1. Introduction

The development of reliable, cost-effective, powerful detection and monitoring strategies for tumor biomarkers is particularly important [1, 2], due to the prevalence of the diseases, high rates of recurrence, and potential lethality. In comparison with other immunologicalmethods based on fluorescence, chemiluminescence, surface-plasmon resonance, or quartz crystal microbalance, electrochemical immunoassay has attracted considerable interest because of the advantages of simple instrumentation, easy signal quantification, low cost of the entire assay, and potential ability for real-time and on-site detection [3, 4]. Bhaskara, et al. summarized the various forms of voltammetry (linear sweep, differential pulse, square-wave, stripping) and amperometry which are the most widely used electrochemical methods for detection of protein biomarkers [3]. Among those methods, anodic stripping voltammetry (ASV) proved to be a very sensitive method for trace determination of metal ions because the method has a pre-concentration procedure [5]. Additionally, electrochemical detection strategies employing nanostructured surfaces, nanoparticle labels, andmagnetic beads have proved to offer new opportunities for highly sensitive protein detection. For instance, Niina et al. indicated that carbon nanotubes (CNTs), graphene, nanowires, magnetic nanoparticles, metal nanoparticles, and quantum dots (QDs) were incorporated into immunosensors and enzymatic biosensors [4]. To our best knowledge, those nanomaterials were used extensively in visual immunoassay [1]. But there has been no work reported by using the GNR reporter probe for electrochemical detection combined with visual immunoassay.

Cotton thread could serve as a support for making microfluidic circuits and rapid diagnostic tests due to low-cost and broad availability, easy handling after use, wicking properties, just like liquid migration in microfluidic channels caused by capillary action [6-10]. Our group, as well as David et al. [11-13] has developed a natural cotton thread immunoassay device combined with gold nanoparticle (GNP) reporter probe. For example, Gina et al. introduced an immunochromatographic assay on thread in a cartridge format to test C-reactive protein with a detection limit of 377 pmol/L [11]. Xun et al. [12] reported the use of a dry-reagent cotton thread-based point-of-care diagnosis device for detection of squamous cell carcinoma antigen (SCCA) and hereditary tyrosinemia type related DNA sequence. Herein, we developed a novel trimer gold nanoparticle reporter probe to detect lung cancer and the cotton thread immunoassay device was capable of meaning 10 ng/mL human ferritin [13]. Compared with GNP, GNR was popular in the design and fabrication of biosensors due to its variable length to diameter ratio, large surface area and excellent optical properties [14]. In this work, we chose the GNR reporter probe to take place of the traditional GNP, since results indicated that the assay performance was improved once the GNR was employed as reporter probe instead of GNP.

Here we demonstrated a natural cotton thread immunoassay device for a rapid, sensitive and quantitative detection of human ferritin with the GNR reporter probe. The immunoassay format of the device was more like a lateral flow strip biosensor with the sample solution and running buffer added dropwise directly on the sample pad to rehydrate the conjugates. It is well known that cetyltrimethylammonium bromide (CTAB), serving as the essential inducer and stabilizer in the fabrication process of the GNR, presents a challenge of surface modification for future functionalization with different biochemical groups [15, 16]. Therefore, we used mercaptopropionic acid (MPA) to displace CTAB to acquire anchor points for the future immobilization of biological molecules. The functionalized GNR reporter probe was modified with detection antibody so as to form the gold nanorod and antibody conjugates (GNR-dAb). For the human ferritin test, GNR-dAb was mixed with human ferritin antigen and then dropped onto the sample pad; the solution would then migrate through the cotton thread. After 20 min, a purple band would appear on the test zone due to the presence of the specific analyte, which could be used in the quantitative assay. Quantitative detection of the target analyte can be realized by testing the dissolved gold ions (III) using ASV after the test zone with purple band was cut from the thread and treated in HBr-Br₂ solution. The detail optimization and attractive performance are reported in the following sections.

2. Experimental 2.1. Reagents and materials

Natural cotton thread (100% mercerized) was purchased from a cotton thread store (Xi'an). Human ferritin antigen and detection antibody (dAb), capture antibody (cAb), carcinoembryonic antigen (CEA) and squamous cell carcinoma antigen (SCCA) were supplied by Shanghai Linc-Bio Science Co., Ltd. Human serum and human immunoglobulin G (H-IgG) were purchased from Dingguo Biological Products (Beijing, China). The 3-mercapto propionic acid (MPA) was purchased from Aladdin Chemistry Co., Ltd. (Shanghai, China). CTAB was purchased from Tianjin Guangfu Fine Chemical Research Institute (Tianjin, China). The 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC), and Nhydroxysuccinimide (NHS) were purchased from Life Science Products & Services (Shanghai, China) while HAuCl₄, HBr, Br₂, phenoxyacetic acid, bovin serum albumin (BSA), and Tween-20 were purchased from Sigma-Aldrich. All other chemicals were at least of analytical reagent grade. All buffer solutions were prepared with ultrapure water.

