Tailored carbon nanotube immunosensors for the detection of microbial contamination.

Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

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

Tailored carbon nanotube immunosensors for the detection of microbial contamination B. Prieto-Simón a,n, N.M. Bandaru b, C. Saint c, N.H. Voelcker a,n a

Mawson Institute, University of South Australia, Australia School of Chemical and Physical Sciences, Flinders University, Australia c SA Water Centre for Water Management and Re-use, University of South Australia, Australia b

art ic l e i nf o

a b s t r a c t

Article history: Received 22 June 2014 Received in revised form 29 September 2014 Accepted 30 September 2014

The use of carbon nanotubes (CNTs) as building blocks in the design of electrochemical biosensors has been attracting attention over the last few years, mainly due to their high electrical conductivity and large surface area. Here, we present two approaches based on tailored single-walled CNTs (SWCNTs) architectures to develop immunosensors for the bacteriophage MS2, a virus often detected in sewageimpacted water supplies. In the first approach, SWCNTs were used in the bottom-up design of sensors as antibody immobilization support. Carboxy-functionalised SWCNTs were covalently tethered onto gold electrodes via carbodiimide coupling to cysteamine-modified gold electrodes. These SWCNTs were hydrazide functionalized by electrochemical grafting of diazonium salts. Site-oriented immobilization of antibodies was then carried out through hydrazone bond formation. Results showed microarray electrode behavior, greatly improving the signal-to-noise ratio. Excellent sensitivity and limit of detection (9.3 pfu/mL and 9.8 pfu/mL in buffer and in river water, respectively) were achieved, due to the combination of the SWCNTs′ ability to promote electron transfer reactions with electroactive species at low overpotentials and their high surface-to-volume ratio providing a favorable environment to immobilize biomolecules. In the second approach, SWCNTs were decorated with iron oxide nanoparticles. Diazonium salts were electrochemically grafted on iron-oxide-nanoparticle-decorated SWCNTs to functionalize them with hydrazide groups that facilitate site-directed immobilization of antibodies via hydrazone coupling. These magnetic immunocarriers facilitated MS2 separation and concentration on an electrode surface. This approach minimized non-specific adsorptions and matrix effects and allowed low limits of detection (12 pfu/mL and 39 pfu/mL in buffer and in river water, respectively) that could be further decreased by incubating the magnetic immunocarriers with larger volumes of sample. Significantly, both approaches permitted the detection of MS2 to levels regularly encountered in sewage-impacted environments. & 2014 Elsevier B.V. All rights reserved.

Keywords: Electrochemical immunoassay Carbon nanotubes Iron oxide nanoparticles Antibody immobilization Virus detection

1. Introduction CNTs have been used as promising nanomaterials in electrochemical biosensor development due to their unique chemical, electronic and mechanical properties (Feng et al., 2008). Incorporation of CNTs into electrode design has been achieved by several methods taking advantage of the strength, size, chemical stability and, more importantly, good electronic conductivity of CNTs (Katz and Willner, 2004). CNT-modified electrodes enable the catalysis of redox reactions by promoting the electron transfer at low overpotentials between electroactive species in solution n

Corresponding authors. E-mail addresses: [email protected], [email protected] (B. Prieto-Simón), [email protected] (N.H. Voelcker).

and the electrode (Merkoçi et al., 2005). This role to mediate electron transfer reactions helps to minimize electrode fouling effects (Thomas et al., 2013). Moreover, as a result of the enhanced electron transfer and the large surface area provided by CNTs, CNTbased sensors are usually characterized by high sensitivity values (Wang, 2004). CNTs are typically functionalized via the carboxyl group on their tips to immobilize various bioreceptors, such as antibodies or enzymes (Liu et al., 2005; Yu et al., 2006). Several approaches have been used to develop electrochemical biosensors aiming to incorporate the electrical properties and biocompatible nanostructure of CNTs. CNT-coated electrodes, CNT-composite electrodes, CNT electrodes based on layer-by-layer protocols and vertically aligned CNT-modified electrodes have been prepared. CNT-based electrodes can be easily prepared by drop coating electrodes with CNT suspensions or by electrostatically assembling CNTs on a positively charged surface. However,

http://dx.doi.org/10.1016/j.bios.2014.09.089 0956-5663/& 2014 Elsevier B.V. All rights reserved.

Please cite this article as: Prieto-Simón, B., et al., Biosensors and Bioelectronics (2014), http://dx.doi.org/10.1016/j.bios.2014.09.089i

