Subscriber access provided by DUESSELDORF LIBRARIES

Article
Graphene-coated surface plasmon resonance interfaces
for studing the interactions between cells and surfaces
Palaniappan Subramanian, Fatiha Barka-Bouaifel, Julie Bouckaert,
Nao Yamakawa, Rabah Boukherroub, and Sabine Szunerits
ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/am405541z • Publication Date (Web): 16 Jan 2014
Downloaded from http://pubs.acs.org on January 22, 2014

Just Accepted

“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 26 ACS Applied Materials & Interfaces

1
2
3 Graphene-coated surface plasmon resonance interfaces for studing the interactions between cells
4 and surfaces
5
6 Palaniappan Subramanian, Fatiha Barka-Bouaifel,1,2 Julie Bouckaert,3* Nao Yamakawa,3 Rabah
1

7
Boukherroub1 and Sabine Szunerits1*
8
9
10
11
12 1
Institut de Recherche Interdisciplinaire (IRI), CNRS USR 3078, Université Lille1, Parc de la Haute
13
14 Borne, 50 avenue de Halley, BP 70478, 59658 Villeneuve d'Ascq, France
15
2
16 Laboratoire de Technologie des Matériaux et de Génie des Procédés (LTMGP), Université
17
Abderrahmane Mira de Béjaia, Targa Ouzemour, 06000 Béjaia, Algérie
18
19
3
20 Unité de Glycobiologie Structurale et Fonctionnelle (UGSF), Université Lille 1, CNRS UMR 8576,
21 59655 Villeneuve d'Ascq, France
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52 *
authors to whom correspondence should be sent: Sabine Szunerits (Sabine.Szunerits@iri.univ-
53 lille1.fr); Julie Bouckaert (Julie.Bouckaert@univ-lille1.fr)
54
55
56
57
58
59
60 1
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 2 of 26

1
2
3 Abstract
4
5 A variety of physical and chemical parameters are of importance for adhesion of bacteria to surfaces.
6
7 In the colonization of mammalian organisms for example, bacterial fimbriae and their adhesins not
8 only seek particular glycan sequences exposed on diverse epithelial linings, they also enable the
9
10 bacteria to overcome electrostatic repulsion exerted by their selected surfaces. In this work we present
11 a new technique based on simplified model systems for studying the adhesion strength of different
12
13 Escherichia coli strains. For this purpose, gold-based surface plasmon resonance (SPR) interfaces
14 were coated with thin films of reduced graphene oxide (rGO) through electrophoretic deposition. The
15
16 rGO matrix was post-modified with polyethyleneimine (PEI), poly(sodium 4-styrenesulfonate) (PSS),
17
mannose, and lactose through π-stacking and/or electrostatic interactions by simple immersion of the
18
19 SPR interface into their respective aqueous solutions. The adhesion behaviors of one uropathogenic
20
21 and two enterotoxigenic Escherichia coli clinical isolates, that each express structurally characterized
22 fimbrial adhesins, were investigated. It was found that the UTI89 cystitis isolate that carries the
23
24 mannose-binding FimH adhesin was most attracted to the PEI- and mannose-modified surfaces,
25 whereas the att25 diarrhoeal strain with the N-acetylglucosamine-specific F17a-G adhesin
26
27 disintegrated the lactose-modified rGO. The highly virulent 107/86 strain interacted strongly with the
28 PSS-modified graphene oxide, in agreement with the polybasic surroundings of the ABH blood group-
29
30 binding site of the FedF adhesin, and showed a linear SPR response in a concentration range between
31 102-109 cfu/mL.
32
33
34
35
36 Keywords: Escherichia coli, reduced graphene oxide, polyethyleneimine (PEI), mannose, lactose,
37
poly(sodium-4-styrenesulfonate (PSS), surface plasmon resonance
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 2
ACS Paragon Plus Environment
Page 3 of 26 ACS Applied Materials & Interfaces

1
2
3 1. Introduction
4
5 The Escherichia coli (E. coli) bacterium has been intensively investigated over the last decades and is
6
7 the most widely studied prokaryotic model organism for its importance in the field of biotechnology
8 and microbiology. Enterohemorrhagic E. coli (EHEC) has been recognized as being the cause of
9
10 serious illness and mortality in outbreaks of foodborne illness.1 E. coli O157:H7 is one of the
11 pathogenic bacteria of most concern, but the problem is at large as other pathogenic E. coli strains are
12
13 known to cause foodborne and waterborne illnesses. The sensitive detection as well as the
14 discrimination of different E. coli strains remains thus of uttermost importance. Pathogenic E. coli
15
16 bacteria possess one or more virulence factors that determine the microbial pathogenicity and increase
17
the fitness of a pathogen during infection.2-4 A well-known virulence factor in uropathogenic E. coli
18
19 (UPEC) is FimH, a mannose-binding adhesin found at the tip of the type 1 pilus produced by most of
20
the enterobacteriaceae family, including UPEC.4-6 It was found that most FimH variants from
21
22 uropathogenic, faecal and enterohaemorrhagic isolates express the same specificity and affinity to
23
mannose structures,6 with the only exception being a FimH from E. coli O157:H7 strains.7 The effect
24
25 of FimH variation and mutation on bacterial adherence has mainly evaluated by means of the
26
27 hemagglutination assay, yeast agglutination assay, and by analysis of the adherence and invasion into
28 bladder epithelial cells.3, 7-7
29
30
Competition binding experiments on SPR interfaces modified with Fab fragments of the monoclonal
31
32 antibody 1C10, recognizing mannose-binding sites of FimH, have in addition been used to determine
33
the solution affinity of FimH-carbohydrate interactions.4, 6, 9
Other available quantitative methods are
34
35 based on culture and colony counting methods, polymerase chain reaction (PCR), enzyme-linked
36
37 immunosorbent assay (ELISA) or on a DNA-based nanobarcode detection strategy.8-10 While the latter
38 two methods have proven to be fast and sensitive, they require pre-treatment and cell lysis to extract
39
40 DNA. Strategies based on the direct detection of the whole microorganism using SPR11-16 or electrical
41 impedance spectroscopy (EIS)12, 13, 16, 17 have shown to give reliable results in much shorter time and
42
43 have consequently drawn a lot of interest. The required selectivity is achieved using surface
44 immobilized polyclonal antibodies,16, 18, 19
lectins,20 carbohydrates such as α-D-mannoside,12
45
46 antimicrobial peptides15, 21 or bacteriophages.22, 23
47
48 In recent years graphene and its derivatives (graphene oxide (GO), reduced graphene oxide (rGO)
49
50 have been investigated for their efficiency in pathogen capture, detection and killing.24, 25 The usability
51 of these 2D nanomaterials results from their heterogeneous plane structures featuring ionic and
52
53 aromatic components that facilitate chemical bonding and non-covalent interactions (π−π stacking,
54 electrostatic, H-bonding, etc) with a variety of molecules.26 We have shown that graphene-coated SPR
55
56 interfaces have several advantages over gold-based SPR interfaces.27, 28 The high surface-to-volume
57
ratio of graphene27, 29 and rGO28 has proven to be beneficial for efficient adsorption of biomolecules
58
59
60 3
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 4 of 26

