution was stored at 4 °C before use. The same batch of AuNPs modified with goat anti-human IgG was used throughout the total assay process to eliminate possible deviations between batches.

Wavelength (nm) Fig.1. UV–Vis absorption spectra of AuNPs, AuNP/goat anti-human IgG conjugates, and Ab-modified AuNP/human IgG aggregates.

The change of UV–Vis absorption spectra indicated that goat antihuman IgG had been effectively bound on AuNPs. After the immunoreaction with human IgG, the UV–Vis absorption intensity of AuNPs increased greatly. A similar result was obtained by Lu’s group [31]. It is known that SPR is the coherent excitation of all the ‘‘free’’ electrons within the conduction band, leading to an inphase oscillation. In fact, SPR of 13-nm AuNPs mainly depends on the dipole moment of particles; that is to say, it is closely related to its surface states and substance adsorbed on the surface of particles [34]. The immunoreaction between Ab and Ag caused the aggregation of AuNPs by crosslinking, which made the substance adsorbed on the surface of nanoparticles change from Ag to Ag–Ab complex. This surface states change of spherical nanoparticles induced a drastic polarization of the (free) conduction electrons of AuNPs and led to the higher dipole moment, thereby increasing SPR absorption intensity. In brief, the strong electric charge accumulation and oscillation effect occurring in nanoparticles’ specific sites could lead to the enhancement of SPR absorption [35]. TEM analysis was also used to explore the change of AuNPs induced by the immunoreaction in this system. Before the immunoreaction with human IgG, AuNPs were highly dispersed (Fig. 2A–a). However, when human IgG with different concentrations was added into the goat anti-human IgG-modified AuNP solution, the immunoreaction between human IgG and goat antihuman IgG resulted in crosslinking of the IgG–AuNPs and their varying degrees of aggregation (Figs. 2A–b and A–c). To validate the change of AuNPs in the presence of different concentrations of human IgG, the TEM images with different concentrations of human IgG (5.3  10 9 and 3.9  10 7 g ml 1) were obtained. The TEM images (Fig. 2A) showed the differences in the degree of AuNP aggregation for different concentrations of human IgG. It also can be seen that the CL enhancement (Fig. 2B) increased with the increase of the aggregation degree of AuNPs (Fig. 2A). In addition, the crosslinking aggregation of AuNPs after immunoreaction led to a great increase in SPR absorption intensity (Fig. 1). Hence, the results of UV–Vis absorption spectra and TEM were uniform. Catalytic activity of AuNPs on luminol–H2O2 CL system

Results and discussion Change of AuNPs induced by immunoreaction As shown in Fig. 1, unmodified AuNPs have a surface plasmon resonance (SPR) absorption peak at approximately 520 nm, consistent with the literature [33]. After the modification of goat anti-human IgG on AuNPs’ surface, the UV–Vis absorption peak of AuNPs was shifted slightly and the absorption intensity increased a little.

As shown in Fig. 3A, the luminol–H2O2 system was a slow CL reaction in the absence of a catalyst and produced a very weak CL signal (Fig. 3A–a). The 13-nm AuNPs could catalyze this CL reaction system, and the CL signal was enhanced to a certain extent (Fig. 3A–b), consistent with Cui and coworkers’ report [36]. Before the immunoreaction with human IgG, goat anti-human IgG-modified AuNPs also showed weak catalytic activity (Fig. 3A–c), which was almost similar to unmodified AuNPs. After immunoreaction,

4

Chemiluminescence metal immunoassay using AuNPs / Y. Qi et al. / Anal. Biochem. 449 (2014) 1–8

A

B CL intensity

c

b

a

Time(s) Fig.2. TEM images (A) and CL signals (B) of AuNP/goat anti-human IgG conjugates in different concentrations of human IgG: (a) 0 g ml 3.9  10 7 g ml 1. Conditions: luminol, 5  10 4 M; H2O2, 5.0  10 2 M.