2

B. Prieto-Simón et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

the ability of CNTs to promote electron transfer strongly depends on the ability to obtain stable CNT suspensions and, thus, homogeneous coatings and CNT multilayer films, respectively (Wang et al., 2003; Kumar et al., 2012). CNT-composite electrodes also rely on the homogeneous dispersion of CNTs within the used binder, as well as on the ability to avoid impairing of the electrocatalytic properties of CNTs by the interaction with the binder (Valentini et al., 2003). Vertically aligned CNT-modified electrodes have received much attention over the last few years (Constantopoulos et al., 2010). The small diameter, high aspect ratio and good conductivity of CNTs are responsible for the high sensitivity achieved (Gooding et al., 2003; Yu et al., 2006), as well as for the excellent performance of electrochemical devices based on vertically-aligned CNTs. This is due to a faster charge transfer rate (e.g., between CNTs and redox centers of immobilized molecules) compared to randomly dispersed CNTs. Vertically-aligned CNTs can be prepared by a physical method, based on direct growth of CNTs using template-assisted or template-free synthesis (Huang et al., 2007); or by a chemical method (Diao and Liu, 2010), based on the assembly of modified CNTs on certain substrates. Chemical assembly has been performed on different substrates via covalent bonds (Shearer et al., 2010), metal-assisted chelation and electrostatic interactions (Chattopadhyay et al., 2001). Magnetic particles have been widely used in the development of biosensors (Zhang and Zhou, 2012). They show a large biocompatible surface area for biomolecule immobilization, providing a large number of binding sites for biorecognition events. Moreover, they can be dispersed in solution, increasing the efficiency of the interactions and lowering the analysis time. Their magnetic properties make them useful for analyte separation, purification and concentration from complex samples, avoiding the need for sample pretreatment. Their capture using magnets facilitates and improves the washing protocol, minimizing matrix effects. These features result in improvements in the sensitivity and limit of detection (Xu and Wang, 2012). As an additional advantage of their use, regeneration of the biosensor surface is easily performed by switching off the magnet used to capture the modified magnetic particles. Incorporation of nanobiotechnology in the design of novel biosensors is essential to maximize their analytical performance (Campàs et al., 2012). Here we present two assemblies of SWCNTs in 3D platforms aiming to detect the bacteriophage MS2 as an indicator of microbiological contamination in water supplies. Both immunoplatforms were used to detect MS2 phage through differential pulse voltammetric measurements after performing an indirect sandwich immunoassay. The use of MS2 as a proof of concept is particularly relevant as it can be used not only as a surrogate for enteric viruses but also as an indicator of the presence of Escherichia coli, its natural host. MS2 has been successfully used as a surrogate to examine the effectiveness of water treatment processes as it shows characteristics that are highly similar to enteric viruses (Brehant et al., 2010) and its detection can be performed by simple and inexpensive methods (Leclerc et al., 2000; Petty et al., 2007). In addition, its presence has been well documented in situations where E. coli (the host) occurs in significant numbers, such as raw sewage and water courses impacted by raw sewage (Keegan et al., 2009; Li et al., 2012). Therefore, we have chosen the F-RNA coliphage MS2 to develop a proof-of-concept biosensor that could be further tailored to detect other viruses and bacteria or immunoreactive analyte species in general. Based on our previous work (PrietoSimón et al., 2014), we aim at exploiting the advantages of the incorporation of SWCNTs in sensing devices as a means to improve the sensitivity and limit of detection, while minimizing nonspecific adsorption events. SWCNTs are expected to work as conductive linkers between gold electrodes and the antibodies

used as bioreceptors. Site-directed chemistry to control the orientation of the immobilized antibodies onto the biosensor surface has a strong impact on the final analytical performance (Trilling et al., 2013). Therefore, antibodies have been immobilized through their Fc moieties as a way to maximize the access to their binding sites. Initially SWCNTs were covalently anchored on gold electrodes via amide bond formation aiming to improve the electron-transfer rate along the tubes. SWCNTs were immobilized through the formation of amide bonds between the amine-modified surface of gold electrodes and the carboxylic groups at the ends and sidewall defects of the nanotubes (Shearer et al., 2010). Hydrazide groups were then introduced onto the immobilized SWCNTs by diazonium salt electrografting. Site-oriented immobilization of anti-MS2 antibodies was performed through formation of hydrazone bonds. Additionally, we investigated iron-oxide-decorated SWCNTs as dual magnetic particles and bioreceptor immobilization platforms. Superparamagnetic iron oxide nanoparticles were grafted onto SWCNTs. The hybrid nanomaterials were successfully labeled with anti-MS2 antibodies through hydrazone bond formation with a previously electrografted diazonium salt. The resulting immunocarriers were dispersed and used to capture MS2 bacteriophage. Finally, the magnetic immunocarriers were captured on an electrode to carry out electrochemical MS2 detection. The possibility to rapidly and easily separate the target analyte from the sample by using an external magnetic field allows the pre-concentration of samples. Additionally, the electrode surface can be easily renewed after each measurement by simply switching off the magnetic field, being ready for the next measurement. Both immunosensors were successfully applied to the sensitive and accurate determination of MS2 in spiked river water samples. These new SWCNT-based platforms offer the possibility to tune the features of the sensing devices accordingly to the requirements of the final application. Significantly, the sensitivity of detection was sufficient to identify the presence of MS2 to levels regularly encountered in sewage-impacted environments. This versatility should translate into broad interest in new uses of CNTs for biosensor development.

2. Material and methods 2.1. Reagents All reagents were used as received. Cysteamine hydrochloride, mercaptoethanol, 4-aminobenzoic hydrazide (ABH), 4-aminophenol (APh), sodium nitrite, sodium periodate, potassium ferrocyanide (K4[Fe(CN)6]), potassium ferricyanide (K3[Fe(CN)6]), α-naphtyl phosphate (α-NP), anti-mouse IgG-alkaline phosphatase (ALP), N-hydroxysuccinimide (NHS), N-(3-dimethylaminopropyl)-N′ -ethylcarbodiimide hydrochloride (EDC), diethanolamine (DEA), 2-(N-morpholino)-ethanesulfonic acid (MES), iron (III) chloride (FeCl3), ferrous sulfate (FeSO4) and components of buffers were purchased from Sigma-Aldrich. Sulfuric acid (98% AR grade) and nitric acid (70% AR grade) were purchased from Merck. SWCNTs (pristine, P-2) were purchased from Carbon Solutions, Inc. (USA). Monoclonal antibody (developed in mouse) and polyclonal antibody (developed in rabbit) against MS2 bacteriophage were obtained from Galahad Sales (Australia). MS2 bacteriophage was kindly supplied by the Australian Water Quality Center, SA Water Corporation. All solutions were prepared using Milli-Q water. Screen-printed gold electrodes were purchased from DropSens (three-electrode system, ref. 250BT, 4 mm diameter, including a platinum counter electrode and a silver reference electrode).



