scieee Science in your language
[en] (orig)
Berlin 2018
Spectroelectrochemical study of
biomolecules in artificial
membrane systems
Vorgelegt von
M.Sc. Biomedizin
Barbara Daiana Gonzalez
Geb. in Buenos Aires, Argentinien
von der Fakultät II – Mathematik und Naturwissenschaften
der Technischen Universität Berlin
zur Erlangung des akademischen Grades
Doktor der Naturwissenschaften
Dr. rer. nat.
genehmigte Dissertation
Promotionsausschuss:
Vorsitzender: Prof. Dr. Nediljko budisa
Berichter: Prof. Dr. Peter Hildebrandt
Berichter: Prof. Dr. Inez Weidinger
Tag der wissenschaftlichen Aussprache: 28.09.2017
2
3
Abstract
This work presents a spectroeletrochemical approach combining electrochemical
impedance spectroscopy (EIS) and surface-enhanced infrared absorption (SEIRA)
spectroscopy to characterise the interaction of proteins and peptides with model
membranes. The artificial membrane systems are assembled on a nanostructured
Au-film, which is deposited on an attenuated total reflectance (ATR) crystal as a
prerequisite for the enhancement of the IR signals of the immobilised molecules. The
Au-film also serves as working electrode to control and vary the potential across the
immobilised membrane model. Here, a nanodisc system was employed to investigate
the membrane protein channelrhodopsin II (ChR II), while tethered bilayer lipid
membrane (tBLM) systems were used to characterise the antimicrobial peptides
enniatin B (EB) and arenicin 1 (A1).
For the first time, the light-induced structural changes of the ―slow‖ C128S mutant of the
light-gated ion channel ChR II embedded into nanodisc systems were observed by
SEIRA spectroscopy. The nanodisc systems containing zwitterionic lipids and His-tag tails
were successfully immobilised using the Ni-NTA (nitrilotriacetic acid) monolayer. The
spectral changes were assigned to structural variations of the protein backbone
associated to the transition from the dark state to the P390 intermediate. This
proof-of-principle study demonstrates that nanodisc represent a promising system to
study the retinal protein ChR II.
Characterisation of EB peptide by SEIRA spectroscopy was performed using a
1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC, zwitterionic lipids) tBLM
system. The unprecedented results, provide a comprehensive evaluation of EB
incorporation in membrane systems in the presence of monovalent cations, as well as a
new understanding of EB:ion complex formation. The spectral features of EB ligand
groups in presence of K+ ion, demonstrated that the K+ complex significantly differs from
the species formed in the presence of Na+ and Cs+ ions, which are spectroscopically
alike. These findings disagree with previous studies, and contribute to the controversial
discussion of EB:ion complex stoichiometry assignment.
Finally, tBLMs containing 1-palmitoyl-2-oleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)]
(sodium salt) (POPG) negatively charged lipids were employed for characterising the
membrane binding of the cationic peptide A1. The unique structure of A1 and its
potential-dependent activity were investigated by SEIRA spectroscopy. The β-sheet
folding exhibited high stability in different environments (solid, solution and membrane)
and temperatures. The highest activity was recorded at 37° C upon application of
negative potentials (-400 mV). Overall, A1 significantly disturbed the lipid membrane in
a non-reversible manner.
The EIS and SEIRA spectroscopy results were complemented with UV-vis, ATR-IR, FT-IR
transmission spectroscopy, as well as theoretical calculations, allowing a sound
structure-function insight. The high reproducibility and stability of the membrane
systems, together with the specific results for the target biomolecules, have shown that
EIS and SEIRA spectroscopy are powerful tools for the structural and functional study of
biomolecules and their interaction with model membranes.
4
Zusammenfassung
Die vorliegende Arbeit beschreibt einen spektro-elektrochemischen Ansatz, in welchem
elektrochemische Impedanzspektroskopie (EIS) und oberflächenverstärkte
Infrarotabsorptions-Spektroskopie (surface-enhanced infrared absorption, SEIRA) zur
Charakterisierung der Interaktion zwischen Proteinen und Peptiden mit Model-Membranen
kombiniert wurden. Die künstlichen Membransysteme wurden auf einen nanostrukturierten
Au-Film adsorbiert, der auf einem Abgeschwächte Totalreflexion (attenuated total
reflectance) ATR-Kristall als Voraussetzung für die Verstärkung des Infrarotsignals der
immobilisierten Moleküle aufgebracht wurde. Die Au-Schicht dient ebenso als
Arbeitselektrode und erlaubt die Kontrolle und Veränderung des Potentials an der
immobilisierten Membran. Im Rahmen dieser Arbeit wurde ein Nanodisc-System zur
Untersuchung des Membranproteins Kanalrhodopsins II (ChR II) genutzt, während zur
Analyse der antimikrobiellen Peptide Enniatin B (EB) und Arenicin 1 (A1) Oberflächen-
adsorbierte Lipiddoppelschichtmembransysteme (tethered bilayer lipid membrane, tBLM)
verwendet wurden.
Die licht-induzierten strukturellen Änderungen der „langsamen― C128S-Mutante des licht-
gesteuerten Ionenkanals ChR II eingebettet in einem Nanodisc-System wurden zum ersten
Mal mit SEIRA-Spektroskopie gezeigt. Das Nanodisc-System, bestehend aus zwitterionischen
Lipiden, wurde mit einem terminalen His-Tag erfolgreich an eine Nickel-Nitrilotriessigsäure-
Monolage (Nit-NTA) immobilisiert. Die spektralen Änderungen ordneten man der strukturellen
Unterschiede des Proteinrückrats zu. Diese erfolgen beim Übergang des Dunkelzsutandes
zum P390 Intermediat. Hierbei wurde gezeigt, das Nanodiscs ein vielversprechendes System
zur Untersuchung des Retinalproteins ChR II sind.
Charakterisierung des EB Peptids mittels SEIRA Spektroskopie erfolgte unter der Verwendung
von einem 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC, zwitterionische Lipide)
tBLM System. Die neuartige Erkenntnisse bieten eine umfassende Auswertung des EB
Einbaues in Membran-Systeme bei der Anwesenheit von monovalenten Kationen sowie der
EB:Ion-Komplexbildung. Die spektralen Merkmale der EB-Ligandengruppen in Anwesenheit
eines K+-Ion zeigten, dass der K+-Komplex Unterschiede im Gegensatz zur Na+- und Cs+-
Spezies aufweist. Diese Ergebnisse wiedersprechen vorherige Untersuchungen, dadurch
tragen sie zur kontroversen Diskussion der stöchiometrischen Zuordnung des EB:Ion-
Komplexes.
Abschließend wurden tBLM-Systeme, bestendend aus 1-palmitoyl-2-oleoyl-sn-glycero-3-
[phospho-rac-(1-glycerol)] (POPG als Natriumsalz) negativ geladenen Lipide, zur
Untersuchung des kationischen Peptids A1 eingesetzt. Die einzigartige A1 Peptidstruktur und
deren Potential-Abhängigkeit wurden mittels SEIIRA-Spektroskopie untersucht. Die
β-Faltblattstruktur zeigte höhe Stabilität in verschiedenen Umgebungen (fester Zustand,
Lösung und Membran) und Temperaturen. Die höchste Aktivität wurde bei 37 °C und einem
negativen Potential (-400 mV) nachgewissen. Insgesamt wurde eine irreversible Störung der
Membran beobachtet.
Die Ergebnisse der EIS und SEIRA spektroskopischen Untersuchungen wurden mit UV-VIS, ATR-
IR und FT-IR Transmission, sowie theoretischen Berechnungen ergänzt, dadurch wurde einen
aussagekräftigen Einblick der Struktur-Funktion-Beziehung erreicht. Die gute
Reproduzierbarkeit der Experimente und Stabilität des Membransystems haben zusammen
mit den getroffenen Aussagen zu den untersuchten Molekülen bewiesen, dass EIS und SEIRA
Spektroskopie aussagekräftige Werkzeuge zur strukturellen und funktionellen Betrachtung
der Wechselwirkung von Biomolekülen mit Model-Membranen darstellen.
5
Poster contributions
1. Antimicrobial peptides, Gordon research conference, May 3rd – 8th 2015. Lucca
(Italy), Spectroelectrochemical study of the antimicrobial peptide enniatin B.
2. School of analytical sciences Adlershof poster session, November 2015. Berlin
(Germany), Spectroelectrochemical study of the antimicrobial peptide enniatin B.
Awarded best poster design.
3. AMP2016-International symposium on antimicrobial peptides, June 6th – 8th 2016.
Montpelier (France), Spectroelectrochemical study of the antimicrobial peptide
enniatin B.
4. IMAP-6th International meeting on antimicrobial peptides, August 31st – September
3rd. Leipzig (Germany), Structural study of the antimicrobial peptide enniatin B in a
biomimetic bilayer membrane system.
6
7
Table of content
Abstract ........................................................................................................................................ 3
Zusammenfassung ...................................................................................................................... 4
Poster contributions ..................................................................................................................... 5
Table of content .......................................................................................................................... 7
Abbreviations ............................................................................................................................ 11
1. Motivation .............................................................................................................................. 13
2. Introduction ............................................................................................................................ 17
2.1. Artificial membrane systems ............................................................................................ 17
2.1.1. Cell membranes and artificial membranes ....................................................... 17
2.1.1.1. Phospholipids ............................................................................................. 18
2.1.2. Nanodisc membrane systems .............................................................................. 20
2.1.3. Tethered bilayer lipid membrane systems ......................................................... 21
2.2. Membrane proteins .......................................................................................................... 22
2.2.1. Infrared spectroscopy of proteins ....................................................................... 24
2.2.2. Channelrhopsin II membrane protein................................................................. 27
2.3. Antimicrobial peptides ..................................................................................................... 29
2.3.1. Enniatin B .................................................................................................................. 32
2.3.2. Arenicin 1 .................................................................................................................. 34
2.4. Theory of vibrational spectroscopy ............................................................................... 36
2.4.1. Basics of vibrational spectroscopy ...................................................................... 37
2.4.1.1. Molecular vibrations and the harmonic oscillator .............................. 37
2.4.1.2. Normal modes ........................................................................................... 39
2.4.2. Infrared absorption spectroscopy ....................................................................... 41
2.4.3. Fourier transform IR ................................................................................................. 42
2.4.4. Attenuated total reflectance IR .......................................................................... 44
2.4.5. Surface-enhanced IR absorption SEIRA spectroscopy ................................... 45
2.5. Theory of electrochemical impedance spectroscopy .............................................. 47
2.5.1. Basic theory of the electrical impedance ......................................................... 47
2.5.2. Graphical representation of impedance data ................................................ 49
8
2.5.3. Graphical evaluation and physical relevance in tBLM systems .................... 50
3. Experimental section ............................................................................................................ 52
3.1. Materials .............................................................................................................................. 52
3.2. Sample preparation .......................................................................................................... 53
3.2.Methods ................................................................................................................................ 54
4. Results ..................................................................................................................................... 60
4.1. Channelrhodopsin II in nanodisc systems ..................................................................... 60
4.1.1. UV-vis study of ChR II protein ................................................................................ 61
4.1.2. FT-IR transmission study of ChR II protein ............................................................ 65
4.1.2.1. Illumination cycles of ChR II protein ...................................................... 66
4.1.3. Spectroelectrochemical study of Ni-NTA monolayers .................................... 68
4.1.4. SEIRA study of the ChR II C128S protein in nanodisc systems ......................... 73
4.1.5 Conclusions ............................................................................................................... 75
4.2. Characterisation study of enniatin B ............................................................................. 76
4.2.1. Incorporation of enniatin B into tBLM system .................................................... 77
4.2.1.1. EIS of POPC membrane systems ............................................................ 77
4.2.1.2. SEIRA spectroscopy of POPC membrane systems ............................. 81
4.2.1.3. Membrane incorporation of enniatin B ................................................ 84
4.2.2. Structural study of ennitain B ion complexes in membrane systems ............ 88
4.2.2.1. Enniatin B ion-exchange in membrane systems ................................. 88
4.2.2.2. IR spectroscopy of enniatin B ................................................................. 91
4.2.2.3. Theoretical calculations of enniatin B ion complexes ....................... 94
4.2.3. Mechanism of membrane incorporation and ion binding ............................ 99
4.2.4. Conclusions ............................................................................................................ 104
4.3. Antimicrobial peptide arenicin1 ................................................................................... 105
4.3.1. IR and Raman spectroscopy of arenicin 1 ...................................................... 105
4.3.1.1. ATR-IR spectroscopy of solid arenicin 1 .............................................. 106
4.3.1.2. FT-Raman spectroscopy of solid arenicin 1 ....................................... 108
4.3.1.3. FT-IR transmission of arenicin1 in solution ............................................ 109
4.3.2. SEIRA and EIS study of the POPC/POPG membrane systems ...................... 111
4.3.3. Peptide-membrane interaction study by SEIRA ............................................. 113
9
4.3.3.1. Arenicin1 interaction with POPC/POPG tHLM system. .................... 113
4.3.3.2. SEIRA spectroscopy of arenicin1 interaction with POPC/POPG tBLM
system: temperature dependence .................................................................. 114
4.3.4. Molecular dynamics simulation of arenicin1 ................................................... 117
4.3.5. Conclusions. ........................................................................................................... 120
5. Conclusions ......................................................................................................................... 122
Outlook ..................................................................................................................................... 124
Bibliography ............................................................................................................................. 126
Acknowledgements ............................................................................................................... 138
Appendix ................................................................................................................................. 141
10
11
Abbreviations
1PrOH
1-Propanol
6MH
6-Mercaptohexanol
A1
Arenicin 1
ACN
Acetonitrile
AMPs
Antimicrobial peptides
ATR
Attenuated total reflectance
Arg
Arginine
BTP
Bis-Tris propane
ChR II
Channelrhodopsin II
ChRs
Channelrhodopsins
CPE
Constant phase element
CsCl
Cesium chloride
Cys
Cysteine
D-Hiv
D-hydroxyisovaleric acid
DMPC
1,2-dimyristoyl-sn-glycero-3-phosphocholine
EB
Enniatin B
EIS
Electrochemical impedance spectroscopy
EM
Electromagnetic
FT
Fourier transform
FT-IR
Fourier transform infrared
HCl
Hydrogen chloride
IR
Infrared
KCl
Potassium chloride
NaCl
Sodium chloride
POPC
1-Palmitoyl-2-oleyl-sn -glycero-3-phosphocholine
POPG
1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(1'-rac-glycerol) (sodium salt)
SAM
Self-assembled monolayer
SEIRA
Surface-enhanced IR absorption
tBLM
Tethered bilayer lipid membrane
Trp
Tryptophan
Tyr
Tyrosine
Val
N-methyl-L-valine amino acid
WK3SH
Dihydrocholesteryl (2-(2-(2-ethoxy)ethoxy)ethanethiol
UV-vis
Ultraviolet-visible
12
13
1. Motivation
The main motivation for this thesis is the increasing demand for a better understanding
of the interaction of biomolecules with the cell envelope, and in particular with the
plasmatic membrane. This membrane is comprised of lipids, complex proteins and
other molecules that contribute to form a barrier for the cell. In this membrane an
astonishing number of reactions take place, namely signalling, molecule and cell
recognition, chemical reactions, transport, in addition to many others. A characteristic
of the cell membrane is that it is considered the first contact-site and entrance for many
pathogenic agents. An example is in the symbiotic relationship
1
that multicellular
organisms share with microbiota, where a deregulation of the homeostasis of this
ecosystem leads to dysfunction or disease in the host organism [1]. The capability of
evolution and adaptation of the microbiota is the challenge and motivation for science
to constantly innovate techniques to investigate the relationship between biomolecules
and membranes. The design of mimetic membrane systems applied to
membrane-interacting molecules has been and still is an evolving field, which is of
great relevance in the study of physiological disorders and drug mechanisms of action.
With the aim of studying membrane-related processes, the three targets used in this
work are chosen based on their implication in novel research fields, such as
optogenetics and antimicrobial peptides (AMPs). The membrane proteins
channelrhodopsins are microbial rhodopsins that govern the phototaxis of the green
algae Chlamydomonas reinhardtii, acting as light-sensitive ion-channels. Several studies
of wild-type and mutant proteins have shown the possibility of controlling the open and
close state of the channel by light. This fact gave rise to the development of the
optogenetic techniques, which mainly use Channelrhodopsin II (ChR II) in neural cells
studies [2]. On the other hand, the discovery of AMPs in the beginning of the 20th
century offered new directions in the development of drugs. The potential of AMPs to
target a wide range of diseases, from bacterial and fungi infections to viral, parasitic
and cancer disorders, aims to overcome the increasing resistance and adaptation of
bacteria to conventional antibiotics. In the context of this work, there is a special
interest in membrane-active AMPs, in their structural features and mechanism of action.
The cyclic enniatin B (mainly from Fusarium scripi fungi) and the distinctive arenicin 1
(from lugworm Arenicola marina) AMPs present interesting structural characteristics,
which are in the focus of controversial discussions. Even though nowadays, more than
thousands of AMPs have been isolated or synthesised, there is a great opportunity for
the development of advanced methods offering more detailed information about their
selectivity and interaction with membranes.
1
The microbiota in multicellular organisms is composed mainly of commensal bacteria that do not cause harmful effect to
the host.