1
2
3 when compared to naked gold due to possible π-stacking interactions between the carbon-based ring
4
structures of biological molecules and the hexagonal cells of graphene.28, 30
5
6
7 The motivation of the present study is to use graphene-based SPR chips as novel interface to
8 investigate in a facilitated manner the interaction preference and strength of E. coli. pathogens. In a
9
10 proof of principle that such simplified model interfaces can provide some answers to fundamental
11 questions concerning cell adhesion, positively, negatively and glycan-modified SPR interfaces were
12
13 fabricated. Positively charged SPR interfaces were formed by modification of the graphene-based SPR
14
interface with polyethyleneimine (PEI). The integration of poly(sodium 4-styrenesulfonate) (PSS)
15
16 resulted in a strongly negatively charged interface. The concept of the surface modification is based on
17
18 π−π stacking interaction between the aromatic structures of the polymers and the graphene matrix.
19 This approach prevents the need of pre-functionalization of the polymers by thiol and other functional
20
21 linkers and makes the concept of interest for the integration of many other organic molecules. In the
22 case of PEI and PSS modifications, different electrostatic interactions with the graphene interface are
23
24 expected to be essentially responsible for discrimination between E. coli strains. On the other hand,
25 glycan-modified interfaces should be able to distinguish bacteria specific for simple monosaccharides,
26
27 such as the mannose-binding UTI89 strain and the N-acetylglucosamine-binding att25 strain. It has
28 been verified that the binding UTI89 E. coli to the mannose-modified interface was mannose-specific
29
30 binding, using a mutation in the mannose-binding pocket (UTI89 Q133K)3 and using a fimH knock-out
31 (UTI89 fimH-), in the otherwise isogenic UTI89 strain (Table 2).8 The potential of the interfaces for
32
33 analytical purposes was also investigated. E. coli 107/86 is known to cause oedema disease in pigs at
34
the age of weaning. It produces a Shiga toxin II variant in the digestive tract that is absorbed into the
35
36 circulation and that causes lesions in the vasculature of the intestine, subcutis and brain.31 Quantitative
37
and qualitative information on the presence of E. coli strain 107/86 is thus of interest.
38
39
40
41
42 2. Experimental part
43
44
2.1. Materials
45
46
47 Graphite powder (<20 microns), hydrogen peroxide (H2O2), sulfuric acid (H2SO4), dimethylsulfoxide
48 (DMSO), mannose, lactose, polyethyleneimine (PEI) (MW=1.8 kDa), poly(sodium 4-
49
50 styrenesulfonate) (PSS) (MW=70 kDa), were purchased from Aldrich and used as such.
51
52
53
54
55 2.2. Formation of rGO-SPR interfaces
56
57
58
59
60 4
ACS Paragon Plus Environment
Page 5 of 26 ACS Applied Materials & Interfaces

1
2
3 Graphene oxide (GO) was synthesized from graphite powder by a modified Hummers’ method32 for
4 which the detailed experimental conditions are reported.33 GO was exfoliated in milli-Q water (0.5
5
6 mg/mL) by ultrasonication for 3 h until the solution became clear with no visible particulate matter.
7
8 Gold-based SPR interfaces were formed by depositing 5 nm of titanium and 50 nm of gold
9
10 successively onto cleaned glass slides by thermal evaporation.
11
12 The electrophoretic deposition was carried out using a two-electrode cell containing the GO aqueous
13
14 dispersion (0.5 mg/mL) by applying DC voltage (150 V) for 20 s. Platinum (Pt) foil (1 × 2 cm2) acts as
15 the cathode and the SPR interface [glass/Ti (5 nm)/Au (50 nm)] substrate as the anode. The two
16
17 electrodes were separated by 1 cm and were placed parallel to each other in the GO dispersion. After
18 deposition, the rGO-SPR interface was washed with deionized water (three times) followed by blow
19
20 drying with nitrogen.
21
22
23
24
2.3. Functionalization of the rGO-SPR interfaces
25
26
27 The rGO-SPR interfaces and gold-based interfaces were immersed for 2h in 100 mM aqueous
28 solutions of PEI, PSS, mannose or lactose before being rinsed three times with water. Diluted rGO/PEI
29
30 interfaces were formed by their immersion into 10 mM, 1 mM and 0.1 mM PEI solutions for different
31 time intervals.
32
33
34 2.4. Determination of the amount of mannose (lactose) on mannose (lactose)/Au/rGO modified
35 interfaces
36
37
The amount of glycan was determined by treatment with a phenol/H2SO4 solution as described
38
39 previously.7 First, a calibration curve for mannose or lactose in solution was established using a
40
41 phenolic aqueous solution (5 wt%, 60 µL), concentrated H2SO4 (900 µL) which was added to an
42 aqueous mannose solution (60 µL), stirred for 10 min and then an absorption spectrum was recorded
43
44 (Perkin Elmer Lambda 950 dual beam) against a blank sample (without glycan). The absorbance of the
45 solution was measured at two wavelengths: λ1=485 and λ2=570 nm and the absorbance difference
46
47 (A485 – A570) plotted against the concentration of the glycan. The glycan-modified Au/rGO SPR
48 interface was treated with phenol/H2SO4 following the same protocol as described above.
49
50
51
52
53 2.5. Bacterial strains
54
55 The E. coli strains have been cultured on LB-agar complemented with the appropriate antibiotics for
56
57 the engineered strains. Upon overnight growth at 37 ºC, colonies were harvested by scraping them off
58
59
60 5
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 6 of 26

1
2
3 the agar and diluted in phosphate-buffered saline to different concentrations. Bacterial concentrations
4 were determined by measuring the optical density at 600 nm using a Perkin Elmer Lambda UV/Vis
5
6 950 spectrophotometer and a conversion of OD600 nm with 0.66 = 1x109 cfu/mL.
7
8 2.6. Instrumentation
9
10
11 2.6.1. X-ray photoelectron spectroscopy (XPS)
12
13 X-ray photoelectron spectroscopy (XPS) measurements were performed with an ESCALAB 220 XL
14
spectrometer from Vacuum Generators featuring a monochromatic Al Kα X-ray source (1486.6 eV)
15
16 and a spherical energy analyzer operated in the CAE (constant analyzer energy) mode (CAE=100 eV
17
for survey spectra and CAE = 40 eV for high-resolution spectra), using the electromagnetic lens mode.
18
19 The detection angle of the photoelectrons is 30°, as referenced to the sample surface. After subtraction
20
21 of the Shirley-type background, the core-level spectra were decomposed into their components with
22 mixed Gaussian-Lorentzian (30:70) shape lines using the CasaXPS software. Quantification
23
24 calculations were performed using sensitivity factors supplied by PHI.
25
26
27
28 2.6.2. SPR Instrumentation
29
30
31 The surface plasmon resonance instrument used was an Autolab ESPRIT instrument (Eco Chemie,
32 Utrecht, The Netherlands) with an incident laser light with 670 nm wavelength. The configuration of
33
34 this equipment is described elsewhere.34, 35 The instrument is equipped with a cuvette of about 100 µL.
35 The prism used had a refractive index of n = 1.52. WinSpall 2.0 software (Max-Planck-Institute for
36
37 Polymer Research, Mainz) was used to calculate the SPR curves and approximate the experimental
38 dependences with an optical model within the framework of Maxwell macroscopic approach. The data
39
40 recorded in this manuscript were collected from five different experiments.
41
42
43
44
2.6.3. Zeta-potential measurements
45
46
47 Zeta-potential measurements were performed on electrophoretically deposited Au/rGO and
48 functionalized Au/rGO matrices after the dissolution of the matrix from the Au interface in DMSO
49
50 and resuspension into water using a Zetasizer Nano ZS (Malvern Instruments S.A., Worcestershire,
51 U.K.) using the electrophoretic mode.
52
53
54
55
56 2.6.4. Confocal Microscope images
57
58
59
60 6
ACS Paragon Plus Environment
Page 7 of 26 ACS Applied Materials & Interfaces

