An electrochemical immunosensor for cardiac Troponin I using electrospun carboxylated multi-walled carbon nanotube-whiskered nanofibres

An electrochemical immunosensor for cardiac Troponin I using electrospun carboxylated multi-walled carbon nanotube-whiskered nanofibres

Author’s Accepted Manuscript An electrochemical immunosensor for Cardiac Troponin I using electrospun carboxylated multiwalled carbon nanotube -whiske...

1MB Sizes 1 Downloads 13 Views

Author’s Accepted Manuscript An electrochemical immunosensor for Cardiac Troponin I using electrospun carboxylated multiwalled carbon nanotube -whiskered nanofibres Babak Rezaei, Ahmad Mousavi Shoushtari, Mohammad Rabiee, Lokman Uzun, Wing Cheung Mak, Anthony P.F. Turner www.elsevier.com/locate/talanta

PII: DOI: Reference:

S0039-9140(18)30052-3 https://doi.org/10.1016/j.talanta.2018.01.046 TAL18269

To appear in: Talanta Received date: 27 October 2017 Revised date: 16 January 2018 Accepted date: 17 January 2018 Cite this article as: Babak Rezaei, Ahmad Mousavi Shoushtari, Mohammad Rabiee, Lokman Uzun, Wing Cheung Mak and Anthony P.F. Turner, An electrochemical immunosensor for Cardiac Troponin I using electrospun carboxylated multi-walled carbon nanotube -whiskered nanofibres, Talanta, https://doi.org/10.1016/j.talanta.2018.01.046 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

An electrochemical immunosensor for Cardiac Troponin I using electrospun carboxylated multi-walled carbon nanotube -whiskered nanofibres

Authors: Babak Rezaei a, Ahmad Mousavi Shoushtari b,*, Mohammad Rabiee c, Lokman Uzun d , Wing Cheung Mak d, Anthony P.F. Turner d

a

Nanotechnology Institute, Amirkabir University of Technology, Tehran, Iran

b

Textile Engineering Department, AmirKabir University of Technology, Tehran, Iran

c

Biomaterials Group, Biomedical Engineering Department, Amirkabir University of Technology, Tehran, Iran

d

Biosensors & Bioelectronics Centre, Dept. of Physics, Chemistry and Biology (IFM), Linköping University,

Linköping, Sweden

Corresponding Author E-mail: [email protected]

ABSTRACT A sandwich-type nanostructured immunosensor based on carboxylated multi-walled carbon nanotube (CMWCNT)-embedded whiskered nanofibres (WNFs) was developed for detection of cardiac Troponin I (cTnI). WNFs were directly fabricated on glassy carbon electrodes (GCE) by removing the sacrificial component (polyethylene glycol, PEG) after electrospinning of polystyrene/CMWCNT/PEG nanocomposite nanofibres, and utilised as a transducer layer for enzyme-labelled amperometric immunoassay of cTnI. The whiskered segments of CMWCNTs were activated and utilised to immobilise anti-cTnT antibodies. It was

1

observed that the anchored CMWCNTs within the nanofibres were suitably stabilised with excellent electrochemical repeatability. A sandwich-type immuno-complex was formed between cTnI and horseradish peroxidase-conjugated anti-cTnI (HRP-anti-cTnI). The amperometric responses of the immunosensor were studied using cyclic voltammetry (CV) through an enzymatic reaction between hydrogen peroxide and HRP conjugated to the secondary antibody. The nanostructured immunosensor delivered a wide detection range for cTnI from the clinical borderline for a normal person (0.5-2 ng mL-1) to the concentration present in myocardial infarction patients ( >20 ng mL-1), with a detection limit of ~0.04 ng mL-1. It also showed good reproducibility and repeatability for three different cTnI concentration (1, 10 and 25 ng mL-1) with satisfactory relative standard deviations (RSD). Hence, the proposed nanostructured immunosensor shows potential for point-of-care testing.

Keywords: Immunosensor; Cardiac Troponin I; Carbon nanotubes; Electrospinning; Whiskered nanofibres;

1. Introduction From an epidemiological point of view, cardiovascular disease (CVD) is one of the leading causes of morbidity and mortality, and loss of disability-adjusted life years worldwide [1,2]. Early and accurate treatment of myocardial defects is very beneficial, preventing irreversible damage to the heart, especially for patients suffering acute myocardial infarction (AMI), and can significantly improve survival rates [3]. Although there are some classical symptoms of AMI such as sudden chest pain (Spontaneous Angina), shortness of breath and nausea, a considerable portion of the AMI cases do not show noticeable symptoms and are

2

“silent” [4]. For this reason, quantification of cardiac biomarkers, which are released into the circulation during the continuum of CVD or immediately after myocardial injury, enable clinicians to accurately diagnose cardiovascular damage and design an appropriate patient-care strategy [3,5]. Cardiac troponin I (cTnI) is exclusively presented in heart tissue, and hence this biomarker is deemed to be the “gold standard” for diagnosis of myocardial damage [6,7]. Normal patient levels of cTnI are in the range of 0.5-2 ng mL-1, but these increase to the range of 20-550 ng mL-1 for patients with AMI at 18 h [8]. Laboratory assay methods are accurate and reliable, but are time-consuming, required well-trained personnel, and cannot be used for pointof-care testing. In this latter case, electrochemical immunosensors offer a viable alternative due to their rapid response and ease of miniaturisation, which facilitates portability [9,10]. Nanostructured materials have offered salient benefits to improve sensitivity and reproducibility of electrochemical immunosensors [11]. During the last two decades, carbon nanotubes (CNTs) have been at the centre of intensive research for use as sensors due to their superb characteristics, which can promote electron transfer rates and catalyse the redox reaction of electroactive species [12]. Since the stability of CNTs on the electrode surface is crucial to deliver sensors with constant and repeatable responses over time, methods to stabilise nanotubes on the surface of electrodes have received attention, such as tightly compressing an adhesive carbon ink containing CNTs on the electrode surface and anchoring the CNTs within a polymer thin film [12,13]. Incorporating CNTs into nanofibrous structures is an innovative approach to stabilising the CNTs on the surface of electrodes. Nanofibrous structures also have merits for biosensing applications, such as a large surface area-to-volume ratio, high porosity and good interconnectivity [14]. Thus, nanocomposite nanofibres can combine the desirable features 3