14
To date, there are several techniques used in life science that can provide detailed
chemical information of biomolecules. However, most of them present restrictions in
size, purity or form (solid, crystalline, etc.) of the sample. Vibrational spectroscopy is one
of the techniques that have evolved significantly to overcome these drawbacks. The
advances in application of non-invasive techniques like Raman and infrared (IR)
spectroscopy, allows scientists to investigate building blocks of biomolecules like
complex proteins or peptides, as well as the composition of the cell membrane. Due to
innovations in IR techniques, the characterisation of many macromolecules in solution,
and also solid state was possible. Yet, membrane proteins and related processes
represent a challenge for structural and functional analysis, since they demand a
membrane environment to avoid denaturation. Membrane-mimetic systems like
micelles, vesicles and solid supported bilayers are used to overcome these problems.
Albeit to date, most of the membrane-mimetic systems are applied to electrochemical
or imaging techniques, the discovery of the surface-enhanced IR absorption (SEIRA)
effect increased significantly the sensitivity for structural studies of biological systems,
especially those using membrane systems. SEIRA spectroscopy exploits the properties of
functionalised nanostructured noble metals, i.e. Au, which amplifies the signal and at
the same time, acts as an electrode for electrochemical analysis. In this way, SEIRA is a
powerful tool for the study of membrane proteins or peptides in a native-like
environment by using supported membrane systems linked to the metal surface. The
surface-vicinity restriction in SEIRA (short-range enhancement), allows detecting
processes close to the surface minimizing the interferences from the bulk solution.
In this work, the main goal is to employ planar and nanodisc membrane systems in an
spectroelectrochemical approach combining SEIRA and electrochemical impedance
spectroscopy (EIS) for the structural study of the targets mentioned above. The use of
EIS allows estimating the quality of the monolayers and membrane-mimetic systems.
The challenge of this combined method lies mainly in the development of a stable,
robust and reproducible system. In this approach there are many parameters to control
simultaneously: membrane fluidity, optimal environment for stability and function of
targets (concentration, ionic strength, pH), monolayer ratio for membrane systems,
among others. In some cases, an initial optimisation stage was performed to determine
the optimal balance between detection of the target spectroscopic features and
membrane stability. Whereas the lipids used in this study stand for a small portion of the
diverse cell membrane composition, the presented membrane systems serve as a
starting point for optimisation towards a more representative membrane mimic.
The first part in this work offers an introduction to the methods and targets, as well as to
the theory behind the applied techniques. The results are presented in three sections,
one for each target and the corresponding membrane system, which were all
investigated by SEIRA and EIS spectroscopy. A nanodisc system is used in the study of a
slow mutant of the light-sensitive membrane protein ChR II. A zwitterionic planar
membrane system serves as platform for the characterisation of the cyclic enniatin B
peptide and its ion complexes. A negatively charged planar membrane system is
15
applied in the investigation of arenicin 1 peptide and its disturbance of the membrane.
A number of vibrational spectroscopic techniques such as IR transmission and UV-vis,
together with preliminary theoretical calculations are used to validate or complement
the results obtained from the spectroelectrochemical approach. At the end, the final
conclusions are presented along with an outlook for each project.
16
17
2. Introduction
This section situates the reader into context, for a better understanding of the work
presented in this dissertation. The introduction to the systems and targets used in this
work, in addition to theoretical background set the basis for comprehension and
discussion of the results.
2.1. Artificial membrane systems
The following subsections offer an introduction to cell membranes and its diverse
composition, as well as to artificial membrane systems. A description of the
phospholipids used in this study, is essential to understand the structures they can adopt
and their fascinating behaviour. These fundamentals are basic to get familiar with the
two membrane systems introduced at the end of the section.
2.1.1. Cell membranes and artificial membranes
Lipid membranes are essential for the architecture, organization and function of living
cells. The cell membrane is involved in processes like signal transduction,
electrochemical gradients or ligand-receptor interactions. The composition of cell
membranes varies with the type of cell, containing structures from small molecules to
complex macrostructures like protein aggregates. In the so called ―tree of life‖ cells are
classified into three kingdoms by their evolution in respect with a common progenitor,
which allows distinguishing cells organizational levels. The architecture and composition
of the cell envelope can differ not only between, but also within the archaebacteria,
eubacteria and eukaryotes kingdoms. A classic example is the difference between
plants and animals in the eukaryotic group; plant cells have a cell wall layer outside the
plasma membrane, while the principal barrier to the environment in the case of animal
cells is the plasma membrane [3]. Besides the differences, all cell envelopes have in
common that they act as a fort for the cell with communication lines between the
external and internal content. Therefore, artificial membrane systems play a key role in
the study of membrane interaction with its components and external agents [3] [4]. The
characterization of plasma membranes as mainly composed by proteins and
phospholipids arranged in bilayer motifs [5], is in constant update with new distinct
membrane domains, like lipid rafts in bacteria membranes recently described by
Nickels and coworkers [6].
The involvement of membrane domains in numerous mechanisms of disease, and the
demanding requirements for the stabilization of its components, have pushed the field
of biophysics to develop biomimetic models to study their structure and function.
18
Artificial membrane systems have evolved significantly since the ―fluid mosaic model‖
(also known as painted membrane) developed by Mueller and coworkers in the early
‗60s of last century [7]. Nowadays, there is a wide range of constructs of diverse
complexity and application, going from vesicles and micelles to supported bilayers and
nanodiscs [8] [9] [10] [11] [12]. The development of these systems has allowed
investigating not only cell membrane assemblies and lipid composition, but also
characterizing structure and function of peptides and membrane proteins [13] [14]. This
kind of study is especially relevant for membrane proteins, which require a native-like
environment for proper conformation and function, as well as for other
membrane-related processes. In several cases, membrane components have been
used as pharmacological targets for the treatment of various diseases and disorders
[15]. The versatile characteristics of supported artificial membrane systems permit its
combination with an extensive number of techniques, including vibrational
spectroscopy [16] [17]. The work presented here focuses on two models, nanodisc and
tethered bilayer lipid membrane (tBLM) systems.
2.1.1.1. Phospholipids
The main lipids in plasma membranes are the phospholipids. The phospholipid structure
can be divided into three parts: a hydrophobic tail, a connecting backbone, and a
polar head group (see figure 2.1). The medium part has an important effect for the
organization of lipids. They form mesophases in aqueous solution by self-assembling in a
variety of structures like micelles, bilayers or hexagonal phases, which can be explained
by their amphiphilic character [18]. The mobility of lipids depends on their shape, and
more importantly on the solid-crystalline phase transition. In biological membranes lipids
are found in the crystalline (also called liquid) phase, where lipids have freedom to
move. The phase transition or melting temperature (Tm) between the gel and fluid
phase is determined by the structure of the lipid. The longer the acyl chain of the lipid
the higher the Tm, while a higher degree of unsaturation of the bonds leads to lower Tm
[19]. The transfer of molecules through the dynamic assembly of fluid bilayer is highly
selective. While water can travel relatively fast (30-40 μms-1), the membrane is
impermeable to ions like Na+ or K+. Nonetheless, the permeability is influenced by
membrane components such as channel proteins or cholesterol molecules. It has been
suggested that cholesterol molecules increase the conformational order of the
membrane at temperatures above the lipids Tm [20].
19
Phospholipids are named following the designation for their three components. This
work describes the construction of membrane systems using the lipids 1,2-Dimyristoyl-sn-
glycero-3-phosphocholine (DMPC), 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine
(POPC) and 1-palmitoyl-2-oleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] (sodium salt)
(POPG), as depicted in figure 2.1. The analysis of lipids by infrared (IR) spectroscopy
allows identifying the functional groups of the different phospholipids [8] (normal mode
theory, vide infra). The normal modes of the (CH2) groups of the acyl chains offer
information about the organization of lipids in the bilayer. These modes correspond to
the symmetric and asymmetric stretching of (C-H), scissoring, wagging, twisting and
Figure 2.1. Structure of the phospholipids used in this work: DMPC, POPC and POPG. The three parts that
define the name of the phospholipid are indicated on the top of the figure. Below each phospholipid there is
the respective phase transition temperature (Tm). The unsaturated degree and length of the acyl chains are
indicated as nº of carbon atoms:nº of double bonds for each chain, so that 18:1 denotes 18 carbon atoms
and one double bond. Additionally, the corresponding charge at neutral pH of the headgroup is shown
below each phospholipid.
20
rocking. While the stretching modes are related to the physical state and hydration of
the lipids, the deformation modes in addition to the phase-dependence, give
information about the organization and interaction of the acyl chains [8] [19]. The
stretching modes of the (CH2) and (CH3) groups from the lipids are intense in the IR
spectrum between 3050 and 2800 cm-1. The carbonyl vibrational modes (C=O) appear
between 1800 and 1700 cm-1. The phosphate group in the lipid headgroup contributes
to the IR spectrum as well, with the asymmetric and symmetric stretching of the (R-O-
-P-O-R‘) modes below 1300 cm-1 [8] [19] [21].
2.1.2. Nanodisc membrane systems
In 2002, Sligar and coworkers introduced a novel membrane system of 10-20 nm in
diameter, composed by patches of lipid bilayers encircled by an amphiphilic
membrane scaffold protein (MSP) that avoids the spontaneous formation of vesicles
[10]. There have been advances in the construction of such models, improving the
control in size, composition and functional modifications [22]. Nanodisc systems are
used as water-soluble nanocarriers for membrane proteins, providing a native-like
environment [12]. The advantage of this kind of model is that the scaffold protein can
be modified by a His-tag, allowing a proper immobilization on a metal surface for
spectroscopic analysis, while the target protein remains in its natural form (see figure
2.2). In this case, a Ni-NTA (Nickel-nitrilotriacetic acid) monolayer was used as a linker to
anchor the nanodisc via the MSP His-tag to a Gold film (Au-film). This monolayer works in
Figure 2.2. Schematic representation of Membrane protein embedded in a nanodisc composed of lipid
bilayer and membrane scaffold protein (a). (b) The membrane protein folds into the nanodisc during or after
in-vitro expression. Then, the protein/nanodisc is tethered to the Ni-NTA SAM modified surface via the His-tag
at the terminus of the scaffold protein. (Reprinted from Biochimica et Biophysica Acta (BBA)-Biomembranes,
1828 (10), K. Ataka, Surface-enhanced infrared absorption spectroscopy (SEIRAS) to probe monolayers of
membrane proteins, 2283-2293, Copyright (2013), with kind permission from Elsevier) [12]. (C) Molecular
structure of the NTA molecule used as SAM to functionalise the Au-film.
(c)
21
such a way that first, the NTA monolayer is assembled onto the Au-film, followed by the
chelation of Ni2+ ions by the three carboxylate groups of the NTA molecule (see figure
2.3). The Ni2+ ions serve as linkers between the NTA and His residues of the MSP to
anchor the nanodisc system. Some of the benefits that nanodisc systems offer
compared to other lipid systems are the MSPs, which avoids fusion and controls the size
of the assemblies enhancing the stability and solubility in aqueous solution of the
nanodisc with the integrated membrane protein [22] [23].
The nanodisc system used in this project was composed of DMPC lipids and MSP1E3D1
as scaffold protein, for the study of the microbial membrane protein channelrhodopsin
isoform II (ChR II). Studies of a variety of membrane targets using nanodisc systems
combined to different techniques prove the nanodisc versatility and stability. Some
examples are the applications in NMR [24] [25], chromatography [26], surface plasmon
resonance (SPR) [27] [28] and surface enhanced IR absorption (SEIRA) [14]
spectroscopy. Therefore, this is a promising mimetic membrane system for membrane
process investigation.
2.1.3. Tethered bilayer lipid membrane systems
Tethered bilayer lipid membrane (tBLM) systems are planar solid-supported constructs
that allow anchoring the membrane to a functionalized electrode, a Au-film in this
case. This membrane system presents the advantage that it can be built with tethered
lipids (see figure 2.4). The tBLM used in this study requires three components and a
Figure 2.3. Schematic representation of the chelating reaction of the Ni2+ ion with NTA molecule and His
residues. For matter of simplification, only the head group of NTA is shown in the figure. When Ni2+ ion is
added to the solution onto the immobilised NTA molecules, it adopts an octahedral coordination geometry.
In this first step, NTA carboxylate groups and nitrogen atom occupy four of the six free-spaces of Ni2+ ion
coordination sphere, leaving the last two to water molecules. The latter are replaced by the imidazole group
of His residues when the nanodisc with His-tagged tail is added. Thus, each Ni2+ ion acts as a bridge between
one NTA molecule and two His residues.
22
two-step process. First, clusters of a hydrophobic part such as cholesterol tethered
molecules and of a hydrophilic part are absorbed to form a mixed self-assembled
monolayer (SAM) on the Au-film. The hydrophilic part creates an aqueous reservoir
between the electrode and the lipid bilayer. Subsequently, lipid vesicles are added,
which spontaneously cover the SAM surface forming islands of bilayer lipid membranes
[9] [29]. The phase separation of the SAM molecules is the reason for the formation of
the lipid bilayer islands. Once the tBLM is stabilized, the system is ready for the study. The
molecules that form such a construct should be chosen and tested for each system in
order to obtain the conditions that fit best to the target. This project uses WK3SH
(synthesized by Wiebalck et al. [29]) as cholesterol tethered molecule,
6-mercaptohexanol (6MH) as spacer, and POPC and POPG for the lipid bilayer.
The WK3SH tethered molecule offers the important advantage of being IR-transparent
in the amide region of the spectra (1800-1500 cm-1), in comparison to other tethered
molecules that often interfere with the target biomolecule [29]. The benefits of tBLM
systems are the water reservoir beneath the lipids providing a native-like environment,
and the wide range of applications in the study of peptides, membrane proteins and
ligands [11] [13] [30].
2.2. Membrane proteins
Proteins are biopolymers (macromolecules) composed by a sequence of amino acids,
where monomers are linked to each other by a peptide bond (amide bond). In a
Figure 2.4. Scheme of a tBLM system used in SEIRA spectroscopy. The nanostructured Au-film is functionalized
with a mixed SAM that forms islands of WK3SH cholesterol and of 6MH spacer molecules. The phospholipid
molecules complete the system by forming areas of lipid bilayer on top of the aqueous reservoir created by
6MH. The thickness of the system is ca. 5 nm. Below, there is the structure of the WK3SH tethered-molecule
[29].
WK3SH
23
peptide bond, a carboxyl group (C=O) of one amino acid is linked to the amino group
(N-H) of the succeeding amino acid releasing a water molecule. There are four
structure-organizational levels in proteins. The primary structure of a protein is
represented by the amino acid sequence, starting from the amino end (N-terminal) to
the carboxyl end (C-terminal) [3]. The interactions between amino acids within the
protein determine the secondary structure, resulting in a three-dimensional (3D)
assembly. The most common secondary structures are α-helix and β-sheet, random coil
and turn. The α-helix is a right-handed spiral with a distance of 0.15 nm between two
adjacent amino acids, a turn angle of 100º with respect to the centre and with 3.6
amino acids per turn [31]. β-sheets are formed by at least two strands of the
polypeptide arranged into pleats. The strands can have a parallel or antiparallel
orientation to each other and are connected by inter-strand hydrogen bonds [3].
Proteins may assemble into tertiary structures containing more than one secondary
structure (same kind or different) interconnected to each other. In many cases, proteins
are organised in aggregates of multiple subunits (homo or heterodimers), which define
the quaternary structure. As in secondary and tertiary structures, also in the quaternary
arrangement the stabilisation depends on the interaction between the side-chain of
amino acids. Some of these interactions may include hydrogen bonds,
disulphide-bridges and salt bridges [3] [12].
The adopted secondary structure can determine the family or type, and consequently,
the function of a protein. Membrane proteins include a wide array of protein families
involved in cellular processes, i.e. metabolites and ion transport (see figure 2.5), cell
signalling, ATP synthesis, and many others [22]. Channel proteins are an important kind
of membrane proteins, which are responsible for ion-trafficking between the intra- and
extracellular medium. The main function of ion channel proteins is gating the ion
Figure2.5. The function of the sodium-potassium pump is an example of primary active transport. The two
carrier proteins on the left are using ATP (adenosine triphosphate) to move sodium out of the cell against the
concentration gradient. The proteins on the right are using secondary active transport to move potassium
into the cell. Energy from hydrolysis of ATP is directly coupled to the movement of a specific substance across
a membrane independent of any other species. On the right side, the concentration gradients of Na+ and K+
between the extracellular and intracellular spaces are indicated. (Scheme adapted from Membrane
transport proteins, Mariana Ruiz Villarreal (2007). With kind permission of Wikimedia Commons). [32].
24
conductance for signal transduction processes [32]. Depending on the way these
proteins are activated they are divided into three classes: ligand-gated channels,
voltage-gated channels or mechanosensitive channels [33]. A distinguished and
relatively new class are channelrhodopsins, which are microbial membrane proteins
that work as light-gated ion channels [33] [34] (vide infra).
In this context, the structural study of membrane proteins helps characterising them,
offering information about the interactions that make their 3D folding possible. For
membrane proteins, a lipid or membrane-like environment is key for its assembly and
function [12] [35]. In most cases, these proteins lack the activity without the membrane.