1
2
3 Confocal images were acquired with a 60× objective (Nikon Plan Apo VC, NA=1.2 on a Nikon A1R
4 confocal microscope, Nikon Biosceince, Confocal Systems, Nikon ECLIPSE FN1 fixed stage France)
5
6 equipped with a 488-nm excitation argon laser and an Eclipse FN1 fixes stage direct microscope. The
7
SIMPLEPCI software (Comprox Imaging Systems, Cranberry Township, PA) was used for image
8
9 capture. Photomultiplier tube back levels were kept as a default of 600 for transmitted light images.
10
11
12
13
14 3. Results and discussion
15
16 3.1. Preparation modified reduced graphene oxide coated SPR interfaces (rGO-SPR)
17
18 The rGO-SPR interface was prepared as described previously (Figure 1A).28 The process consists of
19
20 the electrophoretic depositon of rGO onto gold-SPR interface through the application of a voltage of
21 150 V for 20 sec. The GO platelets with a zeta-potential of –35 mV migrate towards the positively
22
23 biased gold-based SPR interface, when a DC voltage is applied.
24
Figure 1B displays a scanning electron microscopy (SEM) image of the Au/rGO SPR interface. It
25
26 shows that the rGO film is homogenously deposited over the gold surface. The C1s core level XPS
27
spectrum of the Au/rGO SPR interface (Figure 1C) indicates clearly that GO was reduced to rGO
28
29 during the electrophoretic deposition. The XPS spectrum can be deconvoluted into three dominant
30
bands at 283.8 (sp2-hybridized carbon), 285.5 (C-OH) and 287.3 eV (C=O) with a C/O ratio of 3.3
31
32 (Table 1). This is different to the intial GO oxide solution where the C1s core level XPS spectrum
33
34 displayed contributions at 283.8, 284.7, 286.7 and 287.9 eV assigned to sp2-hybridized carbon, C-H/C-
35 C, C-O and C=O species, respectively with a C/O ratio of 1.73.28
36
37
38 The SPR signal of the Au/rGO SPR interface is depicted in Figure 2. From the SPR angle shift of the
39
40 Au/rGO SPR interface, the thickness of the electrophoretically deposited rGO can be estimated to be
41 approximately 2 nm, corresponding to about 5-6 graphene layers.36 Such interfaces were in the
42
43 following modified with polyethyleneimine (PEI, 1.8 KDa), poly(sodium 4-styrenesulfonate) (PSS, 70
44 KDa), D-mannose or D-lactose (Figure 1A). Stable interfaces were formed by incubating the Au/rGO
45
46 SPR interfaces for 2h in the aqueous solution (100 mM) of the reactant. The molecular adsorption onto
47
rGO occurs through π-stacking and/or electrostatic interactions, resulting in stable SPR interfaces. The
48
49 sucessful integration of the different molecules onto Au/rGO and information on the chemical
50
composition of the modified interfaces were obtained by XPS analysis (Table 1). In the case of
51
52 Au/rGO/PEI, besides C1s and O1s contributions N1s with an atomic percentage of 10.3 % was found.
53
54 The integration of PSS onto Au/rGO is evidenced by the presence of S2p (2.3 at. %) (Table 1).
55 The amounts of mannose and lactose loaded onto Au/rGO SPR were determined using a colorimetric
56
57 assay, based on the specfic reaction of carbohydrates with phenol in concentrated sulfuric acid as
58 previously described.37 The method allowed the amount of sugar to be estimated as
59
60 7
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 8 of 26

1
2
3 Γmannose=(5.8±0.6)×1012 molecules cm-2 and Γlactose=(3.4±0.6)×1012 molecules cm-2. These interfaces
4
were in the following used for carbohydrate-specific detection of the different E. coli strains.
5
6
7
8
3.1. Adhesion behavior of different E. coli strains on Au/rGO SPR interfaces
9
10 As charge is known to play an important role in bacterial adhesion, the interactions of five E. coli
11
strains with rGO matrices with a zeta potential of ζ = +52±2 mV and ζ = -57±2 mV (Table 1), formed
12
13 through interaction with polyethyleneimine (PEI) and poly(sodium 4-styrenesulfonate) (PSS)
14
15 respectively, were investigated using SPR. The integration of PEI with a low molecular weight of 1.8
16 kDa did not alter significantly the SPR signal, while integration of the high molecular weight (MW=70
17
18 KDa) PSS resulted in a red shift as seen in Figure 2.
19 We used in this study one uropathogenic (UTI89) and two enterotoxigenic E. coli (107/86 and att25)
20
21 clinical isolates, that each express structurally characterized fimbrial adhesins (Figure 3).
22 Differentiation of E. coli strains is known to evolve due to the presence of virulence factors, such as
23
24 the FimH adhesin, that when lost or mutated result in an altered capability of colonization and avidity
25 to cell surfaces and surfaces in general.4 This “biochemical” perception of bacterial interaction with
26
27 surfaces is additionally originating from an interplay between attractive and repulsive electrostatic
28 forces, hydrogen bonding and other physio-chemical parameters such as hydrophobicity, surface
29
30 roughness, etc.38, 39 Figure 3 shows the distribution of positively (blue color) and negatively (red
31 color) charges present at the very end of the fimbriae of each of the pathogens. The amino acid ligands
32
33 for the butyl α-D-mannose and residues contributing to the binding site of FimH are in addition shown
34
in the ball and-stick model. From this illustration is becomes obvious that UTI89 presents a strongly
35
36 localized negative charged mannose- binding pocket, while 107/89 shows an accumulation of positive
37
38 charge bordering the blood group sugar binding groove at the tip of FedF lectin domain. The F17a-G
39 lectin domain in complex with N-acetylglucosamine, of the E. coli att25 strain, is polar to
40
41 electronegative in charge in the binding site, but this is neutralised by a positively charged handle
42 underneath the monosaccharide-binding site. The adhesion of two control E. coli strains was
43
44 investigated. The UTI89 Q133K and UTI89 fimH- strains serve as negative controls for the otherwise
45 isogenic UTI89 E. coli strain (Table 2). Table 2 shows in addition the experimentally measured zeta
46
47 potential of the whole E. coli cell. While there are local charge changes around the binding site of the
48 lectin domain of the adhesin (Figure 3) the global charge of the bacteria is always negative.
49
50
51 SPR studies were carried out to measure the bacterial capture activity of the functionalized SPR
52 interfaces. Figures 4A shows the change in the SPR angle over time upon addition of E. coli strains
53
54 (1.5×109 cfu/mL) onto negatively charged PSS-coated rGO interfaces. A significantly different
55 adhesion behavior was observed for E. coli 107/86, with a three times higher shift in SPR angle
56
57 compared to the other E. coli strains (Figure 4B). As shown in Figure 3B, the F18 fimbriae of E.coli
58
59
60 8
ACS Paragon Plus Environment
Page 9 of 26 ACS Applied Materials & Interfaces