offered by both nanofibres and CNTs for biosensing applications [15,16]. Since the electrochemical response is directly proportion to the quantity of reactive nanotubes at the electrolyte/electrode interface in aqueous solutions, nanocomposite nanofibres with more accessible nanotubes on their surface would be beneficial [17]. Therefore, in this study, a novel enzyme-labeled electrochemical immunosensor was developed based on polystyrene (PS)/carboxylated multi-walled carbon nanotube (CMWCNT) nanocomposite nanofibres with highly accessible nanotubes on their surface, i.e. “CNTwhiskered nanofibres” (WNFs), to analyse cTnI. Electrospinning was utilised to fabricate the PS/CMWCNTs/PEG nanofibres onto a glassy carbon electrode (GCE), and whiskerisation was followed by dissolving the sacrificial polymer (PEG) in water. A nanotextured surface with very high surface-to-volume ratio and stabilised CMWCNTs was obtained with enhanced electroactive area, which facilitated a more sensitive and reliable sensor response. The immunosensors based on WNFs were highly efficient for the detection of cTnI over a wide range from the clinical borderline of normal person (0.5-2 ng mL-1) to the concentration found in myocardial infarction patients ( >20 ng mL-1).

2. Experimental 2.1.Reagents Horseradish peroxidase-conjugated monoclonal anti-cTnI (anti-cTnI-HRP) and polyclonal anti-cTnI antibodies were purchased from Abcam (Cambridge, USA). The cardiac troponin I from human heart (cTnI), bovine serum albumin (BSA), equine cardiac myoglobin (MB, ≥95%), 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EDC), polystyrene (PS, Mw=190 000 g mol-1), polyethylene glycol (PEG 4000, Mw=3600–4400 g mol-1), carboxylated multiwalled carbon nanotubes (CMWCNTs, extent of labeling: >8% carboxylic acid 4

functionalised, average diameter × length: 9.5 nm × 1.5 µm), N,N,-dimethylformamide (DMF), hydrogen peroxide (H2O2), glycine and all other reagents used were of analytical grade and acquired from Sigma-Aldrich (USA). Sulfo-N-hydroxysulfosuccinimide (sulfo-NHS) was purchased from Thermo Fisher Scientific (Rockford, USA). All these materials were used as received. The Milli-Q water used in all experiments was obtained from a three-stage Millipore Milli-Q Plus 185 purification system. Phosphate buffer saline (PBS) (10 mM , pH 7.4) was prepared by dissolving 0.2 g KCl, 8.0 g NaCl, 0.24 g KH2PO4 and 1.44 g Na2HPO4 in 1000 mL of Milli-Q water, and the pH of PBS was adjusted using HCl and NaOH.

2.2. Electrospinning and fabrication of WNFs

The CMWCNTs/PS nanocomposite nanofibres were fabricated by electrospinning. A suspension of CMWCNTs (50 mg L-1) in DMF solution containing 5 wt% PEG and 5 wt% PS were sonicated with the aid of a bath-type sonicator (Elmasonic S 60 H, Singen, Germany) for different times (from 1 to 12 h) at 10 ± 5 ˚C. The effect of sonication times on the size distribution of CMWCNTs bundles was examined using a Zetasizer (Malvern Zetasizer Nano ZS90, Malvern Instruments Ltd., Worcestershire, UK), and the optimum sonication time was determined. To prepare the electrospinning solutions, the adequate amount of PS was dissolved in the suspensions with different CMWCNTs concentrations (0.5, 1, 2, 3, 4, 5, 6 wt%), ultrasonicated under the optimum sonication time, to obtain a final PS concentration in electrospinning solutions of 15 wt%. The electrospinning solutions were loaded into a 1 mL syringe fitted with a blunt ended needle (inner diameter of 0.4 mm). The syringe was placed in an injection micropump in a horizontal position and the metallic needle was connected to a high voltage DC power supply. The parameters for the electrospinning process were: applied voltage 5

14 kV, injection rate 0.3 mL h-1, needle-collector distance 16 cm. To ensure the structural integrity of the nanofibrous layer and the electrode surface, the GCE was directly utilised as a target to collect the nanofibres during the electrospinning process. Prior to use, the GCEs were polished with a polishing pad with 0.3 and 0.05 µm wet alumina slurry for at least 10 minutes, ultrasonicated for 15 min in ethanol and Milli-Q water, followed by cyclic voltammetric (CV) treatment in 0.5 M H2SO4 solution for 30 cycles to remove any residues and contaminants on the electrode surface. Since the “electric wind” phenomenon, arising from the very high field strength in the vicinity of the needle-shaped collector (i.e. GCE), can disturb the desired collection of nanofibres on the surface of GCE [18], a specific electrospinning setup was designed (as shown in Fig. S1-supplementary Information) to attain better collection of the nanofibres and improve the uniformity of nanofibrous layer on the surface of the GCE. Whiskerisation of the CMWCNTs/PS/PEG nanocomposite nanofibres was followed by dissolving the uninvolved PEG microdomains through submerging the nanofibre-coated GCE into a deionised water bath for 1 h.

2.3. Construction of the cTnI immunosensor

The WNF-based electrodes were used to construct an electrochemical immunosensor for cTnI. Activation of the carboxyl groups of the CMWCNTs was achieved by incubating 5 µL of freshly prepared coupling regent solution containing 8 mM EDC and 10 mM sulfo-NHS (1:1 volume ratio) onto the surface of electrodes for 1 h at room temperature. The activated electrodes were rinsed gently with PBS (10 mM, pH 7.4) three times. Then, 5 μL of anti-cTnI antibody solution (20 μg mL-1 in 10 mM PBS, pH 7.4) was pipetted onto the surface of the activated electrodes and incubated at 4 ˚C for 1 h, followed by washing with PBS (10 mM, pH 7.4) and 6

drying at room temperature. The density of immobilised antibodies on the surface of immunosensor (the number of antibodies per mm2 of the electrode) was subsequently determined by back-calculation from the supernatant concentration after the incubation process. A NanoDrop spectrophotometer (Nano-Drop ND-1000 spectrophotometer, Thermo Scientific; Wilmington, DE, USA) was used to estimate the antibody concentration based on absorbance at 280 nm. Finally, the electrodes were incubated with a glycine solution (50 mM in PBS) for 30 minutes to block the remaining active sites and minimise the effect of non-specific binding.