This is the main reason why membrane proteins are highly difficult to investigate in
solution without a proper native surrounding, and why artificial membrane systems are
essential for their study.
2.2.1. Infrared spectroscopy of proteins
The molecular vibrations of the backbone motif of proteins can be characterise by IR
spectroscopy. One of the advantages of the technique is that it provides a distinctive
signal for the different functional groups. In that way, the vibration of the two main
groups in the protein would result in characteristic bands in the IR spectrum. In the
secondary structure of proteins, the periodicity of the backbone results in an addition of
the representative amide modes (figure 2.6) [36] [37]. The important amide band
appearing in IR spectra are amide I and amide II (description below). However, it is also
possible to identify characteristic vibrational modes from amino acid side chains, as in
the case of tyrosine (Tyr) or tryptophan (Trp) [38] [39]. The analysis of the amide I and II
modes, as well as the modes of the amino acid side chains, considering hydrogen
bonding and the coupling of the transition dipole moments, can be used for the
prediction of the polypeptide structure [39]. The following descriptions of the
polypeptide vibrations are based on the work from Krimm and Bandekar (1986) [36] and
Barth (2007) [39].
NH stretching (3300 and 3070 cm-1). This mode, also known as amide A, is localised in
the NH group and it remains unaffected by the conformation of the polypeptide. It is
the first component of the Fermi resonance with first overtone of the amide II in helices.
The second component is named amide B, and it is observed as a weak absorption in
the region between 3100 to 3050 cm-1.
Amide I (ca. 1650 cm-1). The major contribution to this mode corresponds to the (C=O)
stretching (76%), including minor contributions of the out-of-plane CN stretching, (CNN)
deformation, and (NH) in-plane bending (responsible for N-deuteration of the
backbone) [19]. Since the amide I mode depends on the secondary structure of the
protein backbone and is not significantly affected by the side chains, it is one of the
most relevant vibrations for structure prediction.
25
Amide II (ca. 1550 cm-1). This mode is the out-of-phase combination of the in-plane
(NH) bending (43%) and the (CN) stretching (14%), with small contribution from the
in-plane bending from (CO) (11%), (CC) stretching (9%), and (NC) stretching (8%) [19].
This mode has been less investigated than the amide I.
Amide III (1400-1200 cm-1). This mode is a in-phase combination of the (CN) stretching
(66%) and the out-of-plane (NH) bending (34%) [19]. It is sensitive to the side chain
structure and it frequently overlaps with other modes between 1400 and 1200 cm-1. In
some cases, it can as well contribute to predict the secondary structure of the
polypeptide.
One important fact for the study of biomolecules in aqueous solution by IR
spectroscopy, is the interference of the broad (OH) bending mode (ca. 1645 cm-1) of
water with the amide I and II region [37] [21]. One way to overcome the overlapping of
the OH band is using heavy water (D2O) solutions, which leads to a downshift of the
bending mode to ca. 1210 cm-1 [19]. Nevertheless, in some cases, performing IR
26
difference spectroscopy, where the background is subtracted from the protein
spectrum, can reduce the contribution of water.
27
2.2.2. Channelrhopsin II membrane protein
Channelrhodopsin II (ChR II) membrane protein is a microbial type Rhodopsin that
operates as a light-sensitive ion channel. Ion-channel proteins act as gates for ions in
the cell membrane. Their function is important since, as other transporters, they regulate
the traffic of ions between the inside and outside medium. There is a wide range of
processes where ion channel proteins are involved triggering the electrical signal
transduction, like it is the case of ChR II [40]. Channelrhodopsins (ChRs) are integral
proteins, with seven transmembrane α-helices (see representation in figure 2.7) [41].
ChR I and ChR II proteins were the first ones discovered in the green algae
Chlamydomonas reinhardtii. Both variants are responsible for the algae‘s phototaxis,
each responding to different light intensity. Nowadays, thirteen ChRs from different
algae have been isolated. Upon illumination, ChR II produces an increase of the
conductance of monovalent and divalent ions causing the depolarisation of the cell
membrane within milliseconds [40] [42]. The channel is highly selective to protons, up to
106 times compared to monovalent cations, and it does not diffuse anions [40].
Investigations of the photocurrents of the channel of ChR II suggest a pore diameter of
ca. 6.2 Å, where cations travel predominantly in their dehydrated form [43]. The
photoisomerisation from all-trans to 13-cis (C13=C14) of the retinal chromophore
covalently bonded to L257 (Lysine in 257 position) forming a protonated Schiff base,
induces the opening (or closing) of the channel by conformational and electrostatic
changes in the ChR II (retinal isomers shown in figure 2.8) [44]. This means that opening
and closure of this kind of ion channels can be controlled by light, which is an
exceptional characteristic. Optogenetics techniques, which mainly use ChR II variant,
take advantage of this unique feature to control neural circuits with light [45] [46].
Figure 2.7. Schematic representation of the light-gated ion channel channelrhodopsin II, and its seven
transmembrane helices. The hydrogen bond interaction between C128 and D156 is indicated connecting
transmembrane helices 3 and 4. The inset shows the environment around the retinal binding pocket. The
photoisomerisation of the retinal and consequent open and close reaction are depicted as well. (Reprinted
from Biochemistry, 49, Bamann C. et al. Structural guidance oft he photocycle of channelrhodopsin-2 by an
interhelical bond, 267-278, Copyright (2009), with kind permission from American Chemical Society) [41].
28
There have been several attempts to describe the structural changes of the transitions
involved in the light-cycle of ChR II, however, certain transition states are elusive to
detection due to the fast equilibrium between them. After photo-activation the protein
runs through intermediate states that eventually result in the relaxation of the system to
the dark state. Certain changes during these processes like protein conformation,
retinal interactions and protonation states can be observed by vibrational
spectroscopy [40]. The main features of the photocycle of ChR II are the activation of
the conductance of the channel and the closing induced by blue and green light,
respectively. With the objective to slow down the cycle, different mutants were
produced for characterisation based on its analogy to the well-known
bacteriorhodopsin. In that way, the mutation in the position C128 (Cysteine 128
position) of the ChR II sequence can extend the open state life-time of the channel up
to seconds or even minutes (preferably substitutions with Threonine, T; Arginine, A; or
Serine, S) [34] [44] [47] [48]. In this work, the C128S variant incorporated in nanodiscs was
studied by vibrational spectroscopy. According to Berndt and co-workers, his C128S
mutant displays the longest extension of the conducting state lifetime with 106 s versus
11.9 ms for the wild type, and a significant increase of the protein sensitivity to light [48].
These facts make this mutant (C128S) an interesting target to evaluate the application
of nanodisc membrane systems for the study of light-sensitive membrane proteins.
Figure 2.8. Retinal isomers of retinal proteins. In the bacteriorhodopsin dark-adapted state, either all-trans, 15-
anti or 13-cis, 15-syn retinal is observed, whereas 13-cis, 15-anti retinal occurs in the M intermediate structure.
In bovine rhodopsin, the switch between all-trans, 15-anti and all-trans, 15-syn is capable to switch between
the active form and its thermal predecessor. In the case of ChR II, the predominant transitions are between
all-trans, 15-anti (dark state) and 13-cis, 15-anti (conductive state). (Adapted and reprinted from The Journal
Of Biological Chemistry, 288(15), Ritter E. et al. Light-Dark Adaptation of Channelrhodopsin C128T
Mutant, 10451–10458, Copyright (2013), with kind permission from The American Society for Biochemistry and
Molecular Biology, Inc) [34].
29
2.3. Antimicrobial peptides
Peptides are biomolecules of shorter amino acid chain length than proteins, which as
well can adopt secondary structures. In nature, peptides have a wide range of
functions participating in cells communication, defence, repairing and death [3].
Antimicrobial peptides (AMPs) are those peptides that are implicated in the cell or
organisms defence mechanism. Therefore, the production of AMPs is one of the
processes through which cells protect and/or respond to foreign agents and changes
in the environment. Ageitos et al. (2016) offered a vast review of the diverse natural
sources of AMPs and some synthetic variants [49]. The activity spectrum of AMPs is
constantly increasing, with targets like parasites, bacteria, viruses and cancer cells [31].
There are peptides of different length, structure, species and mode of action. Their
importance has been rising because of their application as antibiotics substitutes to
overcome resistance. The first AMP discovered was gramicidin in 1939, isolated from a
Bacillus strain and effective against pneumococci infection [50]. The first AMP from
animal source was defensin from rabbit leukocytes, in 1956 [51]. Tyrothricin is the first
AMP for clinical use in humans (1939), composed of a mixture of tyrocidins and
gramicidin D produced by Bacillus brevis (Gram-positive aerobic spore-forming
bacteria) [52]. This discovery paved the way for AMPs to clinical use, and after more
than 60 years in use it has shown no microbial resistance [53].
The evolution of certain AMPs across the kingdoms of the phylogenetic tree has shown
that peptides from the same family have evolved depending on the environment of
the species producing them [54]. Controversial, but yet attractive examples are
defensins, for which the origin of this kind of peptides could be traced to a common
ancestor of 0.5 billion years old [54]. There are cationic peptides (mostly antibacterial)
[55], anionic AMPs, host defense peptides [56], α-helical antimicrobial peptides, and
many more that raise the number of discovered and synthesized AMPs to more than
5000 [57]. Due to the diversity in nature, structure and function (activity) of AMPs there is
a lack of a normalized classification of these peptides. However, for the benefit of the
studies presented in this work, the discussion is reduced to two classifications: by
structural features and by activity or mechanism of action.
From the structural point of view, α-helix and β-sheet conformations (see section 2.2. for
structural characteristics) are the most common and studied ones, while extended,
loop and circular structures are found in fewer cases [31]. As explained for proteins,
each assembly has structural features that are possible to assess by IR spectroscopy. In
some cases, the amino acid composition may influence the secondary structure of the
peptide. It has been suggested that there are certain amino acids that are important
for the AMP structural stability and/or activity, i.e. Zou et al. (2007) suggested that the
activity of the α-defensin-1 is significantly reduced with the substitution of arginine (Arg)
residues for lysines (Lys) [58]. Additionally, the positive net charge of cationic AMPs is
essential for the interaction with the negatively charged membrane of bacteria. Some
30
AMPs may adopt more than one of these motifs, or there are peptides that present one
structure in solution and a different one in contact with certain compounds, surfaces or
the cell membrane, like the peptide indolicidin [59]. An interesting and less studied
group of AMPs are the circular ionophores such as valinomycin, beauvericin and
enniatins [60]. The ring structure of enniatin peptides allows them to induce ion-transport
through the target membrane, which as well as for valinomycin and beauvericin is
believed to be their mode of action [61].
Another way to classify AMPs is by their activity or mechanism of action (see figure 2.9).
They have shown antiviral, antibacterial (mostly cationic AMPs), antifungal (mainly polar
and neutral amino acids), antiparasitic or anticarcinogenic activity [62]. However, the
mechanisms of action may differ from interacting with the pathogen or with the host,
modifying the host environment, membrane or gene expression [54]. In this investigation
the focus is mainly on membrane-active AMPs. The different lipid environments in
microorganisms may be one of the important factors that determine the
membrane-binding of AMPs [4]. Most of these peptides are amphipathic, containing
both hydrophilic and hydrophobic parts. In the same way as for membrane proteins,
the hydrophobic character allows the peptides to interact with the aliphatic chains
and allocate within the lipid bilayer. On the contrary, the ionic properties allow
interacting with the negatively charged head groups of some lipids, POPG for instance,
and/or with the aqueous medium in and outside the membrane [62]. Currently, a few
models have been described for membrane active peptides: carpet, aggregate, and
barrel-stave are some of the examples shown in figure 2.9. One of the most common
mechanisms of action is the pore formation followed by the permeabilization and
disruption of the membrane, like it has been suggested for one of the targets of this
thesis arenicin isoform 1 (A1) [63]. The amphipathic character of A1 and its net charge
of +6, explains the attraction to negatively charged lipids and the damage of the
membrane [64]. In general, factors like charge, amphipathicity, hydrophobicity and
H-bonding interactions are of important consideration for mechanism of action
estimation [65].
Gram-positive and Gram-negative bacteria are the most studied targets of AMPs.
Therefore, it is important to briefly describe their envelope composition based mainly on
the study offered by Malanovic and Lohner (2015) (see figure 2.10) [4]. The classification
into these two groups originates from the staining result of the bacteria cells, which is
either positive, if the cells take the stain (purple colour) or negative, if they cannot. Both
variants present a plasma lipid bilayer membrane composed of a combination of
zwitterionic (higher percentage in Gram-positive) and negatively charged
phospholipids (higher percentage in Gram-negative), followed by a stabilising layer of
peptidoglycan (PGN). The thickness of the PGN layer is significantly higher for
Gram-positive (ca. 40-80 nm) in comparison to the thin layer of Gram-negative (ca.
8 nm). Gram-negative bacteria present an outer lipid bilayer membrane, where the
inner layer is mainly composed of phospholipids and the outer layer of
lipopolysaccharides (LPS). The cell envelopes of these bacteria are represented in
31
figure 2.10. This figure also shows the membrane and cell wall of fungi, which is another
relevant target of AMPs. The interaction and role of the components of the bacterial
envelopes is an open and on-going discussion in the AMP community.
32
2.3.1. Enniatin B
For decades, natural and synthetic analogues of enniatins have been studied for their
antimicrobial applications. The attractivity of this family of non-ribosomal peptides lies in
their structure, which is achieved by alternating D-hydroxy acids and N-methyl-L-amino
acids (valine, leucine and isoleucine more common ones), forming a six unit ring.
Enniatins are toxins produced by various species of Fusarium fungi, and they belong to
the group of cyclohexadepsipeptides [61]. The particular N-methylation of the L-amino
acid takes place in the module 2 of the enniatin synthetase (ESyn) enzyme, previous to
the final cyclization of the peptide [66]. Gaumann et al. (1947) reported the discovery
of the first enniatin, enniatin A, found in the fungus Fusarium orthoceras var. enniatinum,
active against bacteria, fungi and plant shoot [67]. Nowadays, there exist
approximately 29 variants, plus many others that have been biosynthesized to
investigate biochemical applications. The most common ones are enniatin A, A1, B and
B1. They act against Gram-positive (thick peptidoglycan layer, but without outer
membrane) and Gram-negative (with outer membrane and thinner peptidoglycan
layer) bacteria, fungi and cancer cells [68]. Laboratories Servier (French company)
developed the first drug with enniatins mixture from Fusarium lateritium WR as active
Figure 2.10. Cell envelopes of various microbial families. (Adapted and reprinted from Biochimica et
Biophysica Acta (BBA)-Biomembranes, 1858, N. Malanovic and K. Lohner, Gram-positive bacterial cell
envelopes:The impact on the activity of antimicrobial peptides, 936-946, Copyright (2015), with kind
permission from Elsevier) [4].
33
agent named fusafungine for oral or nasal administration, which presents antimicrobial
activity against numerous microorganisms that cause respiratory infections like the
Gram-positive Staphylococcus aureus [69] (out-of-market since 03.2016).
Enniatins function as neutral ionophores (ion-carrier without ionisable groups)
complexing ions and transporting them through the cell membrane [60]. This means
that the ―ion‖ is transported from the extracellular medium-lipid interface, through the
membrane, and to the lipid-intracellular medium interface. Therefore, the
concentrations of the ―ion‖ at each side of the membrane have a great impact in the
equilibrium of the reactions taking place [60]. Their high hydrophobic character makes
them membrane-active peptides, considered to be the mechanism of action for
enniatins [61]. Enniatin B (EB) variant is produced mainly by the Fusarium scirpi fungi. It is
composed by N-methyl-L-valine (Val) amino acid combined with the
D-hydroxyisovaleric acid (D-Hiv) (see figure 2.11). This peptide presents similarities of
structure and function with the well-known depsipeptides beauvericin, which contains
aromatic N-methyl amino acids, and valinomycin, which is twice the size of EB and
alternates L- and D- enantiomers of valine residues with D-Hiv and L-lactic acid.
Enniatins cannot fold into a 3D structure due to the dimension of the ring, but it can
form complexes of different stoichiometry (1:1, 2:1 or 3:2, EB:cation) with alkali,
earth-alkali and some transition metals in order to induce the ion-transport through the
membrane [61] (see figure 2.12). EB variant has shown potential antimicrobial [70],
antifungal [71], and anticarcinogenic [72] [73] activity, and against the malaria parasite
[70], which makes it an attractive target for the study of peptide-membrane interaction.
Besides the lipophilic character of EB, it has also shown to inhibit enzymatic function of
calcium-dependent proteins like kinases, by binding and inducing structural changes to
the calcium-binding messenger protein calmodulin [74]. Though its wide range of
suggested applications, EB toxicity to human normal cells is an open debate [75].
Figure 2.11. Structure of the cyclic AMP enniatin B. The D-hydroxyisovaleric acid and N-methyl-L-valine
residues are highlighted in red and blue, respectively. The N-methylation of the valine amino acids, binding
to the hydroxyl acid and cyclization reactions are carried out in a sequential manner by the enniatin
synthetase enzyme.