the aggregation of goat anti-human IgG-modified AuNPs occurred, and the crosslinking aggregated AuNPs induced by immunoreaction led to a significant increase in luminol–H2O2 CL signal (Fig. 3A–d), indicating that the catalytic activity of crosslinking aggregated AuNPs induced by immunoreaction had an obvious enhancement on the luminol–H2O2 CL reaction. It can be found from Fig. 3A that the catalytic efficiency of crosslinking aggregated AuNPs on luminol CL is 6-fold greater than that of dispersed AuNPs. The peak intensity data were subtracted from the controls (no antigen; Fig. 3B–a) and are presented in Fig. 3B. It can be seen that after background subtraction, the increase on aggregation appears to be approximately 4-fold (peak intensity). To further confirm that the crosslinking aggregation of AuNPs does affect the catalytic property of AuNPs, we examined the catalytic activity of the AuNPs after dissociation of crosslinking aggregated AuNPs. After the immunoreaction, the pH value of Ab-modified AuNP/human IgG mixture was adjusted to 2.0 with 2 M H2SO4 solution, and the obtained mixture solution was kept reacting for 5 min to ensure the complete dissociation of Ab–Ag. Under strong acidic conditions, the electrostatic bond between Ab and Ag cracked so as to make the Ab–Ag bioconjugate dissociated [37]. The pH value of dissociated Ab-modified AuNP/Ag mixture was then adjusted to approximately 7.0 (the pH value of immunoreaction) with 1 M NaOH solution, and its catalytic activity on the luminol–H2O2 CL reaction was detected immediately. The result indicated that after the dissociation of the Ab–Ag bioconjugate, the catalytic activity of Ab-modified AuNPs decreased significantly (Fig. 3A–e). However, it was still a little stronger than the CL signal of Ab-modified AuNPs before immunoreaction (Fig. 3A–c). One possible reason was that the Ab–Ag bioconjugate did not dissociate completely and there were still a few crosslinking aggregated AuNPs in the system. To corroborate the strong acid results, 7 M urea was added to the Ab–Ag bioconjugate and acted for 10 min to complete the dissociation of Ab–Ag. As shown in Fig. 3C, the dissociation of the Ab–Ag bioconjugate by the addition

1

; (b) 5.3  10

9

g ml

1

; (c)

of urea resulted in a significant decrease of catalytic activity of Abmodified AuNPs. The catalysis activity of AuNPs was found to be related to their size [36], surface state [26], and morphology [20], indicating that more than one parameter was playing a part to influence the catalytic activity of AuNPs. It is well known that the rate of heterogeneous catalysis increases with increasing available active surface area of the catalyst [38,39]. Therefore, a higher active surface area of smaller size particles would be present in a given mass of catalyst material, leading to a higher rate of catalytic reaction. From Fig. 2A, it could be found that the distance between Ab-modified AuNPs decreased significantly after immunoreaction and many larger AuNP aggregates were formed. Hence, the catalytic activity of crosslinking aggregated AuNPs induced by immunoreaction may decrease in comparison with the dispersed AuNPs due to their increased size after immunoreaction. However, it was interesting that an opposite result was observed in our experiments, indicating that besides the size effect, there must be other ones that played roles. Cui and coworkers [36] found that when AuNPs were below 38 nm in diameter, quantum size effects began to function with an increase in band gap energy, leading to a higher activation energy that was needed for electron transfer, and the ability of electron transfer of AuNPs had a marked influence on their catalytic activity for the luminol–H2O2 CL system [38,39]. A greater ability of electron transfer of AuNPs usually results in a higher rate of heterogeneous catalysis. Accordingly, the catalytic activity of crosslinking aggregated AuNPs induced by immunoreaction for luminol–H2O2 CL increased rapidly compared with the dispersed 13-nm AuNPs, probably because the quantum size effects began to weaken and the activation energy that was needed for electron transfer decreased. In addition, according to the literature [36,38,39], during the process of AuNPs’ catalysis on the luminol–H2O2 CL reaction, the adsorption of H2O2 onto the surface of AuNPs would cause partial electron transfer from AuNPs to the adsorbed H2O2, which was

Chemiluminescence metal immunoassay using AuNPs / Y. Qi et al. / Anal. Biochem. 449 (2014) 1–8

d

CL intensity

A

e

5

the increase of electron density in the conduction bands of crosslinking aggregated AuNPs may result in increasing the catalytic activity of AuNPs for the luminol–H2O2 CL system. In conclusion, crosslinking aggregated AuNPs induced by immunoreaction showed the enhanced catalytic effect because of low activation energy for electron transfer, proper surface area, and high electron density and oscillation effect in the conduction bands.

c b

Optimization of the reaction condition

Time (s)