Phosphate buffered saline (PBS, 0.05 mol/L) was prepared by mixing the stock solutions of NaH₂PO₄ and Na₂HPO₄, and 0.1 mol/L NaCl was added as the supporting electrolyte. The washing buffer was PBS (0.05 mol/L, pH 7.4) containing 0.05% (w/v) Tween 20 (PBST). Dispersion buffer and running buffer was PBS (0.05 mol/L, pH 7.4) containing 1% (w/v) BSA and 0.25% Tween-20 (PBSBT).

2.2. Instrumentation

The ASV measurements were performed with an Electrochemical Workstation CHI660D (CH Instruments, Austin, TX) coupled with a conventional, three-electrode cell. The working electrode was a carbon paste working electrode (graphite:paraffin=1:1). A platinum wire and a saturated calomel electrode were used as counter and reference electrodes, respectively.

2.3. Assay procedure

GNR (with lengths of~35 nm and diameters of~12 nm) was synthesized in aqueous media using CTAB as the shape-inducing surfactant according to a previous published method [17]. And colloidal gold nanoparticle of an average diameter of 15±3 nm was prepared according to the reported methods [18]. As shown in Fig. 1, the transmission electron microscopy (TEM) of purple gold nanorod and gold nanoparticle exhibited average dimensions and nearly monodisperse.

The preparation of the cotton thread immunoassay device for human ferritin test was simple. The length of thread was pasted between two parallel-placed double-faced adhesive tapes. An absorbent pad (qualitative filter paper, 1.5 cm × 10 cm) was applied at the downstream end to serve as a capillary pump to draw. A glass fiber (4.5 mm × 9.0 mm) was attached to the other end of the thread as sample pad to load large amount of sample solution. The test zone was coated with 0.75 L (applied as three aliquots of 0.25 L, with 10 min drying at 37 ℃ after each step) of 2.4 μg/mL capture antibody, cAb, and then dried at 37 ℃ for one h in a drying oven. The capture antibody bound tightly to the surface of the thread through a combination of electrostatic and hydrophobic interactions [11].

Briefly, the GNR-dAb reporter probe was synthesized as follows. MPA solution (100 μL, 375 μmol/L) was added to 1 mL GNR solution and incubated for 2 h to form a self-assembled membrane of MPA-GNR. Residual MPA was removed by high-speed centrifugation; the complexes were washed three times with water and resuspended in one mL sterilization water. Then, 3.3 μL of 40 μg/mL EDC and 2.5 μL of 30 μg/mL NHS were mixed with the above MPAGNR solution for 2 h. Next, 0.05 mol/L NaOH was added to adjust the pH of the above solution to 8.2. Eight microliter of 2.5 μg/mL monoclonal dAb was added to the solution overnight at 4 ℃. After centrifugation, the obtained bio-conjugates GNR-dAb were washed with washing buffer and re-suspended in 250 μL of dispersion buffer.

To carry out the immunoreaction and electrochemical measurement, 25 μL of GNR-dAb solution was first incubated with ferritin standard solution, or serum samples; of different concentration for 30 min at r.t. Then each sample solution was dropped onto the sample pad, followed by adding running buffer to the sample pad to allow the migration of the sample solution. After 20 min, a purple band would appear at the test zone. Finally, the test zone was cut from the thread with scissors and transferred to an electrochemical cell containing 750 μL HBr-Br₂ (6.0 mol/L HBr-0.6 mol/L Br₂) solutions to dissolve the captured GNR. After 5 min, 150 mL of 4 × 10^-2 mol/L phenoxyacetic acid solution was used to react with excess bromine. The released gold ions (Au₃₊) were then quantified by ASV under the following instrumental conditions: 420 s deposition at -0.8 V and positive potential scan at 100 mV/s. A new carbon paste working electrode was used for each measurement.

