Bio-bar-code-based photoelectrochemical immunoassay for sensitive detection of prostate-specific antigen using rolling circle amplification and enzymatic biocatalytic precipitation

Biosensors and Bioelectronics 101 (2018) 159–166

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

Bio-bar-code-based photoelectrochemical immunoassay for sensitive detection of prostate-specific antigen using rolling circle amplification and enzymatic biocatalytic precipitation ⁎

Kangyao Zhang, Shuzhen Lv, Zhenzhen Lin, Meijin Li , Dianping Tang

MARK

⁎

Key Laboratory for Analytical Science of Food Safety and Biology (MOE & Fujian Province), State Key Laboratory of Photocatalysis on Energy and Environment, Department of Chemistry, Fuzhou University, Fuzhou 350116, PR China

A R T I C L E I N F O

A B S T R A C T

Keywords: Photoelectrochemical immunosensor CdS nanorods Rolling circle amplification DNAzyme concatamers Enzymatic biocatalytic precipitation

Methods based on photoelectrochemistry have been developed for immunoassay, but most involve in a low sensitivity and a relatively narrow detectable range. Herein a new bio-bar-code-based split-type photoelectrochemical (PEC) immunoassay was designed for sensitive detection of prostate-specific antigen (PSA), coupling rolling circle amplification (RCA) with enzymatic biocatalytic precipitation. The bio-bar-code-based immunoreaction was carried out on monoclonal anti-PSA antibody (mAb1)-coated microplate using primer DNA and polyclonal anti-PSA antibody-conjugated gold nanoparticle (pDNA-AuNP-pAb2) with a sandwich-type assay format. Accompanying the immunocomplex, the labeled primer DNA on gold nanoparticle readily triggered RCA reaction in the presence of padlock probe/dNTPs/ligase/polymerase. The RCA product with a long singlestranded DNA could cause the formation of numerous hemin/G-quadruplex-based DNAzyme concatamers. With the assistance of nicking endonuclease, DNAzyme concatamers were dissociated from gold nanoparticle, which catalyzed the precipitation of 4-chloro-1-naphthol in the presence of H2O2 onto CdS nanorods-coated electrode (as the photoanode for the generated holes). The formed insoluble precipitate inhibited the electron transfer from the solution to CdS nanorods-modified electrode by using ascorbic acid as the electron donor. Under the optimum conditions, the photocurrent of the modified electrode decreased with the increasing of PSA concentration. A detectable concentration for target PSA with this system could be achieved as low as 1.8 pg mL−1. In addition, our strategy also showed good reproducibility, high specificity and accuracy matched well with commercial PSA ELISA kits for real sample analysis. These remarkable properties revealed that the developed PEC immunoassay has great potential as a useful tool for the detection of PSA in practical application.

1. Introduction Along with improvement of the people's life quality and the development of science and technology, people's health awareness is increasing gradually, which has greatly stimulated interest of researchers to explore various analytical methods for the highly selective detection of low-concentration analyte. Photoelectrochemical (PEC) sensor (as a new analytical technique on the basis of photochemistry and electrochemistry) has developed rapidly in recent years to monitor different analytes, e.g., proteins (J. Wang et al., 2017; X. Wang et al., 2017), antibiotics (Yan et al., 2015), nucleic acids (Wu et al., 2013) and heavy metals (Zang et al., 2014). Derived from the traditional electrochemical methods, the PEC method has the merits of simple operation, good portability and short response time. Due to different energy forms of exciter (light) and detector (electricity), the PEC analysis platform has

⁎

an excellent sensitivity and a low background signal. Despite some significant results, the detectable sensitivity still falls short of the practical requirements and large-scale applications. Generally, the development of highly efficient PEC sensing systems depends on at least three concerns: i) biomolecular immobilization in detection scheme, ii) the signal amplification and iii) photosensitive materials. The first key point is to design a feasible detection protocol since the PEC detection system involves the light illumination as well as the strong oxidation characteristics of photogenerated holes in photoactive materials (Shi et al., 2016; Wang et al., 2014; Wen and Ju, 2016; Zang et al., 2014), which inevitably causes the damage of the immobilized biomolecules on the electrode (especially using UV light) (Zhao et al., 2012; Yan et al., 2015; Ma et al., 2015; Zhang et al., 2016). In this case, design of biomolecular immobilization in the PEC system becomes more and more important. An overwhelming strategy is to

Corresponding authors. E-mail addresses: [email protected] (M. Li), [email protected] (D. Tang).

http://dx.doi.org/10.1016/j.bios.2017.10.031 Received 26 September 2017; Received in revised form 11 October 2017; Accepted 14 October 2017 Available online 16 October 2017 0956-5663/ © 2017 Elsevier B.V. All rights reserved.

Biosensors and Bioelectronics 101 (2018) 159–166

K. Zhang et al.

AuNP-pAb2). Introduction of primer DNA on the gold nanoparticle ensures the progression of RCA reaction. Accompanying with the formation of DNAzyme concatamers, the formed double-stranded DNA is cleaved via Nt.BbvCI (note: Nt.BbvCI is a nicking endonuclease that cleaves only one strand of DNA on a double-stranded DNA substrate) to release the DNAzyme concatamers. Upon addition of H2O2, the dissociated DNAzyme concatamers can catalyze 4-chloro-1-naphthol into an insoluble benzo-4-chlorohexadienone product, and coat the surface of CdS nanorods-modified electrode, thus inhibiting the electron transfer. In this case, the photocurrent of the modified electrode decreases using ascorbic acid as the electron donor.