2.4. Electrochemical immunoassay

Electrochemical measurements were carried out using an IVIUM potentiostat/galvanostat in combination with IviumSoft program version 2.645 (Ivium Technologies, Eindhoven, The Netherlands). A conventional three-electrode system comprising of an Ag/AgCl electrode as reference, a platinum wire as counter and a modified GCE as working electrode, in a 10 mL static electrochemical cell was utilised for all experiments. CV experiments were carried out in a solution of PBS containing 5 mM [Fe(CN)6]-3/-4. The [Fe(CN)6]-3/-4 was applied as a redox probe to detect alterations in the electron exchange capability of the immunosensor surface at different stages of the antibody-biomarker complex formation. A heterogeneous sandwich-type electrochemical immunoassay was used for the detection of cTnI. The immunosensor was initially incubated with 3 μL of various concentrations of cTnI standard solution (range of 0.5-100 ng mL-1 in PBS) for 1 h at room temperature. The electrodes were then rinsed with PBS and incubated with 5 μL of anti-cTnI-HRP antibodies (5 μg mL-1 in PBS) for 20 min at room temperature followed by washing the electrode with PBS to remove

7

excess (loosely attached) antibodies. The amperometric responses of the immunosensors were assessed using CV with the same aforementioned solution plus 1 mM H2O2, at a scan rate 50 mV/s from −0.2 to 0.6 V, at room temperature. All measurements were carried out in triplicate.

2.5 Characterisations

A four-point probe DC conductivity analyser (Keithley 4200, Keithley Instruments, Inc., Cleveland, OH, USA) was employed to measure the electric conductivity of the nanofibrous mat. The thickness profile of the nanofibrous samples was determined using a Dektak 6M Profilometer (Veeco Instruments Inc., Woodbury, NY, USA). Scanning electron microscopy (SEM) (LEO 1550 Gemini, Zeiss, Germany) and high-resolution transmission electron microscopy (TEM) (FEI Tecnai G2 TF20 UT HRTEM, Hillsboro, Oregon, USA)) were used to study the morphology, interior structure and diameter of the nanofibres. Attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR, Vertex 70 spectrometer, Bruker, Germany) measurements were performed to elucidate the chemical structure of the nanofibrous samples. Water contact angle (WCA) measurements were performed with fresh Milli Q water (18.2 MΩ) by means of a CAM200 Optical Contact Angle Meter (KVS Instrument, Helsinki, Finland) based on the static sessile drop method, to evaluate the hydrophobicity of the nanofibrous mat.

3. Results and discussion 3.1. Characterisation of WNFs

8

Nanofibres with uniform diameter and morphology are crucial as a supporting substrate for the construction of biosensors with sufficient repeatability and reproducibility. Therefore, since the quality of nanotube dispersion within the suspension is the determinative factor in the terminal nanofibres’ morphology, this parameter was carefully optimised to obtain the desired nanofibrous mat [19]. The size distribution of the CMWCNT suspensions treated for different sonication times are shown in Fig. 1a. It should be mentioned that the results obtained using the Zetasizer are rigorously valid for spherical particles, but the results for 1D nanostructure materials reflect the hydrodynamic dimension of spherical objects [20]. Thus the results from the Zetasizer cannot be construed as the lengths of individual nanotubes, but can indicate the average size of spherical bundles produced by nanotube aggregates [21]. In this context, lower nanotube bundle size can be ascribed to better nanotube dispersion. On increasing the sonication time up to 6 h, the average nanotube bundle size was considerably decreased, while for longer sonication times the improvement in size distribution and hence dispersion was only gradual. Although long sonication time might lead to some edge-like defects along nanotube sidewalls, which can adversely affect the electrical conductivity of nanotubes, it is well-known that these defects can also improve the electrochemical and electrocatalytic reactivity of nanotubes [22,23]. Ergo, to confine the undesired defects in nanotube sidewalls and the breakage of nanotubes into smaller pieces induced by prolonged sonication, 6 h was taken as the optimum sonication interval. Since the electrochemical signal arises from the specific redox reactions at the surface of transducer, the materials utilised should be sufficiently conductive for electron transfer. The DC conductivities calculated for the nanocomposite nanofibrous mat samples with different nanotube concentrations, are shown in Fig. 1b. According to electrical percolation theory for a conductive nanocomposite, the distributed nanotubes should form a continuous network of interconnected

9

CNTs to provide a pathway for electron conduction and meet the percolation threshold [19]. It is suggested that the conductivity of nanocomposites containing nano-fillers below the percolation threshold, can be described by a non-ohmic contact mechanism. In this mechanism, the conductive nano-fillers are separated from each other through non-conductive polymer materials, so that a small fraction of electrons will have the chance of tunneling to reach the neighbouring filler. Whereas, above the percolation threshold, the electron conduction is via an ohmic contact mechanism, and occurs via tunneling or electron hopping along the physical contacts of nanotubes [24]. Since no significant alterations in the measured values of electrical conductivity were observed up to 3 wt%, it could be concluded that the concentrations of CMWCNT inside the nanofibrous samples are far from the critical concentration for electrical conduction. At a concentration of 4 wt%, a pronounced jump in the electrical conductivity of nanofibrous mat was clearly observed, and this upward trend was followed up until 5 wt%, which unambiguously confirms that the nanocomposites are close to the percolation threshold. At a nanotube concentration of 6 wt%, the variations in electrical conductivity became markedly smaller. Therefore, the desired network of interconnected nanotubes appears to be delivered by a nanofibrous mat containing 5 wt% nanotubes. Formation of a continuous physical network of nanotubes with an appropriate percolating pathway for electron conduction was also confirmed by the TEM micrograph (Fig. 1c). Moreover, it is obvious that the nanotubes are distributed all along the nanofibre axis and there are many nanotubes close to the surface of nanofibres, in a way that several of them protrude out from their surfaces. The low magnification SEM image of electrospun nanofibres containing 5 wt% CMWCNTs, with an average diameter of 328 ± 52 nm, is shown in Fig. 1d. Although the morphology of nanofibres is somewhat adversely influenced by the presence of nano-filler, it seems that these variations are inevitable.

