USING SURFACE PLASMON RESONANCE IMAGING (SPRi)

TO STUDY BIOFILMS AND BIOFOULING

A Dissertation Presented

By

Pegah Naghshriz Abadian

to

The Department of Chemical Engineering

in partial fulfillment of the requirements

For the degree of

Doctor of Philosophy

In the field of

Chemical Engineering

Northeastern University

Boston, Massachusetts

March 2, 2016

i

In loving memory of my grandmother, Zari Najati (1943–2015).

ii

ACKNOWLEDMENT

I would like to express my special appreciation and thanks to my advisor, Professor

Edgar D. Goluch. Thank you Prof. Goluch for providing me guidance, and expertise

whenever I needed, for supporting me very patiently during the last five years. I am really

grateful that you gave me the freedom to work on the area I was more interested in and for

encouraging me in collaborative works. Most importantly for being a person who I would

first count on when I needed support. Your patience, flexibility, genuine caring and

concern made my PhD experience a great one.

I would like to extend my thanks to Professor Thomas Webster, Professor Eno

Ebong of the Chemical Engineering Department and Professor Yunrong of the Biology

Department for being on my dissertation committee. I gratefully acknowledge the members

of my Ph.D. committee for their time and valuable feedback on preliminary versions of

this thesis.

I would like to thank my labmates, with whom I shared a lot of moments, Dr.

Thaddaeus Webster, Nil Tandogan, Hunter Sismaet, and Martin Kimani. Thank you for

making this journey very memorable and enjoyable. A good support system is important

to surviving and staying sane in graduate school and I am proud to call you my friends.

This dissertation would not have been possible without the help of the numerous

undergraduate students, high school students and teachers with whom I had the pleasure

of interacting with during my time in the Goluch Group. While there are many I would like

to highlight Catherine Reiter, John Jamieson, and Chase Kelley for helping me in the

experiments with great enthusiasm.

iii

Thanks to William Fowle of the Biology Department for training me on the proper

protocols of biological sample preparation and helping me grow to be so skilled in the use

of his Scanning Electron Microscope.

I also thank Robert Eagan of the Chemical Engineering Department for helping me

make all of the complicated pieces that I used in this work, and for being very responsive

and enthusiastic in his support.

Last, but certainly not least, I must acknowledge with tremendous and deep thanks

my family, especially my Mom, Shamila. Thank you for guiding me to choose the right

priorities in life and always providing me the best support in all steps of my life. Thank

you for listening to all my stories at 2 am in the morning hundreds of miles away. For

motivating me all the time when I was doubtful about the path I chose. For making me stay

positive in the hardest time. For being an amazing Mother.

iv

ABSTRACT

Surface Plasmon Resonance imaging (SPRi) is a label-free detection method with

the capability of real-time detection of multiple interactions occurring simultaneously on a

gold surface. In this work, SPRi was used for the first time to study bacterial

adhesion/growth, biofilm formation/disassembly, and cleaning of biofouled surfaces.

These processes are important to study because biofilms are reservoirs of bacteria and a

source of endotoxins, which both can enter the circulation system of a patient and cause

systemic disorders. More than 60% of hospital-acquired infections are caused by bacterial

biofilms. Formation of biofilms is the main cause of many bacterial infections.

SPR detection is based on changes in the refractive index at the sensing surface

caused by changes in the composition of the material directly above (~200 nm) the sensor

surface. Unlike in traditional SPR where a single point on a surface is measured, SPR

imaging allows the rapid collection of information about refractive index changes and the

location of these events with high precision (~10 µm) over a large area (~1 cm2)

simultaneously.

Using a SPRi system, physiological behavior of bacterial cells and biofilm

dynamics was monitored in real-time. This information were used to help predict and

control bacteria activity in fluidic systems. Studies were conducted to determine the

effectiveness of different chemicals and antibiotics in removing biofilm from a sensor

surface. The efficacy of antibiotics and surface coatings for preventing biofilm formation

on the surface were also studied. Finally, the effects of fluid dynamics on bacterial suface

adhesion and removal was investigated.

v

Staphyloccocus aureus, a gram positive bacteria and one of the major causes of

hospital aquired infections, Pseudomonas aeruginosa, a gram negative species and model

organism for biofilm studies, Eschericia coli, a gram negative and a model prokaryotic

organism, and Bacillus cereus a gram positive and facultative anerobic bacteria, were used

in this study.

vi

TABLE OF CONTENTS

LIST OF FIGURES…………………………………………………………………….……….ix

1.0 INTRODUCTION……………………………………………………………………....1

2.0 CRITICAL LITERATURE REVIEW……………………………………....………...3

2.1. Technology ……………………………………………………………………………….3

2.1.1. Surface Plasmon Resonance (SPR) Sensors………………...…………………....3

2.1.1.1. SPR phenomena……………………..…………………………………..………..3

2.1.1.2. SPR condition…………………………………………………………………….7

2.1.1.3. SPR sensing method……………………………………………………………..10

2.1.2. Surface Plasmon Resonance imaging (SPRi)…………………………………....13

2.2. Cell Detection with SPR sensors…………………………………………………………14

2.2.1. Bacterial Cell Detection with Surface Plasmon Resonance Imaging………..…..14

2.2.2. Mammalian Cell Detection with Surface Plasmon Resonance Imaging…………36

2.3. Biology….………………………………………………………………………….........44

2.3.1. Biofilms…………………………………………………………………………44

2.3.1.1. E. coli……………………………………………………………………………46

2.3.1.2. P. aeruginosa…………………………………………………………………....47

2.3.1.3. S. aureus…………………………………………………………………...……49

2.3.1.4. Bacillus species (Bacillus cereus)…………………………………………...….50

3.0 DISSERTATION GOALS……………………………………………………………52

4.0 EXPERIMENTAL……………………………………………………………………..54

4.1. System Setup for Monitoring Bacterial Growth and Biofilm Formation (Goal 1)………..54

4.1.1. Monitoring Changes on the Surface with SPRi…………………………………54

4.1.2. Monitoring Bacterial Growth with SPRi…………………………………...…...55

4.1.3. Monitoring Bacterial Biofilm Formation……………………………………….56

4.2. Prevention of Biofilm Formation on the Surface (Goal 2)………………………………58

4.2.1. Surface Coatings………………………………………………………………...58

4.2.2. Loading Antibiotics in Solution………………………………………………...60

4.3. Biofilm Removal from the Surface (Goal 3)…………………………………………….63

vii

4.3.1. Cleaning with Different Chemical Compounds………………………………...63

4.3.2. Disinfection with Antimicrobial Components………………………………......64

4.4. Effect of flow rate on bacterial growth (Goal 4)………………………………………...66

4.4.1. SPRi Experiments…………………………………………………………..…...66

4.4.2. COMSOL Multiphysics Modeling…………………………………………...…67

5.0 RESULTS AND DISCUSSION……………………………………………...………68

5.1. System Setup for Monitoring Bacterial Growth and Biofilm Formation (Goal 1)……...69

5.1.1. Bead Imaging……………………………………………………………………69

5.1.2. Cell Growth and Biofilm Formation……………………………………………71

5.1.2.1 E. coli growth and biofilm formation…………………………………………….71

5.1.2.2 P. aeruginosa growth and biofilm formation……………………………….…….76

5.2. Prevention of Biofilm Formation on the Surface (Goal 2)………………………………81

5.2.1. Surface Coating…………………………………………………………………81

5.2.1.1. Casein and BSA……………………...………………………………………….81

5.2.1.2. Penicillin/Streptomycin and BSA……………………………………………......89

5.2.1.3. Penicillin/Streptomycin and Casein………………………………………….....94

5.2.2. Loading Antibiotics in Solution………………………………………………...95

5.2.2.1 Control Experiment…………………………………………………………......95

5.2.2.2 Penicillin/Streptomycin (S. aureus)……………………………………...……...97

5.2.2.3 Colistin (P. aeruginosa)…………………………………………………...……..99

5.2.2.4 Spectinomycin (B. cereus)………………………………………………….…..101

5.3. Biofilm Removal from the Surface (Goal 3)…………………………………………...104

5.3.1. Cleaning with Different Chemical Compounds……………………………….104

5.3.2. Disinfection with Antimicrobial Components………………………………....108

5.4. Effects of flowrate on Bacterial Adhesion (Goal 4)……………………………………..111

5.4.1. SPRi Experiments……………………………………………………………...111

5.4.2. COMSOL Multiphysics Modeling ………………………………………....…115

6.0 Conclusions and Future work………………………………………………………..117

6.1. Goal 1: System Setup for Monitoring Bacterial Growth and Biofilm Formation……...117

viii

6.2. Goal 2: Prevention of Biofilm Formation on the Surface……………………………….118

6.3. Goal 3: Biofilm Removal from the Surface…………………………………………….120

6.4. Goal 4: Effects of flow rate on Bacterial growth………………………………………...122

7.0. METHODS…………………………………………………………………….…...…124

7.1. PDMS Fabrication……………………………………………………………………...124

7.2. Bacterial Culture Preparation………………………………………………………..…127

7.3. Bacterial Culturing……………………………………………………………………..127

7.4 Sample preparation for Scanning Electron Microscopy (SEM)………………………..129

8.0 REFERENCES……………………………………………………………………….131

ix

LIST OF FIGURES

Figure 1: Refraction of light at different incident angles…………….....................5

Figure 2: (a) Otto Configuration, (b) Kretschmann Configuration….………..…2

Figure 3: Schematics of (a) TE and (b) TM polarization of light……………….…7

Figure 4: Excitation of SPs on the thin metal film…………………………………8

Figure 5: (a) SPR wavelength configuration curve, which shows the intensity of

the reflected light versus the incident wavelength. (b) The graph is

plotted the coupling wavelength versus the refractive index of the

sample above the prism……………..…………......................................11

Figure 6: a) SPR angular configuration curve, which shows the intensity of the

reflected light versus the incident angle. (b) The graph is plotted the

coupling angle versus the refractive index of the sample above the

prism…………………………………………………….………………12

Figure 7: Difference image from the chip surface generated by the SPRi device.

The bright spots represent the sections functionalized with (S) specific

antibodies and (N) non-specific antibodies. The dark section shows the

uncoated gold surface…………………………………………………..13

Figure 8: Steps of biofilm formation…………………………………………...…44

Figure 9: Scanning Electron Microscopy image of E. coli………………………47

Figure 10: Scanning Electron Microscopy image of Pseudomonas aeruginosa.....49

Figure 11: Scanning Electron Microscopy image of Staphylococcus aureus……..50

Figure 12: Scanning Electron Microscopy image of Bacillus subtilis…………….51

Figure 13: Setup for initial Surface Plasmon Resonance imaging (SPRi)

experiments. 50 m beads in DI water placed onto a prism coated with

50 nm of gold……………………………………………………….……55

Figure 14: Setup for Surface Plasmon Resonance imaging (SPRi) experiments.

PDMS with two channels. The left channel was filled with LB growth

media and the right channel will be filled with GFP labeled E.

coli……………………………………………………………………….56

x

Figure 15: Setup for monitoring biofilm formation using Surface Plasmon

Resonance imaging (SPRi). PDMS with three channels. The left

channel was filled with trypticase soy broth, middle channel was filled

with PelA mutant P. aeruginosa PA14 and the right channel was filled

with wild type PA14….............................................................................57

Figure 16: Setup for coating experiments using SPRi. The gold surface inside the

chamber is coated with two biomolecules……………………………...60

Figure 17: Setup for studying the effect of antibiotics on prevention of biofilm

formation using SPRi………………………....………………………...62

Figure 18: Setup for studying bacterial growth under different flowrates using

SPRi……………………………………………………………………...67

Figure 19: (Right column) SPRi images of 50µm beads, (left column) stereo

microscope fluorescent images of the same beads…........……………..70

Figure 20: (Top) Schematic of the setup for E. coli SPRi experiments. A PDMS

chip containing two microchambers is reversibly sealed against the

sensor surface. (Bottom) SPR images of LB, and GFP E.coli filled

channels(a-f) Fluorescence images of the same channels (g,h)………..74

Figure 21: Difference images taken with SPRi at different time points. The arrows

are pointing to the center of the GFP labeled E. coli bacterial media

droplet, where the bacteria preferred to gather. j) A fluorescent image

of the surface of the prism surface after being removed from the SPRi

system……………………………………………………………………75

Figure 22: (Top) Schematic of the biofilm formation experiments using SPRi.

PDMS, with three channels, is reversibly sealed against a high

refractive index glass prism coated with 50 nm of gold. (Bottom)

Difference images of SPRi in after 3hours…..…………………………77

Figure 23: Images of P. aeruginosa PAO1 after being grown overnight in LB

growth media. (a-h) SPRi images after overnight growth. (i) GFP-

filtered fluorescence image of the right side of the dried biofilm on the

sensor surface. (j) SEM images of the right side of the biofilm. (k) SEM

image of the center of the chamber…….......…………………………..80

Figure 24: SPRi difference images of P. aeruginosa (CFP-PA01) growth on the

gold surface coated with BSA (left) and casein (right) after (a) 6, (b) 12,

(c) 18, and (d) 24 hours……………………....………………………….83

xi

Figure 25: SPRi difference images of S. aureus growth on the gold surface coated

with BSA (left) and casein (right) after (a) 6, (b) 12, (c) 18, and (d) 24

hours….....................................................................................................86

Figure 26: Scanning electron microscope (SEM) images of the coated and

uncoated sensor surfaces after 6 hours of exposure to flowing solutions

containing to S. aureus. (a) Low-magnification and (b), (c) high-

magnification images of the boundary between BSA and bare gold. (d)

Low magnification and (e), (f) high-magnification images of the

boundary between bare gold and casein…………………………….…88

Figure 27: SPRi difference images of P. aeruginosa (CFP-PA01) growth on the

gold surface coated with BSA (left) and penicillin/streptomycin (right)

after (a) 6, (b) 12, (c) 18, and (d) 24 hours……………..……………….91

Figure 28: SPRi difference images of S. aureus growth on the gold surface coated

with BSA (left) and penicillin/streptomycin (right) after (a) 6, (b) 12, (c)

18, and (d) 24 hours……………......……………………………………92

Figure 29: SPRi difference images of P. aeruginosa growth on the gold surface

coated with casein (left) and penicillin/streptomycin (right) after (a) 6,

(b) 12, (c) 18, and (d) 24 hours…….......………………………………...94

Figure 30: SPRi difference images of S. aureus growth on the sensor surface with

continuous LB flow over the surface after a) 35’, b) 330’, c) 635’, d)

1170’…………………………………………………………………......96

Figure 31: SPRi difference images of the surface. The left column shows the

images from S. aureus growth on the chamber without having any

antibiotic in the inlet media as a control, right column is shows the

difference images at the same time points by running

penicillin/streptomycin from the beginning of the experiment……….98

Figure 32: SPRi difference images of P. aeruginosa growth with and without

antibiotics. The left column shows the images of P. aeruginosa growth

in the chamber when running Colistin from the beginning of the

experiment, right column is shows the difference images at the same

time points without having any antibiotic added to the inlet

media.......................................................................................................100

Figure 33: SPRi difference images of B. cereus growth with and without antibiotics.

The left column shows the images of B. cereus growth in the chamber

xii

when running Spectinomysin solution from the beginning of the

experiment, right column shows the difference images at the same time

points without having any antibiotic added to the inlet

media………………………………………………………..………….102

Figure 34: Quantitative analysis showing the reflectivity changes over time as B.

cereus bacteria grow on the surface in the presence (orange line) and

absence (blue line) of Spenctinomycin antibiotic…………..………...103

Figure 35: Biomass coverage on the sensor surface at different time point. The

orange columns represent B. cereus growth on the surface in the

presence on antibiotics in the solution, the blue columns represent B.

cereus growth in the normal solution sans antibiotics……………….103

Figure 36: S. aureus growth and removal with 1%SDS from the sensor surface

during the experiment period…………………………………………105

Figure 37: P. aeruginosa growth and removal with 1%SDS from the sensor

surface during the experiment period…………………………..……106

Figure 38: B. cereus growth and removal with 1%SDS from the sensor surface

during the experiment period………………………….……………...107

Figure 39: The effect of penicillin/streptomycin on treatment of S. aureus

biofilm……………………………………………………………...…..109

Figure 40: The effect of spectinomycin on treatment of B. cereus biofilm………110

Figure 41: The average reflectivity change in the difference image at each time

point as a result of B. cereus growth under different flowrates.

Blue=stagnant condition, Orange=10µl/min, and gray=40µl/min

flowrates……………………………………………………………….112

Figure 42: SPRi difference images showing S. aureus growth under 10µl/min (left

column) and 120µl/min (right column) flowrates after (a,b) 6 hours,

(c,d) 12 hours, (e,f) 18 hours, and (g,h) 24 hours……………………..114

Figure 43: COMSOL Multiphysics modeling of the shear stress distribution on the

sensor surface. The red color indicates the highest and the blue color

represents the lowest shear stress on the surface…………………….116

Figure 44: (A) Schematic of the SPRi setup for antibiotic resistance experiments.

(B) The average brightness change in the channels filled with LB

xiii

without cells (diamonds), or 5.4E+6 cells/mL S. aureus with no

antibiotic (purple), with 1000X diluted antibiotic (blue), with 200X

diluted antibiotic (sloping lines)………………….………………...…119

Figure 45: Channel design to study biofilm formation under different flowrate

and in non-uniform structures. Bends are in a) 90 degree angle and b)

30 degree angles…………………………….....……………………….123

Figure 46: Schematic of the fabrication of a PDMS chamber. a) The hexagonal

mold made of Aluminum, b) PDMS is poured on the mold and cured

in the over, c) The cured PDMS in peeled off from the mold, d) each

mold contain 6 hexagons, which in the step are cut separately, e) one

hexagon chamber f) is placed on the gold coated prism……………...125

Figure 47: Schematic of the fabrication of PDMS channel. a) The three and two

linear channels on the silicon wafer, b) PDMS is poured on the mold

and cured in the oven. c) The cured PDMS is peeled off from the mold.

d) Each mold contains several groups of channels, which are cut and

separated. e) A PDMS piece containing two separate linear channels f)

is placed on the gold coated prism…………………….………………126

Figure 48: Staphylococcus aureus bacteria cultured on a LB agar plate….…….128

1

1.0 INTRODUCTION

There is a constant need for rapid detection of pathogenic bacteria in the areas of

food, water, and public health to ensure environmental safety, prevent illness, and

economic loss due to bacterial infection and contamination.

Based on estimates from the Centers for Disease Control and Prevention (CDC), in

the United States, foodborne pathogens cause roughly 79 million illnesses, 325000

hospitalizations, and almost 5000 deaths each year [1]. Among all foodborne pathogenic

microorganisms, bacteria are responsible for 91% of all foodborne illnesses in USA [2, 3].

Outbreaks of foodborne illnesses cost billions of dollars each year. According to the United

States Department of Agriculture (USDA) Economic Research Service (ERS), medical

costs and loss of productivity caused by five major pathogens, E. coli O157:H7, non-O157

STEC (Shiga Toxin Producing Escherichia coli), Salmonella (non-typhoidal serotypes

only), Listeria monocytogenes, and Campylobacter is $6.9 billion annually [4, 5].

In order to decrease the health risks, deaths, and reduce economic losses due to

pathogenic bacteria there is an essential need for rapid, sensitive, and selective detection

methods to sense the disease-causing bacteria in food and beverages [6]. Various

techniques have been employed for pathogenic bacteria detection [7], such as conventional

microbiological culture methods [8], Polymerase Chain Reaction (PCR) [9-16], Enzyme-

linked immunosorbent assays (ELISA) [17, 18], amperometric biosensors [19-23],

piezoelectric biosensors [23-26], potentiometric biosensors [27, 28], bioluminescence [29,

30], fluorescent labeling [31, 32], and ultrasound [33]. All of these methods have concerns,

such as long detection times, enrichment requirements, labelling steps, high costs, and

trained personnel to run them [34, 35]. Therefore, there is a need for alternative rapid, low

2

cost, and sensitive detection techniques. These needs can be fulfilled by Surface Plasmon

Resonance (SPR), which provides label-free, real-time, and quantitative detection and can

monitor interaction between biomolecules continuously [36-38].

SPR sensors are optical sensors that employ variations in surface plasmons on

surface of gold coated sensor chips as their detection principle. Surface plasmons are

sensitive to changes (due to biomolecular interactions on the surface) of the local refractive

index within approximately 200 nm of the sensor surface. The changes on the refractive

index are measured using optical signals, which are detected by the device and the results

expressed as Resonance Units (RU) versus time [39-41].

In this work, bacterial behavior such as bacterial growth and biofilm formation will

be studied in real-time with a surface plasmon resonance imaging (SPRi) instrument. SPRi

sensors allow for multiple detection areas on the sensor surface, and we will use this feature

to study the preventative effects of various coatings on biofilm formation. In addition, the

effect of various antibiotics on prohibiting biofilm formation on the sensor surface will be

investigated in real-time. This information is of great importance to both medical and

industrial applications.

3

2.0 CRITICAL LITERATURE REVIEW

2.1. Technology

In this section, Surface Plasmon Resonance techniques will be discussed, including

the working principles and different common variations of this detection method. Also,

work done by other groups on cell detection with SPR sensors will be presented.

2.1.1. Surface Plasmon Resonance (SPR) Sensors

In the past decade the use of SPR-based biosensors has increased significantly [42].

Surface Plasmon Resonance (SPR) based sensors provide highly sensitive, label-free

detection with the capability for real-time monitoring of surface phenomena at the

molecular level [36-38]. Sensors based on SPR phenomena provide promising tools for

use in biomolecular level applications. SPR techniques are used in various applications for

measuring film thickness, binding kinetics, molar concentrations, and provide a highly

sensitive tool with the potential to be used in wide range of diagnostics devices. These

sensors are commonly used for qualitative and quantitative studies of binding kinetics and

binding affinity between two biomolecules: antigen-antibody [43-49], DNA-DNA [50-54],

DNA-protein [55-59], RNA-DNA [60, 61], and carbohydrate-protein [62-65].

2.1.1.1. SPR phenomena

The SPR phenomenon begins with the formation of Surface Plasmons (SPs). SPs

are charge density oscillations of free conductive electrons. When SPs propagate at the

interface between two media with different refractive indices, they are referred to as

Surface Plasmon Polaritons (SPPs) [39-41]. SPR sensors consist of a prism made of high

4

refractive index glass coated with 50 nm of a noble metal. In SPR experiments, gold and

silver are the most commonly used metals [66, 67]. Gold is an inert and biocompatible

metal usually used for SPR-based biological studies. Silver has better sensitivity, but it has

less stability because it oxidizes readily [68].

To describe SPR phenomena, first consider the situation without having the thin

metal film, which is known as Total Internal Reflection (TIR). When light passes from a

material with higher refractive index (prism) to a material with lower refractive index (air),

based on the incident angle value, one of two conditions occur. If the angle of the incident

light is smaller than the critical angle, some portion of the energy will reflected back and

some portion will exit from the prism surface. TIR also occurs when the incident angle is

larger than the critical angle (Figure 1). In TIR, all of the incident light is reflected back

inside the prism. The critical angle occurs when the incident angle is reflected back at an

angle where the reflected light is parallel to the surface of the material. The value of the

critical angle depends on difference in the refractive index between the two media and can

be calculated from Snell’s law (Equation 1).

5

Figure 1: Refraction of light with different incident angle. When θincident<θcritical

smaller than the critical angle the light will transmit from the medium

with higher refractive index (n1) to less denser medium (n2). TIR occur at

θincident>θcritical.

Snell’s law to calculate the critical angle:

𝑛1 Sin θ1 = 𝑛2 Sin θ2 , 𝜃𝑐 = Sin−1 (𝑛2

𝑛1) (1)

Where: θ1 is the incident angle,

θ2 is the refracted angle,

n1 is the refractive indices of the denser medium

n2 is the refractive indices of the less dense medium.

Now when a thin metal film is coated on the prism surface, the SPR phenomena is

observed when an TIR light orientation is employed. In this case, while all the incident

light is reflecting back, it also generates an evanescent field. An evanescent field is an

6

inhomogeneous electromagnetic wave in which the wavelength is identical to the incident

light. The amplitude of the field decreases exponentially at the rate proportional to 1/e in

the direction perpendicular to the thin metal surface toward the dielectric and extends 200

nm above the metal surface. This evanescent wave can couple with the SPs in the thin

metal film at the metal-dielectric interface and excite them to generate SPPs.

Kretschmann and Otto are two pioneers in demonstrating optical excitation of

Surface Plasmon Polaritons. Based on their respective designs, there are two different

configurations for SPR: Kretschmann and Otto. In the Otto configuration, the thin metal

film is located close to the prism and the sample passes between the metal film and the

prism (Figure 2a) [69, 70]. The Kretschmann set up is the more commonly used

configuration in SPR experiments. In this configuration, which is used in this study as well,

the thin metal film in coated on the prism surface while the fluid passes over the metal. In

both configurations, the incident light excites the SPs by interacting with the evanescent

field within the thin metal film, resulting in SPR (figure 2b) [71-73].

Figure 2: (a) Otto Configuration, (b) Kretschmann Configuration.

7

2.1.1.2. SPR condition

SPs can only oscillate in the direction perpendicular to the metal-dielectric

interface. Therefore, the Transverse Magnetic (TM) polarized light, in which the electric

field oscillates normal to the metal film surface, can excite the SPs. The electric field of

the Transverse Electric (TE) polarization of light, which is parallel to the thin metal plane,

and subsequently parallel to the SPs, cannot excite oscillating free electrons. The schematic

of the two different polarizations of light is illustrated in Figure 3.

Figure 3: Schematics of (a) TE and (b) TM polarization of light.

As mentioned above, at SPR conditions, energy from the incident light in the form

of an evanescent wave will excite the Surface Plasmons. In order to couple the energy with

SPs, the wave vector of the SPs (Ksp) should match the wave vector of the incident light

8

(Kx) which is parallel to the metal-dielectric interface as shown in Figure 4 (Kx=Ksp). The

wave vector of the incident light is expressed by Equation 2 [74-77].

Figure 4: Excitation of SPs on the thin metal film occurs when the wave vector of the

incident light (kx) match the wave vector of the SPs (ksp).

𝑘𝑥 = 𝑘0𝑛𝑝𝑟𝑖𝑠𝑚sin θ𝑖𝑛 (2)

Where: k0=2π/λ is the free space wave vector,

nprism is the refractive index of the prism and

θin is the incident angle.