separate PEC measurement from the biomolecular reaction (e.g., immunoreaction). The emergence of split-type detection mode provides a new idea for development of PEC sensing platform. Yamamoto et al. (1995) described a split-type flow cell of a polarized spectrophotometric detector in HPLC for colored amino acid-copper(II) complexes. Zhuang et al. (2018, 2015) devised two split-type PEC immunosensing systems for sensitive monitoring of disease biomarkers. Such a split-type detection scheme by immobilizing biomolecules in the reaction cell and measuring in PEC cell can efficiently avoid the damage of biomolecules. The signal amplification is another important factor in PEC sensing systems for achieving low limits of detection and quantification. Routine approaches are usually adopted via controlling the orientation and density of biomolecules immobilized on the substrates (e.g., microplate and magnetic beads) (Chen et al., 2014; Li et al., 2017; Wang et al., 2015). Unfavorably, the conformational freedom of the immobilized biomolecules (i.e., their structural configuration) is limited because of the steric-hindrance effect, thus resulting in reduction of the bonding efficiency and rate between the biological probes and the targets (Yu and Lai, 2013). In this regard, improvement of the signal molecules in the detection mode is extremely urgent. Rolling circle amplification (RCA) is a simple and efficient isothermal nucleic-acid signal amplification method mediated by DNA polymerases, in which a long single-stranded DNA containing numerous complementary copies of the short circular template can be synthesized (Ye et al., 2014; Kong et al., 2016). Since the discovery of RCA in the mid-1990s, the power, simplicity and versatility of the DNA amplification technique have made it an attractive tool for biomedical research and nanobiotechnology (Zhao et al., 2008). Typically, the amplification process is carried out in aqueous solution, on the surface of solid supports or even in a sophisticated biological environment, and does not require specialized equipment. Importantly, the function of the RCA products can be manipulated by designing the circular template sequence in a very predictable manner to meet the needs of specific applications (Niu et al., 2016; Cao et al., 2017; Deng et al., 2017; Zhu et al., 2016; Wen et al., 2012), especially to form hemin/G-quadruplex-based DNAzyme (Tang et al., 2012; Dong et al., 2013). To the best of our knowledge, there are few reports focusing on the development of RCA-based PEC immunoassays. To this end, our motivation of this study is to design a split-type PEC immunoassay by coupling with RCA-based formation of hemin/G-quadruplex-based DNAzyme concatamers (note: RCA product is a concatamer containing tens to hundreds of tandem repeats that are complementary to the circular template (Ali et al., 2014)). As the signal-generation tag, cadmium sulfide (CdS; an outstanding narrow band gap of ~2.4 eV because of the requirement of a low excited energy according to the equation: E = hv) (Harakeh et al., 2008) photosensitive semiconductor has been widely studied in the past years owing to its excellent light response. However, the presence of bulk recombination leads to the severe photocorrosion (Zhang et al., 2017a). To decrease the bulk recombination, researchers have made great efforts for the synthesis of CdS nanocrystals with different morphologies (e.g., nanorods, nanowires, quantum dots and nanoparticles). Typically, CdS nanorods increase the PEC activity by reducing the radial transport distance of the charge carriers and expanding the surface area for the reaction (Vaquero et al., 2017). Prostate-specific antigen (PSA) is a protein produced by normal prostate cells. The normal levels of PSA in the blood serum of healthy males are maintained at a low level (< 4 ng mL−1), while rising levels are associated with prostate cancer. Herein, we design a new bio-bar-code-based immunoassay method for photoelectrochemical (PEC) detection of target PSA on CdS nanorodsmodified electrode, coupling with rolling circle amplification and hemin/G-quadruplex-based DNAzyme concatamers (Scheme 1). The immunoreaction and rolling circle amplification are carried out on a microtiter well. In the presence of target PSA, the immobilized capture antibody (mAb1) on the microplate captures the biofunctionalized gold nanoparticle with the primer DNA and detection antibody (pDNA-