B CL intensity

b

c

a

CL intensity

C a

b

Time (s) Fig.3. (A) Effect of AuNPs on the luminol–H2O2 CL reaction: (a) without AuNPs; (b) AuNPs; (c) AuNPs + Ab; (d) AuNPs + Ab + Ag after immunoreaction; (e) AuNPs + Ab + Ag after acid dissociation. (B) Effect of AuNPs on the luminol–H2O2 CL reaction: (a) AuNPs + Ab; (b) AuNPs + Ab + Ag after immunoreaction subtracted from (a); (c) AuNPs + Ab + Ag after acid dissociation subtracted from (a). (C) Effect of AuNPs on the luminol–H2O2 CL reaction: (a) AuNPs + Ab + Ag after immunoreaction; (b) AuNPs + Ab + Ag after urea dissociation. Conditions: luminol, 5  10 4 M; H2O2, 5.0  10 2 M.

known to be particle-mediated electron transfer. Therefore, the electron density in the conduction bands of AuNPs would have a significant effect on the particle-mediated electron transfer processes, and higher electron density would be favorable for the particle-mediated electron transfer processes and, thus, would result in an increase in catalytic activity of AuNPs. In fact, the strong electric charge accumulation and oscillation effect occurring in nanoparticles’ specific sites would lead to the enhancement of SPR absorption [35]. So, SPR absorption intensity from UV–Vis absorption spectra is considered to be able to indicate the electron density in the conduction bands of nanoparticles [34]. UV–Vis absorption spectra analysis of AuNPs, AuNP/goat anti-human IgG conjugates, and Ab-modified AuNP/human IgG aggregates are shown in Fig. 1. UV–Vis absorption spectra showed that the SPR absorption intensity of AuNPs increased significantly after immunoreaction, indicating that the electron density in the conduction bands of crosslinking aggregated AuNPs induced by immunoreaction was much higher than that of dispersed AuNPs. Therefore,

In this assay, AuNPs are taken as the signal element and their size influences not only the Ab–AuNP binding but also the catalytic activity for the luminol–H2O2 CL reaction. The research by Cui and coworkers showed that AuNPs’ catalytic activity was enhanced with the increase of their size [36]. For this immunoassay, the CL intensity from the dispersed AuNPs is the background signal. Therefore, the small-sized AuNPs should be used to achieve low background signal so as to improve the sensitivity. However, the too small-sized AuNPs have relatively small surface area, which is not favorable for the Ab binding. Therefore, 13-nm AuNPs were used in this CL immunoassay. Furthermore, the concentration of AuNPs can also influence the sensitivity of the CL immunoassay. If the AuNP concentration is too high, the system is not sensitive enough to detect a smaller number of analytes and does not differentiate them from the background. In this system, we used 1.7 nM AuNPs (13 nm) for all experiments. The amount of goat anti-human IgG could affect the labeled ratio on the surface of AuNPs. Because the improper ratio for preparing Ab-modified AuNPs can bring nonspecific bindings, it will influence the sensitivity and specificity of the assay. To prevent the nonspecific bindings, a flocculation test was performed to determine the minimum amount of goat anti-human IgG required to coat the exterior of AuNPs according to the literature [15,29]. Both the bare AuNPs and the Ab-modified AuNPs have a maximum absorbance at approximately 520 nm. The salt can induce the aggregation of AuNPs, and the absorbance spectrum of aggregated AuNPs is red shifted. Ab-modified AuNPs can prevent salt-induced aggregation, and its absorbance spectrum will not change. Based on this, a flocculation experiment was carried out. Briefly, 0.1 ml of Ab with different concentrations was added to 0.5 ml of AuNP solution and interacted for 30 min, and then 0.1 ml of NaCl (0.1 M) was added to incubate at room temperature for 5 min. The absorbance at 520 nm was detected. From Fig. 4, it can be seen that the minimum Ab amount required for suitable labeling is

Abs

a

Ab concentration (mg/mL) Fig.4. Flocculation test for the proper ratio of the amount of AuNPs to antibody. Conditions: AuNPs, 1.7 nM; k = 520 nm.