The wave vector for the SP is expressed in Equation 3.

𝑘𝑠𝑝 = Re {𝐾0√𝜀𝑚.𝜀𝑠

𝜀𝑚+𝜀𝑠} (3)

9

Where: K0=2π/λ is the free space wave vector

εm is the complex dielectric constant of the thin metal film

εs is the complex dielectric constant of the sample.

Ksp can be written as shown in Equation 4. By neglecting the imaginary part of the

dielectric constants,

𝑘𝑠𝑝 = 𝑘0√𝑛𝑚

2 .𝑛𝑠2

𝑛𝑚2 +𝑛𝑠

2 (4)

Where: K0=2π/λ is the free space wave vector

𝑛𝑚 = √𝜀𝑚 is the refractive index of thin metal film

𝑛𝑠 = √𝜀𝑠 is the refractive index of the sample

In a standard SPR experiment, if the properties of the incident light, and

consequently Kx vector, do not change, the resonance condition will depend on the optical

properties of the sample and the metal. This means any changes to the refractive index due

to attachments at the surface will change the resonance.

At the resonance condition energy from the photons of the incident light will be

transferred to the metal free electrons to excite the surface plasmons this will reduce the

intensity of the reflected light. Therefor SPR can be measured at certain wavelength or

angle of the incident light. Therefore, the resonance angle/wavelength is the

angle/wavelength at which a drop occurs in the intensity curve of the reflected light and

the reflectivity is at a minimum.

10

2.1.1.3. SPR sensing method

In all SPR experiments based on the measured property of the reflected light, three

different modulations exist:

Wavelength modulation: In wavelength modulation a white light, which includes

different wavelengths, is used. The incident angle is fixed and light with different

wavelengths enters the prism. The reflected light is gathered with a CCD camera and the

spectral properties of it are analyzed. Reflected light exhibits changes in the intensity and

at the wavelength where SPR occurs an adsorption dip in the reflectivity curve is observed.

The wavelength at which the dip in the intensity of the reflected light occurs is known as

the coupling wavelength (λsp). The SPR coupling wavelength shifts when the refractive

index of the sample changes. Figure 5a shows the reflectivity versus wavelength curve and

represents the SPR coupling curve and Figure 5b shows the shift due to variation of the

sample refractive index [78].

11

Figure 5: (a) SPR wavelength configuration curve, which shows the intensity of the

reflected light versus the incident wavelength, in which the minimum

intensity occurs at the coupling wavelength (λspr). (b) The graph is

plottedshows the coupling wavelength versus the refractive index of the

sample above the prism.

Angular modulation: In this modulation, a monochromatic light source is used and

the intensity of the reflected light is measured at a range of incident angles. At the SPR

condition, a dip in the intensity of the reflected light will be observed. This angle is known

as resonance angle (θr). The same way as with the coupling wavelength, when the refractive

0

0.5

1

500 600 700 800 900

No

rm.

Ref

lect

ion I

nte

nsi

ty

Wavelength (nm)

655

660

665

670

675

1.333 1.334 1.335 1.336

Coupli

ng W

avel

ength

(nm

)

Reflection Index

(a)

(b)

12

index of the media on the metal thin film varies a shift in the θr will be observed. Figure 6

illustrates the angle of resonance and its shift due to changes in the refractive index of the

sample.

Figure 6: (a) SPR angular configuration curve, which shows the intensity of the

reflected light versus the incident angle. The angle at which the minimum

intensity occurs is called the coupling angle (θspr). (b) The graph shows the

relationship between the coupling angle and the refractive index of the

sample above the prism.

Intensity Modulation: In the intensity modulation, monochromatic light is used as

a light source at a fixed incident angle. The intensity of the reflected light is measured by

the device as the refractive index of the dielectric material changes the resonance condition

0

0.5

1

50 52 54 56 58 60

Norm

.Ref

lect

ion

In

ten

sity

Angle (deg)

55.4

55.5

55.6

55.7

1.333 1.334 1.335 1.336

Coupli

ng A

ngle

(D

eg)

Refractive Index

(a)

(b)

13

and consequently the amount that intensity of the reflected light varies depends on the new

refractive index [78].

2.1.2. Surface Plasmon Resonance imaging (SPRi)

In a traditional SPR system, the average intensity of the reflected light from the

entire surface is measured and the results show the average refractive index variation of

the sample on the entire surface. In SPR imaging, the intensity of the reflected light is

analyzed at each position on the sensing surface. The output of this sensor is a grey scale

image, which is called a difference image and represents the refractive index change of the

dielectric media above the metal film pixel by pixel. The pixel size determines the

resolution of the device, which is 7 µm2 in our device.

In SPRi sensors, multiple areas on the surface can be monitored simultaneously. In

this sensor, if the surface is functionalized with various ligands, the binding kinetics of

different analytes can be monitored at the same time (Figure 7).

Figure 7: Difference image from the chip surface generated by the SPRi device. The

bright spots represent the sections functionalized with (S) specific

antibodies and (N) non-specific antibodies. The dark section shows the

uncoated gold surface.

14

2.2. Cell Detection with SPR sensors

This section presents a review of investigations where SPR is used to study

bacterial cells. The information is partly adapted from “Cellular Analysis and Detection

Using Surface Plasmon Resonance Techniques” in Analytical Chemistry journal, volume

86, 2014 [79]. The results are presented with permission from Analytical Chemistry

journal. It will be useful in learning about different problems that need to be studied and

how people use SPR detection to approach the problem and get the best results. Different

techniques for modification of sensor surfaces for various cell detection and the results that

different research groups attained are discussed.

2.2.1. Bacterial Cell Detection with Surface Plasmon Resonance Imaging

While SPR can provide significant new fundamental insights about bacteria, thus

far research efforts have justifiably explored sensing applications. In order to decrease

health risks, deaths, and reduce economic losses due to pathogenic bacteria there is a

critical need for rapid, sensitive and selective detection methods to sense the disease-

causing organisms in food and beverages [80].

Among all microorganisms, bacteria cause 91% of all foodborne illnesses n USA

[81, 82]. Based on the estimates from the Centers for Disease Control and Prevention

(CDC), in the United States foodborne pathogens cause roughly 79 million illnesses,

325,000 hospitalizations, and nearly 5000 deaths each year [83]. In addition, outbreaks of

foodborne illnesses result in economic losses totaling several billions of dollars annually.

According to the United States Department of Agriculture (USDA) Economic Research

Service (ERS), medical costs and loss of productivity caused by five major pathogens,

Escherichia coli O157:H7, non-O157 STEC (Shiga Toxin Producing E. coli), Salmonella

15

(non-typhoidal serotypes only), Listeria monocytogenes, and Campylobacter is $6.9 billion

annually [84, 85]. Various detection techniques have been employed for pathogenic

bacterial detection applications [86], such as conventional microbiological culture method

[87], polymerase chain reaction (PCR) [9, 88-94], enzyme-linked immunosorbent assays

(ELISA) [95, 96], amperometric biosensors [97-101], piezoelectric biosensors [101-104],

potentiometric biosensors [105, 106], bioluminescence [107, 108], fluorescent labeling

[109, 110], and ultrasound [111]. All of these methods have concerns such as long

detection time, enrichment requirements, labeling necessity, requirements of trained

personnel, and high cost [35, 112]. Therefore, there is a need for alternative rapid, low cost,

and sensitive detection in complex samples. These needs can be fulfilled by Surface

Plasmon Resonance (SPR), which provides label-free, real-time, and quantitative detection

[36, 113, 114].

In the study done by Choi, et al. [115] SPR device was used for monitoring

environmental pollutants, such as phenol. The surface was functionalized with a self-

assembled synthetic oligopeptide. The self-assembly technique provides reliable control

over packing density of ligands on the surface. In this self-assembled synthesis,

oligopeptide sequences, including modified Arg-Gly-Asp (RGD), were immobilized on

the gold surface. The upper part of modified peptide (RGD) was utilized for

immobilization of cells on the surface. RGD is believed to influence positively target cell

immobilization on the surface. Two different designs of peptides were used for the

immobilization of cells, one is a single stranded oligopeptide (C-(RGD)4) and the other is

a poly-oligopeptide network grafted with four branches of (C-(RGD)4). The angle shift in

SPR was monitored in each step during surface functionalization. The resonance angle for

16

bare gold surface was detected with the instrument. Adsorption of single stranded

oligopeptide (C-(RGD)4) causes a shift in the resonance angle. Then the resonance angle

shifts significantly after immobilization of E. coli O157:H7 on the surface. This shift of

the resonance angle after immobilization of E. coli O157:H7 is higher for surfaces

primarily functionalized with grafted C-(RGD)4 than with C-(RGD)4.

The researchers combined Atomic Force Microscopy (AFM) with SPR to study

surface topography as well as biological interactions on the surface. The AFM results from

the surface before and after oligopeptide immobilization confirm the successful

immobilization step.

The concentration of immobilized bacteria (E. coli O157:H7) plays a significant

role in toxicity detection because a higher amount of immobilized bacteria creates more

sensing elements on the surface, which leads to a larger shift in resonance angle and

determines the limit of detection for the sensor. A greater shift in the resonance angle

occurs with higher concentrations of synthetic oligopeptide on the sensor surface,

indicating that a higher number of bacterial cells are immobilized. The results show the

highest bacterial cell immobilization for the surface functionalized with grafted C-(RGD)4.

The live cells keep their physical integrity, including their cellular membranes,

which help them bind to each other and to the surface modified with oligopeptides. As

soon as the cells die due to any toxicity, they lose physical integrity, which causes the

intracellular material to decrease and results in angular shifts in the plasmon curve. The

researchers used this phenomenon to detect the presence of toxic chemicals with SPR.

Different concentrations of phenol were injected to the modified surface and the angular

17

shift in the plasmon curve was obtained for each concentration. The smallest detectable

shift occurred for 5 ppm of phenol, which determined the limit of detection for this sensor.

Arya, et al. [116] used SPR as a transduction technique to detect a specific strain

of Escherichia coli bacteria, E. coli K12. T4 bactriophages, as specific receptors of the

bacteria, were immobilized on the gold surface using a self-assembled monolayer of

dithiobis(succinimidyl propionate) (DTSP).

All steps of surface functionalization were monitored with SPR. It was shown that

using a higher concentration of T4 bacteriophage solution increases the phage

immobilization on the surface and subsequently results in a higher SPR response for the

same concentration of bacteria. The bioassay platform was also shown to be specific to E.

coli K12 as negligible changes in SPR signal were observed for the non-specific bacteria

strains, E. coli NP10 and NP30. A reproducibility experiment, which was done by injecting

a regeneration solution after each injection (in order to set baseline), showed a stable

platform for different experiments on the same chip. This SPR-based platform was able to

detect K12 bacteria concentrations in the range of 7x102 to 7x108 CFU/mL.

Taylor, et al. [7] used SPR to detect E. coli O157:H7 that were processed in three

different ways: untreated, heat killed then soaked into ethanol, and detergent lysed. The

surface was functionalized with a mixed, self-assembled, monolayer of alkanethiols

followed by immobilization of Mouse anti-E. coli O157:H7 monoclonal antibody (MAb).

After surface modification, SPR experiments were run to determine the detection range of

each bacteria sample. In each experiment, different concentrations of bacteria were flowed

over the surface and direct detection of MAb and bacteria occurred, subsequently

additional MAb was flowed over the surface to amplify the response.

18

The resonance wavelength shift was detected with the device and results show that

for untreated bacteria, heat killed, and detergent lysed bacteria the limit of detection was

107, 106, and 105 respectively. Then a sandwich assay was added to the detection protocol

to increase the LOD for each type of bacteria. After bacteria attached to the MAb on the

sensor surface, subsequent binding of a secondary MAb to the bacteria, amplified the

detection limit by an order of magnitude for all three samples.

The reproducibility test for this antibody immobilized sensor surface was done with

SPR by using three non-specific bacteria, E. coli K12 serotype, S. choleraesuis, and L.

monocytogenes. In this set of experiments, the nonspecific bacteria was flowed over the

sensing surface and any shifts in the resonance wavelength were due to binding of non-

specific bacteria to antibodies on the surface. The results show negligible resonance shift

after running the non-specific bacteria over the surface. Addition of the sandwich assay

protocol after the non-specific bacterial also did not shift the resonance angle.

For the first time, Wang, et al. [117] used a SPR biosensor with lectin modified on

the sensor surface as a receptor to detect E. coli O157:H7. To choose the best lectin

regarding binding to the bacteria, five different lectins from Triticum vulgaris (WGA),

Canavailia ensiformis (Con A), Ulex europaeus (UEA), Arachis hypogaea (PNA),

Maackia amurensis (MAL) were immobilized on separate chips, and bacterial attachment

to different surfaces determined the best lectin option for use as a receptor. When E. coli

O157:H7 binds to the immobilized lectin on the surface, an increase in the refractive index

occurs, which is detected with SPR. SPR also determines kinetic binding parameters (ka=K

association , kd=K dissociation), which allows the determination of KA (affinity parameter

ka/kd), which determines how tightly bacteria bound to lectin. SPR results show highest KA

19

values and greatest changes in the refractive index due to bacterial attachment to the WGA

lection with a detection limit of 3 x 103 CFU/mL. It was proposed that WGA is better

because of its biological structure, which provides more binding sites for bacteria to attach

to than the other lectins. They were also able to detect E. coli O157:H7 in cucumber and

ground meat with a LOD of 3.0 x 104 and 3.0 x 105 CFU/mL, respectively.

The effect of two different surfaces on E. coli detection have studied with SPR by

Bacca, et al. [118]. The first surface was primarily functionalized with an acid-thiol self-

assembled monolayer (SAM) and subsequently modified with anti-E. coli antibody for

detection of E. coli. The second substrate was functionalized by immobilizing modified

gold nanoparticles on the surface, which consequently create a larger surface area.

SPR expriments were performed by flowing E. coli bacteria on modified surfaces

and monitoring changes in the refractive index versus incident angle over time. The

resonance angle shift was monitored with the device for all surface functionalization steps

as well as subsequent bacterial attachment to the ligands. The results showed more

bacterial attachment to surfaces functionalized with gold nanoparticles, which makes this

sensor a better option to use for E. coli detection.

The limit of detection for functionalized gold surfaces and surfaces functionalized

with gold nanoparticles was 103 CFU/mL and 104 CFU/mL, respectively.

Subramanian, et al. [119] studied the effect of different surface chemistry for

Staphylococcus aureus detection with SPR. Surfaces with monothiol and dithiol self-

assembled monolayers were examined in different ways to evaluate the best surface

chemistry to choose for a SPR-based biosensor.

20

In the first step, primary anti-S. aureus antibodies were immobilized on the sensor

surface and different concentrations of S. aureus bacteria were passed over the sensor

surface. Binding of S. aereus against anti-S. aureus antibody was measured by a peak in

the SPR output. Increasing the amount of immobilized bacteria on the surface resulted in

a higher peak in the corresponding SPR signal. The results show greater response for

surfaces functionalized with a monothiol SAM. The sensitivity of the sensor was increased

by doing a sandwich assay detection, where binding of secondary anti-S. aureus antibodies

to the already immobilized S. aureus bacteria on the surface decreased the sensor detection

limit to 105 CFU/mL, which was 100 times better than direct detection alone.

In the study done by Lee, et al. [120], E. coli with auto-displayed Z-domains were

immobilized on modified SPR sensor surfaces for molecular recognition. The outer

membrane of E. coli is negatively charged because of phosphate groups in the

lipopolysaccharide layers, which allowed immobilization on the surface coated with a

positively charged layer by charge interaction.

In this work, the efficacy of cell immobilization on different surfaces, the stability

of the immobilized cells and the sensitivity of the sensor for detection of C-reactive Protein

(CRP) was studied.

The efficacy of immobilizing fluorescently labeled E. coli was determined by

counting the immobilized cells on three different surfaces: bare gold, only poly L-lysin

coated, and parylene-H film with a poly-L-lysine coating. The results showed a greater

number of cells immobilized on the surface coated with parylene-H film with poly-L-lysin

compared to the other two surfaces.

21

The stability of different coatings has been studied by treating the surfaces with salt

at different concentrations, and the SPR response change was been monitored for each

surface. The results indicate much less change in the SPR response after sequential

treatments for the gold surface coated with parylene-H film with poly-L-lysine, which

means that this surface had greater stability than the others.

To measure the sensitivity of the sensor for detecting CRP, first anti-CRP antibody

was injected to the surface to react with auto displayed Z-domains on the sensor surface,

then the SPR response to different concentrations of CRP was monitored. The results show

a higher sensitivity and lower detection limit (1 ng/mL with an SPR response of 25.9±37.9

RU) for the surface coated with parylene-H film with poly-L-lysin compare to the other

two surfaces.

Tawil, et al. [121] created a homemade SPR biosensor to specifically detect E. coli

and methicillin-resistant Staphylococcus aureus (MRSA) without further enrichment

requirements in less than 20 minutes. They functionalized the sensor surfaces by

immobilizing T4 bacteriophages for detection of E. coli and BP14 bacteriophages for

detection of MRSA. The sensor was used to detect different concentrations of bacteria with

a detection limit of 103 CFU/mL. The specificity of the sensor for BP14 MRSA versus

EC12 E. coli was studied with SPR. No SPR response was observed upon E. coli injection,

confirming the specificity of the sensor surface.

Subramanian, et al. [6] fabricated an SPR chip for direct detection of E. coli

O157:H7 by using a mixed alkanethiol self-assembled monolayer. They investigate the

effect of different concentrations of primary polyclonal antibodies for bacteria capture by

using three different concentrations of antibodies on the surface. To enhance the detection

22

signal, they used a sandwich assay approach by passing secondary antibody (anti-E. coli

O157:H7) over the sensor surface after bacterial immobilization. The results indicate that

adding the secondary antibody binding improved the sensitivity by 1000 times. The limit

of detection of the sensor was investigated by varying the concentrations of primary and

secondary antibodies, and was found to be 103 CFU/mL of E. coli O157:H7 with the

sandwich assay format.

The specificity of the sensor against different concentrations of S. enteritidis also

in the cocktail including E. coli O157:H7 (106 CFU/mL), S. enteritidis (106 CFU/mL),and

E. coli O55 (109 CFU/mL) was also studied. They also showed that using a Protein G assay

with anti-E. coli O157:H7 Mabs enhanced the sensitivity of the sensor.

Koubová, et al. [122] created a method based on SPR for rapid, sensitive, and

specific detection of the bacterial pathogens: Salmonella enteritidis and Listeria

monocytogenes, responsible for many common foodborne illnesses in humans. The

specific bacteria were identified through the attachment of antibodies to the sensor surface.

The antibodies used in this study were a monoclonal antibody specific to the somatic

antigen (O) serotype 9 surface lipopolysaccharide of Salmonella and the IgG fraction of

rabbit anti-Listeria.

The surface was functionalized in two distinct ways. In the first method, an

antibody layer was adsorbed on the gold surface from citrate buffer (CB) at a pH of 4.

Dextran sulfate sodium salt (DS) polyanions were electrostatically attached to the

positively charged antibodies. A second layer was then electrostatically adsorbed on the

DS layer. 0.5% glutaraldehyde in CB was used to crosslink the multilayer, connecting the

two layers of antibodies through covalent bonding. Phosphate buffered physiological

23

saline (PBS) was used to wash out the DS and any antibody molecules that were not cross-

linked (which were both negatively charged at a pH of 7.4). In the second functionalization

method, a bovine serum albumin (BSA) layer was adsorbed on the gold surface from CB

at a pH of 4. A 2% glutaraldehyde solution in CB was used to crosslink the amino groups

of the BSA, and the antibody was bound to the aldehyde groups on the BSA layer from

PCB (a mixture of PBS and CB). Throughout both of these processes, the SPR system was

used to monitor the presence and attachment of the various solutions and antibodies. The

first method, which utilized a double layer of antibodies, elicited a larger response from

the optical sensor in the SPR mechanism than did the BSA method, as the probability of

antigen binding increased due to the second layer of antibodies on the surface.

The attachment of the bacterial antigens to the antibodies fixed on the sensor

surface altered the wavelength of the reflected light, and this change in wavelength was

used in the SPR process to detect the presence of these specific bacteria. The resulting data

from the SPR wavelength detection process showed that this method of bacterial detection

and identification was able to detect up to a limit of 106 cells per milliliter of solution for

both Salmonella and Listeria, which is on par with the results of the ELISA identification

technique but is not sensitive enough for practical health applications. In addition, the flow

rate seemed to have a large influence on bacterial attachment and overall reflectivity, and

this problem was not addressed.

Waswa, et al. [123] used a SPR based biosensor to directly detect E. coli O157:H7

spiked into food samples: milk, apple juice, and ground beef. The gold surface of the

sensor was modified with biotinylated Rabbit antisera containing polyclonal antibodies

against the pathogen and food samples spiked with E. coli O157:H7 in different

24

concentrations were flowed over the sensor-modified surface. The sensitivity of above

sensor for bacterial detection was established to be 102–103 CFU/mL. The detection limit

of the sensor calculated from the lowest bacterial concentration that generated a response

signal and was at least three standard deviations larger than the signal from a negative

control in the spiked food samples, which the results show it is comparable with other

detection techniques such as fiber-optics. Specificity of the sensor to E. coli O157:H7 was

tested against genetically similar species, shigella sp. and E. coli K12. The response of the

sensor to non-target pathogens was similar to the negative control in which there were no

bacteria present.

Waswa, et al.[124] also used SPR to detect Salmonella enteritidis and Escherichia

coli. Salmonella enteritidis and Escherichia coli are common bacterial foodborne

pathogens in the United States and abroad, causing approximately 1.7 million cases of

illness annually in the United States. Therefore, a method of rapid detection of these

pathogens is in the best interest of public health; the goal of this study was to accomplish

this objective through the use of surface plasmon resonance (SPR).

In this setup, the gold sensor surface was coated with carboxymethyl dextran,

which provided a binding site for the antibodies used in this experiment. For Salmonella,

mouse polyclonal affinity-purified antiserum was used, and polyclonal rabbit antibodies

for E. coli O26 were purchased. The surface was functionalized by covalently linking

protein A (acquired from S. aureus) to the carboxymethyl dextran layer through an amine

coupling method and binding one of the antibodies to the protein A.

Experiments were performed using bacterial solution in buffer as well as skim milk

spiked with varying concentrations of bacterial cells (from 10 to 106 CFU/mL, with a

25

negative control of 0 CFU/mL). For the bacterial solution tests, each species in various

concentrations (from 102 to 107 CFU/mL) was injected into the SPR system with the sensor

surface functionalized with each species’ respective antibody. The change in the refractive

index was recorded and plotted, and the results were normalized by expressing the

refractive index change for each concentration as a ratio with respect to that of the same

species at the maximum concentration (107 CFU/mL). The results of the tests were

averaged by species and concentration, and the R2 results indicate that the process was very

sensitive for both Salmonella and E. coli. The bacteria could be easily washed off the

surface with sodium hydroxide solution, allowing for the same functionalized surface to

be reused several times. Specificity was tested by cross-testing antibodies against

increasing concentrations of nonbinding bacterial species, and the results showed that each

antibody was very specific to its bacterial species and did not show a significant response

to other species or to the negative control.

Subsequently, tests were conducted with spiked skim milk with bacterial

concentrations of 10 to 106 CFU/mL. The limit of detection was chosen to be three standard

deviations higher than the results of the negative control. Based on the results of the SPR

experiments, the limit of detection was calculated to be 25 CFU/mL for E. coli and 23

CFU/mL for Salmonella enteritidis, which is remarkable in comparison to the detection

limits of other rapid response biosensing techniques for Salmonella (105 CFU/mL). The

sensitivity, reproducibility, and specificity of this assay shows incredible promise for the

field of public health and pathogen detection.

Dudak, et al. [125] developed a SPR-based immunosensor for enumeration of E.

coli in water samples from rivers and E. coli inoculated tap water. The surface was

26

modified by immobilization of streptavidin followed by biotin conjugated polyclonal

antibodies against E. coli. SPR response signal to each step of surface functionalization

was monitored with the device. To show the sensitivity, different concentrations of E. coli

in water samples was passed over the surface and the changes in the RU were determined

by measuring the difference between the sensing signal and the baseline. The sensitivity

was established to be comparable with conventional methods such as plate counting but

requires only 30 minutes, which is much less than the 24-48 hour required for conventional

methods. The specificity of the above sensor was tested against E. aerogenes and E.

dissolvens, which are also found in water, as a result of fecal contamination. The results

show much less response to these species than to the target bacteria.

Oh, et al. [126] developed a SPR based immunosensor for rapid detection of

Salmonella paratyphi. To increase the sensitivity Protein G was coated on the gold surface

by using a self-assembly technique. It was shown previously that coating the sensor surface

with protein G increased antibody immobilization on the surface, which consequently

increased the sensor sensitivity (Oh, et al., 2004). The SPR resonance angle shift was

monitored while coating the surface with Protein G, immobilizing Mab against S.

paratyphi on the layer, and subsequently capturing the target S. paratyphi on the

antibodies. The SPR angle shift increased significantly upon adsorption of antibody and

subsequent capture of bacteria. The relationship between SPR response signal and S.

paratyphi concentration showed that increasing the concentration of bacteria produced a

corresponding linear SPR angle shift. The lower detection limit of 102 CFU/ml is four

orders of magnitude more sensitive than other common detection methods, such as ELISA.

27

The specificity of the sensor was studied by monitoring the cross reaction between

Mab against S.paratyphi and other pathogens that also exist in contaminated water

samples. The results indicated that the SPR angle shift for binding between Mab against S.

paratyphi and non-specific pathogens was much less that the shift due to specific binding

between S. paratyphi and Mab against S. paratyphi.

Jyoung, et al. [127] developed a sensor for detection of Vibrio cholerae O1 with

SPR. To control the orientation of capture antibodies on the sensor surface, the surface was

coated with G protein layer. A self-assembled monolayer (SAM) of 11-

mercaptoundecanoic acid (MUA) was added prior to the G protein layer to help with

protein adhesion. The formation of SAM of 11-MUA, protein G layer, and immobilization

of Monoclonal antibodies (Mab) against Vibrio cholerae O1 were monitored with SPR

spectroscopy and the shift in SPR resonance angle increased in each step, as expected.