2.2. Apparatus Electrochemical measurements were performed on an electrochemical analyzer (CH Instruments, model 600D series) using a three-electrode electrochemical cell, placed into a Faraday cage. Data acquisition and analysis were accomplished using CH Instruments software (CH Instruments, Inc., Austin, TX, USA). 2.3. Synthesis of carboxylic functionalised SWCNTs Purchased SWCNT powders were dispersed in concentrated nitric acid and refluxed for 30 min at 140 °C. The SWCNTs were further shortened by chemical oxidation in a mixture of concentrated sulfuric and nitric acids (3:1 (v/v), 98% and 70%, respectively) under ultra-sonication for 3 h at 20 °C. The water in the sonicator was cooled using ice to maintain a constant temperature of 20 °C. The reaction mixture was then diluted with deionized water and allowed to stand overnight for the SWCNTs to drop out of the suspension. The supernatant was then decanted off and the SWCNT precipitate was diluted with deionized water. This suspension was then filtered through a 0.22 μm poly(tetrafluoroethylene) (PTFE) membrane (Millipore) under vacuum and the SWCNTs were washed until the filtrate pH was close to neutral. The resulting acid-treated SWCNT powders contained carboxyl functionalities and had an average length and diameter of approximately 300–800 nm and 1–3 nm, respectively (Marshall et al., 2006). 2.4. Synthesis of iron-oxide-nanoparticle-decorated SWCNTs The iron oxide-SWCNTs were synthesized by dispersing the acid-treated SWCNTs in a 2:1 aqueous solution of FeCl3 and FeSO4 at pH 11 for 2 h at 80 °C. The iron oxide-SWCNT composite was purified by sonication in EtOH:water (1:1 v/v) followed by filtering through a 0.45 μM PTFE membrane and washing with copious amounts of water. The iron-oxide-nanoparticle-decorated SWCNTs were characterized using thermo gravimetric analysis (TGA) and energy dispersive X-ray spectrometer (EDS) (Fig. S1 in Supplementary data). 2.5. Electrode preparation and electrochemical characterization Gold electrodes were electrochemically cleaned in 0.5 M H2SO4 by scanning the potential between 0 and 1.6 V at 0.1 V/s for eight cycles. All electrochemical data were normalized based on the surface of each electrode (Trasatti and Petrii, 1991; Jensen et al., 2012). Electrodes were characterized at each modification step using cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS). Electrochemical measurements were performed in an unstirred solution of 2 mM ferrocyanide and 2 mM ferricyanide in 0.1 M phosphate buffer, pH 7.4. Cyclic voltammograms were obtained by scanning the potential at 0.1 V/s from 0.2 to 0.6 V. EIS measurements were performed under open circuit potential conditions. Frequencies from 10 kHz to 0.1 Hz in logarithmic spacing were applied. The AC amplitude was 10 mV. 2.6. Sensing platform based on covalently bound SWCNTs 2.6.1. Self-assembly of cysteamine Clean gold electrodes were exposed overnight at 4 °C to 20 μL of 1 mM aqueous cysteamine solution to introduce amine groups on the electrode surface. After modification, the electrodes were rinsed with water to remove weakly adsorbed molecules. 1 h incubation with 1 mM mercaptoethanol was performed to block the vacant places within the self-assembled monolayer.