34
Even though EB structure has been extensively studied and many complexes have
been isolated and characterised in solution, there is scarce evidence about its
behaviour inside the lipid membrane. Ovchinnikov and coworkers (1974) offered a
detailed study on EB proving its ion complexation capability and predictions for the
stoichiometry of EB:ion complexes [61]. The structural predictions from Ovchinnikov were
supported by Kamyar et al. 30 years later by performing patch clamp experiments [76].
Zhukhlistova‘s (2002) X-Ray study of the EB complex with potassium thiocyanate (KNCS)
provided a deeper understanding of EB complexones [77]. The ion-selectivity sequence
for EB collected from previous studies presented in Pressman‘s review (1976) about
ionophores, reveals a preference of monovalent cations over divalent cations [60]. This
is in agreement with the suggestions made by Ovchinnikov (1974) [61] and Kamyar
(2004) [76], with an extremely low K+/Na+ selectivity compared to beauvericin and
valinomycin.
2.3.2. Arenicin 1
Arenicin is a family of AMPs that act against Gram-negative and Gram-positive
bacteria, and fungi. To date, there are three variants known as arenicin1 (A1) and
arenicin2 (A2), discovered in 2004 by Ovchinnikova et al. [64], and arenicin3 discovered
in Denmark by a pharmaceutical firm bond to Novozymes (AdeniumBiotech). These
peptides are derived from the marine polychaeta lugworm, Arenicola marina.
Ovchinnikova and coworkers (2004) claimed that arenicins have not sequence match
with any AMP family known to that date [64]. Both isoforms have 21 residues composing
a two-stranded antiparallel β-sheet, rich in hydrophobic amino acids and arginines
(Arg), and with one disulphide bond between cysteines Cys3 and Cys20 closing the
18-loop providing the peptide of a unique structure with a net charge of +6 [63] (see
figure 2.13 for structure). Due to its rather globular arrangement, and controversial
behaviour in solution and membrane environment, the isoform 1 (A1) of the arenicin
family has presented more challenges to find a comprehensive description. Numerous
Figure 2.12. Schematic representation of the EB:ion complexes stoichiometry, 1:1, 2:1, 3:2 suggested by
literature [61]. The M+ correspond to monovalent metal ions. Divalent ions can form complexes as well, but
were not studied in this work. The hexagons figures represent the EB ring structure.
35
studies have already been carried out to elucidate the 3D conformation of A1,
proposing different mechanism of interaction with the cell membrane [78] [79], and
even conformational changes from solution to lipid bilayer [63]. In initial biological
studies, arenicins showed similar antimicrobial activity compared to the strong antibiotic
peptide protegrin-1 [64] and antifungal activity [80]. Cho and Lee (2011) suggested that
A1 promotes the intracellular production of reactive oxygen species (ROS) and
apoptotic reactions [81]. The same authors offered an interesting study of A1 using
bacteria mimetic liposomes, where they observed that A1 provokes pronounced
leakage in liposomes in comparison to other analogues, estimating a pore formation of
at least 3.3 nm radii [78]. They also explained the role of the Arg residues for the A1
antibacterial activity, with emphasis in the Arg11 and Arg19 of the turn and C-terminus,
respectively. In summary, A1 activity indicates a critical perturbation of the cell
membrane. Andrä and coworkers (2008) suggested that A1 presents higher
antimicrobial activity at 4°C and 37°C, and a strong adaptation to high salt
concentration environments. Contrary to other authors, they found no significant
selectivity of the peptide between zwitterionic and negatively charged lipids [63],
which is believed to be the peptide‘s advantage towards low human cells toxicity [82].
Some of the mechanisms of action suggested for A1 so far are: lipid-microdomains
formation, toroidal pores, carpet model, among others. Because of its relatively recent
discovery, elusive conformation and interesting lipid interaction, there is room for new
hypothesis of modes of action and structure of the antimicrobial peptide A1.
Figure 2.13. 3D representation of the A1 structure (PDB 2JSB [63]). Hydrophobic amino acids are marked in
red, while hydrophilic residues are marked in blue. The yellow line connecting the two strands represents the
disulphide bridge between Cys3 and Cys20. The one letter code amino acid sequence is shown bellow the
representation denoting: R as Arginine, W as Tryptophan, C as Cysteine, V as Valine, Y as Tyrosine, G as
glycine and L as leucine.
36
2.4. Theory of vibrational spectroscopy
The theory provided here is mainly based on the book Vibrational Spectroscopy in Life
Science by Siebert and Hildebrandt (2008), and Dr Jacek Kozuch‘s dissertation,
Structure-Function Relationships of Membrane Proteins - Spectroelectrochemical
Investigation of Artificial Membranes [37] [21]. This section intends to offer the
fundamentals of vibrational spectroscopy, Fourier transform infrared (FT-IR), attenuate
total reflectance infrared (ATR-IR) and surface-enhanced infrared absorption (SEIRA)
spectroscopy, which were used in the projects presented in this work for the study of
biomolecules.
Vibrational spectroscopy probes the interaction of electromagnetic radiation with
matter. This interaction causes transitions between the vibrational states of molecules.
The roots of this technique arose in 1800, when William Herschel discovered the infrared
(IR) radiation. He measured the temperature of the different colors of the rainbow
spectrum that the sun-light created by passing through a glass prism, and found the
highest temperature in the region beyond the red color, which was not visible.
However, IR spectroscopy was not established as a technique after years later,
obtaining the first IR spectra from probing matter with light
2
. In 1920, the Indian physicist
Sir Chandrasekhara V. Raman discovered the phenomenon of light scattering, later
called Raman scattering. The Raman effect causes a change on the wavelength of
part of the deflected light due to its interaction with matter. Initially, vibrational
spectroscopy was used only for the study of small molecules and inorganic materials.
Innovations in interferometers, invention of lasers, and improvements in purification
methods for biological samples allowed the application of vibrational spectroscopy in
life science.
Its wide application scope (organic and inorganic samples in industry and research)
3
and capability to combine with other techniques situates vibrational spectroscopy
within the most important techniques to study biomolecules. For instance, it can be
combined with time-resolved techniques to study biological processes like biocatalysis
or ligand binding [83]. Nowadays, Raman and IR can be used to probe and monitor
structure-function relationship of macromolecules with higher resolution, and with lesser
limitations in sample form and size than for other techniques. These characteristics are
certainly reflected in the use of surface-enhanced Raman (SER) and surface-enhanced
IR absorption (SEIRA) spectroscopy, where the metal surface properties are enhanced,
acting also as an electrode to probe electrochemical processes.
2
It was Herschel‘s son who 40 years later recorded the first IR spectrum from an alcohol-wetted paper.
3
Vibrational spectroscopy applications: quality control, forensic analysis, art and archaeology, NASA experiments, dynamic
and reaction progress analysis, structure elucidation, among many others.
37
2.4.1. Basics of vibrational spectroscopy
The interaction of electromagnetic radiation with matter provokes transitions between
the vibrational states of molecules, which can originate from a resonant absorption of
IR radiation (10-12500cm-1) or from inelastic scattering. If the energy of a photon
from the polychromatic light matches the difference between the initial and final
vibrational state, it results in IR absorption spectroscopy (eq. 2.1).
(2.1)
Raman spectroscopy uses monochromatic light to induce an inelastic scattering,
where the energy of the scattered photon differs from that of the incident light
(eq. 2.2).
(2.2)
The information obtained from both techniques is considered to be complementary,
due to the difference in their underlying mechanisms. IR-active vibrations are those
where the vibrational mode is accompanied by a change of the dipole moment;
Raman-active vibrations require a change of the polarizability. IR and Raman can be
represented in similar way for comparison, taking into account that on the ordinate of
the spectrum the intensities are measured in different way for IR (absorbed light) or
Raman (scattered light). The energy of the vibrational transition represented in
wavenumbers ( in cm-1) on the abscissas, refers to the frequency of the absorbed light
in IR, and to the frequency difference between the exciting and scattered light in
Raman.
2.4.1.1. Molecular vibrations and the harmonic oscillator
The vibrations of a diatomic molecule A-B treated as a harmonic oscillator is frequently
used to explain the fundamentals of molecular vibrations (see figure 2.14). The
harmonic oscillator of two masses mA and mB connected by a spring in equilibrium
position along the x-axis can be described by Hooke‘s law (eq. 2.3), where the restoring
force counters the displacement provoked by the force constant (representing
the strength of the bond between A and B). Accordingly, the potential energy and
kinetic energy , which depends on the reduced mass (eq. 2.6) and the velocity of
their motion
, result in equations 2.4 and 2.5, respectively. The sum of the first
derivatives of and must equal zero (eq. 2.7) to obey the conservation of the
energy, leading to the Newton‘s equation of motion (eq. 2.8). This equation can be
solved by the cosine function (eq. 2.9), where and are the amplitude, circular
frequency (eq. 2.10), and phase, respectively. Thereby, one obtains equation 2.11 from
applying Hooke‘s law to chemical bonds, which indicates that the of a harmonic
vibration increases with the strength (or rigidity) of the bond, but decreases with the
increase of the masses of the atoms involved.
38
(2.6)
(2.7)
(2.8)
(2.9)
√
(2.10)
√
(2.11)
(2.3)
(2.5)
(2.4)
39
2.4.1.2. Normal modes
In a normal mode of a molecule all atoms vibrate in a defined manner with the same
frequency, but different amplitudes. The number of normal modes in a molecule
corresponds to the vibrational degrees of freedom. A non-linear (linear) molecule with
atoms has ( , it can only rotate in two axes) degrees of freedom in a
Cartesian coordinate system.
To describe the normal modes of a given molecule, the displacements of all atoms
have to be considered. The use of the mass-weighted Cartesian displacement
coordinates (eq. 2.13) in the kinetic energy (eq. 2.12), results in 2.14. The potential
energy involves all possible interactions between all atoms (covalent, electrostatic,
van-der-Waals…), resulting in a more complex term that can be expressed as a Taylor
series (eq. 2.15).
Since we are only interested in changes of due to the displacements of the atoms,
the first ( at equilibrium) and second (infinitesimal changes in do not change )
terms in 2.15 are zero. Moreover, higher order terms can be neglected according to the
harmonic approximation, simplifying expression as expressed in 2.16. Therefore, 2.17
can be obtained after substitution into Newton‘s equation (eq. 2.8), which is composed
by linear second order differential equations and its general solution (eq. 2.18).
Thus, one obtains solutions for corresponding to frequencies
. Finally, the
frequencies of the normal modes are ( ), since 6 solutions (5 for linear
molecule) equal zero as they refer to the translation and rotation of the molecule. The
amplitudes of the displacement of each atom for every normal mode can be
calculated with the obtained frequencies. As mentioned in the beginning, a normal
mode represents an in-phase-oscillation of the entire molecule with a given frequency,
but different amplitudes of certain segments of the molecule. As these amplitudes may
∑ *(
) (
) (
) +
∑(
)
∑(
)
∑(
)
√ √ √ )
(2.12)
(2.13)
(2.14)
(2.15)
40
differ substantially, these normal modes may be reduced approximately to the part of
the molecule with the most pronounced motion, in some cases being ascribed to
specific group vibrations or single bonds.
The mass-weighted Cartesian coordinates can be converted into normal coordinates
(for normal modes) by the use of an orthogonal transformation 2.19, in order to
simplify the description of the probability of vibrational transitions. Choosing the
transformation coefficient in such a way that and assume the shape of equations
2.14 and 2.16, and the potential energy does not depend on the cross products
leads to the solution of the Newton equation 2.20.
∑(
)
∑
∑
√
(2.16)
(2.17)
(2.18)
∑
√
(2.19)
(2.20)
Figure 2.15. Schematic representation of the normal modes of water molecule. The water molecule is
composed by one oxygen atom (O, black sphere) covalently bound to two hydrogen atoms (H, golden
sphere). The number of normal modes of a non-linear molecule is given by . Therefore, water exhibits
normal modes. The arrows represent the displacement of the atoms in each normal mode.
41
2.4.2. Infrared absorption spectroscopy
The IR absorption is shown as the absorbance (in OD = optical density) by the use of
the Lambert-Beer law (eq. 2.21). Here, and are the intensity of the IR radiation
after passing through a solution with and without the analyte. The resulting absorbance
is dependent on the conditions during the experiment reflected by the concentration
and the optical path length , as well as on the molar absorption coefficient
including the quantum mechanical probability of the transition between the initial and
final vibrational state (eq. 2.1). The latter is given by the transition dipole moment (eq.
2.22), where and are the wave functions of the final and initial vibrational state
and the operator of the electric dipole moment (eq. 2.23). In 2.23, for each atom ,
and refer to the charge and to its distance to the center of gravity of the
molecule, respectively. The prerequisites for an IR-active absorption can be identified
by expanding the operator of the electric dipole moment in a Taylor series with
respect to the normal coordinates . For a harmonic oscillator only the linear terms of
the Taylor series have to be considered, so that results in 2.24. Thus, the transition
probability transforms into 2.26.
Due to the orthogonality of the wavefunctions and , the first term of equation 2.26
equals zero and, therefore, only a non-zero transition probability (IR-active transition) is
achieved when the second term is non-zero. This is given when (i) the electric dipole
moment of the molecule changes during the vibrational displacement of the atoms
(
), and (ii) the quantum number between the states and differ by one within
the harmonic approximation (⟨
|
| ⟩ ). This consideration holds for all three
Cartesian coordinates ( ), so that the absorbance of unpolarized light of
(
)
[ ] ⟨
|
| ⟩
∑
∑
(
)
[ ] ⟨
|
| ⟩
⟨
| ⟩ ∑
⟨
| | ⟩
(2.21)
(2.22)
(2.23)
(2.24)
(2.25)
(2.26)
42
randomly oriented molecules arises from the sum of the transition probabilities along all
three components (eq. 2.27). However, using linear polarized light on an oriented
sample allows addressing the individual components of the transition dipole moment
[ ] and, by this, obtaining more detailed information about the studied system.
2.4.3. Fourier transform IR
The principle of the Fourier Transform (FT) IR spectroscopy is applied in nearly all of
today‘s IR spectrometers. Due to this approach, the measuring time can be drastically
reduced, in contrast to the previous dispersive technique, leading to an improved
signal-to-noise ratio.
The Michelson Interferometer is the central building block of an FT-IR spectrometer
(figure 2.16), which comprises a beam splitter and two plane mirrors, i.e. a fixed and a
movable mirror, oriented perpendicularly to each other. The beam splitter transmits and
reflects the incoming polychromatic IR radiation, ideally divided in half, onto the two
([ ]
[ ]
[ ]
)
(2.27)
Figure 2.16. Michelson Interferometer: The IR radiation passes a beam splitter where it is divided into two parts
and subsequently recombined directed at the sample. The movable mirror varies the optical path difference
between the two beams generating an interferogram of the detected IR radiation. (Adapted and reprinted
with kind permission from Jacek Kozuch (2013). Structure-Function Relationships of Membrane Proteins -
Spectroelectrochemical Investigation of Artificial Membranes), Technische Universität Belin, Fakultät II [21].
43
mirrors and subsequently recombines both beams directing it at the sample. With an
optical path difference of , the distance between both mirrors and the beam
splitter is equal leading to no phase difference between both beams and, therefore,
the outgoing radiation equals the incoming one. Upon displacement of the movable
mirror, however, the beams interfere with each other, so that each wavelength
undergoes alternately constructive (
) and destructive interference (
), and results in a -dependent cosinusoidal modulation reflecting the respective
wavelength . The signal that is accumulated on the detector is a superposition of the
cosine functions of all frequencies of the polychromatic IR radiation and is referred to as
an interferogram (figure 2.16).
The IR spectrum (frequency domain) is obtained after transforming the optical path
length dependent interferogram (time domain) by use of the Fourier transform. This IR
spectrum displays the attenuation of the IR radiation depending on the wavenumber
(reciprocal wavelength) 2.28.
This integral assumes an infinite motion of the movable mirror. Since the effective
optical path difference is restricted to only a few centimeters, the interferogram has to
be multiplied with an appropriate apodization function (i.e. a triangular function) to
bring the edges of the interferogram smoothly to zero. With this procedure, artifacts in
the spectrum are suppressed, but also the shape of spectral bands is manipulated. The
advantages of FT-IR spectroscopy over the dispersive approach are presented in the
box below.
∫
(2.28)
Multiplex or Felgett advantage. On a FT spectrometer, the signal-
to-noise ratio improves by √ for a spectrum comprised of
elements since the total noise is distributed over the entire
spectral range, contrary to dispersive spectrometers where the
complete noise intensity is recorded at each spectral data point.
Throughput or Jacquinot advantage. In FT-IR, the implementation
of Jacquinot circular apertures (to restrict convergent rays) results
in higher signal-to-noise ratio, in comparison to the slits used in
dispersive spectrometers.
Connes advantage. The laser beam in a FT-IR instrument acts as
an internal calibration of the mirror position providing a more
precise wavenumber of spectral features than in dispersive
spectrometers.
44
2.4.4. Attenuated total reflectance IR
The discovery of the attenuated total reflectance IR (ATR-IR) spectroscopy set the basis
of the development of SEIRA. In this work, ATR-IR was used to investigate the structure of
peptides in solid state by using a diamond-ATR durascope, which is a single-reflectance
ATR with a diamond of ca. 2 mm diameter that is pressed against the sample.
Additionally, SEIRA spectroscopy technique was employed in the ATR mode to perform
the protein/peptide-membrane interaction experiments.