10

Fig. 1. As mentioned before, the amphiphilic and nonionic polymer PEG was utilised for two different purposes: (1) to stabilise the dispersed nanotubes and obtain a homogeneous dispersion of CMWCNTs within the solution, and (2) as a sacrificial material to whiskerise the surface of the nanofibres. PS and PEG are partially missile polymers and in the presence of an appropriate solvent provisionally become miscible, in which upon the solvent evaporation, phase separation would be occurred. Therefore, concurrent with the electrospinning process, polymer phase separation of the PEG molecules which are not involved and did not coat the nanotubes would occur with evaporation of the solvent and lead to the formation of PEG microdomains inside the nanofibres. Whiskerisation was followed by immersing the nanofibre-covered GCEs in water for 1 h. Water acts as a selective solvent to remove the accessible PEG microdomains, which can lead to formation of nanofibres with rougher surfaces and more nanotubes appearing on the surface of nanofibres in the form of whiskers. Since the diameter of the nanotube is much smaller than nanofibres’ diameter, high magnification SEM images (Fig. 1e) were taken to show that the segments of the nanotube network were extruded out and appeared on the nanofibre surfaces. Investigations of the surface wettability of WNFs mats (Fig. 2a) revealed that the water contact angle (WCA) was reduced with increasing nanotube content. Increasing the surface hydrophilicity might be ascribed to the presence of more CMWCNTs with hydrophilic functional groups (carboxyl groups) as well as the formation of protrusions and grooves on the nanofibre’s surface, arising from the presence of nanotubes on the surface and the gaps left after PEG microdomains were removed. Fig. 2b shows the ATR-FTIR spectra of the PS nanofibres and the PS/CMWCNT/PEG nanocomposite nanofibres before and after the whiskerisation process. For the PS nanofibrous sample, the spectrum exhibits the expected peaks, i.e., C-H stretching 11

vibration of the aromatic ring at 3026, 3061 and 3082 cm-1, the aromatic C=C stretching vibration peaks at 1600 and 1583 cm-1, C-H bending vibration at 1452 and 1493 cm-1, and the asymmetric and symmetric stretching vibrations of CH2 at 2851 and 2922 cm-1, respectively. As can be seen from Fig. 2b, almost all the characteristic peaks of PS can be found in the PS/CMWCNTs/PEG nanofibres and WNFs. The FTIR spectra of the nanocomposite nanofibres also showed the characteristic absorption peaks of PEG and CMWCNTs. The peaks at 1101 and 843 cm-1 are assigned to C-O-C starching vibrations of PEG. Moreover, the peaks at 1730 cm-1 and 962 cm−1 are attributed to the C=O stretching vibration and out-of-plane O–H bending of the carboxyl groups of CMWCNTs, respectively. The relatively higher amplitude of the characteristic peak at 1730 cm-1 of WNFs sample could be ascribed to the increased nanotube content in the nanofibres, arising from leaching the accessible PEG microdomains. Fig. 2.

3.2. Electrochemical characterisation of the modified electrodes

Anchoring the CMWCNTs within the polymer nanofibre structure was designed not only to stabilise the nanotubes as transducer on the surface of the GCEs, but also to exploit the merits of nanotubes and nanofibres for biosensing. The stepwise fabrication of the immunosensor as well as the sandwich-type electrochemical immunoassay format are schematically represented in Fig. 3. Since in the designed sandwich-type immuno-complex, the HRP active site is relatively far from the electrode surface, and it is also surrounded by the α-helical structure of enzyme, direct electrical communication between the enzyme active site and the electrode surface is almost impossible. Therefore, use of the reversible redox probe [Fe(CN)6]3-/4- is crucial to

12

facilitate the electron transfer and amplify the signal output. The EDC/sulfo-NHS mixture assisted the conversion of carboxylic acid groups to amino-reactive esters of the sulfo-NHS. Although, many researchers have previously used NHS to immobilise proteins on carboxylic acid groups, here, sulfo-NHS was employed because of: (i) its higher solubility in water; and (ii) its ability to improve the amount of immobilised proteins, owing to the electrostatic attractions between the negatively charged sulfonic acid groups of sulfo-NHS and proteins [25]. Fig. 3.

The cyclic voltammograms of [Fe(CN)6]3−/4− as a redox probe for bare GCE, and WNFsbased nanotextured electrodes, before and after the immobilisation of bioreceptors and blocking agent ( anti-cTnI antibodies and glycine, respectively), are shown in Fig. 4a. The alteration in anodic and cathodic peak currents and peak-to-peak separation (ΔEp) values for different modified electrodes are theoretically associated with the electron transfer rate constant (Ks, which is inversely correlated to the electron transfer resistance) [26]. The control experiment with GCE shows a pair of well-defined reversible redox peaks with ΔEp of 85 mV, indicating that the GCE has a clean and active surface. After directly electrospinning and whiskerising the nanofibres on the surface of GCE, the anodic and cathodic currents increased and the ΔEp also increased to 115 mV. The increased peak currents can be attributed to the presence of numerous segments from the carbon nanotubes network on the surface of conductive nanofibres, which exhibit electro-catalytic activity due to their inherent features such as high electrical conductivity and surface electroactive sites and some topological defects in lattices [23,27]. Whereas, the observed increase in ΔEp might be attributed to the presence of the nonconductive polymeric matrix of the nanofibre’s backbone which can hinder the electron transfer rate in some regions.