After surface modification with Mab, different concentrations of Vibrio cholerae

O1 were injected over the surface, and changes in the minimum SPR angle were monitored

by the device. The results show that increasing the concentration of Vibrio cholerae O1

increased the minimum angle in SPR linearly with a detection range of 3.7x105 to 3.7x109

cells/mL, which are of value for detecting Vibrio cholerae O1 in fecal samples.

The specificity of the sensor to Vibrio cholerae O1 was investigated by using E.

coli O157:H7 and L. pneumophila as non- target samples. Very small shifts in SPR angle

indicated that the above immunosensor was selective and can be used for detection of

Vibrio cholerae O1.

Linman, et al. [128] developed a SPR based biosensor for detection of Escherichia

coli bacteria in fresh spinach. First a SAM of MUA was formed on the surface, then the

28

surface was activated by NHS-EDC solution followed by immobilization of goat anti-E.

coli fused with horseradish peroxidase (HRP) antibody on the surface. Different

concentrations of E. coli in PBS and in extracted samples from spinach were passed over

the sensor surface.

To enhance the sensitivity, a sandwich assay was performed, and HRP labeled anti-

E. coli antibodies were crossed over the surface to bind to already immobilized bacteria on

the surface, followed by injection of undiluted tetramethylbenzidine (TMB) over the

surface. It is worth mentioning that the surface was rinsed with PBS between each step.

SPR response signal was monitored during surface functionalization steps and upon

binding between bacteria and antibodies. This group used TMB to enhance the sensitivity

in bacterial detection for the first time.

For E. coli samples in PBS, SPR results showed very little response upon binding

of bacteria on the modified surface and the signal increased only a small amount after

binding of the anti-E. coli HRP conjugated antibody on the bacteria surface. The author’s

claim that this is because bacteria are large (1-5 µm in diameter) compared to the

evanescent field in SPR (200-300 nm) [129]. The SPR signal increase upon injection of

the TMB was 263% compared to samples without using HRP/TMB. The limit of detection

of E. coli in PBS was calculated to be 6x103 CFU/mL.

For E. coli samples in spinach, the results showed the same behavior, but with a

150% increase upon injection of TMB for signal enhancement, with a detection range of

104 to 106 CFU/mL. The calibration curves indicated a linear relationship with TMB

enhancement in which the SPR response was directly proportional to the concentration of

E. coli.

29

Salmonella (specifically serovars typhimurium and enteritidis) is the most common

bacterial pathogen to cause foodborne gastroenteritis. Detection of this bacteria is crucial

for the prevention of this devastating illness; however, current methods of detection are

time consuming and often require molecular labeling. Barlen, et al. [35] innovated a new

technique to both detect and identify different serovars of Salmonella simultaneously using

SPR. The SPR system measured the change in refractive index of the light reflected caused

by bacterial cell attachment to antibodies on the functionalized hydrophobic gold surface.

The system used was cuvette-based, allowing the researchers to bypass the difficulties

associated with fluid flow and use only 10 µL of sample.

The antibodies used in this experiment were comprised of two categories.

Polyclonal antibodies were used to first attach the bacterial cells to the surface, and O-

specific antibodies were used for specific serovar identification. Polyclonal rabbit

antibodies (IgG) were purchased, as well as O:4 and O:9 specific antibodies. Killed

Salmonella typhimurium and enteritidis cells were also purchased, and E. coli was cultured

on-site and used for cross-reactivity studies. Lipopolysaccharide (LPS) for both serovars

of Salmonella was purchased for further studies on sequential identification in single-

channel SPR.

The first test was to determine the lower detection limit of each bacterium in

separate buffer solutions (PBS). The surface was functionalized by binding polyclonal

antibody to a hydrophobic C18 functionalized gold surface. The bacteria were then

captured. In order to determine the specific bacterium, O-specific antibodies for a certain

serovar (O:4 for typhimurium and O:9 for enteritidis) were then allowed to attach to the

bacteria, increasing the change in refractive index. The detection limits in buffer were 1.25

30

x 105 cells/mL for typhimurium and 2.50 x 108 cells/mL for enteritidis. The same test was

performed with bacteria in spiked milk. The detection limit was unchanged for enteritidis

and increased slightly to 2.50 x 105 cells/mL for typhimurium. O-specific antibodies were

cross-reacted with the opposite strain of Salmonella and with E. coli, and these tests

determined that the polyclonal antibody was specific to Salmonella and the O-specific

antibodies were specific to their respective serovars.

Subsequently, experiments were performed to test the detection of each serovar in

a mixture of both serovars in spiked milk. The O-specific detection signals were additive

when both antibodies were added at the same time, indicating that simultaneous detection

is possible. Multi-channel analysis was used to detect the different serovars (one channel

per serovar). Sequential detection in a single channel was also performed successfully;

however, it was determined that, due to the small field size in the SPR system, the signal

of the second O-specific interaction was reduced. Therefore, the serovar with the lowest

detection signal must be tested for first if the assay is to be performed sequentially in a

single channel. This team of researchers pioneered the use of the surface plasmon

resonance system to both detect and identify different serovars of Salmonella, as well as

developed a new protocol for single-channel sequential serovar detection with SPR.

Acidovorax avenae subsp. citrulli (Aac) is a bacterium responsible for bacterial

fruit blotch in watermelons and cantaloupes, a devastating crop disease that is transmitted

through seed infection. In a series of experiments, Puttharugsa, et al. [130] tried to

determine if surface plasmon resonance imaging (SPRi) is a viable option for the detection

of Aac in naturally infected plants. Antibodies fixed to the gold sensor surface selectively

bound the Aac bacteria, and the SPRi machine interpreted the changes in light reflectivity

31

to determine the presence of the specific bacteria at various concentrations. The antibodies

used were monoclonal antibody MAb 11E5 (produced in mice and found to be very

specific to Aac) and polyclonal antibody rPAb-MPC (purchased from the Department of

Plant Pathology, Kasetsart University, Thailand).

The surface was functionalized by utilizing mixed self-assembled monolayers

(SAMs) which consist of two thiols of different chain lengths. In the direct detection assay,

a mixture of 11-MUA, which contains a carboxyl group that binds to the monoclonal

antibody, and 3-MPOH, containing a hydroxyl group as a spacer, in a concentration ratio

of 1:40 was used as an attachment point for the monoclonal antibody MAb 11E5, produced

in hybridoma mouse cells. Casein was used to prevent non-specific binding, increasing the

specificity of the process for detecting Aac. In addition, a sandwich assay was created by

adding polyclonal antibody rPAb-MPC to the bound MAb/Aac layers, which was found

through SPRi experiments to reduce the amount of cells required for detection.

The SPRi system was first used to identify the optimal antibody concentration for

the highest specific Aac binding. Through a multi-channel experiment, both whole and

broken cells were added to channels with varying amounts of bound antibody, and it was

found that 10 µg/mL was the optimal concentration of antibodies for specific cell binding.

Subsequent SPRi experiments determined that the limit of detection with the direct

detection assay (only monoclonal antibody) was 106 cells/mL; with the polyclonal

antibody added in the sandwich assay, the limit of detection dropped to 5 x 105 cells/mL.

Although these processes do not match up to the level of detection of the ELISA process

(5 x 104 cells/mL), the precision of the SPRi processes is good enough for applications of

infection detection for Aac. In addition, many attributes of the SPRi identification

32

processes described in this paper set SPRi apart from its competition, including the ability

to perform multiple cycles with the same mixed SAM (a wash with 10 mM glycine pH 2.0

washed off the Aac cells but left the SAM intact) and the performance of simultaneous

multichannel analysis. Most importantly, the SPRi process was able to adequately and

selectively detect Aac in a naturally infected plant, which was the main goal of this study.

Ostuni, et al. [131] examined the hypothesis that self-assembled monolayers

(SAMs) that resist protein adsorption also resist the attachment of bacterial and mammalian

cells to the surface. Using SPR, the researchers were able to determine the qualities of

inert SAMs (i.e. SAMs that adequately resist protein adsorption). Inert SAMs generally

contain only hydrogen bond acceptors, have an overall neutral charge, and are polar. Using

these characteristics, the experimenters developed six homologous single-component

SAMs with structurally different terminal groups in order to test their effects on protein

adsorption and cell adhesion on the surface. In the first set of experiments, adsorption of

proteins, specifically fibrinogen and lysozyme, onto the SAMs was tested using surface

plasmon resonance. Hexadecanethiolate (HDT) was used as a reference for this

experiment, and the changes in reflectivity of the chosen SAMs were normalized into a

ratio (%ML) of refractive index change of the SAM vs. refractive index change of HDT.

None of the chosen SAMs resisted adsorption better than the most inert SAM known (tri-

ethylene glycol), %ML = 0.2); however, the protein adsorption for both fibrinogen and

lysozyme was sufficiently low for practical applications.

Subsequently, using the principle that the number of bacterial colonies grown on

an agar plate is proportional to the number of colony forming units (CFU) recovered from

the SAMs, bacterial adhesion to each of the single-component SAMs was tested. The

33

species used in this test were S. aureus and S. epidermidis, as these species are responsible

for 30 – 50% of infections on indwelling medical devices. After adhesion of these species

to the SAM surfaces, agar plate cultures were made of each SAM for each bacterial strain.

The results of the colony formation showed that bacterial adhesion had little to no

correlation with protein adsorption, and that bacterial adhesion must be related to other

factors; this goes against the previously stated hypothesis. In addition, bovine capillary

endothelial (BCE) mammalian cells were used to test mammalian cell adhesion to each

SAM. After the cells in modified Eagle’s medium were allowed to adhere to the SAM

surfaces, the cells were fixed and counted. The results showed that mammalian cell

adhesion also had little to no correlation to protein adsorption, and bacterial adhesion and

mammalian cell adhesion also did not correlate. The results of these experiments did not

agree with the hypothesis; however, the SPR investigations illustrated the use of SPR for

SAM testing and testing protein adsorption, allowing the researchers to quickly screen for

the best possible SAMs for their experiment.

In order to physically defend a wound, the body will often coat the area with

proteins, such as fibronectin. When S. aureus binds to the fibronectin, this can lead to

infection. Holmes, et al. [132] develoed a method using SPR to demonstrate the role of

fibronectin binding protein A (FbpA) in the attachment of S. aureus to fibronectin as well

as the identity of the domain of fibronectin used as a binding site location for S. aureus and

S. epidermidis. SPR was used to experimentally determine the optimal concentrations and

flow rates and the rate of binding for FbpA, S. aureus, and S. epidermidis to fibronectin

and its domain fragments. Fibronectin was purchased and purified, and fibronectin

fragments were isolated. FbpA was purified from S. aureus.

34

The gold-plated surface was functionalized with fibronectin or fragments (30-100

µg/mL) in sodium acetate flowed over the surface for 2 to 7 minutes at 5 µL/min; the

remaining surface was blocked with ethanolamine. During the experiments, the bacteria

were again suspended and flowed over the surface at varying concentrations and flow rates.

Controls were run using gelatin and polyclonal fibronectin antibodies. The experiments

found that the binding of S. aureus to fibronectin required long contact times and the

bonding kinetics were largely unaffected by the flow rate. The flow rate that elicited the

highest response was 2 µL/min. Binding of the S. aureus to immobilized fibronectin had a

limit of detection of 1 x 108 CFU/mL. They also found that FbpA and S. aureus bind with

high affinity to both whole fibronectin and the 27 kDa fragment (containing the N-terminal

of fibronectin), which indicates that FbpA is the protein in S. aureus responsible for

binding and that S. aureus binds to the N-terminal of the fibronectin protein. Their high

affinity is further illustrated by their unwillingness to dissociate. Attachment of S.

epidermidis could not be detected using the standard BIAcore SPR system; instead, the

BIAcore 2000 multichannel system was used, which allowed the researchers to show that

S. epidermidis bound to the C-terminal of fibronectin with a low affinity, requiring a

concentration of 5 x 109 CFU/mL.

This study successfully pioneered the use of surface plasmon resonance in the

detection of binding of whole bacterial cells. The use of intact bacteria and the lack of

molecular labeling and protein modification highlights the value of SPR in the field of cell

detection.

Salmonella bacteria are a major cause of infections and foodborne illnesses in the

United States, and traditional methods of detecting the bacterium are costly, time-

35

consuming, and require large samples of cultured bacteria. Lan, et al. [133] used SPR to

detect the Salmonella typhimurium bacterium in a chicken carcass, a common source of

the foodborne pathogen. The analyte (in this case, the bacteria) was passed over a surface

functionalized with a ligand during SPR measurements. The S. typhimurium was

purchased, cultured, and diluted in buffered peptone water (BPW), and the chickens were

purchased and washed with PBS; that PBS wash was then diluted in BPW for use in the

experiment. The antibody used was a lyophilized affinity-purified antibody to S.

syphimurium common structural antigens (CSA-1), rehydrated in carbonate buffer and

diluted to 500 µg/mL.

A sugar experiment was run in order to find the critical concentration of sugar

required for statistically significant detection. In this experiment, it was determined that a

2.5% analyte solution was required to change the refractive index by an amount that was

statistically significant, identifying a large limitation in their process. For the bacterial

detection, the gold sensor surface for SPR was first immersed in neutravidin dissolved in

PBS solution. In order to functionalize the surface for the experiment, the avidinated

surface was immersed in a biotinylated antibody solution to bind the antibody via an

avidin-biotin interaction. From the subsequent SPR experiment, it was determined that the

limit of detection of this process was 1 x 106 CFU/mL. Although the detection limit is not

groundbreaking, the SPR technique shows promise in the field of rapid, real-time detection

of pathogenic bacteria, saving time and money for the food industry as a whole.

36

2.2.2. Mammalian Cell Detection with Surface Plasmon Resonance Imaging

Hiragun, et al. [134] used a SPR system to examine how antigen stimulation and

epidermal growth factor receptors (EGFR) affected changes in the angle of resonance

(AR). In addition to initially studying how Chinese Hamster Ovary (CHO) cells reacted to

EGFR, some human carcinoma cell lines were examined to determine their SPR signal

patterns. Antibodies were used to immobilize the cells.

The results fell into five distinct categories. The first discovery was as follows:

“CHO cells expressing the wild type EGFR and HaCaT cells show triphasic changes of

AR in response to AGF.” The SPR signals reflected three distinct phases of cell activity

induced by the EGFR. Next, “CHO cells expressing the EGFR mutated on the ATP-

binding domain show the minimal change of AR in response to EGF.” The mutation the

cells experienced on the ATP-binding domain on their membrane does not significantly

affect a change in AR. Thirdly, they reviewed the effect of wortmannin, a PI3K inhibitor,

on the cell phases with SPR. They discovered that PI3K weakened the third phase of AR

shift. Therefore, it can be inferred that PI3K was “involved in the last phase of SPR signal

evoked by EGF simulation.” The next result was that “the pattern of AR change was not

dependent on the concentration of EGF.” The SPR patterns are more likely a function the

cell type and chemical stimulation. Lastly, “carcinoma cell lines exhibited diversities in

changes of AR induced by EGF.” This particular result is significant for the greater purpose

of implementing SPR in functional cancer diagnosis as cancerous cells showed different

SPR patterns than regular cells.

In the other research done by Chabot, et al. [135] the SPR effect was investigated

when living, mammalian cells were stimulated by three specific types of chemical agents:

37

lipopolysaccharides (an endotoxin), sodium azide (a chemical toxin), and thrombin (a

physiological agonist). These specific chemicals were chosen to examine different cell

activities. Lipopolysaccharides “cause an important cellular response often leading to cell

death.” Sodium azide inhibits cellular respiration and can therefore provide insight into

how sensitive the cell was to activation by chemical agents. Thrombin affects cell layer

integrity, which will test how SPR reacts to a change in the cell membrane traits. The

different chemicals each affect SPR by altering cellular morphology, which in turn changes

the cell’s effective refractive index at the interface between the cell membrane and the

metal layer. Poly-d-lysine was used to create adhesion for the kidney cells to the gold

surface.

Increasing lipopolysaccharide concentration resulted in a higher measured SPR

response. The SPR measurements were cross referenced with phase contrast microscopy

to affirm morphological changes in the cells. Sodium azide caused a decrease in SPR

reflectance as the cells shrunk. Thrombin was also observed to show cell contraction as the

intercellular gaps increased upon application. Ultimately, SPR can be used as a real time

detection scheme to collect a dose-response relation in mammalian cells.

Lee, et al. [136] used SPR to characterize molecular interactions. HEK-293 cells

expressing the protein ODR-10 were examined to view intracellular events when these

cells were exposed to diacetyl, an odorant molecule specific to ODR-10. Exposure to

diacetyl causes an opening of the Ca2+ channels within the cell. SPR was used to detect

these intracellular events that were caused by the binding of odorant molecules in diacetyl

on the membrane/surface of the kidney cells. These ODR-10 affected cells were compared

38

to a control experiment in which HEK-293 cells were not genetically altered to express the

ODR-10 protein.

Poly-d-lysine was used to induce cell adhesion to the gold surface. Poly-d-lysine

is an effective adhesive as it contains multiple positive charges. The rho-tag gene within

the HEK cells was used to express ODR-10. Gel electrophoresis was then performed to

affirm the production of ODR-10.

They measured the change in resonance angle with respect to time as the diacetyl

was injected on to the surface. The ODR-10 protein affected cells reached a resonance

angle maximum at 150s. This is when the team ceased diacetyl injection. The control HEK

cell group showed no reaction on the SPR image to diacetyl. This change is was propsed

to be the caused in response to the intracellular change in Ca2+ ions. It was also postulated

that an increase in diacetyl concentration causes an increase in the intensity of the SPR

signal. This experiment successfully used SPR to “identify odorant molecules specific to

each olfactory receptor in a real-time manner and without any labeling.”

Iribe, et al. [137] examined the effect of B and T lymphocyte binding with

antibodies as surface antigens. The effects were detected using SPR signal changes. It has

been proposed that SPR changes can measure not only morphological changes, but also

intracellular events. Lymphocytes were chosen for this study as they cause an intracellular

reaction with a minimal morphological change. The cells utilized in this study were

collected from mouse spleens.

Specifically, anti-IgM and anti-CD19 were used for B lymphocytes and anti-CD3

were used for T lymphocytes. B type cells responded to anti-IgM and anti-CD19, but not

andti-CD3. The spleen cells (which are a mixture of B and T cells) responded to both types

39

of lymphocytes by showing increased SPR signal response. The study concluded that

“specific surface antigens of ‘single’ B and T lymphocyte could be observed without any

labeling by using the high resolution 2D-SPR.”

The reaction of living cells has been studied by Yanase, et al. [138]. This study had

two main purposes: 1) To fix living cells on a gold surface and 2) to recover adherent cells

from the culture dish while preserving their functions for analysis with SPR. They tested

three different ways to adhere living cells to a gold surface. First, they inserted a

biocompatible anchor into the cell membranes. Second, they used a positively charged

amino group that bound to the negatively charged cell membrane. Third, they exploited

covalent, peptide bonds formed between the cell and dithiobis[succinimydylpropionate]

(DSP). To test how to keep adhered cells intact they tested four methods. First, they used

a type of dish where the cells are cultured while floating. Second, they cultured the cells

on a temperature responsive polymers that melted at 32 °C or at lower temperatures. Third,

they used trypsin, an enzyme that breaks down proteins, to dissociate the cells from the

culture plates. Finally, they vigorously pipetted at 4 °C the standard cell-culture dishes.

For this study, the type of cells used were human blood cells, specifically B-

lymphocytes. Amino-alkanethiol, DSP, and the biocompatible anchor for cell membranes

(BAM) were used successfully to fix the basophils to the gold plated chips. In addition this

study “demonstrated that SPR sensors can detect not only reactions of adherent cells, but

also those of non-adherent cells, by locating them on the surface of a SPR sensor chip.”

The most efficient adherent was cysteamine, a type of aminoalkanethiol.

Horii, et al. [139] tested the allergenic response of rat basophilic leukemia cells

(RBL-2H3) with 2D-SPR imager which can obtain 2D-images of local refractive index

40

change on the surface of a gold thin film. The cells were pre-sensitized with anti-DNP IgE.

The response was measured by the change in reflection intensity in the SPR signal. The

SPR angle of the cell region was 52.2° and was 50.8° for the bare gold region. The cell

region angle shifted to 52.4° when antigen stimulation was introduced.

They also tested the degranulation of the rat cells. They concluded that the SPR

measurement was more sensitive for antigen stimulation than for degranulation. They

observed an intensity change near the rat cells and an expansion of adhered area on the 2D-

SPR upon antigen stimulation. This study reaffirmed the use of 2D-SPR as a label-free and

real-time monitoring tool for cellular studies.

Hide, et al. [140] studied RBL cells similarly to the study described above and used

DNP as antigen stimulation. The response of the cells showed a long SPR signal change

that was directly dependent on the dose of DNP applied. The signal may have remained

higher for a long period of time, even after the ligand stimulation was removed, because

of resulting biochemical reactions triggered by the binding. To test this theory, they also

applied other chemicals that affected the cell while the cells where still stimulated by the

DNP. Genistein eliminated the SPR signal. The SPR signal was partially inhibited by

phorbol 12-myristate 13-acetate and wortmannin. This study also tested degranulation by

b-hexosaminidase measurements, reaching the same conclusion as the previous paper.

The control cells, not pre-sensitized by IgE, did not alter SPR signals. This study

established that SPR has to capability to detect biologically significant interactions

between cells and reactive molecules. Their results demonstrated detection of not only

simple binding kinetics between surface receptors on the cells and molecules, but also

reflected intracellular reactions such as the movement of Ca2+ ions.

41

An important finding of this study was that the SPR signal remained high even

after the stimulator was removed from the cell sample. They hypothesized that effect was

caused by other chemical reactions in the cell that were induced by the stimulator. When

analyzing the control, they found that IgE pre-sensitization was required for a cellular

reaction to DNP. Also, the intensity of the signal was dependent on the antigen

concentration. Finally, they proposed that SPR could be used as a way to study the

interaction of molecules with the plasma membrane along with the typical ligand-receptor

sites.

Liu, et al. [141] examined the real time secretion of Vascular Endothelial Growth

Factor (VEGF) using SPR. They cultured living cells on the ceiling of a customized

polydimethylsiloxane (PDMS) SPR flow cell chamber. They used the SKOV-3 (human

ovarian carcinoma) cell line. They coated the SPR chip with a G protein solution containing

antibodies. SPR angle shift indicated the presence of VEGF.

After proving cells could survive in the PDMS flow chamber, the special gasket

was detached from the SPR flow chamber. The control experiments were repeated on an

uncoated PDMS gasket and a tissue culture plate. They concluded that SPR is an effective

tool to measure VEGF biomarker secretion by living SKOV-3 carcinoma cells. Also, their

SPR flow cell chamber set up “mimics the in vivo microenvironment of the VEGF

signaling pathway.” This is important as it could lead to further studies that examine the

cell signaling pathways in regards to drug development.

Baumgarten, et al. [142] studied the effectiveness of SPR for measuring volume

changes in cells. They used two lines of renal epithelial cells: MDCK II and NRK. To fix

cells to the SPR chip, the used glutaraldehyde and non-isotonic media to induce volume

42

changes in cells. They determined, by comparing their results to typical kinetic models,

that SPR can measure the integral, or volume, parameter as a sum of several cellular

reactions.

Upon hypertonic stimulation, the cells showed an increase of reflectivity that

correlated directly with the change in osmolarity. When the signal stabilized, they deduced

that the cells had adapted to the new conditions in their environment. Upon hypotonic

stimulation, they found that the signal decreased as a result of a decrease in osmolarity.

The cells in both tests returned to normal levels after an isotonic solution was added.

Their results concluded that cellular reactions caused by osmotic stress cause a shift

in the cell layer’s refractive index that can be measured with SPR. They found that the

LOD was below 5mOsm/kg (milliosmols per kilogram) and that this is sufficiently

sensitive for bioanalysis.

Yanase, et al. [143] tested how the area of cell adhesion adjusted to the sensor chip

and how the area of the cell reacted to different stimulants by measuring differences in AR

using SPR. They used RBL-2H3 and PAM212 cells and studied the structural changes in

cell membranes when treated with a cell motility inhibitor and antigens. They did observe

a change in the area of cell adhesion, but concluded that “The experimental results

demonstrated that the change in the area of cell adhesion to a sensor chip and that of

membrane structure is insufficient to explain the entire AR change in response to the

activation of living cells.”

Most of their results were expected. They found that antigen and EGF induce large

AR signal changes in both types of cells and that the AR increased proportionally with the

number of cells cultured on a sensor chip. The unexpected result was that there was a higher

43

AR change as a result of an increase in the cell adhesion area in RBL-2H3’s response to

antigen. They concluded that SPR also detects intracellular events and other cellular

changes beyond simply the adhesion area.

44

2.3. Biology

In this section the different bacteria types and strains, which were used in the study,

are described. The literature review highlights the significance of proposed bacteria in this

study.

2.3.1. Biofilms

When bacterial cells stick to the surface, and to each other, and the cellular

concentration reaches a certain threshold, they form colonies. These colonies are

surrounded in a self-produced matrix of exopolysaccharide (EPS) [144-146]. The bacterial

EPS is a complex mixture of polysaccharides, proteins, nucleic acids, lipids, phospholipids

and humic substances, with proteins and polysaccharides contributing up to 89% of the

EPS composition [147-151]. The polysaccharide acts as a shield for microorganisms and

protects them from host immune defense systems; making bacteria inside biofilm up to

1000 times more resistant to any antibacterial chemicals compared to those in suspension

[152-154]. Figure 8 depicts the steps of biofilm formation [77, 155].

Figure 8: Steps of biofilm formation [155].

45

Several factors influence bacterial attachment onto a surface. Some of these factors

are: the condition of substrate surface, the bacterial cell surface properties, the flow rate of

the media passing over the substrate, and the available fresh food for bacterial growth. It

is shown that rougher and more hydrophobic surfaces have higher potential for bacteria to

attach and form biofilm [156-160]. The presence of extracellular appendages and cell

surface hydrophilicity are the key cell surface properties, which play a crucial role in cell-

cell signaling and biofilm formation [157, 161].

Biofilms are reservoirs of bacteria and a source of endotoxins, which both can enter

the circulatory system of a patient and cause systemic disorders. More than 60% of

hospital-acquired infections are caused by bacterial biofilms [162, 163]. Formation of

biofilms is the main cause of many bacterial infections [164].