The phenomenon of total reflectance only takes place at the interface of the internal
reflectance element (IRE) to an optically less dense medium at an angle of incidence
higher than the critical angle . Silicon ( ), germanium ( ), or zinc
selenide ( ) are the most common materials for the IRE, due to a higher
refractive index than biological systems ( ). The propagation of an evanescent
wave through the interface into the less dense medium is a by-product of the total
reflectance. Interestingly, the amplitude of the evanescent wave decays exponentially
in the direction normal to the surface (i.e. in z-direction) eq. 2.29 [84]. In 2.29,
represents the penetration depth of the evanescent wave at which the amplitude
decayed to ca. 37 % ( ) of its initial value. It depends on the wavelength of the
radiation , the ratio of the refractive indices
⁄ (of the IRE and the optically rare
medium, respectively), and it is inversely proportional to the angle of incidence , as
expressed in 2.30.
Figure 2.17. Schematic configuration for attenuated total reflectance-infrared (ATR-IR) spectroscopy. The
incoming IR beam, composed of a perpendicular (s) and parallel (p) component (in respect to the plane of
incidence) of the electromagnetic field, experiences total reflexion at the interface to an optically less dense
medium, i.e. the sample, if the incident angle exceeds the critical angle . (Adapted and reprinted with kind
permission from Jacek Kozuch (2013). Structure-Function Relationships of Membrane Proteins -
Spectroelectrochemical Investigation of Artificial Membranes), Technische Universität Belin, Fakultät II [21].
45
The is in the range of the wavelength of the incident radiation. For instance, in the
case of a Si prism, an incident angle of 60° (as used in this work), and a spectral region
of 1000 to 4000 cm-1 (10-2.5 μm) the penetration depth of the evanescent wave is
between 2.6 and 0.7 μm, respectively. An absorbing medium placed onto the surface
of the IRE can couple with the electric field of the evanescent wave, absorb energy of
the radiation and, thus, attenuate the total reflected beam. The parallel (p) and
perpendicular (s) components of the incident radiation (in respect to the plane of
incidence) causes a polarization of the evanescent wave in the x- and z-directions, as
well as in the y-direction [84].
The important advantage of the ATR technique is the possibility to adsorb the sample
on the IRE surface and study its structure and dynamics. This means that the
supernatant buffer solution can be exchanged very easily to study effects of the
experimental conditions (pH, ionic strength etc.), binding of substrates and ligands.
Additionally, preferably oriented samples can be adsorbed on the IRE to obtain
additional information about the orientation of structural elements, such is the case of
membrane proteins.
2.4.5. Surface-enhanced IR absorption SEIRA spectroscopy
Hartstein et al. discovered in 1980 the SEIRA effect, which was observed from
contaminant hydrocarbons while measuring aromatic carboxylic acids adsorbed on
Ag (silver) and Au films by ATR-IR [85]. The SEIRA effect presents certain mechanistic
analogy to surface-enhanced Raman scattering (SERS). Accordingly, one can define
SEIRA as a spectroscopic technique that uses a nanostructured metal film, in this case
Au, in the interface between the IRE (i.e. a silicon prism) and the sample. The plasmonic
resonance of the nanostructured Au results in an increase of the IR signal in 10-100 times
[86]. This enhancement is restricted, presenting a steep decay of the signal at
approximately 8 nm from the metal surface. Furthermore, there is also an orientation
selection rule with a strong enhancement along the normal of the Au surface. The
Au-film, which also functions as the working electrode, can be functionalized with a
SAM for a wide range of applications, such as oriented proteins, membrane systems,
electrochemical reactions or potential dependence of protein or peptides, among
many others [11] [14] [29] [87] [88] [89].
(
)
√ (
)
(2.29)
(2.30)
46
The electromagnetic (EM) mechanism is believed to contribute to the enhancement of
the IR absorption of the probe molecule, due to similarities to the SERS effect. This
enhancement is conditioned by metal islands morphology, which can be represented
as ellipsoidal metal particles (see figure 2.18) [90]. The box below offers a brief
description of the nature of the enhancement mechanism for the SEIRA effect.
(
)
Electromagnetic (EM) mechanism
The incident photon field polarizes the metal particles through excitation of
collective electron resonance (localized plasmon modes). The induced
dipole in the particles creates a local EM field around them, which is
polarized perpendicularly to the metal surface (surface selection rule,
vibrations associated with changes parallel to the surface are invisible).
The interaction of the adsorbed molecules with the created EM field results
in the transitions between vibrational states. The field enhancement decays
with the distance to the metal surface, which explains the short-range
effective enhancement. The enhancement factor can be expressed at a
distance and a nanoparticle radius .
The dielectric function of the metal can be affected by the oscillating
dipoles of adsorbed molecules. This perturbation of the optical properties of
the metal results in induced dipoles in the metal particles, causing an
effective enhancement of the IR absorption due to the much larger
absorptivity of metals.
(2.31)
Figure 2.18. Schematic representation of the EM mechanism of the SEIRA effect originating from ellipsoidal
metal particles. The electric field component of the incident IR radiation polarizes the metal islands. The dipole
generates an enhanced local electric field around the particles that excites the molecular vibrations of the
absorbed molecules. Furthermore, the molecular vibrations induce an additional dipole and perturb the
optical properties of the metal. (Adapted and reprinted with kind permission from Jacek Kozuch (2013).
Structure-Function Relationships of Membrane Proteins - Spectroelectrochemical Investigation of Artificial
Membranes), Technische Universität Belin, Fakultät II [21].
47
2.5. Theory of electrochemical impedance
spectroscopy
Electrochemical impedance spectroscopy (EIS) is a method that allows measuring the
resistance and the capacitance of interfacial layers. These parameters are related to
the ability of a system to resist the flow of current and to accumulate electrical energy,
respectively. EIS is relevant for characterisation in material science. However, nowadays
it is applied in a much broader scope [91], like the study of monolayers and membrane
systems [29] [92]. The capacitance is associated with the area of the electrode and
the distance :
where and are the constant electrical permittivity in vacuum and the characteristic
ability of a material/medium to store electrical energy [21]. The most common way of
performing EIS is to apply an alternating voltage of small amplitude to the electrode
and record the phase shift and amplitude, or the real and imaginary parts, of the
resulting current. By scanning the frequency (in the range of 1 mHz and 1 MHz) a
complete spectrum can be measured. EIS can be described as the interaction of the
dielectric medium or the analyte with the externally applied and alternating electric
field. In this work, EIS is combined with SEIRA with the aim of characterising the SAMs
and the tBLM systems employed for the study of the targets.
2.5.1. Basic theory of the electrical impedance
Electrical impedance
The electrical impedance is defined as the frequency-dependent resistance [93],
expressed as the ratio of the alternating voltage and the resulting current 2.33. This
current , with a phase difference can be measured when
applying a monochromatic signal with a single frequency of
. Consequently, 2.33 is transformed into 2.34 [94]. For purely resistive behavior
is zero, and the responses of capacitive and inductive elements are
(2.32)
( )
(2.33)
(2.34)
48
[ ] and [ ] , respectively. The analysis of a system with these
differential equations can be simplified by the Fourier transform converting into the
frequency domain (which is also the variable in an EIS measurement). After the
conversion one obtains the voltage and current , as well as
the solution of the differential equations for a resistive 2.35, capacitive 2.36 and
inductive 2.37 behavior ( √ ).
Complex impedance
This impedance is a complex quantity (real and imaginary parts) that can be
represented as 2.38, where the real part (in-phase) and the imaginary part
(out-of-phase) are the resistance and the reactance, respectively. The reactance gives
information about the capacitive and the inductive of the system. The two
coordinate values of the Impedance vector plotted in a Cartesian system can be
written as 2.39 with the phase angle
and the magnitude of the impedance
| | √ . By using the Euler rule the impedance may be rearranged in polar
coordinates and expressed as 2.40.
Admittance
The admittance (eq. 2.41) is the inverse of the impedance. The real part is the
conductance and the imaginary part the susceptance . The partial fraction
decomposition (eq. 2.39) can be used to calculate them. Usually, and are
expressed by distinguishing between resistive and capacitive components in series
or in parallel . From this approach, the reactance
and the susceptance result in 2.43.
(2.35)
(2.36)
(2.37)
( ) ( )
| | | |
| |
(2.38)
(2.39)
(2.40)
49
2.5.2. Graphical representation of impedance data
The graphical representations explained in this work are based on the circuit presented
in figure 2.19 and the mathematical expressions described above. Here, Rsolution
represents the resistance of the electrolyte, and below are the resistance and
capacitance of the bilayer and spacer regions. This circuit serves as an
approximation of the electrical properties of a functionalised electrode in contact with
an electrolyte [93]. In this case, there are three important representations: Nyquist plot,
Bode plot and Cole-Cole plot. The latter is the most relevant for the analysis of the
systems presented in this work.
The Nyquist plot is the most common representation of the EIS data. It is based on the
complex impedance , which plots the imaginary part ( ) versus
the real part ( ) as function of the angular frequency .
The Bode plot, where the magnitude of the impedance | | and the phase difference
are plotted against the frequency , can be divided into three regions. The first and
third regions of the spectra correspond to the very high and low frequencies,
respectively. Here, the impedance is independent of the frequency and shows pure
( )
(2.41)
(2.42)
(2.43)
Figure 2.19. Scheme of the system used to study the dielectric properties of the SAMs and tBLM systems. The
RBilayer-CBilayer element and RSpacer-CSpacer element described the hydrophobic bilayer region and the
hydrophilic spacer region, respectively. The system is supported on a Au electrode (by attachment of the
SAM) and in contact with the supernatant electrolyte above the bilayer region, adapted from [21].
50
resistive behavior (Q = 0). At high frequencies (first region) appears the low resistance
Rsolution and at low frequencies (third region) the sum of all resistances. In the middle
area, the impedance is frequency-dependent due to the presence of Cbilayer/Cspacer.
According to 2.36,
(by the phase approximation) and Cbilayer/Cspacer is
calculated by | | .
The Cole-Cole plot represents a direct analysis of the capacitances (Cbilayer/Cspacer) by
reading out the radius of the first semicircle (see figure 2.20), which is an advantage
with respect to other representations. In this case, the imaginary and real parts are
divided by the angular frequency . Due to the direct relation to the values, this plot
was the one used to characterise the properties of the systems studied in this work.
2.5.3. Graphical evaluation and physical relevance in tBLM
systems
It is important to mention that the resistance and the capacitance are the physical
properties that are used to characterise the membrane system, with the additional
component for the resistance of the electrolyte (in the high frequencies due to its low
magnitude). However, EIS can be used as well to monitor the capacitance of the
electrode, and the properties of the Helmholtz layer or the ionic reservoirs beneath the
bilayer. These terms can be neglected upon modeling the system, since these parts are
usually not found in the same frequency range as the resistance and capacitance of
Figure 2.20. Scheme of the system used to study the dielectric properties of the SAMs and tBLM systems. The
RBilayer-CBilayer element and RSpacer-CSpacer element described the hydrophobic bilayer region and the
hydrophilic spacer region, respectively. The system is supported on a Au electrode (by attachment of the
SAM) and in contact with the supernatant electrolyte above the bilayer region, adapted from [21].
51
the membrane. The presence of defects or pores, however, might intensify the
influence of the sub-membrane region of tBLMs [95].
These electrical properties of a system can be obtained manually or by modeling of the
studied system [96]. There exist two approaches, the continuum theory [94] and the use
of semi-empirical methods. Semi-empirical methods have the advantage of using
already established and optimized models for certain properties of the system. In both
cases, the system is approximated to an equivalent circuit that provides the
mathematical equations for a fit to the data. In this work, a non-linear least squares
fitting based on the Levenberg-Marquardt algorithm was applied [97]. The concepts
presented in the following blue box were used to describe the systems studied here [98]
[99] [94] [100].
(
)
Ohmic resistance of a membrane
The hydrophobic core of a lipid bilayer acts as a barrier for ions and other
molecules. Therefore, it can be treated as an isolator with the Ohmic resistance .
Capacitance of a membrane
The membrane can be described as a capacitor as shown in 2.32. Whereas
depends on the dimensions of the experimental assembly, the thickness (in a
bilayer it is in the range of 4-6 nm) depends on the phospholipid composition. The
dielectric constant of the hydrophobic core is , and of the hydrophilic
head group region (thickness ) is . Neglecting the contribution on
the latter term by the dielectric constant of water ( ) due to the solvation
level of the headgroups, the specific capacitance of a membrane with a
hydrophobic part of 4 nm is shown in 2.44, which matches the experimental results
Constant phase element
In the case of a non-ideal capacitive behavior, the capacitance can be replaced
by a constant phase element (CPE) that treats the dielectric constant as a
complex magnitude on the basis of the Debye theory. The advantage of this
approach is a better fit that describes the system more precisely. The resulting
impedance is expressed in 2.45, where is a parameter with the property
. At , this approach describes an ideal capacitor and equals the
capacitance (see eq. 2.36). All other cases yield the quantity with the unit
, which cannot be compared directly with . Fortunately, both values might
be related to each other when the CPE-element lies in parallel to an ohmic
resistance (2.45). The term is the frequency at which the imaginary part Im reaches
its maximum value.
(2.44)
(2.45)
52
3. Experimental section
3.1. Materials
Chemicals
Provider
1PrOH
Merck
6MH
Sigma Aldrich
Ag/AgCl reference electrode
World Precision Instruments, Inc.
CHCl3
Sigma Aldrich
CsCl
Merck
EtOH
Sigma Aldrich
Extruder
Avestin
HClO4
Merck
HF
Sigma Aldrich
MeOH
Sigma Aldrich
NaCl
Sigma Aldrich
NaOAc
Fluka
Na2S2O35H2O
Sigma Aldrich
Na2SO3 anhydrous
Sigma Aldrich
NaAuCl43H2O
Sigma Aldrich
NH4Cl
Sigma Aldrich
NH4F
Sigma Aldrich
NiSO4
Sigma Aldrich
Dithiobis(C2-NTA)
Dojindo Laboratories
KCl
Sigma Aldrich
POPC/POPG
Avanti Polar Lipids
Polycarbonate filters (100nm)
Avestin
All chemicals were of highest purity grade available.
Nanodisc samples containing the ChR II C128S mutant and the ChR II C128S mutant
samples in solution were provided by Dr. Michael Szczepek from the Institute of Medical
53
Physics and Biophysics, Charité (Berlin, Germany). The AMP samples of enniatin B and
arenicin 1, were provided by Dr. Lennart Richter from the synthetic biotechnology and
antibiotics group (Prof. Roderich Süssmuth), Teschnische Universität Berlin (Berlin,
Germany), and by Prof Dr. Thomas Gutsmann group from the Division of Biophysics,
Research Center Borstel, Leibniz-Center for Medicine and Biosciences (Borstel,
Germany), respectively.
Detailed description of the synthesis and characterization of the tethered molecule
WK3SH can be found in the supplementary information of the article by Wiebalck (2016)
[29].
3.2. Sample preparation
Buffer preparation
All buffers were prepared using MilliQ water with a resistance of >18 MΩ cm, titrated at
required pH with hydrogen chloride (HCl) or sodium hydroxide (NaOH). The pH was
determined using a pHISEmeter (brandmodel) with corresponding pH-electrode.
For the ChR2 experiments, a 50 mM sodium acetate (NaOAc) at pH 5.5 and a 100 mM
sodium chloride 20 mM BTP pH 8.8 were prepared. And in the case of AMPs, all buffers
were prepared with the corresponding chloride salt at 100 mM and 20 mM BTP at
pH 7.4.
Nanodisc containing ChR II C128S membrane protein
The nanodisc samples containing the ChR II C128S protein were stored in 50 μL aliquots
with a concentration of 50 μM at - 80° C. The storage solution was 130mM NaCl 1mM
MgCl 10% Glycerol 20mM BTP pH 7.2. The ChR II C128S protein in solution was stored
using the same conditions at a concentration of 80 μM. All experiments were performed
at 4° C, and a fresh aliquot was thawed prior incubation. The nanodisc samples were
diluted to 1 μM for SEIRA experiments and incubated for ca. 2 hours. For the IR
transmission experiments the samples were concentrated ca. 3 times. For the UV-vis
measurements, samples were diluted with the respective buffer to a concentration of
3 μM. The samples were under red light during incubation, as well as prior and after the
illumination process.
Ni-NTA monolayer assembly for nanodisc systems
Aliquots of 10 mM NTA were prepared from a 100 mM stock solution, and kept at
− 20° C. For each experiment, one aliquot was thawed, diluted with H2O to 1 mM
solution, and incubated overnight onto a fresh prepared Au−film. Then, the excess was
washed away by rising with NaOAc 50 mM pH5.5 buffer. Once stable, a solution of
50mM NiSO4 in NaOAc buffer was added and incubated for 1 hour. The excess was
removed with fresh NaOAc buffer. Prior to the nanodisc system immobilisation, the
buffer was exchanged to the 100 mM NaCl 20 mM BTP at pH 8.8. The nanodisc samples
54
with embedded membrane protein were added at a concentration of 1 μM and
incubated for 2 hours. The activity of the retinal protein was analysed by UV-vis
spectroscopy before performing the SEIRA experiment.