13

The electroactive surface area of the WNFs/GCE was approximated according to the Randles– Sevcik equation [28,29]: (

)







(1)

Where, Ip is the peak current value (ampere, A), n is the number of electrons participating in the redox reaction (1 in this case), A is the electroactive surface area (cm2), D=7.60×10-6 cm2 s-1, which is the diffusion coefficient of [Fe(CN)6]3- in the solution, C corresponds to the concentration of the redox probe (mol cm-3) and v is the scan rate (V s-1). Since the D, n, v and C parameters are all unchanged, the recorded increase in the redox peak currents can be ascribed to higher electroactive surface area of the modified GCE caused by the high surface area of the nanofibres. The electroactive surface area of the bare GCE and modified electrode were calculated to be approximately 4.31±0.02 and 4.79±0.03 mm2, respectively. The measurements were made in triplicate and the relative standard deviations (RSD) for bare GCE and modified electrode were calculated as 0.36 and 0.71%, respectively. The measured value for the modified electrode reflects the electroactive area of accessible nanotubes segments on the surface of the WNFs. After immobilisation of anti-cTnI antibodies and blocking agent (glycine) on the activated carboxyl groups of the CMWCNTs, the redox current decreased due to the insulating nature of the antibodies and glycine, which act as barriers for electron transport. Moreover, the electrostatic repulsions between the [Fe(CN)6] 4− and the negatively charged glycine might also reduce the redox peaks [13]. The stability of WNFs/GCE and the glycine/anti-cTnI/WNFs/GCE (immunosensor) were determined by 20 consecutive voltammetric cycles (as is shown in Fig. S2supplementary information). The redox peaks remained constant through the scanning cycles, 14

and their RSDs were 0.1% and 0.24% for WNFs/GCE and the immunosensor, respectively. These results clearly affirm the good stability of WNFs and the immobilization of the antibodies on the electrode surface. A linearly dependency was observed between the anodic and cathodic peak current values and the square root of the scan rate over the range of 20 – 200 mV s-1 (Fig. 4b), indicating that the immunosensor interface was diffusion controlled. The following linear regression equations then were calculated: Ipa=4.3472(scan rate)½ + 4.1204 (R2=0.9993) and Ipc=4.6322(scan rate)½ – 8.1039 (R2=0.9981). Moreover, the anodic and cathodic peak potentials shifted to more positive and negative values, respectively, as the scan rate was increased, suggesting typical quasi-reversible electron transfer kinetics. Fig. 4. To approximate the density of anti-cTnI antibodies immobilised on the surface of WNFs, a calibration curve was generated from the UV absorbance at 280 nm of standard anti-cTnI antibody solutions (as is shown in Fig. S3- supplementary information, anti-cTnI concentration range of 5-900 μg mL-1, R2 = 0.997). A significant reduction in the concentration of anti-cTnI antibodies in the supernatant (15.24 μg mL-1) after incubation for 1 h on the surface of EDC/sulfo-NHS exposed WNFs/GCE (activated electrode) revealed the successful immobilisation of antibodies on the WNFs/GCE. Whereas, in the case of EDC/sulfo-NHS unexposed WNFs/GCE (non-activated electrode), little reduction in the supernatant antibody concentration (19.23 μg mL-1) was observed, which might be due to the confinement of antibodies due to the nanofibre porosity. Considering the molecular weight of anti-cTnI antibodies (Mw≈24 kDa), the Avogadro’s number and the surface area of GCE as 3 mm2, the density of immobilised antibodies on the surface of immunosensor was estimated to be 15

approximately 1.67 ×1011 molecules/mm2. Therefore, both of the CV and the spectrophotometric results confirmed the immobilisation of antibodies on the surface of the WNFs/GCE.

3.3. Optimisation of conditions for the immunoassay

To obtain the maximum diagnostic sensitivity and specificity, the effect of pH, secondary antibody concentration and its incubation time, and H2O2 concentration on the peak currents of CV measurements were investigated. It is well known that pH can influence not only the electrochemical behavior of immunosensors, but also the bioactivity of the conjugated enzymes of the secondary antibodies. After the incubation of immunosensors with 3 μL of cTnI standard solution (5 ng mL-1) for 60 minutes, the electrodes were incubated with 5 μL of secondary antibody (10 μg mL-1) for 30 minutes. CVs were performed in 10 mM PBS (pH from 5.5 to 8) containing 5 mM [Fe(CN)6]-3/-4 and 1 mM H2O2. The highest cathodic current response was obtained at pH 7.0 (Fig. 5a), which is proportional to the maximum bioactivity of enzymatic catalysis, and was chosen as the optimum pH for all further experiments [11,12,30]. The secondary antibody (HRP-labeled antibody) concentration and incubation time have a critical influence on the analytical performance of the immunosensor, owing to the dependence of the cathodic peak current on the density of secondary antibodies attached to the cTnI. To gain an optimal response with minimal anti-cTnI-HRP antibodies, different concentrations of antibody (from 1 to 10 μg mL-1) were evaluated. Similar concentrations and incubation conditions were used to those described in the previous step, and the CVs were obtained. The cathodic peak current increased gradually with increasing the secondary antibody concentration (as shown in Fig. 5b) up to 5 μg mL-1, and then increased slightly with further increase of the

16

secondary antibody concentration. Thus, the anti-cTnI-HRP concentration of 5 μg mL-1 was chosen as the optimal concentration for later experiments. Different incubation times (from 10 to 40 minutes) were also evaluated. Fig. 5c represents the catalytic reduction effect of HRP enzymes on the surface of the immunosensor by measuring the variation in cathodic peak currents. The response curve shows a reduction peak plateau at 20 minutes. Finally, the effect of hydrogen peroxide concentrations (from 0.2 to 2 mM) on the catalytic performance of the immunosensor was evaluated. As can be seen in Fig. 5d, the cathodic peak currents were proportional to H2O2 concentration up to 1 mM, followed by a plateau. Therefore, the optimal H2O2 concentration was taken as 1 mM. In summary, optimal conditions for subsequent measurements were taken as: pH 7.0, secondary antibody concentration of 5 μg mL-1 with an incubation time of 20 minutes, and H2O2 concentration of 1 mM. Fig. 5.