In medical applications, bacterial biofilms can contaminate implants and

indwelling medical devices such as tracheotomy tubes (TTs) [165], urinary catheters,

venous catheters [144], and can cause many diseases from lung and kidney infections to

tooth decay [166]. Biofilm formation increases the resistance of infections to treatment

procedures and can lead to implant failure and the need for replacement as a result. In the

United States, 80% of nosocomial infections are related to medical implants and indwelling

devices; within this number the fatality rate is 60% [155].

In the food industry, biofilm is a major concern in different sections such as

brewing, meat, poultry, and dairy processing [151, 167-169]. Among all foodborne

pathogenic microorganisms, bacteria cause 91% of total foodborne illnesses in USA [2, 3].

Based on the estimation of the Centers for Disease Control and Prevention (CDC), in the

United States, foodborne pathogens cause roughly 79 million illnesses, 325000

46

hospitalizations and almost 5000 deaths each year [1]. Also, outbreaks of foodborne

illnesses cost billions of dollars each year.

In industry, biofouling and microbial corrosion in marine vessels and oil, water,

and gas pipelines cost significant amounts in both money and time [170-173]. In order to

decrease the health risks, deaths, and reduce economic losses due to pathogenic bacteria,

there is an essential need for rapid, sensitive, and selective detection methods to sense

disease-causing bacteria [6].

2.3.1.1. E. coli

A model prokaryote organism, Escherichia coli is gram-negative, rod-shaped, and

can usually be found in the lower intestine of warm-blooded organisms [174]. A scanning

electron microscope image of several E. coli cells attached to a surface is shown in Figure

9. E. coli can cause serious food poisoning in humans [174]. E. coli in one of the main

food-borne pathogenic microorganisms, inadequate sanitizing of fresh cut food help the

survival of this bacteria. E. coli infections are mainly related to ingesting of ready-to-eat

foods [175-177].

Some strains of E. coli such as E. Coli O157:H7 are responsible for serious

gastrointestinal diseases like diarrhea, hemorrhagic colitis, and hemolytic uremic

syndrome, with a mortality rate of 50% in children and seniors [121, 178]. In the United

States, annually about 270000 cases of illnesses are caused by E. coli [124].

The ability to grow in both aerobic and anaerobic environments, fully identified

gene sequence and the high rate of growth make it the best option in the field of

biotechnology. Having peritrichous flagellation, motility mechanism of E. coli has been

47

studied as a model for microrobotics, which consequently will be used in drug delivery

systems [179, 180].

In addition to avoiding economic damage to food companies, studying E. coli

growth is providing information for future biological problems. Most experiments in this

proposal will use genetically modified Green Fluorescent Protein (GFP) E. coli K12,

obtained from the laboratory of Prof. Veronica Godoy-Carter in the Department of Biology

at Northeastern University. The Goluch Group also has m-cherry labeled E. coli K12, and

access to other strains via on-campus collaborations.

Figure 9: Scanning Electron Microscope image of E. coli [174].

2.3.1.2. P. aeruginosa

Pseudomonas aeruginosa is a model organism for investigating biofilm formation

and pathogenesis [181]. P. aeruginosa is a gram-negative, rod-shaped bacterium (Figure

10) [182] that has incredible nutritional flexibility, capable of consuming more than

seventy-five different organic compounds, which it is why it is one of the most abundant

48

organisms on earth [183]. P. aeruginosa has been found in environments such as soil,

water, humans, animals, plants, sewage, and hospitals.

P. aeruginosa is an opportunistic human pathogen, meaning it rarely infects healthy

individuals. It is considered a serious problem for patients whose immune system has been

compromised, like those hospitalized with severe burns, cancer, AIDS and cystic fibrosis

[184]. P. aeruginosa is the most commonly found gram-negative bacterium in hospital

acquired infections, carrying a 40-60% mortality rate and listed as one of most frequent

gram-negative pathogens [185-188]. Early detection of P. aeruginosa growth will

potentially decrease the high rate of deadly infections in immunocompromised patients.

P. aeruginosa has many strains: P. aeruginosa PA01 and P. aeruginosa PA14 are

two common strains, which have had their complete genomes sequenced. A comparison

between the two shows that although the genome of PA14 (6.5 Mbp) is slightly larger than

that of PA01 (6.3 Mbp), both genomes are very similar. There are 58 gene clusters from

PA14 that are missing in PA01 and it is proposed that some of these genes are what make

PA14 more infectious than PA01 [189, 190].

49

Figure 10: Scanning Electron Microscope image of Pseudomonas aeruginosa [182].

2.3.1.3. S. aureus

Staphylococcus aureus is a gram positive and facultative anaerobic bacteria, which

has spherical shape (Figure 11) [191]. S. aureus causes hospital-acquired diseases such as

endocarditis, osteomyelitis and abscesses. Staphylococcus bacteria is able to adhere to the

surface of the indwelling medical devices and form biofilm on different surfaces, such as

plastics and metals [192, 193]. This capability makes it one on the most serious blood

stream pathogens responsible for about 38% of this type of infection [194].

Early stage detection of bacterial attachment on the surfaces plays crucial role in

prohibiting further growth and biofilm formation. SPRi provides kinetic information of

biofilm formation on the surface and is able to monitor the surface in real-time, as bacteria

grow, attach the gold surface and form biofilms on it.

50

Figure 11. Scanning Electron Microscope image of Staphylococcus aureus [191].

2.3.1.4. Bacillus species (Bacillus cereus)

B. cereus is a gram positive, spore forming, and rod shape bacteria (Figure 12)

[195]. B. cereus is facultative anaerobic food bacterium [196]. The ability of this species

to form spores make them resistance to chemicals, a wide range of pH and temperatures;

the spores can also survive for decades. B. cereus causes two forms of food poisoning: the

diarrhoeal form and emetic form [196-201]. The emetic form is caused by cells growing

and depositing toxins in food, the diarrhoeal syndrome is the result of different toxins that

can be formed in the food and in the small intestine [196, 201, 202]. B. cereus bacteria are

the main contamination source of milk and dairy products and are also known as school

related pathogens because many schools provide dairy products on a daily basis for

51

children [201, 203]. B. cereus is one of the most damaging pathogens in ocular infections

which progress in 24-48 hours and in many cases lead to vision loss [202, 204-206].

n biological studies and development of biosensors, B. cereus spores are used as

simulants for the dangerous spores of B. anthracis [207]. B.anthracis exists in the form of

rod shape bacteria and spores. B.anthracis is a dangerous pathogen, which causes anthrax

in humans through injection and inhalation [208-211].

Figure 12: Scanning Electron Microscope image of Bacillus cereus bacteria[212].

52

3.0 DISSERTATION GOALS

Goal 1. System Setup for Monitoring Bacterial Growth and Biofilm Formation (Goal 1)

The Changes due to particle attachments were monitored with SPRi

Growth of GFP E.coli bacteria was monitored and compared with a control

channel

Mutant P. aeruginosa (PA14) and Wild type P. aeruginosa (PA14) was

monitored with SPRi to compare and differentiate bacterial growth with

bacterial biofilm formation

Goal 2. Prevention of Biofilm Formation on the Surface (Goal 2)

The preventative effect of different surface coatings were evaluated with

SPRi system

The effect of different antibiotics at various concentrations were monitored

in prevention of bacterial growth

Goal 3. Biofilm Removal from the Surface (Goal 3)

SDS was used to clean the biofilm from the substrate and the process was

monitored with SPRi

The effect of different antibiotics were studied in disinfecting bacterial

biofilms using SPRi

Goal 4. Effects of flow rates on Bacterial Growth (Goal 4)

The effect of flow rate and consequently shear stress on bacterial growth

was studies in real-time with SPRi

53

The shear stress distribution profile was simulated with COMSOL

multiphysics modeling software

54

4.0 EXPERIMENTAL

The overall goal of this research was to develop a label-free methodology to study

bacterial behavior and biofilm formation under different conditions. In the first level, the

effect of different surface coatings, different antibiotics in solution and various flowrates

on initial bacterial attachment and further biofilm formation on the substrate are studied

with SPRi. Next, biofilm removal with different chemicals and biofilm disruption with

antibiotics are also investigated. In this section, experimental setups were developed to

combine microfluidic devices with a SPRi instrument so that it can be used to study general

bacterial behavior such as bacterial growth and biofilm formation. The following

experiments were designed to provide an analytical approach for studying biofilm related

problems including growth kinetics, biofilm formation prevention, and biofilm removal.

4.1. System Setup for Monitoring Bacterial Growth and Biofilm Formation

(Goal 1)

The overall goal of this research was to study bacterial behavior and biofilm

formation under different conditions. In the first goal, the capability of SPRi to real-time

monitor changes on the surface as biological interactions took place were investigated.

4.1.1. Monitoring Changes on the Surface with SPRi

When something attaches to the sensor surface in SPRi, it will change the resonance

condition in the binding spot, this change will prevent the incident light from reacting with

the surface free electron and as a result it will prevent conversion of light to SPs.

Consequently, a portion of the incident light will be reflected back, which causes a bright

spot to appear in the SPRi difference image. In the first step, the ability of the device to

monitor the changes on the surface as relatively large objects (much greater than 200 nm)

55

attach to it was studied. For this purpose, microscale beads was used on the chip surface.

It was expected that the attached beads would change the reflectivity and would appear as

bright points in the difference image.

Figure 13: Setup for initial Surface Plasmon Resonance imaging (SPRi) experiments.

50 m beads in DI water was placed onto a prism coated with 50 nm of

gold. The setup was placed inside of a SPR imaging system (Horiba).

4.1.2. Monitoring Bacterial Growth with SPRi

Once detection of beads was accomplished with SPRi, bacterial growth was

examined. This was carried out by loading bacteria in solution in micro channels on the

sensor surface. This would allow the growth and movement of bacteria, and the formation

of biofilm was monitored for the first time with a SPRi device. To study bacterial growth,

E. coli labeled with Green Fluorescent Protein (GFP) was injected in a linear microchannel

placed on the sensor surface and the changes on the surface as bacteria attach were

monitored continuously with the SPRi device. At the same time, a parallel microchannel

was filled with only growth media as a control. There was no growth in this channel and

nothing would attach the surface during the experiment. To study bacterial movement and

to identify the preferred areas for initial aggregation of bacteria prior to forming biofilm, a

56

droplet of GFP-labeled E. coli (of an order of magnitude in the microliter range) was placed

on the prism and covered by a PDMS chamber. While bacteria swam around, some of them

touched the surface, moved and touched another location. This entire process was

monitored with the instrument.

Figure 14: Setup for Surface Plasmon Resonance imaging (SPRi) experiments.

PDMS with two channels was reversibly sealed against a high refractive

index glass prism coated with 50 nm of gold. The left channel was filled

with LB growth media and the right channel was filled with GFP labeled

E. coli. The setup was placed inside of a SPR imaging system (Horiba).

4.1.3. Monitoring Bacterial Biofilm Formation

Bacterial biofilm formation was monitored with the SPRi device. In biofilms,

bacterial cells surround themselves with a complex matrix of exopolysaccharides (EPS)

and other biomolecules, which result in the appearance of much larger spots on the SPR

image than individual cell attachment events. For these experiments, different strains of P.

aeruginosa bacteria were injected in separate linear microchannels placed on the prism

57

surface, as shown in Figure 15. To all direct comparison across experiments, one of the

channels was filled with only growth media. The images of this channel was used as a

control in the experiment. To compare the biofilm formation and bacterial growth, two

different strains of P. aeruginosa were used: PelA mutant PA14 and Wild Type (WT)

PA14. The PelA gene in P. aeruginosa is responsible for producing biofilm, and when

knocked-out this strain performs its natural bacterial functions with the exception of

producing biofilm. The wild type PA14 forms biofilm as it grows during the experiment.

The results of this experiment provided useful information to compare bacterial growth

and biofilm formation.

Figure 15: Setup for monitoring biofilm formation using Surface Plasmon Resonance

imaging (SPRi). PDMS with three channels was reversibly sealed against

a high refractive index glass prism coated with 50 nm of gold. The left

channel was filled with trypticase soy broth, middle channel was filled

with PelA mutant P. aeruginosa PA14 and the right channel was filled

with wild type PA14. The setup was placed inside of a SPR imaging system

(Horiba).

58

4.2. Prevention of Biofilm Formation on the Surface (Goal 2)

Considering the fact that prevention of biofilm formation is more ideal than treating

it, the goal of this aim was to study the effects of different surface coatings on bacterial

growth and biofilm formation as well as the effect of antibiotics on growth kinetics of

bacterial cells. In all the following experiments bacteria were cultures in 6 mL of fresh LB

at 37⁰C for 18 hours. Then they were diluted in fresh LB by 1:100 ratio. Different amounts

of this diluted media were used for each experiment.

4.2.1. Surface Coatings

In this experiment, three different biomolecules were chosen as surface coatings to

study their effect on biofilm formation on the surface. Penicillin/Streptomycin, Casein and

Bovine Serum Albumin (BSA) were chosen for this aim because of their specific properties

mentioned below.

The prism surface was coated with two different biomolecules at the same time to

allow a direct comparison between coatings. Growth media containing bacteria was loaded

in the PDMS chamber placed on top of the prism, and the effect of the coating on

preventing bacterial adhesion to the surface was studied.

Penicillin is a well-known antibiotic for treatment of Gram positive bacteria such

as Staphylococcus aureus. Penicillin belongs to the class of β-lactam antibiotics and, as

other antibiotics in this class, it works by inhibiting cell wall synthesis in bacterial cells.

Casein is the main protein found in milk and is known to decompose slowly. It has

the ability to release amino acids for up to 7 hours. The effect of casein as a coating on

implants to prevent biofilm formation and further infections is important, as it is

inexpensive and known to be compatible with human body.

59

Bovine Serum Albumin (BSA) is a common blocking agent, which is used in many

biochemical experiments to prevent non-specific bindings [213]. Casein and BSA are both

commonly used proteins in microfluidics to avoid non-specific bindings on the surface.

During the biofilm formation process, bacteria first attach to the surface to form

colonies. In this stage, the refractive index of the attachment spot changes and this variation

is detected by the SPRi sensor. The changes in the refractive index appear as bright spots

on the initially dark difference image. SPRi provides difference images of the surface every

3 seconds and offers real-time monitoring of bacterial growth on the gold surface.

The SPRi difference images at different time intervals provide useful information

about biofilm formation on the surface. As bacteria grow and attach to the surface, bright

spots start to appear on the surface. The more bacteria attach to the surface, the larger the

change in the refractive index above the sensor surface, and as a result, the bright area

expands and gets brighter.

The mean value of the contrast change for the three different surface sections will

be calculated for each experiment. A graph displaying the contrast change during the

experimental period will present the bacterial growth on the sections of the surface coated

with different biomolecules. When bacteria attach to the surface, the contrast of the

grayscale SPRi difference image will increase; this rise in contrast represents a

proportional increase in the amount of biofilm on the surface.

Having both SPRi difference image and graphs showing contrast variation during

for a set of experiments will provide useful information about bacterial growth kinetics and

biofilm formation on different surfaces.

60

Figure 16: Setup for coating experiments using SPRi. The PDMS chamber is placed

on the gold-coated prism surface, and the gold surface inside the

chamber was coated with two biomolecules. The setup was placed inside

of a SPR imaging system (Horiba).

4.2.2. Loading Antibiotics in Solution

Penicillin with streptomycin, a well-known antibiotic combination for treatment of

gram positive bacteria such as Staphylococcus aureus, Colistin, an antibiotic for treatment

of P. aeruginosa, and Spectinomycin, an alternative antibiotic for patients who are allergic

to penicillin, for treatment of Bacillus cereus were used in this study. Antibiotic effects

were being tested in this fashion to simulate conditions for indwelling devices that may

remain in the body for a few hours or days and cannot be coated or modified in different

ways.

The effect of penicillin/streptomycin on growth and biofilm formation of S. aureus

was studied by placing the bacteria solution on the surface and inside the PDMS chamber.

Penicillin/streptomycin solution was added to the growth media at the inlet, which would

61

flow continuously during the experiment period to provide fresh food for the bacteria in

the chamber. SPRi provided difference images every three seconds and let us monitor

bacterial adhesion and biofilm formation on the surface in real-time.

The effect of Colistin on the adhesion and biofilm formation of P. aeruginosa was

studied by placing the bacterial solution on the surface and inside the PDMS chamber.

Colistin was added to the inlet media, which would then pass over the surface continuously

during the entire experiment to provide fresh media for bacteria inside the chamber. SPRi

difference images was let us monitor biofilm formation on the surface in real-time and

study the adhesion of bacteria under antibiotic treatment.

The effect of Spectinomycin on B. cereus growth on the surface was studied

following the same procedure. PDMS made chamber was placed on the gold surface and

200 µL of solution of B. cereus in LB (1:100 v:v) was loaded inside the chamber. The

entire setup was then loaded into the SPR system and the spectinomycin solution in fresh

LB was passed over the surface continuously with peristaltic pump. SPR imaging provided

real-time images from the surface, which represented B. cereus growth under the effect of

spectinomycin antibiotic. Figure 17 shows the system setup for this part of the research.

62

Figure 17: Setup for studying the effect of antibiotics on prevention of biofilm

formation using SPRi. The PDMS chamber was placed on the gold-

coated prism surface, bacteria media in loaded inside the chamber and

antibiotic solution was flowed over the surface. The setup was placed

inside of a SPR imaging system (Horiba).

63

4.3. Biofilm Removal from the Surface (Goal 3)

Currently, there is no established method for permanently and completely

preventing biofilm formation. In this section, biofilm removal from a contaminated surface

was studied with SPRi. After early stage detection of biofilm formation, which was

discussed in Goal 1, the efficacy of different chemicals to remove biofilm and the rate of

removal from the surface were studied in real-time with SPRi. It was the first time SPRi

has been used for biofilm removal studies and would provide kinetics and real-time

information for cleaning procedures. In this aim, the effects of cleaning with chemicals,

and disinfection using antimicrobial agents was studied with the SPRi device.

4.3.1. Cleaning with Different Chemical Compounds

Effective cleaning of biofilm is the first step in eliminating bacterial infection.

Ineffective cleaning will cause subsequent steps to be unsuccessful, because antimicrobial

products cannot easily reach the cells living inside a biofilm to kill them [151]. Commonly

used chemicals for removing the biofilm from a contaminated surface are surfactants and

alkali products. These compounds act as cleaning agents by denaturing proteins and

reducing surface tension, which leads to dissolution of biofilm [129, 214, 215]. However,

these processes have never been observed in real-time. The effects of well-known cleaning

chemical, Sodium dodecyl sulfate (SDS) on biofilm removal from contaminated surfaces

was studied. SDS is and anionic surfactant which is used in many cleaning detergents. SDS

has been chosen because of its common use in industry and the laboratory. The goal is to

study the effectiveness of tis chemical on biofilm removal.

For this study, bacteria media was first placed on a gold-coated prism surface and

inside the PDMS chamber. Lysogeny Broth (LB) growth media was run over the surface

64

for 24 hours to provide fresh food for the bacteria and allow the biofilm to form on the

surface. After biofilm formation on the surface for 24 hours the inlet solution was switched

from fresh LB to SDS. SDS was then flowed over the surface for few hours to remove the

biofilm. In this process when the solution reaches the surface, it chemically affects the

biofilm on the surface and, if the chosen chemical performs as expected, it removes the

biofilm. To eliminate the effect of running solution in reflectivity variation, at the end of

SDS run, fresh LB was run over the surface for few more hours. This step allowed having

accurate comparison on the reflectivity levels before and after SDS effect. The procedure

was repeated several times.

In these experiments, the position where biofilm is formed appeared as large bright

areas in the SPRi difference image due to its different refractive index and its effects on

the resonance condition on that area. When bacteria were removed from the surface by the

loading chemicals, the running media replaced them and the refractive index changed back

to its first value. This resulted in the disappearance of the bright spots in the difference

image as a result of biofilm removal. The whole procedure was monitored with SPRi. The

results provided useful information about effectiveness of SDS on biofilm removal and the

kinetics of cleaning.

4.3.2. Disinfection with Antimicrobial Components

A cleaning procedure can remove 95% or more of a biofilm; however, as for killing

the microorganisms, disinfection of bacteria are required. This is a crucial step in both

science and industry; if the bacterial cells are not all killed after removal of the bulk

biofilm, they can move and deposit in a new location and form another biofilm, which will

increase the time and cost required [151, 216]. Disinfection products should be safe,

effective, and easy to use, and they should leave no toxic residues behind. Disinfecting is

65

defined here as the use of antimicrobial products to kill the microorganisms within the

biofilm. When bacteria inside the biofilm are killed they will detach from the surface,

biofilm structure will be disrupted. This would make biofilm removal with other techniques

(mechanical and chemical ways) more effective. At the same time disinfection step will

reduce the number of viable cells on the surface and decrease the chance of biofilm

reformation and subsequent infections.

In this part of the research, the effects of different antibiotics on killing

microorganisms and eliminating biofilm attachment on the surface was studied.

Penicillin/streptomycin solution for S. aureus bacteria, Colistin for P. aeruginosa, and

Spectinomycin for B. cereus was used. It is believed that a large, rapidly administered dose

of antibiotics will inactivate the susceptible living microorganisms and will prevent

regrowth of biofilm even though bacterial cells can increase resistance toward antibiotics

by mutation or genetic exchange [151, 217-220].

To test the effectiveness of disinfection, after biofilm is formed on the surface for

24 hours, solution of different concentrations of antibiotics in fresh LB was run over the

surface for another 24 hours to provide enough time for antibiotics to diffuse through the

new biofilm on the surface and effect bacteria inside the biofilm. The changes in the

reflectivity as antibiotics effect biofilm was continuously monitored with SPRi.

66

4.4. Effect of flow rate on bacterial growth (Goal 4)

In this part of the research the effect of various flowrates on biofilm formation was

studied. The first step toward biofilm formation is bacterial irreversible attachment on the

surface. Theoretically different flowrates over the surface will result in various shear

stresses and will interfere with bacterial initial attachment on the surface. Also mechanical

force is considered the most effective method for biofilm removal and is used whenever

possible. In this investigation, the effect of shear stress as a physical method for prevention

of initial bacterial attachment and further biofilm formation was studied by modeling the

system with COMSOL Multiphysics and also experimentally with SPRi system.

4.4.1. SPRi Experiments

The SPRi device includes a fluidic flow system, which allows the operator to set

the flow rate of the inlet solution. In this part, hexagon shape PDMS made chamber was

placed on the gold surface. Then bacterial media was loaded inside the chamber with

pipette. The entire setup was then placed inside the SPRi system. Fresh LB at various flow

rates was flowed over the surface for 24 hours. Figure 18 shows the system setup for SPRi

experiments. Flow rate affect the fluidic shear stress on the surface where bacteria grew

and formed biofilm. The SPRi technique was used to study these effects in real-time and

monitor the surface as bacteria attach and form biofilm under different flow rates. The

section where biofilm was formed on the surface had a different refractive index and

appeared as a bright area on the difference image. SPRi difference images, which were

taken every three seconds, let us monitor the effect of flow rate, and therefore shear stress,

on bacterial attachment and biofilm formation over the period of 24 hours.

67

Figure 18: Setup for studying bacterial growth under different flowrates using

SPRi. The PDMS chamber was placed on the gold-coated prism surface,

and bacterial media was loaded inside the chamber. The setup was

placed inside of a SPR imaging system (Horiba).

4.4.2. COMSOL Multiphysics Modeling

In this part the system setup was modeled using COMSOL Multiphysics modeling.

COMSOL Multiphysics software was used to simulate and model shear stress profile in

the conduits. Simulation will be used to study the effect of flowrate on the consequent

shear stress of the flow on the surface. Shear stress itself will influence the biofilm

formation mechanism on the sensor surface. In this model, the non-slip wall boundary

condition was chosen for the hexagon shape PDMS chamber. The fluid was following

under the laminar flow condition. The results of this model provided useful information

about the shear stress distribution over the entire surface of our system.

68

5.0 RESULTS AND DISCUSSION

This section will outline the preliminary results obtained thus far. The results

include monitoring of bacterial growth and biofilm formation with SPRi, using surface

coatings and antibiotics to prevent biofilm formation, and using chemicals for biofilm

removal.

69

5.1. System Setup for Monitoring Bacterial Growth and Biofilm Formation (Goal

1)

The results presented in this section are partly from “Using surface plasmon

resonance imaging to study bacterial biofilms” in Biomicrofluidics, volume 8, 2014. The

results are presented with permission from Biomicrofluidics journal.

5.1.1. Bead Imaging

To first demonstrate the ability of SPRi to detect microscale objects, a control

experiment was performed using 50-µm-diameter beads. A drop of deionized (DI) water

containing a very dilute amount of 50-µm-diameter beads was placed directly onto the gold

coated prism surface (Figure 19) and a single image was obtained using SPRi. Knowing

that the majority of the surface would not have beads on it and should therefore remain

dark when imaged, it was possible to find the critical angle for imaging the surface. The

prism was then removed from the instrument and a series of images of the surface were

acquired using a CCD camera mounted on a stereo microscope with illumination from an

oblique angle. Since it is not possible to obtain high magnification images of a square

centimeter using a microscope, the SPRi image was cropped for comparison against

micrographs with similar surface areas (Figure 19: b, d, f, h). A comparison of the SPRi

images with the micrographs of corresponding areas of the surface (Figure 19: (a, b) (c, d)

(e, f) (g, h)) reveals that it is possible to detect individual 50-µm-diameter beads using

SPRi. Bead aggregates appear as larger bright areas in both sets of images. It is interesting

to note that a halo is observed around the beads in the SPRi image; it is believed that this

is caused by SPPs reflecting from the spherical beads. This effect is also observed at

chamber walls.

70

Figure 19: (Right column) SPRi images of 50µm beads, (left column) Stereo

microscope fluorescent images of the same beads (a, e), (b, f), (c, g), (d, h)

microscope and SPRi images of same spots.

71

5.1.2. Cell Growth and Biofilm Formation

Images of the surface were acquired every three seconds, making it possible to

observe movement at the bottom surface of a chamber. A large chamber was filled with E.

coli in LB and allowed to grow over night. The chamber was several millimeters tall,

allowing cells to move far away from the surface.

