SPR Sensor Surfaces based on Self-Assembled Monolayers Anna Bergström

SPR Sensor Surfaces based on Self-Assembled Monolayers Anna Bergström
Department of Physics, Chemistry and Biology
Master’s thesis
SPR Sensor Surfaces based on
Self-Assembled Monolayers
Anna Bergström
Linköping, January 2009
LiTH-IFM-x-EX--09/2044--SE
Department of Physics, Chemistry and Biology
Linköping University
SE-581 83 Linköping, Sweden
Department of Physics, Chemistry and Biology
Master’s thesis
SPR Sensor Surfaces based on
Self-Assembled Monolayers
Anna Bergström
Linköping, January 2009
LiTH-IFM-x-EX--09/2044--SE
Supervisor
Per Kjellin
Division of Advanced Systems
GE Healthcare Life Sciences, Uppsala
Supervisor and Examiner
Bo Liedberg
Division of Molecular Physics
Department of Physics, Chemistry and Biology
Linköping University, Linköping
Department of Physics, Chemistry and Biology
Linköping University
SE-581 83 Linköping, Sweden
Abstract
The study and understanding of molecular interactions is fundamentally important in
today’s field of life sciences and there is a demand for well designed surfaces for
biosensor applications. The biosensor has to be able to detect specific molecular
interactions, while non-specific binding of other substances to the sensor surface
should be kept to a minimum.
The objective of this master´s thesis was to design sensor surfaces based on selfassembled monolayers (SAMs) and evaluate their structural characteristics as well as
their performance in Biacore systems. By mixing different oligo (ethylene glycol)
terminated thiol compounds in the SAMs, the density of functional groups for
biomolecular attachment could be controlled. Structural characteristics of the SAMs
were studied using Ellipsometry, Contact Angle Goniometry, IRAS and XPS. Surfaces
showing promising results were examined further with Surface Plasmon Resonance in
Biacore instruments.
Mixed SAM surfaces with a tailored degree of functional COOH groups could be
prepared. The surfaces showed promising characteristics in terms of stability,
immobilization capacity of biomolecules, non-specific binding and kinetic assay
performance, while further work needs to be dedicated to the improvement of their
storage stability. In conclusion, the SAM based sensor surfaces studied in this thesis
are interesting candidates for Biacore applications.
Acronyms and abbreviations
SAM Self-Assembled Monolayer
ELISA Enzyme-Linked ImmunoSorbent Assay
SPR
Surface Plasmon Resonance
IFC
Integrated microFluidic Cartridge
RU
Resonance Units
EDC
N-Ethyl-N’-(3-Dimethylaminopropyl) Carbodiimide
NHS
N-Hydroxy-succinimide
IRAS Infrared Reflection Absorption Spectroscopy
XPS
X-ray Photoelectron Spectroscopy
OEG
Oligo (Ethylene Glycol)
FC
Flow Cell
P20
Polyoxyethylene (20) Sorbitan Monolaurate
SDS
Sodium Dodecyl Sulfate
Table of contents
1
2
Introduction ..................................................................................................... 1
1.1
Background ........................................................................................................ 1
1.2
Aim ..................................................................................................................... 2
1.3
General approach .............................................................................................. 2
Theory .............................................................................................................. 3
2.1
Self-assembled monolayers (SAMs) .................................................................. 3
2.2
Surface Plasmon Resonance ............................................................................. 4
2.2.1
The principle of Surface Plasmon Resonance ............................................. 4
2.2.2
Biacore systems .......................................................................................... 5
2.2.3
Ligand immobilization ................................................................................ 7
2.3
2.3.1
The principle of infrared spectroscopy ....................................................... 8
2.3.2
Infrared Reflection Absorption Spectroscopy (IRAS) .................................. 9
2.4
3
Infrared spectroscopy ....................................................................................... 8
Null ellipsometry ............................................................................................. 10
2.4.1
Polarization of light .................................................................................. 10
2.4.2
The principle of Null Ellipsometry ............................................................. 10
2.5
Contact angle goniometry ............................................................................... 11
2.6
X-ray Photoelectron Spectroscopy (XPS) ........................................................ 12
Experimental section ...................................................................................... 15
3.1
Materials .......................................................................................................... 15
3.1.1
Substrates ................................................................................................. 15
3.1.2
Thiol compounds ....................................................................................... 15
3.1.3
Reagents ................................................................................................... 15
3.2
SAM preparation ............................................................................................. 16
3.3
Characterization of SAM quality and structure............................................... 17
3.3.1
Null ellipsometry ....................................................................................... 17
3.3.2
Contact angle goniometry ........................................................................ 18
3.3.3
XPS ............................................................................................................ 18
3.3.4
IRAS ........................................................................................................... 19
3.4
Characterization of sensor chip performance in Biacore ............................... 20
3.4.1 Surface stability ......................................................................................... 21
4
3.4.2
Immobilization capacity and analyte binding .......................................... 21
3.4.3
Non-specific binding – plasma .................................................................. 22
3.4.4
Non-specific binding – proteins ................................................................ 22
3.4.5
Ligand stability.......................................................................................... 23
3.4.6
Kinetic assay performance........................................................................ 23
3.4.7
Detergents ................................................................................................ 24
3.4.8
Storage stability ........................................................................................ 24
Results and discussion .................................................................................... 27
4.1
Characterization of SAM quality and structure............................................... 27
4.1.1
Null ellipsometry ....................................................................................... 27
4.1.2
Contact angle goniometry ........................................................................ 28
4.1.3
XPS ............................................................................................................ 30
4.1.4
IRAS ........................................................................................................... 31
4.2
Characterization of sensor chip performance in Biacore ............................... 35
4.2.1
Surface stability ........................................................................................ 35
4.2.2
Immobilization capacity and analyte binding .......................................... 36
4.2.3
Non-specific binding – plasma .................................................................. 39
4.2.4
Non-specific binding – proteins ................................................................ 40
4.2.5
Choice of surfaces for further characterization ........................................ 42
4.2.6
Ligand stability.......................................................................................... 43
4.2.7
Kinetic assay performance........................................................................ 44
4.2.8
Detergents ................................................................................................ 45
4.2.9
Storage stability ........................................................................................ 47
4.2.10 Summary ................................................................................................... 49
5
Conclusions .................................................................................................... 53
6
Future aspects ................................................................................................ 55
7
Acknowledgements ........................................................................................ 57
8
References ...................................................................................................... 59
Appendix A – Reagents ......................................................................................... 61
Appendix B – Experimental details ........................................................................ 63
1 Introduction
1.1
Background
Understanding the nature of molecular interactions is fundamentally important in all
areas of life science. Binding of an antibody to its antigen, a drug compound to its
target or a growth factor to its receptor – they are all examples of the dynamic
interactions that drive and regulate all biological processes. As these processes are
real-time events, there is consequently a need for real-time analysis. End-point
assays, like radio labeling or ELISA, only provide a snapshot of the interaction at
equilibrium. However, if you can study the molecular interaction over time, a fuller
understanding of the process can be achieved and questions like the following can be
answered:
Do the potentially interacting molecules bind to each other?
How specific is the interaction?
How strong is the binding (affinity)?
How fast do the interactants bind to and dissociate from each other (kinetics)?
What is the concentration of interactant in the sample?
Systems on which to perform such analytical studies are provided by GE Healthcare
Life Sciences, where this Master’s thesis work was performed. The company was
formerly known under the name Biacore, but since 2006 it is part of the GE
Healthcare organization and Biacore is now a product line name only. Biacore
systems offer label-free interaction analysis in real time, and information about
concentration, affinity and kinetics of biomolecules can be obtained. The systems are
used in areas such as drug discovery, proteomics, food analysis and in many life
science and academic research applications. The technology is based on an optical
phenomenon called surface plasmon resonance (SPR), whereby biomolecular
interactions can be detected as they occur.
A vital part of the Biacore instrument system is the sensor surface, where the
interactions to be studied take place. It consists of a glass surface coated with a thin
layer of gold, on top of which a hydrophilic layer, called the matrix, is created. The
matrix contains functional groups for attachment of biomolecules and its chemistry
affects the detection sensitivity and specificity. A commonly used matrix is the
carboxymethylated dextran, a flexible unbranched carbohydrate polymer that forms
a three-dimensional structure on the sensor surface. As the technical development
advances, with a rising sensitivity of the SPR instrumentation hardware, demands on
1
the sensor surface chemistry are increased. It becomes particularly important to
minimize drift and non-specific binding of biomolecules, in order to be able to detect
low interaction signals. One possible approach to reduce such noise would be to
minimize matrix fluctuations on the sensor surface by using a much shorter and more
rigid matrix than the commonly used dextran. Self-assembled monolayers (SAMs) of
oligo (ethylene glycol) terminated alkanethiols could potentially combine the benefits
of a shorter matrix with the excellent resistance towards non-specific binding that is
associated with dextran.
1.2
Aim
The aim of this Master’s thesis was to create sensor surfaces based on mixed selfassembled monolayers of different oligo(ethylene glycol) terminated alkanethiol
compounds and evaluate their quality and structural characteristics as well as their
performance in Biacore systems. SAM stability, immobilization capacity of
biomolecules and the SAM surfaces’ capability to reduce drift and non-specific
binding should be compared to existing Biacore sensor surfaces.
1.3
General approach
Firstly, SAM sensor surfaces were prepared and quality and structural characteristics
were studied. This first part of the experimental work was performed at the
Department of Physics, Chemistry and Biology at Linköping University. Secondly,
surfaces that presented promising characteristics were examined further with surface
plasmon resonance at the Division of Advanced Systems, GE Healthcare Life Sciences,
Uppsala.
2
2 Theory
2.1 Self-assembled monolayers (SAMs)
Self-assembled monolayers are thin, organic films that form spontaneously on solid
surfaces.[1] A typical molecule used for self-assembly consists of a three parts; a head
group, for interaction with the surface, an alkyl chain and finally a tail group,
providing functionality to the SAM. (see figure 2.1) The substrate surface is simply
immersed into a dilute (µM-mM) solution of appropriate molecules whereby strong
specific interactions between the molecules’ reactive head groups and the substrate
surface take place, resulting in chemisorption. Once pinned to the surface, the
molecules start organizing themselves into an ordered, densely packed layer. [2]
Organization is driven by van der Waals interactions between the chains, as the
system aims to optimize lateral interactions and reach potential energy minimum.
This organization process, which takes several hours to complete, is illustrated in
figure 2.1 below.
Figure 2.1 Schematic illustration of SAM formation. Thiol molecules are rapidly chemisorbed to the
surface, but the organization into a densely packed layer takes several hours to complete.
Organic monolayers can be prepared using a broad range of compound/surface
combinations. Organosulfur compounds, such as thiols and disulfides, form SAMs
with a high internal order and structural stability on noble metals, and this
combination is the most extensively studied. [1] Gold is often the substrate material
of choice due to its general inertness. [3] The terminal groups giving functionality to
the SAM can be varied widely. This results in a broad range of available
functionalities, from simple methyl groups to much more complex structures. For
3
example, poly- or oligo(ethylene glycol) surfaces, like the ones designed in this thesis,
are of interest in biosensor applications due to their excellent protein repelling
properties. [4]
Mixed monolayers of compounds with different chain lengths and/or terminal groups
provide a route for surface engineering at the molecular level. Chemical reactivity of
the SAM can be designed by controlling the types of functionalities present, as well as
their lateral separation. [5] However, it should be noted that a certain ratio between
molecules in the preparation solution not necessarily results in a SAM with that same
ratio. To gain better control over the mixed SAM composition and to minimize the
risk of phase segregation, the molecules should be as similar as possible and
preferably only differ in the tail group. [6]
2.2 Surface Plasmon Resonance
2.2.1
The principle of Surface Plasmon Resonance
Surface plasmon resonance, SPR, is an optical phenomenon that arises when light is
reflected at metallic films under specific conditions. [7] The SPR phenomenon can be
used for sensing purposes, for example to monitor biomolecular interactions. A
surface plasmon is a charge density wave that propagates along the interface
between a metal surface and the ambient medium. The wave originates from a
collective oscillation of electrons in the metal surface region. [8] Surface plasmon
resonance occurs when energy and momentum are being transferred from incident
photons to the surface plasmon. [7] A surface plasmon wave is exceptionally sensitive
to changes in refractive index near the metal surface, caused, for instance, by
adsorption of biomolecules. SPR is a surface sensitive technique and typically,
interactions within the first few hundred nanometers from the surface can be
detected. [8]
The most commonly used setup for SPR is called the Kretschmann configuration,
schematically illustrated in figure 2.2. It is based on total internal reflection in a glass
prism coated with a thin metal film. Although no propagating light beam is refracted
into the metal, a small part of the light, called the evanescent field, may penetrate
outside the glass and excite a surface plasmon wave. For this excitation to occur, the surface-parallel component of the incident light’s wave vector, and - the wave
vector of the surface plasmon, must be equal. By simply varying the incident angle
(θ), one can find the resonance condition, at which = and surface plasmon
resonance occurs. SPR is observed as a characteristic dip in the intensity of the
reflected light, as energy is transferred from the light into the surface plasmon. [7, 9]
4
Figure 2.2 Schematic illustration of the Kretschmann setup used for optical excitation of surface
plasmons. At a certain angle θ, energy is transferred from the light into the surface plasmon and
surface plasmon resonance occurs. For this to happen, the surface parallel component of the incident
light’s wave vector, kx must be equal to the wave vector of the surface plasmon, ksp.
