Selective Quantification of Bacteria in Mixtures by Using Glycosylated Polypyrrole/Hydrogel Nanolayers

Abstract

Here, we present a covalent nanolayer system that consists of a conductive and biorepulsive base layer topped by a layer carrying biorecognition sites. The layers are built up by electropolymerization of pyrrole derivatives that either carry polyglycerol brushes (for biorepulsivity) or glycoside moieties (as biorecognition sites). The polypyrrole backbone makes the resulting nanolayer systems conductive, opening the opportunity for constructing an electrochemistry-based sensor system. The basic concept of the sensor exploits the highly selective binding of carbohydrates by certain harmful bacteria, as bacterial adhesion and infection are a major threat to human health, and thus, a sensitive and selective detection of the respective bacteria by portable devices is highly desirable. To demonstrate the selectivity, two strains of Escherichia coli were selected. The first strain carries type 1 fimbriae, terminated by a lectin called FimH, which recognizes α-d-mannopyranosides, which is a carbohydrate that is commonly found on endothelial cells. The otherE. coli strain was of a strain that lacked this particular lectin. It could be demonstrated that hybrid nanolayer systems containing a very thin carbohydrate top layer (2 nm) show the highest discrimination (factor 80) between the different strains. Using electrochemical impedance spectroscopy, it was possible to quantify in vivo the type 1-fimbriated E. coli down to an optical density of OD₆₀₀ = 0.0004 with a theoretical limit of 0.00005. Surprisingly, the selectivity and sensitivity of the sensing remained the same even in the presence of a large excess of nonbinding bacteria, making the system useful for the rapid and selective detection of pathogens in complex matrices. As the presented covalent nanolayer system is modularly built, it opens the opportunity to develop a broad band of mobile sensing devices suitable for various field applications such as bedside diagnostics or monitoring for bacterial contamination, e.g., in bioreactors.

Keywords: biorepulsive; distinction of bacteria; electrochemical sensor; glycosides; nanolayer polymerization; polypyrrole.

Figure 1

Schematic overview of the investigated nanolayer systems. Top: The basic nanolayer (PPG), which provides the biorepulsive background in the sensing setup, is formed by electropolymerization of the monomer PyPG. PPG is conductive due to the PPy backbone (red) and at the same time biorepulsive due to the PG polymer (light blue) surrounding the PPy strands. By exchange of the electroactive monomer (from PyPG to PyMan or PyGlc, respectively), it is possible to vary the functionalization of the PPy backbone with either α-d-mannopyranosyl (Man, dark green, center) or β-d-glucopyranosyl (Glc, dark blue, bottom) residues to generate selective recognition sites for bacterial adhesion. Nomenclature: with a total thickness of 50 nm, the identity and thickness (X nm) of the glycosylated part of the layer system are described as ManX and GlcX, respectively. E. coli strain, which carries type 1 fimbriae and thus selectively adheres to mannosylated surfaces (green, E.c.(g)), will bind to ManX layers, while other E. coli strains (orange, E.c.(o)) lacking mannose-specific fimbriae and lectins will not recognize any of the surfaces.

Figure 2

Formation and composition of the nanolayer systems. Left: reaction conditions for the deposition of the nanolayer systems by sequential electropolymerization of PyPG followed by PyGlc and PyMan, respectively. The digit behind the glycoside acronym (Glc/Man) describes the thickness of the glycoside layer on top of the PPG layer. All polymer films were fabricated to be 50 nm thick. The polymerizations were conducted under constant-current conditions with a current density of 5 mA/cm² and a concentration of 5 mM for all electroactive monomers. All reactions were conducted in water with NaClO₄ (0.3 M) as the supporting electrolyte as well as HClO₄ (0.2 M) as dopant. Right: Ellipsometric data for the different nanolayer systems: the thickness of the PPG layer is indicated in blue, the glycosylated ones in yellow (Glc) and red (Man).

Figure 3

Admittance plot (inverse Nyquist plot) of all nanolayers (left: GlcX, right: ManX) and Au as a reference in 0.3 M NaClO₄ solution. All measurements were conducted at E_start: 0.1 V (vs Ag/AgCl) with an amplitude of 0.01 V, and the impedances were measured for 181 frequencies logarithmically distributed between 0.1 and 100 kHz.

Figure 4

Top row: Number of adhered E. coli bacteria (green: E. coli PKL1162 = E.c.(g), orange: E. coli S17–1 pMRE-Tn5–134 = E.c.(o)) on the different functionalized surfaces, with bare Au as reference. SD = standard deviation. All data sets were tested positively for normal distribution (α = 0.05) with the Kolmogorov–Smirnov and the Shapiro–Wilk test, and a significance level of p = 0.001 was confirmed for all data in reference to the Au data set. Bottom left: Highlighted comparison of the biorepulsive performance (pure strains) of Glc2 and Man2 compared to that of Au and PPG. Bottom right: Results of competitive adhesion on a Man2 surface using mixtures of the E.c.(g) and E.c.(o) strains.

Figure 5

Left: Admittance plot of Man2 in suspensions of E.c.(g) (E. coli PKL1162) in PBS/CASO broth of different optical densities at 600 nm (OD1–9: OD1 = 0.1, OD2 = 0.05, OD3 = 0.025, OD4 = 0.0125, OD5 = 0.00625, OD6 = 0.00313, OD7 = 0.00156, OD8 = 0.00081, and OD9 = 0.00039). E.c.(o) was present at a constant density of OD = 0.01 to demonstrate selectivity during the sensing process. All measurements were reproducibly conducted under the same conditions as the ones described in Figure 3. The data of Man2 without any E.c.(g) ("None") are presented as reference. Right: Logarithmic plot of the OD of the E.c.(g) bacteria against the measured charge transfer conductance (S_ct). The error bars indicate the standard deviation.