1
2
3 107/86 carry the FedF adhesin that is positively charged at the tip. Enhanced electrostatic interaction
4 with the negatively charged PSS matrix and the positive charge at the tip of the FedF adhesin 4 of E.
5
6 coli stain 107/86 is most likely at the origin for enhanced affinity and thus larger capture density
7
(Figure 4A). A somewhat larger capture activity was also observed with E.coli att25, specific strain
8
9 presenting some positive charges underneath the N-acetylglucosamine-binding pocket as illustrated in
10
Figure 3C (Lys90, Arg93). The increased changes in SPR angle correlates to increased baterial
11
12 adhesion, as confirmed by confocal microscopy images (Figure 4C).
13
14 In addition to electrostatic interactions, π−π stacking has also to be considered in this system.
15 However, as seen in Figure 5A summarizing SPR data on the different rGO interfaces for each E. coli
16
17 strain, bacterial adhesion to rGO is independent of the used E. coli. being the same order of magnitute
18 (≈80 millidegrees) for all tested strains. Furthermore, the importance of rGO is furthermore evidenced
19
20 when comparing the adhesion behavior of the different E. coli strains to gold interfaces modified in
21 the same way as rGO. As can be decuded from Figure 5B, no differentiation is obverved on such SPR
22
23 interfaces thus addressing the role of rGO in bacterial adhesion. In addition, bacterial adhesion to rGO
24
is comparable to non-specific adsorption on gold. It can be concluded, that π−π stacking interactions
25
26 are minimaly involved in the differential binding of the E. coli strains under study. In contrast, the
27
degree of charge seems to be an important parameter defining bacterial adhesion. The experimental
28
29 data indicate that a SPR interface with a zeta potential of -32±3 mV is not sufficient negatively
30
31 charged to allow strong electrostatic interactions to take place. Indeed, the zeta potential of rGO is
32 negative as is the case of rGO/PSS, but its absolute value is approximately half that of Au/rGO/PSS
33
34 (Table 1). The overall zeta potential over each of the E. coli strains is negative, the least so for the
35 107/85 strain (-24±2 mV) (Table 2) that is the first one to overcome the electrostatic repulsion by
36
37 specific bacterial adhesion via the FedF adhesin binding pocket.
38
39
40 The SPR results indicate that a PSS-modified Au/rGO SPR interface allows for a partially selective
41 detection of the E. coli strain 107/86. As mentioned previously, specific interaction of bacteria with
42
43 SPR interfaces is currently achieved using monoclonal antibodies,16, 40
protein G,19 and
44 bacteriophages.22, 41 Major disadvantages of antibodies are their relative instability, limiting their long-
45
46 term storage and increasing the production cost. Bacteriophages, viruses recognising and binding
47 specific receptors on host bacteria, are somewhat easier to produce and have a longer storage time, but
48
49 they suffer from poor surface coverage and subsequent bacterial capture.23 To determine the sensor
50
sensitivity, the PSS-modified rGO surface was tested with a concentration series of E. coli 107/86
51
52 (Figure 4D). The SPR angle changes linearly with the 107/86 concentration with a linear response
53
between 102-109 cfu/mL range and a detection limit of ≈102 cfu/mL. In contrast, the detection limits of
54
55 the other E. coli strains were around 104 cfu/mL. The detection limit for E. coli 107/86 is comparable
56
57
to other methods used of pathogen analysis based on fluorescence or electrochemical impedance
58 spectrosopy.17 It is also comparable to bacteriophage-modified SPR chips for detection of O157:H7 E.
59
60 9
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 10 of 26

1
2
3 coli,22 or E. coli K-12,41 but is more sensitive than other SPR-based E. coli platforms.16, 19
4 Parallel experiments were performed on positively charged Au/rGO/PEI interfaces (Figure 6A). The
5
6 strongest interaction was observed with UTI89 (Figure 6B) with a detection limit of around 100
7
cfu/mL (Figure 6C), sufficient for a high positive predictive value for cystitis in symptomatic women.
8
45
9 The discrimination of UTI89 from the other E. coli strains is less pronounced than was the case for
10
E. coli strain 107/86 on Au/rGO/PSS (Figure 4B), however it is the only E. coli strain that stands out,
11
12 indicating a specific interaction of the UTI89 strains with the PEI-modified interface. This agrees with
13
14
the local, negative charge accumulation at the lectin domain end of the FimH adhesin at the edge of
15 type-1 fimbriae (Figure 3A). The specificity of the interaction proper to the adhesin was confirmed
16
17 using two mutations of UTI89: either glutamine 133 to the basic lysine (Q133K, which would render
18 the binding pocket more positively charged however specific binding is abolished,7 or by the deletion
19
20 of the adhesin (UTI89 fimH-), that equally nihilates specific binding. Both showing a decreased
21 affinity to rGO/PEI when compared to UTI89, with a difference that corresponds to the difference
22
23 between UTI89 binding to mannose-modified surface and these of the UTI89 non-binding mutants.
24 Because the Q133K and fimH- mutation and deletion abolish specific recognition by the FimH
25
26 adhesin, the negative charge localized in the binding site (Figure 3A) only configures a role in the
27 case of fimbrial adhesin-mediated specific bacterial adhesion.
28
29
30 To investigate in more details the effect of surface charge on the change of bacteria affinity, rGO/PEI
31
32 interfaces with varing zeta-potentials were formed by varing the amount of PEI integrated onto the
33
rGO matrix (Figure 6D). In the case of UTI89, decreasing the zeta-potential of the matrix results in a
34
35 decrease in UTI89 affinity to rGO/PEI. These results support the observation that the specific,
36
adhesion-mediated, bacterial adhesion UTI89 to rGO/PEI is enhanced by local charges in and around
37
38 the binding site. There is indeed no significant effect of dilution of PEI on the affinity of E. coli strain
39
107/86. Some binding independent of PEI concentration and thus charge of the surface is observed
40
41 (Figure 6D). The adhesion of strains Q133K and UTI89 fimH- to rGO/PEI was also always higher
42
43 than on unmodified rGO (-32 mV) (≈80 millidegrees), mannose-(-39mV) on PSS-modified rGO (-57
44 mV) (Figure 5A). Considering that the overall charge of each of the E. coli strains is negative (Table
45
46 2), favoring electrostatic interactions between the E. coli and the positively charged interface, this
47 might not seem suprising. The imine-modified Au/rGO/PEI leads to increased background or basal
48
49 level of bacterial binding and is not likely to be mediated by the fimbrial adhesin. The PEI
50 modification itself seems responsible for some interactions in a charge independent way as the affinity
51
52 to even slighly negatively charged rGO/PEI (zeta potential =-12 mV) is stronger (Figure 6D) than on
53 negatively charged rGO (Figure 5A). E. coli is possibly attracted to a killer surface: the viability of E.
54
55 coli has been reported to decrease dramatically on positively charged, amine-coated surfaces.39 The
56 increases in basal level of binding moreover differ between the different E. coli strains, for each of the
57
58 three E. coli strain expresses, besides the fimbrial type, different biopolymers on its surface such as
59
60 10
ACS Paragon Plus Environment
Page 11 of 26 ACS Applied Materials & Interfaces