The wave vector, , of the surface plasmon is dependent on the refractive index of
the sample medium close to the metal surface. Thus, if the refractive index is
changed, due to association or dissociation of biomolecules, is altered.
Consequently, - the angle of incidence that results in SPR, changes. For a given
experimental setup (light source, prism, metal and buffer), the angular shift is
proportional to the local change in refractive index, which in turn depends on the
mass concentration of biomolecules at the surface. During an SPR experiment, the
change in is measured as a response signal. [9]
2.2.2
Biacore systems
By employing the SPR technique, Biacore systems can monitor biomolecular
interactions in real-time, without the need of labeling. Quantitative information
about concentration, specificity, kinetics and affinity can be obtained. The
biomolecular interaction occurs on a sensor chip surface, to which one of the
interacting molecules, called the ligand, is immobilized. The sensor chip consists of a
glass surface coated with a thin layer of gold, on top of which a hydrophilic matrix is
attached. The matrix contains functional groups for covalent immobilization of ligand
molecules. An integrated microfluidic cartridge (IFC) of silicone rubber is brought in
contact with the sensor surface, whereby the sensor surface itself forms one wall of
each flow cell, as illustrated in figure 2.3. Firstly, ligand molecules are immobilized to
the surface, using appropriate coupling chemistry. Secondly, interacting partners in
solution, called analytes, are delivered to the sensor surface via the flow cell system
(see figure 2.4). Flow cells can be addressed separately or in series, depending on
application. In the A100 instrument, it is possible to access individual spots within the
5
Figure 2.3 Schematic illustration of how flow cells (1-4) are formed when an integrated microfluidic
cartridge (IFC) is brought in contact with the sensor chip surface.
same flow cell via hydrodynamic addressing. Between sample injections, a
regeneration solution is injected to dissociate remaining analyte molecules. Thus, the
sensor chip surface may be reused several times. [10] In Biacore systems, the
resonance angle, θ, is followed as function of time and presented in a so called
sensorgram (see figure 2.4) The angle is expressed in arbitrary units called RU
(resonance units) with 1 RU corresponding to a resonance angle shift of
approximately 0,0001°. Experiments have shown that a 1000 RU response
corresponds to an increase in surface concentration of 1 ng/mm2. [11]
Figure 2.4 Illustration of the setup used in Biacore instruments, with the flow cell amplified in size.
Incident light is reflected at the glass/gold interface. The intensity drops for light with a certain angle,
θ, due to SPR. Interactions between ligand molecules on the surface and analytes delivered in
solution result in a change of SPR angle (from position 1 to position 2). The shift, Δθ, is detected by a
photodiode array and presented as the output signal.
6
2.2.3 Ligand immobilization
There are a number of different coupling chemistries to enable ligand immobilization
to the sensor chip matrix. The most common route is to use amine coupling, whereby
carboxyl groups in the matrix form covalent amide bonds with primary amine groups
in proteins (ligands). However, this process does not occur spontaneously, the
carboxyl groups need to be activated. Activation is performed with a 1:1 mixture of
N-Ethyl-N’-(3-Dimethylaminopropyl) Carbodiimide (EDC) and N-Hydroxy-succinimide
(NHS). EDC reacts with the carboxyl group and forms a reactive intermediate which in
turn reacts with NHS to form an active NHS ester. As ligand is passed over the sensor
chip surface, the NHS-moiety (which is a good leaving group) reacts spontaneously
with a primary amine group in the ligand and covalent bond between ligand and
matrix is formed. The process is illustrated in figure 2.5. Most proteins contain several
primary amines, and thus, immobilization can be achieved without seriously affecting
the ligand’s biological activity.
Figure 2.5 To covalently couple ligand molecules to the surface, carboxyl groups in the matrix need
to be activated. EDC reacts with the carboxyl group and mediates the formation of an active NHS
ester which subsequently reacts with a primary amine in the ligand.
To facilitate ligand immobilization, attractive electrostatic forces are employed in a
process called pre-concentration. The ligand is dissolved in a coupling buffer with pH
below the isoelectric point (pI) of the protein, but above the pI of the matrix. Hereby,
the ligand and matrix obtain opposite net charges. Positively charged ligand
molecules are electrostatically attracted to the negative surface and a high ligand
concentration near the surface results in more efficient immobilization
After ligand immobilization, excess NHS-activated carboxyl groups are deactivated
with ethanolamine that caps residual NHS esters so that no more protein can be
immobilized to the surface matrix during analyte injection. [12]
7
2.3 Infrared spectroscopy
2.3.1
The principle of infrared spectroscopy
Infrared (IR) spectroscopy is an optical technique that can provide information about
chemical composition, functional groups and orientation. IR radiation excites
molecular vibrations, causing atoms and groups of atoms to vibrate with increased
amplitude about the covalent bonds that connect them. A certain bond vibration is
excited by a certain wavelength of IR radiation, resulting in absorption of that
wavelength. Hence, the pattern of wavelengths absorbed by a molecule is
characteristic of the types of bonds and atoms present in its structure. [13] A certain
criterion has to be fulfilled in order for a vibration to be detectable with IR
spectroscopy: the dipole moment of the molecule must change during the vibration.
This implies that asymmetric molecules containing polar groups are IR active,
whereas monoatomic and homodiatomic symmetric substances are not. [14]
Unfortunately, air contains carbon dioxide and water vapour which are IR active
compounds. In order to minimize interference during measurements, the sample is
contained either in a vacuum chamber or a chamber purged with IR inactive nitrogen
gas.
Today, Fourier transform spectrometers are the most commonly used. Instead of
probing each wavelength component sequentially, a Fourier transform spectrometer
examines all wavelengths simultaneously. This enables a rapid collection of many
sample spectra. There are three basic components in the spectrometer – an IR
radiation source, an interferometer and a detector. Radiation from the IR source
reaches a Michelson interferometer, which consists of a fixed mirror, a moving mirror
and a semi reflecting device called a beamsplitter. At the beamsplitter, part of the IR
beam is transmitted to the fixed mirror and the other part is transmitted to the
moving one. By changing the position of the moving mirror, a difference in optical
path length between the beams is introduced. After reflection at the two mirrors, the
beams are recombined and an interference pattern is generated. The resulting beam
is then passed through the sample and certain wavelengths matching the sample’s
vibrational modes are absorbed. Finally, the beam reaches the detector and an
interferogram is acquired. The interferogram is a spectrum displaying intensity versus
time within the mirror scan. A mathematical Fourier transform converts the
interferogram to the final IR spectrum, which shows intensity versus wavenumber.
[15]
8
There exists a variety of infrared spectroscopical techniques that can be used for
characterization of thin organic layers, such as SAMs. In this thesis, Infrared
Reflection Absorption Spectroscopy (IRAS) has been used.
2.3.2
Infrared Reflection Absorption Spectroscopy (IRAS)
IRAS can be used not only to investigate chemical composition, but also to extract
information about molecular orientation on a surface. It is an extremely sensitive
technique that has proven to be very useful for studies of thin layers, such as SAMs,
on metal surfaces. Central for this technique is the so called surface dipole selection
rule. It states that only those molecular vibrations giving rise to a transition dipole
moment perpendicular to the surface will yield IR absorption and be detected. [6] The
intensity (I) of a vibration is proportional to the transition dipole moment (M)
associated with the molecular vibration and the electric field (E), which for metals
during current conditions is perpendicular to the surface. This can be written as:
∝ | ∙ | = || ∙ || ∙ (Equation 1)
with being the angle between the surface normal and the transition dipole
moment. Maximum intensity is achieved when the vibration occurs perpendicular to
the surface, i.e. along the surface normal ( = 0°) whereas a vibration parallel to the
surface ( = 90°) has zero intensity and cannot be detected, as illustrated in figure
2.6. [16]
Figure 2.6 Illustration of the IRAS principle of detection. Molecular vibrations perpendicular to the
surface give maximum peak intensity. If the vibrations are tilted, intensity is lowered, and vibrations
parallel to the surface cannot be detected at all.
9
2.4 Null ellipsometry
2.4.1
Polarization of light
Polarization characterizes light by specifying the direction of its electrical field. Light
from common light sources, such as light bulbs or the Sun, is not polarized. The
electrical field at any given point is perpendicular to the direction of propagation, but
the orientation of the field in that perpendicular plane is arbitrary. However, if the
light is polarized, the orientation of the electric field vector at any given point is
known. Light can be polarized in different ways, for example elliptically, linearly or
circularly. The electrical field vector of elliptically polarized light both rotates and
changes in magnitude whereas it only changes in magnitude for linearly polarized
light and only rotates if the polarization is circular. [17]
2.4.2
The principle of Null Ellipsometry
Null ellipsometry is an optical, non-destructive technique widely used for determining
the thickness of thin films. The basic principle of ellipsometry is that the polarization
of light changes upon reflection at a surface. [6] The change of polarization in the
reflected light is affected by properties of the reflecting surface. If a thin isotropic film
is deposited on the surface, phase and amplitude of the reflected light is altered.
These changes, which are different for light polarized perpendicular (s) or parallel (p)
to the plane of incidence, are detected as changes in the ellipsometric angles ∆ and
. Provided that the refractive indices of the substrate, the ambient and the film are
known, ∆ and can be used in combination with an optical model to calculate the
film thickness. The refractive index of the film is usually unknown, though, but set to
1.5 when measuring on organic films. Null ellipsometry is an averaging method where
the result obtained is an average of the results within the beam area of
approximately 1 mm2. [16, 6]
Figure 2.7 shows the instrumental setup of a polarizer compensator sample analyser
(PCSA) ellipsometer. Monochromatic light from a He-Ne laser passes a rotatable
polarizer and a compensator before reaching the sample surface. The light coming
out of the polarizer is linearly polarized. When passing the compensator, a phase shift
between the light’s s- and p-polarized components is induced, resulting in elliptical
polarization. When the light is reflected at the sample surface, its polarization
changes, as mentioned earlier. The reflected light passes through an analyzer, which
is a second rotatable polarizer. Finally, the intensity of the light transmitted through
the analyzer is measured with a detector. [6] When measuring, the angle of the (first)
10
polarizer is adjusted so that the incident elliptically polarized light becomes linearly
polarized after reflection at the surface. The analyzer is then rotated to the single
certain angle where all linearly polarized light is quenched and no light reaches the
detector. Thus, a “null” condition is satisfied and this is why the technique is called
null ellipsometry. [18] The values of ∆ and can be determined from the angular
positions of the polarizer and analyzer. [6]
Figure 2.7 The instrumental principle of a null ellipsometer. Light from a He-Ne light source passes a
rotable polarizer and a compensator and is elliptically polarized when reaching the sample surface.
The reflected light is passed through an analyzer before reaching the detector.
2.5 Contact angle goniometry
Contact angle goniometry is a straightforward technique used to obtain information
about a surface’s energy and wettability. Such information can for example be used
to indicate the orientation of molecules in a SAM or study an adsorption process
which affects the surface energy. [19] A liquid droplet is placed on a solid surface and
the contact angle, θ , between solid and liquid is measured. Surface energy and
contact angle are related through Young’s equation:
Θ =
(Equation 2)
where ! is the free energy of solid in contact with air (vapour), " the solid/liquid
interfacial free energy and "! the free energy of liquid in contact with air. (see figure
2.8) Water is the most common liquid used for contact angle measurements. The
method is surface sensitive and approximately the outermost 5 Å of the sample
surface are analyzed. [20] Hydrophilic, high-energy surfaces are characterized by low
11
contact angles as the water droplet wets (spreads over) the surface. Conversely, high
contact angles are encountered when measuring on hydrophobic, low-energy
low
surfaces. When measuring, a water droplet is placed on the sample surface and
observed through a horizontal camera arrangement coupled to a computer software
program. The intersection between solid, liquid and vapour is defined manually or
automatically and the contact angle θ is measured. [19]
Figure 2.8 The droplet profile depends on the properties of the underlying surface. The contact
angle θ is determined by the relation between the interfacial energies of solid/vapour (SV),
solid/liquid (SL) and liquid/vapour (LV).
Besides from recording the static contact angle, dynamic measurements are often
conducted. The value of the contact angle varies, depending on whether the liquid is
advancing over a dry surface or receding from a wetted one, and two different angles
are noted during dynamic measurements. These
These are called the advancing contact
angle θa and the receding contact angle θr. The hysteresis between θa and θr is
influenced by factors such as surface roughness and heterogenecity and provides
further information regarding the surface. [6]
2.6 X-ray Photoelectron
otoelectron Spectroscopy (XPS)
X-ray
ray Photoelectron Spectroscopy is a powerful surface analysis technique that can
provide information about elemental composition,, orientation and the atoms’
chemical environment in a sample.
sample The technique is based
ed on the photoelectric
effect, whereby photons can induce emission of electrons from
m a solid. Provided
P
that
the photon energy, hv, is greater
gr
than the energy (Eb + #) that attracts the electrons
to the nucleus,, emission occurs.
occurs The remaining excess energyy is transformed into
kinetic energy, Ek, of the photoemitted electrons,
electrons, according to equation 3 below.
$ = %& ' () * #)
12
(Equation 3)
# is the work function, the minimum energy required to remove an electron from the
highest occupied energy level (the Fermi level) to the vacuum level. Eb is the binding
energy of the electron, defined as the difference in potential energy between the
electron’s actual level and the Fermi level.
In XPS, a beam of monochromatic X-rays induces electron emission from both
valence and core levels of the atoms in the sample. The key to chemical identification
is that core electrons, that do not take part in chemical bonding, are largely
unaffected by their surroundings and retain binding energies Eb that are signatures of
the atom type. During an XPS-experiment, X-rays with fixed photon energy are
directed on the sample and the kinetic energies of emitted electrons are measured.