2. Experimental 2.1. Chemicals and reagents PSA standards were acquired from Biocell Biotechnol. Inc. (Zhengzhou, China). Monoclonal anti-PSA capture antibody (A45180, designated as mAb1) and monoclonal anti-PSA detection antibody (A45190, designated as pAb2) were the products of BiosPacific, Inc. (CA, USA). All high-binding polystyrene 96-well microplates were obtained from Greiner Bio-One (Frickenhausen, 705071, Germany). Gold nanoparticles with an average diameter of 16 nm were prepared according to our previous report (Zhang et al., 2012). T4 DNA ligase and Phi29 DNA polymerase were obtained from Thermo Fisher Scientific Inc. (Shanghai, China). Nt.BbvCI was acquired from New England Biolab (Beijing, China). Oligonucleotides and dNTPs were purchased from Sangon Biotech. Co., Ltd. (Shanghai, China). Bovine serum albumin (BSA) was acquired from Dingguo Biotechnol. Co., Ltd. (Beijing, China). Tris(2-carboxyethyl)phosphine hydrochloride (TCEP) and 4chloro-1-naphthol (4-CN) were obtained from Sigma-Aldrich (USA). Cadmium nitrate tetrahydrate [Cd(NO3)2·4H2O], thiourea, ethlylenediamine, 3,3′,5,5′-Tetramethylbenzidine (TMB) and ascorbic acid (AA) were purchased from Sinopharm Chem. Re. Co., Ltd. (Shanghai, China). All reagents (analytical grade) were used as received without further purification. Ultrapure water obtained from a Millipore water purification system (18.2 MΩ cm−1, Milli-Q, Millipore). The sequences of oligonucleotides used in this work: Primer DNA (pDNA): 5′-SH-TTAAT CCTCA GCTTA TTTGA GGTAG TAGGT TGTAT AGTT-3′ Linear padlock DNA: 5′-phosphate-CTACT ACCTC AAAAA CATCC CAACC CGCCC TACCC ACGCA ACTAT ACAAC-3′ Cleaved DNA: 5′-AAGCT GAGGA TT-3′ G-rich DNA: 5′-SH-TTAAT CCTCA GCTTA TTGGG TAGGG CGGGT TGGGA TGTT-3′ Buffer solutions were used as follows: DNA ligation buffer: 40 mM Tris-HCl, 10 mM MgCl2, 10 mM dithiothreitol (DTT), 0.5 mM ATP, pH 7.8 Reaction buffer: 33 mM Tris-HAc, 10 mM Mg(Ac)2, 66 mM KAc, 0.1% (v/v) Tween 20, 1.0 mM DTT, pH 7.9 CutSmart buffer: 50 mM KAc, 20 mM Tris-HAc, 10 mM Mg(Ac)2, 100 μg mL−1 BSA, pH 7.9 Phosphate buffered saline (PBS, 10 mM, pH 7.4): 2.9 g Na2HPO4·12H2O, 0.24 g KH2PO4, 0.2 g KCl, 8.0 g NaCl Washing buffer: 10 mM PBS, 0.05% Tween 20, pH 7.4 Blocking buffer: 10 mM PBS, 1.0 wt% BSA, pH 7.4 2.2. Fabrication of CdS nanorods-based photoanode CdS nanorods were synthesized by using a hydrothermal method according to the literature with minor modification (Yin et al., 2016). Generally, Cd(NO3)2·4H2O (8.1 mmol) was initially added into 40-mL ethlylenediamine to form a homogeneous solution under sonication and 160

Biosensors and Bioelectronics 101 (2018) 159–166

K. Zhang et al.

Scheme 1. (A) Schematic illustration of split-type photoelectrochemical (PEC) immunosensing platform for target prostate-specific antigen (PSA) on CdS nanorods-modified FTO electrode with a sandwich-type assay format between monoclonal anti-PSA antibody (mAb1)-coated microplate and primer DNA/polyclonal anti-PSA antibody-labeled gold nanoparticle (pDNA-AuNP-pAb2) accompanying rolling circle amplification (RCA) and hemin/G-quadruplex-based enzymatic catalytic precipitation; and (B) mechanism on the photocurrent generation of CdS nanorods under the light irradiation (CB: conduction band; VB: valence band; AA: ascorbic acid).

of the blank polystyrene 96-well microplate and then incubated overnight at 4 °C (To avoid the evaporation of the incubation solution, the microplates were covered with a sealing film). On the next day, the excess mAb1 was fully washed with washing buffer. To reduce unspecific adsorption, the microplates were incubated with the blocking buffer (300 μL per well) for 1 h at 37 °C with gentle shaking. After washing, 50 μL of PSA standard or sample was injected into the microplate, and then shaken for 30 min at 37 °C. After that, the resulting plates were washed three times with washing buffer to remove the unbound PSA. Following that, 50 μL of the as-prepared pDNA-AuNPpAb2 was pipetted into the well and incubated for another 30 min under the same conditions to form the sandwiched structure of mAb1/PSA/ pDNA-AuNP-pAb2.

continuous stirring at room temperature. Then, thiourea (24.3 mmol) was added into the resulting mixture. After being stirred for 30 min, the mixture was transferred to a 50-mL Teflon-lined stainless steel autoclave. The autoclave was heated in an electric oven at 160 °C for 24 h. After cooling to room temperature, the bright yellow precipitates were collected by centrifugation, rinsed with distilled water and ethanol for several times, and finally dried at 60 °C. The as-synthesized CdS nanorods were characterized by using high-solution transmission electron microscopy (HRTEM), X-ray diffraction (XRD) and UV–visible diffuse reflectance spectrum (DRS) (Please see Supporting Information). Prior to fabrication, fluorine-doped tin oxide (FTO) electrode was treated by successive sonication in water, acetone, ethanol and water, and then dried in oven. Next, CdS powder was ultrasonically dispersed in a mixed solution of H2O and ethlylenediamine (v/v, 25:1). Following that, 25 μL of CdS suspension (1.0 mg mL−1) was dropped onto the surface of electrode [note: a perforated (r = 3 mm) transparent tape was stuck on the cleaned FTO glass], and then dried in oven at 60 °C for further use.

2.5. Bio-bar-code assay and PEC measurement Initially, 25 μL of 0.5 μM padlock DNA was injected into the well including mAb1/PSA/pDNA-AuNP-pAb2 and then incubated at 37 °C for 30 min. Thereafter, 25 μL of DNA ligation buffer containing 20 U T4 DNA ligase was added in the resulting mixture. The reaction was carried out for 60 min at 37 °C. After completion of the reaction, the solution was removed. Afterwards, the reaction buffer (50 μL), phi29 DNA polymerase (10 U) and dNTPs (5 μL, 10 mM) were added and incubated for 60 min at 37 °C to conduct the RCA reaction. After that, 25 μL of solution (containing 5.0 μM hemin, 20 mM KCl and 0.5 μM cleaved DNA) was added to the plate, and incubated for 60 min at 37 °C (note: The double-stranded DNA containing a specific sequence that could be recognized by the endonuclease and the peroxidase-mimicking DNAzyme were formed during this process). After being washed with PBS to remove the excess hemin and DNA, CutSmart buffer (25 μL) and 2 U Nt.BbvCI were added to the well and incubated for 60 min at 37 °C. Following that, 25 μL of incubation solution containing 10 mM 4-CN and 0.6 mM H2O2 was pipetted into the well, and the mixture then rapidly transferred to CdS nanorods-modified FTO photoanode. After the precipitation reaction, the resulting electrode was rinsed with PBS and transferred to the detection cell for photocurrent measurement. The PEC measurements were carried out in 0.1 M Na2SO4 containing 10 mM AA under the irradiation of 500 W Xe lamp at 0 V on an AutoLab electrochemical workstation (μAutIII.Fra2.v, Eco Chemie, the Netherlands) with a classical three-electrode system including a CdS nanorods-modified FTO working electrode, a Pt-wire counter electrode and an Ag/AgCl reference electrode.

