Article pubs.acs.org/ac
Ultrasensitive Electrogenerated Chemiluminescence Peptide-Based Method for the Determination of Cardiac Troponin I Incorporating Amplification of Signal Reagent‑Encapsulated Liposomes Honglan Qi,*,† Xiaoying Qiu,† Danping Xie,† Chen Ling,‡ Qiang Gao,† and Chengxiao Zhang*,† †
Key Laboratory of Applied Surface and Colloid Chemistry, Ministry of Education, School of Chemistry and Chemical Engineering, Shaanxi Normal University, Xi’an 710062, P.R. China ‡ Xianyang Central Hospital, Xi’an 713199, P.R. China S Supporting Information *
ABSTRACT: An ultrasensitive electrogenerated chemiluminescence peptide-based (ECL-PB) method for the determination of cardiac troponin I (TnI) incorporating amplification of signal reagent-encapsulated liposome was reported for the first time. A synthesized short linear specific binding peptide (FYSHSFHENWPSK) was employed as a molecular recognition element for TnI, which was a reliable biomarker for detecting cardiac injury. Liposomes assembled using a standard sonication procedure were designed as the carrier of ECL signal reagents [bis(2,2′-bipyridine)-4,4′-dicarboxybipyridine ruthenium-di(N-succinimidyl ester) bis(hexafluorophosphate)] for signal amplification. The magnetic capture peptides for the enrichment of the target protein and magnetic separation were synthesized by covalently attaching the peptides to the surface of magnetic beads via an acylation reaction, and the liposome peptides were synthesized by covalently attaching the peptides to the signal reagent-encapsulated liposomes. In the presence of TnI, sandwich-type conjugates were generated in incubation of the magnetic capture peptides and the liposome peptides. After a magnetic separation, the sandwich-type conjugates were treated with ethanol and, thus, a great number of the ECL reagents were released and measured by the ECL method at a bare glassy carbon electrode with a potential pulse of +1.15 V versus Ag/AgCl in the presence of tri-n-propylamine. The increased ECL intensity has good linearity with the logarithm of the TnI concentration in the range from 10 pg/mL to 5.0 ng/mL, with an extremely low detection limit of 4.5 pg/ mL. The proposed ECL-PB method was successfully applied to the detection of TnI in human serum samples. This work demonstrated that the employment of the magnetic capture peptides for the enrichment of the target proteins and magnetic separation and the liposome peptides for the signal amplification and polyvalent binding motifs may open a new door to ultrasensitive detection of proteins in clinical analyses. e propose, herein, a novel and ultrasensitive “sandwich” electrogenerated chemiluminescence peptide-based (ECL-PB) method for protein detection. Cardiac troponin I (TnI), with a molecular weight of 24 kDa, is a known reliable biomarker for detecting cardiac injury and is selected to be our target protein.1 A synthesized short linear specific-binding peptide (FYSHSFHENWPSK, Figure S-1 of the Supporting Information)2 was employed as a molecular recognition
W
© 2013 American Chemical Society
element and bis(2,2′-bipyridine)-4,4′-dicarboxybipyridineruthenium di(N-succinimidyl ester) bis(hexafluorophosphate) (abbreviated as Ru1) served as an ECL signal reagent. A liposome assembled using a standard sonication procedure was Received: October 23, 2012 Accepted: March 18, 2013 Published: March 18, 2013 3886
dx.doi.org/10.1021/ac4005259 | Anal. Chem. 2013, 85, 3886−3894
Analytical Chemistry
Article
Figure 1. Schematic diagram of liposome peptide-based electrogenerated chemiluminescence method for the determination of cardiac troponin I.