Eb values are then calculated, thus enabling identification of the elements present in
the sample.
Although an atom’s core level electrons do not take part in chemical bonding, their
exact binding energies are affected by the species to which the atom is bonded.
Charge transfer between atoms of different electronegativity may alter the columbic
attraction between core electrons and their nucleus. Small shifts in binding energies,
called chemical shifts, can be detected, that provide information about the atom’s
local chemical environment.
In the XPS instrument, kinetic energies are measured using an electrostatic energy
analyzer that consists of two concentric hemispheres. Between the hemispheres, a
potential difference is applied that gives rise to an electric field. Electrons that enter
the analyzer are deflected by the electric field. Only electrons of a certain chosen
kinetic energy, the pass energy, reach the detector. Electrons with higher or lower
energies are not deflected to the right extent and either hit the outer or the inner
hemisphere. Before entering the analyzer, electrons are retarded to by a negative
electrode – the retard plate. By changing the negative voltage on the retard plate,
electrons with different kinetic energies are retarded to the pass energy and allowed
through to the detector, and a spectrum can be obtained. [21]
13
14
3 Experimental section
3.1 Materials
3.1.1
Substrates
Glass slides coated with gold were obtained from GE Healthcare, Uppsala, Sweden.
The thickness of the evaporated gold film varied, depending on application. 2000 Å
was used for IRAS, ellipsometry and contact angle goniometry, whereas a thinner film
of 440 Å was used for SPR measurements.
3.1.2
Thiol compounds
Four different oligo(ethylene glycol) (OEG) derivatized alkanethiol compounds, shown
schematically in figure 3.1 below, were purchased. SAMs of each separate thiol
compound were studied, as well as mixed SAMs of monothiols and dithiols,
respectively. Bulk products were dissolved in 99,5 % ethanol to 10 mM stock
solutions, which were stored frozen until use.
Figure 3.1 The different alkanethiol compounds used in this thesis. To the left: OH-terminated and
COOH-terminated monothiols. To the right: OH-terminated and COOH-terminated dithiols.
3.1.3 Reagents
Milli-Q water with a resistivity of 18.2 MΩ∙cm was obtained from a Millipore system
(Milli-Q Academic) with a 0.22 µm filter. All other reagents used at Linköping
University and GE Healthcare are listed in Appendix A.
15
3.2 SAM preparation
Prior to monolayer assembly, all gold surfaces, were cleaned in a 5:1:1 mixture of
Milli-Q water, 30% hydrogen peroxide and 25% ammonia (TL-1 wash) at 85°C for 10
minutes, to remove organic contamination. Tweezers used to handle the surfaces
were washed according to the same procedure. After washing, surfaces were rinsed
thoroughly in Milli-Q water and one surface was examined with ellipsometry. This
was done in order to verify cleanliness and to obtain reference Δ and values for the
bare gold substrate, necessary for subsequent calculations of SAM thickness.
Ellipsometric angles Δ> 110° were considered to indicate a satisfactory cleaning
procedure. Surfaces were then rinsed in ethanol and immersed in a 50 µM or 500 µM
ethanolic thiol solution. Incubations were carried out for at least 20h. Shortly before
analysis, the surfaces were rinsed and ultrasonicated in ethanol for 5 min, rinsed
again and blown dry with N2 gas. When establishing a method for SAM preparation,
the influence of incubation time on SAM thickness was evaluated. Surfaces were
incubated for 1 day or 3 days, and the resulting SAMs were characterized with
ellipsometry. No significant difference in thickness could be detected for the different
incubation times, and the shorter time was chosen for practical reasons.
In this thesis, mixed SAMs, containing both OH- and COOH- terminated molecules
were prepared. Schematic illustrations are shown in figures 3.2 and 3.3. COOHterminated compounds provided active functional groups for immobilization of
biomolecules to the surface, as described in section 2.3. OH-terminated compounds
served as background filling molecules, controlling the surface density of functional
groups. Henceforward, OH:COOH ratios (shown in parenthesis) will be used to denote
each SAM type. Initially, 5 mixtures were studied for both mono- and dithiols:
100% OH, 0% COOH
95% OH, 5% COOH
90% OH, 10% COOH
70% OH, 30% COOH
0% OH, 100% COOH
(100:0)
(95:5)
(90:10)
(70:30)
(0:100)
Following basic characterization, a pilot SPR study of ligand immobilization capacity
and non-specific binding was performed at IFM, Linköping. On basis of the results,
promising mixtures were chosen for studies at GE Healthcare, Uppsala and a few new
mixtures were introduced, as shown below.
16
Monothiols:
100% OH, 0% COOH
90% OH, 10% COOH
70% OH, 30% COOH
50% OH, 50% COOH
30% OH, 70% COOH
0% OH, 100% COOH
(100:0)
(90:10)
(70:30)
(50:50)
(30:70)
(0:100)
Figure 3.2 Schematic illustration of a
monothiol SAM. Not to scale.
Dithiols:
100% OH, 0% COOH
95% OH, 5% COOH
90% OH, 10% COOH
70% OH, 30% COOH
50% OH, 50% COOH
(100:0)
(95:5)
(90:10)
(70:30)
(50:50)
0% OH, 100% COOH (0:100)
Figure 3.3 Schematic illustration of a
dithiol SAM. Not to scale.
3.3 Characterization of SAM quality and structure
To evaluate the quality of formed SAMs, different analytical techniques were used.
This part of the thesis work was performed at Linköping University.
3.3.1
Null ellipsometry
Ellipsometric measurements of SAM thicknesses were performed on a Rudolph
Research AutoEL ellipsometer with a He-Ne laser light source, λ= 632.8 nm and an
incidence angle of 70°. To calculate SAM thickness, a three phase model (ambient –
organic film – gold) was used. The ellipsometric parameters Δ and of the bare gold
substrate (obtained previously) were entered as setvalues and the refractive index of
the organic film was assumed to be 1.5. This is a common assumption, preprogrammed in the instrument’s automatic measurement file. Each sample surface
was measured at five detection spots and the average SAM thickness was calculated
by the instrument, using the McCrackin algorithm. [22]
17
3.3.2
Contact angle goniometry
Contact angle goniometry was used to determine the hydrophilicity of the different
SAMs. Measurements were performed at room temperature with a CAM200 Optical
Contact Angle Meter from KSV Instruments Ltd. Before measuring, the syringe needle
was TL-1 washed and the syringe was rinsed and filled with Milli-Q water. Advancing,
receding and static contact angle values were recorded on three sites of each sample
surface.
Firstly, the advancing angle was measured. A small water droplet was placed on the
surface and the amount of water was constantly increased, causing the droplet to
expand and advance over the dry surface surrounding it, as illustrated to the left in
figure 3.4. Pictures of the droplet profile were taken during the advancing phase by a
horizontally mounted camera and 10 frames were recorded with a frame interval of 1
sec. Contact angles were then determined by a numerical curve fit of the droplet
profile, using the instrument software. An average advancing contact angle value
was calculated for the current site, based on the 10 measurements. Secondly, the
receding angle was determined. In this case, water was withdrawn from the droplet
(the same droplet as above), causing it to recede from the wetted surface, see figure
3.4 to the right. Multiple frames of the droplet were recorded and processed as
described above. Finally, the static contact angle was measured. One frame was
recorded and the contact angle was determined by curve fitting, as previously
described.
Figure 3.4 Schematic illustration of the measurement procedure for advancing (left) and receding
(right) contact angles. When measuring advancing angle, the amount of water in the droplet is
constantly increased. Conversely, water is withdrawn to obtain the receding angle.
3.3.3 XPS
XPS measurements were performed to examine the chemical state of dithiol sulfur
atoms. Measurements were performed on a VG Scientific instrument with a twin
Mg/Al X-ray source. In this thesis, the Al anode was used, producing X-rays with a
18
photon energy of 1486.6 eV. The pressure in the analysis chamber was 3.1 ∗ 10/
mbar and the pass energy was set at 50 eV. First, a widescan, showing all peaks
between 1-1300 eV, was recorded. Then, a sulfur scan (158-172 eV) was performed.
Curve fitting of the obtained sample peaks was done with the XPSPEAK software.
3.3.4
IRAS
Infrared Reflection Absorption Spectroscopy was used to study orientation and
chemical composition of the SAMs. Spectra were recorded on a Bruker IFS 66 system,
equipped with a liquid nitrogen cooled mercury cadmium telluride (MCT) detector.
Spectra were recorded in the range of 5000 cm-1 – 400 cm-1 with 2 cm-1 resolution
and 3000 scans. Prior to Fourier transformation, a three-term Blackman Harris
apodization function was applied to the interferogram, in order to zero-value the
signal outside the wavenumber interval of interest.
The measurement chamber was continuously purged with nitrogen gas to minimize
interference from water and carbon dioxide. A deuterated hexadecanethiol
(HS(CD2)15CD3) SAM on gold was used to record a reference spectra. Also, a water
spectrum was recorded, by letting air into the measurement chamber.
After inserting a sample surface, the chamber was closed and purged with N2 for 30
minutes before the actual measurement started. Background and water was
subtracted from each sample spectrum, using the OPUS instrument software.
19
3.4 Characterization of sensor chip performance in Biacore
After the first basic characterization of the different SAM surfaces, their performance
as sensor chips was studied with the Biacore instrument. New SAM surfaces were
prepared on site in Uppsala and evaluated according to the scheme in figure 3.5
below. All OH:COOH mixtures, for both monothiols and dithiols, were analyzed
according to “Chip 1” and “Chip 2”. For those OH:COOH combinations showing the
most promising characteristics, additional surfaces were prepared and examined
according to “Chip 3”, “Chip 4” and “Chip 5”.
SAM preparation
Chip 1
Chip 2
Chip 3
Surface stability
Non-specific
binding - plasma
Ligand stability
Immobilization
capacity and
analyte binding
Non-specific
binding - proteins
Kinetic assay
performance
Chip 4
Detergents
Chip 5
Storage
stability
Non-specific
binding - plasma
Figure 3.5 Scheme describing the test procedure when evaluating sensor chip performance. All
different SAM mixtures were evaluated according to “Chip 1” and “Chip 2”. For those mixtures
showing the best characteristics, additional surfaces were prepared and analyzed according to “Chip
3”, “Chip4” and “Chip 5”.
All experiments were performed on a Biacore3000 instrument, equipped with a liquid
degasser and an in-house built chip exchanger that enables the analysis of several
surfaces without manual chip exchange. HBS-EP was the running buffer of choice for
all experiments except for “Detergents” when a surfactant-free HBS-N buffer was
20
used. After docking a new surface in the instrument, the resonance signal was
normalized with BIAnormalizing solution (70% glycerol), to compensate for small
differences between individual sensor chips. A flat surface, equipped with another
type of two-dimensional SAM matrix, was used as a reference in all experiments.
Activation reagents EDC and NHS were diluted according to instructions in the
Biacore amine coupling kit and dispensed in aliquots that were frozen until use.
Specific information regarding injection times and flow rates will not be reported in
this chapter. The interested reader is referred to Appendix B for additional
information.
3.4.1 Surface stability
To examine surface stability, the sensor chip surface was subjected to regeneration
pulses at high flow (100 µl/min). Regeneration is normally the process whereby
bound analyte molecules are removed from the sensor chip so that it can be reused.
However, in this case, the procedure gave information about stability of the SAM
surface, i.e. if the thiol molecules were firmly attached so that the SAM matrix
remained stable at harsh conditions. If not, there would be a substantial baseline drift
in the sensorgram. Two types of regeneration solutions were used - 10 mM glycineHCl, pH 1.5 (low pH) in flow cells 1 and 2 and 50 mM NaOH (high pH) in flow cells 3
and 4. 10 injections of regeneration solution were performed in each of the four flow
cells on the surface. After each pulse, absolute baseline values were recorded, to
enable monitoring of a potential baseline drift.
3.4.2 Immobilization capacity and analyte binding
After the surface stability test, ligand molecules were immobilized to the sensor chip.
Flow cells 2 and 4 were used to study specific ligand immobilization and subsequent
analyte binding. First, the SAM surface was activated with EDC/NHS. Then, ligand was
injected, followed by deactivation with ethanolamine. After two regeneration pulses
with 10 mM glycine pH 1.7, the amount of immobilized ligand was recorded.
Thereafter, analyte solution was injected to examine ligand activity and the amount
of bound analyte was recorded.
Dithiol surfaces: Anti-myoglobin (~150 kDa) served as ligand and was diluted
from stock solution to 50 µg/ml in sodium acetate buffer pH 5.0, which made
the protein positively charged to facilitate pre-concentration. Myoglobin (~17
21
kDa), the analyte, was diluted to 5 µg/ml in HBS-EP running buffer. This ligandanalyte system was used in both flow cells (FC 2 and FC 4).
Monothiol surfaces: The ligand-analyte system described above was used in FC
2. In FC 4, the ligand-analyte system was reversed so that myoglobin served as
ligand and anti-myoglobin as analyte.