Vesicle Preparation
For the POPC vesicles, 10 µL of a 25 mg mL−1 POPC lipid (in chloroform stock solution)
and 100 µL chloroform were mixed in a test tube. To prepare the mixed POPC/POPG
vesicles, in a test tube were added 8 µL of the 25 mg mL−1 POPC lipid, 5 µL of the
10 mg mL−1 POPC lipid (in chloroform stock solution) and 100 µL of a 50:50 CHCl3:MeOH
mix. In both cases, the solution was dried with N2 stream and set under vacuum
overnight. The next day, lipids were re−dissolved with 500 µL of the corresponding
100 mM chloride salt (Na+, Cs+ or K+ depending on the experiment) 20 mM BTP buffer at
pH 7.4. The vesicles were prepared by repeating 3 cycles of 30 s vortexing and 10 min
resting in between. Afterwards, the solution was extruded 31 times through a 100 nm
pore size filter in order to obtain POPC unilamellar vesicles of 100 nm size.
Antimicrobial peptides
The purified solid of enniatin B peptide was dissolved in EtOH and divided in equal
aliquots of a final concentration of 100 μM. The aliquots were dried in a speedvac
centrifuge and stored at - 20° C. Prior experiment, aliquots were dissolved in 100 mM
chloride salt 20 mM BTP buffer pH 7.4 with 2% EtOH. All experiments with enniatin B were
carried out at 25° C.
The purified solid of arenicin 1 peptide was dissolved in a 3:1 EtOH NaCl−BTP buffer to a
final concentration of 100 µM and stored at - 20° C. The experiments with arenicin1
were carried out at three fixed temperatures: 4° C, 25° C and 37° C.
3.2.Methods
SEIRA cell
The experiments were performed using a homemade spectroelectrochemical SEIRA
setup and ATR prism coated with a freshly prepared nanostructured Au-film. This setup
was used to perform SEIRA and EIS experiments. It allows the use of a thermostat and
electrochemical measurements in a three-electrode configuration as shown in figure
3.1. A trapezoidal Si crystal (L:25 mm, W:20 mm and H:10 mm) was used as ATR prism.
The IR signal irradiates the prism in an incident angle of 60° giving a measuring area of
7 mm x 3 mm. The OPUS 5.5 software was employed to evaluate the spectra acquired
using a Bruker IFS66v/s FT-IR-spectrometer. The IFS spectrometer is equipped with an ATR
setup in the Kretschmann configuration, a photoconductive liquid N2-cooled
MCTdetector (HgCdTe), and a globar as the IR radiation source. The globar, the
Michelson Interferometer, and the detector were operated under vacuum; solely the
sample chamber was purged continuously with nitrogen gas. SEIRA spectra were
55
recorded between 4000 and 1000 cm-1 with a spectral resolution of 4 cm-1. Each
spectrum comprises 400 scans.
Au deposition and electrochemical cleaning
The Si prism was polished with alumina powder (Microgrit WCA-9, grain size ca. 6 µm)
and abundantly rinsed with H2O to obtain a hydrophobic surface. The Au-film was
prepared following the electroless deposition procedure [101]. Then, a solution of
400 g/L NH4F was added to the prism‘s surface for 2 min. After rinsing with H2O, the prism
was dried and placed in a H2O bath at 65° C. Meanwhile, the Au solution was
prepared with equal volumes of a 2% (w/w) HF solution, a 0.03 M NaAuCl43H2O
solution, and a plating solution containing 0.3 M Na2SO3 anhydrous, 0.1 M Na2S2O35H2O
and 0.1 M NH4Cl. Subsequently, the Au solution was deposited onto the prism. The
reaction was stopped after 1 min by washing with H2O. Lastly, the prism with the Au−film
was dried with N2 stream.
3.1
3.2
For the electrochemical cleaning, the prism was properly assembled into the SEIRA
setup. The Au surface was rinsed with H2O before adding the 0.1 M HClO4 solution. The
reaction was conducted by running 6 oxidation/reduction cycles between 0.1 and
0.4 V, under constant Ar purge to avoid the formation of reactive oxygen species. The
area of the Au electrode was calculated by using the area of the single reduction peak
at ca. 920 mV obtained during the sixth reduction cycle, and comparing it to the
specific charge density of 400 µC cm−2 [102]. The cyclic voltammetry (CV) measurement
was monitored using the GPES software, and a CHI 600E series potentiostat from CH
Instruments, Inc with a three-electrode configuration. The Au-film with a real area of ca.
1.65 cm2; geometric area of 0.79 cm2 corrected by the roughness factor of 2.1; newly
Figure 3.1. Schematic representation of the spectroelectrochemical SEIRA cell. The relevant features are
indicated in the figure. The Au metal-film acts as IR signal amplifier and working electrode. The IR beam
irradiates the silicon prism with an incident angle of 60° with respect to the surface normal.
56
determined for each Au-film using the Au-oxide reduction charge density method) is
used as working electrode. The Pt-mesh, and Ag/AgCl (3 M KCl) electrode serve as
counter, and reference electrodes, respectively (all potentials are referred to the
Ag/AgCl electrode).
Tethered bilayer lipid membrane construction
The Au-film was functionalized with a mixed self-assembled monolayer (SAM) of
WK3SH/6MH 80:20, which was incubated overnight in a solution of 0.6 mM WK3SH and
0.4 mM 6MH in 1-PrOH. For the bilayer formation, unilamellar POPC or POPC/POPG
vesicles were added onto the SAM after rinsing in order with 1-PrOH, H2O and the
corresponding buffer. The lipids were incubated for 2 hours or until a stable difference
spectrum was achieved. Then, the SEIRA cell was rinsed abundantly with fresh buffer.
Impedance and SEIRA spectra of the SAM and lipids were recorded to characterise the
step-wise construction of the tBLM. Detailed information of the construction of such a
system can be found in the supplementary information of the article by Wiebalck and
coworkers (2016) [29].
Tethered hybrid lipid membrane construction
The tethered hybrid lipid monolayer (tHLM) was built following the same step-wise
procedure as for the tBLM, but using a pure WK3SH monolayer instead of the mixed
SAM. It was monitored by EIS and SEIRA spectroscopy as well.
Illumination procedure
SEIRA and FT-IR transmission spectroscopy were used to follow the illumination assay of
the ChR II protein in the nanodisc system and in solution, using in both case the same
procedure. The illumination assay was performed using a blue (λ~460 nm) LED and a UV
(λ~390 nm) LED. A macro for the OPUS software was designed to repeat 8 cycles of the
following process: Dark/ 460 nm/ 390 nm/ Dark. The dark steps were recorded for ca.
30-40 min and the light steps for ca. 10 min.
FT-IR transmission and ATR-IR spectroscopy
Bruker IFS28 and Tensor27 spectrometers were used for these experiments,
accumulating 200 scans instead of the 400 in SEIRA experiments. A
sandwich-transmission cell with CaF2 windows was used for the sample preparation. The
ChR II and Nanodisc samples were deposited on the sample window by 5 cycles of 2 µL
of sample and drying with N2 gas stream, distributing the sample homogeneously on
the surface. The sample was deposited onto a CaF2 window with a 2 µm deepening.
Afterwards, a thin layer of silicone grease was applied to seal the cell, before closing
with a plain CaF2 window. The cell was placed into the corresponding holder to perform
the experiments. In the case of the AMP samples, 2 µL of sample were placed onto the
deep window and sealed as mentioned before (no drying step). All experiments were
performed at room temperature. Additionally, a diamond-ATR durascope setup was
used to measure the dried samples of the peptides. The setup can be used in both
57
spectrometers mentioned above, and consists of a single-reflectance ATR with a
diamond of ca. 2 mm diameter that is pressed against the sample.
UV-vis spectroscopy
A Cary 50 Bio spectrophotometer (Varian Inc.) or Cary 4000 UV-vis spectrophotometer
(Agilent Technologies) was used to measure stationary or time-resolved (in seconds to
minutes) absorption spectra. The ChR II and nanodisc samples of 300 μL (3 μM) were
measured in a UV quartz cuvette with a total volume of 500 μL (Blaubrand QS 1.000)
and a path length of 1 cm. The absorption spectra were recorded between 200 and
900 nm with a 1 mm resolution.
The buffer solution was measured as the black, and it was used together with the
zero-transmission to carry out the standard baseline correction. A similar illumination
procedure as the one described in the SEIRA experiments was applied for the UV-vis
using the blue (λ~460 nm) LED and a UV (λ~390 nm) LED.
The kinetic measurements of the ChR II C128S protein embedded into the nanodisc
system were performed following a similar procedure as described by Velazquez (2015)
[103]. The intensity was monitored at λ=480 nm in the dark, with the 460 nm light on, and
with the 460 nm light off for a total time of 40 min. The resulting data was evaluated
using the ORIGIN 7 software package or higher (OriginLab, Northampton, MA), and the
curves were simulated using exponential decay functions.
Electrochemical impedance spectroscopy
The electrochemical impedance (EIS) measurements were performed using a
μAutolabIII/FRA2 instrument, and controlled with the frequency response analyser
software. The Au-film served as the working electrode in a three-electrode
configuration with a Pt-mesh as counter electrode and a Ag/AgCl (3 M KCl) reference
electrode. EIS spectra were recorded in the frequency range of 0.05 Hz to 100 kHz at a
DC potential of 250 mV (vs Ag/AgCl) and amplitude of 25 mV (rms). The same software
was used to carry out the fitting of the EIS data using the equivalent circuit described in
the theory section.
Potential-dependence assay
The μAutolabIII/FRA2 instrument and the three-electrode configuration were used to
perform the potential-dependent (+ 400 mV to - 400 mV) experiment of the arenicin 1
peptide, which was monitored by SEIRA spectroscopy. The experiment was performed
at 25º C and 37º C.
DFT calculations
In the density functional theory (DFT), the total energy E of an electronic system is
determined by its electron density r. The total energy is written in terms of the energy of
n non-interacting electrons and a term Eex that takes into account the complicated
correlated motion of the electrons. Here, DFT is used to find the most stable
stoichiometry for a given structure of the EB peptide, and calculate its vibrational
58
frequencies. The method of calculation used was [BP86 functional with basis sets
LanL2DZ for metals and 6-31 g* for N, C, O, H]. Initially, only the dihedral angles of the
ring backbone were fixed to 180° during geometry optimization. After this initial
optimization, the geometry was optimized without constraints.
The density functional theory (DFT) calculations for the enniatin B complexes were
carried out by Dr. Jacek Kozuch, Postdoctoral Researcher at Boxer Lab (Prof. Steven
Boxer), Stanford University (California, United States). The FT-Raman measurement of the
solid sample of arenicin 1 peptide was kindly performed by Dr. Francisco Velazquez
from the Department of Physical Chemistry / Biophysical Chemistry (Prof. Peter
Hildebrandt), Technische Universität Berlin (Berlin, Germany). The molecular dynamic
(MD) simulations of the arenicin 1 peptide in the POPC/POPG tBLM system were carried
out by Ahn Duc Nguyen, Biomolecular Modelling group (Prof. Maria Andrea Mroginski),
Techniche Universität Berlin (Berlin, Germany)
59
60
4. Results
4.1. Channelrhodopsin II in nanodisc systems
In this study, channelrhodopsin II (ChR II) was used as a proof of concept for the
nanodisc membrane system. The final goal was to analyse the system by SEIRA
spectroscopy. Like other rhodopsins, the ChR II protein has seven transmembrane
α-helices, and a retinal chromophore in its active site. The protein used in this work had
a mutation in the position 128, where the cysteine residue was substituted by a serine
(C128S). The mutation at this position leads to an extension of the channel open state
lifetime of 104-fold [40]. Reports of similar mutations showed that the equilibrium
between the P390 and P520 states, closed and open channel respectively, can be
controlled with blue and green light. As a starting point in the study of the ChR II C128S
variant, UV-vis and FT-IR transmission experiments were performed to evaluate the
photocycle of this mutant. These two techniques provided the necessary information to
set the parameters for the characterisation of the nanodisc system with ChR II C128S by
SEIRA spectroscopy.
The microbial rhodopsin ChR II rules the phototaxis of the green algae Chlamydomonas
reinhardtii, and it is a light-gated cation channel. The potential of channelrhodopsins to
control the depolarization of the cell membrane by light is widely used in optogenetics.
Optogenetics is a technique based on the control of certain cells, usually neurons, using
integrated light sensitive proteins. In view of its importance for biomedical applications,
the molecular function of ChR II, including the photocycle and its intermediate states,
has been extensively investigated. The present study builds upon previous work by
Stehfest et al (2010) [47], by Berndt et al (2011) [48] and by Bruun et al (2015) [42].
In their study, Berndt and co-workers highlighted that the variant C128S presented the
longest extension of the conducting state lifetime with 106 s versus 11.9 ms for the wild
type. Additionally, this mutant showed up to 300-fold more sensitivity for light. The
underlying hypothesis was based on the equilibrium between states, inspired by certain
analogies to bacteriorhodopsin. The mutation of bR in the residue Threonine90 of the
helix 3 slowed down the photocycle kinetics. This residue is well conserved and
corresponds to C128 in ChR II. In both proteins, these amino acids are in close
interaction with the retinal chromophore. Thus, it was concluded that the C128S mutant
provoked controlled depolarization of the membrane in neurons after 10 ms of blue
light (470 nm) illumination, which could be terminated by 50 ms of green light (530 nm).
Furthermore, the study by Stehfest brought more clarity on the photocycle adapted by
these mutants by focusing in the C128T and determining side reactions depending on
61
the illumination. The results obtained by Stehfest et al allowed them to postulate an
alternative, yet more complex photocycle in comparison to the one described for the
wild type. Bruun and coworkers demonstrated that low light intensities provoke
equilibrium between two dark-adapted states (D480 and D470) in a ratio 3:1, also
known as apparent dark-adapted state (DAapp), and they discussed the challenge that
this equilibrium represents for the characterisation of the initial dark state (IDA) of ChR II
by spectroscopic techniques. The equilibrium explained in Bruun‘s light-dark adaptation
study is represented in a mirror photocycle for each DA state interconnected by the
two DA states.
Figure 4.1 shows a simplified representation of the photocycle of C128 mutants of ChR II.
The starting point is the D480 (DA) state. After blue light illumination, D480 is
photoconverted to the P500 state, which rapidly decays to P390. The P390 state is
photoactive, and it is in equilibrium with the open state P520. In this stage, it is possible
to take a shortcut by illuminating with green light and transition to P480b, which can
decay back to D480 [47]. There is another shortcut to the P480b by illumination of the
P390 state with UV light. The prevailing conformations of the retinal chromophore in the
DA and P390 states are also depicted in the figure.
4.1.1. UV-vis study of ChR II protein
Taking into account the information described above, a light-induced UV-vis
spectroscopic analysis was performed with the ChR II C128S in solution (3 μM). UV-vis
allowed defining the different states of this mutant by identifying the maximum
absorbance at each step. The DA state was recorded under red light to minimize other
photoreactions. Figure 4.2A shows the spectra corresponding to the DA state (black
Figure 4.1. Schematic and simplified representation of the ChR II protein photocycle. Each state is named
after the respective UV-vis maxima, and the D and P letters indicate dark or photo state, respectively. The
blue and green arrows depict the transitions after illumination with blue and green light, and the purple arrow
indicates the transition induced by UV light. The black of arrows represent the transition, equilibrium or decay
of the different states. The preferential retinal chromophore conformations are indicated as well.
62
line), after 460 nm light illumination (blue line) and after 390 nm light illumination (purple
line). The DA state has two intrinsic maxima vibronic side bands at 450 nm and 480 nm.
The spectrum after blue light illumination (460 nm, blue line) showed one maximum at
390 nm that can be assigned to the P390 state, and residual absorption at 485 nm. The
P500 state cannot be observed with this technique, since it does not accumulate [47].
From the difference between DA and P390 spectra, it is possible to see that the
conversion is not complete. The transition after illumination with UV light (390 nm, purple
line) presented a high conversion of the P390 species to a state with the maximum of
absorbance at 485 nm (slightly red-shifted from the DA state), which corresponds to the
P480b state (see figure 4.2A).
Analysing the results obtained for the protein in solution, three distinct states were
characterised: D480, P390 and P480b. According to the recent detailed analysis of the
dark state of ChR II [42], in the D480 state, the all-trans, anti conformation of the
deprotonated Schiff base chromophore prevails (ca. 70%) whereas the remaining
fraction is in the 13-cis, 15-syn configuration. Furthermore, the formation of the P390
state, associated with a double bond photoisomerisation and the subsequent
deprotonation of the Schiff base linked to a tertiary structure change of the protein,
corresponds to a yet closed channel. The last conversion, from P390 to DA state with
maximum at 485 nm induced by UV light illumination, represented the return to the dark
state via P480b. Therefore, the studies in this work focused on the section of the
photocycle involving the states D480, P500, P390 and P480 b (see figure 4.1).
The very similar illumination procedure was followed for the protein ChR II C128S
embedded in the nanodisc system. The corresponding spectra are shown in figure 5.2B.