TnI bioassays have also been reported using colorimetric,7 electrochemical,4,8 fluorescent,9,10 chemiluminescence,11 and ECL techniques.12,13 Among these, ECL immunoassays have attracted considerable interest for protein detection, due to its high sensitivity and selectivity and wide dynamic range.14,15 For example, several ECL bioassays have been developed for the determination of proteins with high sensitivity.16,17 Much effort has been devoted to improve the sensitivity of protein assays, employing an appropriate carrier for a heavily labeled probe, in which a large number of signal reagents and molecular recognition elements, such as antibodies,18−20 aptamers,21 or antimicrobial peptides22 and carbohydrates,23 are loaded. A lot of the work has been reported, such as a luminol multilabeled carrier protein,18 glucose oxidase multilabeled gold nanorods,19 and Ru(bpy)32+-encapsulated liposomes for the ECL immunoassay,20 the ruthenium(II) complex derivative multilabeled dendrimer for ECL aptamerbased assay,21 fluorophore multilabeled gold nanoparticle for a fluorescence peptide-based assay,22 and mannose multilabeled gold nanoparticles for surface plasmon resonance assay of Con A.23 We developed an ECL immunoassay for detection of digoxin by using BSA as a carrier protein for luminol-labeled digoxin18 and Ru(bpy)32+-labeled digoxin,24 as an ECL probe. A 34:4:1 label ratio for luminol:digoxin:BSA was obtained, and a detection limit of 2.8 × 10−10 g/mL digoxin was achieved.18 Bard and colleagues successfully developed an ultrasensitive ECL DNA hybridization method using polystyrene microspheres/beads (PSB) as carriers of ECL labels.25 The loading capacity of the ECL labels per PSB were as high as 7.5 × 109 molecules, thus a very large amplification factor of Ru(bpy)32+labeled molecules for each target molecule of single-stranded DNA (t-ssDNA). Bard and colleagues also developed an ECL sandwich-type immunoassay of human C-reactive protein (CRP), in which multiple Ru(bpy)32+ molecules were added to a single antibody by encapsulating Ru(bpy)32+ compounds in liposomes. This approach demonstrates enormous signal amplification when compared with the conventional ECL immunoassays with one label per antibody. A detection limit of 100 ng/mL for CRP was obtained.20 However, the antibody drawbacks associated with their production was stability. Therefore, the sensitivity of protein assay should be further
designed as the carrier of the ECL signal reagent for signal amplification. As shown in Figure 1, the magnetic capture peptides (MBs-peptide, I) for the enrichment of the target protein and magnetic separation were synthesized by covalently attaching the peptides to the surface of isothiocyanic acidcoated magnetic beads (MBs) via an acylation reaction, and the liposome peptides (Ru-L-peptide, II) were synthesized by covalently attaching the peptides to the surface of Ru1encapsulated liposomes. In the presence of TnI, sandwich-type conjugates (Ru1-L-peptide < TnI > MBs-peptide) were formed in incubation of the magnetic capture peptides and the liposome peptides. After a magnetic separation, the sandwichtype conjugates were treated with a selected releasing agent (ethanol) and, thus, a great number of the ECL reagents were released and measured by the ECL method at a bare glassy carbon electrode with a potential pulse to +1.15 V versus Ag/ AgCl (saturated KCl) in the presence of tri-n-propylamine. The increased ECL intensity has good linearity with the logarithm of the TnI concentration in the range from 10 pg/mL to 5.0 ng/mL, with an extremely low detection limit of 4.5 pg/mL. Sensitivity in protein assays is important in clinical diagnostics, environmental pollution evaluation, and biological warfare agent detection. In particular, the detection of cancer biomarkers has become an increasingly significant issue for early stage diagnosis and highly reliable predictions.3 Cardiac troponin I is part of the troponin complex that is present in cardiac muscle tissues. The concentration of TnI in blood rises rapidly within 4−6 h after the onset of an acute myocardial infarction and reaches the maximum at approximately 12 h. After 6−8 days, the TnI level returns to normal and, thus, the concentration of TnI in blood can provide a long diagnostic window for detecting cardiac injury.4,5 Immunoassays that employ antibody−antigen interactions are some of the most significant analytical methods in the quantitative detection of biomarkers, due to the highly specific molecular recognition of the antibody/antigen. Traditional methods for the quantitative detection of TnI, such as enzyme-linked immunosorbent assays and radioimmunoassays, are most frequently used. However, most of the current reported traditional methods are limited by the sensitivity, sample size, and equipment cost and do not meet the urgent needs in many situations.6 Additionally, the 3887
dx.doi.org/10.1021/ac4005259 | Anal. Chem. 2013, 85, 3886−3894
Analytical Chemistry
Article