3.4.3
Non-specific binding – plasma
Components in complex sample fluids, such as plasma, often bind non-specifically to
the sensor chip surface. A high degree of such binding complicates the analysis of
specific binding events. This test was performed in the unused flow cells 1 and 3 on
Chip 1 and in FC 4 on Chip 2. Frozen human plasma from two different donors, 1953
and 1955, was brought to room temperature and centrifuged at 4000 rpm for 20
minutes. The supernatants were filtered through 1 µm + 0.2 µm disposable syringe
filters, coupled in series on the same syringe. Two regeneration pulses with 50 mM
NaOH were run in each flow cell prior to plasma injection. On Chip 1, plasma 1953
was injected in one flow cell and plasma 1955 in the other. On chip 2, only plasma
1953 was used, since injection was only performed in one flow cell. Following plasma
injection, the surface was regenerated with two pulses of NaOH. The amount of nonspecific binding to the surface was recorded both before and after the regeneration
step.
3.4.4 Non-specific binding – proteins
Non-specific binding of proteins to the sensor chip was evaluated in flow cells 1, 2
and 3 on Chip 2. Stock solutions of Protein A, which represented a “normal” protein,
and HSA, which represented a “sticky” protein, were prepared and diluted to 100
µg/ml in sodium acetate buffer pH 5.0. Hence, a positive net charge was applied to
Protein A and HSA. Also, non-specific binding from thyroglobulin diluted in the
neutral running buffer (pH 7.4) was studied.
The same procedure as for plasma was used when studying non-specific binding of
protein A, HSA and thyroglobulin, i.e. the sensor chip surface was subjected to two
regeneration pulses, protein injection and two additional regeneration pulses. The
amount of non-specific binding to the surface was recorded both before and after the
last regeneration step.
22
3.4.5
Ligand stability
A criterion for good sensor chip performance is that the ligand remains bound to the
surface during multiple analyte/regeneration cycles. The ligand stability test was
designed to investigate if ligand molecules dissociated from the surface during
repeated regeneration. This would indicate that some ligands were attached noncovalently to the SAM matrix. Anti-myoglobin was immobilized using two slightly
different procedures:
FC 1: Maximum level immobilization: The surface was activated with EDC/NHS
and 50 µg/ml anti-myoglobin in Acetate pH 5.0 was injected. After ligand
injection, the surface was deactivated with ethanolamine. This is the “ordinary”
immobilization procedure used in previous tests, resulting in a high level of
immobilized anti-myoglobin to mono- and dithiol surfaces. (~3000 RU).
FC 2: “Aim for ligand level”-immobilization: The Biacore3000 instrument has a
software wizard that performs immobilization, aiming for a chosen target ligand
level. [23] The target level is specified in RU by the user. To study if the amount
of immobilized ligand affects ligand stability, the wizard was used to immobilize
~1300 RU of anti-myoglobin to the surface in FC 2. This lower ligand level is in
comparison with the level achieved on the reference chip.
Following ligand immobilization, the sensor chip surface was subjected to 20
regeneration pulses of 50 mM NaOH at high flow (100 µl/min). After each pulse,
absolute baseline values were recorded, to enable monitoring of a potential baseline
drift due to ligand dissociation.
3.4.6
Kinetic assay performance
The Biacore instrument is widely used for measuring kinetics and affinities of binding
interactions. Therefore, it was relevant to study the performance of mono- and
dithiol SAM surfaces with a kinetic assay. A well known model system from the
Biacore Getting Started Kit was used, with Anti-β2-microglobulin as ligand and β2microglobulin as analyte. The surface was activated with EDC/NHS, Anti-β2microglobulin diluted to 40 µg/ml in Acetate pH 5.0 was injected, followed by
deactivation with ethanolamine. Injection of analyte was performed using the
instrument’s kinetic analysis wizard. [23] A concentration series of 0 nM, 2 nM, 4 nM,
8 nM, 16 nM, 32 nM and 64 nM β2-microglobulin in HBS-EP was prepared. Analyte
was injected over the surface from low to high concentration and dissociation was
monitored for 5 minutes after completed injection. Between each injection, the
23
surface was regenerated with glycine pH 2.5. The resulting sensorgrams were
evaluated with the BIAevaluation software, and kinetic constants were derived.
3.4.7 Detergents
Detergents are often added to the running buffer in Biacore experiments to prevent
proteins from adsorbing to tubings and IFC. Detergents contain both a hydrophilic
and a hydrophobic segment. As they attach to the walls of the flow system via their
hydrophobic tails, their hydrophilic parts are exposed outwards, towards the
solution. This hydrophilic coating of the flow system prevents protein adsorption.
However, detergents that bind to and dissociate from the sensor chip surface is a
potential source of noise. In this experiment, association and dissociation of three
common non-ionic detergents was studied. All experiments were performed with a
detergent free HBS-N buffer. The detergents of choice - Brij 35, Polyoxyethylene (20)
sorbitan monolaurate (P20) and Pluronic F-127 – were dissolved in HBS-N to 0,05%. A
similar but more extensive study of detergent binding has been performed by Essö
[24]. The detergents were sequentially injected in separate flow cells. Injection was
performed during 10 minutes, followed by 13 minutes dissociation time. Thereafter,
three pulses of HBS-N were injected, to speed up the dissociation process. Finally, the
system (tubing, needle etc.) was thoroughly washed with SDS and glycine to remove
bound molecules before injection of the next detergent.
3.4.8
Storage stability
For a sensor surface to be interesting in commercial applications, it has to be able to
stand the strain of storage. Fridge storage is normally recommended for Biacore
sensor surfaces, as this prolongs their shelf-life, but one has to take into account that
the surface may be subjected to elevated temperatures during transport. To study
the effect of such an event, mono- and dithiol surfaces with 50% OH and 50% COOH
were stored at elevated temperatures and sensor chip performance after storage was
evaluated. Moreover, this test also served as an accelerated simulation of long-time
storage at lower temperatures. Half the surfaces were stabilized with proprietary
stabilization solution, and the other half were stored without any pre-treatment. The
surfaces were blown dry with N2, mounted on chip carriers and packed individually
under N2 in sealed foil bags. After 6 days storage at 25 °C or 40 °C, immobilization
capacity, surface stability, ligand stability and non-specific binding of plasma was
studied. FC 2 was activated with EDC/NHS and anti-myoglobin (50 µg/ml in Acetate
pH 5.0) was immobilized, followed by deactivation with ethanolamine. Thereafter, 10
24
regeneration pulses of NaOH were injected in both FC 1 (empty) and FC 2 (with
ligand), thus providing information on both surface stability and ligand stability.
Finally, plasma was injected in FC 4 according to the procedure described above in
section 3.4.3.
25
26
4 Results and discussion
4.1 Characterization of SAM quality and structure
During the basic characterization, two different incubation solution concentrations
were used – 500 µM and 50
50 µM, to investigate the influence of solution
concentration on SAM quality.
4.1.1
Null ellipsometry
The thickness of formed SAMs was measured with ellipsometry and results are shown
below in figure 4.1.
Monothiols 500 µM
30
25
20
15
10
22,5
24,3
24,6
25,7
29,6
5
30
25
20
15
10
25,9
27,3
28,3
29,5
31,7
0
100:0
95:5
90:10 70:30 0:100
100:0 90:10 70:30 50:50 30:70 0:100
OH : COOH ratio (%)
OH : COOH ratio (%)
Dithiols 500 µM
35
Dithiols 50 µM
35
30
25
20
15
16,4
16,9
17,6
18,4
22,9
5
SAM thickness (Å)
SAM thickness (Å)
24,2
5
0
10
Monothiols 50 µM
35
SAM thickness (Å)
SAM thickness (Å)
35
30
25
20
15
10
17,3
16,9
17,2
18,5
19,8
22,1
5
0
0
100:0 95:5 90:10 70:30 0:100
100:0 95:5 90:10 70:30 50:50 0:100
OH : COOH ratio (%)
OH : COOH ratio (%)
Figure 4.1 Thicknesses of the different SAMs, measured with null ellipsometry. Measurments were
performed on at least five surfaces of each SAM type, and each sample surface was measured in five
detection spots. Error bars represent 95 % confidence intervals.
27
The trend that a higher COOH content in the incubation solution results in a thicker
SAM could be observed for both mono- and dithiol surfaces (although the differences
were not statistically significant between all adjacent mixtures). This is according to
expectations, as COOH-terminated thiol compounds contain more ethylene glycol
units and are longer than their corresponding OH-terminated compound. However,
the results are not as obvious as they may seem. As mentioned previously in the
theory section, a certain ratio between molecules in the preparation solution do not
necessarily result in a SAM with that same ratio, as one type of molecule may be
favoured over the other during the adsorption process. In this case, though, the
ellipsometric results indicate that as the COOH concentration in solution increased,
more COOH-terminated compounds were actually incorporated in the SAM.
For monothiol surfaces, measured SAM thicknesses correspond well to theoretical
estimations, based on the length of the two compounds. For dithiol surfaces, the
SAMs are shorter than theoretically expected. This is possibly explained by the shape
of the dithiols. Assuming that the thiols are attached to the surface by both sulfur
atoms, they are considerably wider at the base than at the top. This implies that
while the molecules are close packed at the bottom, there is still considerable space
available for each OEG-chain on top to adapt a conformation less upright than the
one demonstrated in figure 3.3. If the OEG-chains are tilted or coiled, the SAM will
consequently be shorter than theoretically expected.
4.1.2
Contact angle goniometry
Results from contact angle measurements are presented in figure 4.2. Contact angles
were higher on SAMs containing 100 % OH than on those containing 100 % COOH.
This might be due to the fact that the carboxyl group, with its two oxygen atoms,
contains more sites for hydrogen bonding than OH, and is thus more hydrophilic.
The trend that a higher COOH content results in a more hydrophilic SAM was clearly
observed 50 µM surfaces, but hardly distinguishable for the intermediate mixtures
on 500 µM surfaces. Generally, the lower concentration seems to result in more
densely packed and oriented, and thus more well defined SAMs. Several studies, for
example by Valiokas et al. [25], support the hypothesis that a low concentration in the
incubation solution is beneficial for SAM quality.
Normally, ethylene glycol terminated surfaces display contact angles around 30°. The
higher angles observed in this study, especially on dithiol surfaces at high
concentrations, suggest that some chemical moiety other than OEG is exposed
28
towards the surface and affect its wetting properties. One possible explanation would
be that some dithiols are not attached with both
both sulfur atoms to the surface, but
standing on one “leg” only. The other alkane chain may thus be exposed towards the
exterior of the SAM. This hypothesis was investigated further with XPS, see section
4.1.3.
The rather large hysteresis between advancing and receding contact angles indicate
that the thiol SAMs are not homogenous at the molecular level.
Contact angle (°)
Monothiols 500 µM
Monothiols 50 µM
50
45
40
35
30
25
20
15
10
5
0
50
45
40
35
30
25
20
15
10
5
0
0:100
95:5 90:10 70:30
OH : COOH ratio (%)
advancing receding
0:100
0:100 90:10 70:30 50:50 30:70 0:100
OH : COOH ratio (%)
advancing
receding
Contact angle (°)
Dithiols 500 µM
µ
Dithiols 50 µM
50
45
40
35
30
25
20
15
10
5
0
50
45
40
35
30
25
20
15
10
5
0
0:100
95:5 90:10 70:30
OH : COOH ratio (%)
advancing receding
0:100
0:100 95:5 90:10 70:30 50:50 0:100
OH : COOH ratio (%)
advancing
receding
Figure 4.2 Advancing and receding contact angles measured on monomono and dithiol surfaces. The
trend
rend that a higher COOH content results in more hydrophilic SAMs was clearly distinguishable on 50
µM surfaces, but not as obvious on SAMs formed from 500 µM incubation solutions.
29
4.1.3
XPS
XPS measurements were performed to examine the chemical state of dithiol sulfur
atoms. An XPS study of monothiol compounds has been conducted previously by
Valiokas [25], where monothiols similar to those used in this thesis were
demonstrated to form highly organized SAMs on gold, with a low or non-existing
degree of unbound sulfur, depending on incubation concentration. In this
experiment, “pure” OH and COOH SAMs as well as 50:50 mixtures were prepared
using 500 µM and 50 µM incubation solutions. A typical curve is presented in figure
4.3. Due to spin-orbit coupling, two peaks are visible for each chemical state. The
highest peak in each pair is assigned to 2p3/2 and the lowest to 2p1/2. The black peak
pair with 2p3/2 situated at 162 eV corresponds to sulfur bound to the gold surface (SAu). The grey peaks at 164 eV correspond to sulfur bound to hydrogen (S-H), i.e. not
bound to the surface. The percentage of unbound sulfur (S-H) on the different
surfaces is presented in table 4.1. When studying the results, it is clear that a
significant amount of thiol sulfur atoms are not bound to the gold surface. Since the
SAM surfaces are thoroughly washed and sonicated after incubation, most noncovalently attached SAM molecules are expected to be removed, and it is hence not
likely that they alone are responsible for the S-H peak. More likely, some dithiol
molecules are not attached with both their “legs” (sulfur atoms) pinned to the
surface, as suspected previously. If the dithiol is attached by one S-Au bond only, the
sulfur atom at the end of the other alkane chain can still remain bound to hydrogen.
Figure 4.3 A typical sulfur XPS-curve. Raw data is shown as dots and the fitted curve in bold black.
The black doublet peak represents S-Au and the grey S-H (i.e. sulfur atoms that are not surface
bound).
30
Surface
100 % OH, 500 µM
100 % OH, 50 µM
50% OH, 50% COOH, 500 µM
50% OH, 50% COOH, 50 µM
100 % COOH, 500 µM
100 % COOH, 50 µM
% unbound S
26
23
32
25
35
29
Table 4.1 Percentage of unbound sulfur (S-H) in the different dithiol SAMs, in relation to the total S
content.
