Abstract

The objective of this work is to elaborate an immunosensing system which will detect and quantify Staphylococcus aureus bacteria. A gold electrode was modified by electrografting of 4-nitrophenyl diazonium, in situ synthesized in acidic aqueous solution. The immunosensor was fabricated by immobilizing affinity-purified polyclonal anti S. aureus antibodies on the modified gold electrode. Cyclic voltammetry (CV) and Faradaic Electrochemical Impedance Spectroscopy (EIS) were employed to characterize the stepwise assembly of the immunosensor. The performance of the developed immunosensor was evaluated by monitoring the electron-transfer resistance detected using Faradaic EIS. The experimental results indicated a linear relationship between the relative variation of the electron transfer resistance and the logarithmic value of S. aureus concentration, with a slope of 0.40 ± 0.08 per decade of concentration. A low quantification limit of  CFU per ml and a linear range up to  CFU per mL were obtained. The developed immunosensors showed high selectivity to Escherichia coli and Staphylococcus saprophyticus.

1. Introduction

Staphylococcus aureus is a major human pathogen responsible for a broad range of diseases because it is able to produce heat-resistant toxins in food [1]. Foods that are frequently incriminated in staphylococcal poisoning include meat and meat products, milk, and water. The toxic chemical released by S. aureus is enterotoxin that has been found to be produced at a hazardous level for 108 cells per kg of food [2]. Therefore rapid, sensitive, and reliable methods to detect foodborne pathogens are increasingly necessary in health care today. Traditional methods for foodborne bacteria detection including polymerase chain reaction (PCR) [3], enzyme-linked immunosorbent assay [4], and those based on culture and colony counting [5] have been widely applied due to their high reliability. However, these methods suffer from obvious disadvantages: they are time consuming and require expensive equipment and complicated pretreatment, so they are difficult to use on site.

Biosensor technologies play an increasingly important role in the detection of pathogenic bacteria because they present the great potential of satisfying the practical need for rapid, wearable, and low-cost detection [6]. Among biosensors, immunosensors are widely investigated for bacteria detection due to their specific advantages, such as high affinity and simple fabrication [6, 7]. In the literature, many recent studies focus on E. coli and Salmonella bacteria detection with different transducing techniques: QCM [8], SPR [9], electrochemical techniques: capacitive [10] and amperometric measurements [11]. Although several electrochemical immunosensors for the detection of foodborne pathogenic bacteria have been reported in the literature, only a few of them have been used for the quantification of S. aureus [1214]. Escamilla-Gómez et al. reported an amperometric immunosensor for the determination of S. aureus cells, obtaining a detection limit of cells per mL [12]. Tan et al. have developed a PDMS-based microfluidic immunosensor chip for the detection of foodborne pathogens S. aureus and E. coli O157: the detection limit of this impedimetric immunosensor modified with a nanoporous alumina membrane was around 102 CFU per mL [13]. Braiek et al. [14] have used a self-assembled monolayer of alkanethiols for the antibody immobilization on gold surface and EIS for the immunodetection of S. aureus and have reported a sensor sensitivity of 127  per decade of CFU with a detection limit of 10 CFU per mL [14].

The originality reported in this present study is the development of an immunosensor based on a diazonium electrografted gold electrode for the detection of S. aureus, which has never been tested until now. The method for elaborating stable thin organic layers on conductive substrates such as glassy carbon (GC), carbon nanotubes, HOPG, gold, metals, and on semiconductive substrates is now recognized as being an efficient procedure [1522]. The mechanism by which aryl groups attach themselves to gold is not fully understood, although there is a good evidence for the formation of Au-C bonds [23] which, in the case of aryldiazonium, results in quite a stable layer. By comparison, alkanethiol layers, that are highly stable [24], can only be applied to a rather limited number of materials (), whereas aryldiazonium layers can also be applied to carbonaceous materials. The use of electroreduction of aryl diazonium salts provides higher stability with respect to long-term storage in air, potential use in cyclic voltammetry under acidic conditions, and a wider potential window for subsequent electrochemistry [25]. Diazonium salt molecules, modified with various functional groups, have been used for the immobilization of nanoparticles [26], DNA [27], enzymes [28], carbon nanotubes [29, 30], proteins, antibodies and electroactive species [3134] on GC, Si, and Au surfaces. Moreover, applying an electrochemical potential to the specific electrode (chip) allows biomolecule immobilization which should be compatible with the production of biochip microarrays for multidetection.