3

2.6.2. Covalent binding of carboxylic-functional SWCNTs via carbodiimide coupling Initially, SWCNTs were dispersed in MES buffer by sonication until a homogeneous suspension was obtained. The carboxyl groups on the SWCNTs were activated by incubating them in a 26 mM EDC and 35 mM NHS solution in 0.1 M MES buffer (pH 4.5) for 30 min. 100 mL of activated SWCNT solution at different concentrations (from 28 to 1400 mg/mL) was incubated at room temperature during 1 h with the amine-modified electrodes to form amide bonds. Finally, unreacted active esters were blocked by 15 min incubation in a 0.1 M ethanolamine solution. 2.6.3. Hydrazide-phenyl diazonium salt electrografting SWCNTs covalently attached to gold electrodes were functionalized with hydrazide groups by diazonium salt electrografting. The in situ generation of the aryl diazonium followed a protocol previously described based on the addition of 0.5 equivalents of sodium nitrite to an acidic solution (0.5 M aqueous HCl) of 10 mM ABH and 10 mM APh (Prieto-Simón et al., 2014). APh was used to prevent diazo coupling, to improve the spatial distribution and to minimize non-specific adsorptions. These solutions were degassed and left to react for about 30 min in ice, prior to the electrografting process. The electrochemical reductive modification of the immobilized SWCNTs with in situ generated diazonium salts was conducted by cycling the potential between 0.6 V and 0.6 V. Subsequently, the electrodes were rinsed with copious amounts of Milli-Q water and then subjected to potential scanning between 0.2 V and 0.6 V for 10 cycles at 0.1 V/s in 0.1 M phosphate buffer to remove physically adsorbed compounds onto CNTs. 2.6.4. Oriented immobilization of anti-MS2 antibody through hydrazone bonds Sugar moieties attached to the Fc region of polyclonal anti-MS2 antibody were oxidized to aldehydes, following a procedure previously described (Hermanson, 1996). Briefly, a solution of 0.3 mg/mL anti-MS2 antibody was mixed with 0.2 mM sodium periodate. The mixture was incubated for 30 min protected from light and then diluted 10 times with phosphate buffer to stop the reaction. Residual periodate was removed from the oxidized antibody solution using spin-desalting columns (7 K MWCO, Thermo Scientific). The oxidized anti-MS2 antibody was incubated during 30 min at room temperature with hydrazide-modified SWCNTs to form stable hydrazone bonds. Electrodes modified with hydrazidemodified SWCNTs covalently bound, were incubated with 20 mL oxidized antibody. 2.7. Sensing platform based on the use of SWCNTs as magnetic immunocarriers Iron oxide-SWCNTs were dispersed in 0.5 M HCl by sonication until a homogeneous suspension was obtained. 20 mL of suspended iron oxide-modified SWNTs were dropped onto the gold working electrode, where they were trapped by placing an external magnet at the underside of the electrode. 2.7.1. Hydrazide-phenyl diazonium salt electrografting The in situ generation of the diazonium cations and their electrografting onto the iron oxide-SWCNTs were performed as described above for covalently bound SWCNTs. 2.7.2. Oriented immobilization of anti-MS2 antibody through hydrazone bonds Iron oxide-modified SWCNTs electrografted with diazonium salts were released from the electrode surface by removing the external magnet, to allow the antibody coupling to take place in solution. The immobilization of the antibody was performed



4

following the same protocol described for covalently bound SWCNTs. 2.8. Immunosensor protocol Polyclonal anti-MS2 antibody-modified SWCNTs, covalently attached to the electrode surface or dispersed in solution were used to build on them a sandwich-based immunosensor for MS2 bacteriophage. 1 h incubation of MS2 bacteriophage solutions in a wide range of concentrations (from 1 to 1011 pfu/mL) was carried out. Then, as a required step for a sandwich assay, 1 h incubation of a 2 mg/mL MAb anti-MS2 bacteriophage was performed. Finally, a 0.2 mg/mL anti-mouse IgG-ALP solution was incubated during 1 h to introduce the label required for the voltammetric detection. The SWCNT-modified electrodes were thoroughly rinsed with phosphate buffer between each step. SWCNTs in solution were placed after each incubation and washing step on the magnetic separation stand until supernatant clearing and then the supernatant was removed. Controls in the absence of each one of the components were systematically performed. 2.9. Electrochemical detection protocol Differential pulse voltammetry (DPV) was used as electrochemical detection technique, to measure the activity of the enzyme ALP used as label (Prieto-Simón et al., 2008). Both, SWCNTmodified electrodes and clean gold electrodes with iron oxidemodified SWCNTs trapped on their surface using a magnet on the electrode underside, were exposed for 20 min to a 2 mM α-NP solution in 0.1 M DEA buffer, pH 9.5. Then, the potential was scanned at 0.05 V/s from 0 to 0.4 V to measure the oxidation current of the enzymatic product α-naphthol. DPV data was corrected by subtraction of the voltammogram obtained in the absence of α-NP. Electrode assays were performed in triplicate. 2.10. River water sample preparation Aliquots of alum-treated water (from Morgan reservoir, Murray river) were spiked with MS2 bacteriophage to obtain dilutions from 1 pfu/mL to 1011 pfu/mL MS2. To perform electrochemical measurements, either samples of 20 mL were incubated onto the electrodes or samples of different volumes were incubated with iron oxide-modified SWCNTs in solution. Scheme 1 illustrates the stepwise procedure for the preparation of sensing platforms based on covalently bound SWCNTs and on the use of SWCNTs as magnetic immunocarriers for the detection of MS2 bacteriophage.

3. Results and discussion There are several precedents of incorporation of CNTs into electrode configurations aiming to develop ultrasensitive electrochemical biosensors (Chen et al., 2013; Balasubramanian, et al., 2012; Vashist et al., 2011). Those studies mainly focused on the good electrocatalytic properties and high electrical conductivity of CNTs, responsible for their ability to mediate electron transfer reactions with electroactive species in solution when used as the electrode material. Our work combines the excellent features of SWCNTs that provide high sensitivity and efficiency, and a biocompatible environment, with the merits of the site-oriented antibody immobilization required to preserve the antibodies folding and functionality. Additionally, to the best of our knowledge, we exploit for the first time the magnetic properties of the iron oxide-SWCNTs as magnetic immunocarriers able to perform the immunoassay in dispersion.

Scheme 1. Illustration of the preparation of the sensing platforms based on (A) covalently bound SWCNTs and (B) the use of SWCNTs as magnetic immunocarriers.