2.3. Conjugation of gold nanoparticles with primer DNA and pAb2 (pDNAAuNP-pAb2) The pDNA-AuNP-pAb2 conjugates were synthesized and prepared similarly to our previous report (Zhang et al., 2017b). Prior to functionalization, gold colloids were initially adjusted to pH 9.0–9.5 by using Na2CO3 solution (0.1 M). Then, pAb2 (200 μL, 0.5 mg mL−1) was injected to 4.0-mL gold colloids and incubated for 2 h at 4 °C with slight shaking. After that, the thiolated primer DNA (pDNA, 0.5 OD, reduced by TCEP) was added into the above mixture and incubated for another 2 h under the same conditions. Following that, 100 μL of 0.1 M Tris-HCl buffer (containing 0.5 M NaCl, pH 7.5) was injected to the resultant gold colloids. After being incubated overnight at 4 °C, the resulting mixture was obtained by centrifugation for 10 min at 13,000g (4 °C), then resuspended in 1.0 mL of PBS (containing 1.0 wt% BSA and 0.1 wt % sodium azide, pH 7.4), and stored at 4 °C when not in use. 2.4. Immunoreaction with target PSA and pDNA-AuNP-pAb2 Scheme 1 A displays the principle and procedure toward target PSA by coupling with the pDNA-AuNP-pAb2-based RCA and the DNAzymecatalyzed precipitation amplification. Initially, 50 μL of 10 μg mL−1 mAb1 in carbonate buffer solution (pH 9.6) was injected into one well 161

Biosensors and Bioelectronics 101 (2018) 159–166

K. Zhang et al.

3. Results and discussion

state of RCA product in the absence and presence of nicking endonuclease by using gel electrophoresis (Fig. 1D). As shown in lane 'a', no band was observed in the absence of endonucleases. Upon addition of endonucleases, a bright electrophoresis band was observed at the beginning of lane 'b'. Besides, the migrant rate of the biggest fragment of the DNA marker (1500 bp in size) was much faster than that of the RCA product, indicating that the base number of RCA product was far more than 1500 bp. These results revealed that the product of the amplification reaction was a very long single-stranded DNA, and the RCA product/DNAzyme could be released from pDNA-AuNP-pAb2 in the assistance of nicking endonuclease. As is well-known to all, the Faradic electrochemical impedance spectroscopy (EIS) is a valid and useful tool to characterize the interface properties of electrodes. To demonstrate the third issue, EIS was used to study the interfacial electrochemical characteristics of the modified electrodes at different stages. All the impedimetric measurements were performed in 10 mM PBS (pH 7.4) containing 5.0 mM K3Fe(CN)6/K4Fe (CN)6 and 0.1 M KCl. Fig. 1E shows the corresponding Nyquist plots (inset: a Randles equivalent circuit), four parameters constituted the equivalent circuit: the electrolyte resistance (Rs), the double-layer capacitance (Cdl), the interfacial charge-transfer resistance (Ret) and Warburg element (Zw) (Katz and Willner, 2003). Typically, the chargetransfer resistance represents the electron transfer kinetics of redox couple at the electrode surface, and the value of Ret can be read directly from the diameter of the semicircle in the Nyquist plots. As seen from the curve 'a', the Ret value of the bare FTO electrode was very small. A significant increase in the Ret value was observed after the introduction of the CdS nanorods into the bare FTO electrode surface (curve 'b'), indicating that the presence of cadmium sulfide hindered the electron exchange between the redox probe [K3Fe(CN)6/K4Fe(CN)6] and the electrode. In the presence of incubation solution containing 4-CN and H2O2, the hemin/G-quadruplex released from the well initiated the precipitation reaction and generated an insoluble/insulating product on the electrode surface. The precipitate further obstructed the electron transfer between the ferricyanide ion and the electrode. As expected, the Ret value greatly increased after the biocatalytic precipitation reaction (curve 'c'). Finally, we confirmed the feasibility of our design by investigating the photocurrent responses of the FTO electrode at different stages in 0.1 M Na2SO4 solution containing 10 mM AA (Fig. 1F). As depicted in curve 'a', no apparent photocurrent was generated on bare FTO electrode under the light irradiation. After the introduction of CdS nanorods onto the FTO surface, a significant photocurrent response was observed (curve 'b'). Curves 'c-e' display the photocurrents of CdS nanorodsmodified FTO electrode after incubation with the deposition solution and DNAzyme obtained through the immunoreaction and amplification reaction at different-concentration PSA. Obviously, the observed photocurrent decreased with the increase of target PSA concentration. It was easy to explain this phenomenon. The insoluble precipitate coated on the electrode hindered the transfer of electrons from AA to the photogenerated holes of CdS, which accelerated the electron-hole pairs recombination, thereby resulting in the decreasing photocurrent. The yield of precipitate was related to the amount of hemin/G-quadruplex, and the amount of mimetic enzyme depended on the concentration of target PSA. After incubation with the deposition solution in the absence of target PSA, the photocurrent response (curve 'c') basically stayed the same (curve 'b' vs. curve 'c'). The results in Fig. 1F were in consonance with the previous EIS results. On the basis of the above-mentioned results, we might get a conclusion that the CdS nanorods-based split-type PEC sensor proposed in this work could be a preliminary tool for the detection of PSA.

