Electrochemical immunosensor with surface-confined probe for sensitive and reagentless detection of breast cancer biomarker

Sensitive and reliable detection of tumour markers is of great significance for early diagnosis and monitoring recurrence of cancers. Herein, a simple electrochemical immunosensor is developed with an integrated electrochemical probe on the sensing surface, which is able to sensitively and reagentlessly detect the breast cancer biomarker, human epidermal growth factor receptor 2 (ErbB2). Ferrocene (Fc) is chosen as the signal indicator and covalently grafted on cationic polyelectrolyte poly(ethylene imine) (Fc-PEI). The redox Fc-PEI could alternately assemble with carboxyl functionalized single-walled carbon nanotubes (SWNTs) on an indium tin oxide electrode through layer-by-layerelectrostatic assembly. After Anti-ErbB2 antibody is covalently immobilized onto the outermost SWNTs layer followed by blocking the electrode with bovine serum albumin, a sensing interface with recognitive probe and electrochemical probe is obtained. In the presence of ErbB2, the formed antigen–antibody complex makes a barrier to inhibit electro-transfer of inner Fc, leading to a decreased electrochemical response. Owing to the SWNTs-facilitated charge transfer and abundant surface-bound probes, the developed sensor demonstrates outstanding performance for reagentless detection of ErbB2 in terms of wide detection range (1.0–200.0 ng mL−1) and low detection limit (0.22 ng mL−1). The developed immunosensor also exhibits good selectivity, reproducibility and stability. Real analysis of ErbB2 in human serum samples is also demonstrated.


Introduction
Early diagnosis of cancers is crucial for providing the appropriate treatment processes and increasing the survival rate. Sensitive and reliable detection of low-levels of tumour markers is of great signicance in early diagnosis and monitoring recurrence of cancers. 1,2 Human epidermal growth factor receptor 2 (ErbB2), as an important tumour biomarker, [3][4][5] is over-expressed in around 20-25% of invasive breast cancers and plays a major role in promoting breast cancer cell proliferation and malignant growth. 6,7 Currently, several methods have been used for diagnostic tests of ErbB2, including enzyme-linked immuno-sorbent assay and in situ uorescent hybridization. 3,8 However, most of these techniques provide only semi-qualitative results separating patients into HER2 positive and HER2 negative. Moreover, they are usually complicated and time-consuming, need exhaustive sample pre-treatment, and require specially trained personnel to perform the complex procedures. In comparison to these techniques, electrochemical biosensors have received particular attention because of high selectivity and sensitivity, simple and rapid detection, and no need of expensive and complex instruments. [9][10][11][12][13] In addition, easy integration and miniaturization of electrochemical sensors exhibit potential in point-of-care test devices. 14 Although several electrochemical immunosensors have been reported for the detection of ErbB2, [15][16][17][18] most detection requires external probe to produce electrochemical signals (e.g., enzymatic substrates or Ag + probes for stripping voltammetry or electrochemical redox probes). However, the use of solutionphase probes may compromise the detection efficiency due to diffusion limit and may bring side-effects (e.g. contamination to the target system) especially for long-time or repetitious detections. [19][20][21][22][23] In addition, the complicated operation is not benecial for the construction of integrated and miniaturized sensors. To tackle these problems, development of reagentless electrochemical immunosensors with surface-conned probe is of great signicance.
In this work, we demonstrate an electrochemical immunosensor with surface-conned electrochemical probe for sensitive and reagentless detection of breast cancer biomarker ErbB2. Multilayers with ferrocene (Fc) as electrochemical signal indicator and single-walled carbon nanotubes (SWNTs) as conductive enhancement materials were fabricated on an indium tin oxide electrode via layer by layer self-assembly. Immunosensing interface was constructed by covalent immobilization of anti-ErbB2 antibodies on the SWNTs at the outmost layer of the self-assembled lm. Owing to the electronic wire action of nanotubes and abundant surface-bound probes, the developed sensor demonstrates outstanding performance for reagentless detection of ErbB2 in terms of wide detection range, low detection limit, good selectivity, reproducibility and stability.