3.4. Analytical response to the cTnI

The capability of the designed immunosensor to amplify the catalytic response towards H2O2 is shown in Fig. 6a. As expected for a sandwich-type immunoassay mechanism, by incubating the immunosensor with cTnI biomarker, and subsequently with secondary antibody, the reduction peak currents were enhanced in proportion to the biomarker concentration, indicating the enzymatic reduction of H2O2. A calibration curve for the immunosensor was plotted by measuring the reduction current in response to various concentrations of cTnI standard solutions (0.5 to 100 ng mL-1). As shown in Fig. 6a, with incremental increase in the cTnI biomarker concentration, the reduction peak currents rose proportionally, showing a good linear

17

correlation to the logarithm of cTnI concentration with a linear regression equation y = 12.175x + 4.4883 and a correlation coefficient of 0.989. The detection limit (LoD) for cTnI, defined as the lowest assayable concentration differentiated from the signal of zero concentration (buffer solution) by 3SD/m, was calculated to be ~0.04 ng mL-1, which is considerably lower than clinical analytical systems such as ELISA (~5 ng mL-1) [31,32]. The immunosensor exhibited a wide diagnostic range from the clinical borderline of a normal person (0.5-2 ng mL-1) to the concentration of cTnI found in myocardial infarction patients ( >20 ng mL-1). The obtained results for detection limit and linear range confirm that the designed immunosensor is comparable or even better than some recent results reported in literature, which are summarised in Table 1. Table 1

Given the other advantages of 1D nanostructured materials, such as very good flexibility and bendability, the immunosensor can be considered as a promising candidate for future investigations and specialist applications (e.g. wearable sensors). Fig. 6.

3.5. Reproducibility, repeatability and stability

To evaluate the reproducibility and repeatability of the immunosensors, their inter- and intra- assay precisions were assessed, respectively. The inter-assay precision of the immunosensor for cTnI detection was measured by using 5 different electrodes with three cTnI concentration levels (1, 10 and 25 ng mL-1). The RSDs of inter-assay were 3.54, 3.26, and 3.17% for 1, 10, and 25 ng mL-1, respectively, indicating acceptable reproducibility. The intra-assay 18

precision was assessed using 5 replicates at three different concentration levels with the same electrode. The RSDs of intra-assays were 0.63, 0.84, and 1.14% For 1, 10, and 25 ng mL-1, respectively, illustrating the repeatability of the immunosensor. The sensor also possessed satisfactory storage stability with 89.47% of the initial cathodic peak current remaining after the storage for 10 days in pH 7.4 PBS at 4 °C.

3.6. Interference study To appraise the selectivity and specificity of the designed immunosensor, BSA and MB were employed as a model protein resembling human blood plasma proteins and a non-specific cardiac biomarker, respectively. Fig. 6b shows the variations in cathodic peak currents after adding different concentrations of the interfering compounds, providing evidence for the good specificity of the immunosensor for cTnI detection and the lack of binding between the anti-cTnI antibodies and the two non-specific compounds. The observed decrease in cathodic peak currents with increasing both interfering compounds concentrations is thought to be due to physical entrapment of protein molecules within the WNFs arising from the very high surface area and porosities of the nanofibrous structure. The results obtained confirm that the designed immunosensor possesses remarkable selectivity and specificity for detection of cTnI and provide evidence that there is no significant adsorption after incubating with the non-target proteins. Overall, these results indicated that the proposed immunosensor is selective, sensitive and reliable enough to be compared with routine immunoassay techniques.

19

4. Conclusions A highly sensitive nanostructured electrochemical immunosensor based on WNFs as the transducer was developed for cTnI as a basis for a rapid and accurate diagnostic (point-of-care testing) for cardiovascular diseases. The formation of a sandwich type (anti-cTnI/cTnI/HRP-anticTnI) immuno-complex delivered excellent amperometric responses over a wide range of cTnI concentrations. The principal feature of the developed immunosensor is the incorporation of CNTs within the nanofibres in the form of whiskers to deliver high surface area nanofibres simultaneously with the superb merits of CNTs for biosensing applications. The comparatively low detection limit, wide detection range, good selectivity, acceptable reproducibility, repeatability, and stability provides a promising candidate for future point-of-care diagnostics.

Acknowledgment BR would like to gratefully acknowledge the Ministry of Science, Research and Technology of Iran for financial support that allowed him to work at Linköping University. LU thanks the European Commission for a Marie Curie Fellowship (PIEF-GA-2013-629251).

Notes The authors declare that they have no conflict of interest.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at:

20

References [1]

M. Hasanzadeh, N. Shadjou, M. Eskandani, M. de la Guardia, E. Omidinia, Electrochemical nanoimmunosensing of effective cardiac biomarkers for acute myocardial infarction, TrAC - Trends Anal. Chem. 49 (2013) 20–30. doi:10.1016/j.trac.2013.04.009.

[2]

A.A. Shumkov, E. V. Suprun, S.Z. Shatinina, A. V. Lisitsa, V. V. Shumyantseva, A.I. Archakov, Gold and Silver Nanoparticles for Electrochemical Detection of Cardiac Troponin I Based on Stripping Voltammetry, Bionanoscience. 3 (2013) 216–222. doi:10.1007/s12668-013-0090-9.

[3]

B. Rezaei, M. Ghani, A.M. Shoushtari, M. Rabiee, Electrochemical biosensors based on nanofibres for cardiac biomarker detection: A comprehensive review, Biosens. Bioelectron. 78 (2016) 513–523. doi:10.1016/j.bios.2015.11.083.

[4]

E. V. Suprun, A.A. Saveliev, G.A. Evtugyn, A. V. Lisitsa, T. V. Bulko, V. V. Shumyantseva, A.I. Archakov, Electrochemical approach for acute myocardial infarction diagnosis based on direct antibodies-free analysis of human blood plasma, Biosens. Bioelectron. 33 (2012) 158–164. doi:10.1016/j.bios.2011.12.045.

[5]

A. Qureshi, Y. Gurbuz, J.H. Niazi, Biosensors for cardiac biomarkers detection: A review, Sensors Actuators, B Chem. 171–172 (2012) 62–76. doi:10.1016/j.snb.2012.05.077.

[6]

L. Tang, J. Casas, Quantification of cardiac biomarkers using label-free and multiplexed gold nanorod bioprobes for myocardial infarction diagnosis, Biosens. Bioelectron. 61 (2014) 70–75. doi:10.1016/j.bios.2014.04.043.

[7]

H. Guo, N. He, S. Ge, D. Yang, J. Zhang, Molecular sieves materials modified carbon paste electrodes for the determination of cardiac troponin i by anodic stripping voltammetry, Microporous Mesoporous Mater. 85 (2005) 89–95. doi:10.1016/j.micromeso.2005.06.022. 21

[8]

S. Ko, B. Kim, S.S. Jo, S.Y. Oh, J.K. Park, Electrochemical detection of cardiac troponin I using a microchip with the surface-functionalized poly(dimethylsiloxane) channel, Biosens. Bioelectron. 23 (2007) 51–59. doi:10.1016/j.bios.2007.03.013.

