Analytical Methods PAPER
Cite this: Anal. Methods, 2018, 10, 2390
An efficient method for the synthesis of a polymer brush via click chemistry and its ultrasensitive electrochemical detection of AFP† Dongcheng Yang,‡ Jing Wang,‡ Hou Chen, Huawei Yang, Donglei Wei and Liangjiu Bai
* Hui Xu, Wenxiang Wang, *
A sensitive sandwich-type electrochemical immunosensor based on polyacrylonitrile-g-poly(hydroxyethyl methacrylate) (PAN-g-PHEMA)/graphene oxide (GO) composites was prepared for the ultrasensitive detection of tumor markers (alpha fetal protein, AFP as a model). The functional polymer brush was efficiently prepared by the union of cyano click chemistry and controlled/“living” radical polymerization (CRP). The designed immunosensor showed a low limit of detection (LOD, 0.285 pg mL1), with a wide Received 18th March 2018 Accepted 30th April 2018
detection range (2.5 101 to 2.5 104 pg mL1) (R2 ¼ 0.9876). The electrochemical immunosensor also exhibited high sensitivity, favorable stability, excellent reproducibility and good selectivity towards AFP. Moreover, the immunosensor could be applied to the analysis of AFP in serum samples with satisfactory
DOI: 10.1039/c8ay00583d
results. This present work may provide a general model for the detection of tumor markers based on
rsc.li/methods
sandwich-type immunosensors.
1. Introduction Cancer is one of the greatest threats to human health and it is one of the main causes of human death.1 Many factors can cause cancer, such as radiation, exposure to carcinogenic chemicals, and genetic and environmental factors.2,3 There are more than 200 kinds of cancer-related diseases affecting different parts of the body, so the clinical test of cancer is very complex.4–8 Early detection and diagnosis of tumors play a crucial role in the prevention and treatment of cancers. To meet the requirements for early diagnosis of cancers, tumor markers must be tested effectively. Compared with healthy people, the concentrations of tumor markers in a patient's body are much higher. The content of tumor markers represents the current state of tumors.9 Therefore, sensitive and rapid diagnosis methods are very important for their detection. Tumor markers are potentially one of the most valuable tools for early cancer detection, accurate pretreatment staging, determining the response of cancer to chemotherapy treatment, and monitoring disease progression.10 But in the early stages of tumor formation, tumor markers are at low levels. Thus for early screening of small tumors, the limit of detection (LOD) is quite important.
Shandong Key University Laboratory of High Performance and Functional Polymer, School of Chemistry and Materials Science, Ludong University, Yantai 264025, China. E-mail:
[email protected];
[email protected] † Electronic supplementary information (ESI) available: Fig. S1: SEM surface images of GO and GO–DETA. Table S1: Comparison of methods for the detection of AFP. See DOI: 10.1039/c8ay00583d ‡ These authors contributed equally to the work.
2390 | Anal. Methods, 2018, 10, 2390–2397
An electrochemical immunoassay utilizes the highly specic binding between the antigen and antibody, which combines the immunoassay and electrochemical sensor. It is a widely used method for the early detection of tumor markers. Improving the sensitivity of detection and signal amplication has been a hot research topic.11 It has been reported that the sensitivity of detection can be improved through increasing the amount of antibodies in the early stages of research.12 In recent years, nanomaterials have been used as carriers to prepare nanoprobes that enable signal amplication, and the sensitivity could also be greatly improved.13–15 Chen et al. developed an ultrasensitive electrochemical immunosensor based on Au/Pt modied C60 bimetallic nanoclusters as labels for the detection of Vangl 1, which exhibited a wide detection range from 0.1 pg mL1 to 450 pg mL1.16 Guo et al. used Pt/Pd nanoparticles loaded on reduced graphene oxide (rGO) as labels for the quantitative detection of tumor markers, which exhibited a high sensitivity and low detection limit.17 GO as a typical nanomaterial, which has the advantages of a huge surface area, unique two-dimensional structure and excellent mechanical properties, has caused widespread concern.18,19 In addition, GO has a large amount of oxygen groups on its surface, and it is very conducive to the loading of signal labels. Thus, GO is an excellent nanomaterial available for electrochemical immunosensors. Xu et al. prepared a biofunctional nanoprobe with GO as a carrier, and signal amplication was realized by using the nanoprobe.20 Li et al. reported an enzyme-linked immunosorbent based on allochroic molecule modied carboxyl GO, which successfully achieved the simultaneous colorimetric detection of diagnostic biomarkers.21
This journal is © The Royal Society of Chemistry 2018
Paper
Analytical Methods