According to Spangler [26], an advantageous property with dithiols is that they adsorb
faster than monothiols to gold. However, this fast adsorbtion behavior may actually
be a disadvantage if the thiols are rapidly and firmly attached in a less favourable
configuration and do not undergo structural rearrangement.
A visible trend for all SAM types was that the higher concentration resulted in more
unbound thiol molecules. This is in accordance with the monothiol results presented
by Valiokas et al. [25] and supports the hypothesis that a high concentration of thiols
in the incubation solution results in SAMs with lower quality.
4.1.4
IRAS
All SAMs displayed distinctive peaks in the region 3000 - 2800 cm-1, assigned to the
asymmetric and symmetric CH stretching of alkyl chains. Mono- and dithiol SAMs
display different peak patterns in the fingerprint region 1800 - 900 cm-1, as shown in
figures 4.4 and 4.5 where spectra from single component SAMs, formed from OHand COOH-terminated compounds, are displayed. The fingerprint region for COOHterminated monothiol SAMs contains features typical for ethylene glycol chains in
helical phase (as opposed to the less stable all-trans configuration) and the peaks
correspond very well to those reported by Valiokas. [25] Vibration mode assignments
are shown in table 4.2. The transition dipole moments of these vibrational modes are
all oriented along the helical axis of PEG and their appearance in the spectra confirms
a dominating orientation of axises along the surface normal, i.e. the PEG-chains are
mainly upright. It is obvious that the helical OEG phase is dominating for the COOHterminated SAM whereas the OH-terminated SAM is more disordered (contains a
mixture of helical and all-trans phases).
31
Figure 4.4 IRAS spectra showing the fingerprint region for monothiol SAMs. SAM with 100% OHterminated thiols is marked in red whereas 100% COOH is shown in black.
Peak position (cm-1)
1764
1464
1349
1243
1130
1115
964
Assigned to
COOH, only seen in SAMs containing carboxylated thiols
CH2 scissoring, OEG chain
CH2 wagging, OEG chain, helical phase
CH2 twisting, OEG chain, helical phase
Skeletal C-O-C stretching, OEG chain, disordered phase
Skeletal C-O-C stretching, OEG chain, helical phase
CH2 rocking, OEG chain, helical phase
Table 4.2 Assignment of vibrational modes to the peaks seen in monothiol spectra.
Dithiol compounds are more complex than monotiols and hence, their spectra
contain more peaks. Vibrational mode assignments of the peaks shown in figure 4.5
are presented in table 4.3. The fact that the OEG peaks have moved and not
correspond as well to upright helical conformation indicates that the chains are tilted
or mainly in all-trans configuration, as suggested earlier by the ellipsometric results
presented in section 4.1.1. It is known [27] that the conformation of end-tethered
chains depends on spacing.
32
Figure 4.5 IRAS spectra showing the fingerprint region for dithiol SAMs. SAM with 100% OHterminated thiols is marked in red whereas 100% COOH is shown in black.
Peak position (cm-1)
1764
1609, 1598
1460
1350
1295
1151
Assigned to
COOH, only seen in SAMs containing carboxylated thiols
C-C stretching, aromatic ring
CH2 scissoring, OEG chain
Probably from the aromatic moiety
Probably from the aromatic moiety
Skeletal C-O-C stretching, OEG chain, all trans phase
Table 4.3 Assignment of vibrational modes to the peaks seen in dithiol spectra.
A large spacing between molecules, which can be expected in dithiol SAMs due to the
shape of the thiols, brings about a greater probability for randomly coiled chains,
whereas the end-chains are more extended when the spacing is smaller, as for
monothiols.
At an early stage, peak features for COOH-containing SAMs suggested that the
carboxylic group might be present both as a carboxylic acid (COOH) and as a
carboxylate group (COO-). To investigate this, surfaces were examined both before
and after immersion in a HCl solution, where the low pH ensures that all COO- groups
are converted to COOH. After HCl treatment, the peak corresponding to the
33
carboxylate ion disappeared while the carboxylic acid peak increased in size, thus
confirming the hypothesis. The fact that the carboxyl group existed in two states was
not expected to have any negative influence on the surfaces’ performance in Biacore
applications. However, to avoid multiple peaks and to facilitate the interpretation of
IRAS spectra, surfaces were treated with HCl before analysis.
Mixed SAMs displayed peak features intermediate of those seen for the
corresponding pure OH and COOH terminated SAMs. No clear differences could be
seen between spectra for SAMs formed from 50 µM and 500 µM incubation
solutions. When studying the COOH peak at 1764 cm-1, it was clear that an increased
content of carboxylated thiols in the incubation solution lead to a stronger peak, as
shown for dithiols 50 µM in figure 4.6. This applied for both mono- and dithiol SAMs
and the result, together with ellipsometric results, clearly proved that as the COOH
concentration in solution increased, more COOH-terminated compounds were
actually incorporated in the SAM.
Figure 4.6 IRAS spectra showing how an increased concentration of COOH-terminated dithiols
in the incubation solution (increasing from bottom to top) leads to an increased COOH-peak
(marked by an arrow).
34
4.2 Characterization of sensor chip performance in Biacore
Based on the results presented in section 4.1, where a low concentration in the
incubation solution was beneficial for SAM quality, only 50 µM SAMs were used in
the Biacore characterization. To facilitate comparison between results obtained on
mono- and dithiol surfaces, a blank 95:5 position has been introduced in monothiol
diagrams and a blank 30:70 position for dithiols. It should be emphasized that most
results presented in this section are based on single measurements, with one surface
per SAM type and test. Hence, conclusions are based on trends rather than statistics.
4.2.1
Surface stability
To examine surface stability, the sensor chip surfaces were subjected to 10
regeneration pulses at high flow. After each pulse, absolute baseline values were
recorded. Surface stability results are shown in figures 4.7 and 4.8, with the different
mixed SAMs denoted by their OH:COOH content. For monothiol surfaces, mixtures
30:70 and 0:100 showed notably lower stability than the other OH:COOH mixtures
when subjected to NaOH.
10 mM glycine, monothiol SAMs
Accumulated baseline response (RU)
50 mM NaOH, monothiol SAMs
25
20
15
10
5
0
-5
-10
-15
-20
-25
0
-20
-40
-60
-80
-100
-120
-140
-160
0
1
2
3 4 5 6 7 8
Regeneration pulse
100:0
90:10
70:30
30:70
0:100
Ref.
9 10
0
1
2
3
4
5
6
7
8
9 10
Regeneration pulse
50:50
100:0
90:10
70:30
30:70
0:100
Ref.
50:50
Figure 4.7 Illustration of surface stability during regeneration with NaOH or glycine pH 1.5. Accumulated
baseline responses are shown. For monothiols, mixtures 30:70 and 0:100 were less stable than the others,
when subjected to glycine. In the diagram to the left, the topmost curve contains curves for mixtures
100:0, 70:30 and 50:50, superimposed on each other.
35
10 mM glycine, dithiol SAMs
Accumulated baseline response (RU)
50 mM NaOH, dithiol SAMs
0
-5
-10
-15
-20
-25
-30
-35
-40
-45
-50
40
35
30
25
20
15
10
5
0
-5
-10
0
1
2
3 4 5 6 7 8
Regeneration pulse
100:0
95:5
90:10
50:50
0:100
Ref.
9 10
70:30
0
1
2
3 4 5 6 7 8
Regeneration pulse
100:0
95:5
90:10
50:50
0:100
Ref.
9 10
70:30
Figure 4.8 Illustration of surface stability during regeneration with NaOH or glycine pH 1.5.
Accumulated baseline responses are shown. All ditihiol surfaces display excellent surface stability
4.2.2 Immobilization capacity and analyte binding
Immobilization capacity of ligand and subsequent analyte binding was investigated on
the different thiol surfaces. When regarding the results in figure 4.9, a common
feature for both mono- and dithiol surfaces is that the maximum ligand
immobilization capacity is very high, more than 4 times higher than on the reference
chip. Immobilization levels of ~4000 RU are in close correspondence with the
theoretical maximum of anti-myoglobin molecules that can be fitted on a surface of
the sensor chip size. (i.e. a full monolayer) According to expectations, the
immobilization capacity increases with increasing SAM COOH content, as more sites
for immobilization are introduced. However, for dithiols, immobilization capacity is
slightly lower on the 0:100 surface than on 50:50. This is most likely due to steric
reasons, as the 50:50 surface is already crowded with ligand. Immobilized antimyoglobin molecules are biologically active and accessible for myoglobin binding and
the more ligand present on the surface, the more analyte binds.
36
5000
4500
4000
3500
3000
2500
2000
1500
1000
500
0
300
250
200
150
100
Myoglobin (RU)
Anti-myoglobin (RU)
Anti-myoglobin/myoglobin
myoglobin/myoglobin - monothiol SAMs
50
0
100:0
95:5
90:10
70:30
50:50
30:70
0:100
Ref.
OH : COOH ratio (%)
ligand - anti-myoglobin
analyte - myoglobin
5000
4500
4000
3500
3000
2500
2000
1500
1000
500
0
300
250
200
150
100
Myoglobin (RU)
Anti-myoglobin (RU)
Anti-myoglobin/myoglobin
myoglobin/myoglobin - dithiol SAMs
50
0
100:0
95:5
90:10
70:30
50:50
30:70
0:100
Ref.
OH : COOH ratio (%)
ligand - anti-myoglobin
analyte - myoglobin
Figure 4.9 Results of ligand immobilization and analyte binding. As can be expected, immobilization
levels rise with increasing COOH content
content in the SAM. Immobilization capacity and subsequent analyte
binding is much higher on
n thiol surfaces than on the reference surface. Analyte response is shown on
the Y-axis to the right. A blank 95:5 position has been introduced for monothiols and a blank 30:70
position for dithiols,, to facilitate comparision between diagrams.
However, the analyte response approaches saturation rather fast and the highest
ligand level does not necessarily result in the best ligand accessibility. By employing
equation 4,, one can determine the number of analyte molecules bound to each
ligand molecule. For high efficiency, this ratio should be as high as possible, i.e. close
to 1, which means that most of the surface bound ligands are active and accessible
for analytes. Calculated accessibilities are presented in table 4.4,
4.4 with ligand
immobilized to the maximum level.
level
37
012345 36171819: =
OH:COOH ratio
100:0
95:5
90:10
70:30
50:50
30:70
0:100
Reference
;<=>[email protected] ⁄CD=>[email protected]
;<?FG=>H ⁄CD?FG=>H
Ligand accessibility –
monothiol surfaces
0
0,69
0,74
0,62
0,59
0,57
0,83
(Equation 4)
Ligand accessibility –
dithiol surfaces
0,86
0,91
0,91
0,73
0,51
0,52
0,83
Table 4.4 Ligand accessibility factors on the different surfaces. A ratio close to 1 means that most of
the surface bound ligands are active and accessible for analytes.
Maximum accessibilities are achieved on 70:30 for monothiols and 95:5 or 90:10 for
dithiols. It is remarkable that there is a substantial amount of ligand binding, and a
high accessibility on the 100:0 dithiol surface, which is not supposed to contain any
carboxylic groups at all. One reason might be that the binding is non-specific, but the
response value is recorded after two regeneration pulses which are expected to
remove non-covalently bound molecules. More likely, the purchased OH-terminated
compounds are not completely pure and do actually contain some carboxylic
moieties. This theory is supported by the fact that binding of anti-myoglobin was
more than 8 times higher to 100:0 surfaces activated with EDC/NHS than to nonactivated surfaces (not shown). If the interaction was purely non-specific, it would
probably not be as affected by the activation procedure.
When studying the reversed ligand/analyte system on monothiol surfaces, the same
trends could be observed as with anti-myoglobin/myoglobin. However, ligand
immobilization levels were much lower as myoglobin is a considerably smaller ligand.
Accordingly, analyte responses were larger as the high molecular weight antimyoglobin now acted as analyte. Accessibility factors were generally low, between
0.1 and 0.3, which is realistic for steric reasons. The large analytes can hardly bind to
every small ligand molecule present on the surface.
38
4.2.3
specific binding – plasma
Non-specific
When evaluating non-specific
specific binding, plasma from two different donors
donor (# 1953 and
# 1955), was used. Since both plasmas resulted in scaled but similar binding patterns
p
to the different SAM mixtures, only results obtained with plasma
plasma 1953 are presented
in figures 4.10 and 4.11 below. Non-specific
Non specific binding was recorded both before and
after regeneration with NaOH. The lowest degree of non-specific
specific binding was
wa
achieved
ed with a 50:50 SAM. This applies for both mono- and dithiol surfaces,
although the trend was more pronounced for monothiols.
Non-specific binding wass consistently high to surfaces with a high OH-content,
OH
which
can be explained in terms of hydrophobicity. Contact angle measurements revealed
that these surfaces have higher contact angles, i.e. are more hydrophobic, than
surfaces with a high degree of COOH.
CO
Hence, it is plausible that plasma components
are adsorbed to the OH-rich
rich surfaces via hydrophobic interaction.
inte
. On hydrophobic
surfaces in fluid environments,
environments surface associated water is readily replaced by more
hydrophobic molecules such as proteins. The process is energetically favourable as
water
ater that was previously bound is released into solution,
solution leading to a rise in entropy.
Plasma 1953 - monothiol SAMs
4000
Plasma response (RU)
3500
3000
2500
2000
before reg.
1500
after reg.
1000
500
0
100:0
95:5
90:10 70:30 50:50 30:70 0:100
Ref.
OH : COOH ratio
Figure 4.10 Non-specific
specific binding from plasma to the different monothiol SAM surfaces. Minimal
response is achieved with 50:50 SAMs.