Chemiluminescence metal immunoassay using AuNPs / Y. Qi et al. / Anal. Biochem. 449 (2014) 1–8

Analytical performance of this CLIA for human IgG This CL analytical protocol based on the aggregation of Ab-modified AuNPs induced by immunoreaction was assessed by measuring the dependence of the CL intensity on the concentration of human IgG. As shown in Fig. 6, the CL intensity increased with

(A)

(A)

k j

CL intensity

approximately 0.45 mg ml 1, consistent with the report of Li and coworkers [29]. Hence, the concentration of goat anti-human IgG (Ab) used for labeling in this work was controlled at 0.5 mg ml 1, which was approximately 10% more than the minimum amount required to ensure that the amount of Ab was enough for preventing AuNPs from aggregation. To make the immunoreaction more efficient and more complete, the immunoreaction time was optimized. It is generally known that the longer the reaction time, the better the Ab–Ag bioconjugate. As shown in Fig. 5A, at the beginning of the immunoreaction, the CL signal was weaker. The CL signal increased gradually with the time was prolonged, then more rapidly when the time exceeded 20 min, and remained almost constant after 30 min. Therefore, 30 min was chosen as the optimal immunoreaction time for this assay. In general, a certain concentration of salt would induce AuNPs to aggregate very easily due to the electrostatic screening effect [40,41] and to further influence the CL signal. In addition, appropriate electrolyte solution is also beneficial to the immunoreaction. Therefore, the salt concentration in the immunoreaction was also optimized in this work. As shown in Fig. 5B, the CL intensity was weak when the NaCl concentration was less than 0.5 mM, indicating that AuNPs were monodispersed in the lower salt concentration. But the CL intensity increased sharply as a result of the aggregation of AuNPs when the NaCl concentration was greater than 0.5 mM. The CL intensity decreased with further increases of the NaCl concentration when the NaCl concentration was greater than 1 mM. It is likely that the higher ionic strength can cause the denaturation of proteins (Ag and Ab), which is unfavorable for the immunoreaction and may further affect the aggregation of AuNPs induced by immunoreaction. Therefore, the optimal concentration of NaCl is 1.0  10 3 M in the system for immunoassay. In addition, the important experimental parameters influencing the luminol–H2O2–AuNPs (13 nm) CL system, including the concentrations of luminol and H2O2 and medium pH, were optimized. It was found that both signal and background increased as the luminol concentration was raised, and the signal/noise ratio was highest at 0.5 mM luminol. The effect of the H2O2 concentration on the CL was studied in the range from 0.01 to 100 mM, and the maximum CL emission was observed at 5  10 2 M H2O2 solution. The experimental results showed that the optimal pH for this CL system was 12.0, in agreement with the results of our previous studies [26,27].

d

h

g

f

e

i

c a

b

Time(s)

(B)

k

CL intensity

6

j i h e

g

f

d c b

Ag Fig.6. CL signals for triplicate measurement (A) and CL intensity subtracted from the controls (no antigen) (B) of AuNPs in different concentrations of human IgG: (a) 0 g ml 1; (b) 2.6  10 10 g ml 1; (c) 1.3  10 9 g ml 1; (d) 2.6  10 9 g ml 1; (e) 3.9  10 9 g ml 1; (f) 5.3  10 9 g ml 1; (g) 2.6  10 8 g ml 1; (h) 1.3  10 7 g ml 1; (i) 2.6  10 7 g ml 1; (j) 3.9  10 7 g ml 1; (k) 5.3  10 7 g ml 1. Conditions: luminol, 5  10 4 M; H2O2, 5.0  10 2 M.

the increase of concentration of human IgG. Under the above optimized experimental conditions, the CL intensity of this CL system was linear with the concentration of human IgG in the range from 2.6  10 10 to 5.3  10 7 g ml 1 (Fig. 7). The detection limit (3r) was 3.2  10 11 g ml 1. The detection limit of the current protocol is two orders of magnitude lower than that of the reported non-stripping AuNP-based CL immunoassay [20,23]. In addition, a series of five repetitive measurements of 6.2  10 9 g ml 1 human IgG was used for estimating the precision, and the relative standard deviation (RSD) was 4.6%, indicating good precision of the proposed CL immunoassay method. The immunoreaction has high specificity in theory, but some nonspecific reactions would reduce the sensitivity of the assay. The specificity of the proposed CL immunoassay was investigated by detecting the CL signals arising from human IgG, goat IgG, rabbit IgG, and BSA. The AuNPs modified with goat anti-human IgG were used for the immune recognition, and goat IgG, rabbit IgG, and BSA served as contrast reagents. The concentration of all the analytes was kept the same (1  10 7 g ml 1). As presented in Fig. 8, a distinct CL signal was observed from the human IgG, whereas the CL intensities from goat IgG, rabbit IgG, and BSA were very weak. The

CL intensity

CL intensity

(B)

-3

Time (min)

NaCl concentration (×10 mol/L)

Fig.5. Effect of immunoreaction time (A) and NaCl concentration (B) on the CL intensity. Conditions: luminol, 5  10

4

M; H2O2, 5.0  10

2

M.