It should be noted that the use of GO alone as the carrier may not provide enough effective sites for the loading of signal molecules due to the limited active functional groups and inevitable aggregation. To overcome these shortcomings, functional polymers can be used to pre-modify GO.22 It has been reported that a series of polymers with good biocompatibility were used to detect tumor markers, such as poly(ethylene glycol), poly[styrene-alter-(maleic acid)], polydiacetylene and polydopamine.23–25 Polymer brushes, which were dened as polymer chains attached to a surface or an interface, were further suitable in the elds of medical implants, drug delivery and biosensors compared to ordinary polymers.26–28 There are multiple monomer units in the polymer brush, and repeated units can be connected to signal molecules to greatly improve the sensitivity of the electrochemical immunosensor. However, the application of polymer brushes has been limited due to the complex synthesis process and low conversion rate.29–32 Click chemistry, as a rising, simple and efficient synthesis method, has been rapidly applied to the synthesis of polymer brushes.33–35 Tsarevsky and co-workers successfully prepared a tetrazole polymer with a narrow molecular weight distribution (Mw/Mn < 1.10) using polyacrylonitrile (PAN) and sodium azide under the action of catalyst ZnCl2.36 The research showed that the nitrile group is a suitable group for click chemistry. And acrylonitrile is an excellent raw material for cyano click chemistry due to the large amount of nitrile groups. Our group has carried out a great deal of research on the controlled/“living” radical polymerization (CRP) of PAN in the previous studies.37–39 Thus it is feasible to prepare a composite which was constructed using functional polymer brushes and graphene oxide to improve the sensitivity of electrochemical detection. In this paper, we designed a novel electrochemical immunosensor for the quantitative detection of alpha fetal protein (AFP). High sensitivity was achieved using polyacrylonitrileg-poly(hydroxyethyl methacrylate) (PAN-g-PHEMA, polymer brush) connected to a large number of signal molecules, which were rapidly synthesized by cyano click chemistry. And due to the huge surface area of GO, it can be modied with multiple functional polymer brushes. This is the rst time that GO has been combined with a polymer brush for the detection of tumor markers. The immunosensor assembly process is shown in Scheme 1. The immunosensor showed a low limit of detection (LOD, 0.285 pg mL1). It also provided potential applications in clinical monitoring of tumor markers.
2.
Experimental
2.1
Reagents and materials
2-Hydroxyethyl methacrylate (HEMA, $99%, J&K SCIENTIFIC LTD) was passed through a column of Al2O3 to remove the inhibitor and stored at 4 C. Acrylonitrile (AN, $98%, Tianjin FuChen Chemical Reagents, China) was distilled and stored at 4 C. Ethyl 2-bromoisobutyrate (EBiB, $98%) was from Aladdin Chemistry Co., Ltd and copper bromide (CuBr2, $98.5%) was from Tianjin DaMao Chemical Reagents Co., Ltd, China. 1,1,4,7,7-Pentamethyldiethylenetriamine (PMDETA, $98%, Beijing HWRK Chem Co., Ltd) and azobis-isobutyronitrile
This journal is © The Royal Society of Chemistry 2018
(AIBN) were recrystallized in ethyl alcohol for purication. 4Cyanopentanoic acid dithiobenzoate (CPADB) was prepared by the method reported in the literature.40 Anthraquinone-2carboxylic acid ($99.5%, Zhengzhou Alfachem Co., Ltd), tetrahydrofuran (THF, $99.5%, Tianjin Bodi Chemical Company, China), ascorbic acid (Vc, $99.7% Tianjin Bodi Chemical Company, China), methanol (CH3OH, $99.5%, Tianjin Chemical Reagents, China), N,N-dimethylformamide (DMF, $99.5%, Tianjin Chemical Reagents, China), and bovine serum albumin (BSA) were purchased from Sigma-Aldrich. GO was prepared via Hummer's method.41 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride, N-hydroxysuccinimide (EDC and NHS, $99.5%, Shanghai Yuanye Biological Technology Company, China), sodium azide (NaN3, $98%, ChengDu Micxy Chemical Co., Ltd, China), and ammonium chloride (NH4Cl, $99.5%, Sinopharm Chemical Reagent Co., Ltd, China) were all used as received. EDC/NHS, AFP, primary antibodies (Ab1) and secondary antibodies (Ab2) were all dissolved in phosphate buffered saline (10 mM, pH ¼ 7.4). All solutions were prepared with ultra-pure water (18 MU cm). Live subject statement. All experiments on human serum were approved by the Animal Ethics Committee of College of School of Chemistry and Materials Science, Ludong University. All experiments concerning a human patient were performed aer obtaining informed consent. In this study, one of the coauthors, Mr Yang voluntarily donated blood (10 mL). The School of Chemistry and Materials Science is committed to the protection and safety of human subjects involved in research. 2.2
Characterization methods
The number-average molecular weight (Mn) and molecular weight distribution of the synthesized PHEMA and PAN were determined by using a gel permeation chromatograph (GPC, Waters 1515) equipped with a refractive-index detector (Waters 2414) using a HR column (7.8 300 mm). FT-IR spectra measurements were performed on a Nicolet iS50 (Thermo Fisher Nicolet, United States) Fourier transform infrared spectrometer equipped with Thermo Nicolet corporation OMINIC 32 soware. The 1H NMR spectrum of the prepared polymers was recorded via an INOVA 400 MHz nuclear magnetic resonance instrument; (CD3)2SO was used as a solvent and the internal standard was tetramethylsilane (TMS). The morphologies of the GO and GO–DETA were characterized via a scanning electron microscope (SEM), BRUKER SU8010. GO, the polymer brush and the composite were characterised by UV-vis absorption spectroscopy using a Shimadzu UV-visible spectrophotometer (UV2550). Thermogravimetric analysis (TGA) was performed on a NETZSCH STA409PC instrument. The CHI 660C electrochemical workstation (CH Instruments, Shanghai) with a traditional three-electrode system was used to carry out square wave voltammetry (SWV). 2.3
Preparation of the functional polymer brush
PHEMA was synthesized by atom transfer radical polymerization (ATRP). 1.0 mL of monomer (HEMA, 8.25 mmol), 1.0 mL of solvent (DMF), 1.2 mL of initiator (EBiB, 0.082 mmol), 8.5 mL of Anal. Methods, 2018, 10, 2390–2397 | 2391