In this work, anti-S. aureus antibody is covalently anchored onto a gold electrode surface through grafting onto p-aminophenyl groups, in situ generated in the aqueous deposition solution. The aryl diazonium covalently binds to the surface by the spontaneous release of nitrogen [35, 36], and the nitro group is then reduced. The different steps for immunosensor elaboration have been characterized using electrochemical measurements. The EIS technique is an efficient electrochemical technique for the transduction of biosensing events on electrodes [37, 38], and is therefore used here for affinity assays: antibody-cell binding. By modeling Nyquist plots, a calibration curve for S. aureus has been determined, and the analytical characteristics of this immunosensor have been deduced and compared to those of previous relevant immunosensors.

2. Experimental

2.1. Reagents

Polyclonal antibodies (developed in rabbit) against Staphylococcus aureus were obtained from BIOtech RDP (Sfax, Tunisia). P-nitroaniline (PNA), hydrochloric acid (37%), sodium nitrite, phosphate buffered saline (PBS), potassium ferrocyanide [K4Fe(CN)6], potassium ferricyanide [K3Fe(CN)6], hydrogen peroxide (30%), glutaraldehyde (25%), and bovine serum albumin (BSA) were purchased from Sigma Aldrich. All solutions were prepared with ultra-pure Milli-Q water.

2.2. Bacteria Cultivation

S. aureus reference strain (ATCC25923) was used in the present study. The bacteria strain was cultured in a tryptic soy broth [TSB, (Difco)] or on TSB agar plates for 24 hrs at 37°C. A highly concentrated bacteria suspension was prepared as follows: the liquid culture medium was inoculated with 100 μL of pre-culture solution and then cultivated at 37°C for 18–24 hrs. The bacterial culture was centrifuged for 5 min at 6400 rpm and then washed twice. Finally, it was resuspended in sterile phosphate-buffered saline (PBS). The proportions of viable cells and bacterial concentration were determined by the spread-plate technique. Optical density (OD), measured using a spectrophotometer, was used as a measure of the concentration of bacteria in a suspension. The cultures of S. aureus strain were grown to the stationary growth phase for . At the stationary growth phase, bacterial concentration did not change with time, and, subsequently, all the operations to determine bacterial concentration can be performed accurately [14].

2.3. Electrochemical Instrumentation

Electrochemical measurements were performed using a Voltalab 40 potentiostat-galvanostat with a standard three-electrode configuration. The measurement set-up consisted of a three-electrode system, with a gold electrode (surface area ) as the working electrode, a saturated calomel electrode (SCE) as the reference electrode and a platinum plate as the counter electrode.

All electrochemical measurements were carried out in a measuring chamber of volume (chamber) = 5 mL and inside a Faraday cage, at room temperature .

Two-electrochemical techniques were used in this work as follows.(i)Cyclic voltammetry was performed in 5 mM ferro/ferricyanide PBS solution (8 mM) at a scan rate of 100 mV/s.(ii)Faradaic EIS was used by applying a small sinusoidal signal (amplitude 10 mV; frequency range 100 mHz to 100 kHz) to the system at open circuit potential in 5 mM ferro/ferricyanide PBS solution (8 mM). (iii)The modeling software compatible with Windows (provided by Scribner Associate Inc, Southern Pines, NC, USA) was used for fitting the Faradaic impedance spectra.