3.1. Characteristics of bio-bar-code-based PEC immunosensing platform on CdS nanorods Scheme 1 shows the schematic illustration of bio-bar-code-based split-type PEC immunosensing platform and the generation mechanism of anodic photocurrent on the CdS nanorods. The purpose of designing the split-type sensing platform is to promote the bonding efficiency and rate between the pre-immobilized mAb1 and the subsequent antigen/ antibody. Also, the design avoids potential factors (such as the light irradiation and the PEC system itself) that may cause damage to protein molecules. The successful construction of pDNA-AuNP-pAb2 benefited from the existence of the following two issues: i) the S-Au bond between the thiolated oligonucleotides and gold nanoparticles; and ii) the interaction between cysteine or NH3+-lysine residues of proteins and gold nanoparticles (Zhuang et al., 2015). In the presence of target PSA, the sandwiched immunocomplex (mAb1/PSA/pDNA-AuNP-pAb2) is formed on the microplate. After the immunoreaction, the primer DNA on the pDNA-AuNP-pAb2 can hybridize with padlock DNA to form a template, which triggers the subsequent RCA reaction with the help of relevant enzymes. In our design, the obtained long single-stranded DNA contains thousands of G-rich fragments, which can bind hemin to form hemin/Gquadruplex-based DNAzyme concatamer. Owing to the introduction of cleave DNA (5′-AAGCTGAGGATT-3′), a short double-helix strand is generated at the 5′ end of primer DNA. Upon addition of Nt.BbvCI, the endonuclease recognizes the specific sequence in this double-helix strand and then cleaves it. In this case, the dissociated DNAzyme concatamers can efficiently oxidize 4-CN into the insoluble benzo-4chlorohexadienone precipitate on the surface of CdS photoanode. As shown in Scheme 1, the detection signal (photocurrent change) derives from the insoluble/insulating product of the enzymatic biocatalytic reaction. By recording the change in the photocurrent, we indirectly get the concentration of target analyte PSA in the sample. The photoactive semiconductors could generate electron-hole pairs under light irradiation, while the photocurrent originated from the separation of these photogenerated electron-hole pairs (Zhao et al., 2015). Thus, we investigated the PEC response of CdS photoanode. The role of AA in this work was to inhibit the recombination of electronhole pairs and enhance the photocurrent of photoanode. As shown in Fig. 1A, CdS photoanode could maintain a high and stable photocurrent under the repeated irradiation. To verify the feasibility of our design, three major issues should be carefully investigated in this work: (i) whether the peroxidase-mimicking DNAzyme could be produced after the amplification reaction, (ii) whether the formed DNAzyme could be released from the surface of gold nanoparticles, and (iii) whether the resulting DNAzyme could induce the deposition reaction of 4-CN to suppress the generation of photocurrent. To this end, we used the classical TMB-H2O2 system to clarify the first issue. As seen from the photograph in Fig. 1B, a bluecolored solution only appeared in the 'e' well, while the solution color of the remaining wells (wells 'a-d') was colorless (the table below the photograph gave the substance that was absent in the whole reaction). The blue color of well 'e' was derived from the oxidation of the TMB caused by the formed DNAzyme. In contrast, when target PSA (well 'a'), pDNA-AuNP-pAb2 (well 'b'), padlock DNA (well 'c') or hemin (well 'd') was missing in the reaction, the color of the solution remained colorless, these results indicated that the mAb1, PSA, pDNA-AuNP-pAb2 or the pure RCA amplification product (the single-stranded DNA that did not bind to hemin) could not induce the oxidation of TMB. The results were in accordance with those obtained by UV–vis absorption spectroscopy (Fig. 1C). Thus, we could get a conclusion that the DNAzyme could generate in the presence of PSA, pDNA-AuNP-pAb2, padlock DNA and hemin. The absence of any substances would not produce the DNAzyme we needed. In addition, the release of the formed DNAzyme is also an important issue of our design. Therefore, we monitored the

3.2. Optimization of experimental conditions To obtain a high analytical performance of the split-type PEC immunosensing platform, several experimental parameters containing the 162

Biosensors and Bioelectronics 101 (2018) 159–166

K. Zhang et al.

Fig. 1. (A) The stability of CdS nanorods-modified FTO electrode in 0.1 M Na2SO4 containing 10 mM AA by repeatedly controlling the light irradiation 'on↔off'; (B) photographs of TMBH2O2 system in different components (+: with; −: without); (C) UV–vis absorption spectra of the corresponding figure 'B'; (D) agarose gel electrophoresis images for the obtained RCA products in the (a) absence and (b) presence of nicking endonuclease; (E) Nyquist plots of (a) bare FTO, (b) CdS nanorods-modified FTO electrode, and (c) electrode 'b' after enzymatic biocatalytic precipitation (inset: equivalent circuit); and (F) photocurrents of (a) bare FTO, (b) CdS nanorods-modified FTO electrode and (c-e) the CdS photoanode after immunoreaction with different-concentration PSA standards (c: 0 ng mL−1; d: 0.1 ng mL−1; e: 1.0 ng mL−1).