5.1.2.1 E. coli growth and biofilm formation

We first used SPRi to detect the growth and movement of GFP-labeled E. coli along

the sensor surface inside a closed microchamber. To perform experiments simultaneously,

two separate rectangular chambers were made in one piece of PDMS. One of the chambers

was filled with 5 µl of LB growth media as a control while the second chamber was filled

with GFP-labeled E. coli in LB as shown in Figure 20, top. Prior to filling, the bacteria

were grown at 37⁰C for 7 h to reach exponential growth phase. Difference images of the

two chambers were obtained simultaneously using SPRi for 6 h (Figures 20(a)–20(f)) at

room temperature, and were compared with fluorescence images obtained with a stereo

microscope of the same areas on the sensor chip (Figures 20(g) and 20(h)).

As shown in Figure 20(b), after 6 min, bright spots begin to appear in the bacteria

containing chamber as cells begin to attach to the surface, while the control chamber

(Figure 20(a)) remains completely dark. Bright spots indicate a change in the refractive

index of the surface over time, which, in the context of this experiment, translates to

biomass accumulation and bacterial growth on the surface of the prism. After 1 h, Figure

20(d) shows an increase in the number of bright spots, as new cells divide and attach to the

surface. The spot size also increases as an extracellular matrix is created around the

adherent cells. A few bright lines appeared in the control chamber (Figure 20(c)). The lines

correspond to features, such as scratches, on the gold. This phenomenon is observed in

72

SPR images after a few minutes when the bulk fluid is stagnant. In this experiment, the

chamber access holes were not sealed and after 6 h, the fluid in the chambers evaporated.

Since air has a different index of refraction than water, images of the chamber surface

became bright. The resulting surface morphology that is observed in the chambers after

they were dried is quite interesting. The control chamber in Figure 20(e) is much smoother

than the chamber containing cells (Figure 20(f)) where a biofilm formed on the surface.

Finally, the chip was removed from the instrument, the PDMS chambers removed

from the prism, and fluorescence images of the surface were taken. The surface inside the

control chamber remained dark (Figure 20(g)) while the chamber containing cells had

many fluorescent cells on the surface (Figure 20(h)).

Figure 20(f) shows that the entire surface is covered with biomass, while only a

few fluorescent areas are visible in Figure 20(h). There are multiple reasons why the

fluorescent image in Figure 20(h) does not match what is observed in Figure 20(f). Most

notably, the extracellular matrix and dead cells do not fluoresce and hence are not visible

in Figure 20(h). In addition, a relatively low camera exposure time was used to obtain the

images to highlight the location of live cells and minimize photo-bleaching as multiple

images across the surface had to be acquired to confirm our findings. Unfortunately, the

opaque gold substrate prevents transmission bright field imaging.

Next, we inoculated GFP-labeled E. coli in LB. We placed 200µL of bacterial

culture on the gold-coated prism and put a large PDMS chamber on top of it to isolate the

media from the surrounding environment. The solution did not touch the PDMS chamber

in these experiments.

73

The chamber was several millimeters tall, allowing cells to move far away from the

surface. The bacteria in this experiment were not incubated at 37⁰C prior to loading; instead

they were grown the entire time at room temperature which increased the amount of time

needed to form a biofilm. Difference images taken every 3 s with SPRi let us monitor

bacterial growth and biofilm formation in real time. Figures 21(a)–21(i) show the initial

biofilm formation around 7 h after the start of the experiment. We took fluorescence images

(Figure 21(e)) of the same location on the sensor surface to confirm the SPRi results.

Similar to the images shown in Figure 19, many cells died when the solution in the chamber

was dried prior to fluorescence microscopy and therefore Figure 21(j) does not match the

features shown in Figure 21(i). The drying effects can be seen on the lower right portion

of Figure 21(e), where salt crystal dendrites have formed. The general size and shape of

the fluorescent region in Figure 21(e) do match what was observed with SPRi.

To confirm that the location of biofilm formation was not affected by external

factors, such as impurities on the sensor surface or exposure to non-uniform light intensity,

we repeated the experiment several times by placing the large PDMS chamber on different

ends of the prism. We also performed the experiment with multiple gold sensor surfaces.

In all experiments, the bacteria consistently gathered in the middle of the chamber. We

hypothesize that the center of the hemispherical shaped droplet provides a higher

concentration of available nutrients and signaling molecules for the bacteria. Control

experiments without bacteria added to the chamber did not show significant contrast

change after 24 h.

74

Figure 20: (Top) Schematic of the setup for E. coli SPRi experiments. A PDMS chip

containing two microchambers was reversibly sealed against the sensor

surface. (Bottom) SPR images of LB filled channels at (a) 6 min, (c) 1 hr,

(e) 6 hr, GFP E.coli filled channels at (b) 6 min, (d) 1 hr, (f) 6 hr.

Fluorescence images of (g) channel filled only with LB, and (h) GFP E-

coli filled channel. Each image is at the same magnification. White lines

have been added to the images to highlight the location of channel

sidewalls.

75

Figure 21: Difference images taken with SPRi at a) 6,45’; b) 6,47’; c) 6, 49’; d) 6,51’;

e) 6,53’; f) 6,55’; g) 7 hours; h) 7,6’; i) 7,10’ are shown. The arrows are

pointing to the center of the GFP labeled E. coli bacterial media droplet,

where the bacteria preferred to gather. j) A fluorescent image of the

surface of the prism surface after being removed from the SPRi system.

76

5.1.2.2 P. aeruginosa growth and biofilm formation

The robust biofilms that P. aeruginosa forms aid significantly in its pathogenesis

and a better understanding of the initial cell adhesion and biofilm formation processes will

provide significant insight into strategies for preventing infections. We monitored biofilm

formation of two P. aeruginosa PA14 strains, wild type and mutant pelA. A PDMS chip

containing three linear chambers was placed on the sensor surface. One chamber contained

no bacteria, the second chamber was filled with pelA mutant, which cannot produce a

robust biofilm, and the third chamber was filled with wild type, which produces biofilm.

Trypticase soy broth was used as growth media in all three chambers. The pelA gene is

responsible for biosynthesis of cellulase-susceptible polysaccharide that is essential for

formation of robust biofilms, but has no influence on the cellular adhesion [221].

Prior to being loaded into the microchambers, the two bacterial strains were incubated at

37⁰C for 5 h to initiate exponential growth. Difference images of the loaded chambers were

collected for 3 h using SPRi (Figure 22), and afterwards, the sensor surface was imaged

with a stereo microscope to confirm the formation of biofilm. We repeated this experiment

three times and obtained consistent results. The chamber without cells remained dark

(Figure 22(a)). Some growth was observed in the chamber containing the pelA strain

(Figure 22(b)). The small bright spots in the image are cells that attached to the surface and

began growing. Biofilm growth in the chamber with wild type cells was extensive after 3

h in this small (approximately 3 µL) fluid volume (Figure 22(c)).

77

Figure 22: (Top) Schematic of the biofilm formation experiments using SPRi. PDMS

with three channels was reversibly sealed against a high refractive index

glass prism coated with 50 nm of gold. The left channel was filled with

trypticase, middle channel was filled with PelA mutant PA14and the right

channel is filled with wild type PA14. The setup was placed inside of a

SPR imaging system (Horiba). (Bottom) Difference images of SPRi in

after 3hours. The left column is the difference images taken with SPRi of

channel filled with just trypticase soy broth, the middle column is the

difference images of channel filled with PelA mutant PA14 and the right

channel is difference images of Wild type PA14.

78

Next, we inoculated CFP-PA01 in LB. We placed 200µL of bacterial culture on the

gold-coated prism and put a large PDMS chamber on top of it to isolate the media from

the surrounding environment, similar to the procedure in the last experiment with GFP-

E.coli. The solution did not touch the PDMS chamber in these experiments.

The chamber was several millimeters tall, allowing cells to move far away from the

surface. The bacteria in this experiment were not incubated at 37⁰C prior to loading;

instead, they were grown the entire time at room temperature, which increased the amount

of time needed to form a biofilm. Difference images taken every 3 s with SPRi let us

monitor bacterial growth and biofilm formation in real time. Shown in Figure 23(a)-23(h)

are SPRi images of P. aeruginosa PAO1 grown in a large chamber overnight at the same

conditions as GFP-labeled E. coli shown in the last experiment. The PAO1 cells behave

differently than E. coli cells. PAO1 cells initially formed a biofilm at the edges of the

droplet, and only later began forming a biofilm in the center. This knowledge of biofilm

assembly may potentially be exploited to identify bacterial species in unknown samples.

The spots outside of the fluid region are caused by contact of the PDMS chamber with the

sensor surface. They were present throughout the experiment and did not change in size or

shape. After SPRi, the prism was removed from the system, the biofilm was dried, and the

PDMS chamber was removed. Figure 23(i) is a GFP-filtered fluorescence image of the

right side of the biofilm, which crystallized during the drying process. A camera exposure

time of 50 ms was used, which makes the image very overexposed. A clear boundary is

visible where the biofilm ends and the PDMS chamber begins. The image is the brightest

directly next to the PDMS where a large amount of cells are located and the fluorescence

intensity decreases when moving away from chamber wall. Figure 23(j) is a SEM image

79

of a portion of the biofilm on the right side of the chamber. The gold surface on the right

side of the image, which was not exposed to bacteria, is very clean. The drying process

potentially adds more cells to the surface than there were adhered in the SPRi, however,

this can be distinguished partly as cells originally immobilized on the surface have

extracellular matrix formed around them. The fixing process can also remove cells that are

not tightly bound to the surface, thus preventing an exact comparison. The SEM image

shows a biofilm geometry that is consistent with the SPRi and fluorescence microscopy

images. On the left side of Figure 23(j), after the cellular biofilm ends, there is biomaterial

on the surface that is measured with SPRi with lower contrast than the cells at the chamber

perimeter. This biomaterial is distinguishable in the SEM from the clean gold surface on

the right side of Figure 23(j). Figure 23(k) is a SEM image from the center of the chamber

area where a biofilm was forming and contrast was increasing in the SPRi at the end of the

experiment.

80

Figure 23: Images of P. aeruginosa PAO1 after being grown overnight in LB growth

media. (a-h) SPRi images after overnight growth. The red arrow on the

right side of (h) points to the area that is shown in (i,j). The green arrow

in the center of (h) points to the area shown in (k). (i) GFP-filtered

fluorescence image of the right side of the dried biofilm on the sensor

surface. (j) SEM images of the right side of the biofilm. (k) SEM image of

the center of the chamber.

81

5.2. Prevention of Biofilm Formation on the Surface (Goal 2)

The results presented in this section are from “Surface plasmon resonance imaging

(SPRi) for multiplexed evaluation of bacterial adhesion onto surface coatings” in

Analytical Methods, volume 7, 2015. The results are presented with permission from

Analytical Methods journal.

The goal of this study was evaluating the effectiveness of different parameters on

prevention of bacterial growth and biofilm formation. In this regard, the preventative effect

of various surface coatings were simultaneously studies and compared. Also the effect of

certain concentrations of antibiotics in decreasing biofilm formation on the surface was

monitored in real time.

5.2.1. Surface Coating

In this section, the effects of different surface coatings, such as Casein, BSA, and

Penicillin/streptomycin on preventing bacterial biofilm formation on the sensor surface are

presented. As it is possible to monitor a 1 cm square are of the sensor surface with the SPRi

device, the surface was coated with two different biomolecules at the same time, and the

results are compared with the non-coated gold surface.

5.2.1.1. Casein and BSA

In the next set of experiments, the effects of casein, a well-known family of proteins

for preventing biomolecular attachment in microfluidic applications [222], and BSA, a

hydrophilic protein frequently used to prevent non-specific biofouling [223, 224], were

investigated. First, P. aeruginosa adhesion was evaluated on a sensor surface with BSA

coated on the left side, casein coated on the right, and bare gold in the middle. The graph

in Figure 24 shows the normalized mean reflectivity for each coating, with error bars

82

representing the standard deviations. After 24 hours of exposure to flow containing P.

aeruginosa, the BSA coating had about 15% less biomass and the casein coating had over

80% less biomass than bare gold.

83

Figure 24: SPRi difference images of P. aeruginosa (CFP-PA01) growth on the gold

surface coated with BSA (left) and casein (right) after (a) 6, (b) 12, (c) 18,

and (d) 24 hours. Differences in brightness are clearly distinguishable

between surface coatings. The regions selected for analysis in each of the

images are outlined in panel (a) using yellow dashed lines. (e) Changes in

the mean value of the reflectivity caused by binding of P. aeruginosa to

portions of the surface coated with BSA, casein and non-coated (red brick

= BSA, solid brown = bare gold, horizontal green lines = casein). Error

bars show the standard deviations of three separate experiments. *

indicates a change from the control bare gold at each time (p < 0.05, 2-

tailed t-test with unequal variance).

84

Next, BSA and casein were exposed to flow containing planktonic S. aureus cells.

Images in Fig. 25a–d show the difference images from one experiment at six-hour intervals

and reveal the formation of biofilm on the surface. Fig. 25e shows the results of the data

analysis, which was performed the same way as for the P. aeruginosa experiments. After

24 hours, the BSA coating had 20% less biomass and the casein coating had 60% less

biomass than bare gold. The increased standard deviation for the results using S. aureus,

shown in Fig. 25, is attributed to the greater variability in attachment behavior for this

species. S. aureus cells tend to cluster together and attach to surfaces in clumps at the tested

flow conditions. The greater contrast increase for bare gold, shown in Fig. 24 and 25,

indicates more biofilm formation in comparison to the two coatings. Fig. 24 and 25

indicate, both qualitatively and quantitatively, that casein prevented biofilm attachment

onto the surface. BSA provided a statistically significant decrease in biofouling for P.

aeruginosa, but was not effective at preventing S. aureus attachment. Given that bacteria

with hydrophobic cellular membranes are generally attracted to hydrophobic surfaces and

repelled by hydrophilic surfaces, the results support a conclusion that the casein coating

remains hydrophilic throughout the experiment. The BSA was perhaps partially denatured

by the gold surface, exposing some of its hydrophobic amino acids to the solution [225,

226]. We ran a contact angle test and the casein-coated surface wetted significantly more

than the BSA-coated side. A drop of water on the casein surface exhibited a contact angle

of less than 20⁰ while the contact angle on BSA was ~45⁰, indicating that the BSA coating

creates a less hydrophilic surface. A control SPRi experiment without the addition of

bacteria showed negligible change in brightness for both coatings, indicating that the

85

material was not removed or degraded by the flow, and thus we do not expect cross

contamination of the surfaces.

86

Figure 25: SPRi difference images of S. aureus growth on the gold surface coated

with BSA (left) and casein (right) after (a) 6, (b) 12, (c) 18, and (d) 24

hours. Differences in brightness are clearly distinguishable between

surface coatings. The regions selected for analysis in each of the images

are outlined in panel (a) using yellow dashed lines. (e) Changes in the

mean value of the reflectivity caused by binding of S. aureus to portions

of the surface coated with BSA, casein and non-coated (blue checker =

BSA, solid orange = bare gold, upward sloped grey lines = casein). Error

bars show the standard deviations of three separate experiments. *

indicates a change from the control bare gold at each time (p < 0.05, 2-

tailed t-test with unequal variance).

87

To validate the SPRi results, one of the S. aureus experiments was stopped after

six hours and imaged using an SEM. Figure 26a and d show the borders between the

BSA/bare gold and bare gold/casein, respectively, at a low magnification. There is

significantly more biomass on the bare gold than on either of the coatings. Figure 26b and

c show two sections of the surface at the border between the BSA and bare gold at a higher

magnification. Figure 26e and f show the border between bare gold and casein. S. aureus

cells are attached to the bare gold surface and beginning to form biofilms while the BSA

and casein surfaces are still relatively uncontaminated. There were no visible edge effects

in the coating materials on the surface, and the cell distribution is uniform across each

individual coating. The boundaries between coatings are sharply defined and match the

shape of the initial droplets that were placed on the sensor surface prior to drying.

88

Figure 26: Scanning electron microscope (SEM) images of the coated and uncoated

sensor surfaces after 6 hours of exposure to flowing solutions containing

to S. aureus. (a) Low-magnification and (b), (c) high-magnification images

of the boundary between BSA and bare gold. (d) Low magnification and

(e), (f) high-magnification images of the boundary between bare gold and

casein.

89

5.2.1.2. Penicillin/Streptomycin and BSA

In this set of experiments, we studied the effects of a penicillin/ streptomycin

antibiotic cocktail as a surface coating for inhibition of P. aeruginosa and S. aureus biofilm

formation. The experiments were repeated three times and the data was analyzed the same

way as in the BSA/casein experiments. Figure 27 shows the results of experiments with P.

aeruginosa. The antibiotic cocktail was consistently less effective than BSA at preventing

bacterial adhesion over the course of 24 hours for species. Figure 28a–d shows the SPRi

difference images at six hour intervals for the antibiotic and BSA coatings when they are

exposed to S. aureus. For the right side of the chamber, which was coated with

penicillin/streptomycin, bacterial adhesion is initially suppressed; however, the coating

ceases to be effective after a few hours. After 24 hours, the area with the antibiotic coating

is nearly indistinguishable from the bare gold surface.

Figure 28e shows the normalized contrast changes due to bacterial growth on

different parts of the surface. The quantitative analysis shows that the difference between

the antibiotic coating and the bare gold surface is not statistically different after 24 hours.

It is suspected that the bacteria rapidly degraded the antibiotics as only a small dose was

present on the surface versus dissolving it in the growth medium continuously [227]. The

antibiotics on their own did not appear to degrade, as a control SPRi experiment without

bacteria showed minimal changes in brightness on the coated surface over 24 hours. At the

conclusion of the experiments, the BSA coating had 25% less biomass than the bare gold

region when exposed to P. aeruginosa and 50% less biomass when exposed to S. aureus.

The perceived improvement in BSA effectiveness versus the BSA/ casein experiments is

attributed to a slightly lower initial concentration of bacterial cells. The absolute values for

90

the reflectivity changes on the bare gold surface, for the BSA/casein and BSA/antibiotic

experiments after 24 hours, were within 10% of each other. The larger standard deviation

for the S. aureus results at the 6 hour time point is the result of minor changes in initial

concentration and growth rates between experiments that resulted in changes in the rate at

which cells initially attached to the coatings. These variables cannot be controlled with

very high precision, and the results highlight the additional issues faced when testing

individual coatings in separate experiments.

91

Figure 27: SPRi difference images of P. aeruginosa (CFP-PA01) growth on the gold

surface coated with BSA (left) and penicillin/streptomycin (right) after (a)

6, (b) 12, (c) 18, and (d) 24 hours. The region of the chip coated with BSA

is darker than the antibiotic-coated region, which is nearly

indistinguishable from the bare gold region. (e) Changes in the mean

value of the reflectivity caused by binding of P. aeruginosa to portions of

the surface coated with BSA, antibiotics, and non-coated gold (red brick

= BSA, solid brown = bare gold, vertical pink lines = antibiotics).

92

Figure 28: SPRi difference images of S. aureus growth on the gold surface coated

with BSA (left) and penicillin/streptomycin (right) after (a) 6, (b) 12, (c)

18, and (d) 24 hours. The region of the chip coated with BSA is darker

than the antibiotic-coated region, which is nearly indistinguishable from

the bare gold region. (e) Changes in the mean value of the reflectivity

caused by binding of S. aureus to portions of the surface coated with BSA,

antibiotics, and non-coated gold (blue checker = BSA, solid orange = bare

gold, downward sloped purple lines = antibiotics).

93

5.2.1.3. Penicillin/Streptomycin and Casein

This set of experiments was conducted to directly compare the preventative effect

of penicillin/streptomycin antibiotic cocktail and casein coatings on P. aeruginosa biofilm

formation. The left side of the surface was coated with casein and the right side was coated

with antibiotics while the center of the gold surface was not coated. Bacterial solution was

placed in the chamber and the SPRi experiment was run for 24 hours. Figure 29 show

difference images generated at the surface by the SPRi sensor show much lower attachment

for the section coated with casein, while the region coated with antibiotic showed almost

no effect on preventing the bacterial growth (Figure 29).

94

Figure 29: SPRi difference images of P. aeruginosa growth on the gold surface coated

with casein (left) and penicillin/streptomycin (right) after (a) 6, (b) 12, (c)

18, and (d) 24 hours. Differences in brightness are clearly distinguishable

between surface coatings.

95

5.2.2. Loading Antibiotics in Solution

In this section, the efficacy of penicillin/Streptomycin to prevent S. aureus

infection, Colistin to prevent P. aeruginosa growth and Spectinomycin to prevent B. cereus

growth were investigated. A control experiment was first completed to determine the

normal bacterial growth on the gold surface without antibiotics. Then the growth rate and

biofilm formation kinetics were monitored in the presence of antibiotics in the inlet growth

media. The results of both experiments provided useful information about the kinetics of

biofilm formation under the effect of antibiotics.

5.2.2.1 Control Experiment

In this experiment, a PDMS chamber was first placed on the sensor surface. LB

growth media was run for 3 minutes at a high flow rate in the SPRi system to make sure

all of the tubing system was filled with LB. 200 µL of LB growth media containing S.

aureus was placed in the PDMS chamber on the sensor surface and the prism was placed

in the SPR device. The SPRi experiment was run for 24 hours. Fresh LB was provided for

the bacteria during the entire experiment by flowing LB over the sensor surface at a flow

rate of 10 µL/min.

Figure 30 shows the SPRi difference images at different time intervals. As bacteria

grew and attached to the surface, bright spots started to appear on the surface. The more

bacteria attached to the surface, the larger the change in the refractive index above the

sensor surface; as a result, the chamber area increased in brightness over time.

96

Figure 30: SPRi difference images of S. aureus growth on the sensor surface with

continuous LB flow over the surface after a) 35, b) 330, c) 635, and d) 1170

min.

97

5.2.2.2 Penicillin/Streptomycin (S. aureus)

In this section, 200 µL of the diluted media was placed in the PDMS chamber on

the sensor surface and the prism was placed in the SPR device. To study the preventive

effect of penicillin/streptomycin on the growth of S. aureus, penicillin/streptomycin

solution was added to the inlet LB growth media (1:100). The flow system ran during the

entire 24-hour experimental period in order to provide fresh food for the S. aureus cells

initially present in the chamber. The SPRi device provided difference images of the sensor

surface every 3 seconds. Figure 31 shows the comparison between the results of this

experiment and the control experiment. The results indicate that the

penicillin/streptomycin reduced bacterial growth.

98

Figure 31: SPRi difference images of the surface. The left column shows the images

from S. aureus growth on the chamber without having any antibiotic in

the inlet media as a control, right column shows the difference images at

the same time points by running penicillin/streptomycin from the

beginning of the experiment. Growth in the two chambers is shown at

(a,b) 35, (c,d) 330, (e,f) 635, and (g,h) 1414 min from the start of the

experiment. This image compares the bacterial growth in the two

experiments to study the effect of penicillin/streptomycin on prevention

of bacterial growth.

99

5.2.2.3 Colistin (P. aeruginosa)

In this section, 200 µL of diluted (1:100) P. aeruginosa bacteria in LB, which was

initially inoculated in 6 mL of LB growth media and incubated overnigh for 18 hours, was

placed in the PDMS chamber on the gold surface of the sensor. 50 mg of Colistin was

added to 50 mL of the inlet LB growth media to make a final concentration of 1g/L. The

experiment was run for one day, and the SPRi system monitored the bacterial growth. The

results showed that Colistin prevented bacterial growth completely, as shown in Figure 32.

100

Figure 32: SPRi difference images of P. aeruginosa growth with and without

antibiotics. The left column shows the images of P. aeruginosa growth in

the chamber when running Colistin from the beginning of the experiment.

The right column shows the difference images at the same time points

without having any antibiotic added to the inlet media. Images are shown

after (a,b) 3 hrs , (c,d) 9 hrs, (e,f) 17 hrs.

101

5.2.2.4 Spectinomycin (B. cereus)

In this section, 200 µL of diluted (1:100) B. cereus bacteria in LB, which was

initially inoculated in 6 mL of LB growth media and incubated over night for 18 hours,

was placed in the PDMS hexagon shape chamber on the gold surface of the sensor. 10 mg

of Spectinomycin was added to 50 mL of the inlet LB growth media to make a final

concentration of 100 µg/mL. The experiment was run for 24 hours, and the SPRi system

monitored the bacterial growth. Images at 6 hour intervals are presented infigure 33.

Comparing B. cereus growth with and without antibiotics shows that Spectinomycin

decreased overall bacterial growth and stopped bacterial growth on the surface after 6

hours. Figure 34 shows the reflectivity changes over time as B. cereus bacteria grow on

the surface in the presence and absence of Spenctinomycin antibiotic and provides more

quantitative analysis. Reflectivity change in directly proportional to the biomass

accumulation on the surface and it was calculated to show the biomass accumulation on

the surface each 6 hours. Figure 35 shows the amount of biomass coverage on the surface

every 6 hours. As presented in this graph in the presence of antibiotics stopped bacterial

further growth after around 6 hours and the total mass coverage in the absence of antibiotics

(~3000 pg/mm2) was almost five times more that when antibiotics was present (600

pg/mm2).

102

Figure 33: SPRi difference images of B. cereus growth with and without antibiotics.

The left column shows the images of B. cereus growth in the chamber

when running Spectinomysin solution from the beginning of the

experiment, right column shows the difference images at the same time

points without having any antibiotic added to the inlet media. Images

shown at (a,b) 1 hr, (c,d) 6 hrs, (e,f) 12 hrs, (g,h) 18 hrs, and (i,j) 24 hrs.

103

Figure 34: Quantitative analysis showing the reflectivity changes over time as B.

cereus bacteria grow on the surface in the presence (orange line) and

absence (blue line) of Spenctinomycin antibiotic.

Figure 35: Biomass coverage on the sensor surface at different time point. The

orange columns represent B. cereus growth on the surface in the presence

on antibiotics in the solution, the blue columns represent B. cereus growth

in the normal solution sans antibiotics.

104

5.3. Biofilm Removal from the Surface (Goal 3)

Currently, there is no established method for permanently and completely

preventing biofilm formation. In this section, biofilm removal from a contaminated surface

was studied with SPRi. After early stage detection of biofilm formation, the efficacy of

different chemicals to remove biofilm and the rate of removal from the surface were

studied in real-time with SPRi. It was the first time SPRi has been used for biofilm removal

studies and would provide kinetics and real-time information for cleaning procedures. In

this aim, the effects of cleaning with chemicals, and disinfection using antimicrobial agents

was studied with the SPRi device.