1
2
3 flagellae and lipopolysaccharides. The latter LPS can explicitely carry negatively charged compounds,
4 such as kdo, that is a bacterial variant of sialic acid. The basal level of binding, or background, can
5
6 thus vary accordingly, and is generally higher on the PEI interface than on the PSS and on the other
7
modified rGO and Au surfaces.
8
9
10
To widen the concept of using modified rGO SPR interfaces for the study of E. coli tropism, glycan-
11
12 modified interfaces were in addition investigated. Stronger capture densities of UTI89 and att25 are
13
14
expected on mannose- or lactose-modified interfaces respectively (Table 2). Figure 7 shows a rather
15 different SPR signature for both pathogenic strains. An interesting behavior is observed in the case of
16
17 UTI89 (Figure 7A). After an increase in the SPR angle for the first 15 min, the SPR signal declines,
18 giving E. coli UTI89 a specific SPR signature. This decline is believed to be a consequence of the non-
19
20 covalently linked mannose to the Au/rGO SPR interface. Once in contact with the mannose-surface,
21 UTI89 integrates mannose into its pocket and detaches it from the surface. Confocal microscope
22
23 images recorded on mannose-terminated Au/rGO SPR interface after 15 min and 80 min confirm this
24 hypothesis (Figure 7C), where a rather different bacterial density can be visualized.
25
26 On the lactose-modified rGO, binding of the E. coli strain att25, specific to N-acetylglucosamine, a
27 comparable result was obtained in the sense that the bacteria disintegrate the SPR interface (Figure
28
29 7D), resulting in an overall negative shift of the SPR angle (Figure 7E). Disintegration appears very
30
rapidly with this adhesin, possibly due to the powerfull π−π stacking of the lactose molecule with
31
32 tryptophane (Trp) 108 and phenylalanine (Phe) 106 (Figure 3C). The other strains had a comparable,
33
basal level of bacterial adhesion on the lactose-modified interface. The overall level of SPR change is
34
35 also comparable to the other interfaces (about 80 milli) (Figure 5).
36
37
38
39 4. Conclusion
40
41 We have developed in this work a novel experimental concept for the investigation of pathogen
42
43 tropism in an easy manner. The approach is based on the use electrophoretically deposited reduced
44 graphene oxide films onto gold SPR chips with subsequent chemical modification of the rGO with
45
46 differently charged ligands such as positively-charged polyethyleneimine (PEI) and negatively-
47 charged poly(sodium 4-styrenesulfonate) (PSS) as well as with glycans such as mannose or lactose.
48
49 The summary of adhesion results reported in Figure 5A suggests that it is difficult to pinpoint the
50
observed different interactions towards charge discrimination alone. However, it could be concluded
51
52 that electrostatic interactions seem to be of higher importance than π−π stacking interactions and that
53
54 local charge variations close to the binding pockets of the bacteria seem to determine bacteria affinity.
55 Chemically based affinity effects have to be considered in addition. Glycan interfaces show indeed
56
57 specific binding events with sugar-specific E. coli strains such as UTI89 and att25. In the case of PSS
58 and PEI modified Au/rGO SPR interface, a detection limit of ≈102 cfu/mL was achieved for E. coli
59
60 11
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 12 of 26

1
2
3 107/86 and UTI89, respectively. The advantage of such an approach is thus multiple: in contrast to
4 bacteria sensing with bacteriophage where the size of the phage has an important implication on the
5
6 proportion of the target bacteria which can be probed by the evanescent field of the SPR,22 the
7
modified rGO SPR interfaces allow probing of the entire bacteria without any constrain. The
8
9 integration of affinity targets for different E. coli strains can be easily achieved through non-covalent
10
interactions between the rGO matrix and various organic ligands. The strategy is thus adaptable to any
11
12 other bacteria and more specifically if one is interested in examining bacterial affinities to molecules
13
14
on a surface and quantitative detection of pathogen in solution. It should be possible to design novel
15 experiments based on simplified model systems to answer many fundamental questions using this
16
17 technique combing a plasmonic read out and graphene based surface functionalization.
18
19
20
21 Acknowledgements
22
23
24 R.B. and S.S. gratefully acknowledge financial support from the Centre National de la Recherche
25 Scientifique (CNRS), the University Lille 1 and the Nord Pas de Calais region. S.S thanks the Institut
26
27 Universitaire de France (IUF) for financial support. P.S. acknowledges the financial support by the
28 European Commission in RTN-project MATCON (contract no. 238201). J.B. and N.Y. thank the
29
30 CNRS and the French Agence National de la Recherche for grant ANR-12-BSV5-0016-01.
31
32
33
34 References:
35
36 (1) Viazis, S.; Diez-Gonzalez Adv. Agronomy 2011, 111, 1-50.
37
38 (2) Sokurenko, E. V.; Chesnokova, V.; Dykhuizen, D. E.; Ofek, I.; Wu, X.-R.; Krogelt, K. A.;
39 Struve, C.; Schembri, M. A.; Hasty, D. L. Proc. Nat. Acad. Sci 1998, 95, 8922.
40
41 (3) Chen, S. L.; Hung, C. S.; Pinkner, J. S.; Walker, J. N.; Cusumano, C. K.; Li, Z.; Bouckaert, J.;
42 Gordon, J. I.; Hultgren, S. J. Proc. Nat. Acad. Sci 2009, 106, 22439.
43
44 (4) Londardi, E.; Moonens, K.; Buts, L.; de Boer, A. R.; Olsson, J. D. M.; Weiss, M. S.; Fabre, E.;
45
Guérardel, Y.; Deelder, A. M.; Oscarson, S.; Wuhrer, M.; Bouckaert, J. Biology 2013, 2, 894.
46
47
(5) Nishiyama, M.; Ishikawa, T.; Rechsteiner, H.; Glockshuber, R. Science 2008, 320, 376.
48
49
(6) Bouckaert, J.; Mackenzie, J.; de Paz, J. L.; Chipwaza, B.; Choudhury, D.; Zavialov, A.;
50
51 Mannerstedt, K.; Anderson, J.; Pierard, D.; Wyns, L.; Seeberger, P. H.; Oscarson, S.; De
52 Greve, H.; Knight, S. D. Mol. Microbiol. 2006, 61, 1556.
53
54 (7) Hung, C. S.; Bouckaert, J.; Hung, D.; Pinkner, J. S.; N., W. J.; Widberg, C.; DeFusco, A.;
55 Gale Auguste, C.; Strouse, R.; Langermann, S.; Waksman, G.; Hultgren, S. J. Mol. Microbiol.
56 2002, 44, 903.
57
58
59
60 12
ACS Paragon Plus Environment
Page 13 of 26 ACS Applied Materials & Interfaces