Analytical Methods
ligand (PMDETA, 0.041 mmol), 7.3 mg of reductant (ascorbic acid, 0.041 mmol) and 3.7 mg of catalyst (CuBr2, 0.0165 mmol) were added to a 5.0 mL ampoule in the following order: CuBr2, ascorbic acid, EBiB, PMDETA, DMF and HEMA. Aer being degassed with nitrogen at 0 C for 10 min, the ampoule was sealed using a sealing device and reacted at 65 C with magnetic stirring. Aer the end of the reaction, the ampoule was cooled in ice water. The product was puried with the addition of 3 mL of THF : methanol (1 : 1) mixture. Then, the solution was poured into 10 mL distilled water and moved into a dialysis tube. The solution was dialysed for 48 h. And then, the product was obtained by freeze-drying. 0.1 g of PHEMA powder, 0.01 g of anthraquinone-2carboxylic acid, 0.075 g of DCC, 0.001 g of DMAP and 4 mL of THF were added to a 5.0 mL ampoule. The mixture was reacted at 25 C for 24 h with magnetic stirring. Aer completion of the reaction, the mixture was puried by dialysis for 48 h. Finally, the solid was obtained by freeze-drying. 2.0 mL of monomer (AN, 0.031 mol), 2.0 mL of solvent (DMF), 12.3 mg of initiator (AIBN, 0.0775 mmol), and 42 mg of RAFT-agent (CPADB, 0.155 mmol) were added to a 5.0 mL ampoule in the following order: CPADB, AIBN, AN and DMF. Aer being degassed with nitrogen and sealed, the mixture was reacted at 75 C with magnetic stirring. When the required reaction time was up, the ampoule was placed in ice water and cooled. Aerward, the mixture was dissolved in 5.0 mL DMF. Then the product was puried using methanol as a precipitant. The PAN powder was obtained by suction ltration and drying. About 0.53 g of PAN powder and 10 mL of DMF were added to a 50 mL three-necked ask with stirring at room temperature. And then, 0.535 g of NH4Cl and 0.65 g of NaN3 were added. The ask was immediately placed into an oil-bath pan equipped with magnetic stirrers and heated to 120 C, respectively, and kept at this temperature for 12 h. This solution was added into distilled water and treated with 0.5 M HCl as the precipitant to get poly(5-vinyltetrazole) (PVT). The PVT was washed with deionized water several times and dried until constant weight under vacuum. About 0.025 g of PVT, 0.05 g of PHEMAanthraquinone-2-carboxylic acid, 0.5 mL of triethylamine and 1 mL of DMF were added to a 5.0 mL ampoule. Then the ampoule was reacted at 65 C for 3 h with stirring. Aer completion of the reaction, the mixture was puried by dialysis for two days. And nally, the polymer brush was obtained by freeze-drying.
Paper
1 h at ambient temperature. The excess EDC and NHS were removed by centrifugation aer the shaking. Then 0.1 mL Ab2 (0.01 mg mL1) was added and the mixture was immediately shaken at 37 C for 2 h. The excess antibody was removed by centrifugation and washed with PBS. 6 mg of polymer brush was dissolved in 1 mL DMSO, and then 1 mL of activation reagents were added to the solution. The solution was subjected to shaking for 1 h at ambient temperature. Soon 100 mL of the solution was immediately mixed with GO–DETA and subjected to shaking for 2 h at 37 C. Then aer centrifugation and washing with PBS, the GO composite was obtained and stored at 4 C for future use. 2.5
3. 2.4
Preparation of GO composites
50 mg of GO was dispersed in 50 mL ethanol. Next the mixture was sonicated for several days to make the GO disperse fully. Then 0.52 mL DETA was added. The mixture was allowed to react at room temperature for 24 h with stirring. The products were puried by centrifugation and washed several times with ethanol, methanol and acetone. Finally, the solid was dried until constant weight under vacuum to get GO–DETA.42 1 mg of GO–DETA was dispersed in 1 mL distilled water, and then 1 mL of activation reagents (0.4 M EDC and 0.1 M NHS) were added to the mixture. The mixture was kept under shaking conditions for
2392 | Anal. Methods, 2018, 10, 2390–2397
Fabrication of the immunosensor
Before each experiment, the bare glassy carbon electrode (GCE) was successively polished with 0.3 mm and 0.05 mm alumina slurry, and it was then rinsed with distilled water, and cleaned ultrasonically sequentially in water and 95 % ethanol for 5 min. The electrode was modied with reducing graphene oxide (rGO) and gold nanoparticles (AuNPs) aer being dried in a nitrogen atmosphere. 5 mL GO dispersions (1 mg mL1) and 100 mM HAuCl4 were added to a small beaker and degassed with nitrogen for 15 min. Then, the GCE was immersed in the solution and cyclic voltammetry was performed. The potential was set to 0 to 1.5 V and the scan rate was 0.05 V s1 for 300 s. The GCE was stored at 4 C for future use. To load Ab1, the modied GCE was immersed in Ab1 solution (0.1 mg mL1) for 30 min at 4 C, and then the GCE was immersed in BSA solution (1%) to block non-specic sites for 30 min at ambient temperature. Subsequently, AFP and the nanoprobe were assembled onto the electrodes in turn by incubation at 30 C for half an hour, respectively. It is necessary for each step to remove the excess materials by careful washing. Thus, the electrochemical immunosensor with multiple signal amplication was ultimately obtained. In order to explore the selectivity of the immunosensor, 1 mg mL1 prostate-specic antigen (PSA), 1 mg mL1 carcinoembryonic antigen (CEA) and 25 ng mL1 AFP were detected using the immunosensor under the same conditions, respectively. And their signal values were compared. To evaluate the practicability of the immunosensor, different concentrations of AFP were added to normal human serum (diluted 10 times) and immersed for 20 min, respectively.