3. Results and discussion 3.1. Electrochemical immunoassay based on cotton thread immunoassay device combined with GNR reporter probe

In this study, an electrochemical immunoassay strategy was developed by using a natural cotton thread immunoassay device combined with a GNR reporter probe. The structure of the device and the entire assay procedures are shown in Fig. 2. Firstly, the GNR was functionalized by MPA to obtain GNR-MPA. The functionalized GNR was activated by EDC and NHS for the attachment of antibodies to yield GNR-dAb conjugates and form GNR-dAb-Ferritin complexes with Ferritin antigen. Secondly, the solution of GNR-dAb-Ferritin complexes was applied to the sample pad, and immediately migrated along the cotton thread to cAb, which was pre-immobilized as the test zone of the cotton thread, where the ferritin antigen is captured to form sandwich GNR-dAb-FerritincAb complexes. Because of the accumulation of GNR complex, a purple band would appear on the test zone. For performing the quantitative detection, the test zone was cut from the thread with scissors and transferred to an electrochemical cell containing HBr-Br₂ solutions to dissolve the captured GNR and equipping with a three-electrode system to perform ASV detection. The ASV has been proved to be a very sensitive method for trace determination of metal ions due to the pre-concentration procedure of the method. Therefore, gold ions (III) could be tested at a concentration as low as five nmol/L by ASV at a carbon paste electrode [19]. A GNR reporter probe contains thousands of gold ions, so the released Au₃₊ could be detected with the typical ASV method. The electrochemical signal was indirectly proportional to the amount of human ferritin in the standards, or samples. The results of ASV that determined the device based on GNR and on GNP at a concentration of human ferritin of 1000 ng/mL are illustrated in Fig. 3. Anodic peak could not be observed in the absence of ferritin (Curve a). And an obvious anodic peak would emerge at about 0.6 V once the GNR reporter probe was used in the presence of ferritin (Curve c) which could be explained at the advantages of GNR that due to their variable length to diameter ratio and large surface area, while a weak anodic peak would emerged once GNP was employed as reporter probe in the presence of ferritin (Curve b). Additionally, the performance of other methods reported in the literature has been compared with this method for ferritin analysis with the characteristics such as the linear range and detection limit summarized in Table 1. Based on these data, the method exhibited a wide linear range and low detection limit for ferritin assay and the results show that employment of the GNR reporter probe greatly improved the sensitivity of the cotton thread immunoassay device. The method made it possible to detect a low concentration of human ferritin.

3.2. Analytical performances of the method

Under optimal experimental conditions, we examined the performance of the cotton thread immunoassay device with different concentrations of ferritin (Fig. 4). The typical ASV responses increase with the increase of the concentrations of the human ferritin in the range of 5-5000 ng/mL. The specific response, coupled with high reproducibility and a series of measurements of 500 ng/mL ferritin with well-prepared cotton thread immunoassay device and the GNR reporter probe yielded a reproducible signal with a relative standard deviation (RSD) of 8.6% (data not shown).

3.3. Detection of ferritin in human serum

For the detection of human ferritin in complex biological matrixes, the GNR-dAb conjugates were dispersed in the solution containing a mixture of human serum and running buffer PBSBT (v:v, 1:1). Fig. 5 presents the response of the device in the presence of ferritin in the solution mixture. The resulting calibration indicated that the logarithm of the peak currents were proportional to the logarithm of ferritin concentration in the range of 5-5000 ng/mL with a detection limit of 1.58 ng/mL (based on S/N=3) and a total assay time of 30 min. So the proposed device could be used to develop diagnostic medical systems for clinical applications.

3.4. Effect of other proteins

To determine the specificity of the strategy, the sample solutions were prepared by adding 1 μg/mL of human IgG, CEA, SCCA and ferritin to the PBSBT buffer. In Fig. 6, a visible oxidation peak current value was recorded when 1 μg/mL ferritin was tested, and negligible signals were obtained in the presence of 1 μg/mL of human IgG, CEA or SCCA. Because the concentration of BSA in the running buffer was nearly in a 10000-fold excess with respect to 1 μg/mL ferritin, the presence of a high concentration of BSA undoubtedly did not affect the device performance.

4. Conclusion

We have successfully developed a rapid, sensitive, electrochemical cotton thread immunoassay device based on the GNR reporter probe. The approach takes advantage of the rapid detection, low-cost of the cotton thread immunoassay device and high sensitive of electrochemical detection. The results indicate that the employment of the GNR reporter probe, as opposed to the GNP reporter probe, greatly improved the assay performance. Firstly, GNR has special optical characters due to their variable length to diameter ratio and large surface area. Secondly, instead of physical adsorption, the detection antibody was modified on GNR through chemical cross-linking reactions after GNR was functionalized with MPA, which made the GNR-dAb conjugate more stable. Consequently, the GNR reporter probe offers a strong motivation for the improvement of detection sensitivity in future bioassay applications, such as pathogens detection, cancer diagnosis and environmental quality evaluation. Different sizes of the GNR reporter probe with various colors could be synthesized according to the reported seed-mediated growth method [17]. Therefore, our future work will focus on multiplexing detection of protein by employing different colored GNR as reporter probe.

Acknowledgments

This work was financially supported by the National Natural Science Foundation of China (No. 21205094), NFFTBS (Nos. J1103311, J1210057) and the New Faculty Startup Funds of Northwest University in Shaanxi Province (No. PR12011).