Materials and reagents
ErbB2 and rabbit anti-ErbB2 monoclonal antibody (ErbB2-Ab) were purchased from Beijing ACROBiosystems Co. Ltd. (China). Poly(ethylene imine) (PEI), ferrocenecarboxaldehyde (98%), N-(3-dimethylaminopropy)-N 0 -ethylcarbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), bovine serum albumin (BSA), and indium-tin-oxide (ITO) coated glass were purchased from Sigma-Aldrich. Carboxyl functionalized singlewall carbon nanotubes (SWNTs) were obtained from Shenzhen Nanotech Port Co. Ltd. (China). Fresh human serum samples were provided by the Affiliated Tumor Hospital of Guangxi Medical University. The Fc-PEI complex was synthesized according to the literature. 27 All other chemicals and reagents were of analytical grade and used without further purication. Ultrapure water (18.2 MU cm) from a Milli-Q Plus system (Millipore) was used to prepare all aqueous solutions throughout the work.

Instrumentation
Atomic force microscopy (AFM) images were obtained on a SPI3800N microscope (Seiko Instruments, Inc.). Cyclic voltammetry (CV) and differential pulse voltammetry (DPV) measurements were performed on a CHI 660D electrochemical analyzer (Shanghai CH Instrument Company, China). A conventional three-electrode system was adopted with a modi-ed ITO as the working electrode, Ag/AgCl electrode (saturated with KCl) as the reference electrode, and a Pt sheet (1 cm Â 1 cm) as the counter electrode. The ITO electrode with a certain area (3 mm in diameter) was prepared by photolithography as previously reported. 19 Phosphate buffered solution (PBS, 0.1 M, pH 7.0) was used as the electrolyte for all electrochemical experiments. All measurements were carried out at room temperature.

Fabrication of sensing interface
The preparation of electrochemical immunosensor included four steps. Firstly, a clean ITO electrode with negatively charged surface was produced. ITO electrode was ultrasonicated in ethanol/water solution (v/v ¼ 1 : 1, saturated with NaOH) for 5 min followed by successive rinsing by acetone, ethanol and ultrapure water under ultrasonication. Secondly, the multilayers with Fc as electrochemical signal indicator and SWNTs as conductive enhancement materials were fabricated. The treated ITO electrode was alternately immersed into Fc-PEI (2 mg mL À1 ) and SWNTs (0.1 mg mL À1 ) solutions for 30 min, respectively. The obtained electrode was rinsed thoroughly with ultrapure water and dried in a nitrogen stream aer each assembly. This process was repeated four times until the (Fc-PEI/SWNTs) 4 multilayer was obtained. Thirdly, covalent immobilization of antibody was performed to construct the immuno-recognition interface. The (Fc-PEI/SWNTs) 4 multilayer modied ITO electrode was immersed into an aqueous solution containing EDC (50 mM) and NHS (50 mM) for 2 h at room temperature to activate the carboxyl groups of SWNTs, followed by rinsing with ultrapure water. Then, the activated electrode was dipped into a solution of ErbB2-Ab (100 mg mL À1 ) in PBS (0.1 M, pH 7.0) at room temperature for 2 h. As a result, ErbB2-Ab was graed onto the outmost SWNTs layer through the formation of covalent bonds between the -COOH groups of the SWNTs and the -NH 2 groups of ErbB2-Ab. Aer rinsing with PBS (0.1 M, pH 7.0) to remove excessive and physically adsorbed ErbB2-Ab, the electrode was immersed into a BSA solution (1%) and kept for 1 h to block any possible nonspecic bonding sites. Finally, the resultant BSA blocked ITO/(Fc-PEI/SWNTs) 4 /ErbB2-Ab electrode was stored at 4 C for detection of ErbB2.

Electrochemical detection of ErbB2
The prepared immunosensor was incubated with ErbB2 solution for 30 min at room temperature. Aer rinsing with PBS, the sensing electrode was transferred into an electrochemical cell for electrochemical measurements. CV measurements were performed by scanning from 0 V to 0.7 V at a scan rate of 50 mV s À1 . DPV were carried out from 0.7 to 0 V under the following conditions: modulation amplitude, 25 mV; modulation time, 50 ms; potential step, 5 mV; interval time, 0.5 s.