3.1
Results and discussion Preparation and characterization of the nanoprobe
The fabrication process of the immunosensor is described in Scheme 1. Scheme 1(a) shows the synthesis of the polymer brush via cyano click chemistry. PAN was synthesized by RAFT polymerization using 4-cyanopentanoic acid dithiobenzoate (CPADB) as the RAFT regent, which was used to form the polymer brush as the backbone. And as the side chains, PHEMA was prepared by ATRP. Anthraquinone-2-carboxylic acid, as the signal molecule, was used to modify PHEMA through its unique interaction with the wealth of hydroxyl groups. The rst-order kinetic plots for the ATRP of PHEMA are shown in Fig. 1(a)
This journal is © The Royal Society of Chemistry 2018
Paper
Scheme 1 (a) Preparation process of the polymer brush; (b) assembling process toward the nanoprobe and (c) schematic presentation of the immunosensor fabrication.
and (b). Consistent with expectations, the conversion of the monomer increased signicantly with an increase of reaction time. And (ln([M0]/[M]) ¼ kpapp[R]t) showed an extremely signicant linear relationship with reaction time. This indicated the rst-order kinetics of the polymerization process. The plots of the number-average molecular weight (Mn) and molecular weight distributions (Mw/Mn) (Fig. 1(b))
Fig. 1 (a) First-order kinetic investigation of PHEMA with DMF as the solvent at 65 C. (b) Number-average molecular weight (Mn) and molecular weight distribution (Mw/Mn) vs. monomer conversion for the ATRP of PHEMA. (c) First-order kinetic investigation of PAN with DMF as the solvent at 60 C. (d) Number-average molecular weight (Mn) and molecular weight distribution (Mw/Mn) vs. monomer conversion for the RAFT polymerization of PAN.
This journal is © The Royal Society of Chemistry 2018
Analytical Methods
indicated that the Mn of PHEMA increased linearly with the monomer conversion. And the Mw/Mn values remained quite low (Mw/Mn < 1.4). For PAN, the results of rst-order kinetic plots are summarized in Fig. 1(c) and (d). The rst-order kinetic plots visually indicated that the polymerization proceeded smoothly and strictly obeyed the rst-order kinetics. Namely, the number of radicals taking part in the RAFT polymerization was persistent. In Fig. 1(d), the obtained PAN possessed a well-controlled molecular structure and comparatively low molecular weight distributions (Mw/Mn < 1.5). It can also be seen that the molecular weight of PAN measured via a GPC increased linearly versus the conversion of the monomer and was close to the theoretical values. The preparation procedure of the GO nanocomposite is presented in Scheme 1(b). The polymer brush and GO were combined to form the composite. GO was used to immobilize antibodies and amplify multiple signals. Scheme 1(c) shows the process of immunosensor assembly. The 1H NMR spectrum (Fig. 2(a)) was used to analyse the polymer brush. The characteristic peaks of PHEMA and PAN can be distinguished clearly. The two protons of PAN in the polymer brush gave two different resonances. The signal at 1.99 ppm (a) corresponded to the methylene protons of PAN in the polymer brush. And the signal at 3.10 ppm (b) derived from the methylidyne protons of PAN in the polymer brush. The signal at d ¼ 1.77–1.99 ppm belonged to the displacement of methyl (c) and methylene (d) in HEMA. The signal at d ¼ 3.90 ppm (e) was attributed to the methylene group which was connected to the ester group. The signal at d ¼ 3.54 ppm (f) was due to the methylene group which was connected to the hydroxide radical. And the signal at d ¼ 4.80 ppm (g) was attributed to the hydroxide radical. In summary, these chemical shis clearly indicated the successful preparation of the polymer brush.