[9]

E. V. Suprun, A.L. Shilovskaya, A. V. Lisitsa, T. V. Bulko, V. V. Shumyantseva, A.I. Archakov, Electrochemical immunosensor based on metal nanoparticles for cardiac myoglobin detection in human blood plasma, Electroanalysis. 23 (2011) 1051–1057. doi:10.1002/elan.201000668.

[10]

D. Brondani, J. V Piovesan, E. Westphal, H. Gallardo, R.A. Fireman Dutra, A. Spinelli, I.C. Vieira, A label-free electrochemical immunosensor based on an ionic organic molecule and chitosan-stabilized gold nanoparticles for the detection of cardiac troponin T, Analyst. 139 (2014) 5200–5208. doi:10.1039/C4AN00993B.

[11]

B.V.M. Silva, I.T. Cavalcanti, A.B. Mattos, P. Moura, M.D.P.T. Sotomayor, R.F. Dutra, Disposable immunosensor for human cardiac troponin T based on streptavidin-microsphere modified screen-printed electrode, Biosens. Bioelectron. 26 (2010) 1062–1067. doi:10.1016/j.bios.2010.08.051.

[12]

S.L.R. Gomes-Filho, A.C.M.S. Dias, M.M.S. Silva, B.V.M. Silva, R.F. Dutra, A carbon nanotubebased electrochemical immunosensor for cardiac troponin T, Microchem. J. 109 (2013) 10–15. doi:10.1016/j.microc.2012.05.033.

[13]

B. Silva, I. Cavalcanti, M. Silva, R. Dutra, A carbon nanotube screen-printed electrode for labelfree detection of the human cardiac troponin T, Talanta. 117 (2013) 431–437. http://www.sciencedirect.com/science/article/pii/S0039914013007261.

[14]

S. Sadir, M.P. Prabhakaran, D.H.B. Wicaksono, S. Ramakrishna, Fiber based enzyme-linked immunosorbent assay for C-reactive protein, Sensors Actuators B Chem. 205 (2014) 50–60. doi:10.1016/j.snb.2014.08.051.

22

[15]

L.S. Wan, B.B. Ke, Z.K. Xu, Electrospun nanofibrous membranes filled with carbon nanotubes for redox enzyme immobilization, Enzyme Microb. Technol. 42 (2008) 332–339. doi:10.1016/j.enzmictec.2007.10.014.

[16]

Z.G. Wang, Z.K. Xu, L.S. Wan, J. Wu, C. Innocent, P. Seta, Nanofibrous membranes containing carbon nanotubes: Electrospun for redox enzyme immobilization, Macromol. Rapid Commun. 27 (2006) 516–521. doi:10.1002/marc.200500885.

[17]

S.H. Park, H.R. Jung, B.K. Kim, W.J. Lee, MWCNT/mesoporous carbon nanofibers composites prepared by electrospinning and silica template as counter electrodes for dye-sensitized solar cells, J. Photochem. Photobiol. a-Chemistry. 246 (2012) 45–49. doi:DOI 10.1016/j.jphotochem.2012.07.013.

[18]

D. Lukáš, A. Sarkar, L. Martinová, K. Vodsed’álková, D. Lubasová, J. Chaloupek, P. Pokorný, P. Mikeš, J. Chvojka, M. Komárek, Physical principles of electrospinning (Electrospinning as a nano-scale technology of the twenty-first century), Text. Prog. 41 (2009) 59–140. doi:10.1080/00405160902904641.

[19]

S. Mazinani, A. Ajji, C. Dubois, Morphology, structure and properties of conductive PS/CNT nanocomposite electrospun mat, Polymer (Guildf). 50 (2009) 3329–3342. doi:10.1016/j.polymer.2009.04.070.

[20]

E.J. Weydemeyer, A.J. Sawdon, C.-A. Peng, Controlled cutting and hydroxyl functionalization of carbon nanotubes through autoclaving and sonication in hydrogen peroxide SI, Chem. Commun. 51 (2015) 5939–5942. doi:10.1039/C5CC01115A.

[21]

D. Lin, N. Liu, K. Yang, L. Zhu, Y. Xu, B. Xing, The effect of ionic strength and pH on the stability of tannic acid-facilitated carbon nanotube suspensions, Carbon N. Y. 47 (2009) 2875– 2882. doi:10.1016/j.carbon.2009.06.036.

23

[22]

A.T. Lawal, Synthesis and utilization of carbon nanotubes for fabrication of electrochemical biosensors, Mater. Res. Bull. 73 (2016) 308–350. doi:10.1016/j.materresbull.2015.08.037.

[23]

K.M. Manesh, H.T. Kim, P. Santhosh, A.I. Gopalan, K.P. Lee, A novel glucose biosensor based on immobilization of glucose oxidase into multiwall carbon nanotubes-polyelectrolyte-loaded electrospun nanofibrous membrane, Biosens. Bioelectron. 23 (2008) 771–779. doi:10.1016/j.bios.2007.08.016.

[24]

S. Kumar, B. Lively, L.L. Sun, B. Li, W.H. Zhong, Highly dispersed and electrically conductive polycarbonate/oxidized carbon nanofiber composites for electrostatic dissipation applications, Carbon N. Y. 48 (2010) 3846–3857. doi:10.1016/j.carbon.2010.06.050.

[25]

Z. Pei, H. Anderson, A. Myrskog, G. Dunér, B. Ingemarsson, T. Aastrup, Optimizing immobilization on two-dimensional carboxyl surface: pH dependence of antibody orientation and antigen binding capacity, Anal. Biochem. 398 (2010) 161–168. doi:10.1016/j.ab.2009.11.038.

[26]

S. Singal, A.K. Srivastava, B. Gahtori, Rajesh, Immunoassay for troponin I using a glassy carbon electrode modified with a hybrid film consisting of graphene and multiwalled carbon nanotubes and decorated with platinum nanoparticles, Microchim. Acta. 183 (2016) 1375–1384. doi:10.1007/s00604-016-1759-x.