ECL measurements were performed with a MPI-A ECL detector (Xi’an Remax Electronics). A commercial cylindroid glass cell was used as an ECL cell, which contained a conventional three-electrode system that consisted of a glassy carbon electrode (2.0 mm diameter) as the working electrode, a platinum wire as the counter electrode, and an Ag/AgCl (saturated KCl) as the reference electrode. ECL emissions were detected with a photomultiplier tube (PMT) that was biased at −900 V, unless otherwise stated. Electrochemical experiments were performed with a CHI 660 electrochemical workstation (Chenhua Instruments Company). Fluorescence imaging was conducted with a Nikon Eclipse TE 300 inverted microscope (Nikon Instruments Inc., Melville, NY) coupled with a Magnafire model S99806 Olympus America CCD camera (Olympus America, Melville, NY). Preparation of Signal Reagent-Encapsulated Liposomes (Ru1-L) and Liposome Peptide (Ru1-L-peptide). Signal reagent-encapsulated liposomes (Ru1-L) were prepared according to the procedure from the literature36,37 with slight modifications. Phosphocholine, cholesterol, and phosphoethanolamine (6:6:1 molar ratio, 65 μmol, and 0.0370 g in total) were dissolved in 4 mL of chloroform, followed by the addition of 1 mL of 0.1 M phosphate buffer (0.1 M NaH2PO4, 0.1 M Na2HPO4, PB, pH 7.4) containing 1 mM Ru1. After sonicating for 5 min, the organic solvent was removed by rotary evaporation under reduced pressure at 45 °C to form a thin lipid film, leaving a yellow viscous film of liposomes. Afterward, the film was hydrated in 1 mL of 0.1 M PB for 1 h with vigorous shaking at 45 °C, and then, sonicated using a bathtype sonicator (100 Hz, KQ5200DE Kun Shan Ultrasonic Instruments Company, Ltd.). Finally, the liposome fraction was dialyzed to remove the unencapsulated reagents at 4 °C for 12 h using the MD34−2 Da molecular weight cutoff membrane (Viskase) against 0.1 M PB (pH 7.4), and stored in 0.1 M PB (pH 7.4) at 4 °C. The average size and the distribution of the liposomes were measured using a Malvern Zetasizer (BI90Plus). Transmission electron micrograph (TEM) images were obtained with a JEM-2100 transmission electron microscope (JEOL), operating at 200 kV with an ultrahigh-resolution pole piece, providing a point resolution of 2 Å. Carbon-coated copper grids were used for the preparation of the specimens. The samples were negatively stained with 0.1% phosphotungstic acid and left at room temperature until a dried film was obtained.38 The liposome peptides were synthesized by covalently coupling the peptide to the signal reagent-encapsulated liposome, according to ref 36 with slight modification. Briefly, 500 μL of 2.5% (v/v) glutaraldehyde was added into 500 μL of a signal reagent-encapsulated liposome prepared above, and the mixture was stirred gently for 2 h, and dialyzed to remove an excess of glutaraldehyde for 12 h at 4 °C, using a MD34−2 Da molecular weight cutoff membrane. Then, 60 μL of the peptide solution (FYSHSFHENWPSK, 1.2 mM) was added into the dialyzed liposome suspension and incubated for more than 2 h. Finally, 500 μL of 1% (w/w) BSA was added into the above suspension and kept overnight to block the residual aldehyde groups and nonspecific binding sites on the surface of the liposome. The excess of the peptide and BSA were separated from the resulting suspension upon centrifugation at 3000 rpm for 10 min. The resulting liposome peptides (Ru1-L-peptide) were resuspended and adjusted with 0.1 M PB (pH 7.4) to the
improved and an alternative molecular recognition element should be explored. Short linear binding peptides obtained using phage display can function as substitute antibodies and have several advantages when compared with antibodies, including the fact that peptides can be created synthetically in a reliable and costeffective manner. Peptides are more stable, resistant to harsh environments, and more amenable to engineering at the molecular level than antibodies.26 Several peptide-based methods have been developed for the detection of proteins, such as a colorimetric peptide-based assay for TnI, 2 fluorescence peptide-based assay for protein kinase,22 electrochemical peptide-based assay for prostate-specific antigen,27,28 protein kinase,29 human immunodeficiency virus,30,31 ECL peptide-based assay for prostate-specific antigen,32 and troponin I.33 Recently, Park et al. reported a new peptide (FYSHSFHENWPS) that selectively bound to TnI with a disassociation constant of the complex at the nanomolar level.2 We developed an ECL peptide-based method for the determination of TnI using this peptide as a molecular recognition element.33 However, the detection limit of the previous ECL method (1.2 × 10−10 g/mL) is limited, since only one ECL molecule is attached directly to each peptide. Binding interactions can be increased by the binding of multiple ligands on one entity to multiple receptors.34,35 Thus, the employment of the polyvalent binding motif of the probe on the appropriate platform for the target protein is promising in the development of sensitive peptide-based bioanalytic methods. The principle described here, which is based on the idea of encompassing heavy labels in larger carriers and on polyvalent binding motifs, employing the Ru1-encapsulated liposome peptides and the magnetic capture peptides, is applied to the development of a highly sensitive ECL peptide-based method for the determination of proteins. In this paper, as a proof-ofprinciple, the characteristics of the Ru1-encapsulated liposomes and the liposome peptide, optimization of the releasing agent, and the analytical performance for the analysis of TnI using this liposome peptide are presented.