Fig. 2 (a) 1H NMR spectrum of the polymer brush recorded at 25 C with DMSO-d6 as solvent; (b) FT-IR spectra of PAN, PHEMA and the polymer brush; (c) FT-IR spectra of the polymer brush, anthraquinone2-carboxylic acid and polymer brush–anthraquinone-2-carboxylic acid; (d) the TG spectrum of PHEMA, PAN and the polymer brush.
Anal. Methods, 2018, 10, 2390–2397 | 2393
Analytical Methods
PHEMA, PAN and the polymer brush were also characterized by FT-IR. As shown in Fig. 2(b), strong absorption peaks corresponding to the stretching vibration of CH2 could be found at 2930 cm1. The peak at 2240 cm1 was the characteristic peak of C^N in PAN which disappeared in the polymer brush, suggesting the decrease of C^N groups. And the peaks at 1490, 1450, 1390, and 1075 cm1 of the polymer brush are attributed to the characteristic C]N stretching vibrations of the tetrazole ring. This is because only some of the tetrazole groups were involved in the reaction. Thus, the IR spectral results demonstrated the successful synthesis of polymer brushes. In order to investigate whether the anthraquinone-2-carboxylic acid was connected to the polymer brush, the products were also characterized by FT-IR (Fig. 2(c)). A set of weak peaks appeared at 1400–1650 cm1 in the spectra of polymer brushanthraquinone-2-carboxylic acid. The result corresponds to the stretching vibration of the benzene ring which was consistent with that of anthraquinone-2-carboxylic acid. In addition, the other peaks of the polymer brush did not change signicantly aer the reaction. The synthesis of the polymer brush was further characterized by TGA. As shown in Fig. 2(d), PHEMA has a mass loss of about 10% below 150 C due to the evaporation of water absorbed, similar to the polymer brush. The mass then remained almost constant until the temperature rose to 300 C. The sharp weight loss at 300 C was attributed to the decomposition of lowmolecular segments into volatile small molecules. In contrast, the thermal stability of PAN is much better than that of PHEMA. PAN only had a partial weight loss between 300 C and 450 C. And its nal weight was much higher than that of PHEMA. The TGA curve of the polymer brush was between the other two curves roughly. The share of PHEMA in the polymer brush can be estimated using the following formula: Masspolymer brush ¼ massPHEMA CPHEMA + massPAN (1 CPHEMA) where masspolymer brush, massPHEMA and massPAN are all shown in the TG curves and CPHEMA is the share of PHEMA in the polymer brush.43,44 The PHEMA weight content relative to the polymer brush was calculated to be 75.7%. These, to some extent, demonstrated the successful synthesis of the polymer brush. The FT-IR spectra were used to characterize the functional group changes aer GO was modied to GO–DETA (Fig. 3(a)). GO exhibited representative peaks at 1720, 1613, 1206 and 1048 cm1, attributed to C]O stretching, C]C stretching of the aromatic ring, epoxy C–O stretching, and alkoxy C–O stretching vibration, respectively. The peak at 860 cm1, which was attributed to the stretching vibration of epoxy C–O in GO, decreases for GO–DETA. In addition, GO–DETA exhibited new peaks at 1363 and 1120 cm1, corresponding to O–H bending and C–N stretching vibrations, respectively. These results indicated that DETA successfully reacted with the epoxy groups of GO. Furthermore, the broad peak at 3000–3300 cm1 also showed the generation of amino groups. Thus, GO was successfully modied to GO–DETA. The scanning electron microscopy results of GO and GO–DETA could also conrm this
2394 | Anal. Methods, 2018, 10, 2390–2397
Paper
(a) FT-IR spectra of GO and GO–DETA; (b) the TG spectrum of GO, the polymer brush and the composite; (c) UV-vis absorption spectra of GO, the polymer brush and the composite; (d) EDS of GO and the composite.