39
Plasma 1953 - dithiol SAMs
4000
Plasma response (RU)
3500
3000
2500
2000
before reg.
1500
after reg.
1000
500
0
100:0
95:5
90:10 70:30 50:50 30:70 0:100
Ref.
OH : COOH ratio
Figure 4.11 Non-specific
specific binding from plasma to the different dithiol SAM surfaces. Minimal response
is achieved with 50:50 SAMs, where the non-specific
non
binding is very low.
On monothiol surfaces, binding levels rise as the COOH content is increased above
50%.. This may be explained by an increased electrostatic attraction between
b
the
surface and plasma components as more negatively charged COO groups are
introduced. As binding is low to both 50:50 and 0:100 dithiol SAMs, it would also
have been interesting to investigate the intermediate 30:70 mixture.
Taking both plasmas (1953
1953 and 1955) in consideration, the results obtained with
monothiols 50:50 are equivalent to those obtained on the reference chip. Dithiols
50:50 present better resistance towards non-specific
non
binding from plasma than the
reference.
4.2.4
Non-specific binding – proteins
Non-specific
specific binding of proteins to the sensor surfaces was evaluated using protein A,
which represented a “normal” protein, and HSA, which represented a “sticky”
protein. Both proteins were diluted in acetate buffer pH 5.0,, and had thus
thu a positive
net charge. Also, non-specific
specific binding from thyroglobulin diluted in neutral HBS-EP
running buffer (pH 7.4) was studied. Results are presented in figure 4.12 below.
below
Binding of thyroglobulin was rather high to the monothiol surface containing OHOH
terminated compounds only (100:0). The reason for this is not clear. It might be a
40
result of hydrophobic interaction, as the 100:0 surface is more hydrophobic than the
other mixtures.
xtures. However, contact angle measurements showed that the
ProteinA, HSA, Thyroglobulin (RU)
Nonspecific protein binding - monothiol SAMs
400
350
300
250
200
Protein A
150
HSA
100
Thyroglobulin
50
0
100:0
95:5
90:10 70:30 50:50 30:70 0:100
Ref.
OH : COOH ratio (%)
Protein A, HSA, Thyroglobulin (RU)
Nonspecific protein binding - dithiol SAMs
400
350
300
250
200
Protein A
150
HSA
100
Thyroglobulin
50
0
100:0 95:5 90:10 70:30 50:50 30:70 0:100 Ref.
OH : COOH ratio (%)
Figure 4.12 Non-specific
specific binding of different proteins to the mixed monomono and dithiol SAMs.
corresponding 100:0 dithiol surface is equally hydrophilic/hydrophobic, but it is not at
all as affected by thyroglobulin
hyroglobulin binding. Apart
art from the unexpectedly high
thyroglobulin
hyroglobulin response, non-specific
non specific binding from proteins was generally rather low
to the different surfaces. Lahiri et al.
al [28] have previously reported on the good
41
protein resistance of SAMs with ethylene glycol terminated thiols, and the binding
responses observed here are in the same range as those presented in the article. The
low non-specific binding is explained by the high protein repellence of the terminal
OEG chains. The exact mechanisms underlying their protein resistance are a matter of
debate, but factors such as OEG-water interactions, high motility and charge
neutrality are considered to be important.
On most surfaces, protein A gave the highest binding response. However, when
monitoring binding levels after regeneration (not shown), the sticky HSA was
dominating on the surface whereas the other proteins were washed away to a large
extent.
4.2.5
Choice of surfaces for further characterization
On basis of previous results, the following surfaces were chosen for further characterization with Biacore:
Monothiol surface 50:50
•
•
•
Very good surface stability (baseline changes < 15 RU during regeneration)
Good immobilization capacity
Low non-specific binding of proteins and plasma
Dithiol surface 50:50
•
•
•
Very good surface stability (baseline changes < 11 RU during regeneration)
Maximum immobilization capacity (although rather low accessibility)
Low non-specific binding of proteins and very low non-specific binding from
plasma
42
4.2.6
Ligand stability
Following ligand immobilization, either by maximum immobilization or by using the
“Aim for ligand level” software wizard, the sensor chip surface was subjected to 20
regeneration pulses of 50 mM NaOH in both flow cells. Accumulated baseline
decrease due to ligand loss is presented in figure 4.13 and ligand loss in relation to
corresponding immobilization level is presented in table 4.5. For both mono- and
dithiol surfaces, the higher immobilization level resulted in lower ligand stability. This
implies that maximum level immobilization entails a significant degree of noncovalently bound ligand molecules that dissociate from the surface during
regeneration. The majority of dissociation occured during the first regeneration, but
the baseline continued to decline during all cycles. With the lower ligand level, good
ligand stability was achieved on both thiol surfaces. The results obtained on the
dithiol surface were the most promising since the baseline remains stable after the
first regeneration cycle and does not continue to decline. It should be noted that
both thiol surfaces presented far better ligand stability than the reference surface,
where 38% of immobilized anti-myoglobin is lost during the 20 cycles of regeneration.
The corresponding ligand loss is 4% for monothiols and 2% for dithiols, when using a
similar immobilization level, and even with maximum level immobilization, the thiol
surfaces present better ligand stability than the reference.
Accumulated baseline response (RU)
Ligand stability
0
-50
-100
-150
-200
-250
-300
-350
-400
-450
-500
-550
Monothiols high (3676 RU)
Monothiols - low
(1571 RU)
Dithiols - high
(4476 RU)
Dithiols - low
(1523 RU)
Ref. - (1360 RU)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Regeneration cycle
Figure 4.13 Ligand stability on mono- and dithiol surfaces 50:50. Ligand molecules were attached
either by maximum immobilization, resulting in a high ligand level, or by using an instrument wizard
to achieve a lower level and the surface was then regenerated with NaOH. For both mono- and
dithiol surfaces, the higher immobilization level resulted in lower ligand stability.
43
Surface
Monothiols 50:50
Dithiols 50:50
Reference
Ligand loss – high level
14 %
8%
-
Ligand loss – low level
4%
2%
38 %
Table 4.5 Ligand loss displayed in % of original immobilization value. Both thiol surfaces present far
better ligand stability than the reference surface.
4.2.7
Kinetic assay performance
Kinetic assay performance was tested using the Anti-β2-microglobulin/β2microglobulin protein system. Rate constants ka and kd, affinity KD and maximum
analyte level Rmax were obtained from the kinetic evaluation of data using a 1:1
interaction model, where molecule A is assumed to bind reversibly to molecule B,
forming an AB complex. ka is the rate constant for the formation of new complexes
and kd is the rate of complex dissociation. At equilibrium, when the change in
concentration of AB is zero, equation 5 is valid:
IJ =
$H
(Equation 5)
$=
Results were compared with results from a run on the flat reference surface and
report data on a three-dimensional reference surface with a dextran matrix, see table
4.6. Numbers that differ by less than a power of ten are considered “equal” in this
context.
Parameter Monothiols 50:50
ka (1/Ms)
8.39 ∙ 105
kd (1/s)
1.45 ∙ 10-3
KD (M)
1.72 ∙ 10-9
Rmax (RU)
56.4
Dithiols 50:50
8.17 ∙ 105
1.43 ∙ 10-3
1.75 ∙ 10-9
59.2
Flat reference
8.7 ∙ 105
1.37 ∙ 10-3
1.57 ∙ 10-9
37.1
3D reference
1.2 ∙ 106
2.2 ∙ 10-3
1.9 ∙ 10-9
206
Table 4.6 Kinetic parameters obtained on mono- and dithiol surfaces 50:50, on the flat reference
surface and on a 3D reference surface, using the Anti-β2-microglobulin/β2-microglobulin protein
system. Parameters are very well corresponding between the different surface types.
No differences between the mono- and dithiol surface could be detected and the
kinetic parameters are very well corresponding with those obtained on the reference
surfaces. More ligand was immobilized to thiol surfaces than to the flat reference and
in consequence, the maxium analyte level Rmax is higher. The highest immobilization
level and Rmax is obtained with the surface carrying a three-dimensional dextran
matrix, according to expectations.
44
4.2.8
Detergents
Detergent binding to mono- and dithiol surfaces was investigated using Brij 35, P20
and Pluronic F-127. Detergents were injected during 10 minutes, whereupon
dissociation was monitored for 13 minutes. With these rather short injection and
dissociation times, one cannot be sure that the system has had time to reach
equilibrium, which may affect the results negatively. However, in Essö’s similar study
(where 2h dissociation times were used) [24], detergent levels seemed to approach
saturation during the first 5 minutes of injection, and the majority of dissociation
occurs within the first minutes after injection is completed. Thus, this shorter
experiment should be regarded as an initial check, showing general trends of
detergent binding, whereas a longer study would be needed for a more thorough
understanding of the binding events. Both flat and three-dimensional reference
surfaces were included in this experiment. In figure 4.14, binding responses to
different sensor surfaces are presented for Brij 35, P20 and Pluronic, respectively.
Each detergent/surface combination was examined twice, on separate surfaces. For
all three detergents, binding was lower to the thiol surfaces than to the references,
and dissociation occured more rapidly. The monothiol 50:50 surface presents the
best resistance towards detergent binding. The short and rigid SAM matrix on thiol
surfaces possibly contains fewer binding sites for detergents than the large and
porous dextran matrix on the 3D reference. Moreover, detergent molecules
dissociating from the interior of the dextran matrix might easily re-bind on their way
out, thus slowing down the dissociation process. On thiol surfaces, detergent binding
probably only occurs “on top” of the SAM and the dissociation is thus faster.
However, this reasoning does not apply to the flat reference surface, which also has a
SAM matrix. Instead, the very highest detergent binding and slowest dissociation is
encountered with the flat reference and the reason for this unexpectedly high
binding is not clear. For all surface types, Pluronic binds to a lower extent than Brij 35
and P20. The slight downward slope during the injection phase indicates that Pluronic
might wash away some surface bound material. The same trend is observed with Brij
35 on dithiol surfaces.
45
Response (RU)
320
280
240
200
160
120
80
40
0
-40
-80
Brij 35
Flat
reference
Monothiols
50:50
Dithiols
50:50
3D reference
Response (RU)
0
200
400
600
800
1000
Time (s)
1200
1400
1600
320
280
240
200
160
120
80
40
0
-40
-80
P20
Flat
reference
Monothiols
50:50
Dithiols
50:50
3D reference
0
200
400
600
800 1000
Time (s)
1200
1400
1600
320
280
Pluronic F-127
Response (RU)
240
200
Flat
reference
Monothiols
50:50
Dithiols
50:50
3D reference
160
120
80
40
0
-40
-80
0
200
400
600
800
1000
1200
1400
1600
Time(s)
Figure 4.14 Detergent binding to different surfaces. Binding was lower to the thiol matrices and
dissociation occurred more rapidly than on the reference surfaces.
46
4.2.9
Storage stability
Mono- and dithiol surfaces with 50:50 SAMs were stored dry at 25 °C or 40 °C for 6
days, whereupon surface stability, immobilization capacity, ligand stability and nonspecific binding of plasma were studied. Half the surfaces were stabilized with
proprietary stabilization solution, and the other half were stored without any pretreatment. Generally, it could be observed that storage had a negative influence on
all the studied aspects of sensor surface performance. The surfaces were more
damaged when stored at 40 °C than at 25 °C, according to expectations. A summary
of the tests is given below.
Surface stability: All surfaces displayed impaired SAM stability after storage, i.e. more
surface material dissociated during regeneration with 50 mM NaOH. The surfaces
presenting the best results were non-stabilized mono- and dithiols and stabilized
dithiols, all stored at 25 °C. Treatment with stabilizing solution had no positive effect
on monothiol surfaces. For dithiols however, stabilization seems to enhance surface
stability, especially for storage at 40 °C. Accumulated baseline decrease after 10
regeneration cycles is shown below in table 4.7.
Monothiols 50:50
Stabilization
No
Yes
Storage
25°C 40°C 25°C 40°C
temperature
Baseline
37
236 104 255
decrease (RU)
Flat
reference
No
Dithiols 50:50
No
25°C
27
Yes
40°C 25°C 40°C
324
15
93
25 °C
40°C
36
100
Table 4.7 Accumulated baseline decrease after 10 regeneration cycles with NaOH, shown for the
different surface/storage combinations. Treatment with stabilization solution had no positive effect
on monothiol surfaces. On dithiol surfaces, stabilization appeared to enhance stability, especially at
40°C.
Immobilization capacity: Immobilization capacity was lowered on all surfaces except
on non-stabilized monothiols 50:50, stored at 25 °C, where the result was
comparable with that obtained on the reference surface. Monothiol surfaces present
better immobilization capacities after storage than dithiol surfaces, as shown in figure
4.15. Immobilization levels are presented in % of the original immobilization level on
surfaces used directly after preparation.
47
Percentage of original immobilization
capacity (%)
Immobilization capacity after 6 days storage
110
100
90
80
70
60
50
40
30
20
10
0
102,9
99,9
87,4
81,2
85,3
60,9
33,8
25 °C
40 °C
monothiols
50:50, not
stabilized
25 °C
64,2
59,6
25 °C
40 °C
C
46,4
40 °C
25 °C
40 °C
25 °C
dithiols 50:50,
monothiols
dithiols 50:50,
not stabilized 50:50, stabilized
stabilized
40 °C
flat reference
Figure 4.15 Immobilization capacities after storage, presented as % of the original immobilization
capacity on surfaces used directly after preparation. All surfaces, except non-stabilized
stabilized monothiols at
25 °C, present lower immobilization capacity after storage.