■
EXPERIMENTAL SECTION Reagents and Apparatus. The peptide FYSHSFHENWPSK (13 mer, MW = 1665.80) chemically synthesized was purchased from Sinoasis Pharmaceuticals, Inc. Cardiac troponin I (TnI, human heart) was obtained from Abcam Inc. (Cambridge, United Kingdom). Phosphocholine, phosphoethanolamine, cholesterol and bis(2,2′-bipyridine)-4,4′-dicarboxybipyridine-ruthenium di(N-succinimidyl ester) bis(hexafluorophosphate) (Ru(bpy)2(dcbpy-NHS)(PF6)2, abbreviated as Ru1) were obtained from Sigma-Aldrich. Isothiocyanic acid-coated MBs were obtained from Shaanxi Lifegen Company, Ltd. Human alpha-fetoprotein (AFP) was obtained from Fitzgerald Industries International, Inc. Human immunoglobulin G (IgG) and bovine serum albumin (BSA) were obtained from Beijing Biosynthesis Biotechnology Company, Ltd. Albumin chicken egg protein was obtained from SinoAmerican Biotechnology Company, Ltd. The serum samples were provided by Xianyang Central Hospital. Phosphate buffered saline (PBS) consisted of 0.1 M NaH2PO4, 0.1 M Na2HPO4, and 0.1 M KCl (pH 7.4). The other reagents used in this work were of analytical grade and directly used without additional purification. Millipore Milli-Q water (18.2 MΩ cm) was used to prepare all solutions. 3888
dx.doi.org/10.1021/ac4005259 | Anal. Chem. 2013, 85, 3886−3894
Analytical Chemistry
Article
Figure 2. (A) TEM image of Ru1-encapsulated liposome. (B) Fluorescence image of Ru1-encapsulated liposome. Excitation occurred at 490 nm; exposure time was 30 s. (C) ECL intensity-potential profiles of the Ru1-encapsulated liposome in 0.10 M PBS (pH 7.4) containing 50 mM TPA with a scan rate of 50 mV/s (a) before release, (b) after release with 60 °C water bath, (c) in 1% (v/v) Triton X-100, and (d) in ethanol. (D) TEM image of Ru1-encapsulated liposome after release with ethanol.
concentration of 18.5 mg/mL, according to the amount of the lipid, and maintained at 4 °C until use. The signal reagent-labeled peptides (Ru1-peptide) were synthesized according to literature,33,39 with some modifications for comparison with the liposome peptide. Briefly, 1 mg of the peptide (FYSHSFHENWPSK, 0.0006 mmol) was dissolved in 500 μL of 0.1 M PB (pH 7.4). This peptide solution was added to a 10-fold molar excess of Ru1 and stirred overnight. The Ru1-labeled peptides were purified by dialysis at 4 °C for 12 h, using the MD34−2 Da molecular weight cutoff membrane with 0.1 M PB (pH 7.4). The concentration of the Ru1-peptide solution was calculated to be 3.96 × 10−5 M, according to the value of UV−vis absorption of Ru1 at 457 nm. Preparation of Magnetic Capture Peptide (MBsPeptide). The magnetic peptide captures (MBs-peptide) were synthesized according to the protocol provided by Shaanxi Lifegen Company, Ltd., and the MBs and “solutions” used were provided by Shaanxi Lifegen Company, Ltd. First, 0.4 mg of the peptide (FYSHSFHENWPSK, 0.00024 mmol) was dissolved in 200 μL of the “coupling solution”. And then, 200 μL of isothiocyanic acid-coated MBs (5 mg/mL) was added into the above peptide solution, followed by incubation at 37 °C for 20 min. After being rinsed with a “washing solution” three times, the peptide-coated MBs were resuspended in a “blocking solution” and incubated at 37 °C for 2 h to block any remaining active surfaces. Finally, the resulting MBs-peptide were rinsed three times with “washing solution” and resuspended in a final volume of 600 μL of “preserving fluid” and stored at 2−8 °C. ECL Measurement. Ten microliters of standard solutions of TnI or the samples were mixed with 10 μL of MBs-peptide and 10 μL of Ru1-L-peptide, incubated for 60 min to form the Ru1-L-peptide < TnI > MBs-peptide sandwich-type conjugate.