Fig. 3
conclusion. Fig. S1† shows the smooth surface of the GO sheet and the wrinkles on the GO–DETA sheet via SEM. As a reducing agent, DETA could destroy the carbon skeleton of GO and introduce defects (pentagon–heptagon rings) thus leading to the loss of long range ordering on the surface of GO; these reasons cause the GO surface to wrinkle.45 These wrinkles reduced the mechanical strength and electrochemical properties of GO, but had a benecial effect on the composite of the polymer brush and GO.46–48 The SEM image of GO–DETA also demonstrates that the modication of GO with DETA did not cause damage to the structure of the GO sheet. The TGA curves of GO and the composite are shown in Fig. 3(b). The weight loss of GO was about 44% between 150 C and 450 C, while the composite exhibited a rapid weight loss of about 80% between 150 and 450 C. In addition, the polymer brush weight content relative to the GO was roughly calculated to be 82.3%. To further demonstrate the successful synthesis of the composite, the UV-vis spectra of GO, the polymer brush and the composite (Fig. 3(c)) were also studied. For the polymer brush, there were two strong peaks at 290 nm and 340 nm. And the characteristic absorption band of GO is shown at 260 nm. The peak position has a signicant shi aer the composite was constructed, due to the increase of particle size. Moreover, the GO and composite were also characterized by element distribution spectra (EDS). As shown in Fig. 3(d), the contents of C and O elements in GO were about 85.3% and 14.7%. And a new N element appeared in the composite due to the tetrazole groups generated aer click chemistry. Another difference was that the O content increases to 24.3%, which was attributed to the abundant O element in PHEMA. All the results illustrated that the composite was successfully constructed. 3.2
Optimization of the immunosensor response
To demonstrate the detection of AFP using the immunosensor constructed using the GO/polymer brush composite, the
This journal is © The Royal Society of Chemistry 2018
Paper
electrochemical behavior of the immunosensor was investigated. Fig. 4(a) shows the typical square wave voltammetry curve of the immunosensor with a peak at 0.52 V. Compared to the signal value of the blank experiment, the presence of AFP (2.5 ng mL1) led to a signicantly high current value. The response was attributed to the construction of a sandwich-type immunosensor when AFP exists. Anthraquinone-2-carboxylic acid as a signal molecule in the nanoprobe was oxidized to generate electrochemical signals. The resulting electrochemical impedance spectroscopy (EIS) spectra of the modied electrode were compared to conrm the validity of each modication step of the electrode. As shown in Fig. 4(b), the electron transfer resistance (Ret) of the rGO/AuNPs GCE was smaller than that of the bare GCE, which was consistent with the high conductivity and electron transfer rate of AuNPs. Aer immobilization of Ab1 and BSA, Ret increased, for these macromolecules hindered the conduction of current. And aer loading AFP and the nanoprobe, the value of Ret further increased. The results indicated that rGO/AuNPs, Ab1, BSA, AFP and the nanoprobe were all successfully loaded on the surface of the GCE.49 The content of BSA has a remarkable inuence on electrochemical behavior. To optimize the content of BSA, a series of BSA solutions with concentrations from 0% to 2% were prepared and the immunosensors were tested by SWV. As shown in Fig. 4(c), the signal value increased with increasing
Analytical Methods
BSA concentration up to 1%. When the BSA content was less than 1%, the non-specic sites cannot be completely blocked. But the signal value decreased when the BSA content exceeded 1%. This is because the excessive BSA hindered the transfer of electrons on the surface of the electrode. Thus, 1% is the optimum concentration of BSA.50 Another key inuencing factor is the incubation time. The effect of incubation time was investigated in the time range of 10–35 min. Fig. 4(d) shows that the signal value rapidly increased within the rst 30 min and then remained stable due to the saturation of the antigen–antibody binding. The proportion of the polymer brush in composites was also investigated to explore its effect on electrochemical behavior with the amount of GO (m ¼ 1 mg) being constant. Fig. 4(e) shows that the current response rapidly increased when the amount of polymer brush was less than 6 mg. And then the signal value tended to be stable due to the saturation of the GO/ polymer brush. It can be seen from the thermogravimetric results that the polymer brush in the composite accounted for a large proportion. As an important parameter of the polymer brush, the molecular weight of PHEMA also has a great impact on the electrochemical behavior. We prepared different composites constructed from a series of PHEMA with different molecular weights (Mn1 ¼ 25 800 g mol1, Mn2 ¼ 27 400 g mol1, Mn3 ¼ 29 300 g mol1, Mn4 ¼ 32 800 g mol1 and Mn5 ¼ 44 000 g mol1), respectively. The detection of AFP (2.5 ng mL1) using these composites was explored, and the results are shown in Fig. 4(f). With the increase of PHEMA molecular weight, the signal value increased signicantly due to more signal molecules connected to PHEMA. 3.3
Fig. 4 (a) Square wave voltammetry (SWV) curves of the immunosensor in the absence and presence of AFP (25 pg mL1); (b) electrochemical impedance spectra (EIS) of different modified electrodes in 0.1 M KCl containing 2.5 mM K3Fe(CN)6 and K4Fe(CN)6; (c) peak currents against the content of BSA; (d) the plots of the peak currents against the incubation time of AFP during sandwich immunoassay; (e) the plots of the peak currents against the graft density of composites; (f) peak currents against the molecular weight of PHEMA.
This journal is © The Royal Society of Chemistry 2018
Selectivity, reproducibility and stability
PSA, CEA and AFP were used to evaluate the selectivity of the immunosensor. As shown in Fig. 5(a), the signal values of PSA and CEA were all obviously lower than that of AFP, while the concentrations of PSA and CEA were signicantly higher. The results conrmed the acceptable specicity of the proposed immunoassay for the determination of AFP.51 To investigate the reproducibility of the immunosensor, we used a series of electrodes to detect AFP (2.5 ng mL1) under the same conditions; the results are shown in Fig. 5(b). The relative standard deviation (RSD) of the measurements for the electrodes was 1.6%, proving the good reproducibility of the immunosensor. To measure the stability of the immunosensor, the working electrodes were prepared and stored at 4 C when not in use. As shown in Fig. 5(c), the current response of the immunosensor decreased by 2.5% aer 6 days. The current response still remained at 91.2% aer 15 days. These results showed that the immunosensor constructed using the GO/polymer brush composite possessed good stability. 3.4
Electrocatalytic detection of AFP
Under the optimum conditions, different concentrations of AFP were detected using the immunosensor, and the results are shown in Fig. 6. The immunosensor exhibited a wider detection
Anal. Methods, 2018, 10, 2390–2397 | 2395
Analytical Methods
Paper
from Table 1 that the recovery and relative standard deviation values were acceptable. The results implied that the accuracy of the immunosensor was good, which also indicated that the immunosensor could be used for analytical applications in complex human serum.