Ligand stability: In table 4.8, normalized ligand stability is presented. As shown in
section 4.2.6, ligand stability is affected by the immobilization level, with higher levels
resulting in more ligand loss. Therefore, ligand loss is reported in % of immobilized
ligand level in the table below. Thee three poorest stabilities were encountered on
surfaces that had been stored at 40 °C. However, the stabilized dithiol surface at 40
°C was once again presenting unexpectedly good results. As noted previously (see
(se
figure 4.13),
), ligand stability was rather low on the reference surface.
Monothiols 50:50
Stabilization
Storage
temperature
Ligand loss
(% of original
immobilization level)
No
Flat
Reference
No
Dithiols 50:50
Yes
No
Yes
25°C 40°C 25°C 40°C 25°C 40°C 25°C 40°C 25°C 40°C
16
28
15
17
4
46
3
8
38
35
Table 4.8 Ligand loss, presented in % of the original immobilization level, for the different
surface/storage combinations. The three poorest stabilities were encountered on surfaces that had
been stored at 40 °C.
48
specific binding from plasma was 2-10 times
Non-specific binding – plasma: Non-specific
higher to surfaces that have been stored
stored than to those used directly,
directly as can be seen
in figure 4.16.. It is likely that storage introduces conformational changes in the OEGOE
chains of the SAM compounds, thus turning
turn
the surfaces less resistant towards nonnon
specific binding. Binding was lower to dithiol surfaces than to monothiol surfaces,
surfaces in
line with previous results presented in section 4.2.3.
4.2.3 Unfortunately, no data on nonnon
specific binding from plasma could be recorded
recorded on the reference surface due to
instrumental issues.
Non-specific
specific binding of plasma 1953 - 6 days storage
Plasma response (RU)
2500
2000
1500
1000
500
monothiols 50:50, dithiols 50:50, not monothiols 50:50,
not stabilized
stabilized
stabilized
before reg.
dithiols
monothiols
40 °C
25 °C
40 °C
25 °C
40 °C
25 °C
40 °C
25 °C
0
dithiols 50:50, results after direct
stabilized
use
after reg.
Figure 4.16 Non-specific
specific binding of plasma, before and after regeneration. To facilitate result
interpretation, previously obtained results on non-stored
non
mono- and dithiol surfaces are included in
the two rightmost columns.
4.2.10 Summary
This thesis work has demonstrated the use of mixed oligo(ethylene glycol) SAMs as
sensor surface matrices in Biacore applications. Both mono- and dithiol compounds
were investigated. By simply varying the ratio of
o OH- and COOH-terminated
terminated thiols in
solution,, it was possible to create SAMs with varying thickness, wettability and a
tailored degree of carboxylation. When increasing the content of carboxylated thiols,
thiols
49
thicker and more hydrophilic SAMs were achieved. Carboxylated thiols are longer,
with more OEG units than their corresponding hydroxyl terminated compounds, and
it was hence expected that an increased COOH content should lead to thicker SAMs.
The increased hydrophilicity can probably be explained by the presence of more sites
for hydrogen bonding. Contact angle goniometry indicated that the SAMs were not
homogenous at the molecular level, as the hysteresis between advancing and
receding angle was rather high. One reason for this might be that not all dithiol
molecules were anchored to the gold surface by both sulfur atoms, but exposing one
alkane “leg” towards the SAM exterior. This hypothesis was investigated with XPS and
results confirmed that approximately 25-30 % of the sulfur atoms were not bound to
gold but to hydrogen.
During the basic characterization, two different incubation solution concentrations
were investigated - 50 µM and 500 µM. The more dilute solution rendered SAMs of
higher quality, as shown by the results from ellipsometry, contact angle goniometry
and XPS. The self-assembly process is affected by mass transport of molecules from
solution, and when decreasing the thiol concentration and thus slowing down the
kinetics, the reorganization process producing a well-ordered monolayer appears to
be more efficient. It has been stated earlier, by other authors, that a low thiol
concentration is beneficial for SAM quality, and the results presented in this thesis
support that statement.
Characterization with the Biacore instrument revealed that the SAM matrices were
generally stable, with a low amount of material dissociating from the surfaces during
regeneration. All dithiol surfaces displayed excellent surface stability, whereas slightly
more dissociation occurred from monothiol surfaces, especially from those with
OH:COOH ratios 30:70 and 0:100. The immobilization capacity of ligand was high to
the mono- and dithiol surfaces - considerably higher than to the flat reference
surface. As expected, a higher COOH content in the SAM enabled the attachment of
more ligands. However, when studying ligand stability on 50:50 surfaces, it became
clear that not all ligand molecules were covalently attached to the surface when
maximum level immobilization was used. A ligand level somewhat below the possible
maximum is probably beneficial in terms of ligand stability and availability. This can
be achieved by reducing the activation time with EDC/NHS and/or using a shorter
ligand injection time.
Results showed that the degree of non-specific binding from plasma and proteins
varied with the SAM’s OH:COOH ratio. Minimum plasma responses were obtained
with 50:50 mixtures, both for mono- and dithiols. The dithiol surface was especially
50
good at withstanding binding from plasma. When increasing the OH content, nonspecific binding increased due to more hydrophobic interactions between surface and
plasma components. Non-specific binding from proteins was generally rather low to
the mono- and dithiol surfaces, but it was more difficult to observe any clear trends
in the results.
50:50 mono- and dithiol surfaces were chosen for further examination and
performed well in terms of kinetics and detergent binding. The kinetic parameters
obtained on thiol SAM surfaces corresponded very well to those obtained on
reference surfaces (with flat and 3D matrix). Promising results from the detergent
experiment showed that binding of Brij 35, P20 and Pluronic F-127 was substantially
lower to the thiol surfaces than to the two reference surfaces. The low binding of
detergents makes the SAM surfaces promising candidates for high sensitivity
applications. However, as mentioned earlier, a more thorough study of detergent
interaction might be relevant.
The main problem associated with SAM surfaces was their significantly impaired
performance after storage. It is clear that further work needs to be dedicated to the
improvement of storage stability before the surfaces can be interesting for
commercial purposes. It is plausible that storage induces conformational changes
within the outermost OEG parts of the thiol molecules, thus altering the surface
properties. However, it would be highly relevant to investigate these changes in
detail, by repeating the basic characterization performed on non-stored surfaces.
IRAS and ellipsometry would be suitable techniques for the investigation of OEG
chain conformation. It is possible that storage causes a transformation from helix to
all trans conformation, where the latter has been considered disadvantageous in
terms of protein resistance [25]. With the use of contact angle goniometry, one could
study if the hydrophilicity of the surfaces is altered upon storage. A decrease in
hydrophilicity could for example explain the observed increase of non-specific
binding. In conclusion, a deeper understanding of the molecular events that occur
would make it easier to suggest an appropriate way to improve the thiol surface’s
storage stability.
51
52
5 Conclusions
Main conclusions drawn from the thesis work are here summarized in short:
•
It was possible to create mixed mono- and dithiol SAMs with reproducible
thickness and wettability. IRAS and ellipsometry showed that a tailored degree of
COOH functionalities could be introduced in the SAMs by controlling the
proportion of carboxylated thiols in the incubation solution. Results from
ellipsometry, contact angle goniometry and XPS indicated that a dilute incubation
solution (50 µM) is beneficial for SAM quality.
•
Dithiol SAMs show very good surface stability, i.e. the thiol molecules were firmly
attached to the gold surface. However, not all molecules appear to be anchored
by both sulfur atoms. This might be an effect of the dithiols’ fast adsorption
behavior.
•
The immobilization capacity of ligand is considerably higher (4-5 times) to the
thiol SAM surfaces than to the flat reference, which is beneficial in terms of
detection sensitivity. It is likely, though, that a ligand density slightly below the
possible maximal capacity is preferable, as this improves ligand stability and
availability.
•
The degree of non-specific binding from plasma is comparable on the reference
and the monothiol 50:50 surface. On the dithiol 50:50 surface, plasma binding
was very low - approximately 25% of the binding levels obtained with the flat
reference. This is valid both before and after regeneration.
•
Mono- and dithiol surfaces with 50% OH-terminated thiols and 50% COOHterminated present promising sensor chip properties. The thiol surfaces are equal
to the flat reference chip in terms of non-specific binding of proteins and kinetic
assay performance, superior in terms of surface stability and far superior in terms
of ligand stability and detergent influence.
•
The thiol surfaces have problems to stand the strain of storage and are inferior to
the reference surface in terms of storage stability. The surfaces are more
negatively affected by storage at 40 °C than at 25 °C, according to expectations.
53
54
6 Future aspects
In the future, the most promising surfaces studied in this thesis might be useful in
high sensitivity applications. When studying the interactions of small molecules, with
rapid and mass transport limited kinetics, the dextran matrix may actually contain too
many binding sites. As the ligand density is so high, dissociating analytes can re-bind
on their way out. This scenario gives rise to false impressions of the actual kinetics.
The SAM sensor surface, with its lower ligand density and two dimensional geometry,
may reduce such effects of mass transport. Moreover, a dextran matrix may shrink or
swell due to changes in salt concentration or pH. Such matrix fluctuations, that are
detectable as baseline drift, will probably be greatly reduced by using a short and
rigid SAM matrix. Furthermore, the flat SAM matrix allows for interactions to take
part close to the gold surface, where the sensitivity is high. Since small molecules give
rise to low signals, a high sensitivity is necessary in order to be able to detect the
interactions that occur. 50:50 mono- and dithiol surfaces have been prepared for GE
Healthcare’s account and are to be analyzed with a small molecular assay within the
near future.
However, there are many parameters to study before the SAM sensor surfaces would
be ready for commercial applications. Firstly, the Biacore experiments performed in
this thesis need to be repeated on several surfaces of each type, in order to evaluate
reproducibity and collect statistical data. Moreover, the mixing ratio of OH- and
COOH-terminated thiols can most likely be optimized further to obtain surfaces with
even better performance. Based on the results presented in this thesis, it would be
interesting to investigate a 30:70 mixture of dithiols, as this surface might present
even lower non-specific binding than the surfaces studied in this thesis. Also, it would
be relevant to optimize the protocol for activation with EDC/NHS and subsequent
ligand immobilization.
The major problem to be addressed is the impaired performance after storage, in
particular in terms of increased non-specific binding. One solution might be to store
the surfaces in a liquid environment, but that would probably be too laborious and
expensive to work in practice. Instead, the surfaces might need to be treated with
some kind of stabilization solution (other than the one tested in this thesis), that
preserves the initial conformation of the SAM OEG chains.
55
56
7 Acknowledgements
First of all, I would like to acknowledge Per Kjellin, my supervisor at GE Healthcare.
Thank you for your encouraging support, for all your help during laboratory work and
report writing and for showing such a dedicated interest in my work.
Prof. Bo Liedberg, my examiner and supervisor at IFM, thank you for all valuable
discussions and guidance and for sharing your great knowledge in this line of
research.
Tobias Ekblad, thank you for assisting me with practical laboratory details and
answering many questions.
Thanks to Hans Sjöbom for good advices regarding my project and my report.
Thanks to Jos Buijs for teaching me about detergents and helping me with planning
and interpretation of the detergent experiment.
Linnéa Selegård, thank you for performing the XPS experiments with me.
I wish to collectively express my gratitude to the people I met at GE Healthcare and to
the employees at IFM. Thank you for being so helpful and nice to me! A special
thanks to Annica Myrskog for patiently answering all my questions about IRAS.
Last but not least, I would like to put across a heartfelt “thank you” to Mattias¸ Mum
and Dad for your continuous support during the ups and downs of this project. It has
been invaluable!
57
58
8 References
[1]
Valiokas, R. Interfacial Design and Characterization of Oligo(ethylene glycol) Self-Assembled
Monolayers, Templates for Biomolecular Architechtures; Dissertation No. 666, 2000,
Department of Physics and Measurement Technology, Linköping University, Sweden
[2]
Ulman, A. An Introduction to Ultrathin Organic Films, from Langmuir-Blodgett to SelfAssembly; Academic Press, San Diego, 1991.
[3]
Ulman, A. Self-Assembled Monolayers of Thiols; Academic Press, San Diego, 1998
[4]
Prime, K.L. ; Whitesides, G.M. Adsorption of Proteins onto Surfaces Containing End-Attached
Oligo(ethylene oxide): A Model System Using Self-Assembled Monolayers ; Journal of the
American Chemical Society, 1993, 115, 10714-10721
[5]
Ulman, A. Formation and Structure of Self-Assembled Monolayers; Chemical Reviews, 1996,
96, 1533-1554
[6]
Larsson, A. Biochip design based on tailored ethylene glycols; Dissertation No. 1111, 2007,
Department of Physics, Chemistry and Biology, Linköping University, Sweden
[7]
GE Healthcare/Biacore, Technology note 1- Surface plasmon resonance, available from
www.biacore.com , cited 2008-07-08
[8]
Liedberg, B.; Lundström, I.; Laricchia Robbio, L.; Revoltella, P. In Biomedical Opical
Instrumentation and Laser-Assisted Biotechnology, Kluwer Academic Publishers, the
Netherlands, 1996
[9]
Liedberg, B.; Johansen, K. In Methods in Biotechnology, Vol 7: Affinity Biosensors: Techniques
and Protocols. Humana Press Inc., Totowa, 1997
[10]
GE Healthcare/Biacore, Technology note 23 – Label-free interaction analysis in real-time
using surface plasmon resonance, available from www.biacore.com, cited 2008-06-24
[11]
Jönsson, U.; Fägerstam, L.; Ivarsson, B.; Johnsson, B.; Karlsson, R.; Lundh, K.; Löfås, S.;
Persson, B.; Roos, H.; Rönnberg, I.; Sjölander, S.; Stenberg, E.; Ståhlberg, R.; Urbaniczky, C.;
Östlin, H.; Malmqvist, M. Real-Time Biospecific Interaction Analysis Using Surface Plasmon
Resonance and a Sensor Chip Technology. BioTechniques, 1991, 11, 5
[12]
Biacore Sensor Surface Handbook, Version AA, October 2003
[13]
Solomons, T.W.G.; Fryhle, C.B. Organic Chemistry 8th ed., John Wiley & Sons Inc., Hoboken,
2004
[14]
Kariis, H. Adsorption of Organic Phosphines and Thiols on Metal Surfaces; Dissertation No.