Chemiluminescence metal immunoassay using AuNPs / Y. Qi et al. / Anal. Biochem. 449 (2014) 1–8

(A)

7

y = 43.39x + 88.43 r = 0.9902

Human IgG concentration (×10-9 g/mL)

CL intensity

CL intensity

(B)

y = 77.32x + 263.4 r = 0.9912

Human IgG concentration (×10 -7g/mL)

Fig.7. Calibration curves of CL intensity versus human IgG concentration. The concentrations of human IgG: (A) 2.6  10-10 ~ 5.3  10-9 g mL-1, (B) 2.6  10-8 ~ 5.3  10-7g mL-1.

CL immunoassay is reliable and can be used to detect human IgG in human serum samples.

Relative CL intensity

a

Conclusions

c

b

d

Analyte

-8

ELISA (×10 g/mL)

Fig.8. Specificity for the determination of human IgG using the proposed immunoassay: (a) human IgG; (b) BSA; (c) goat IgG; (d) rabbit IgG.

y = 1.0243x - 0.2864 r = 0.9905

A novel and sensitive strategy to convert the Ab–Ag recognition event into CL signals by employing Ab–AuNPs as signaling probes has been proposed in this study. This strategy has several significant advantages. First, the assay developed does not require the stripping of metal nanoparticles, which is usually rigorous to ensure that nanoparticles dissolve fully. So, it avoids the high background from the tedious and difficult stripping treatment, and it offers high sensitivity and good reproducibility. Second, the preparation of 13-nm AuNPs used as labels in the proposed method by the normal citrate reduction method is much easier than that of irregular AuNPs used in other non-stripping CL immunoassays. Third, it is a noncompetitive one-step operation, thereby avoiding the multiple steps of immunoreactions and washings. So, it is time saving, with the whole assay process able to be finished within 0.5 to 1 h. Fourth, the assay is homogeneous and occurs in the liquid phase, making it easy to automate by standard robotic manipulation of microwell plates. This study initiates a new concept for the CL metalloimmunoassay based on metal nanoparticles. Acknowledgments

CL method (×10 -8 g/mL)

Fig.9. Correlation of the proposed CL immunoassay and the ELISA in the detection of human IgG for human serum samples.

results indicated that AuNPs coated with goat anti-human IgG could specifically recognize the human IgG and that the specificity of the current CL immunoassay had an acceptable level.

This work was supported financially by the Natural Science Foundation of Fujian Province of China (2011J05016), the Foundation of Fujian Educational Committee (JA11177 and JA12237), the Special Science Funding of Provincial Universities of Fujian (JK2012032), and the Scientific Research Foundation of Fujian University of Technology (GY-Z10055). References

Analytical application A series of human serum samples was analyzed simultaneously with the proposed method and ELISA to investigate the applicability and reliability of the proposed CL immunoassay for clinical diagnostics. The comparative results are shown in Fig. 9. The IgG contents in human serum samples obtained by the proposed method were identical to the data of ELISA. A good correlation is obtained between the proposed method and the ELISA, and the correlation equation is y = 1.0243x – 0.2864, where y is the determination results of ELISA and x is that of the proposed method. The correlation coefficient is 0.9905, indicating that the proposed

[1] I.A. Darwish, A.M. Mahmoud, A.A. Al-Majed, Generic simple enzyme immunoassay approach to avert small molecule immobilization problems on solid phases: application to the determination of tobramycin in serum, Talanta 72 (2007) 1322–1328. [2] T. Pulli, M. Höyhtyä, H. Söderlund, K. Takkinen, One-step homogeneous immunoassay for small analytes, Anal. Chem. 77 (2005) 2637–2642. [3] A. Bromberg, R.A. Mathies, Homogeneous immunoassay for detection of TNT and its analogues on a microfabricated capillary electrophoresis chip, Anal. Chem. 75 (2003) 1188–1195. [4] S.A. Berson, R.S. Yalow, Assay of plasma insulin in human subjects by immunological methods, Nature 184 (1959) 1648–1649. [5] H.L. Qi, C.X. Zhang, Homogeneous electrogenerated chemiluminescence immunoassay for the determination of digoxin, Anal. Chim. Acta 501 (2004) 31–35. [6] M. Bruchez Jr., M. Moronne, P. Gin, S. Weiss, A.P. Alivisatos, Semiconductor nanocrystals as fluorescent biological labels, Science 281 (1998) 2013–2016.