4. Conclusion
Fig. 5 (a) Selectivity investigation of the electrochemical immunosensor for AFP (25 ng mL1), CEA (1 mg mL1) and PSA (1 mg mL1); (b) the reproducibility study of the immunosensor; (c) the time stability study of the immunosensor.
A sensitive electrochemical immunosensor for the detection of AFP has been prepared with GO/polymer brush composite signal amplication. GO can connect with a large amount of polymer brushes and capture primary antibodies to achieve signal amplication, due to its huge surface area. The designed immunosensor has high sensitivity, favorable stability and good selectivity towards AFP. It showed a wider detection range of 2.5 101 to 2.5 104 pg mL1, with a low detection limit of 0.285 pg mL1. In addition, the excellent reproducibility and accuracy of the immunosensor indicated promising applications in clinical diagnosis.
Conflicts of interest The authors declare no competing nancial interest.
Acknowledgements The study was nancially supported by the National Natural Science Foundation of China (No. 51573075 and 51773086), the Natural Science Foundation of Shandong Province (No. ZR2016BM27), the Project of Shandong Province Higher Educational Science (No. J16LC20) and the Technology Program and the Program for Scientic Research Innovation Team in Colleges and universities of Shandong Province. Quantitative measurements of the peak currents as a function of the concentration of AFP.
Fig. 6
range from 2.5 pg mL1 to 2.5 104 pg mL1 (R2 ¼ 0.9876) with a lower detection limit of 0.285 pg mL1 (S/N ¼ 3). Table S1† shows the comparison of different methods for the detection of AFP. It can be seen that the immunosensor we prepared has a signicantly low detection limit and an acceptable detection range. To evaluate the accuracy of the immunosensor, the standard addition method was used. Different samples were prepared by adding different concentrations of AFP to human serum. The analytical results for AFP are shown in Table 1. It can be seen
Table 1
Recovery of AFP in serum samples
Entry
Standard value (ng mL1)
Average value (ng mL1)
Recovery (%)
1 2 3
1.75 0.175 0.0175
1.67 0.181 0.0185
95.4 103.4 105.8
2396 | Anal. Methods, 2018, 10, 2390–2397
References 1 B. E. Burcu and S. M. Kemal, Talanta, 2015, 132, 162–174. 2 B. Bohunicky and S. A. Mousa, Nanotechnol., Sci. Appl., 2010, 4, 1–10. 3 I. A. Darwish, T. A. Wani, N. Y. Khalil, T. Aboul-Fadl, A. A. Kadia and A. R. A. Al-Majeda, Anal. Methods, 2013, 5, 1502–1509. 4 I. F. Akyildiz, J. M. Jornet and C. Han, Phys. Commun., 2014, 12, 16–32. 5 Y. Li, R. Yuan, Y. Chai and Z. Song, Electrochim. Acta, 2011, 56, 6715–6721. 6 H. Su, R. Yuan, Y. Chai, L. Mao and Y. Zhuo, Biosens. Bioelectron., 2011, 26, 4601–4604. 7 X. B. Sun and Z. F. Ma, Anal. Chim. Acta, 2013, 780, 95–100. 8 S. Wu, J. S. Chen, Y. J. Tian, X. M. Tang, W. Li and J. M. Li, Anal. Methods, 2015, 7, 1843–1848. 9 X. Liu, Y. Qin, C. Deng, J. Xiang and Y. Li, Talanta, 2015, 132, 150–154. 10 H. K. Choi and H. L. Ji, Anal. Methods, 2013, 5, 6964–6968. 11 C. M. Cheng, A. W. Martinez, J. Gong, C. R. Mace, S. T. Phillips, E. Carrilho, K. A. Mirica and
This journal is © The Royal Society of Chemistry 2018
Paper
12
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