Fig. 2. Effects of (A) incubation time for the immunoreaction; (B) RCA reaction time and (C) hemin/G-quadruplex-based enzymatic catalytic precipitation time for 4-CN (1.0 ng mL−1 PSA used in all cases).

deposition time of 4-CN on ΔI. As depicted in Fig. 2C, we observed that ΔI increased with the increasing deposition time, and an optimal signal could be obtained at 25 min. To save the assay time, 25 min was used as the catalytic deposition time of 4-CN.

incubation time for the antigen-antibody immunoreaction, the RCA reaction time, and the reaction time for hemin/G-quadruplex catalyzed precipitation of 4-CN were systematically investigated (1.0 ng mL−1 PSA used as an example) and evaluated based on the change of photocurrent (ΔI, where ΔI = I0 - I, I0 and I represent the photocurrent of CdS photoanode before and after the biocatalytic precipitation reaction). As seen from Fig. 2A, ΔI increased with the prolongation of incubation time for the antigen-antibody immunoreaction, and tended to a plateau after 30 min. A longer immunoreaction time did not cause the significant increase in the ΔI. Therefore, 30 min was chosen as the incubation time for immunoreaction. As described above, the amount of hemin/G-quadruplex would seriously influence the signal (ΔI) of the PEC immunosensing platform. The polymerases needed enough time to produce the long single DNA strand. As shown in Fig. 2B, ΔI increased with the increment of RCA reaction time and reached a plateau after 60 min. Thus, 60 min was selected for the RCA reaction time. Next, we studied the effect of the

3.3. Analytical performance of bio-bar-code-based PEC immunosensing platform Under the optimum conditions, we studied the analytical performance of the proposed split-type PEC immunosensing platform toward different-concentration PSA standards. As seen from Fig. 3A, the photocurrent responses decreased with increasing target PSA concentration in the sample. A good linear relationship between ΔI and the logarithm of the analyte level was acquired in the range from 0.005 ng mL−1 to 50 ng mL−1 (Fig. 3B), and the linear regression equation was ΔI (μA) = 0.806 × lg(C[PSA]/ng mL−1) + 2.2552 (R2 = 0.996, n = 7). The detection limit (LOD) was 1.8 pg mL−1, as calculated at the 3sblank 163

Biosensors and Bioelectronics 101 (2018) 159–166

K. Zhang et al.

Fig. 3. (A, C) Photocurrents and (B, D) the corresponding calibration curves of the split-type PEC immunosensing platform toward different-concentration PSA standards by using different G-rich fragments generation mechanism: (A, B) the products of RCA reaction and (C, D) the pre-synthesized products, respectively; (E) the specificity of the developed PEC immunoassay against target PSA (1.0 ng mL−1) and non-target analytes (e.g., 10 ng mL−1 CEA, 10 ng mL−1 AFP, 10 U mL−1 CA 125 and 10 U mL−1 CA 15-3); and (F) comparison of the results obtained by the developed PEC immunoassay and the referenced PSA ELISA kit for human PSA serum samples.

developed bio-bar-code-based PEC immunoassay were comparable with other PEC immunosensing strategies (Table 1). These results indicated that the introduction of RCA would increase the sensitivity of the PEC biosensor and lower the LOD. The improvement of sensor performance was mainly attributed to the powerful signal-amplification capability of RCA. Next, we monitored the reproducibility of the developed immunosensing platform by detecting three different-concentration PSA standards (0.05, 1.0 and 20 ng mL−1). As analyzed from the experimental data, the coefficients of variation (CVs) by using the same-batch

criterion (i.e., from the expression of 3 S/K, where 'K' and 'S' represent the slope of the calibration plot and the standard deviation of 15 determination of blank solution, respectively). For comparison, we simplified the sensing platform by removing RCA from the original design procedures and investigated its analytical properties (note: in the simplified method, G-rich DNA replaced primer DNA in the synthesis of DNA-AuNP-pAb2). As depicted in Fig. 3C-D, the linear range was 0.1 – 20 ng mL−1 and the LOD was 0.04 ng mL−1. Obviously, the LOD obtained from the proposed immunosensing platform was lower than that of using DNA-AuNP-pAb2. Moreover, the analytical properties of the 164

Biosensors and Bioelectronics 101 (2018) 159–166

K. Zhang et al.

Table 1 Comparison of analytical properties of differently PEC immunosensing strategies toward target PSA. Materialsa

Type

Linear range

LOD

Refs.

SnS2@mpg-C3N4 rGO/FeOOH rGO/Ca: CdSe CdS:Mn/g-C3N4 Ag@Au CdSe/TiO2 CdSe/TiO2 ZnCdHgSe/gold nanorods TiO2/CdS:Mn BiVO4-rGO CdS/TiO2 CdS nanorods

Signal-off Signal-on Signal-on Signal-off Signal-off Signal-off Signal-off Signal-off Signal-off Signal-off Signal-off Signal-off

0.00005–10 ng mL−1 0.001–100 ng mL−1 0.005–50 ng mL−1 0.01–20 ng mL−1 0.01–100 ng mL−1 0.01–100 ng mL−1 0.05–100 pg mL−1 0.001–50 ng mL−1 0.001–100 ng mL−1 0.01–80 ng mL−1 0.001–3.0 ng mL−1 0.005–50 ng mL−1

21 fg mL−1 0.3 pg mL−1 2.6 pg mL−1 3.8 pg mL−1 3.0 pg mL−1 2.7 pg mL−1 17 fg mL−1 0.1 pg mL−1 0.32 pg mL−1 3.0 pg mL−1 0.32 pg mL−1 1.8 pg mL−1