The main differences were the lower intensity (especially in the DA state) and the
absence of the 485 nm maximum. Besides that, the phototransitions in the nanodisc
Figure 4.2. UV-vis spectra of ChR II C128S protein in solution A and embedded in the nanodisc system B. In
both cases the samples concentration was of approximately 3 µM. The black line depicts the DA state
measured under red light. The blue line represents the spectrum after illumination with 460 nm (blue) for
10 minutes and the purple line, the spectrum after illuminating with 390 nm light (UV) for 10 minutes. The
maxima for each spectrum are indicated in the picture.
63
spectra were similar enough to assume that the protein behaves in the same way. The
DA state presents two maxima at 450 nm and 480 nm, and a similar transition to the
P390 state after illumination with blue light (460 nm). Moreover, the illumination with UV
light (390 nm) yielded in the disappearance of the P390 state and the transition to the
DA state with a maximum at 480 nm. Overall, these results support the main features of
the nanodisc system; it not only offers a native-like environment for the protein, but also
allows the organization and orientation of the protein in the membrane.
As mentioned before, the mutant C128S has shown reversible cycles between states
controlled by light. Therefore, to evaluate this effect in the section of the photocycle
studied here, an analysis on the reversibility of the states was performed with cycles of
blue and UV light excitation of the ChR II C128S embedded in nanodisc systems (figure
4.3). The UV-vis spectra showed a higher conversion in the first cycle from the DA to the
P390 state compared to the following steps. The reversible character with blue (460 nm)
and UV light (390 nm) irradiation proved to convert approximately 60% of the species
back and forth. This conversion needs to be higher in order to obtain a reasonable
signal-to-noise ratio and to be able to visualise these changes in the IR light-induced
difference spectrum. Nevertheless, this information is crucial for the structural
characterisation of the different states by SEIRA spectroscopy. The protocol described
above was used in the following experiments of the ChR II C128S in solution and
embedded in the nanodisc system.
Figure 4.3. UV-vis spectra of the illumination cycles of ChR II C128S protein embedded in the nanodisc system.
The black line depicts the spectrum of the DA state; the blue and the purple lines refer to the spectra
obtained after illumination with 460 nm and 390 nm (UV), respectively. The maxima of each spectrum are
indicated in the picture.
64
UV-vis spectroscopy: kinetics analysis of ChR II C128S in nanodisc system
Additionally, the UV-vis analysis allowed measuring the kinetics of the light-induced
cycling between D480 and P390, monitoring the temporal evolution of the absorption
at 480 nm maximum, induced by illumination with 460 nm (blue) light, and the recovery
after the illumination.
The results were represented into two graphs, along with fits of exponential functions to
the data (figure 4.4). In the first step, the sample was under red-light in the DA state and
after one minute the 460 nm light was turned on. Graph A shows two processes. Firstly,
there was a rapid transition represented by a steep decay (1), indicating that after
12 ms 40% of the DA state has been converted. The second term of the equation
denoted a slower transition with a 2 of 1.2 minutes. Figure 5.4B shows the recovery of
the species at 480 nm with the blue light off. The recovery was evaluated under red
light for approximately 1 hour. The equation for the recovery transition resulted in a 1 of
4 minutes.
The kinetic analysis provided important information about the transition time between
these states. The results shown here suggest that the P390 state forms almost
instantaneously in the beginning of the illumination and, after 3 minutes the curve
reaches a plateau. On the contrary, the process of recovery of the dark state is much
slower and it needs at least 4 minutes to transition to the dark state. Thus, these times
were taken into account for the illumination protocol of the ChR II C128S protein.
Figure 4.4. Kinetic representation for the evolution of the 480 nm species of the DA state. Graph A
corresponds to the illumination with the 460 nm light, and graph B to the dark recovery phase after the
illumination. Underneath each graph there are the respective fitting equations and parameters. The symbol
represents the mean life time of the state; and are the function constants.
65
4.1.2. FT-IR transmission study of ChR II protein
IFT-R transmission spectroscopy of the ChR II C128S protein was performed in solution
and embedded in the nanodisc system. This technique offered structural information of
the system, which sets the basis for the SEIRA measurements. As in the UV-vis, the first
step was to analyse the protein in solution, and to perform the illumination process
described in the previous section. The same experiment was done with the protein
embedded in the nanodisc system. Figure 4.5 shows the IR spectra of the ChR II C128S
protein in solution (black line) and the protein in the nanodisc system (grey line). These
FT-IR difference spectra correspond to the protein in the DA state, using the buffer
spectrum as reference. The amide I band (mainly the C=O stretching mode) appeared
at 1657 cm-1 and the amide II band (N-H bending and C-N stretching modes) at
1550 cm-1. The band positions and the relative area ratio of ca. 3 of the amide I and II
confirm the α-helical structure of the protein. The spectrum of the protein in the
nanodisc (grey line) showed similar amide I and II bands, though higher intensities. The
second derivative of the spectra (not shown) proved that the amide bands appeared
at the same position in both cases. The difference in intensity may be attributed to the
contribution of the scaffold protein of the nanodisc, which adopts an α-helical structure
as well. The band at 1740 cm-1 in the grey spectrum corresponds to the CO vibration of
the DMPC phospholipids in the nanodisc system. This band is actually an overlap
between the ester groups and the carboxylic acid groups of the phospholipids at ca.
1740 cm-1 and at ca. 1730 cm-1, respectively. These results indicate that the nanodisc
system does not modify the protein secondary structure, maintaining its α-helical
organization.
Figure 4.5. FT-IR spectra of the ChR II C128S protein in solution (black line), and embedded in the nanodisc
system (grey line). The spectra correspond to the DA state of the protein measured under red light and at
room temperature.
66
4.1.2.1. Illumination cycles of ChR II protein
The illumination protocol was applied to both cases: the protein in solution and the
protein embedded in the nanodisc. The protocol was based on three phases that were
repeated in five illumination cycles. The first stage was the transition from DA to the P390
state induced by the 460 nm (blue) light. After the formation of the P390 state, the
sample was illuminated with UV light, and later it was let to rest under red light to
achieve the DA state. From here, the cycle started again with the formation of the P390
state. Figure 4.6 offers a scheme of the illumination protocol to visualise the process. The
whole illumination process was followed by FT-IR spectroscopy at room temperature
(RT). Figure 4.7 and 4.8 show the spectra for the ChR II C128S protein after blue light and
UV light illumination, respectively. The part A of each figure corresponds to the protein
in solution, and the part B to the protein embedded in the nanodisc system. Figure 5.7
presents the 460 nm minus DA difference spectra and figure 5.8 the difference spectra
between the 390 nm and 460 nm lights. The ChR II C128S mutant presented low
reversibility for the evaluated photocycle section, which means that after the third
conversion with the 460 nm light the changes were minimal (data not shown).
The first cycle (black line) showed the most significant difference between the spectra
in figure 4.7. In this first cycle, the protein in solution showed a more prominent change
than the protein in the nanodisc. Besides this, the protein displayed similar light induced
Figure 4.6. Schematic representation of the illumination protocol followed in the IR study of the ChR II C128S
protein. D480, P390 and P480 are depicted as the main states involved in this cycle. The first step consisted of
the illumination of the sample with the 460 nm blue light, and then with the UV light. Each illumination was
performed for approximately ten minutes. Lastly, the protein was let ca. 30min to rest under red light to
achieve the DA state before the new illumination cycle starts. There were performed five illumination cycles.
67
Figure 4.7. FT-IR transmission difference spectra of the ChR II C128S protein photoinduced transition from dark
to P390 state after 460 nm (blue light) illumination. The DA state spectrum was used as reference. There are
three cycles shown: first in black, second in red and third in blue. A: Spectra of the protein in solution. B:
Spectra of the protein embedded in the nanodisc system.
Figure 4.8. FT-IR transmission difference spectra of the ChR II C128S protein photoinduced transition from P390
to P480b state after 390 nm (UV light) illumination. The P390 state spectrum was used as reference. There are
shown two cycles: first in black and second in red. A: Spectra of the protein in solution. B: Spectra of the
protein embedded in the nanodisc system.
68
structural changes in both cases for this transition step. The spectra revealed the most
distinctive bands at (-) 1663 cm-1 and (+) 1650 cm-1 in the amide I region. There was a
small variation in the amide II region at approximately (-)1550 cm-1 . These spectra are
characteristic of the P390 state formation, which is related to major structural changes
in the protein backbone. This observation is in line with results in the literature with the
C128T mutant [34] [47]. In the case of the protein in the nanodisc, the spectra showed
the same behaviour than the protein in solution. A tentative explanation for the
difference in intensity between the spectra in A and B for the first transition may be a
lower percentage of the chromophore in the protein embedded in the nanodisc
system.
Overall, the conversion to the P390 state was reproduced with the ChR II C128S mutant
using the nanodisc system. This means that the function of the protein is not greatly
affected by the nanodisc environment. The decrease in intensity between the black,
red and blue spectra suggests a decay of the changes with the number of cycles. This
decay, observed as well observed in the UV-vis spectra, limits the detection by IR
transmission spectroscopy. This effect might be of stronger influence in SEIRA, which
means that the higher the cycle number, the less probable to observe changes in the
SEIRA difference spectrum. In figure 4.8, which presents the conversion from the P390
into a DA state, the spectra A showed smaller changes than graph B. Interestingly,
comparing A and B, the band at (+) 1663 cm-1 is more defined in the spectra of the
protein with the nanodisc. The characteristic bands appeared at the same positions as
in figure 4.7, but less intense and inversed sign. The decay between cycles was also
present in this transition, which allowed observing changes only until the second cycle.
To conclude this section, it was possible to observe analogous structural changes of the
protein in solution and embedded in the nanodisc system. Nevertheless, slight
differences in intensities were observed between ChR II C128S in solution and
integrated in the nanodisc. Similar conclusions were drawn from the UV-vis conversion
analysis. This fact might influence the SEIRA experiments in terms of intensity: the lesser
intensity of changes between the transitions, the less probable their observation by
SEIRA spectroscopy. Even though three cycles were the maximum to distinguish
changes in IR transmission, the illumination protocol proved to be effective enough to
differentiate the states. Therefore, this protocol using 460 nm (blue) and 390 nm (UV)
irradiation was applied to the ChR II C128S protein embedded in the nanodisc
analysed by SEIRA spectroscopy.
4.1.3. Spectroelectrochemical study of Ni-NTA monolayers
This subsection presents the comparison of two SAMs for the optimisation of the Ni-NTA
monolayer to the anchor His-tagged nanodisc system. The results are presented in three
parts. First, there is the SEIRA spectroscopic evaluation of the NTA SAM versus a mixed
SAM of NTA and 3MP. The second part shows the EIS analysis of the respective SAMs.
And the last part provides the results from the Ni2+ coordination of the NTA in the two
69
SAMs. The information obtained from this subsection allows determining the optimal
conditions for the proper assembly and stability of the final system.
SEIRA spectroscopy of NTA monolayers.
The assembly of a proper self-assembled monolayer (SAM) is crucial for the
immobilization of His-tagged systems like the nanodisc. Therefore, the first stage of this
SEIRA study consisted in the optimization of the SAM. In the literature, it has been shown
that the nitrilotriacetic acid (NTA) is the molecule of choice to anchor His-tagged
systems. The NTA molecule can resemble a trident fork with eight atoms in its backbone.
Despite the high affinity of NTA for this system, a mixed SAM may offer advantages in
terms of charge distribution and accessibility [104] [105] [106]. For this reason, two
different scenarios were tested: pure NTA SAM and a mixture of NTA with
3-mercaptopropanol (3MP).
The pure NTA SAM spectrum (see figure 4.9) shows an intense band at 1740 cm-1 that
corresponds to the CO stretching vibration of the protonated carboxylic groups. The
bands at 1606 cm-1 and 1403 cm-1 are assigned to the COO- stretching vibrations of the
asymmetric and symmetric deprotonated carboxylic groups, respectively. The fact that
the immobilization was carried out at pH 5.5 may explain the mix of protonated and
deprotonated carboxylic species. Another important band appears at 1553 cm-1,
which is the C-N stretching vibration coupled to the N-H bending of the carboxamide
group (amide II) of NTA. The bands in the lower frequency region are assigned to the
C-H bending and rocking of the molecule backbone. The band at 1432 cm-1 is an
Figure 4.9. SEIRA difference spectrum of the NTA SAM after overnight incubation at 4º C minus the spectrum
of the Au-film. The relevant modes of the NTA molecule are depicted in the spectrum as: stretching and
bending. The denominations and correspond to the symmetric and asymmetric vibrations, respectively.
The C=O band is assigned to the protonated carboxylic groups, while the COO- bands correspond to the
deprotonated carboxylic groups. The N-H corresponds to the vibration in the carboxamide group of the
molecule. The immobilization was carried out at pH 5.5.
70
overlap of the C-H bending and a symmetric stretching of the COO-. At 1354 cm-1 there
is a broad band corresponding to the hydrogen bond bending of carboxylate groups.
Overall, the spectrum displays intense bands and presents minimal changes in
intensities after washing off the excess of NTA. The band positions confirm the formation
of the NTA SAM on the Au-film surface [87] [107] [83].
The mixed SAM was prepared aiming a ratio of 80:20 of NTA to 3MP. The SEIRA
difference spectrum is shown in figure 4.10 (blue spectrum). There are slight changes
compared to the pure NTA SAM spectrum. The main difference is the broad band at
1645 cm-1 that overlaps and hides the 1606 cm-1 band of the carboxylate groups from
NTA. Water can find its way through when having a mixed SAM, which results in a
broad -OH band, at 1645 cm-1. In the spectrum of the pure NTA the important bands
are of higher intensity than in the mixed SAM. This can be explained by the lower
percentage of NTA molecules contributing to the different vibrational modes.
EIS spectroscopy analysis of NTA monolayers.
EIS spectroscopy was performed to evaluate the SAM-coated electrode. For the
analysis of the quality of the SAMs, the data was represented in a Cole-Cole plot. The
spectra in Figure 4.11 correspond to the pure NTA, pure 3MP and mixed SAM. All
spectra show only one semicircle, which means there is one region defined by the NTA
and/or 3MP molecules and another one by the solution on top. The lower capacitance
in the NTA spectrum is the result of a thicker layer with a higher resistance. The fact that
the mixed SAM spectrum was more similar to the 3MP than to the NTA might point out a
Figure 4.10. SEIRA difference spectrum of the NTA-3MP mixed SAM after overnight incubation at 4º C minus
the spectrum of the Au-film. In grey there is the spectrum of the NTA SAM for comparison. The relevant modes
of the NTA molecule are depicted in the spectrum as: stretching and bending. The denominations and
correspond to the symmetric and asymmetric vibrations, respectively. The C=O band is assigned to the
protonated carboxylic groups, while the COO- bands correspond to the deprotonated carboxylic groups.
The N-H corresponds to the vibration in the carboxamide group of the molecule. The immobilization was
carried out at pH 5.5.
71
poorer quality of the mixed monolayer, since 3MP can introduce defects into the SAM.
For the analysis, a Rsolvent(Rmonolayer · Cmonolayer)Q circuit was employed. This circuit involves
a constant phase element (CPE) used for non-ideal capacitive behaviour (section
2.5.3). The parameters obtained from the fitting are presented in table 4.1.
Table 4.1. EIS information obtained from the fitting of the NTA, 3MP and mixed SAMs data using the
equivalent circuit Rsolvent(Rmonolayer · Cmonolayer)Q.
SAM
R solvent
Ω
R monolayer
KΩ cm2
C solvent
µF cm-2
Q
α
NTA
475
898
15.3
1.25 10-4
0.60
3MP
722
642
16.8
1.21 10-4
0.64
NTA-3MP
354
1.17 103
16.9
2.18 10-4
0.55
The value gets closer to 0.5 in the mixed SAM than NTA alone. The ~0.5 represents
the Warburg impedance resulting from the diffusive process of ions within the SAM. A
tentative interpretation is that the mixed SAM permits an easier diffusion of water and
ions through the monolayer than the pure SAMs. The experiments with NTA showed a
higher reproducibility in both SEIRA and EIS.
SEIRA spectroscopy of the Ni-NTA system.
In this sophisticated construction, there is a necessary ingredient to facilitate the
immobilization of the nanodisc, the nickel ion (Ni2+). This ion forms a complex with the
carboxylate groups and the N-atom of the NTA, and with the N-atoms of the His-tags of
the nanodisc. After addition of the nanodisc with the His-tag, the Ni2+ experiences a
coordination change that stabilizes the system (see figure 3.12 materials and methods).
Figure 4.11. EIS spectra of the NTA (red), 3MP (black) and 80:20 NTA:3MP mixed (turquoise) SAMs. The data is
presented in a Cole-Cole plot.
72
For this reaction, the Ni2+ was incubated for at least two hours to ensure the
complexation of the ion with the carboxylate groups of the NTA. The spectrum in figure
4.12 shows the changes of NTA when complexing the Ni2+ ion using a pure NTA SAM
(black line) and a mixed SAM (blue line). The bands at 1591 cm-1 and at 1416 cm-1 are
characteristic bands that confirm the complexation of the NTA with the Ni2+ ion. These
bands correspond to the asymmetric and symmetric stretching of the carboxylate
groups, respectively. The band at 1740 cm-1 assigned to the protonated carboxylic
species disappeared. The intense and broad band at 1591 cm-1 overlaps with the N-H
band at 1553 cm-1. Additionally, the band at 1453 cm-1 corresponds to the C-H bending
mode.