G. M. Whitesides, Angew. Chem., Int. Ed., 2010, 46, 4771– 4774. C. Y. Yean, B. Kamarudin, D. A. Ozkan, L. S. Yin, P. Lalitha, A. Ismail, M. Ozsoz and M. Ravichandran, Anal. Chem., 2008, 80, 2774–2779. S. Song, Y. Qin, Y. He, Q. Huang, C. Fan and H. Y. Chen, ChemInform, 2010, 39, 4234–4243. H. Xu, Q. Li, L. Wang, Y. He, J. Shi, B. Tang and C. Fan, Chem. Soc. Rev., 2014, 43, 2650–2661. J. Wang, D. C. Yang, M. F. Chen, B. F. Liu, H. Chen, H. Xu, W. X. Wang and L. J. Bai, Anal. Methods, 2018, 10, 526–532. Q. Chen, C. Yu, R. Gao, L. Gao, Q. Li, G. Yuan and J. He, Biosens. Bioelectron., 2016, 79, 364–370. A. Guo, Y. Li, W. Cao, X. Meng, D. Wu, Q. Wei and B. Du, Biosens. Bioelectron., 2015, 63, 39–46. J. Wang, Y. T. Wang, H. Chen, H. Xu, W. X. Wang and D. C. Yang, Sens. Actuators, B, 2018, 258, 998–1007. Y. Liu, L. Xu, J. S. Liu, X. Y. Liu, C. H. Chen, G. Y. Li and Y. F. Meng, Chem. Eng. J., 2016, 285, 698–708. H. Xu, D. Wang, S. He, J. Li, B. Feng, P. Ma, P. Xu, S. Gao, S. Zhang and Q. Liu, Biosens. Bioelectron., 2013, 50, 251–255. C. Li, Y. Yang, D. Wu, T. Li, Y. Yin and G. Li, Chem. Sci., 2016, 7, 3011–3016. D. Feng, L. Li, X. Han, X. Fang, X. Li and Y. Zhang, Sens. Actuators, B, 2014, 201, 360–368. M. Cui, Z. Song, Y. Wu, B. Guo, X. Fan and X. Luo, Biosens. Bioelectron., 2016, 79, 736–741. W. Hu, G. He, H. Zhang, X. Wu, J. Li, Z. Zhao, Y. Qiao, Z. Lu, Y. Liu and C. M. Li, Anal. Chem., 2014, 86, 4488–4493. F. Ma, R. J. Qu, C. M. Sun, C. H. Wang, C. N. Ji, Y. Zhang and P. Yin, J. Hazard. Mater., 2009, 172, 792–801. M. P. Weir and A. J. Parnell, Polymers, 2011, 3, 2107–2132. M. Zhang, L. Liu, C. Wu, G. Fu, H. Zhao and B. He, Polymer, 2007, 48, 1989–1997. B. Zhao and W. J. Brittain, Prog. Polym. Sci., 2000, 25, 677– 710. M. Schappacher and A. Deffieux, Macromolecules, 2005, 33, 7371–7377. J. Ding, C. Xiao, Z. Tang, X. Zhuang and X. Chen, Macromol. Biosci., 2011, 11, 192–198.
This journal is © The Royal Society of Chemistry 2018
Analytical Methods
31 S. S. Sheiko, B. S. Sumerlin and K. Matyjaszewski, Prog. Polym. Sci., 2008, 33, 759–785. 32 J. Zhao, G. Zhang and S. Pispas, J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 2320–2328. 33 H. F. Gao and K. Matyjaszewski, Macromolecules, 2006, 39, 4960–4965. 34 R. Ranjan and W. J. Brittain, Macromol. Rapid Commun., 2007, 28, 2084–2089. 35 B. Wang, B. Dong, B. Li, B. Zhao and C. Y. Li, Polymer, 2010, 51, 4814–4822. 36 N. V. Tsarevsky, K. V. Bernaerts, B. Dufour, F. E. Du Prez and K. Matyjaszewski, Macromolecules, 2004, 37, 9308–9313. 37 H. Chen, Y. Liang, M. Wang, P. Lv and Y. Xuan, Chem. Eng. J., 2009, 147, 297–301. 38 X. Liu, H. Chen, C. Wang, R. Qu, C. Ji, C. Sun and Y. Zhang, J. Hazard. Mater., 2010, 175, 1014–1021. 39 G. Zong, C. Hou, R. Qu, C. Wang and N. Ji, J. Hazard. Mater., 2011, 186, 614–621. 40 S. Thang, Y. Chong, R. Mayadunne, G. Moad and E. Rizzardo, ChemInform, 1999, 30, 2435–2438. 41 A. Yang, J. Li, C. Zhang, W. Zhang and N. Ma, Appl. Surf. Sci., 2015, 346, 443–450. 42 Z. Su, H. Wang, K. Tian, F. Xu, W. Huang and X. Tian, Composites, Part A, 2016, 84, 64–75. 43 G. L. Li, D. Wan, K. G. Neoh and E. T. Kang, Macromolecules, 2010, 43, 10275–10282. 44 Q. Li, L. Zhang, Z. Zhang, N. Zhou, Z. Cheng and X. Zhu, J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 2006–2015. 45 A. Kaniyoor, T. T. Baby, T. Arockiadoss, N. Rajalakshmi and S. Ramaprabhu, J. Phys. Chem. C, 2011, 115, 17660–17669. 46 X. Shen, X. Lin, N. Youse, J. Jia and J. K. Kim, Carbon, 2014, 66, 84–92. 47 J. Zhang, J. Zhao and J. Lu, ACS Nano, 2012, 6, 2704–2711. 48 W. Zhu, T. Low, V. Perebeinos, A. A. Bol, Y. Zhu, H. Yan, J. Tersoff and P. Avouris, Nano Lett., 2012, 12, 3431–3436. 49 A. Mostafa, R. Mahdi, N. Navid, A. Khadijeh and N. M. Hossein, Biosens. Bioelectron., 2016, 77, 99–106. 50 Y. Wang, H. Chen, H. Xu, W. Wang and L. Bai, Mater. Chem. Phys., 2016, 183, 534–541. 51 D. Wang, N. Gan, J. Zhou, P. Xiong, Y. Cao, T. Li, D. Pan and S. Jiang, Sens. Actuators, B, 2014, 197, 244–253.
Anal. Methods, 2018, 10, 2390–2397 | 2397