3. Results and Discussion

3.1. Functionalization of the Electrode Surface with p-Nitrophenyldiazonium Cation
3.1.1. Formation of p-Nitrophenyl Diazonium Cation

The gold surface was cleaned in acetone, in an ultrasonic bath for 10 min, dried under a nitrogen flow, and then dipped for 2 min into a piranha solution 4 : 1 (v/v) [98% H2SO4/30% H2O2], rinsed with ultrapure water and dried under a nitrogen flow. The diazonium cations were synthesized in situ in accordance according to previously described procedure [35]: 20 mM of NaNO2 was added to the electrolytic solution containing 20 mM p-nitroaniline (PNA) in 0.5 M HCl under stirring at room temperature. The solution was cooled in an ice bath (the temperature did not exceed 0°C).

3.1.2. Grafting of p-Nitrophenyl Diazonium Salt

The electrochemical grafting on golf electrode was immediately performed in the mixture above described, by cyclic voltammetry (three cycles from 0.6 to −0.4 V/SCE at a scan rate of 100 mV/s). The consecutive cyclic voltammograms (CVs) of p-nitrophenyl diazonium at the gold electrode are presented in Figure 1(a).

The CVs are characterized by the first cycle exhibiting a well-defined, reproducible, and irreversible reduction peak located at 0.4 V/SCE. This feature corresponds to the typical electro-reduction reaction of the diazonium function, leading to the elimination of a nitrogen molecule and the production of highly reactive radicals. This radical was previously shown to attack the surface and to form a covalent bond between the aryl group and the Au electrode [39, 40].

Very low currents were observed during the second and third voltammetric cycles, evidencing that surface saturation of grafted molecules has been achieved. Although Brooksby and Downard have pointed out that the electrochemical determination of surface concentration must be interpreted with caution [41, 42], they assume a surface concentration of nmol/cm2 for one monolayer [43].

The surface concentration, , of the grafted nitrophenyl (NP) groups can be calculated according to the following formula:

is the charge corresponding to the integral , which is the area of the current time reduction peak located at 0.4 V/SCE (cf Figure 1(a)). It was calculated using Origin 8 software. is the number of electrons exchanged per reactant molecule (), is the Faraday constant (96.500 C/mol) and is the surface area of the working electrode.

A surface concentration (mole coverage) of p-NP:  nmol/cm2 was found, when the experiments were repeated three times in the same conditions. The surface concentration of p-NP on the Au surface is then found as a significantly higher coverage than one monolayer for the first voltammetric half cycle.

3.1.3. Reduction of Nitrophenyl Group to Aminophenyl Group

The electrode was then washed and transferred to 0.1 M KCl solution and subjected to five potential scans between 0.4 and −1.25 V/SCE at 100 mV/s in order to reduce the nitro group and obtain a modified film of 4-aminophenyl on the electrode surface. Figure 1(b) shows a cathodic sweep of the p-nitrophenyl modified gold electrode with an irreversible reduction peak at around −0.9 V/SCE. During the second scan, the reduction peak is drastically diminished, indicating that nearly all the electroactive -Ph-NO2 groups are reduced in the first scan. However, as pointed out in previous works [1517], the reduction of -Ph-NO2 to -Ph-NH2 is incomplete as evidenced by the reversible couple that appears at . This couple is assigned to the hydroxyaminophenyl/nitroso phenyl interconversion.

Once the p-aminophenyl layer had been formed, the electrochemical behavior of the modified gold electrode surface was investigated by cyclic voltammetry in the presence of the Fe2+/Fe3+ redox couple. Figure 2(a) shows the cyclic voltammogram for bare Au electrode and for aminophenyl-Au electrode, in the presence of the Fe2+/Fe3+ redox couple. This voltammogram is highly affected by the deposited layer; the redox peaks have entirely disappeared as expected for completely blocked surfaces. The aminophenyl (pAP) group is assumed to be covalently grafted to the Au surface through the formation of an Au-C bond [43, 44]. From the literature, after diazonium cation reduction, the stability of a diazonium compound on gold electrodes is weaker than on glassy carbon electrodes [45].

Electrochemical impedance spectroscopy measurements were also performed to further characterize the modified surfaces. Impedance is a complex number, according to (2) where is the angular frequency, is the imaginary number and , the angular frequency is equal to , being ac-frequency.