1
2
3 (8) Wellens, A.; Garofalo, C.; Nguyen, H.; Van Gerven, N.; Slättegård, R.; Hernalsteens, J. P.;
4 Wyns, L.; Oscarson, S.; De Greve, H.; Hultgren, S. J.; Bouckaert, J. PloS One 2008, 3(4):
5 e2040.
6
7 (9) Durka, M.; Buffet, K.; Iehl, J.; Holler, M.; Nierengarten, J.-F.; Taganna, J.; Bouckaert, J.;
8 Vincent, S. P. Chem. Commun. 2011, 47, 1321.
9
10 (10) Barras, A.; Martin, F. A.; Bande, O.; Baumann, J. S.; Ghigo, J.-M.; Boukherroub, R.; Beloin,
11
C.; Siriwardena, A.; Szunerits, S. Nanoscale 2013, 5, 2307.
12
13
(11) Lazcka, O.; Del Campo, F. J.; Munoz, F. X. Biosens. Bioelectron. 2007, 22, 1205.
14
15
(12) Belgrade, P.; Benett, W.; Hadley, D.; Richards, J.; Stratton, P.; Mariella, R.; Milanovich, F.
16
17 Science 1999, 284, 449.
18
19 (13) Li, Y.; Cu, Y. T. H.; Luo, D. Nat. Biotechnol. 2005, 23, 885.
20
21 (14) Yang, L.; Li, Y.; Erf, G. F. Anal. Chem. 2004, 76, 1107.
22
23
(15) Guo, X.; Kulkarin, A.; Doepke, A.; Hlsall, H. B.; Iyer, S.; Heinemann, W. R. Anal. Chem.
24 2012, 84, 241.
25
26 (16) Wang, Y.; Ye, Z.; Ying, Y. Sensors 2012, 12, 3449.
27
28 (17) Yang, L.; Bashir, R. Biotechnol. Adv. 2008, 26, 135.
29
30 (18) Mannoor, M. S.; Tap, H.; Clyton, J. D.; Sengupta, A.; Kaplan, D. L.; Naik, R. R.; Verma, N.;
31 Omenetto, F. G.; McAlpine, M. C. Nature Commun. 2012, 3, 763.
32
33 (19) Maalouf, R.; Fournier-Wirth, C.; Coste-J.; Chebib, H.; Saikali, Y.; Vittori, O.; Errachid, A.;
34 Cloarec, J.-P.; Martelet, C.; Jaffrezic-Renault, N. Anal. Chem. 2007, 79, 4879.
35
36 (20) Wan, Y.; Lin, Z.; Zhang, D. W.; Wang, Y.; Hou, B. Biosens. Bioelectron. 2011, 26, 1959.
37
38 (21) Yang, L. J.; Li, Y. B. Biosens. Bioelectron. 2005, 20, 1407.
39
40 (22) Oh, B.-K.; Lee, W. Y.; Chun, B. S.; Bae, Y. M.; Lee, W. H.; Choi, J.-W. Biosens. Bioelectron.
41 2005, 20, 1847.
42
43 (23) Gamella, M.; Campuzano, S.; Parrado, C.; Reviejo, A. J.; Pingarrón, J. M. Talanta 2009, 78,
44 1303.
45
46 (24) Mannoor, M. S.; Zhang, S.; Link, A. J.; McAlpine, M. C. Proc. Nat. Acad. Sci 2010, 107,
47 19207.
48
49 (25) Tawill, N.; Sacher, E.; Mandeville, R.; Meunier, M. Biosens. Bioelectron. 2012, 37, 24.
50
51 (26) Singh, A.; Glass, N.; Tolba, M.; Brovoko, L.; Griffiths, M.; Evoy, S. Biosens. Bioelectron.
52
2009, 24, 3645.
53
54
(27) Wu, M. C.; Deokar, A. R.; Lao, J.-H.; Shih, P.-Y.; Ling, Y.-C. ACSNano 2013, 7, 1281.
55
56
57
58
59
60 13
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 14 of 26

1
2
3 (28) Das, M. R.; Sarma, R. K.; Borah, S. C.; Kumri, R.; Saikia, R.; Deshmukh, A. B.; Shelke, M.
4 V.; Sengupta, P.; Szunerits, S.; Boukherroub, R. Colloids and Surfaces B: Biointerfaces 2012,
5 105, 128.
6
7 (29) Dreyer, D. R.; Park, S.-M.; Bielawski, C. W.; Ruoff, R. S. Chem. Soc. Rev. 2010, 39, 228.
8
9 (30) Szunerits, S.; Maalouli, N.; Wijaya, E.; Vilcot, J.-P.; Boukherroub, R. Anal. Bioanal. Chem.
10 2013, 405, 1435-1443.
11
12 (31) Subramanian, P.; Lesniewski, A.; Kaminska, I.; Vlandas, A.; Vasilescu, A.; Niedziolka-
13
Jonsson, J.; Pichonat, E.; Happy, H.; Boukherroub, R.; Szunerits, S. Biosens. Bioelectron.
14
15 2013, 50, 239-243.
16
17 (32) Salihoglu, O.; Balci, O.; Kocabas, C. App. Phys. Lett. 2011, 100, 213110.
18
19 (33) Wu, M.; Kempaiah, R.; Jimmy Hunag, P.-J.; Maheshwari, V.; Liu, J. Langmuir 2011, 27,
20 2731.
21
22 (34) Nielsen, N. O.; Clugston, R. E. Ann. N. Y. Acad. Sci. 1971, 176, 176.
23
24 (35) Hummers, W. S.; Offerman, J. R. E. J. Am. Chem. Soc. 1958, 80, 1339.
25
26 (36) Das, M. R.; Sarma, R. K.; Saikia, R.; Kale, V. S.; Shelke, M. V.; Sengupta, P. Colloid Surf. B:
27 Biointerfaces 2011, 83, 16.
28
29 (37) Wink, T.; Van Zuilen, S. J.; Bult, A.; Van Bennekom, W. P. Anal. Chem. 1998, 70, 827.
30
31 (38) Kooyman, R. P. H.; Lenferink, A. T. M.; Eenik, R. G.; Greve, J. Anal. Chem. 1991, 63, 83.
32
33 (39) Szunerits, S.; Maalouli, N.; Wijaya, E.; Vilcot, J. P.; Boukherroub, R. Anal. Bioanal. Chem.
34 2013, 405, 1435.
35
36 (40) Kaminska, I.; Barras, A.; Coffinier, Y.; Lisowski, W.; Niedziolka-Jonsson, J.; Woisel, P.;
37 Lyskawa, J.; Opallo, M.; Siriwardena, A.; Boukherroub, R.; Szunerits, S. ACS Appl. Mater.
38 Interfaces 2012, 4, 5386.
39
40 (41) Busscher, H. J.; van der Mei, H. PLOS Phathogens 2012, 8, e1002440.
41
42 (42) Terada, A.; Yuasa, A.; Kushimoto, T.; Tsuneda, S.; Katakai, A.; Tamada, M. Microbiol. 2006,
43
152, 3575.
44
45 (43) Arya, S. K.; Singh, A.; Naidoo, R.; Wu, P.; McDermott, M. T.; Evoy, S. Analyst 2011, 136,
46
47 486.
48
49 (44) Oh, B.-K.; Kim, Y. K.; Lee, W. H.; Bae, Y. M.; Lee, W. H.; Choi, J. W. Biosens. Bioelectron.
50 2003, 18, 605.
51
52 (45) Wilson, M.L.; Gaido, L. Clin Infect Dis. 2004, 38, 1150.
53
54
55
56
57
58
59
60 14
ACS Paragon Plus Environment
Page 15 of 26 ACS Applied Materials & Interfaces