8

Chemiluminescence metal immunoassay using AuNPs / Y. Qi et al. / Anal. Biochem. 449 (2014) 1–8

[7] J.L. Yuan, K. Matsumoto, H. Kimura, A new tetradentate-diketonate-europium chelate that can be covalently bound to proteins for time-resolved fluoroimmunoasaay, Anal. Chem. 70 (1998) 596–601. [8] F.N. Lamari, E.D. Anastassiou, F. Kolonitsiou, G. Dimitracopoulos, N.K. Karamanos, Potential use of solid phase immunoassays in the diagnosis of coagulase-negative staphylococcal infections, J. Pharm. Biomed. Anal. 34 (2004) 803–810. [9] G.L. Wang, J.L. Yuan, B.L. Gong, K. Matsumoto, Z.D. Hu, Immunoassay by graphite furnace atomic absorption spectrometry using a metal chelate as a label, Anal. Chim. Acta 448 (2001) 165–172. [10] J. Reichert, A. Csaki, J.M. Kohler, W. Fritzsche, Chip-based optical detection of DNA hybridization by means of nanobead labeling, Anal. Chem. 72 (2000) 6025–6029. [11] L.N. Lewis, Chemical catalysis by colloids and clusters, Chem. Rev. 93 (1993) 2693–2730. [12] D.D. Awschalom, D.P. DiVincenzo, J.F. Smyth, Macroscopic quantum effects in nanometer-scale magnets, Science 258 (1992) 414–421. [13] H. Weller, Self-organized superlattices of nanoparticles, Angew. Chem. Int. Ed. Engl. 35 (1996) 1079–1081. [14] C. Dodeigne, L. Thunus, R. Lejeune, Chemiluminescence as diagnostic tool: a review, Talanta 51 (2000) 415–439. [15] Z.P. Li, Y.C. Wang, C.H. Liu, Y.K. Li, Development of chemiluminescence detection of gold nanoparticles in biological conjugates for immunoassay, Anal. Chim. Acta 551 (2005) 85–91. [16] D. Hu, H. Han, R. Zhou, F. Dong, W. Bei, F. Jia, H. Chen, Gold(III) enhanced chemiluminescence immunoassay for detection of antibody against ApxIV of Actinobacillus pleuropneumoniae, Analyst 133 (2008) 768–773. [17] S. Gupta, S. Huda, P.K. Kilpatrick, O.D. Velev, Characterization and optimization of gold nanoparticle-based silver-enhanced immunoassays, Anal. Chem. 79 (2007) 3810–3820. [18] A. Fan, C. Lau, J. Lu, Magnetic bead-based chemiluminescent metal immunoassay with a colloidal gold label, Anal. Chem. 77 (2005) 3238–3242. [19] Z.P. Li, C.H. Liu, Y.S. Fan, Y.C. Wang, X.R. Duan, A chemiluminescent metalloimmunoassay based on silver deposition on colloidal gold labels, Anal. Biochem. 359 (2006) 247–252. [20] Z.P. Wang, J.Q. Hu, Y. Jin, X. Yao, J.H. Li, In situ amplified chemiluminescent detection of DNA and immunoassay of IgG using special-shaped gold nanoparticles as label, Clin. Chem. 52 (2006) 1958–1961. [21] X.Y. Yang, Y.S. Guo, A.G. Wang, Luminol/Antibody labeled gold nanoparticles for chemiluminescence immunoassay of carcinoembryonic antigen, Anal. Chim. Acta 666 (2010) 91–96. [22] Y.M. Liu, L. Mei, L.J. Liu, L.F. Peng, Y.H. Chen, S.W. Ren, Sensitive chemiluminescence immunoassay by capillary electrophoresis with gold nanoparticles, Anal. Chem. 83 (2011) 1137–1143. [23] C.F. Duan, Y.Q. Yu, H. Cui, Gold nanoparticle-based immunoassay by using non-stripping chemiluminescence detection, Analyst 133 (2008) 1250–1255. [24] S. Bi, Y. Yan, X. Yang, S. Zhang, Gold nanolabels for new enhanced chemiluminescence immunoassay of a-fetoprotein based on magnetic beads, Chem. Eur. J. 15 (2009) 4704–4709. [25] X.Y. Yang, Y.S. Guo, S. Bi, S.S. Zhang, Ultrasensitive enhanced chemiluminescence enzyme immunoassay for the determination of a-