Then, the solid phase of the sandwich-type conjugates were obtained by magnetically separating from the reaction media and rinsed three times with 0.1 M PB (pH 7.4). Then, 30 μL of ethanol were added into the conjugates and maintained for 15 min to lyse the liposomes. Finally, 970 μL of 0.10 M PBS (pH 7.4) containing 50 mM TPA was added to the above mixture. After the magnetic beads were removed by magnetic separation, the ECL measurements were performed by leaving the solution at room temperature and a constant potential of +1.15 V. The concentration of TnI was quantified by an increased ECL intensity (ΔI = Is − I0), where I0 and Is are the ECL peak intensity in the absence and presence of TnI, respectively.
■
RESULTS AND DISCUSSION
Characterization of Signal Reagent-Encapsulated Liposome and Optimization of the Releasing Agent. Bard’s group employed Ru(bpy)32+-encapsulated liposome in an ECL sandwich-type immunoassay of human C-reactive protein.20 In our work, Ru1 was employed as the ECL signal reagent to compare the analytical performance of the Ru1-labeled peptide with that of the Ru1-encapsulated liposome-labeled peptide. As expected, for the application of the developed method in our work, Ru(bpy)32+ is better than Ru1 because the former is less expensive and higher in the luminescence yield. The signal reagent-encapsulated liposomes synthesized in this work were employed as an ECL label and characterized by TEM, fluorescence, and ECL techniques, as showed in Figure 2. The TEM images of Ru1-encapsulated liposome demonstrate that the size of the liposome prepared in our work is on the order of 100 nm and is dependent on the sonication time in the preparation process (as shown in Figure 2A and Figure S-2 of the Supporting Information). After sonication for 5, 15, and 20 3889
dx.doi.org/10.1021/ac4005259 | Anal. Chem. 2013, 85, 3886−3894
Analytical Chemistry
Article
Figure 3. (A) TEM image of MBs, (B) TEM image of Ru1-L-peptide < TnI > MBs-peptide conjugates, (C) TEM image of Ru1-L-peptide < TnI > MBs-peptide conjugates after release with ethanol. (D) ECL intensity-potential profiles in 0.10 M PBS (pH 7.4) containing 50 mM TPA with a scan rate of 50 mV/s (a) before and (b) after reaction with 2.0 × 10−10 g/mL cardiac troponin I and (c) 2.0 × 10−9 g/mL cardiac troponin I, respectively.
In this work, three commonly used release agents, including ethanol, 1% (v/v) Triton X-100, and a 60 °C water bath were checked. As shown in Figure 2C, these three commonly used release agents can increase the ECL peak intensity in sequence from low to high, in a 60 °C water bath (ECL intensity is 700), in 1% (v/v) Triton X-100 (1036), and in ethanol (1264), indicating that ethanol has the highest releasing efficiency. Additionally, it was found that 3% (v/v) ethanol had no effect on the ECL intensity of Ru1 (see Figure S-4 of the Supporting Information), and the TEM image of Ru1-encapsulated liposome after release by ethanol was evidence that the liposomes were observed to be lysed (Figure 2D). Therefore, ethanol was selected as the releasing agent in the following experiment. Feasibility of the Liposome Peptide-Based Method for the Determination of TnI. In this work, a specific peptide with a sequence of FYSHSFHENWPSK as a recognition molecular element was covalently coupled onto the surface of the Ru1-encapsulated liposomes via glutaraldehyde to construct the ECL probes (II) and was covalently coupled onto the surface of the isothiocyanic acid-coated MBs, to construct the magnetic capture peptides (I), in order to provide a platform of polyvalent binding motif, as shown in Figure 1. In the presence of TnI, sandwich-type conjugates (Ru1-L-peptide < TnI > MBs-peptide conjugates) were formed in incubation of the magnetic capture peptides and the liposome peptides. After a magnetic separation, the sandwich-type conjugates were treated via ethanol and thus a great number of the ECL reagents were released and measured by the ECL method at a bare glassy carbon electrode with a potential pulse of +1.15 V versus Ag/ AgCl in the presence of tri-n-propylamine. The concentration of TnI was qualified by the increased ECL intensity. Figure 3A shows TEM images of magnetic beads, demonstrating that the magnetic beads were well-dispersed