Preparation and characterization of the self-assembled multilayers
The preparation and detection mechanism of the immunosensor are shown in Scheme 1. As seen, through layer-by-layer (LBL) electrostatic assembly between Fc-PEI (positively charged) and carboxyl functionalized SWNTs (negatively charged), abundant Fc signal molecules are conned in Fc-PEI/ SWNTs multilayers supported by ITO electrode. The immune recognition interface was constructed by the carbodiimide Scheme 1 Illustration for fabrication of the immunosensor with surface-confined electrochemical probe by LBL assembly. catalysed antibody coupling method, which has been proven to be an effective method for the covalent immobilization of active antibodies on various functional nanomaterials containing carboxyl groups. [24][25][26] In addition to the four terminal amino groups in the "Y" shaped antibody molecule, the peptide chains also possess abundant free amino sites derived from special amino acid residues (e.g. lysine and ornithine), which can also act as binding sites for the linkage with carboxylated SWNT. Briey, the -COOH of SWNTs activated by EDC and NHS can couple with the free -NH 2 of anti-ErbB2 antibody (ErbB2-Ab) molecules via the formation of amide bond, and thus xing the anti-ErbB2 antibodies onto the outmost SWNTs layer. The sensing interface are nally obtained aer blocking nonspecic adsorption sites with BSA. When the target antigens, ErbB2 proteins, are captured by the anti-ErbB2 antibodies on the sensing interface, the formed antigen-antibody complex would increase the coverage of the electrode surface, which would insulate the interface from the electrolyte solution to some extent and therefore inhibit the electrochemical redox process of inner Fc molecules, leading to a decreased electrochemical signal.
The assembly process of Fc-PEI/SWNTs multilayers on ITO electrode was characterized by cyclic voltammetry. As shown in Fig. 1A, even one bilayer of Fc-PEI/SWNTs modied ITO exhibits a pair of well-dened redox peaks located at +0.401 V and +0.349 V, which represent the oxidation and reduction of the xed Fc, respectively. The redox peak currents gradually increase with increasing the bilayer number of PEI-Fc/SWNTs from 1 to 4, suggesting abundant redox Fc could be controllably conned on ITO electrode through SWNTs mediated LBL assembly. A linear dependence between the peak currents and the number of the bilayer is revealed (inset of Fig. 1A), indicating the controllability of the assembling process. On the other hand, the peak-to-peak potential difference has no obvious change with increasing the number of assembly bilayer, suggesting efficient charge transfer promoted by SWNTs. In comparison with the electrode modied through MWNTs mediated LBL lm, ITO/(Fc-PEI/SWNTs) 4 exhibits lower background current signal. 28 These advantages are beneted to the construction of electrochemical sensors with high sensitivity and low detention limit. To obtain the highest sensitivity, ITO/ (Fc-PEI/SWNTs) 4 was chosen for further investigation.
The effect of scan rate on the cyclic voltammograms (CVs) obtained by ITO/(Fc-PEI/SWNTs) 4 electrode was also investigated. As shown in Fig. 1B, both the anodic and cathodic peak currents linearly increase with increasing the scan rate from 20 mV s À1 to 100 mV s À1 , indicating a surface-controlled electrochemical process of the conned Fc. AFM was used to characterize the surface morphology of ITO electrode before and aer assembly of (Fc-PEI/SWNTs) 4 multilayers. As revealed in Fig. 2A, ITO electrode shows a rough surface with many uniformly distributed ITO nanoparticles, whereas no ITO nanoparticle can be seen on (Fc-PEI/SWNTs) 4 /ITO, conrming the ITO has been fully covered with the polymer-SWNTs multilayers aer the LBL assembly (Fig. 2B).

Electrochemical characterization of the immunosensing interface
Electrochemical experiments were carried out to characterize the preparation of immunosensing interface and the following binding of ErbB2. As shown in Fig. 3, the ITO/(Fc-PEI/SWNTs) 4 / ErbB2-Ab electrode exhibits a decreased peak current in comparison to ITO/(Fc-PEI/SWNTs) 4 due to the insulative properties of macromolecules. When such modied electrode is blocked by BSA, the resultant electrode displays a further reduction in peak current signal. Aer incubation with ErbB2 solution (100 ng mL À1 ) for 30 min, the sensing electrode presents a signicant current decrease with increased peak-to-peak potential difference, suggesting the effective recognition and capture of ErbB2 by the sensing interface. As large amount of antibody-antigen complex covers the electrode surface, the electrochemical process of inner Fc probes is signicantly inhibited.