[26]

[27]

[28]

[29]

[30]

[31] [32]

[33] [34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

fetoprotein amplified by double-codified gold nanoparticles labels, Biosens. Bioelectron. 24 (2009) 2707–2711. Y. Qi, B. Li, Z. Zhang, Label-free and homogeneous DNA hybridization detection using gold nanoparticles-based chemiluminescence system, Biosens. Bioelectron. 24 (2009) 3581–3586. Y. Qi, B. Li, A sensitive, label-free, aptamer-based biosensor using a gold nanoparticle-initiated chemiluminescence system, Chem. Eur. J. 17 (2011) 1642–1648. Y. Qi, B. Li, Enhanced effect of aggregated gold nanoparticles on luminol chemiluminescence system and its analytical application, Spectrochim. Acta A 111 (2013) 1–6. B. Du, Z. Li, Y. Cheng, Homogeneous immunoassay based on aggregation of antibody-functionalized gold nanoparticles coupled with light scattering detection, Talanta 75 (2008) 959–964. N.T.K. Thanh, Z. Rosenzweig, Development of an aggregation-based immunoassay for anti-protein A using gold nanoparticles, Anal. Chem. 74 (2002) 1624–1628. C.X. Zhang, Y. Zhang, X. Wang, Z.M. Tang, Z.H. Lu, Hyper-Rayleigh scattering of protein-modified gold nanoparticles, Anal. Biochem. 320 (2003) 136–140. C. Xie, F. Xu, X. Huang, C. Dong, J. Ren, Single gold nanoparticles counter: an ultrasensitive detection platform for one-step homogeneous immunoassays and DNA hybridization assays, J. Am. Chem. Soc. 131 (2009) 12763–12770. K.C. Grabar, R.G. Freeman, M.B. Hommer, M.J. Natan, Preparation and characterization of Au colloid monolayers, Anal. Chem. 67 (1995) 735–743. S. Link, M.A. El-sayed, Shape and size dependence of radiative, non-radiative, and photothermal properties of gold nanocrystals, Int. Rev. Phys. Chem. 19 (2000) 409–453. J. Fang, S. Du, S. Lebedkin, Z. Li, R. Kruk, M. Kappes, H. Hahn, Gold mesostructures with tailored surface topography and their self-assembly arrays for surface-enhanced Raman spectroscopy, Nano Lett. 10 (2010) 5006– 5013. Z.F. Zhang, H. Cui, C.Z. Lai, L.J. Liu, Gold nanoparticle-catalyzed luminol chemiluminescence and its analytical applications, Anal. Chem. 77 (2005) 3324–3329. X. Chu, C.H. Lin, G.L. Shen, L.Q. Yu, Kinetic studies of immunogobulin M immunoreaction using quartz crystal microbalance methodology, Chem. J. Chin. Univ. 17 (1996) 1025–1029. R.K. Sharma, P. Sharma, A. Maitra, Size-dependent catalytic behavior of platinum nanoparticles on the hexacyanoferrate(III)/thiosulfate redox reaction, J. Colloid Interface Sci. 265 (2003) 134–140. A. Henglein, Physicochemical properties of small metal particles in solution: Microelectrode reactions, chemisorption, composite metal particles, and the atom-to-metal transition, J. Phys. Chem. 97 (1993) 5457–5471. H. Li, L.J. Rothberg, Label-free colorimetric detection of specific sequences in genomic DNA amplified by the polymerase chain reaction, J. Am. Chem. Soc. 126 (2004) 10958–10961. C.A. Mirkin, Programming the assembly of two- and three-dimensional architectures with DNA and nanoscale inorganic building blocks, http:// pubs.acs.org/doi/abs/10.1021/ic991123r – ic991123rAF2#ic991123rAF2Inorg. Chem. 39 (2000) 2258–2272.