3.1. Sensing platform based on covalently bound SWCNTs On the basis of a chemical assembly strategy, SWCNTs were covalently attached to a cysteamine-modified gold electrode via amide bond formation. It has been previously reported that there is electron communication between tethered SWCNTs and gold substrates through short self-assembled monolayers that in the absence of immobilized SWCNTs would block the electron transfer (Diao et al., 2002; Yu et al., 2007). Cyclic voltammetry and EIS were used to confirm the covalent immobilization of SWCNTs on the electrode surface (Fig. S2 in Supplementary data). The results confirmed the electron communication between SWCNTs and the gold electrode, and thus the expected improvement of the electrochemical properties of the modified electrodes. Most of the CNT-immunosensors developed to date have not considered the preservation of the binding potential of antibodies by pursuing an oriented immobilization strategy. Recently, Marrakchi et al. (2013) developed an immunosensor for atrazine, showing that combination of site-directed immobilization of antibodies on the electrode surface and, separately, incorporation of CNTs into the electrode configuration decreases the limit of detection obtained by a factor of 100 compared to non-oriented antibodies. Another approach is that reported by Puertas et al. (2013) based on the oriented covalent attachment of anti-horseradish peroxidase antibodies on the surface of multi-walled CNT (MWCNT)-polystyrene composite material. Results showed a signal increase up to 10 times compared to an equivalent bioplatform prepared by random orientation of the antibodies on the composite material. We present here for the first time the oriented immobilization of antibodies directly onto SWCNTs covalently bound to the electrode. To that purpose, tethered SWCNTs were functionalised in one step by electrochemical grafting of diazonium cations generated in situ from ABH. Cyclic voltammetry was used to monitor the stepwise fabrication. As previously reported,



5

4.5

4 4.0 3.5

0

I (µA)

I (µA)

2

3.0 2.5

-2 0 µg/mL 28 µg/mL 140 µg/mL 700 µg/mL 1400 µg/mL

-4

-6 -0.2

0.0

0.2

0.4

0.6

E (V) Fig. 1. Cyclic voltammograms recorded at SWCNT-modified electrodes, prepared with different concentrations of SWCNTs, after ABH electrografting. Measurements were performed at 10 mV/s in a 2 mM ferrocyanide/2 mM ferricyanide solution in phosphate buffer.

the cyclic voltammogram peak current decreased and peak separation increased after electrografting (Prieto-Simón et al., 2014). Nevertheless, the presence of the tethered SWCNTs significantly reduced the redox potential for the ferrocyanide/ferricyanide couple by more than 0.1 V compared to the previously reported gold electrodes modified by ABH electrografting (Prieto-Simón et al., 2014). More interestingly, as shown in Fig. 1, cyclic voltammograms showed a sigmoidal shape that can be attributed to microelectrode behavior (Chen et al., 2013). The reversibility of the electron transfer reaction usually depends on the surface density of SWCNTs (Diao and Liu, 2005; Chou et al., 2009). Electrodes modified with a relatively high surface density of SWCNTs exhibit cyclic voltammetry features of macroelectrodes, due to the presence of a large uniform diffusion layer. In our case, a radial diffusion, typical of microelectrodes, seems to be prevalent. These results indicate that once the SWCNTs have been electrografted, the oxidation/reduction rate is approximately equal to the rate of diffusion of the electroactive species from solution to the SWCNTs. The main advantage of the microelectrode behavior is that the capacitive current is dramatically reduced, greatly improving signal-to-noise ratio. The peak shape of the cyclic voltammograms in ferrocyanide/ferricyanide shown in Fig. 1 indicates that, even at the highest concentration of SWCNTs used to prepare the electrodes (1400 mg/mL), there was no overlap among the diffusion layers of adjacent SWCNTs, and thus the immobilized SWCNTs still act as an array of independent microelectrodes. We suggest that the microelectrode behavior can be the result of the length of the SWCNTs (between 300 and 800 nm) that exceeds the reported length that enable homogeneous vertically aligned CNTs surfaces (Diao and Liu, 2010), combined with the roughness of the electrodes. Functionalization was followed by oriented immobilization of the capture antibody (polyclonal anti-MS2), previously oxidized, onto the hydrazide groups introduced by electrografting of diazonium salts onto SWCNTs (Prieto-Simón et al., 2014). We investigated the performance of the immunosensors based on a

2.0 1.5 1.0 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 9 10 10 10 11 10 12

[MS2] (pfu/mL) Fig. 2. Calibration curves of peak current (DPV) vs. [MS2] obtained for MS2 bacteriophage in phosphate buffer (black symbols) and in river water samples (white symbols) using sandwich-based electrochemical immunosensors prepared by diazonium salt electrografting on covalently bound SWCNTs.

sandwich immunoassay. Electrochemical measurements of the ALP activity used as label were performed using DPV. Good electrocatalytic activity of SWCNTs allowed an effective lower potential DPV determination of α-naphthol (136 mV) than that obtained using a diazonium salt-modified gold electrode without SWCNTs (176 mV) (Fig. S3 in Supplementary data), proving that the covalent immobilization of SWCNTs improves the electrochemical properties of the modified electrode. Fig. 2 shows the plots obtained from the DPV measurements of immunosensors prepared by electrografting on covalently bound SWCNTs followed by polyclonal anti-MS2 antibody immobilization, incubation with buffer or river water samples spiked with different concentrations of MS2 bacteriophage, interaction with monoclonal anti-MS2 antibody and labeling with anti-mouse IgGALP. The presence of covalently immobilized SWCNTs has greatly improved our previous results using diazonium salt-electrografted electrodes by reducing the limit of detection (LOD) 1.5 105 times (Table S1 (Supplementary data)). The developed immunosensor showed a 1.6-fold increase of the sensitivity towards MS2. The LODs for MS2 in buffer and river water have greatly decreased from 1.4 106 pfu/mL and 2.2 105 pfu/mL to 9.3 pfu/mL and 9.8 pfu/mL, respectively. Moreover, the percentages of slope deviation between calibration curves in buffer and in river water samples were lower than 4%, showing their ability to detect MS2 in spiked river water samples, without significant matrix effects. Nevertheless, higher current intensities were measured in river water compared to buffer, indicating that the high sensitivity of the platform and the high surface area also attract some electroactive interfering compounds present in river water. That limitation can be overcome by performing standard additions. 3.2. Sensing platform based on the use of SWCNTs as magnetic immunocarriers As an alternative to the covalent immobilization of SWCNTs on the electrode surface, we used iron oxide nanoparticle decorated SWCNTs