5.3.1. Cleaning with Different Chemical Compounds

In this part of the research the effect of SDS a well-known surfactant used in

detergents was studied on biofilm removal of different bacterial species. The chosen

bacteria were S. aureus, P. aeruginosa, and B. cereus. The bacteria was cultured in 6 mL

fresh LB for 18 hours at 37 °C. Then this media was diluted in fresh LB (1:100 v:v). All

experiments were run using the diluted cultures. SPRi experiments began by running fresh

LB first to let the signal stabilize, then bacteria media was flowed over the surface for 24

hours, this allows bacteria attach on the surface, growth, and form biofilm. After this period

of time, 1%SDS was flowed over the surface for 3 hours to remove the biofilm, as biofilm

removes from the surface the reflectivity returns to the original state, because the biomass

in replaced by liquid. In these experiments it is important to remember, the reflectivity

change is partly due to biofilm removal and partly due to the variation in the refractive

index of the running solution, from LB to SDS. To eliminate the effect of refractive index

of the solution and to compare the exact amount of biomass removal, after SDS was run

105

for few hours, the experiment was ended by running fresh LB as a reference solution. The

signal level of the LB before and after SDS run represented the biomass removal, and as a

result the efficacy of SDS.

Figure 36 shows S. aureus growth for 24 hours on the surface, bacteria covered

4000pg/mm2 on the surface after 24 hours. Then 1%SDS was run over the surface for 3

more hours, the sharp decrease in biomass coverage is partly due to the solution variation

from bacterial media (LB) to SDS. After SDS run, fresh LB was run again on the surface,

this allowed the accurate comparison of biomass removal. The results showed ~80%

biomass removal for S. aureus bacteria.

Figure 36: S. aureus growth and removal on the sensor surface during the

experiment period. Arrows indicate the time when the mentioned solution

was loaded through the system. The dashed lines compare the level of

mass coverage value at each step of the experiment. The double-sided red

arrow represents the amount of biomass removal after SDS run.

106

In the next experiment the effect of SDS was studied in biofilm removal of P.

aeruginosa. The experiment was run following the exact same procedure as for S. aureus.

The results for this bacteria showed almost 100% biofilm removal (Figure 37).

Figure 37: P. aeruginosa growth and removal on the sensor surface during the

experiment period. Arrows indicate the time when the mentioned solution

was loaded through the system. The dashed lines compare the level of

mass coverage value at each step of the experiment. The double-sided red

arrow represents the amount of biomass removal after SDS run.

107

Finally 1% SDS was used to remove B. cereus biofilm, bacteria was growth of the

surface following the same protocol for 24 hours, then 1% SDS was run over the surface

for 3 hours to remove the biomass. The level of LB before and after SDS run was compared

and showed 87.5% biofilm removal for B. cereus bacteria. (Figure 38)

Figure 38: B. cereus growth and removal on the sensor surface during the

experiment period. Arrows pointing at the time when the mentioned

solution was loaded through the system. The dashed lines compare the

level of mass coverage value at each step of the experiment. The double

head red arrow represents the amount of biomass removal after SDS run.

108

5.3.2. Disinfection with Antimicrobial Components

In this section, the effect of antibiotics on biofilm was investigated. Antibiotic

treatment is one of the main options to treat infections related to medical devices and

implants. It is known that bacteria inside biofilm are up to 1000 times more resistant to

different antimicrobial components. Biofilm matrix acts as a barrier for any antibacterial

molecule to diffuse through and kill the organism inside it. This is the first time a SPRi

instrument was used to real-time study the effect of various antibiotics on killing bacteria

inside the biofilm and disrupting biofilm.

First, the selected bacteria were grown on the sensor surface for 24 hours following

the procedure mentioned in the last section. After 24 hours of biofilm formation antibiotics

specific to the chosen bacteria at the minimum inhibitory concentration (MIC) were flowed

over the surface for another 24 hours. This would allow enough time for the antibiotic

molecules to diffuse through the biofilm and potentially kill bacteria inside the biofilm. If

the antibiotic at the MIC concentration is effective, the biofilm will start to detach from the

surface as a result of degradation and any changes in the attachment of biofilm on the

surface will be observed with SPRi.

In the first experiment, S. aureus bacteria was grown on the surface for 24 hours

then, penicillin/Streptomycin solution in LB (0.06 µg/mL) was flowed over the surface for

another 24 hours. The results are presented in figure 38 and show that mass coverage did

not change after treating S. aureus biofilm with penicillin/streptomycin (0.06 µg/ml)

solution for 2 hours. Penicillin/streptomycin at MIC was not sufficient for treatment of the

biofilm; the concentration of antibiotics or the treatment time was not enough for

antibacterial molecules to diffuse through the biofilm matrix, kill the bacteria, and remove

the biomass attachment from the surface.

109

Figure 39: The effect of penicillin/streptomycin on treatment of S. aureus biofilm.

The black arrow points at the time point when the bacterial media was

flowed over the sensor surface (t=0). The red arrow points at the time

point when antibiotic solution was flowed over the surface to treat the

biofilm (t=24h).

In the next experiment, the effect of spectinomycin antibiotic on B. cereus biofilm

was investigated. The concentration of antibiotic was 200 µg/mL. Initially, B. cereus

bacterial was cultured in 6 mL of fresh LB at 37⁰C for 18 hours. Then the bacterial media

was diluted 100 times in fresh LB (v:v). This diluted culture was flowed over the gold

surface of the prism for 24 hours and B. cereus growth was monitored in real time.

After 24 hours, the bacterial media was switched with the solution of 200 µg/mL

specitinomycin in fresh LB. The antibiotic solution was then run over the B. cereus biofilm

for another 24 hours at the same flow rate of 10 µL/min. The graph in figure 40 represents

the B. cereus growth profile over the period of 24 hours. The red arrow indicates the time

at which the antibiotic solution was loaded through the system. The results show almost an

800 pg/mm2 reduction in the attached biomass density on the surface ~21 hours after the

110

antibiotic started being introduced. At 45 hours, which means 21 hours after starting the

antibiotic, bacteria start to regrow on the surface which resulted in an increase in the

biomass accumulation. This is believed to show bacterial natural behavior of developing

resistance toward antibiotics. B. cereus biofilm was treated with the consistent MIC of

spectinomycin, the results indicated this antibiotic is effective at reducing bacterial

attachment for around 21 hours only, and after that bacteria will form a resistance and will

grow again, to avoid that the concentration of antibiotic need to be increased or another

type of effective antibiotics need to be used.

Figure 40: The effect of spectinomycin on treatment of B. cereus biofilm. The black

arrow points at the time point when the bacterial media was flowed over

the sensor surface (t=0). The red arrow points at the time point when

antibiotic solution was flown over the surface to treat the biofilm (t=25h).

111

5.4. Effects of flowrate on Bacterial Adhesion (Goal 4)

Changes in the flowrate directly in a channel affect the flow velocity at the channel

surface. Shear stress is proportional to the change in velocity (u) with height above the

surface (z) multiplied by the fluid's dynamic viscosity (v), which is expressed as:

𝜏 = 𝜈𝜕𝑢

𝜕𝑧

In our system, the channel height and the fluid’s dynamic viscosity was constant, so the

shear stress on the surface is only related to the fluid’s velocity and flowrate.

Changes in flowrate affect bacterial attachment and growth on the surface. The

SPRi system was used to study the effect of flowrate on bacterial growth and biofilm

formation on the surface in real time. COMSOL Multiphysics was used to simulate the

fluidic flow system and investigate the shear stress distribution on the surface in our

system.

5.4.1. SPRi Experiments

B. cereus growth was first monitored on the entire surface over a period of 24 hours

in stagnant fluid, which represented the 0 flowrate. The average reflectivity change over

the entire surface was calculated and results are presented at 6 hour intervals.

Next, B. cereus growth was monitored under 10 µL/min and 40 µL/min flowrates,

respectively, with the same procedure mentioned above. The average reflectivity changes

were obtained following the same procedure and the results are presented for each flowrate

at 6 hours time intervals. Figure 41 shows the changes in bacterial growth on the surface

as flowrate increases at each time point.

112

Figure 41: The average reflectivity change in the difference image at each time point

as a result of B. cereus growth under different flowrates. Blue=stagnant

condition, Orange=10µL/min, and gray=40µL/min flowrates.

0

5

10

15

20

25

30

1 6 12 18 24

Ref

lect

ivit

y (

A.U

.)

Time (hour)

Stagnant condition 10 ul/min 40 ul/min

113

Next, the effect of fluidic flowrate on S. aureus growth was investigated. Bacteria

were cultured in fresh LB for 18 hours at 37 °C prior to the experiment. This media was

then diluted in fresh LB (1:100 v:v) to provide bacteria enough food to perform normal

activities. This media was flowed over the sensor surface at different flowrates. SPRi

provided images of the surface every 3 seconds. S. aureus bacterial growth was studied at

the slow flowrate of 10 µL/min for 24 hours. The results presented in figure 42-left column

shows S. aureus growth over the period of 24 hours, with images at every 6 hours presented

here. S. aureus at a 10 µL/min flowrate formed a uniform biofilm on the entire surface.

Bacterial growth increased gradually over time, and after 24 hours, biofilm covered the

entire surface.

Next, the S. aureus experiment was repeated using a higher flowrate of 120 µL/min.

The results in figure 42-right column show non-uniform bacterial growth over this period

of time. The results show that bacteria tend to grow more on the sides on the channel rather

than the middle. Also higher bacterial growth was detected toward the bottom part of the

channel where the fluid exits through the outlet tube. The reason for higher biofilm

formation toward the bottom of the channel is because of the flow direction, which is from

top toward the bottom of the image. The flow direction pushes bacteria toward the bottom

of the channel and then media leaves upward from that site, this gives bacteria more

residence time on the bottom part of the channel rather than top part. In order to understand

the physical properties of the solution over the surface, the fluidic system was simulated

with COMSOL Multiphysics modeling software, which is discussed later in this section.

114

Figure 42: SPRi difference images showing S. aureus growth under 10µl/min (left

column) and 120µl/min (right column) flowrates after (a,b) 6 hours, (c,d)

12 hours, (e,f) 18 hours, and (g,h) 24 hours.

115

5.4.2. COMSOL Multiphysics Modeling

COMSOL Multiphysics modeling was carried out to simulate the shear stress

distribution on the entire sensor surface. Shear stress is directly related to flowrate of the

fluid passing over the surface and it affects bacterial attachment and further biofilm

formation.

In this model, the non-slip wall boundary condition was assumed for the hexagon-

shaped PDMS chamber. The Reynold’s number for this simulation indicates a laminar

flow. The results of this simulation are shown in figure 43, where the color code represents

the shear stress distribution over the surface. Red represents the highest and blue the lowest

shear stress on the surface. The results clearly show higher shear stress at the boundaries,

where there was no slip condition applied. The higher shear stress at the boundaries means

lower flow velocity. The lower flow velocity results in a higher residence time for bacteria

in those regions. Residence time determines the time bacteria can sit on a spot and

potentially form attachments. The higher residence time consequently increases the chance

of bacterial attachment on those regions. When bacteria form irreversible attachments they

will then start to produce exopolysaccharide matrix around themselves and form biofilm.

That explains the higher biofilm formation on the sides of the PDMS hexagon chamber.

116

Figure 43: COMSOL Multiphysics modeling of the shear stress distribution on the

sensor surface. The red color indicates the highest and the blue color

represents the lowest shear stress on the surface. The Color shows highest

shear stress at the boundaries where the chamber walls are, and the lowest

shear stress was detected toward the middle of the channel.

117

6.0 Conclusions and Future work

6.1. Goal 1: System Setup for Monitoring Bacterial Growth and Biofilm

Formation

These experiments show for the first time that SPRi is a viable technique for real-

time, label-free imaging of biofilm formation and removal on a surface. The technique

provides spatial information about where cells are adhering within a chamber or channel

that is not available with standard SPR. We used SPRi to image biofilms produced by two

common bacterial species, E. coli and P. aeruginosa. This straightforward methodology

allows researchers to begin using SPRi for high resolution large-area studies of bacteria on

surfaces. This type of population level analysis of bacterial response may provide new

insights for medicine, biotechnology, and ecology. Further, the gold sensing surface used

in these experiments lends itself to chemical functionalization, which makes this an ideal

approach for adhesion experiments. Finally, this approach complements other methods,

such as confocal microscopy for studying biofilms, and is the first to offer real-time, high-

resolution analysis of biofilm removal.

118

6.2. Goal 2: Prevention of Biofilm Formation on the Surface

This set of experiments demonstrates as a proof-of-concept that SPRi can be used

to investigate the effects of different surface coatings on the inhibition of biofilm formation

by imaging multiple coatings in a single experiment in real-time. This side–by-side

comparison minimized variability between runs and increased throughput. This is

important because slight variations in the starting concentration of bacteria and

experimental conditions can significantly change the results, making comparisons between

experiments difficult. SPRi difference images indicate that bacteria adhere the least to

casein-coated gold surfaces, suggesting that casein is a better candidate than BSA as a

surface coating for the prevention of P. aeruginosa and S. aureus biofilm formation. The

results also show that penicillin/streptomycin solution does not have much long-term effect

in the prevention of biofilm formation on a surface, as the surface was nearly

indistinguishable from bare gold after 24 hours. The results of one experiment were further

investigated using SEM to show that the biomass surface coverage is indeed proportional

to the intensity change observed with SPRi. While only three surfaces were tested per run

in these experiments, the multiplexing capabilities for coatings can be extended further by

using a template to selectively pattern smaller regions on the 1 cm2 sensor area. However,

care must be exercised when adjusting the size of the coated region so that there are

sufficient interactions with the bacterial cells during the course of the experiment to test

the antifouling properties.

Further, interactions between coatings can be minimized, if needed, by using a

fluidic setup consisting of multiple smaller flow channels that are fed by a single inlet.

Given that the evanescent field of the surface plasmons extends a few hundred nanometers

119

from the surface, it is possible to coat the gold with multiple layers of coatings, creating,

for example, functionalized polymer and oxide coatings for analysis. As the next step, we

plan to develop standardized testing protocols that will evaluate the effectiveness of

antimicrobial surface coatings exposed to varying fluid shear stress and other common

pathogens.

Further studies can be done by integrating other microfluidic geometries with the

SPRi system. The preliminary studies on this area were presented at the Micro Total

Analysis Systems Conference, 2014. In this system, microfluidic channels (Figure 44) can

be placed on top of the gold surface and bacteria can be exposed to multiple concentrations

and antibiotics in a single run.

Figure 44: (A) Schematic of the SPRi setup for antibiotic resistance experiments. (B)

The average brightness change in the channels filled with LB without cells

(diamonds), or 5.4E+6 cells/mL S. aureus with no antibiotic (purple), with

1000X diluted antibiotic (blue), with 200X diluted antibiotic (sloping

lines). The experiment was run four times. The error bars show the

standard deviation of 3 data sets.

120

6.3. Goal 3: Biofilm Removal from the Surface

As completely preventing bacterial growth and biofilm formation is not possible

for long periods of time, we did several studies to evaluate the effectiveness of various

methods on removing bacterial biofilm from the substrates.

In the first, part we looked at the effect of SDS which is a well-known chemical

compound used in detergents in removing biofilm from the surface. The effectiveness of

1% SDS was studied on biofilm removal of P. aeruginosa, S. aureus, and B. cereus. The

results showed that a few hours of running 1% SDS at the slow flowrate of 10 µL/min

effectively removes 80-100% of the biomass from the surface. SPRi provided real-time

analysis of the biofilm formation and removal over the entire 1cm2 surface for the entire

experimental period of 24 hours. The SPRi difference images allowed for visual study of

biomass accumulation and removal on the surface. The reflectivity graph also provided

thorough quantitative analysis of the exact biomass coverage on the surface.

In the next part, antibiotic treatment was evaluated using SPRi. Antibiotic treatment

is still the main treatment for infections related to medical devices and implants. These

processes usually consist of treatment of the infected part with high doses of antibiotics.

We investigated the capability of SPRi to measure the efficacy of biofilm treatment with

antibiotics. For this purpose the effect of Colistin, Penicillin/streptomycin, and

Spectinomycin was investigated in treatment of biofilms of P. aeruginosa, S. aureus, and

B. cereus. The results, as expected, showed no effect when biofilms were treated for 24

hours at MIC. The MIC for treatment of biofilms was not sufficient, while being effective

for killing planktonic bacteria. One explanation of this observation is that the biofilm

matrix acts as a shield and impedes antibiotic diffusion.

121

This methodology provided a sensitive approach for studying, in real time, the

effectiveness of different kinds of antibacterial and chemical components in treating

biofilms.

122

6.4. Goal 4: Effects of flow rate on Bacterial growth

In the final part of the study, bacterial growth was investigated under different flow

rates. Flow rate is directly related to shear stress on the surface. Shear stress interfered with

initial bacterial attachment on the surface and further biofilm formation. The results

indicated that higher flowrate decrease bacterial growth over the same period of time on

the surface.

To analyze the relationship between shear stress and flowrate on the entire surface,

COMSOL Multiphysics modeling was used to simulate the setup. The results showed

higher shear stress at the boundaries, which caused lower flow rates in those locations. The

effect of flow rate, and as a result shear stress, was more pronounced at higher flow rates

and the bacterial growth profile for the bacterial species tested. At slow flow rates uniform

biofilm formation on the entire surface was detected, however uneven growth was

observed at faster flows, indicating a bias for bacterial attachment.

In future, the same setup and simulation can be used to study the effect of different

channel geometries and binding angles in bacterial growth and streamer formation.

Streamers are a special morphology of biofilms which are formed under certain flow

conditions. Streamers bridge between corners in non-uniform environments, such as filters,

porous materials, and medical devices. Bacterial growth and streamer formation has been

studied under different laminar flow condition [228-230] using fluorescence and other

microscopy techniques.

It is possible to study bacterial growth in channel geometries and bending angles

and investigate the kinetics of steamer formation using SPRi. The suggested channel design

123

and the related geometries are shown in figure 45, where Figure 45-a shows a channel

design with 90 degree angle bends and figure 45-b shows 150 degree bends to study

streamer formation.

Figure 45: Channel design to study biofilm formation under different flowrate and

in non-uniform structures. Bends are in a) 90 degree angle and b) 150

degree angles.

124

7.0. METHODS

7.1. PDMS Fabrication

Polydimethylsiloxane (PDMS) is a commonly used polymer in biofilm research

[231]. PDMS channels and chambers with different shapes were made. The channels and

chambers were placed on a gold-coated prism surface and bacterial cells were loaded in

them.

The two major components of PDMS are silicone elastomer base, which is a

monomer, and a silicone elastomer curing agent. The mixing ratio of these two components

plays a crucial role in the final properties of the PDMS polymer, such as stiffness, which

increases by decreasing the base to curing agent ratio [232]. The ratio which was used in

this study was 10:1, which is the most common ratio used for microfluidics experiments.

To fabricate the channels, first the two components, at the mentioned ratio, was mixed

together thoroughly and the solution was poured on a mold containing the designs for a

desired experiment. Then the mold was placed in oven at 70 °C for approximately two

hours to cure the polymer. The cured polymer was then be peeled off from the mold and it

was cut into pieces that have the same size as the prism. Before each experiment, the

channel was placed on the prism surface. In these experiments, two different channels was

used. The fabrication process for each is represented in Figure 46 and 47.

125

Figure 46: Schematic of the fabrication of a PDMS chamber. a) The hexagonal mold

is made of aluminum. b) PDMS was poured on the mold and cured in an

oven. c) The cured PDMS was peeled off from the mold. d) each mold

contains 6 hexagons, which were cut into separate pieces. e) One hexagon

chamber f) was placed on the gold coated prism at a time.

126

Figure 47: Schematic of the fabrication of PDMS channel. a) The silicon mold has

three and two linear raised feature groups on it. b) PDMS was poured on

the mold and cured in the oven. c) The cured PDMS was peeled off from

the mold. d) Each mold contains several groups of channels, which were

cut and separated. e) A PDMS piece containing two separate linear

channels f) was placed on the gold coated prism.

127

7.2. Bacterial Culture Preparation

Different bacteria species was used in this research, which have been provided

kindly by our collaborators. The two growth media used for culturing the bacteria were

Lysogeny Broth (LB) growth media and trypticase soy broth (TSB) growth media. To

prepare 500 mL of LB growth agar media, first 12.5g of LB broth and 7.5g of Agar

(solidifying agent) were mixed with distilled water completely. To prepare 500 mL of

trypticase agar media, 7.5 g of solidifying agent and 15 g of trypticase soy broth were

mixed thoroughly with distilled water. The solutions were then autoclaved at high pressure

and temperature. The agar solution solidifies as it cools down, so when the temperature

was around 50 °C the agar solution was poured in petri dishes and was left to solidify. The

petri dishes containing the agar media were stored in the fridge at -20 °C.

7.3. Bacterial Culturing

Bacteria was cultured on agar plates by streak culturing. In this method, a sterile

inoculation loop was used. The loop was dipped in the bacterial stock solution and then the

sample on the loop was spread on the agar plate. The plate was then incubated at 37 °C for

24 hours. Figure 48 shows the agar plate after 24 hours of incubation when the bacteria

grew and formed colonies. After the bacteria grew on the plates, a colony of the bacteria

was selected with the inoculation loop and cultured in 6 mL of liquid growth media. The

liquid culture was incubated at 37 °C for 18 hours to make sure the bacteria reach

exponential growth phase. These bacterial cultures were diluted at a 1:100 ratio in fresh

LB growth media before being used in the experiments.

128

Figure 48: Staphylococcus aureus bacteria cultured on a LB agar plate.

129

7.4 Sample preparation for Scanning Electron Microscopy (SEM)

1. Fixation of the specimen:

Initially, bacteria was grown on the surface under different conditions. For primary

fixation, the biological samples on the surface were soaked into 2.5% glutaraldehyde, 2.5%

formaldehyde in 0.1 M Na cacodylate buffer solution at pH 7.2 for 2 hours at 4⁰C.

Formaldehyde penetrates into the sample quickly and glutaraldehyde reacts rapidly. Cold

temperature is important to avoid lysis of cells by autolytic enzymes.

2. Washing step:

The sample was washed three times with 0.1M Na cacodylate buffer at pH 7.2 for 45

minutes at 4 °C at 15 minutes intervals.

3. Post fixation of the sample:

The sample was treated with 1.0% osmium tetroxide in 0.1 M Na canodylate buffer at pH

7.2 at 4 °C for 2 hours. After this step, temperature changes and osmolarity are not

important.

4. Washing step:

The sample was again washed three times with 0.1 M Na cacodylate buffer at pH 7.2 for

45 minutes at 15 min intervals.

5. Dehydration step:

Samples were placed into ethanol solutions with various concentrations at room

temperature to replace water molecules with ethanol in the specimen.

130

30% ethanol solution for 10-15 minutes

50% ethanol solution for 10-15 minutes

70% ethanol solution for 10-15 minutes

85% ethanol solution for 10-15 minutes

95% ethanol solution for 10-15 minutes

100% ethanol solution for 1 hour (the sample was washed with fresh 100% ethanol

three times)

6. Drying step:

The sample was placed in vacuum chamber for 2-3 hours

7. Sputter coating step:

In order to make the surface conductive for SEM procedure, the sample was sputter coated

with 5nm of Cr.

131

8.0 REFERENCES

1. Bresee, J.S., Dietz, V., Griffin, P.M., McCaig, L.F., Mead, P.S., Shapiro, C.,

Slutsker, L. and Tauxe, R.V., Food-Related Illness and Death in the United States,

Journal of Environmental Health, 62, pp. 9, 2000.

2. Potter, M., Gonzalez-Ayala, S.and, Silarug, N., Food Microbiology: Fundamentals

and Frontiers, Epidemiology of foodborne diseases, 1997.

3. Beran;, G., Shoeman;, H. and Anderson, K., Food safety: an overview of problems,

Dairy Food Environ Sanit, 11, pp. 189–194, 1991.

4. USDA/ERS (Economic Research Service), Economics of foodborne disease,

http://www.ers.usda.gov/briefing/FoodborneDisease, 2003.

5. Yang, L. and Bashir, R., Electrical/electrochemical impedance for rapid detection

of foodborne pathogenic bacteria, Biotechnology Advances, 26, pp. 135-150,

2008.

6. Subramanian, A., Irudayaraj, J. and Ryan, T., A mixed self-assembled monolayer-

based surface plasmon immunosensor for detection of E. coli O157:H7, Biosensors

and Bioelectronics, 21, pp. 998-1006, 2006.

7. Taylor, A.D., Yu, Q., Chen, S., Homola, J. and Jiang, S., Comparison of E. coli

O157:H7 preparation methods used for detection with surface plasmon resonance

sensor, Sensors and Actuators B: Chemical, 107, pp. 202-208, 2005.

8. June, G.A., Relative effectiveness of selenite cystine broth , tetrathionate broth ,

and rappaport - vassiliadis medium for recovery of Salmonella spp. from raw flesh

, highly contaminated foods , and poultry feed : collaborative study, Journal of

AOAC International, 79, pp. 1307, 1996.

9. Zhou, X. and Jiao, X., Polymerase chain reaction detection of Listeria

monocytogenes using oligonucleotide primers targeting actA gene, Food Control,

16, pp. 125-130, 2005.

10. Hoshino, T., Kawaguchi, M., Shimizu, N., Hoshino, N., Ooshima, T. and Fujiwara,

T., PCR detection and identification of oral streptococci in saliva samples using

GTF genes, Diagnostic Microbiology and Infectious Disease, 48, pp. 195-199,

2004.

11. Rodrı, amp, x, guez-Lázaro, D., Hernández, M., Esteve, T., Hoorfar, J. and Pla, M.,

A rapid and direct real time PCR-based method for identification of Salmonella

spp, Journal of Microbiological Methods, 54, pp. 381-390, 2003.

132

12. Kong, R.Y.C., Lee, S.K.Y., Law, T.W.F., Law, S.H.W. and Wu, R.S.S., Rapid

detection of six types of bacterial pathogens in marine waters by multiplex PCR,

Water Research, 36, pp. 2802-2812, 2002.

13. Meng, J., Zhao, S., Doyle, M.P., Mitchell, S.E. and Kresovich, S., Polymerase

chain reaction for detecting Escherichia coli O157 : H7, International Journal of

Food Microbiology, 32, pp. 103, 1996.