Impedance is represented in the complex plane, according to a Nyquist plot: imaginary part versus Real part . Figure 2(b) represents the Nyquist plots for bare Au electrode and for aminophenyl-Au electrode. Indeed, the charge-transfer resistance, estimated from the diameter of the semi-circle of the Nyquist plot, is higher for the p-AP-Au electrode than for the bare gold electrode. This is due to the lower rate of electron transfer through the film, showing high coverage at the electrode.

An electrical equivalent circuit for both a bare and a modified electrode, shown in Figure 2(c), comprises a series resistance (resistance of solution and of ohmic contacts), a resistance related to the charge transfer rate of redox reactions at the functionalized electrode , a constant phase element (CPE) related to the capacitance of the functionalized Au electrode/electrolyte interface. The capacitance of the interfacial film in parallel with the resistance of the interfacial film are also included in the equivalent circuit.

It is worth noting that the CPE reflects the nonideality of the double-layer at the functionalized gold electrode/electrolyte interface due to the roughness and porosity of the interfacial film.

The CPE is defined as where is the CPE constant and has the quality of a susceptance or -capacitance; is the angular frequency; is the factor describing deviation from ideal capacitor [46].

A specific electrochemical element of diffusion, the Warburg element () is defined as

Here, denotes the Warburg coefficient. This equation requires that .

3.2. Electrochemical Monitoring of Antibody Immobilization
3.2.1. Procedure of Covalent Grafting of the Antibody

Following deposition of the p-aminophenyl (p-AP) film on the gold electrode, the terminal amine was activated by incubation for 60 min with glutaraldehyde (25%) at pH 4, at room temperature [39]. The value of pH was chosen close to pKa of aminophenyl group.

After rinsing with PBS, the electrode was incubated for 1 hour in a 0.2 mg/mL solution of anti-S. aureus antibodies. The amino groups on the antibody molecule enable the covalent bonding with the glutaraldehyde activated amino functions on Au electrode. The total time necessary for the whole process of preparation of the immunosensor is 3 h.

Finally, after rinsing with PBS, the electrodes were incubated for 20 min in BSA (1%) to block the unreacted aldehyde groups. Cyclic voltammetry (CV) and EIS were employed to characterize the p-AP gold electrode (a), p-AP-GA gold electrode (b) and p-AP-GA-Ab gold.

Figure 3(a) shows the CV curves and Figure 3(b) Nyquist plots obtained with these surfaces.

The Faradaic EIS measurements are in good agreement with CV measurements; the current density decreases (Figure 3(a)) and the diameter of the semi-circles of the Nyquist diagrams increases significantly after the different stages of the elaboration of the immunosensor (Figure 3(b)). This indicates that following the observed blocking effect of the p-AP film, a further blocking effect was observed after the binding of glutaraldehyde and the covalent immobilization of the antibody.

3.2.2. Number of Immobilized Antibodies (Binding Sites)

In order to calculate the number of binding sites we calculate the antibody surface coverage , using the following equation: Charge transfer resistance after antibody immobilization, Charge transfer resistance after grafting of the aminophenyl layer.

55% of the surface is found to be covered with antibodies. Considering the surface density for a dense antibody layer determined in [47], the number of immobilized antibodies on the aminophenyl layer, it should be  Ab/cm2.

3.3. Detection of S. aureus Cells Using Electrochemical Impedance Spectroscopy

Having developed our sensor, we studied its response towards different concentrations of the S. aureus bacteria. The electrodes were incubated in different aliquots corresponding to different bacterial concentrations in CFU per mL. Figure 4(a) shows the Nyquist diagrams of the immunosensor obtained after incubation in increasing concentrations of the target bacterium S. aureus. The impedance spectra obtained in the presence of different concentrations of bacteria were modeled by the equivalent electrical circuit presented in Figure 2(c).

3.3.1. Surface Coverage by Bacteria

The percentage of surface coverage by bacteria was determined, according to the following equation: Charge transfer resistance after binding antibody, Charge transfer resistance after coverage with bacteria.

Assuming that the surface area of one S. aureus bacteriumis 3.14 μm2 [48], it follows that in presence of a concentration of 10 CFU per mL, the determined surface coverage of the immunosensor is 29%, corresponding to bacteria per cm2 on the Ab functionalized gold surface. In the presence of a concentration of 107 CFU per mL, a surface coverage of the immunosensor is 59%, corresponding to bacteria per cm2 has been determined.

3.3.2. Apparent Equilibrium Dissociation Constant

The reaction flow between immobilized antibody and bacteria is as follows:

Mass action applied to the equilibrium gives where represents the apparent equilibrium dissociation constant; [Ab] and [Bact] respectively are the concentrations of antibody at electrode surface and of bacteria in solution (CFU per mL). is the surface concentration of complex.

If and .

The expression of becomes

The measured relative variation of , , after equilibrium with different concentrations of bacteria [Bact], is related to the amount of complex formed. is the maximum variation obtained corresponding to saturation. It then comes

The measured relative variation of , , after equilibrium with different concentrations of bacteria [Bact] is presented in Figure 4(b).

is equal to [Bact] when . The determined value of , from Figure 4(b), is 102.6 CFU per mL.

3.4. Analytical Performance
3.4.1. Detection Limit, Dynamic Range

The charge transfer resistance increases gradually as the bacteria concentration increases after consecutive incubations from 10 up to 107 CFU per mL. The impedance immunosensor shows a linear relationship between the relative variation of charge transfer and the decimal logarithmic value of S. aureus concentration [Bact] (Figure 4(b)) in a range of 10 to 107 CFU per mL with a slope of 0.4 and a correlation coefficient of 0.994, the limit of quantification being 10 CFU per mL.

3.4.2. Reproducibility

The reproducibility of the immunosensor was investigated by repeating the experiments with three different immunosensors prepared in the same conditions. The response for three different immunosensors has an average standard deviation of 20%, which verifies the reproducibility of the system for the detection of S. aureus. The drawback is that our immunosensor is not reusable, because we were not able to regenerate our antibodies for other detections.

3.4.3. Specificity

To ensure that the response of our immunosensor was not due to the adsorption of bacteria on the Au surface modified by p-aminophenyl, different electrodes were elaborated in the same conditions, without the antibody incubation step. The results show that there is no significant variation of the relative variation of the charge transfer resistance of the modified electrode after incubation of the electrode in S. aureus solutions (Figure 4(b)).

The specificity of the impedimetric immunosensor system was determined for different bacteria strains at 106 CFU per mL. The relation variation of charge transfer resistance of the immunosensor for Staphylococcus saprophyticus (ATCC 15305) and Escherichia coli increased less than 0.2, whereas a marked increase by a factor 2 was found after incubation with S. aureus cells. Therefore, the specificity of the immunosensor is 10% for Staphylococcus saprophyticus (ATCC 15305) and Escherichia coli, compared to 100% for S. aureus. The characteristics of previous relevant immunosensors for S. aureus detection in terms of the limit of detection (LOQ) and of the dynamic range are given in Table 1. In this work, the LOQ obtained for the developed immunosensor is the lowest value and the largest dynamic range is obtained.

4. Conclusion

An impedance-based immunosensor for the detection of S. aureus has been developed. Electrochemical grafting of the gold electrode surface by p-aminophenyl (p-AP) was demonstrated as a suitable, simple, and versatile platform for antibody immobilization, giving stable and robust biomolecular layers. It allows rapid, inexpensive, sensitive, and selective detection of these bacteria. This method permits antibodies to be spatially addressed on a single gold chip in a microarray and then allows simultaneous detection of different bacteria. Further studies are in progress for the use on real samples.

Acknowledgments

The authors would like to thank EGIDE for its support through UTIQUE Program no. 09G 1128 and through PHC Maghreb no. 12 MAG 088. Amina Chrouda is grateful to the Tunisian Ministry of Higher Education and Technology for the “Mobility Grant”.