3.5

3.0

I (µA)

as platforms for highly sensitive electrochemical immunosensing. Zarei et al. (2012) previously reported a magnetic nanocomposite based on the capture of Fe3O4 nanoparticle-modified COOH-MWCNTs on a gold electrode for the detection of human tetanus IgG. Nevertheless, the working protocol relied on the permanent capture of the magnetic nanocomposite during all the steps for a sandwich immunoassay. The authors highlighted as advantage the possibility to easily renew the electrode surface by switching off the magnet, but they did not take advantage of the use of the Fe3O4-modified MWCNTs as magnetic immunocarriers able to perform the immunoassay in dispersion. Here, we exploited the magnetic properties of the iron oxideSWCNTs. As a result, our approach involves new advantages to the immunosensors based on diazonium salt-electrografted electrodes (Prieto-Simón et al., 2014). Those advantages result from the use of SWCNTs as enhancers of the electron transfer combined with magnetic particles, such as (1) the simplicity in the preparation of antibody-modified iron oxide-SWCNTs and their high stability over time that eliminates the need for electrode modification; (2) the simplified analytical protocol for the incubation of the modified-SWCNTs with the sample solution, followed by incubation with monoclonal anti-MS2 antibody and labeling with antimouse IgG-ALP; (3) the increased surface-to-volume ratio, able to provide numerous antibody-binding sites, greatly improving the sensitivity; and (4) the possibility to separate the binding and detection steps, greatly minimizing interferences from complex matrixes. As in the previous biosensing platform, special attention was given on the site-directed immobilization of the antibodies. To functionalize the magnetic SWCNTs by diazonium salt electrografting, the iron oxide-SWCNTs were first trapped onto the electrode surface using an external magnet. Once they were electrografted, the hydrazide groups were reacted with the aldehyde groups on the Fc region of the oxidized polyclonal anti-MS2 antibody providing antibody-modified SWCNTs. The sandwich immunoassay was then carried out in solution, greatly improving the kinetics of binding of analyte to the immunocarrier compared to the tethered SWCNTs. After incubation with buffer and river water samples spiked with MS2 bacteriophage, interaction with monoclonal MS2 antibody and labeling with anti-mouse IgG-ALP, SWCNTs were captured onto gold electrodes using a magnet on the electrode underside. This step generated an electrode ready for electroanalytical measurements. Similarly to what has been already described for CNT-modified electrodes (Chou et al., 2005), the presence of iron oxide-SWCNTs onto the electrode surface improved the electrocatalytic behavior by enhancing the measured currents for α-naphthol and shifting the oxidation potential to lower values compared to those found for gold electrodes (Fig. S3 in Supplementary data). Table S1 (Supplementary data) summarizes the results for the non-linear four parameter logistic regression fitting of the plots obtained for the immunosensor with iron-oxide SWCNTs (Fig. 3) compared to the other systems without SWCNTs and tethered SWCNTs. The immunosensor without SWCNTs showed an LOD of 1.4 106 pfu/mL, 1.2 105 times higher than that obtained by using iron oxide-SWCNTs (12 pfu/mL). The comparison of the curves in Fig. 3 for MS2 bacteriophage in phosphate buffer and in river water samples shows that the two curves almost overlap, indicating that the iron oxide SWCNTimmunosensing strategy minimizes non-specific adsorptions and matrix effects caused by interfering species that might be present in complex samples. Nevertheless, the LOD values obtained using magnetically captured SWCNTs are slightly higher than those obtained using covalently tethered SWCNTs. In order to push the performance of this immunoassay to obtain lower LOD values, we used the iron oxide-SWCNTs to pre-concentrate higher volumes of

2.5

2.0

1.5 100

101

10 2

10 3

10 4

10 5

10 6

10 7

[MS2] (pfu/mL) Fig. 3. Calibration curves of peak current (DPV) vs. [MS2] obtained for MS2 bacteriophage in phosphate buffer (black symbols) and in river water samples (white symbols) performing a sandwich immunoassay in solution with anti-MS2 antibody-modified iron oxide-SWCNTs, followed by capture of those modified SWCNTs on a gold electrode with an external magnet at the electrode′s underside.

8

10 µL 100 µL 500 µL 1000 µL 5000 µL

6

I (µA)

6

4

2

0

0

1

2

3

4

5

6

Volume MS2 (µL) Fig. 4. Intensity current (DPV) obtained using different volumes of 104 pfu/mL MS2 bacteriophage in a sandwich immunoassay performed in solution with diazonium salt electrografted iron oxide-SWCNTs, followed by capture of those modified SWCNTs on a gold electrode with an external magnet at the underside of the electrode.

MS2 spiked water samples. Several sample volumes at the same MS2 concentration were evaluated as shown in Fig. 4. Results suggest that a 4.3-fold increase of the sensor current can be achieved by increasing the sample volume 50 times. The versatility of the use of iron oxide-SWCNTs as magnetic immunocarriers to



separate and concentrate samples, seems to be very promising in the development of new electrochemical assays for analytes present in samples with complex matrixes.