Zhang et al. (2017c) Zhou et al. (2017) X. Wang et al. (2017) Zhang et al. (2017b) Zhu et al. (2017) Dong et al. (2017) Fan et al. (2017) Wang et al. (2016) Fan et al. (2016) Shu et al. (2016) Zhuang et al. (2015) This work

a

rGO: reduced graphene oxide.

whereas ascorbic acid was used as the electron donor to enhance the photocurrent response. Compared with common “signal-off” photoelectrochemical sensing method, the developed protocol could efficiently avoid the reduction of background photocurrent and improve the stability of the sensing platform. In addition, the introduction of RCA technique could enhance the sensitivity and lower the detection limit of the sensing system. Experimental results showed that our protocol displayed a wide linear range with good reproducibility, high specificity and good anti-interference capability. Nevertheless, the disadvantages of this sensing platform are extra reagents (e.g., additional enzymes to release the generated DNAzyme) and a long reaction time. Moreover, the entire sensing process is very tedious which involves many steps (because any mistakes made in one of steps will have strong impact on the final results). Thus, our future works should be centered on simplification of the steps/reagents and the improvement of the practical application.

pDNA-AuNP-pAb2 and the PEC platforms were 8.9%, 6.7% and 7.2% toward 0.05, 1.0, and 20 ng mL−1, respectively. By using differentbatch pDNA-AuNP-pAb2 and CdS nanorods-modified FTO electrodes, the CVs were 9.7%, 8.5% and 10.1% toward the above-mentioned concentrations. Hence, the reproducibility of the proposed method was acceptable. Furthermore, we also investigated the specificity of the PEC immunosensor by challenging this sensing platform against other common cancer biomarkers, for example, carcinoembryonic antigen (CEA), alpha fetoprotein (AFP), cancer antigen 125 (CA 125) and cancer antigen 15-3 (CA 15-3). As shown in Fig. 3E, the interfering proteins including CEA, AFP, CA125 and CA15-3 only caused a similar signal compared with the blank sample, while a significant photocurrent change was achieved toward target PSA. In addition, the coexistence of non-target proteins with target PSA did not cause an obvious change in ΔI relative to target analyte PSA alone, suggesting that the specificity of the developed method was quite satisfactory. Next, the stability of CdS nanorods-modified FTO electrodes and pDNA-AuNP-pAb2 were also studied over an eight-week period. When they were stored at 4 °C and measured intermittently (every 3–5 days) for 1.0 ng mL−1 PSA (used as an example), they retained 98.9%, 98.1% and 90.3% (n = 3) of the initial signal after being stored for 2, 4 and 8 weeks, respectively. Hence, the stability of our system was acceptable. To demonstrate the accuracy and the application potential of our designed platform in real sample analysis, several human serum specimens with different-concentration target PSA collected from the local hospital were tested by the proposed sensing strategy and a commercial ELISA kit (Sigma-Aldrich), respectively. The obtained results by two methods were displayed in Fig. 3F, and the regression equation was fitted to y = 0.972x + 0.002 (R2 = 0.988, n = 8). Further, the developed PEC immunoassay was utilized for the detection of 6 spiked new-born calf serum samples. Prior to measurement, these six samples including 0.01, 0.1, 1.0, 5.0, 20 and 40 ng mL−1 were prepared by spiking PSA standards into new-born calf serum (Dingguo Biotechnol. Inc., Beijing, China; note: The aim of using new-born calf serum is to avoid the PSA interfering in the normal human serum). The contents obtained by the developed PEC immunoassay were 0.092, 0.12, 1.13, 4.81, 22.4 and 43.5 ng mL−1 PSA toward the above-mentioned five samples, respectively. The recovery was 92–113%. These assayed results indicated that the developed method could be a reliable tool for the determination of PSA in the biological fluids.

Acknowledgments This work was financially supported by the National Natural Science Foundation of China (21675029 & 21475025), the National Science Foundation of Fujian Province (2014J07001), and the Program for Changjiang Scholars and Innovative Research Team in University (IRT15R11). Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2017.10.031. References Ali, M., Li, F., Zhang, Z., Zhang, K., Kang, D., Ankrum, J., Le, X., Zhao, W., 2014. Chem. Soc. Rev. 43, 3324–3341. Cao, J., Shi, H., Chen, X., Wang, Z., Chen, Y., Shu, Y., Zhu, X., 2017. Sens. Actuators B 238, 331–336. Chen, X., Zhou, G., Song, P., Wang, J., Gao, J., Lu, J., Fan, C., Zuo, X., 2014. Anal. Chem. 86, 7337–7342. Deng, K., Li, C., Huang, H., Li, X., 2017. Sens. Actuators B 238, 1302–1308. Dong, H., Wang, C., Xiong, Y., Lu, H., Ju, H., Zhang, X., 2013. Biosens. Bioelectron. 41, 348–353. Dong, Y., Cao, J., Liu, Y., Ma, S., 2017. Biosens. Bioelectron. 91, 246–252. Fan, D., Ren, X., Wang, H., Wu, D., Zhao, D., Chen, Y., Wei, Q., Du, B., 2017. Biosens. Bioelectron. 87, 593–599. Fan, G., Shi, X., Zhang, J., Zhu, J., 2016. Anal. Chem. 88, 10352–10356. Harakeh, M., Alawieh, L., Saouma, S., Halaoui, L., 2008. Phys. Chem. Chem. Phys. 11, 5962–5973. Katz, E., Willner, I., 2003. Electroanalysis 15, 913–947. Kong, X., Wu, S., Cen, Y., Yu, R., Chu, X., 2016. Biosens. Bioelectron. 79, 679–684. Li, R., Zhang, Y., Tu, W., Dai, Z., 2017. ACS Appl. Mater. Interfaces 9, 22289–22297. Ma, Z., Ruan, Y., Zhang, N., Zhao, W., Xu, J., Chen, H., 2015. Chem. Commun. 51, 8381–8384. Niu, C., Song, Q., He, G., Na, N., Ouyang, J., 2016. Anal. Chem. 88 (22), 11062–11069. Shi, X., Fan, G., Shen, Q., Zhu, J., 2016. ACS Appl. Mater. Interfaces 8, 35091–35098.