Optimization of analytical parameters
The experimental parameters including the concentration of reacted ErbB2-Ab and immune reaction time were optimized to obtain the highest sensitivity. During optimization of the ErbB2-Ab concentration, the activated ITO/(Fc-PEI/SWNTs) 4 electrode was incubated in ErbB2-Ab solutions at different concentrations (25,50,75,100,150, and 200 mg mL À1 ) for 2 h to evaluate the immobilization amount of ErbB2-Ab. As shown in  This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 22291-22296 | 22293 Fig. 4A, the current signal gradually decreases with increasing the concentration of ErbB2-Ab and reaches to a plateau as the concentration is over 100 mg mL À1 , indicating a saturated immobilization amount of ErbB2-Ab on the electrode. Therefore, the optimized concentration of ErbB2-Ab was set at 100 mg mL À1 for the fabrication of sensing interface. To determine the optimum immune reaction time for ErbB2 detection, kinetic experiments were carried out and the results were shown in Fig. 4B. The DPV peak current gradually decreases with increasing the reaction time and reaches a plateau as the time goes beyond 30 min due to the saturation effect. Thus, the optimal reaction time was set at 30 min. . As the ErbB2 concentration increases from 1.0 ng mL À1 to 200.0 ng mL À1 , the DPV peak current gradually decreases. The ratio of I/ I 0 was used as calibration, where I 0 represents the original DPV peak current of ITO/(Fc-PEI/SWNTs) 4 /ErbB2-Ab, and I represents the peak current aer incubation with ErbB2. The inset of Fig. 5B shows the corresponding calibration curve, which displays a good linearity between I/I 0 and logarithm of ErbB2 concentration in a dynamic range of 1.0 ng mL À1 to 200.0 ng mL À1 with a correlation coefficient of 0.9991. The limit of detection (LOD) was calculated to be 0.22 ng mL À1 at a signal/ noise ratio of 3, which is $68 times lower than the clinical cut-off value of ErbB2 established at 15 ng mL À1 . 29 On the other hand, the use of surface-conned Fc as signal indicator for reagentless detection could avoid the addition of external probes in measurement solution, providing facile and efficient strategy for detection and reducing possible contamination of the samples. The selectivity of the developed immunosensor was also investigated. The DPV current responses of the immunosensor towards ErbB2 and other different protein solutions including human immunoglobulin G (IgG), human serum albumin (HSA), carcinoembryonic antigen (CEA), or prostate specic antigen (PSA) were measured. As illustrated in Fig. 6, no obvious change of electrochemical signals is observed for even high concentration (ten times higher than that of ErbB2) of IgG, HSA, CEA, or PSA. This result suggests that the immunosensor has a good selectivity towards ErbB2 detection. To evaluate the reproducibility of the sensor, ve sensing electrodes were fabricated independently for the detection of ErbB2 (100 ng mL À1 ). The    relative standard deviation (RSD) of the measured currents is within 4.2%, indicating good reproducibility of the immunosensor. In order to evaluate the stability, the developed immunosensor was stored in PBS (0.1 M, pH 7.0) for 2 weeks at 4 C. Aerwards, the detection for ErbB2 (100 ng mL À1 ) was performed. In comparison with the result obtained on the freshly prepared immunosensor, only 3.4% decrease of DPV current is observed, suggesting good stability of the immunosensor.

Analysis of real samples
Standard addition experiments were carried out in human serum matrices (diluted by 20 times with 0.1 M PBS, pH 7.0) to evaluate the analytical reliability and potential for practical applications of the immunosensor. Three serum samples spiked with increasing amounts of ErbB2 were analysed by the immunosensor (parallel determination of each sample for 3 times). As demonstrated in Table 1, the recoveries range from 96.7% to 103% and the RSDs are less than 4.1%. Satisfactory recoveries demonstrate good precision and potential application of the proposed immunosensor in real sample analysis.

Conclusions
In summary, electrochemical immunosensor with surface-conned redox probe is simply constructed for reagentless detection of breast cancer biomarker. Layer-by-layer assembled multilayers with ferrocene (Fc) as electrochemical signal indicator and SWNTs as conductive enhancement materials are fabricated on the supporting electrode. Owning to the presence of SWNTs and abundant conned ferrocene, the fabricated immunosensor presents good performance for the detection of ErbB2 in terms of wide detection range, low detection limit, high stability and good reliability in analysis of real samples. On the other hand, the use of surface-conned Fc as signal indicator for reagentless detection could avoid the addition of external probes in measurement solution, thus efficient reducing possible contamination of the samples. The strategy for the fabrication of immunosensors could be easily extended for the detection of other tumour biomarkers and exhibits potential in clinical diagnostics and biochemical analysis.

Conflicts of interest
There are no conicts to declare.