1
2
3 Table 1: Physicochemical properties of rGO-modified SPR interfaces
4
5
6 Surface Surface 1 Surface 2 Surface 3 Surface 4
7
rGO rGO/PEI rGO/PSS rGO/mannose rGO/lactose
8
9
XPS C1s 77.3 64.9 68.6 56.7 57.3
10
11
12 O1s 22.7 24.7 29.1 43.3 42.7
13
14 other - 10.4 (N1s) 2.3 (S2p) - -
15
16
17 Water contact 58±1 80±1 70±1 15±3 12±3
18 angle / °
19
20 Zeta potential -32±3 +52±2 -57±2 -39±2 -38±2
21 of matrix/ mV
22
23
24
25
26
27
28 Table 2: Pathogenicity and characteristics of the five different E. coli strains investigated
29
30 E. coli strains Pathogenicity of E. coli strains Zeta-potential
31 (mV)
32 Characteristics of the fimbriae and fimbrial adhesins
33
34 UTI89 Isolate from patient with an acute bladder infection. -35±2
35
36 Type-1 fimbriae with mannose-binding FimH adhesin.
37
38 107/86 Causes oedema disease in just-weaned piglets. -24±2
39
40 F18 fimbriae with the FedF adhesin as the colonization factor:
41 basic residues juxtaposed to the binding site.
42
43 att25 Bovine isolate of enterotoxigenic, diarrhoeal E. coli. -30±2
44
45 F17 fimbriae with N-acetylglucosamine-specific adhesin.
46
47 UTI89 fimH- Isogenic to UTI89 apart from deletion of fimH. -36±2
48
49 No adhesion, thus abrogating mannose binding.
50
51 UTI89 Q133K UTI89 with Gln133 to Lys mutation in the binding pocket of -35±2
52 FimH, thus abrogating mannose binding
53
54
55
56
57
58
59
60 15
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 16 of 26

1
2
3 (A)
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34 e (B) (C)
35
36
37
38
39
40
41
Au/rGO
42
43
44
45
46
GO
47
48 292 290 288 286 284 282 280
Binding energy / eV
49
50
51
Figure 1: (A) Schematic illustration of the different SPR interfaces investigated; (B) SEM image of
52 the gold-based SPR interface coated with reduced graphene oxide through electrophoretic deposition
53
54 of GO; (C) C1s core-level XPS spectra of initial GO and of a Au/rGO-SPR interface.
55
56
57
58
59
60 16
ACS Paragon Plus Environment
Page 17 of 26 ACS Applied Materials & Interfaces

1
2
3
4
1
5
6
7 0,8
8
9
10
Reflectivity

11 0,6
12
13
14
15 0,4
16
17
18
19 0,2
20
21
22
23 0
24 66 68 70 72 74 76 78 80
25
26
Angle / °
27
Figure 2: SPR curves of the sensor before (black), after electrophoretic modification with rGO
28
29 forming Au/rGO (blue) and post-treatment of Au/rGO with mannose (green), lactose(orange), PEI
30
31 (violet) and PSS (red); electrophoretic deposition time: 20 s, DC voltage: 150 V. The dotted lines are
32 the experimental data and the solid lines are SPR fits: [nprism= 1.52, nTi=2.36+i3.47 (d=5 nm);
33
34 nAu=0.197+i3.67 (d=50 nm); ngraphene=3.0+i1.216; n(PSS)=1.45; λ=670 nm]
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 17
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 18 of 26

1
2
3 (A)
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21 (B)
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46 (C)
47
48
49
50
51
52
53
54
55
56
57
58
59
60 18
ACS Paragon Plus Environment
Page 19 of 26 ACS Applied Materials & Interfaces

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26 Figure 3. Schematic illustration of charge distribution on the different E. coli strains; (A) UTI89
27
28 Electrostatic surface of the mannose-binding site is depicted in blue (for positive charge and for
29 nitrogen atoms) and red (for negative charge and for oxygen atoms) for the FimH – butyl α-D-
30
31 mannose complex. (PDB entry 1TR7, that has an open gate (Tyr48-Tyr137) conformation where the
32
butyl chain travels through). Gln133 is indicated in cyan, this is an amino acid crucial for mannose
33
34 binding and was mutated to lysine in the Q133K strain. Amino acid ligands for the butyl α-D-mannose
35
and residues contributing to the negatively charged binding site of FimH are shown in ball-and-stick
36
37 model; (B) View on the electrostatic surface of the FedF lectin domain where it makes a complex with
38
39 blood group A type 1 hexasaccharide (PDB entry 4B4Q). The tip of FedF near the binding site is
40 highly seeded with positive charges (blue) from lysine and arginine residues. The fucose is in π-
41
42 stacking with Tyr49. Bound sulfate ions (sulfur in yellow) are shown (PDB entry 4BWO).; (C) F17a-
43 G in complex with N-acetylglucosamine (PDB entry 1O9W) is polar to electronegative (red) in charge
44
45 in the binding site, compensated by a positively charged patch (blue) underneath for which currently
46 no ligand is known. Aromatic π-stacking happens with Trp108.
47
48
49
50
51
52
53
54
55
56
57
58
59
60 19
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 20 of 26