4. Conclusions In summary, we report on a new split-type photoelectrochemical immunosensing platform for the sensitive detection of low-abundance cancer biomarkers (PSA acted as the model analyte in this work) by using the polymerase-triggered rolling circle amplification, accompanying with the enzymatic biocatalytic precipitation reaction. The photocurrent derived from the CdS nanorods-modified FTO electrode, 165

Biosensors and Bioelectronics 101 (2018) 159–166

K. Zhang et al.

Zang, Y., Lei, J., Hao, Q., Ju, H., 2014. ACS Appl. Mater. Interfaces 6, 15991–15997. Zhang, B., Liu, B., Tang, D., Niessner, R., Chen, G., Knopp, D., 2012. Anal. Chem. 84, 5392–5399. Zhang, N., Ruan, Y., Ma, Z., Zhao, W., Xu, J., Chen, H., 2016. Biosens. Bioelectron. 85, 294–299. Zhang, J., Li, W., Li, Y., Zhong, L., Xu, C., 2017a. Appl. Catal. B-Environ. 217, 30–36. Zhang, K., Lv, S., Lin, Z., Tang, D., 2017b. Biosens. Bioelectron. 95, 34–40. Zhang, Y., Liu, Y., Li, R., Khan, M., Gao, P., Zhang, Y., Wei, Q., 2017c. Sci. Rep. 7 (4629), 1–7. Zhao, W., Ali, M., Brook, M., Li, Y., 2008. Angew. Chem. Int. Ed. 47, 6330–6337. Zhao, W., Ma, Z., Yu, P., Dong, X., Xu, J., Chen, H., 2012. Anal. Chem. 84, 917–923. Zhao, W., Xu, J., Chen, H., 2015. Chem. Soc. Rev. 44, 729–741. Zhou, Q., Lin, Y., Shu, J., Zhang, K., Yu, Z., Tang, D., 2017. Biosens. Bioelectron. 98, 15–21. Zhu, Y., Wang, H., Wang, L., Zhu, J., Jiang, W., 2016. ACS Appl. Mater. Interfaces 8, 2573–2581. Zhu, Y., Xu, F., Zhang, N., Zhao, W., Xu, J., Chen, H., 2017. Biosens. Bioelectron. 91, 293–298. Zhuang, J., Han, B., Liu, W., Zhou, J., Liu, K., Yang, D., Tang, D., 2018. Biosens. Bioelectron. 99, 230–236. Zhuang, J., Tang, D., Lai, W., Xu, M., Tang, D., 2015. Anal. Chem. 87, 9473–9480.

Shu, J., Qiu, Z., Zhou, Q., Lin, Y., Lu, M., Tang, D., 2016. Anal. Chem. 88, 2958–2966. Tang, L., Liu, Y., Ali, M., Kang, D., Zhao, W., Li, J., 2012. Anal. Chem. 84, 4711–4717. Vaquero, F., Navarro, R., Fierro, J., 2017. Appl. Catal. B-Environ. 203, 753–767. Wang, W., Hao, Q., Wang, W., Bao, L., Lei, J., Wang, Q., Ju, H., 2014. Nanoscale 6, 2710–2717. Wang, G., Shu, J., Dong, Y., Wu, X., Li, Z., 2015. Biosens. Bioelectron. 66, 283–289. Wang, J., Long, J., Liu, Z., Wang, W., Hu, C., 2017a. Biosens. Bioelectron. 91, 53–59. Wang, X., Xu, R., Sun, X., Wang, Y., Ren, X., Du, B., Wu, D., Wei, Q., 2017b. Biosens. Bioelectron. 96, 239–245. Wang, Y., Yu, X., Ye, X., Wu, K., Wu, T., Li, C., 2016. Anal. Chim. Acta 943, 106–113. Wen, Y., Xu, Y., Mao, X., Wei, Y., Song, H., Chen, N., Huang, Q., Fan, C., Li, D., 2012. Anal. Chem. 84, 7664–7669. Wen, G., Ju, H., 2016. Anal. Chem. 88, 8339–8345. Wu, Y., Zhang, B., Guo, L., 2013. Anal. Chem. 85, 6908–6914. Yamamoto, A., Matsunaga, A., Mizukami, E., Hayakawa, K., Miyazai, M., Nishimura, M., Kitaoka, M., Fujita, T., 1995. Analyst 120, 1137–1139. Yan, K., Liu, Y., Yang, Y., Zhang, J., 2015. Anal. Chem. 87, 12215–12220. Ye, T., Chen, J., Liu, Y., Ji, X., Zhou, G., He, Z., 2014. ACS Appl. Mater. Interfaces 6, 16091–16096. Yin, X., Li, L., Jiang, W., Zhang, Y., Zhang, X., Wan, L., Hu, J., 2016. ACS Appl. Mater. Interfaces 8, 15258–15266. Yu, Z., Lai, R., 2013. Anal. Chem. 85, 3340–3346.

166