min, multilayered vesicles with average diameters of 51.3, 170.8, and 339.9 nm were obtained, respectively. The effective loading of Ru1 in the liposome can be visualized via verified fluorescent microscopy, as showed in Figure 2B. The encapsulation efficiency of Ru1 in the Ru1encapsulated liposomes and the loading capacity of Ru1 molecules per liposome are important parameters for the liposome function. The former was estimated using the ratio of the amount of encapsulated Ru1 in the liposomes to the amount of Ru1 added in the preparation of the liposomes, while the latter was calculated according to previous literature reports.20 The amount of encapsulated Ru1 in the prepared liposomes was calculated using ECL to measure the concentration of Ru1, which was released from the prepared liposomes with ethanol, and a calibration curve of a standard Ru1 solution in the presence of TPA (Figure S-3A of the Supporting Information). A good correlation between the ECL intensity and the number of Ru1 released from Ru1encapsulated liposomes using ethanol was also obtained (Figure S-3B of the Supporting Information). The encapsulation efficiency was calculated to be 29.6%, 59.3%, and 84.7% for 51.3, 170.8, and 339.9 nm liposome samples prepared, respectively. The loading capacity was calculated to be 6.4 × 106, 1.9 × 107, and 2.23 × 107 Ru1 molecules/liposome for the respective 51.3, 170.8, and 339.9 nm liposome samples prepared. These results are consistent with the data obtained on the basis of the “bulk liposome”25 and indicate that the Ru1encapsulated liposomes prepared in this work have a high effective loading. Furthermore, the releasing agent was also optimized since it decided the release efficiency of Ru1 from the liposome, which mediated the ECL performance of the liposomes. 1% (v/v) Triton X-100,20 ethanol,40 a 60 °C water bath,41 and melittin42 have all been employed as releasing agents in previous reports. 3890
dx.doi.org/10.1021/ac4005259 | Anal. Chem. 2013, 85, 3886−3894
Analytical Chemistry
Article
and that their diameter sizes ranged from 8−10 nm. Figure 3B represents the morphology of the Ru1-L-peptide < TnI > MBspeptide conjugate, which is distinct from that of the Ru1-L. The surface of the liposome becomes rough and one Ru1-L-peptide was surrounded by MBs-peptide, indicating the formation of sandwich conjugates between the liposome peptides and the magnetic capture peptides in the presence of TnI and the polyvalent binding of one Ru1-L-peptide with TnI and MBspeptide. Figure 3C demonstrates that the Ru1-L-peptide < TnI > MBs-peptide conjugate is released by ethanol. An increase in diameter can be obtained, and irregular shapes can clearly be observed. Therefore, a liposome peptide-based method was proposed to detect TnI. The ECL probe was first characterized by the ECL method at a glassy carbon electrode using a triangular potential scan (Figure S-5 of the Supporting Information). The result showed that the ECL behavior of the Ru1-L-peptide was similar to that of the released Ru1-L-peptide after ethanol treatment. Additionally, the ECL intensity of the Ru1-L-peptide after being released with ethanol was much higher than that of the Ru1-Lpeptide. This confirms that large amounts of Ru1 were released from the liposome. To illustrate the feasibility of this sandwich method, the ECL intensity of the Ru1-L-peptide in the absence and presence of protein was examined as described in ECL Measurement, employing the liposome of 170.8 nm. Figure 3D shows that the ECL peak intensity increases from 1672 to 2676 as the TnI concentration is elevated from 2.0 × 10−10 to 2.0 × 10−9 g/mL. These data indicate that the ECL method is feasible for the detection of TnI. The effect of the size of the liposome peptides prepared (51.3, 170.8, and 339.9 nm) on the ECL intensity for the determination of TnI was also examined. It was found that the slope of the linear relationship between the increased ECL intensity and the logarithm of the TnI concentration increased with an increase in the size of the liposome peptides from 1450 (51.3 nm) to 1715 (170.8 nm) and to 1911(339.9 nm) (see Figure S-6 of the Supporting Information). This result may be attributed to the fact that the large-sized liposome peptides have large loading capacities of Ru1 molecules per liposome and larger number of binding sites.43 However, it was found that the reproducibility of the method in this work using large liposome peptides (339.9 nm) was lower than those using small liposome peptides. This may be induced by the appearing broken liposomes. Therefore, liposome peptides with a size of 170.8 nm were employed as the ECL probe in the following experiments. Optimization of Detection Conditions. Applied potential is an important parameter because it modulates the sensitivity of an ECL bioassay. Figure 4A indicates that the ECL intensity increased when the applied potential was raised from +0.9 V to +1.15 V and reached a maximum at +1.15 V. This potential is generally observed at a glassy carbon electrode, associated with the direct oxidation of Ru(bpy)32+ derivatives in homogeneous solution,44 and is negatively shifted about 0.20 V compared with that of Ru1 modified at the electrode (+1.35 V) in our previous report.45 A low working potential is of great advantage to the electrode protection and the reproducibility of the developed method. Therefore, a constant potential of +1.15 V was chosen in the following experiments. Figure 4B displays the effects of the binding time between MBs-peptide and TnI. From Figure 4B, the ECL intensity sharply increases with increasing binding time from 20 to 40