1
2
3
4
5 (A) (B)
6 400
7 107/86
400
8 att25 107/86
350 att25
9 300 UTI89 UTI89
10 Q133K 300 Q133K
/ m°

-
11 UTI89 fimH
- UTI89 fimH

/ m°
250
SPR

12 200

SPR
∆Θ

13 200

∆Θ
14
150
15 100
16 100
17
50
18 0
19 0 10 20 30 40 50 60 70 80 0
20 time / min. interaction of different E. coli strain with rGO/PSS
21
22 (C)
23
24
25
26
27
28
29
30
31
32
33
(D)
34
35 400
36 107/86
37 350 att25
38
300 UTI89
39 Q133K
/ m°

40 250 -
UTI89 fimH
41
SPR

200
∆ΘB

42
43 150
44
45 100
46
47 50
48 0
49 4 6 8
1 100 10 10 10
50 [E. coli ] / cfu/ml
51
52 Figure 4: (A) SPR responses recorded on Au/rGO/PSS upon addition of different E. coli strains
53
54 (c=1.5×109 cfu/mL): 107/86 (blue), att25 (green), UTI89 (red), UTI89 Q133K (grey), UTI89 fimH-
55 (black); (B) Bar diagram of SPR responses on Au/rGO/PSS after 80 min for different E. coli strains;
56
57 (C) Confocal images of different bacteria strains (bright rod-shaped spots) on Au/rGO/PSS; (D)
58
59
60 20
ACS Paragon Plus Environment
Page 21 of 26 ACS Applied Materials & Interfaces

1
2
3 Calibration curve on Au/rGO/PSS SPR interfaces for different concentrations of the different E. coli
4 strains.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 21
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 22 of 26

1
2
3 (A)
4
400
5
107/86
6 att25
350
7 UTI89
8 Q133K
300
9 -
UTI89 fimH
10
/ m° 250
11
12
SPR

200
13
∆Θ

14
150
15
16
100
17
18
50
19
20
0
21
22
rGO/PSS

rGO/lactose
rGO/mannose
23
rGO/PEI

Au

rGO
24
25
26
27
28
29 (B)
30 400
31
107/86
32 att25
350
33 UTI89
34 Q133K
300
35 -
UTI89 fimH
36
37 250
/ m°

38
SPR

39 200
40
∆Θ

41 150
42
43 100
44
45 50
46
47 0
48
49
Au/PSS

Au/mannose

Au/lactose

50
Au/PEI
Au

51
52
53
54
55
56 Figure 5. Bar diagramm resuming all the recorded change of SPR responses upon addition of
57 1.5× 109cfu/mL E. coli strains onto differently modified (A) rGO and (B) gold interfaces
58
59
60 22
ACS Paragon Plus Environment
Page 23 of 26 ACS Applied Materials & Interfaces

1
2
3 (Five replicate measurements were performed for each experiment): 107/86 (blue), att25
4
5 (green), UTI89 (red), UTI89 Q133K (grey), UTI89 fimH- (black).
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 23
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 24 of 26

1
2
3
4
(A) (B)
5
6
7 400
400
8 107/86 PEI
9 350 att25 107/86
350 att25
10 300 UTI89 UTI89
11 Q133K 300 Q133K
/ m°

12 250 UTI89 fimH- UTI89 fimH

-

/ m°
250
SPR

13
200

SPR
14
∆Θ

200

∆Θ
15 150
16 150
17 100
100
18 50
19 50
20 0
0 10 20 30 40 50 60 70 80 0
21
22 time / min. interaction of different E. coli strains with rGO/PEI
23
24
(B) (D)
25
26
27 350
400
28 107/86 107/86
29 300 att25 350 att25
30 UTI89 UTI89
250 Q133K 300 Q133K
31
/ m°

-
UTI89 fimH
32 -
200 UTI89 fimH
/ m°

250
SPR

33
SPR
∆ΘB

34 150
200
∆Θ

35
150
36 100
37 100
38 50
50
39
40 0 4 6 8 0
41 1 100 10 10 10 52 40 15 5 -12
42 [E. coli ] / cfu/ml zeta-potential / mV
43
44
Figure 6. (A) SPR responses recorded on Au/rGO/PEI interfaces upon addition of different E. coli
45
46 strains (c=1.5×109 cfu/mL): 107/86 (blue), att25 (green), UTI89 (red), Q133K (grey), UTI89 fimH-
47
48 (black); (B) Bar diagram of SPR responses on Au/rGO/PEI after 80 min for different E. coli strains;
49 (C) Calibration curve on Au/rGO/PEI SPR interfaces for different concentrations of the different E.
50
51 coli strains; (D) Change in SPR responses on Au/rGO/PEI by changing the amount of PEI integrated
52 onto the interface and thus the zeta-potential of the matrix
53
54
55
56
57
58
59
60 24
ACS Paragon Plus Environment
Page 25 of 26 ACS Applied Materials & Interfaces

1
2
3
4
5 (A) (B)
6
7
400
8 400
107/86 107/86
9 350 att25 mannose 350 att25
10 UTI89

11 300 UTI89 300 Q133K

Q133K -
/ m°

UTI89 fimH
12

/ m°
250 UTI89 fimH- 250
13
SPR

SPR
200 200
14
∆Θ

∆Θ
15 150
150
16 100
17 100

18 50
50
19 0
0 10 20 30 40 50 60 70 80
20 0
21 time / min. interaction of different E. coli strains with rGO/mannose
22
23 (C)
24
25
26
27
28
29
30
31
32
33
34
35
36
37 (D) (E)
38
39
40 240 400
107/86 lactose
41 107/86
200 att25 350
42 UTI89 att25
UTI89
43 160 Q133K 300
/ m°

Q133K
44 -
UTI89 fimH 250 UTI89 fimH
-
/ m°

120
SPR

45 200
SPR
∆Θ

46 80
∆Θ

150
47
40
48 100
49 0 50
50
-40 0
51 0 10 20 30 40 50 60 70 80
52 time / min. interaction of different E. coli strains with rGO/lactose
53
54
55 Figure 7. (A) SPR responses recorded on Au/rGO/mannose interfaces upon addition of different E.
56
coli strains (c=1.5×109 cfu/mL): 107/86 (blue), att25 (green), UTI89 (red), Q133K(grey), UTI89 fimH-
57
58 (black); (B) Bar diagram of SPR responses on Au/rGO/mannose; (C) Confocal images of E. coli strain
59
60 25
ACS Paragon Plus Environment
 ACS Applied Materials & Interfaces Page 26 of 26

1
2
3 UTI89 after interaction for 15 min (left) and 80 min (right) on Au/rGO/mannose; (D) SPR responses
4 recorded on Au/rGO/lactose interfaces upon addition of different E. coli strains (c=1.5×109 cfu/mL):
5
6 107/86 (blue), att25 (green), UTI89 (red), UTI89 mut (grey), UTI89 fimH- (black); (E) Bar diagram of
7
SPR responses on Au/rGO/lactose.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 26
ACS Paragon Plus Environment