14. Sperveslage, J., Stackebrandt, E., Lembke, F.W. and Koch, C., Detection of

bacterial contamination , including Bacillus spores , in dry growth media and in

milk by identification of their 16S rDNA by polymerase chain reaction, Journal of

Microbiological Methods, 26, pp. 219, 1996.

15. Mozola, M.A., Genetics-based methods for detection of Salmonella spp. in foods,

Journal of AOAC International, 89, pp. 517, 2006.

16. D'Souza, D.H. and Jaykus, L.A., Nucleic acid sequence based amplification for the

rapid and sensitive detection of Salmonella enterica from foods, Journal of Applied

Microbiology, 95, pp. 1343-1350, 2003.

17. Andresen, L.O., Klausen, J., Barfod, K. and Sørensen, V., Detection of antibodies

to Actinobacillus pleuropneumoniae serotype 12 in pig serum using a blocking

enzyme-linked immunosorbent assay, Veterinary Microbiology, 89, pp. 61-67,

2002.

18. Prior, J.L. and Titball, R.W., Monoclonal antibodies against Yersinia pestis

lipopolysaccharide detect bacteria cultured at 28°C or 37°C, Molecular and

Cellular Probes, 16, pp. 251-256, 2002.

19. Abdel-Hamid, I., Ivnitski, D., Atanasov, P. and Wilkins, E., Flow-through

immunofiltration assay system for rapid detection of E. coli O157:H7, Biosensors

and Bioelectronics, 14, pp. 309-316, 1999.

20. Hasebe, Y., Yokobori, K., Fukasawa, K., Kogure, T. and Uchiyama, S., Highly

sensitive electrochemical determination of Escherichia coli density using

tyrosinase - based chemically amplified biosensor, Analytica Chimica Acta, 357,

pp. 51, 1997.

21. Brewster, J.D., Filtration capture and immunoelectrochemical detection for rapid

assay of Escherichia coli O157 : H7, Journal of Immunological Methods, 211, pp.

1, 1998.

22. Rishpon, J. and Ivnitski, D., An amperometric enzyme - channeling immunosensor,

Biosensors and Bioelectronics, 12, pp. 195, 1997.

133

23. Su, X.-L. and Li, Y., A self-assembled monolayer-based piezoelectric

immunosensor for rapid detection of Escherichia coli O157:H7, Biosensors and

Bioelectronics, 19, pp. 563-574, 2004.

24. Mo, X.-T., Zhou, Y.-P., Lei, H. and Deng, L., Microbalance-DNA probe method

for the detection of specific bacteria in water, Enzyme and Microbial Technology,

30, pp. 583-589, 2002.

25. Su, X., Low, S., Kwang, J., Chew, V.H.T. and Li, S.F.Y., Piezoelectric quartz

crystal based veterinary diagnosis for Salmonella enteritidis infection in chicken

and egg, Sensors and Actuators B: Chemical, 75, pp. 29-35, 2001.

26. Plomer, M., Guilbault, G.G. and Hock, B., Development of a piezoelectric

immunosensor for the detection of enterobacteria, Enzyme and Microbial

Technology, 14, pp. 230, 1992.

27. Gehring, A.G., Use of a light - addressable potentiometric sensor for the detection

of Escherichia coli O157 : H7, Analytical biochemistry, 258, pp. 293, 1998.

28. Libby, J. and ., H.W., Detection of Neisseria meningitidis and Yersinia pestis with

a novel silicon-based sensor, Journal of Clinical Microbiology, pp. 1989.

29. Foley-Thomas, E.M., Phage infection , transfection and transformation of

Mycobacterium avium complex and Mycobacterium paratuberculosis,

Microbiology, 141, pp. 1173, 1995.

30. Blasco, R., Murphy, M.J., Sanders, M.F. and Squirrell, D.J., Specific assays for

bacteria using phage mediated release of adenylate kinase, Journal of Applied

Microbiology, 84, pp. 661-666, 1998.

31. Pyle, B., Broadaway, S. and McFeters, G., A rapid, direct method for enumerating

respiring enterohemorrhagic E. coli O157:H7 in water., Applied and

Environmental Microbiology, pp. 1995.

32. Chowdhury, M.A.R., Xu, B., Montilla, R., Hasan, J.A.K., Huq, A. and Colwell,

R.R., A simplified immunofluorescence technique for detection of viable cells of

Vibrio cholerae O1 and O139, Journal of Microbiological Methods, 24, pp. 165,

1995.

33. Hawkes, J.J., Long, M.J., Coakley, W.T. and McDonnell, M.B., Ultrasonic

deposition of cells on a surface, Biosensors and Bioelectronics, 19, pp. 1021-1028,

2004.

34. Ivnitski, D., Abdel-Hamid, I., Atanasov, P. and Wilkins, E., Biosensors for

detection of pathogenic bacteria, Biosensors and Bioelectronics, 14, pp. 599-624,

1999.

134

35. Barlen, B., Mazumdar, S., Lezrich, O., Kämpfer, P. and Keusgen, M., Detection of

Salmonella by Surface Plasmon Resonance, Sensors, 7, pp. 1427-1446, 2007.

36. Paul, S., Vadgama, P. and Ray, A.K., Surface plasmon resonance imaging for

biosensing, IET Nanobiotechnology, 3, pp. 71-80, 2009.

37. Malic, L., Veres, T. and Tabrizian, M., Nanostructured digital microfluidics for

enhanced surface plasmon resonance imaging, Biosensors and Bioelectronics, 26,

pp. 2053-2059, 2011.

38. Yanase, Y., Hiragun, T., Kaneko, S., Gould, H.J., Greaves, M.W. and Hide, M.,

Detection of refractive index changes in individual living cells by means of surface

plasmon resonance imaging, Biosensors and Bioelectronics, 26, pp. 674-681, 2010.

39. Kodoyianni, V., Label-free analysis of biomolecular interactions using SPR

imaging, BioTechniques, 50, pp. 32-40, 2011.

40. Knoll, W., Interfaces and thin films as seen by bound electromagnetic waves,

Annual Review of Physical Chemistry, 49, pp. 569-638, 1998.

41. Abdulhalim, I., Zourob, M. and Lakhtakia, A., Surface Plasmon Resonance for

Biosensing: A Mini-Review, Electromagnetics, 28, pp. 214-242, 2008.

42. Folch, A., Introduction to BioMEMS, pp. 2013.

43. Fischer, B., Heyn, S.P., Egger, M. and Gaub, H.E., Antigen binding to a pattern of

lipid-anchored antibody binding sites measured by surface plasmon microscopy,

Langmuir, 9, pp. 136-140, 1993.

44. Berger, C.E.H., Beumer, T.A.M., Kooyman, R.P.H. and Greve, J., Surface plasmon

resonance multisensing, Analytical chemistry, 70, pp. 703-706, 1998.

45. Ladd, J., Taylor, A., Piliarik, M., Homola, J. and Jiang, S., Label-free detection of

cancer biomarker candidates using surface plasmon resonance imaging,

Analytical and bioanalytical chemistry, 393, pp. 1157-1163, 2009.

46. Liu, W., Chen, Y. and Yan, M., Surface plasmon resonance imaging of limited

glycoprotein samples, The Analyst, 133, pp. 1268-1273, 2008.

47. Rebe Raz, S., Bremer, M.G.E.G., Giesbers, M. and Norde, W., Development of a

biosensor microarray towards food screening, using imaging surface plasmon

resonance, Biosensors and Bioelectronics, 24, pp. 552-557, 2008.

48. Wong, C.L., Ho, H.P., Suen, Y.K., Kong, S.K., Chen, Q.L., Yuan, W. and Wu,

S.Y., Real-time protein biosensor arrays based on surface plasmon resonance

differential phase imaging, Biosensors and Bioelectronics, 24, pp. 606-612, 2008.

135

49. Yuk, J.S., Kim, H.-S., Jung, J.-W., Jung, S.-H., Lee, S.-J., Kim, W.J., Han, J.-A.,

Kim, Y.-M. and Ha, K.-S., Analysis of protein interactions on protein arrays by a

novel spectral surface plasmon resonance imaging, Biosensors and Bioelectronics,

21, pp. 1521-1528, 2006.

50. Jordan, C.E., Frutos, A.G., Thiel, A.J. and Corn, R.M., Surface Plasmon Resonance

Imaging Measurements of DNA Hybridization Adsorption and Streptavidin/DNA

Multilayer Formation at Chemically Modified Gold Surfaces, Analytical

chemistry, 69, pp. 4939-4947, 1997.

51. Frutos, A.G. and Corn, R.M., Peer Reviewed: SPR of Ultrathin Organic Films,

Analytical chemistry, 70, pp. 449A-455A, 1998.

52. Hayashi, G., Hagihara, M. and Nakatani, K., Genotyping by allele-specific l-DNA-

tagged PCR, Journal of Biotechnology, 135, pp. 157-160, 2008.

53. Mannelli, I., Courtois, V., Lecaruyer, P., Roger, G., Millot, M.C., Goossens, M.

and Canva, M., Surface plasmon resonance imaging (SPRI) system and real-time

monitoring of DNA biochip for human genetic mutation diagnosis of DNA

amplified samples, Sensors and Actuators B-Chemical, 119, pp. 583-591, 2006.

54. Nelson, B.P., Grimsrud, T.E., Liles, M.R., Goodman, R.M. and Corn, R.M.,

Surface Plasmon Resonance Imaging Measurements of DNA and RNA

Hybridization Adsorption onto DNA Microarrays, Analytical chemistry, 73, pp. 1-

7, 2000.

55. Nelson, B.P., Frutos, A.G., Brockman, J.M. and Corn, R.M., Near-infrared surface

plasmon resonance measurements of ultrathin films. 1. Angle shift and SPR

imaging experiments, Analytical chemistry, 71, pp. 3928-3934, 1999.

56. Brockman, J.M., Frutos, A.G. and Corn, R.M., A multistep chemical modification

procedure to create DNA arrays on gold surfaces for the study of protein-DNA

interactions with surface plasmon resonance imaging, Journal of the American

Chemical Society, 121, pp. 8044-8051, 1999.

57. Jeong, E.J., Jeong, Y.S., Park, K., Yi, S.Y., Ahn, J., Chung, S.J., Kim, M. and

Chung, B.H., Directed immobilization of DNA-binding proteins on a cognate DNA-

modified chip surface, Journal of Biotechnology, 135, pp. 16-21, 2008.

58. Bouffartigues, E., Leh, H., Anger-Leroy, M., Rimsky, S. and Buckle, M., Rapid

coupling of Surface Plasmon Resonance (SPR and SPRi) and ProteinChip™ based

mass spectrometry for the identification of proteins in nucleoprotein interactions,

Nucleic acids research, 35, pp. e39, 2007.

59. Wegner, G.J., Lee, H.J., Marriott, G. and Corn, R.M., Fabrication of Histidine-

Tagged Fusion Protein Arrays for Surface Plasmon Resonance Imaging Studies of

Protein−Protein and Protein−DNA Interactions, Analytical chemistry, 75, pp.

4740-4746, 2003.

136

60. Nelson, B.P., Liles, M.R., Frederick, K.B., Corn, R.M. and Goodman, R.M., Label-

free detection of 16S ribosomal RNA hybridization on reusable DNA arrays using

surface plasmon resonance imaging, Environmental Microbiology, 4, pp. 735-743,

2002.

61. Li, Y., Lee, H.J. and Corn, R.M., Fabrication and characterization of RNA aptamer

microarrays for the study of protein–aptamer interactions with SPR imaging,

Nucleic acids research, 34, pp. 6416-6424, 2006.

62. Grant, C.F., Kanda, V., Yu, H., Bundle, D.R. and McDermott, M.T., Optimization

of Immobilized Bacterial Disaccharides for Surface Plasmon Resonance Imaging

Measurements of Antibody Binding, Langmuir, 24, pp. 14125-14132, 2008.

63. Mercey, E., Sadir, R., Maillart, E., Roget, A., Baleux, F., Lortat-Jacob, H. and

Livache, T., Polypyrrole Oligosaccharide Array and Surface Plasmon Resonance

Imaging for the Measurement of Glycosaminoglycan Binding Interactions,

Analytical chemistry, 80, pp. 3476-3482, 2008.

64. Wakao, M., Saito, A., Ohishi, K., Kishimoto, Y., Nishimura, T., Sobel, M. and

Suda, Y., Sugar Chips immobilized with synthetic sulfated disaccharides of

heparin/heparan sulfate partial structure, Bioorganic &amp; Medicinal Chemistry

Letters, 18, pp. 2499-2504, 2008.

65. Smith, E.A., Thomas, W.D., Kiessling, L.L. and Corn, R.M., Surface Plasmon

Resonance Imaging Studies of Protein-Carbohydrate Interactions, Journal of the

American Chemical Society, 125, pp. 6140-6148, 2003.

66. Aslan, K. and Geddes, C.D., Directional Surface Plasmon Coupled Luminescence

for Analytical Sensing Applications: Which Metal, What Wavelength, What

Observation Angle?, Analytical chemistry, 81, pp. 6913-6922, 2009.

67. Koutsioubas, A.G., Spiliopoulos, N., Anastassopoulos, D., Vradis, A.A. and Priftis,

G.D., Nanoporous alumina enhanced surface plasmon resonance sensors, Journal

of Applied Physics, 103, pp. 094521-094526, 2008.

68. Li, P.C.H., Microfluidics ad Lab on a Chip For Biological Analysis and Discovery,

pp. 2010.

69. Zou, Q., Development of surface plasmon resonance biosensors for resistance of

matrix effects, 3557184, pp. 159, 2013.

70. Otto, A., Excitation of nonradiative surface plasma waves in silver by the method

of frustrated total reflection, Zeitschrift für Physik, 216, pp. 398-410, 1968.

71. Sau, T.K. and Murphy, C.J., Seeded High Yield Synthesis of Short Au Nanorods in

Aqueous Solution, Langmuir, 20, pp. 6414-6420, 2004.

137

72. Huang, X., El-Sayed, I.H., Qian, W. and El-Sayed, M.A., Cancer Cell Imaging and

Photothermal Therapy in the Near-Infrared Region by Using Gold Nanorods,

Journal of the American Chemical Society, 128, pp. 2115-2120, 2006.

73. Lee, K.J., Nallathamby, P.D., Browning, L.M., Osgood, C.J. and Xu, X.-H.N., In

Vivo Imaging of Transport and Biocompatibility of Single Silver Nanoparticles in

Early Development of Zebrafish Embryos, ACS Nano, 1, pp. 133-143, 2007.

74. Masson, J.-F., Battaglia, T.M., Kim, Y.-C., Prakash, A., Beaudoin, S. and Booksh,

K.S., Preparation of analyte-sensitive polymeric supports for biochemical sensors,

Talanta, 64, pp. 716-725, 2004.

75. Masson, J.-F., Barnhart, M., Battaglia, T.M., Morris, G.E., Nieman, R.A., Young,

P.J., Lorson, C.L. and Booksh, K.S., Monitoring of recombinant survival motor

neuron protein using fiber-optic surface plasmon resonance, The Analyst, 129, pp.

855-859, 2004.

76. Masson, J.-F. and Booksh, K.S., Qualitative analysis of excess dielectric properties

of binary mixtures, ternary mixtures and mixing dynamics measurement using

surface plasmon resonance, Thermochimica Acta, 432, pp. 83-90, 2005.

77. Masson, J.-F., Kim, Y.-C., Obando, L.A., Peng, W. and Booksh, K.S., Fiber-Optic

Surface Plasmon Resonance Sensors in the Near-Infrared Spectral Region, Appl.

Spectrosc., 60, pp. 1241-1246, 2006.

78. Law, W.-C., Nanostructure materials for biosensing and bioimaging applications,

pp. 2010.

79. Abadian, P.N., Kelley, C.P. and Goluch, E.D., Cellular Analysis and Detection

Using Surface Plasmon Resonance Techniques, Analytical Chemistry, 86, pp.

2799-2812, 2014.

80. Subramanian, A., Irudayaraj, J. and Ryan, T., A mixed self-assembled monolayer-

based surface plasmon immunosensor for detection of E. coli O157:H7, Biosens.

Bioelectron., 21, pp. 998-1006, 2006.

81. Potter, M., Gonzalez-Ayala, S. and Silarug, N., Epidemiology of foodborne

diseases, Food Microbiology: Fundamentals and Frontiers, pp. 376-390, 1996.

82. Beran, G.W., Shoeman, H.P. and Anderson, K.F., Food safety: an overview of

problems, Dairy Food Environ. Sanit., 11, pp. 189–194, 1991.

83. Bresee, J.S., Dietz, V., Griffin, P.M., McCaig, L.F., Mead, P.S., Shapiro, C.,

Slutsker, L. and Tauxe, R.V., Food-Related Illness and Death in the United States,

J. Environ. Health, 62, pp. 9, 2000.

84. USDA/ERS, Economics of foodborne disease, 2014, pp. 2003.

138

85. Yang, L. and Bashir, R., Electrical/electrochemical impedance for rapid detection

of foodborne pathogenic bacteria, Biotechnol. Adv., 26, pp. 135-150, 2008.

86. Taylor, A.D., Yu, Q., Chen, S., Homola, J. and Jiang, S., Comparison of E. coli

O157:H7 preparation methods used for detection with surface plasmon resonance

sensor, Sens. Actuators, B, 107, pp. 202-208, 2005.

87. June, G.A., Relative effectiveness of selenite cystine broth , tetrathionate broth ,

and rappaport - vassiliadis medium for recovery of Salmonella spp. from raw flesh

, highly contaminated foods , and poultry feed : collaborative study, J. AOAC Int.,

79, pp. 1307, 1996.

88. Hoshino, T., Kawaguchi, M., Shimizu, N., Hoshino, N., Ooshima, T. and Fujiwara,

T., PCR detection and identification of oral streptococci in saliva samples using

GTF genes, Diagn. Microbiol. Infect. Dis., 48, pp. 195-199, 2004.

89. Rodrıguez-Lázaro, D., Hernández, M., Esteve, T., Hoorfar, J. and Pla, M., A rapid

and direct real time PCR-based method for identification of Salmonella spp, J.

Microbiol. Methods, 54, pp. 381-390, 2003.

90. Kong, R.Y.C., Lee, S.K.Y., Law, T.W.F., Law, S.H.W. and Wu, R.S.S., Rapid

detection of six types of bacterial pathogens in marine waters by multiplex PCR,

Water Res., 36, pp. 2802-2812, 2002.

91. Meng, J., Zhao, S., Doyle, M.P., Mitchell, S.E. and Kresovich, S., Polymerase

chain reaction for detecting Escherichia coli O157 : H7, Int. J. Food Microbiol.,

32, pp. 103, 1996.

92. Sperveslage, J., Stackebrandt, E., Lembke, F.W. and Koch, C., Detection of

bacterial contamination, including Bacillus spores, in dry growth media and in

milk by identification of their 16S rDNA by polymerase chain reaction, J.

Microbiol. Methods, 26, pp. 219, 1996.

93. Mozola, M.A., Genetics-based methods for detection of Salmonella spp. in foods,

J. AOAC Int., 89, pp. 517, 2006.

94. D'Souza, D.H. and Jaykus, L.A., Nucleic acid sequence based amplification for the

rapid and sensitive detection of Salmonella enterica from foods, J. Appl.

Microbiol., 95, pp. 1343-1350, 2003.

95. Andresen, L.O., Klausen, J., Barfod, K. and Sørensen, V., Detection of antibodies

to Actinobacillus pleuropneumoniae serotype 12 in pig serum using a blocking

enzyme-linked immunosorbent assay, Vet. Microbiol., 89, pp. 61-67, 2002.

96. Prior, J.L. and Titball, R.W., Monoclonal antibodies against Yersinia pestis

lipopolysaccharide detect bacteria cultured at 28°C or 37°C, Mol. Cell. Probes,

16, pp. 251-256, 2002.

139

97. Abdel-Hamid, I., Ivnitski, D., Atanasov, P. and Wilkins, E., Flow-through

immunofiltration assay system for rapid detection of E. coli O157:H7, Biosens.

Bioelectron., 14, pp. 309-316, 1999.

98. Hasebe, Y., Yokobori, K., Fukasawa, K., Kogure, T. and Uchiyama, S., Highly

sensitive electrochemical determination of Escherichia coli density using

tyrosinase - based chemically amplified biosensor, Anal. Chim. Acta, 357, pp. 51,

1997.

99. Brewster, J.D., Filtration capture and immunoelectrochemical detection for rapid

assay of Escherichia coli O157:H7, J. Immunol. Methods, 211, pp. 1, 1998.

100. Rishpon, J. and Ivnitski, D., An amperometric enzyme-channeling immunosensor,

Biosens. Bioelectron., 12, pp. 195, 1997.

101. Su, X.-L. and Li, Y., A self-assembled monolayer-based piezoelectric

immunosensor for rapid detection of Escherichia coli O157:H7, Biosens.

Bioelectron., 19, pp. 563-574, 2004.

102. Mo, X.-T., Zhou, Y.-P., Lei, H. and Deng, L., Microbalance-DNA probe method

for the detection of specific bacteria in water, Enzyme Microb. Technol., 30, pp.

583-589, 2002.

103. Su, X., Low, S., Kwang, J., Chew, V.H.T. and Li, S.F.Y., Piezoelectric quartz

crystal based veterinary diagnosis for Salmonella enteritidis infection in chicken

and egg, Sens. Actuators, B, 75, pp. 29-35, 2001.

104. Plomer, M., Guilbault, G.G. and Hock, B., Development of a piezoelectric

immunosensor for the detection of enterobacteria, Enzyme Microb. Technol., 14,

pp. 230, 1992.

105. Gehring, A.G., Use of a light - addressable potentiometric sensor for the detection

of Escherichia coli O157:H7, Anal. Biochem., 258, pp. 293, 1998.

106. Libby, J.M. and Wada, H.G., Detection of Neisseria meningitidis and Yersinia

pestis with a novel silicon-based sensor, J. Clin. Microbiol., 27, pp. 1456-1459,

1989.

107. Foley-Thomas, E.M., Phage infection, transfection and transformation of

Mycobacterium avium complex and Mycobacterium paratuberculosis,

Microbiology, 141, pp. 1173, 1995.

108. Blasco, R., Murphy, M.J., Sanders, M.F. and Squirrell, D.J., Specific assays for

bacteria using phage mediated release of adenylate kinase, J. Appl. Microbiol., 84,

pp. 661-666, 1998.

140

109. Pyle, B.H., Broadaway, S.C. and McFeters, G.A., A rapid, direct method for

enumerating respiring enterohemorrhagic E. coli O157:H7 in water., Appl.

Environ. Microbiol., 61, pp. 2614-2619, 1995.

110. Chowdhury, M.A.R., Xu, B., Montilla, R., Hasan, J.A.K., Huq, A. and Colwell,

R.R., A simplified immunofluorescence technique for detection of viable cells of

Vibrio cholerae O1 and O139, J. Microbiol. Methods, 24, pp. 165, 1995.

111. Hawkes, J.J., Long, M.J., Coakley, W.T. and McDonnell, M.B., Ultrasonic

deposition of cells on a surface, Biosens. Bioelectron., 19, pp. 1021-1028, 2004.

112. Ivnitski, D., Abdel-Hamid, I., Atanasov, P. and Wilkins, E., Biosensors for

detection of pathogenic bacteria, Biosens. Bioelectron., 14, pp. 599-624, 1999.

113. Malic, L., Veres, T. and Tabrizian, M., Nanostructured digital microfluidics for

enhanced surface plasmon resonance imaging, Biosens. Bioelectron., 26, pp. 2053-

2059, 2011.

114. Yanase, Y., Hiragun, T., Kaneko, S., Gould, H.J., Greaves, M.W. and Hide, M.,

Detection of refractive index changes in individual living cells by means of surface

plasmon resonance imaging, Biosens. Bioelectron., 26, pp. 674-681, 2010.

115. Eun, Y.J., Utada, A.S., Copeland, M.F., Takeuchi, S. and Weibel, D.B.,

Encapsulating bacteria in agarose microparticles using microfluidics for high-

throughput cell analysis and isolation, ACS Chem Biol, 6, pp. 260-266, 2011.

116. Griffiths, A.D. and Tawfik, D.S., Directed Evolution of an Extremely Fast

Phosphotriesterase by in vitro Compartmentalization, EMBO Journal, 22, pp. 24-

35, 2003.

117. Bouguelia, S., Roupioz, Y., Slimani, S., Mondani, L., Casabona, M.G., Durmort,

C., Vernet, T., Calemczuk, R. and Livache, T., On-chip microbial culture for the

specific detection of very low levels of bacteria, Lab on a chip, pp. 2013.

118. Bai, Y., Patil, S.N., Bowden, S.D., Poulter, S., Pan, J., Salmond, G.P., Welch, M.,

Huck, W.T. and Abell, C., Intra-species bacterial quorum sensing studied at single

cell level in a double droplet trapping system, Int J Mol Sci, 14, pp. 10570-10581,

2013.

119. Kiss, M.M., Ortoleva-Donnelly, L., Beer, N.R., Warner, J., Bailey, C.G., Colston,

B.W., Rothberg, J.M., Link, D.R. and Leamon, J.H., High-Throughput

Quantitative Polymerase Chain Reaction in Picoliter Droplets, Analytical

chemistry, 80, pp. 8975-8981, 2008.

120. Baret, J.C., Miller, O.J., Taly, V., Ryckelynck, M., El-Harrak, A., Frenz, L., Rick,

C., Samuels, M.L., Hutchison, J.B., Agresti, J.J., Link, D.R., Weitz, D.A. and

Griffiths, A.D., Fluorescence-activated droplet sorting (FADS): efficient

141

microfluidic cell sorting based on enzymatic activity, Lab on a chip, 9, pp. 1850-

1858, 2009.

121. Tawil, N., Sacher, E., Mandeville, R. and Meunier, M., Surface plasmon resonance

detection of E. coli and methicillin-resistant S. aureus using bacteriophages,

Biosensors and Bioelectronics, 37, pp. 24-29, 2012.

122. Koubová, V., Brynda, E., Karasová, L., Škvor, J., Homola, J., Dostálek, J., Tobiška,

P. and Rošický, J., Detection of foodborne pathogens using surface plasmon

resonance biosensors, Sensors and Actuators B: Chemical, 74, pp. 100-105, 2001.

123. Waswa, J., Irudayaraj, J. and DebRoy, C., Direct detection of E. Coli O157:H7 in

selected food systems by a surface plasmon resonance biosensor, LWT - Food

Science and Technology, 40, pp. 187-192, 2007.