Figure 4. (A) Dependence of the ECL intensity on applied potential, (B) binding time between the MBs-peptide and cardiac troponin I (2.0 × 10−9 g/mL), and (C) binding time between the Ru1-L-peptide and cardiac troponin I (2.0 × 10−9 g/mL). Measurement conditions are the same as those in Figure 3D, except for an applied potential of +1.15 V in (B) and (C).
min and reaches a maximum at about 50 min. This increase suggests that the binding reaction is complete within 50 min. Figure 4C demonstrates the effect of the binding time between TnI and the Ru1-L-peptide on the ECL intensity. The results showed that the ECL intensity sharply increased with an increasing binding time from 30 to 60 min, and then, nearly kept a stable value. This indicates that the binding reaction mostly completes within 60 min. Additionally, the binding time needed for the complete binding between the TnI and Ru1-Lpeptide is longer than that between the MBs-peptide and TnI, attributed to the fact that the MBs-peptide is much smaller and more rigid than the liposome peptide. In consideration of both these two binding processes in one step in our ECL measurement protocol, we chose a 60 min binding time for the incubation processes in the following experiments. Linear Range and Detection Limit. Under the optimized conditions, the quantitative behavior of the proposed ECL method for the determination of TnI was assessed according to 3891
dx.doi.org/10.1021/ac4005259 | Anal. Chem. 2013, 85, 3886−3894
Analytical Chemistry
Article
employing the liposomes as amplification carriers is feasible. Additionally, compared with homogeneous ECL method in our previous report,33 the MBs-peptide used in this work having the enrichment of the target protein is evident. An evaluation of the selectivity of the ECL method developed was performed, according to the protocol described in the ECL measurement section by examining 2.0 × 10−9 g/ mL (8.3 × 10−11 M) TnI and 1.5 × 10−7 M other proteins including BSA, albumin chicken egg, IgG, and a prostatespecific antigen, respectively. Compared with the ECL intensity obtained in 0.1 M PBS containing a 50 mM TPA (pH7.4) blank, a great increase in the ECL intensity was observed for TnI (83%), while very slight increases in the ECL intensity were found for the other tested proteins such as IgG (1.5%), BSA (5.4%), PSA (4.1%), and albumin chicken egg protein (2.4%), respectively (see Figure S-7 of the Supporting Information). A good selectivity is evident and could be ascribed to the sandwich assay, employing both capture probe and signal probe in this work. Detection of TnI in Clinical Serum Samples. The potential clinical application of the developed ECL method for the determination of TnI was evaluated by analyzing clinical serum samples, including four positive sera and six negative sera, which were provided by Xianyang Central Hospital, according to the rules of the local ethical committee. Prior to measurement, all of the samples were gently shaken at room temperature (Note: all sample handling and processing were performed carefully, and all tools in contact with patient specimens and immunoreagents were disinfected after use). The assay results for positive sera are listed in Table 1. The
the protocol described in Experimental Section. Figure 5A shows the ECL intensity versus time profiles with different
Figure 5. ECL intensity−time profiles and the calibration curve in 0.10 M PBS (pH 7.4) containing 50 mM TPA. (A) ECL peptide-based method employing the Ru1-L-peptide as the ECL probe and the concentration of cardiac troponin I: (a) 1.0 × 10−11 g/mL, (b) 2.0 × 10−11 g/mL, (c) 1.0 × 10−10 g/mL, (d) 3.0 × 10−10 g/mL, (e) 8.0 × 10−10 g/mL, (f) 2.0 × 10−9 g/mL, and (g) 5.0 × 10−9 g/mL. Inset, calibration curve of TnI. (B) ECL peptide-based method employing Ru1-peptide as an ECL probe and the concentration of cardiac troponin I: (a) 5.0 × 10−11 g/mL, (b) 1.0 × 10−10 g/mL, (c) 5.0 × 10−10 g/mL, (d) 1.0 × 10−9 g/mL, and (e) 5.0 × 10−9 g/mL. Inset, calibration curve of cardiac troponin I. Experimental conditions: applied potential, +1.15 V; binding time, 1 h.