[27]

J. Li, A. Cassell, L. Delzeit, J. Han, M. Meyyappan, Novel Three-Dimensional Electrodes : Electrochemical Properties of Carbon Nanotube Ensembles, J. Phys. Chem. B. 106 (2002) 9299– 9305. doi:10.1021/jp021201n.

[28]

D.G.A. Cabral, E.C.S. Lima, P. Moura, R.F. Dutra, A label-free electrochemical immunosensor for hepatitis B based on hyaluronic acid-carbon nanotube hybrid film, Talanta. 148 (2016) 209– 215. doi:10.1016/j.talanta.2015.10.083.

[29]

S. Singal, A.K. Srivastava, S. Dhakate, A.M. Biradar, R. Rajesh, Electroactive graphene-multi-

24

walled carbon nanotube hybrid supported impedimetric immunosensor for the detection of human cardiac troponin-I, RSC Adv. 5 (2015) 74994–75003. doi:10.1039/C5RA15449A. [30]

T.A. Freitas, A.B. Mattos, B.V.M. Silva, R.F. Dutra, Amino-Functionalization of Carbon Nanotubes by Using a Factorial Design: Human Cardiac Troponin T Immunosensing Application, Biomed Res. Int. 2014 (2014). doi:10.1155/2014/929786.

[31]

A. Periyakaruppan, R.P. Gandhiraman, M. Meyyappan, J.E. Koehne, Label-Free Detection of Cardiac Troponin-I Using Carbon Nano fi ber Based Nanoelectrode Arrays, Anal. Chem. 85 (2013) 3858–3863. doi:10.1021/ac302801z.

[32]

V. Bhalla, S. Carrara, P. Sharma, Y. Nangia, C. Raman Suri, Gold nanoparticles mediated labelfree capacitance detection of cardiac troponin i, Sensors Actuators, B Chem. 161 (2012) 761–768. doi:10.1016/j.snb.2011.11.029.

[33]

A.J. Saleh Ahammad, Y.H. Choi, K. Koh, J.H. Kim, J.J. Lee, M. Lee, Electrochemical detection of cardiac biomarker troponin I at gold nanoparticle-modified ITO electrode by using open circuit potential, Int. J. Electrochem. Sci. 6 (2011) 1906–1916.

[34]

F. Zhou, M. Lu, W. Wang, Z.P. Bian, J.R. Zhang, J.J. Zhu, Electrochemical immunosensor for simultaneous detection of dual cardiac markers based on a poly(dimethylsiloxane)-gold nanoparticles composite microfluidic chip: A proof of principle, Clin. Chem. 56 (2010) 1701– 1707. doi:10.1373/clinchem.2010.147256.

[35]

G. Liu, M. Qi, Y. Zhang, C. Cao, E.M. Goldys, Nanocomposites of gold nanoparticles and graphene oxide towards an stable label-free electrochemical immunosensor for detection of cardiac marker troponin-I, Anal. Chim. Acta. 909 (2016) 1–8. doi:10.1016/j.aca.2015.12.023.

25

Table 1 Comparison of composition, detection limit and linear range of different modified working electrodes for electrochemical detection of cTnI.

Electrode type Screen printed electrode (SPE) Indium tin oxide (ITO) GCE Modified carbon paste electrode (MCPE) Poly(dimethylsiloxane) (PDMS)-Gold nanoparticles composite Microfluidic Chip Silicon wafer Screen printed electrode (SPE) Interdigitated array (IDA) gold microelectrode chip and a PDMS channel GCE GCE

Detection limit ng mL-1

Linear range ng mL-1

Refs

0.2

0.2-12.5

[32]

1 0.001

1-500 0.001-10

[33] [26]

Colloidal gold (cAu)

0.2

0.5-5

[7]

Quantum dots (QDs)

0.004

0.01-50

[34]

0.2

5-100

[31]

0.1

0.1-32

[2]

0.148

0.2-10000

[8]

0.05

0.05-3

[35]

0.04

0.5-100

this work

Modification Gold nanoparticles (GNPs) GNPs G-MWCNT/PtNPs

Vertically aligned carbon nanofibres (VACNFs) Gold/Silver nanoparticles Silane films and biomolecular layers Nanocomposites of gold nanoparticles and graphene oxide WNFs

26

Fig. 1. (a) Average size and size distribution of dilute nanotube suspensions with different ultrasonication times (the x-axis is logarithmic), (b) dependence of the electrical conductivity (σ) of the WNFs on the nanotube concentrations in electrospinning solution, (c) TEM image of the PS/CMWCNTs/PEG as-spun nanofibres, and (d) SEM micrograph of PS/CMWCNTs/PEG as-spun nanofibres, and (e) high magnification SEM images of WNFs from

different sections of sample (e1-e4).

27

Fig. 2. (a) Measurement of the contact angles of water droplets on the WNFs mat with different nanotube contents, and (b) ATR-FTIR spectra of PS nanofibres, PS/CMWCNT/PEG nanocomposite nanofibres and WNFs.

28

Fig. 3. Schematic view of the stepwise fabrication of immunosensor and the mechanism of detection.

29

Fig. 4. (a) CV profile of the immunosensor at each step of the fabrication (A) bare GCE, (B) WNFs/GCE and (C) immunosensor, and (b) CV profile of the immunosensor at different scan rates. Effect of different scan rate (20-200 mV s-1) on the cathodic and anodic currents and plots of the currents as a function of square root of the scan rate confirm that the system is diffusion controlled.

30

Fig. 5. Influence of (a) pH values, (b) secondary antibody concentration, (c) the incubation time of secondary antibody and (d) H2O2 concentration, on the cathodic peak current of the immunosensor.

31

Fig. 6. (a) Calibration curve for cTnI over the 0.5 to 100 ng mL-1 range under the optimal conditions (pH 7.0 PBS containing 1 mM H2O2, the inset picture shows the cyclic voltammograms: top to bottom from low to high concentrations), and (b) cathodic current changes of the immunosensor toward different concentration of cTnI as specific target, and MB and BSA as non-specific targets.

32

Graphical abstract

Highlights    

An immunosensor based on CNT-whiskered nanofibres is developed for cTnI detection. Whiskerisation is done to exploit the unique merits of nanotubes and nanofibres. Sandwich-type immunoreactions is identified using an amperometric technique. The sensor is selective and reliable enough to employ for point-of-care testing.

33