523, 1998, Department of Physics and Measurement Technology, Linköping University
59
[15]
Sherman, C.P. In Handbook of Instrumental Techniques for Analytical Chemistry, chapter 15;
Settle, F., Ed; Prentice-Hall Inc., New Jersey, 1997
[16]
Tengvall, P.; Lundström, I.; Liedberg, B.; Protein adsorption studies on model organic
surfaces: an ellipsometric and infrared spectroscopic approach, Biomaterials 1998, 19, 407422
[17]
Nave, C.R., Department of Physics and Astronomy, Georgia State University; Classification of
Polarization. Available from http://hyperphysics.phy-astr.gsu.edu/hbase/phyopt/polclas
.html, cited 2008-12-03
[18]
Elwing, H. Protein absorption and ellipsometry in biomaterials research; Biomaterials, 1998,
19, 397-406
[19]
Davies, J. Surface Analytical Techniques for Probing Biomaterial Processes; CRC Press, Boca
Raton, 1996
[20]
Kabza, K.; Gestwicki, J.E. ; McGrath, J.L. Contact Angle Goniometry as a Tool for Surface
Tension Measurements of Solids, Using Zisman Plot Method; Journal of Chemical Education,
2000, 77 (1), 63
[21]
Attard, G.; Barnes, C. Surfaces; Oxford University Press, New York, 1998.
[22]
McCrackin, F.L.A, NBS Technical Note 479, Washington DC, 1969
[23]
Biacore Instrument Handbook, Version AE, October 2003
[24]
Essö, C.; Modifying Polydimethylsiloxane (PDMS) surfaces; Mälardalen University,
Department of Biology and Chemical Engineering, 2007. Available from www.divaportal.org/mdh/abstract.xsql?dbid=491, cited 2008-11-11
[25]
Valiokas, R.; Interfacial Design and Characterization of Oligo(ethylene glycol) Self-Assembled
Monolayers; Dissertation No. 666, 2000, Department of Physics and Measurement
Technology, Linköping University
[26]
Spangler, B.D.; High Affinity Self-Assembled Monolayers on Gold: Why more is better;
Material Matters, Aldrich Chemical Company, 2006 1 (2), 15-17
[27]
Unsworth, L.D.; Sheardown, H.; Brash, J.L. Protein Resistance of Surfaces Prepared by
Sorption of End-Thiolated Poly(ethylene glycol) to Gold: Effect of surface chain density.
Langmuir 2005, 21, 1036-1041
[28]
Lahiri, J.; Isaacs; L.; Tien, J.; Whitesides, G.M. A strategy for the Generation of Surfaces
Presenting Ligands for Studies of Binding Based on an Active Ester as a Common Reactive
Intermediate: A Surface Plasmon Resonance Study. Analytical Chemistry, 1999, 71 (4)
60
Appendix A – Reagents
Reagents used at IFM, Linköping:
Reagent
Supplier
Ethanol 99,5 % (CH3CH2OH)
Hydrogen peroxide (H2O2) 30%
Ammonia solution (NH3) min 25%
Human Serum Albumin (HSA)
HSA
Rabbit Anti Human Serum Albumin
Rabbit Anti-Human IgG, A0424
Sodium acetate 10mM pH 4.5
Sodium acetate 10mM pH 5.0
HBS-EP buffer (10mM HEPES pH 7.4,
150 mM NaCl, 3.4 mM EDTA, 0,005%
Surfactant P20)
1-Ethyl-3-(3-dimethylaminopropyl)
carbodiimide hydrochloride (EDC)
N-Hydroxysuccinimide (NHS)
Ethanolamine-HCl 1,0M pH 8.5
Deuterated hexadecanethiol
(HS(CD2)15CD3)
Kemetyl AB, Haninge, Sweden
Merck KGaA, Darmstadt, Germany
VWR International, Fontenay sous Bois, France
Sigma Aldrich Sweden AB, Stockholm, Sweden
Sigma-Aldrich Sweden AB, Stockholm, Sweden
Sigma-Aldrich Sweden AB, Stockholm, Sweden
Dako, Glostrup, Denmark
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
Laborana Fine Chemicals, Malmö, Sweden
Reagents used at GE Healthcare, Uppsala:
Reagent
Supplier
Ethanol 99,5 % (CH3CH2OH)
Hydrogen peroxide (H2O2) 30%
Ammonia solution (NH3) min 25%
Myoglobin
Anti-myoglobin 2F9.1
ProteinA
Human Serum Albumin (HSA)
Rabbit anti Mouse Subclass kit
(ramFC)
Ovalbumin
Kemetyl AB, Haninge, Sweden
Fluka, Sigma-Aldrich, Buchs, Switzerland
Merck KGaA, Darmstadt, Germany
Sigma-Aldrich Sweden AB, Stockholm, Sweden
LabAs, Tartu, Estonia
Sigma-Aldrich Sweden AB, Stockholm, Sweden
Sigma-Aldrich Sweden AB, Stockholm, Sweden
Pharmacia Biosensor AB, Uppsala, Sweden
Thyroglobulin
Ferritin
Catalase
Amersham Pharmacia Biotech Inc., Piscataway,
USA
Amersham Pharmacia Biotech Inc., Piscataway,
USA
Amersham Pharmacia Biotech Inc., Piscataway,
USA
Amersham Pharmacia Biotech Inc., Piscataway,
61
Monoclonal mouse-anti-human β2microglobulin
Human β2-microglobulin
Sodium acetate 10mM pH 5.0
HBS-EP buffer (10mM HEPES pH 7.4,
150 mM NaCl, 3.4 mM EDTA, 0,005%
Surfactant P20)
HBS-N buffer (10mM HEPES pH 7.4,
150 mM NaCl)
1-Ethyl-3-(3-dimethylaminopropyl)
carbodiimide hydrochloride (EDC)
N-Hydroxysuccinimide (NHS)
Ethanolamine-HCl 1,0M pH 8.5
10 mM Glycine-HCl pH 1.5
10 mM Glycine-HCl pH 2.0
10 mM Glycine-HCl pH 2.5
Sodium hydroxide, 50mM
Dimethyl sulfoxide (CH3)2SO
Human plasma #1953
USA
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
GE Healthcare, Uppsala, Sweden
Riedel-de Haën, Seelze, Germany
Blodcentralen, Uppsala Academic Hospital,
Uppsala, Sweden
Human plasma #1955
Blodcentralen, Uppsala Academic Hosptial,
Uppsala, Sweden
Pluronic F-127
Sigma Aldrich Sweden AB, Stockholm, Sweden
Brij 35
Sigma Aldrich Sweden AB, Stockholm, Sweden
Polyoxyethylene sorbitan monolaurate GE Healthcare, Uppsala, Sweden
(P20) 10% solution
BIAdesorb solution 1 (0.5% (w/v)
GE Healthcare, Uppsala, Sweden
sodium dodecyl sulphate)
BIAdesorb solution 2 (50mM glycine
GE Healthcare, Uppsala, Sweden
pH 9.5)
62
Appendix B – Experimental details
In this appendix, experimental conditions for the tests evaluating Biacore sensor chip
performance are presented in detail. HBS-EP buffer was used for all tests except
“Detergents” when detergent free HBS-N was chosen. The temperature was 25 °C.
SAM preparation
Chip 1
Chip 2
Chip 3
Surface stability
Non-specific
binding - plasma
Ligand stability
Immobilization
capacity and
analyte binding
Non-specific
binding - proteins
Kinetic assay
performance
Chip 4
Detergents
Non-specific
binding - plasma
Surface stability
Flow:
100 µl/min
Flow cell/cells: 1, 2, 3, 4
Injections:
10 ∙ 30 sec 50 mM NaOH in FC 1 and FC 2
10 ∙ 30 sec 10 mM glycine-HCl pH 1.5 in FC 3 and FC 4
63
Chip 5
Storage
stability
Immobilization capacity and analyte binding
Flow:
10 µl/min
Flow cell/cells: 2, 4
Injections:
30 sec HBS-EP
5 min ligand
2 ∙ 30 sec 10 mM glycine-HCl pH 1.7
7 min EDC/NHS
7 min ligand
7 min ethanolamine
2 ∙ 1 min 10mM glycine-HCl pH 1.7
5 min analyte
1 min 10 mM glycine-HCl pH 1.7
Ion exchange
Ligand immobilization
Analyte injection
Comments:
For dithiol surfaces, identical injections were run in FC 2 and 4 with
Anti-myoglobin (50 µg/ml in Acetate pH 5.0) as ligand and Myoglobin (5 µg/ml in
HBS-EP) as analyte. On monothiol surfaces, the system described above was studied
in FC 2. Moreover, an additional analyte, ramFC, was injected after myoglobin. In FC
4, the ligand/analyte system was reversed so that Myoglobin (24 µg/ml in Acetate pH
5.0) served as ligand and Anti-myoglobin (50 µg/ml in HBS-EP) served as analyte.
Non-specific binding – plasma
Flow:
10 µl/min
Flow cell/cells: 1 and 3 for Chip1, 4 for Chip2
Injections:
2 ∙ 30 sec 50 mM NaOH
5 min plasma, centrifuged (4000 rpm, 20 min) and filtered (1 µm +
0.2 µm)
2 ∙ 30 sec 50 mM NaOH
Non-specific binding – proteins
Flow:
10 µl/min
Flow cell/cells: 1, 2, 3
Injections:
2 ∙ 30 sec 50 mM NaOH
5 min protein
2 ∙ 30 sec 50 mM NaOH
64
Comments:
Non-specific binding from ProteinA, (100 µg/ml in acetate buffer pH
5.0) was studied in FC 1, from HSA (100 µg/ml in acetate buffer pH 5.0) in FC 2 and
from Thyroglobulin (150 µg/ml in HBS-EP) in FC 3.
BME
Flow:
15 µl/min
Flow cell/cells: 1, 2, 3, 4
Ligand stability
Flow:
10 µl/min during protein immobilization, 100 µl/min during
regeneration.
Flow cell/cells: 1, 2
Injections:
7 min EDC/NHS
7 min Anti-myoglobin 50 µg/ml in Acetate pH 5.0
7 min ethanolamine-HCl
7 min EDC/NHS
Pulses of Anti-myoglobin 25 µg/ml in Acetate pH 5.0
7 min ethanolamine-HCl
FC 1
FC 2, Wizard
20 ∙ 30 sec 50 mM NaOH in FC 1 and 2
Kinetic assay performance
Flow:
10 µl/min during protein immobilization, 30 µl/min during analyte
injection
Flow cell/cells: 4, with 3 as an on-line reference cell
Injections:
7 min EDC/NHS
7 min Anti-β2-microglobulin 40 µg/ml in Acetate pH 5.0
7 min Ethanolamine-HCl
2 min β2-microglobulin
5 min dissociation time
Analyte injection
30 sec glycine-HCl, pH 2.5
Comments: The analyte injection step was repeted for β2-microglobulin with
concentrations 0 nM, 2 nM, 4 nM, 8 nM, 16 nM, 32 nM and 64 nM.
65
Detergents
Flow:
30 µl/min during initial washing steps (with BIAdesorb1, BIAdesorb2
and NaOH), 10 µl/min during detergent injection
Flow cell/cells: 2, 3, 4
Injections:
1 min BIAdesorb1 (0.5% (w/v) sodium dodecyl sulphate)
1 min BIAdesorb2 (50mM Glycine pH 9.5)
1 min 50mM NaOH
10 min detergent 0,05%
13 min dissociation time
3 ∙ 1 min HBS-N buffer
Comments:
After normalization, the sensor chip surface was washed with
BIAdesorb solution 1, BIAdesorb solution 2 and 50 mM NaOH to remove any possibly
unbound material from the surface. After washing, the baseline was allowed to
stabilize for three minutes prior to detergent injection. Brij 35 was injected in FC 4,
P20 in FC 3 and Pluronic F-127 in FC 2. After each detergent injection, needle and
sample line were thoroughly washed with BIAdesorb 1 and 2.
Storage stability
Flow:
10 µl/min during protein immobilization and plasma injection, 100
µl/min during regeneration.
Flow cell/cells: 1, 2, 4
Injections:
7 min EDC/NHS
7 min Anti-myoglobin 50 µg/ml in Acetate pH 5.0
FC 2
7 min ethanolamine
10 ∙ 30 sec 50 mM NaOH in FC 1 and 2
2 ∙ 30 sec 50 mM NaOH
5 min plasma, centrifuged (4000 rpm, 20 min) and filtered
FC 4
(1 µm + 0.2 µm)
2 ∙ 30 sec 50 mM NaOH
Comments: Via regeneration in both FC 1, which was empty, and FC 2, with
immobilized ligand, information about surface stability as well as ligand stability could
be obtained.
66
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement