Reduced graphene oxide/polyethylenimine based immunosensor for the selective and sensitive electrochemical detection of uropathogenic Escherichia coli

Highlights

• Electrophoretic deposition of reduced graphene oxide/polyethylenimine thin films on Au.

• Functionalization with anti-fimbrial E. coli antibodies.

• Uropathogenic E. coli electrochemical detection with high sensitivity and selectivity.

• Bacteria detection in serum samples.

Abstract

Fast, reliable and selective detection of microorganisms is of uttermost importance in clinical analysis, but also in food and water quality monitoring. In this study, we report on the construction of an immunosensor for sensitive and selective electrochemical detection of uropathogenic Escherichia coli (E. coli) UTI89 bacteria in aqueous and serum samples. We took benefit of electrophoretic deposition (EPD) to prepare, in a simple, controllable and cost effective way, gold electrodes modified with thin active layers of reduced graphene oxide/polyethylenimine (rGO/PEI). While rGO exhibits high surface area and favourable electrochemical properties, the presence of abundant NH₂ groups on PEI offers a plethora of opportunities for the sensor’s surface functionalization. To achieve selectivity of detection, the electrode surface was covalently modified with anti-fimbrial E. coli antibodies via amide bond formation. To minimize non-specific adsorption, the immunosensor was additionally modified with pyrene-polyethyleneglycol (pyrene-PEG) moieties prior to antibody immobilization. The detection of E. coli was based on the restriction of electron transfer of a redox mediator, in our case potassium ferrocyanide, to the rGO/PEI modified electrical transducer due to the formation of an immune complex. The developed immunosensor displayed a sigmoidal shape with a linear range of 1 × 10¹–1 × 10⁴ cfu mL⁻¹ (R² = 0.995) according to i(μA) = −16.66–20.5 × log[E. coli] (cfu mL⁻¹) and a detection limit of 10 cfu mL⁻¹. Additionally, the sensor performed well both in aqueous, serum and urine media, which is essential for its potential use for clinical diagnosis of pathogenic diseases. Selectivity studies showed that the immunosensor was able to discriminate between E. coli UTI89 wild-type strain and UTI89 Δfim, without fim operon.

Introduction

Urinary tract infections (UTIs) represent the most common bacterial infectious diseases for humans, and Escherichia coli (E. coli) are the most predominant pathogens responsible for 30–50% of hospital acquired and 80–90% of community-acquired UTIs [1]. Pathogenic E. coli bacteria produce virulence factors, able to infect and cause disease to the host tissues. Thus, it is of high importance to detect these microorganisms at a very low level and be able to discriminate between different bacterial strains. The common techniques for bacteria identification and detection rely on conventional culturing techniques [2]. However, these techniques are time-consuming, very elaborate and necessitate microbiology laboratory to be completed. Also these standard methods take up to a full day to rule out a negative sample, and the analysis may require up to several days to confirm a positive result. As such, these techniques are not suitable for on-site monitoring. Therefore, there is continuously a huge demand for the implementation of rapid, reliable, specific and highly sensitive detection of bacteria at a low cost. In this context, biosensors occupy a central position to achieve these goals [[3], [4], [5]].

The last decade has witnessed an emergence of a plethora of 2D materials [6]. Particularly, graphene has attracted a big deal of interest for its outstanding properties such as good conductivity, high surface area, ease of production and functionalization using different routes,… [6]. These properties have been exploited in various applied fields and particularly for sensing small organic molecules, metal ions and biological species [7,8].

Detection of pathogens with graphene based sensors represents thus an appealing technology. A vast amount of work has been performed using graphene-based field effect transistors (GFETs) [[9], [10], [11]]. Huang et al. investigated CVD graphene derivatized with anti-E. coli antibodies for sensing E. coli bacteria with a detection limit of 10 cfu mL⁻¹ [11]. Chang et al. described a GFET comprising anti-E. coli modified thermally-reduced graphene oxide sheets (rGO) semiconducting channels for the selective and sensitive detection of E. coli O157:H7 [9]. A low detection limit of 10 cfu mL⁻¹ was achieved under optimized conditions. Chen et al. utilized holey rGO chemically functionalized with an antimicrobial peptide, Magainin I, as the transducer element in a GFET for E. coli O157:H7 detection [10]. The device exhibited a good selectivity and a detection limit of 803 cfu mL⁻¹.

Electrochemical detection based on impedimetric [[12], [13], [14], [15], [16], [17]] approaches is another particular appealing technique for fast and selective detection of pathogens, due to its low cost, simplicity of operation, and the possibility of miniaturization and portability [18]. Voltammetric and amperometric-based immunoassays for pathogen sensing are in addition popular label-free approaches [[19], [20], [21], [22]]. Ahmed et al. reported on an antibody-based immunosensor for the specific detection of S. pyogenes pathogenic bacteria in human saliva with a linear response of 10²–10⁵ cells [19]. The integration of graphene nanostructures has proven to be advantageous in enhancing sensitivity of electrochemical sensors. Consequently, some recent works focused on the use of graphene composite electrodes for the construction of electrochemical immunosensors for the detection of pathogenic bacteria [23].

In a recent work by our group, we have studied the influence of the chemical composition and charge of graphene-gold interfaces on the adhesion strength of different E. coli strains through surface plasmon resonance (SPR) measurements [24]. Our results revealed that uropathogenic E. coli UTI89 interacts strongly with polyethyleneimine (PEI) modified graphene surfaces through electrostatic interactions. Given that these interactions are not specific, in this study, we propose a new strategy for rapid and specific detection of uropathogenic E. coli UTI89 in aqueous and serum samples using an electrochemical read out. The sensitive and selective detection of E. coli UTI89 is achieved through chemical functionalization of the NH₂ groups of PEI modified reduced graphene oxide (rGO) electrodes with anti-fimbiral E. coli antibodies along with poly(ethylene glycol) (PEG) modified pyrene units to inhibit non-specific adsorption of E. coli bacteria onto the sensor surface. The change in the peak current of potassium ferrocyanide redox probe using differential pulse voltammetry (DPV) was used for the quantitative analysis of E. coli UTI89 contaminated solutions. It relies on the fact that the formation of an immunocomplex between anti-fimbiral E. coli antibodies and E. coli UTI89 induces a restriction of electron transfer of potassium ferrocyanide to the electrical transducer, which scales linearly with E. coli UTI89 concentration. The designed pathogen chip can specifically detect E. coli UTI89 down to 10 cfu mL⁻¹ in a highly specific manner even in blood and urine samples.