4. Conclusions In summary, we present immunosensors that feature oriented immobilization of the capture antibody onto diazonium saltelectrografted SWCNTs. Two approaches have been used to immobilize SWCNTs onto gold electrodes: chemical assembly and magnetic capture. Covalent tethering of SWCNTs onto electrode surfaces enabled very low LODs thanks to the high signal-to-noise ratio provided by their microelectrode behavior. Moreover, a wide linear range was obtained, attributed to the alignment of the SWCNTs that enhances electron transfer and the spatial orientation of the binding sites of the site-directed immobilized antibodies. The possibility to tune the density of SWCNTs arrays opens up the possibility to tailor any sensor for specific working conditions. Alternatively, iron oxide-SWCNTs featuring site-oriented antibodies were captured onto gold electrodes by an external magnetic field. The larger amount of SWCNTs captured on the surface compared to those covalently tethered, translates into a higher sensitivity. The excellent analytical performance combined with the easy working protocol based on an ON/OFF switch of the magnetic field as a simple way to regenerate the electrode surface, suggests the potential of this approach to be applied as a platform in water quality controls. Our results confirm that SWCNT-modified electrodes, either by covalent tethering or magnetic capture, can work as effective and sensitive transducers. Moreover, the oriented immobilization of antibodies ensures the minimization of steric hindrance and, thus, facilitates the antigen–antibody interaction, enhancing the sensitivity of the analysis. Finally, and most significantly, both platforms have been demonstrated as detecting MS2 bacteriophage to a sensitivity comparable to the levels seen in the real-world situation, e.g. in wastewater (101–104 pfu/ml; Keegan et al., 2009) and sewage impacted wetlands (3 101 pfu/ml; Li et al., 2012). This is an important advance in the application of such biosensors for environmental analysis. Future work is envisaged to combine the developed sensing approaches with label-free strategies to enable measurements in real-time.

Acknowledgment We are grateful for financial support from the Australian Research Council′s Linkage Project Scheme (LP130100032).

Appendix A. Supplementary material

7

References Balasubramanian, K., Kurkina, T., Ahmad, A., Burghard, M., Kern, K., 2012. J. Mater. Res. 27, 391–402. Brehant, A., Glucina, K., Le Moigne, I., Laine, J.-M., 2010. Desalination 250, 956–960. Campàs, M., Garibo, D., Prieto-Simón B., 2012. Analyst 137, 1055–1067. Chattopadhyay, D., Galeska, I., Papadimitrakopoulos, F., 2001. J. Am. Chem. Soc. 123, 9451–9452. Chen, R., Li, Y., Huo, K., Chu, P.K., 2013. R. Soc. Chem. Adv. 3, 18698–18715. Chou, A., Böcking, T., Singh, N.K., Gooding, J.J., 2005. Chem. Commun., 842–844. Chou, A., Eggers, P.K., Paddon-Row, M.N., Gooding, J.J., 2009. J. Phys. Chem. C 113, 3203–3211. Constantopoulos, K.T., Shearer, C., Ellis, A., Voelcker, N.H., Shapter, J.G., 2010. Adv. Mater. 22, 557–571. Diao, P., Liu, Z., 2010. Adv. Mater. 22, 1430–1449. Diao, P., Liu, Z.F., 2005. Phys. Chem. B 109, 20906–20913. Diao, P., Liu, Z.F., Wu, B., Nan, X.L., Zhang, J., Wei, Z., 2002. Chem. Phys. Chem. 3, 8991. Feng, M., Han, H., Zhang, J., Tachikawa, H., 2008. Electrochemical sensors based on carbon nanotubes. In: Zhang, X., Ju, H., Wang, J. (Eds.), Electrochemical Sensors, Biosensors and Their Biomedical Applications, Elsevier Inc., pp. 460–501. Gooding, J.J., Wibowo, R., Liu, J., Yang, W., Losic, D., Orbons, S., Mearns, F.J., Shapter, J.G., Hibbert, D.B., 2003. J. Am. Chem. Soc. 125, 9006–9007. Hermanson, G.T., 1996. Bioconjugate Techniques. Academic Press, San Diego, CA. Huang, L.M., Jia, Z., O’Brian, S., 2007. J. Mater. Chem. 17, 3863–3874. Jensen, U.B., Ferapontova, E.E., Sutherland, D.S., 2012. Langmuir 28, 11106–11114. Katz, E., Willner, I., 2004. Chem. Phys. Chem. 5, 1084–1104. Keegan, A., Monis, P., Jagals, P., Toze, S., Blackbeard, J., 2009. Pathogen Risk Indicators for Wastewater and Biosolids. Final Report of the Water Environment Research Foundation. IWA Publishing, London, UK. Kumar, B., Park, Y.T., Castro, M., Grunlan, J.C., Feller, J.F., 2012. Talanta 88, 396–402. Leclerc, H., Edberg, S., Pierzo, V., Delattre, J.M., 2000. J. Appl. Microbiol. 88, 5–21. Li, Y.L., Deletic, A., Alcazar, L., Bratieres, K., Fletcher, T.D., McCarthy, D.T., 2012. Eco. Eng. 49, 137–145. Liu, J., Chou, A., Rahmat, W., Paddon-Row, M.N., Gooding, J.J., 2005. Electroanalysis 17, 38–46. Marrakchi, M., Helali, S., Soto Camino, J., González-Martínez, M.A., Abdelghani, A., Hamdi, M., 2013. Int. J. Nanotechnol. 10, 496–507. Marshall, M.W., Popa-Nita, S., Shapter, J.G., 2006. Carbon 44, 1137–1141. Merkoçi, A., Pumera, M., Llopis, X., Pérez, B., del Valle, M., Alegret, S., 2005. Trends Anal. Chem. 24, 826–838. Petty, N.K., Evans, T.J., Fineran, P.C., Salmond, G.P., 2007. Trends Biotechnol. 25, 7–15. Prieto-Simón, B., Campàs, M., Marty, J.L., Noguer, T., 2008. Biosens. Bioelectron. 23, 995–1002. Prieto-Simón, B., Saint, C., Voelcker., N.H., 2014. Anal. Chem. 86, 1422–1429. Puertas, S., de Gracia Villa, M., Mendoza, E., Jimenez-Jorquera, C., de la Fuente, J., Fernandez-Sanchez, C., Grazu, V., 2013. Biosens. Bioelectron. 43, 274–280. Shearer, C., Ellis, A., Shapter, J.G., Voelcker, N.H., 2010. Langmuir 26, 18468–18475. Thomas, T., Mascarenhas, R.J., Martis, P., Mekhalif, Z., Swamy, B.E.K., 2013. Mater. Sci. Eng. C 33, 3294–3302. Trasatti, S., Petrii, O., 1991. Pure Appl. Chem. 63, 711–734. Trilling, A.K., Beekwilder, J., Zuilhof, H., 2013. Analyst 138, 1619–1627. Valentini, F., Amine, A., Orlanducci, S., Terranova, M.L., Palleschi, G., 2003. Anal. Chem. 75, 5413–5421. Vashist, S.K., Zheng, D., Al-Rubeaan, K., Luong, J.H.T., Sheu, F.-S., 2011. Biotechnol. Adv. 29, 169–188. Wang, J., 2004. Electroanalysis 17, 7–14. Wang, J., Musameh, M., Lin, Y., 2003. J. Am. Chem. Soc. 125, 2408–2409. Xu, Y., Wang, E., 2012. Electrochim. Acta 84, 62–73. Yu, J.X., Shapter, J.G., Quinton, J.S., Johnston, M.R., Beattie, D.A., 2007. Phys. Chem. Chem. Phys. 9, 510–520. Yu, X., Munge, B., Patel, V., Jensen, G., Bhirde, A., Gong, J.D., Kim, S.N., Gillespie, J., Gutkind, J.S., Papadimitrakopoulos, F., Rusling, J.F., 2006. J. Am. Chem. Soc. 128, 11199–11205. Zarei, H., Ghourchian, H., Eskandari, K., Zeinali, M., 2012. Anal. Biochem. 421, 446–453. Zhang, Y., Zhou, D., 2012. Expert Rev. Mol. Diagn. 12, 565–571.