Table 1. Analytical Results of TnI in Clinical Positive Serum Samplesa sample number
CL method (ng/mL)b
1 2 3 4
11.49 3.45 0.38 0.44
this method (n = 3, P = 0.9, ng/mL)c
relative error (%)
± ± ± ±
−7.2 2.9 2.6 −2.3
10.66 3.55 0.39 0.43
1.58 0.80 0.02 0.02
a
A level of >0.25 ng/mL TnI is considered a troponin positive serum sample.46 bThe CL results of TnI in patients’ serum samples from clinical reports provided by Xianyang Central Hospital. cConfidence interval.
concentrations of TnI. The ECL intensity increases with increasing TnI concentration. The increased ECL intensity was linearly proportional to the logarithm of the TnI concentration in the range from 1.0 × 10−11 to 5.0 × 10−9 g/mL. The linear regression equation was ΔI = 1715 log C + 19867 (unit of C is g/mL), and the correlation coefficient was 0.9883. The limit of detection was 4.5 pg/mL TnI. This is 26-fold lower than that obtained by the homogeneous ECL method in our previous report using a Ru1-labeled peptide.33 The relative standard deviation for 2.0 × 10−9 g/mL TnI was 4.3% (n = 5). To illustrate the good sensitivity of our proposed method, an ECL measurement for the determination of TnI was performed according to the protocol described in Experimental Section, employing Ru1-peptide instead of Ru1-L-peptide. As showed in Figure 5B, the increased ECL intensity was linearly proportional to the logarithm of the TnI concentration in the range from 5.0 × 10−11 g/mL to 5.0 × 10−9 g/mL. The regression equation was ΔI = 823.7 lgC +8962 (unit of C is g/mL), and the correlation coefficient was 0.9742. The limit of detection was 24 pg/mL, which was 5.5-fold higher than that of the Ru1L-peptide as the ECL probe and 5-fold lower than that obtained by the homogeneous ECL method in our previous report using the Ru1-labeled peptide.33 This indicates that
obtained results showed an acceptable agreement with the data provided by Xianyang Central Hospital, using a standard chemiluminescence (CL) method with an Abbott Immunoanalyzer (Abbott Axsym, i1000). Since employing the Student’s t test at P = 0.9, no statistical significance between the result obtained using the ECL method developed in this work and that using the CL method was observed, therefore signifying that the proposed ECL method is reliable for clinical testing. The analytic characteristics of the proposed ECL method were also obtained in negative TnI clinical serum samples. The increased ECL intensity was linearly proportional to the logarithm of the standard TnI concentration in the range from 1.0 × 10−11 to 1.0 × 10−9 g/mL. The regression equation was ΔI = 1686 log C + 18836 (unit of C is g/mL) with a correlation coefficient of 0.9428. The limit of detection was 3.7 pg/mL (Figure S-8 of the Supporting Information). This regression equation is nearly the same as the regression equation obtained with standard solutions of TnI presented above, due to the 3892
dx.doi.org/10.1021/ac4005259 | Anal. Chem. 2013, 85, 3886−3894
Analytical Chemistry
Article
Table 2. Analytical Results of TnI in Clinical Negative Serum Samplesa sample number
CL method (ng/mL)b
5 6 7 8 9 10
0 0 0 0 0 0
this method (n = 3, P = 0.9, ng/mL)c
TnI added (ng/mL)
± ± ± ± ± ±
0.050 0.050 0.010 0.10 0.050 0.010
0.0046 0.0049 0.0053 0.0053 0.0047 0.0052
0.0008 0.0005 0.0005 0.0015 0.0009 0.0019
a
A level of