124. Waswa, J.W., Debroy, C. and Irudayaraj, J., Rapid detection of Salmonella

enteritidis and Escherichia coli using surface plasmon resonance biosensor,

Journal of Food Process Engineering, 29, pp. 373-385, 2006.

125. Dudak, F.C. and Boyacı, İ.H., Development of an immunosensor based on surface

plasmon resonance for enumeration of Escherichia coli in water samples, Food

Research International, 40, pp. 803-807, 2007.

126. Alain, K. and Querellou, J., Cultivating the uncultured: limits, advances and future

challenges, Extremophiles, 13, pp. 583-594, 2009.

127. Jyoung, J.-Y., Hong, S., Lee, W. and Choi, J.-W., Immunosensor for the detection

of Vibrio cholerae O1 using surface plasmon resonance, Biosensors and

Bioelectronics, 21, pp. 2315-2319, 2006.

128. DeLong, E.F., Microbial community genomics in the ocean, Nat Rev Microbiol, 3,

pp. 459-469, 2005.

129. Leonard, P., Hearty, S., Quinn, J. and O’Kennedy, R., A generic approach for the

detection of whole Listeria monocytogenes cells in contaminated samples using

surface plasmon resonance, Biosensors and Bioelectronics, 19, pp. 1331-1335,

2004.

130. Puttharugsa, C., Wangkam, T., Huangkamhang, N., Gajanandana, O., Himananto,

O., Sutapun, B., Amarit, R., Somboonkaew, A. and Srikhirin, T., Development of

surface plasmon resonance imaging for detection of Acidovorax avenae subsp.

citrulli (Aac) using specific monoclonal antibody, Biosensors and Bioelectronics,

26, pp. 2341-2346, 2011.

131. Ostuni, E., Chapman, R.G., Liang, M.N., Meluleni, G., Pier, G., Ingber, D.E. and

Whitesides, G.M., Self-Assembled Monolayers That Resist the Adsorption of

Proteins and the Adhesion of Bacterial and Mammalian Cells, Langmuir, 17, pp.

6336-6343, 2001.

142

132. Holmes, S.D., May, K., Johansson, V., Markey, F. and Critchley, I.A., Studies on

the interaction of staphylococcus aureus and staphylococcus epidermidis with

fibronectin using surface plasmon resonance (BIAcore), Journal of

Microbiological Methods, 28, pp. 77-84, 1997.

133. Lan, Y.-b., Wang, S.-z., Yin, Y.-g., Hoffmann, W.C. and Zheng, X.-z., Using a

Surface Plasmon Resonance Biosensor for Rapid Detection of Salmonella

Typhimurium in Chicken Carcass, Journal of Bionic Engineering, 5, pp. 239-246,

2008.

134. Aoi, Y., Kinoshita, T., Hata, T., Ohta, H., Obokata, H. and Tsuneda, S., Hollow-

fiber membrane chamber as a device for in situ environmental cultivation, Appl

Environ Microbiol, 75, pp. 3826-3833, 2009.

135. Chabot, V., Cuerrier, C.M., Escher, E., Aimez, V., Grandbois, M. and Charette,

P.G., Biosensing based on surface plasmon resonance and living cells, Biosensors

and Bioelectronics, 24, pp. 1667-1673, 2009.

136. Markx, G.H., Dyda, P.A. and Pethig, R., Dielectrophoretic Separation of Bacteria

Using a Conductivity Gradient, Journal of Biotechnology, 51, pp. 175-180, 1996.

137. Iribe. Y., Shinohara, H. and Suzuki, M., Label-free detection of B and T cell

responses by using high resolution 2D-SPR imaging sensor, 14th International

Conference on Miniaturized Systems for Chemistry and Life Sciences, 2010.

138. Yanase, Y., Suzuki, H., Tsutsui, T., Uechi, I., Hiragun, T., Mihara, S. and Hide,

M., Living cell positioning on the surface of gold film for SPR analysis, Biosensors

and Bioelectronics, 23, pp. 562-567, 2007.

139. Pohl, H.A., The Motion and Precipitation of Suspensoids in Divergent Electric

Fields, Journal of Applied Physics, 22, pp. 869, 1951.

140. Hide, M., Tsutsui, T., Sato, H., Nishimura, T., Morimoto, K., Yamamoto, S. and

Yoshizato, K., Real-Time Analysis of Ligand-Induced Cell Surface and

Intracellular Reactions of Living Mast Cells Using a Surface Plasmon Resonance-

Based Biosensor, Analytical biochemistry, 302, pp. 28-37, 2002.

141. Pohl, H.A. and Crane, J.S., Dielectrophoresis of Cells, Biophysical journal, 11, pp.

711-727, 1971.

142. Hughes, M.P., Strategies for Dielectrophoretic Separation in Laboratory-on-a-

Chip Systems, Electrophoresis, 23, pp. 2569-2582, 2002.

143. Li, H. and Bashir, R., Dielectrophoretic Separation and Manipulation of Live and

Heat-Treated Cells of Listeria on Microfabricated Devices with Interdigitated

Electrodes, Sensors and Actuators B, 86, pp. 215-221, 2002.

143

144. ying fang, Z.S., Lan Li, Yong Cao, Li-Ying Gu, Qing Gu, Xiao-Qi Zhong, Chao-

Hui Yu, You-Ming Li, A study of the efficacy of bacterial biofilm cleanout for

gastrointestinal endoscopes, 16, pp. 1019-1024, 2010.

145. ZOBELL, C.E. and ANDERSON, D.Q., Observation on the multiplication of

bacteria in different volumes of stored sea water and the influence of oxygen

tension and solid surfaces, The Biological Bulletin, 71, pp. 324-342, 1936.

146. Vickery, K., Pajkos, A. and Cossart, Y., Removal of biofilm from endoscopes:

evaluation of detergent efficiency, American Journal of Infection Control, 32, pp.

170-176, 2004.

147. Lewis, K., Persister cells, Annu Rev Microbiol, 64, pp. 357-372, 2010.

148. Sutherland, I.W., The biofilm matrix an immobilized but dynamic microbial

environment, Trends in Microbiology, 9, pp. 222-227, 2001.

149. Shi, X. and Zhu, X., Biofilm formation and food safety in food industries, Trends

in Food Science & Technology, 20, pp. 407-413, 2009.

150. Flemming, H.-C. and Wingender, J., The biofilm matrix, Nat Rev Micro, 8, pp.

623-633, 2010.

151. Simões, M., Simões, L.C. and Vieira, M.J., A review of current and emergent

biofilm control strategies, LWT - Food Science and Technology, 43, pp. 573-583,

2010.

152. Cochran, W.L., McFeters, G.A. and Stewart, P.S., Reduced susceptibility of thin

Pseudomonas aeruginosa biofilms to hydrogen peroxide and monochloramine,

Journal of Applied Microbiology, 88, pp. 22-30, 2000.

153. Gilbert, P. and McBain, A.J., Biofilms: Their impact on health and their

recalcitrance toward biocides, American Journal of Infection Control, 29, pp. 252-

255, 2001.

154. Ntsama-Essomba, C., Bouttier, S., Ramaldes, M., Dubois-Brissonnet, F. and

Fourniat, J., Resistance of Escherichia coli growing as biofilms to disinfectants,

Vet Res, 28, pp. 353-363, 1997.

155. Bryers, J.D. and Ratner, B.D., Bioinspired Implant Materials Befuddle Bacteria,

American Society for Microbiology, pp. 232-237, 2004.

156. Chae, M.S., Schraft, H., Truelstrup Hansen, L. and Mackereth, R., Effects of

physicochemical surface characteristics of Listeria monocytogenes strains on

attachment to glass, Food Microbiology, 23, pp. 250-259, 2006.

157. Donlan, R.M., Biofilms: Microbial life on surfaces, Emerging Infectious Diseases,

8, pp. 881-890, 2002.

144

158. Millsap, K.W., Reid, G., van der Mei, H.C. and Busscher, H.J., Adhesion of

Lactobacillus species in urine and phosphate buffer to silicone rubber and glass

under flow, Biomaterials, 18, pp. 87-91, 1997.

159. Oulahal, N., Brice, W., Martial, A. and Degraeve, P., Quantitative analysis of

survival of Staphylococcus aureus or Listeria innocua on two types of surfaces:

Polypropylene and stainless steel in contact with three different dairy products,

Food Control, 19, pp. 178-185, 2008.

160. Simões, M., Simões, L.C., Machado, I., Pereira, M.O. and Vieira, M.J., Control of

Flow-Generated Biofilms with Surfactants: Evidence of Resistance and Recovery,

Food and Bioproducts Processing, 84, pp. 338-345, 2006.

161. Parsek, M.R. and Greenberg, E.P., Sociomicrobiology: the connections between

quorum sensing and biofilms, Trends in Microbiology, 13, pp. 27-33, 2005.

162. Bloß, R. and Kampf, G., Test models to determine cleaning efficacy with different

types of bioburden and its clinical correlation, Journal of Hospital Infection, 56,

Supplement 2, pp. 44-48, 2004.

163. Chicurel, M., Slimebusters, Nature, 408, pp. 284-286, 2000.

164. Costerton, J.W., Stewart, P.S. and Greenberg, E.P., Bacterial Biofilms: A Common

Cause of Persistent Infections, Science, 284, pp. 1318-1322, 1999.

165. Silva, R.C., Carver, R.A., Ojano-Dirain, C.P. and Antonelli, P.J., Efficacy of

disinfecting solutions in removing biofilms from polyvinyl chloride tracheostomy

tubes, The Laryngoscope, 123, pp. 259-263, 2013.

166. Potera, C., Forging a Link Between Biofilms and Disease, Science, 283, pp. 1837-

1839, 1999.

167. Chen, J., Rossman, M.L. and Pawar, D.M., Attachment of enterohemorrhagic

Escherichia coli to the surface of beef and a culture medium, LWT - Food Science

and Technology, 40, pp. 249-254, 2007.

168. Jessen, B. and Lammert, L., Biofilm and disinfection in meat processing plants,

International Biodeterioration & Biodegradation, 51, pp. 265-269, 2003.

169. Somers, E.B. and Wong, A.C.L., Efficacy of two cleaning and sanitizing

combinations on Listeria monocytogenes biofilms formed at low temperature on a

variety of materials in the presence of ready-to-eat meat residue, Journal of food

protection, 67, pp. 2218-2229, 2004.

170. Frimat, J.P., Becker, M., Chiang, Y.Y., Marggraf, U., Janasek, D., Hengstler, J.G.,

Franzke, J. and West, J., A microfluidic array with cellular valving for single cell

co-culture, Lab on a chip, 11, pp. 231-237, 2011.

145

171. Haber, C., Microfluidics in commercial applications; an industry perspective, Lab

on a chip, 6, pp. 1118-1121, 2006.

172. Ismagilov, R.F., Rosmarin, D., Kenis, P.J.A., Chiu, D.T., Zhang, W., Stone, H.A.

and Whitesides, G.M., Pressure-Driven Laminar Flow in Tangential

Microchannels: an Elastomeric Microfluidic Switch, Analytical chemistry, 73, pp.

4682-4687, 2001.

173. Lapizco-Encinas, B.H., Simmons, B.A., Cummings, E.B. and Fintschenko, Y.,

Dielectrophoretic Concentration and Separation of Live and Dead Bacteria in an

Array of Insulators, Analytical chemistry, 76, pp. 1571-1579, 2004.

174. Kaewnopparat, S., Dangmanee, N., Kaewnopparat, N., Srichana, T., Chulasiri, M.

and Settharaksa, S., In vitro probiotic properties of Lactobacillus fermentum SK5

isolated from vagina of a healthy woman, Anaerobe, 22, pp. 6-13, 2013.

175. Brackett, R., Incidence, contributing factors, and control of bacterial pathogens in

produce, Postharvest Biology and Technology, 15, pp. 305-311, 1999.

176. Francis, G.A., Thomas, C. and O'Beirne, D., The microbiological safety of

minimally processed vegetables, International Journal of Food Science &

Technology, 34, pp. 1-22, 1999.

177. Ölmez, H. and Temur, S.D., Effects of different sanitizing treatments on biofilms

and attachment of Escherichia coli and Listeria monocytogenes on green leaf

lettuce, LWT - Food Science and Technology, 43, pp. 964-970, 2010.

178. Easton, L., Escherichia coli O157 : occurrence , transmission and laboratory

detection, British Journal of Biomedical Science, 54, pp. 57, 1997.

179. Edd, J., Payen, S., Rubinsky, B., Stoller, M.L. and Sitti, M., Biomimetic propulsion

for a swimming surgical micro-robot, Intelligent Robots and Systems, 2003. (IROS

2003). Proceedings. 2003 IEEE/RSJ International Conference on, 3, pp. 2583-2588

vol.2583, 2003.

180. Sitti, B.B.M., E. coli inspired propusion for swimming microrobots , American

Society of Mechanical Engineering, pp. 2004.

181. Ryder, C., Byrd, M. and Wozniak, D.J., Role of polysaccharides in Pseudomonas

aeruginosa biofilm development, Current Opinion in Microbiology, 10, pp. 644-

648, 2007.

182. Deligianni, E., Pattison, S., Berrar, D., Ternan, N., Haylock, R., Moore, J., Elborn,

S. and Dooley, J., Pseudomonas aeruginosa Cystic Fibrosis isolates of similar

RAPD genotype exhibit diversity in biofilm forming ability in vitro, BMC

Microbiol, 10, pp. 1-13, 2010.

146

183. Costerton, J.W. and Anwar, H., Pseudomonas aeruginosa: The microbe and

pathogen, Infectious Disease and Therapy; Pseudomonas aeruginosa infections and

treatment, 12, pp. 1-20, 1994.

184. Botzenhardt, K. and Doring, G., Ecology and epidemiology of Pseudomonas

aeruginosa. Pseudomonas aeruginosa as an Opportunistic Pathogen, pp. 419,

1993.

185. Fick, R.B., Pseudomonas aeruginosa: The Opportunist, Chapter1: Pseudomonas

aeruginosa—the Microbial Hyena and Its Role in Disease, pp. 272, 1993.

186. Chastre, J. and Fagon, J.-Y., Ventilator-associated Pneumonia, Am. J. Respir. Crit.

Care Med., 165, pp. 867-903, 2002.

187. Gibson, R.L., Burns, J.L. and Ramsey, B.W., Pathophysiology and Management of

Pulmonary Infections in Cystic Fibrosis, American Journal of Respiratory and

Critical Care Medicine, 168, pp. 918-951, 2003.

188. Van Delden, C. and Iglewski, B.H., Cell-to-cell signaling and Pseudomonas

aeruginosa infections, Emerging Infectious Diseases, 4, pp. 551-560, 1998.

189. Rothberg, J.M., Hinz, W., Rearick, T.M., Schultz, J., Mileski, W., Davey, M.,

Leamon, J.H., Johnson, K., Milgrew, M.J., Edwards, M., Hoon, J., Simons, J.F.,

Marran, D., Myers, J.W., Davidson, J.F., Branting, A., Nobile, J.R., Puc, B.P.,

Light, D., Clark, T.A., Huber, M., Branciforte, J.T., Stoner, I.B., Cawley, S.E.,

Lyons, M., Fu, Y., Homer, N., Sedova, M., Miao, X., Reed, B., Sabina, J.,

Feierstein, E., Schorn, M., Alanjary, M., Dimalanta, E., Dressman, D., Kasinskas,

R., Sokolsky, T., Fidanza, J.A., Namsaraev, E., McKernan, K.J., Williams, A.,

Roth, G.T. and Bustillo, J., An integrated semiconductor device enabling non-

optical genome sequencing, Nature, 475, pp. 348-352, 2011.

190. Stover, C.K., Pham, X.Q., Erwin, A.L., Mizoguchi, S.D., Warrener, P., Hickey,

M.J., Brinkman, F.S.L., Hufnagle, W.O., Kowalik, D.J., Lagrou, M., Garber, R.L.,

Goltry, L., Tolentino, E., Westbrock-Wadman, S., Yuan, Y., Brody, L.L., Coulter,

S.N., Folger, K.R., Kas, A., Larbig, K., Lim, R., Smith, K., Spencer, D., Wong,

G.K.S., Wu, Z., Paulsen, I.T., Reizer, J., Saier, M.H., Hancock, R.E.W., Lory, S.

and Olson, M.V., Complete genome sequence of Pseudomonas aeruginosa PAO1,

an opportunistic pathogen, Nature, 406, pp. 959-964, 2000.

191. Fischer, E., Battling MRSA - Superdrug for the Superbug, pp. 2011.

192. Zmantar, T., Chaieb, K., Miladi, H., Mahdouani, K. and Bakhrouf, A., Detection

of the intercellular adhesion loci (ica) in clinical Staphylococcus aureus strains

responsible for hospital acquired auricular infection, Annals of Microbiology, 56,

pp. 349-352, 2006.

193. Pfaller, M.A., Jones, R.N., Doern, G.V., Sader, H.S., Kugler, K.C. and Beach,

M.L., Survey of blood stream infections attributable to gram-positive cocci:

147

frequency of occurrence and antimicrobial susceptibility of isolates collected in

1997 in the United States, Canada, and Latin America from the SENTRY

antimicrobial surveillance program, Diagnostic Microbiology and Infectious

Disease, 33, pp. 283-297, 1999.

194. Iorio, N.L.P., Lopes, A., Schuenck, R.P., Barcellos, A.G., Olendzki, A.N., Lopez,

G.L. and dos Santos, K.R.N., A combination of methods to evaluate biofilm

production may help to determine the clinical relevance of Staphylococcus in blood

cultures, Microbiology and Immunology, 55, pp. 28-33, 2011.

195. Abadian, P.N. and Goluch, E.D., Surface plasmon resonance imaging (SPRi) for

multiplexed evaluation of bacterial adhesion onto surface coatings, Analytical

Methods, 7, pp. 115-122, 2015.

196. Neuman, K.C., Chadd, E.H., Liou, G.F., Bergman, K. and Block, S.M.,

Characterization of Photodamage to Escherichia coli in Optical Traps,

Biophysical journal, 77, pp. 2856-2863, 1999.

197. Di Carlo, D., Aghdam, N. and Lee, L.P., Single-Cell Enzyme Concentrations,

Kinetics, and Inhibition Analysis Using High-Density Hydrodynamic Cell Isolation

Arrays, Analytical chemistry, 78, pp. 4925-4930, 2006.

198. Li, P.C., de Camprieu, L., Cai, J. and Sangar, M., Transport, retention and

fluorescent measurement of single biological cells studied in microfluidic chips,

Lab on a chip, 4, pp. 174-180, 2004.

199. Rowat, A.C., Bird, J.C., Agresti, J.J., Rando, O.J. and Weitz, D.A., Tracking

lineages of single cells in lines using a microfluidic device, Proceedings of the

National Academy of Sciences of the United States of America, 106, pp. 18149-

18154, 2009.

200. Dickson, J.S. and Koohmaraie, M., Cell Surface Charge Characteristics and Their

Relationship to Bacterial Attachment to Meat Surfaces, Applied and Environmental

Microbiology, 55, pp. 832-836, 1989.

201. Ashkin, A., Applications of Laser Radiation Pressure, Science, 210, pp. 1081-

1088, 1980.

202. Kenndler, E., Capillary Electrophoresis of Macromolecular Biological

Assemblies: Bacteria and Viruses, Trends in Analytical Chemistry, 20, pp. 543-

551, 2001.

203. Lee, H.L., Boccazzi, P., Ram, R.J. and Sinskey, A.J., Microbioreactor arrays with

integrated mixers and fluid injectors for high-throughput experimentation with pH

and dissolved oxygen control, Lab on a chip, 6, pp. 1229-1235, 2006.

204. Markx, G.H. and Pethig, R., Dielectrophoretic Separation of Cells: Continuous

Separation, Biotechnology and Bioengineering, 45, pp. 337-343, 1995.

148

205. Andersson, H. and van den Berg, A., Microfluidic devices for cellomics: a review,

Sensors and Actuators B: Chemical, 92, pp. 315-325, 2003.

206. Ashkin, A., Dziedzic, J.M. and Yamane, T., Optical Trapping and Manipulation of

Single Cells Using Infrared Laser Beams, Nature, 330, pp. 769-771, 1987.

207. Zare, R.N. and Kim, S., Microfluidic platforms for single-cell analysis, Annu Rev

Biomed Eng, 12, pp. 187-201, 2010.

208. Ashkin, A., Dziedzic, J.M., Bjorkholm, J.E. and Chu, S., Observation of a Single-

Beam Gradient Force Optical Trap for Dielectric Particles, Optics Letters, 11, pp.

288-290, 1986.

209. Ericsson, M., Hanstorp, D., Hagberg, P., Enger, J. and Nystrom, T., Sorting Out

Bacterial Viability with Optical Tweezers, Journal of bacteriology, 182, pp. 5551-

5555, 2000.

210. Markx, G.H., Huang, Y., Zhou, X.-F. and Pethig, R., Dielectrophoretic

Characterization and Separation of Micro-organisms, Microbiology, 140, pp. 585-

591, 1994.

211. Chiou, P.Y., Ohta, A.T. and Wu, M.C., Massively parallel manipulation of single

cells and microparticles using optical images, Nature, 436, pp. 370-372, 2005.

212. Chen, Z., Huang, Z., Cheng, Y., Pan, D., Pan, X., Yu, M., Pan, Z., Lin, Z., Guan,

X. and Wu, Z., Cr(VI) uptake mechanism of Bacillus cereus, Chemosphere, 87, pp.

211-216, 2012.

213. Bollmann, A., Palumbo, A.V., Lewis, K. and Epstein, S.S., Isolation and

physiology of bacteria from contaminated subsurface sediments, Appl Environ

Microbiol, 76, pp. 7413-7419, 2010.

214. Mosteller, T.M. and Bishop, J.R., Sanitizer efficacy against attached bacteria in a

milk biofilm, Journal of food protection, pp. 1993.

215. Maukonen, J., Mättö, J., Wirtanen, G., Raaska, L., Mattila-Sandholm, T. and

Saarela, M., Methodologies for the characterization of microbes in industrial

environments: a review, Journal of Industrial Microbiology and Biotechnology, 30,

pp. 327-356, 2003.

216. Gram, L., Bagge-Ravn, D., Ng, Y.Y., Gymoese, P. and Vogel, B.F., Influence of

food soiling matrix on cleaning and disinfection efficiency on surface attached

Listeria monocytogenes, Food Control, 18, pp. 1165-1171, 2007.

217. Gilbert, P., Allison, D.G. and McBain, A.J., Biofilms in vitro and in vivo: do

singular mechanisms imply cross-resistance?, Journal of Applied Microbiology,

92, pp. 98S-110S, 2002.

149

218. Gilbert, P. and McBain, A.J., Potential Impact of Increased Use of Biocides in

Consumer Products on Prevalence of Antibiotic Resistance, Clinical Microbiology

Reviews, 16, pp. 189-208, 2003.

219. David, D., Understanding biofilm resistance to antibacterial agents, Nature

Reviews Drug Discovery, 2, pp. 114-122, 2003.

220. Elzoghby, A.O., Helmy, M.W., Samy, W.M. and Elgindy, N.A., Spray-dried

casein-based micelles as a vehicle for solubilization and controlled delivery of

flutamide: Formulation, characterization, and in vivo pharmacokinetics, European

Journal of Pharmaceutics and Biopharmaceutics, 84, pp. 487-496, 2013.

221. Vasseur, P., Vallet-Gely, I., Soscia, C., Genin, S. and Filloux, A., The pel genes of

the Pseudomonas aeruginosa PAK strain are involved at early and late stages of

biofilm formation, Microbiology, 151, pp. 985-997, 2005.

222. Roman, J.A. and Sgarbieri, V.C., The hydrophilic, foaming and emulsifying

properties of casein concentrates produced by various methods, International

Journal of Food Science & Technology, 41, pp. 609-617, 2006.

223. Ying, P., Jin, G. and Tao, Z., Competitive adsorption of collagen and bovine serum

albumin—effect of the surface wettability, Colloids and Surfaces B: Biointerfaces,

33, pp. 259-263, 2004.

224. Pascual, A., Fleer, A., Westerdaal, N. and Verhoef, J., Modulation of adherence of

coagulase-negative staphylococci to teflon catheters in vitro, European Journal of

Clinical Microbiology, 5, pp. 518-522, 1986.

225. Jeyachandran, Y.L., Mielczarski, E., Rai, B. and Mielczarski, J.A., Quantitative

and Qualitative Evaluation of Adsorption/Desorption of Bovine Serum Albumin on

Hydrophilic and Hydrophobic Surfaces, Langmuir, 25, pp. 11614-11620, 2009.

226. Reimhult, K., Petersson, K. and Krozer, A., QCM-D Analysis of the Performance

of Blocking Agents on Gold and Polystyrene Surfaces, Langmuir, 24, pp. 8695-

8700, 2008.

227. Yurchenco, J.A., Hopper, M.W. and Warren, G.H., Degradation of penicillin G

and three semisynthetic penicillins by bacillus sereus and staphylococcus aureus

penicillinase, Canadian Journal of Microbiology, 12, pp. 35-42, 1966.

228. Rusconi, R., Lecuyer, S., Autrusson, N., Guglielmini, L. and Stone, Howard A.,

Secondary Flow as a Mechanism for the Formation of Biofilm Streamers,

Biophysical journal, 100, pp. 1392-1399, 2011.

229. Rusconi, R., Lecuyer, S., Guglielmini, L. and Stone, H.A., Laminar flow around

corners triggers the formation of biofilm streamers, Journal of The Royal Society

Interface, 7, pp. 1293-1299, 2010.

150

230. Drescher, K., Shen, Y., Bassler, B.L. and Stone, H.A., Biofilm streamers cause

catastrophic disruption of flow with consequences for environmental and medical

systems, Proceedings of the National Academy of Sciences, 110, pp. 4345-4350,

2013.

231. Kim, J., Hegde, M., Kim, S.H., Wood, T.K. and Jayaraman, A., A microfluidic

device for high throughput bacterial biofilm studies, Lab on a chip, 12, pp. 1157,

2012.

232. Lötters, J.C., The mechanical properties of the rubber elastic polymer

polydimethylsiloxane for sensor applications, Journal of Micromechanics and

Microengineering, 7, pp. 145-147, 1997.