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.09.089.


Horizontally Aligned Carbon Nanotube Based Biosensors for Protein Detection.

WIMP detection and slow ion dynamics in carbon nanotube arrays.

Conducting polymer coated single-walled carbon nanotube gas sensors for the detection of volatile organic compounds.

Conductive carbon nanotube hydrogel as a bioanode for enhanced microbial electrocatalysis.

Microbial contamination of toothbrushes.

Signal Amplification Strategy Based on TiO2-Nanotube Layers and Nanobeads Carrying Quantum Dots for Electrochemiluminescent Immunosensors.

An oligonucleotide-functionalized carbon nanotube chemiresistor for sensitive detection of mercury in saliva.

Probe Sensor Using Nanostructured Multi-Walled Carbon Nanotube Yarn for Selective and Sensitive Detection of Dopamine.

Aqueous nanosilica dispersants for carbon nanotube.

A Thermal Model for Carbon Nanotube Interconnects.

Carbon nanotube-based lateral flow biosensor for sensitive and rapid detection of DNA sequence.

Oligomer-coated carbon nanotube chemiresistive sensors for selective detection of nitroaromatic explosives.

Carbon nanotube-enhanced polarization of fluorescent peptides: a novel amplification strategy for homogeneous detection of proteases.

Fluorescence-based method and a device for rapid detection of microbial contamination.

Colloidal methods for the fabrication of carbon nanotube-manganese dioxide and carbon nanotube-polypyrrole composites using bile acids.

Carbon nanotube signal amplification for ultrasensitive fluorescence polarization detection of DNA methyltransferase activity and inhibition.

Surfactant hydrogels for the dispersion of carbon-nanotube-based catalysts.

poly [Ni-(Protoporphyrin IX)] composite for amperometric detection of long chain aliphatic amines.

Carbon nanotube-polyamidoamine dendrimer hybrid-modified electrodes for highly sensitive electrochemical detection of microRNA24.

Single walled carbon nanotube-based junction biosensor for detection of Escherichia coli.

Laser Treated Carbon Nanotube Yarn Microelectrodes for Rapid and Sensitive Detection of Dopamine in Vivo.

Biosensor based on tyrosinase immobilized on a single-walled carbon nanotube-modified glassy carbon electrode for detection of epinephrine.

Methodology for modeling the microbial contamination of air filters.

Performance and microbial community of carbon nanotube fixed-bed microbial fuel cell continuously fed with hydrothermal liquefied cornstalk biomass.