The spectrum of the mixed SAM complexation of the Ni2+ ion is the blue spectrum in
figure 4.12. This spectrum shows bands at the same positions as in the pure NTA SAM.
The lower NTA percentage on the metal surface corresponds to the lower intensity in
the spectrum of the mixed SAM. Even though the mixed SAM spectrum appears more
defined, the NTA SAM provided more reproducible spectra. In many cases, there was
bulk water leaking through the monolayer of the mixed SAM, resulting in a broad water
band at ca. 1650 cm-1. Therefore, the pure NTA SAM was used for the subsequent
experiments due to a reduced interference of the water band that overlaps with bands
of the target protein, and ensures a higher reproducibility.
Figure 4.12. SEIRA difference spectra of the NTA complexation with Ni2- ion. The black and blue spectra
correspond to the NTA and the mixed SAM, respectively. Both spectra show the characteristic bands of the
Ni2+ coordinated by the carboxylate groups of the NTA molecule. The relevant modes are depicted in the
spectra as: stretching and bending. The denominations and correspond to the symmetric and
asymmetric vibrations, respectively. The C=O band is assigned to the protonated carboxylic groups, and the
COO- bands correspond to the deprotonated carboxylic groups that coordinate the Ni2+ ion. The N-H
corresponds to the vibration in the carboxamide group of the molecule. The complexation reaction was
carried out at pH 5.5.
73
4.1.4. SEIRA study of the ChR II C128S protein in nanodisc
systems
The scaffold protein of the nanodisc has a His-tagged tail that allows binding to the
Ni-NTA SAM. After changing to pH 8, the imidazole groups from the His amino acids bind
to the Ni2+ ions. In this way, the Ni2+ ion remains packed between the NTA and the
His-tag, anchoring the nanodisc system to the surface. The spectrum measured after
the incubation is shown in figure 4.13. The amide I (mainly the C=O stretching of the
protein backbone) and amide II (predominantly the N-H bending and C-N stretching)
bands confirmed a successful immobilization of the nanodisc system. The amide I
appears at 1653 cm-1 and the amide II at 1550 cm-1, these bands are characteristic of
the -helix structures. The broad bands might correspond to the contribution of the
scaffold protein, which is also an -helix. The bands between 3000 and 2850 cm-1 and
at 1740 cm-1 are distinctive for the (CHn) stretching and the CO stretching modes of the
lipids, respectively. The asymmetric and symmetric stretching vibrations of the aliphatic
chains of the DMPC lipids are observed at 2958 and 2871 cm-1 for the -CH3 and at 2921
and 2852 cm-1 for the -CH2. In figure 4.13, the binding process is illustrated at different
time points. The intensity of the spectrum changed only slightly after washing away the
excess of nanodisc in solution.
In the SEIRA spectrum, the amide I band is down-shifted compared to the FT-IR
transmission measurements (1657 cm-1). The oriented immobilization in SEIRA explains
not only this slight-shift in the amide I, but more importantly the difference of the
amide I/II band ratio. By area, the amide I/II ratio was 1.18 in SEIRA and 3.74 in FT-IR
transmission. The FT-IR transmission measurements offered an orientation average of the
bands in solution, while in SEIRA the protein bands and the nanodisc have a restricted
orientation due to the immobilization of the system. Zaitseva and co-workers suggested
an angle of about 24° between the horizontal plane of the nanodisc and the metal
surface, using a Ni-NTA SAM [14]. In an -helix, the amide I mode is oriented parallel to
the helix axis and perpendicular to the amide II mode. As a result, the amide I and II
modes are enhanced similarly in SEIRA with an angle of ca. 60° with respect to the
normal of the Au surface. Thus, this explains that the ratio between the two bands was
close to 1. It is worth noticing the contribution of the scaffold protein in the SEIRA and
FT-IR transmission spectra. The bands that correspond to the lipid patch were
characterised by SEIRA as well. The CO vibration from the lipid head groups appears at
the same position as in the IR transmission experiment.
Overall, the spectrum obtained by SEIRA presents the characterisation of the nanodisc
system containing the membrane protein. The secondary structure of the ChR II protein
with seven transmembrane -helices was conserved when incorporated into the
nanodisc system. These results show the potential of this membrane system, which
allows the identification of its composition and the target protein.
74
Light-induced SEIRA study of the ChR II C128S protein in nanodisc system
The illumination process was performed following the same protocol applied in the
UV-vis and FT-IR transmission measurements. Figures 4.14 shows the changes
corresponding to the first transition from the DA to the P390 state probed by SEIRA and
FT-IR spectroscopy. The IR transmission spectrum (B) was described in section 4.1.2, using
the DA state as reference. In the SEIRA experiments there were five cycles recorded as
well, involving the 460 nm, 390 nm, and dark resting period (see figure 5.6). Following this
protocol, it was possible to obtain the same changes observed in FT-IR transmission
spectroscopy. Even though the intensity of the changes was 100 times lower in the case
of SEIRA, the two spectra showed similar band-shape and positions. The first transition in
SEIRA presented the most significant changes in the amide I region at (-) 1664 cm-1 and
(+) 1653 cm-1, which were assigned to the structural changes in the backbone of the
retinal protein. The rest of the conversions presented minimal changes that could not
be distinguished from the noise.
The short-range restriction in SEIRA spectroscopy, which permits enhancement of the
modes close to the surface, may explain the lower intensity of the transition spectrum in
SEIRA compared to the transmission results. Even so, the assembly of the nanodisc
system containing the membrane protein ChR II C128S was successful, and it was
possible to observe the photo-transition of the light-sensitive channel as well.
Figure 4.13. SEIRA difference spectra of the immobilization of the ChR II C128S protein embedded in the
nanodisc system. The spectrum of the previous step (Ni2+-NTA) was used as reference. A: (CHn) stretching
band region, which reveals the contribution of the lipid in the nanodisc. B: CO band corresponding to the
lipids, as well as the amide I and amide II bands corresponding to the ChR II C128S protein. In both cases, the
different shades of blue represent the evolution of the incubation over time. The nanodisc binding reaction
was carried out at pH 8.8.
75
4.1.5 Conclusions
EIS and SEIRA spectroscopy were used to characterise and optimise the Ni-NTA SAM.
This monolayer served to anchor the nanodisc membrane system embedding the ChR II
C128S light-sensitive protein. It was possible to observe the transition from the DA to the
P390 state of the protein by SEIRA spectroscopy. However, the rest of the reactions
involved in the illumination process did not have significant impact in the corresponding
spectra. The spectra obtained in both FT-IR and SEIRA experiments were similar to the
spectra described in the literature [34] [42] [47]. However, the yield of the reactions
seems to decrease with the cycle number, decreasing as well the probability to detect
these changes. This decrease is reflected in the intensity of the transition spectra
observed by FT-IR transmission and UV-vis, and even more critical in SEIRA.
UV-vis and FT-IR transmission measurements offered supporting information for the
interpretation and understanding of the results obtained by SEIRA. UV-vis provided the
identification of the different states involved in the portion of the ChR II C128S
photocycle analysed here. Moreover, the FT-IR transmission spectroscopy allowed the
structural evaluation of the mutant and its photo-induced changes. Both techniques
served to evaluate the protein in solution and embedded in the nanodisc membrane
system. This comparison demonstrated that the protein structure and function were not
greatly affected when inserted in the nanodisc system. These findings confirm the
relevance of nanodisc membrane systems to investigate transmembrane proteins by
SEIRA spectroscopy. This study sets a base for future improvements for applications of
this system to similar targets.
Figure 4.14. A: SEIRA difference spectrum of the first illumination transition from DA to P390 state with the
460 nm (blue) light of the ChR II C128S protein embedded into the nanodisc system. B: FT-IR transmission
spectrum of the same first transition after illumination with the blue (460 nm) light. There was a 100 fold
absorbance difference between spectra A and B.
76
4.2. Characterisation study of enniatin B
This section presents the results of the structural study of the AMP enniatin B (EB) and its
interaction with the POPC membrane systems. Enniatins are depsipeptide
4
produced
by various species of Fusarium fungi, though Fusarium Scripi preferably makes EB. The
antimicrobial activity of enniatins has been proven against a wide range of diseases,
like cancer and respiratory infections [70]. Even though most of the studies suggest a
membrane-active mechanism of action, details are yet not known. The isopropyl
residues and methylated amide nitrogens render this cyclodepsipeptide highly
hydrophobic, which indicates it can easily diffuse through the cell membrane. Due to its
circular shape, EB can form ion complexes with mono and divalent ions creating
ion-transports within lipid bilayers.
Mueller and Rudin (1967) provided the first evidence of the membrane interaction of
enniatins and other depsipeptides, by monitoring the increase in conductance through
a bimolecular lipid membrane in presence of Na+ and K+ ion-solutions [108].
Ovchinnikov et al. (1974) offered a model for the conformation as well as stoichiometry
of the different ion-complexes [61]. They suggested a difference in the C=O ligands
contribution depending on the EB:ion complex stoichiometry. While for the 1:1 and 3:2
ratios both CO and amide I ligands coordinate the ion/s, for the 2:1 ratio the amide I
ligands are pointing inwards forming a sandwich with the central ion. Accordingly, and
taking into account the ion size, they proposed that Na+ and K+ form a 2:1 complex and
Cs+ with bigger radii would form a 3:2 complex. The results obtained from the
patch-clamp study by Kamyar et al 30 years later [76], led to a similar conclusion for the
complexes stoichiometry extending it to the evaluation with divalent ions and
suggesting a cation selectivity order for EB according to K+>Ca2+≥Na+>Mg2+>Li+.
Interestingly and against expectations, IR spectra of EB in absence and presence of
different ions show very similar position of the CO and amide absorptions at 1746-1744
cm-1 and 1666-1665 cm-1 [61]. This contrary to results found for beauvericin, valinomycin,
and other peptides like gramicidin A [61], and suggests that either these normal modes
are insensitive to the interaction with different ions or that only a fraction of EB formed
complexes.
The advantage of this project is that it presents a method combining the tBLM system
with SEIRA spectroscopy to evaluate the peptide structure in a membrane-mimetic
environment. The main goal of this work was to addresses the following questions: i)
Does EB incorporate into membrane systems? Is it an ion carrier? ii) What is the structure
and stoichiometry of EB:ion complexes? iii) What is EB binding mechanism to ions and to
membranes? The diversity of methods employed here are essential to complement the
SEIRA results, and offer valuable information about the EB-membrane and ion
interaction. The results are presented with focus on the questions. First, the
characterisation of the POPC tBLM system by EIS and SEIRA spectroscopy is shown,
4
Depsipeptides are peptides that have one or more amide groups substituted by an ester group.
77
together with the analysis of the membrane incorporation of EB. Then, the results of the
ion-exchange process in tBLMs and theoretical calculations of the respective ion
complexes will be presented, as well as a critical discussion of the related findings from
literature. Lastly, there will be an evaluation of the kinetics of EB membrane binding and
ion-exchange processes, using the data from the SEIRA measurements. This last part
offers supplementary information that support the structural models proposed in this
work, in addition to an evaluation of EB mobility within the membrane. The collection of
conclusions obtained in this study is presented in the summary at the end of the section.
4.2.1. Incorporation of enniatin B into tBLM system
This study shows for the first time the structural characterization of EB by SEIRA
spectroscopy using a POPC tBLM system. The beginning of this subsection describes the
results of the step-by-step construction of the planar membrane systems. The evaluation
of the tBLM system is followed by the independent incubation of EB in presence of three
metal ions. From literature [109] it is known that EB can form complexes of different
stoichiometry with several mono and divalent ions. Here, the ion complex formation of
EB with Na, K and Cs are characterised by SEIRA spectroscopy. All three metals belong
to the alkali metals group and form monovalent positive ions. The size of the ion
increases when going down in the group, from Na+ to K+ and then, Cs+. This increase in
the ion radii is considered one of the features for EB to form either 1:1, 2:1 or 3:2 EB:ion
complexes [61]. The independent incubation of the complexes shall be taken in
consideration for the discussion offered in the following subsection.
4.2.1.1. EIS of POPC membrane systems
The construction of the tBLM composed of areas of lipid bilayers requires a mixed SAM
to anchor the system. The tBLM has advantages in comparison to other membrane
systems. The most significant ones are the covalent immobilization of the tBLM on a
Au-electrode, which increases its mechanical stability considerably, and at the same
time provides space for an aqueous reservoir between the membrane and the
electrode. This construct is bound to the surface of the nanostructured Au-film using
tether molecules, which allows investigating the system by SEIRA and EIS spectroscopy.
The tBLM used in this study contains the WK3SH tethered-cholesterol molecule and
POPC zwitteronic lipids, which assures minimal spectral overlap with the EB normal
modes of interest.
The first step consisted in optimizing the mixed SAM to form hydrophilic and
hydrophobic islands on the Au-film surface. The tethered cholesterol molecule WK3SH
formed the hydrophobic parts, and the 6MH formed the hydrophilic parts, which
facilitate the formation of the aqueous reservoir beneath the lipids. The SAM was
prepared to aim 80% of WK3SH and 20% of 6MH molecules on the surface of the Au
according to Wiebalck et al. (2016) [29]. The EIS offers an analysis in terms of quality of
the tBLM construct, which in this kind of system can prove whether or not the SAM and
78
the lipid bilayer are formed. The approach using EIS to evaluate the tBLM system is
based on former work developed by Jeuken (2007) [9]. For this project, the
capacitance of the system was determined by reading out the diameter of the first
semicircle in the Cole-Cole plot (see figure 4.15). The capacitance of the mixed SAM
can be expressed as a molar fraction relationship of the capacitances of the pure
monolayers. Figure 4.15 top shows the EIS spectra of the WK3SH and 6MH pure
monolayers assembled on a nanostructured Au-film. The capacitances determined for
each monolayer are (0.94 ± 0.13) µF cm-2 for WK3SH and (3.51 ± 0.23) µF cm-2 for 6MH.
Accordingly, the SAM spectrum (black) depicted in figure 4.15 bottom indicates a total
capacitance (CSAM) of (1.28 ± 0.01) µF cm-2. Taking into account the capacitances of
the pure SAMs, the total capacitance of the system can be represented as follow:
(4.1)
where
X
and
C
are the respective molar fraction and capacitance for each monolayer.
In this equation, it is possible to determine
XWK3SH
and estimate the surface fraction of
the cholesterol-tethered molecule. Thereby, it is possible to control the ratio of both
SAM molecules at each experiment. As shown in previous work by our group and
Jeuken (2006), the assembly of these molecules on the metal surface leads to a phase
separated SAM with islands of WK3SH and islands of 6MH. This phase separation is
based on the hydrophobicity of the C6-chain of 6MH and the hydrophilicity of the
triethylene glycol linker [-(CH2-CH2-O)3-] of WK3SH, so that under selected conditions a
cooperative binding behaviour at the Au-electrode is observed [9] [29].
After the addition of the vesicles, the capacitance of the system decreased to
(0.62 ± 0.03) µF cm-2 for the POPC tBLM (red spectrum figure 4.15 bottom). The decrease
of the capacitance is a marker for the tBLM formation due to an increase of the
thickness of layers on the surface of the electrode. This membrane construct is
represented by the equivalent circuit RSolvent(RSpacerCSpacer)(RbilayerCbilayer)Q introduced in
section 2.5.3. The fitting of this circuit to the data, allowed calculating the physical
parameters shown in table 4.2. The Csystem refers to the CSAM in the case of pure WK3SH,
pure 6MH and mixed monolayers, and to CtBLM in the case of the mixed SAM with the
lipids. The capacitor Q describes the non-ideal electrical behaviour of the system.
Since this circuit has been well described by Wiebalck et al. (2016) [29], only the
parameters relevant for the quality of the different steps are discussed, i.e. the
difference in the measured capacitance values respective to each step. This study uses
the results as a proof of the proper assembly of the system on the surface of the
electrode.
79
80
Table 4.2. Results from the EIS analysis obtained by the fitting of the spectra to the equivalent circuit
described for the full monolayers, SAM and the tBLM system.
RSolvent
Ω
Rbilayer
KΩ cm2
Cbilayer
µF cm-2
Rspacer
KΩ cm2
Cspacer
µF cm-2
Q
µF s α-1
cm-2
α
Csystem
µF cm-2
6MH
184.6
1.34
13.93
192.79
18.81
21.23
0.82
3.51
WK3SH
219.9
5.30
2.35
364.96
4.32
12.99
0.79
0.94
SAM
266.7
2.21
5.03
483.13
6.55
9.97
0.82
1.28
tBLM
170.1
1.57
1.12
389.03
2.34
3.65
0.83
0.62
The POPC tethered hybrid lipid membrane
The tethered hybrid lipid membrane (tHLM) system was built in similar step-wise
procedure as the tBLM, but instead of using a mixed SAM it uses a pure WK3SH
monolayer. Consequently, the vesicles formed a lipid monolayer on top of the
cholesterol SAM.
The impedance spectra of the pure WK3SH SAM (turquoise) and the tHLM (purple) are
shown in figure 4.16. The interpretation used for the tBLM applies as well to the tHLM,
therefore, the same equivalent circuit is used to obtain the physical parameters that
describe the system (see table 4.3). The tHLM shows a capacitance of (0.70 ± 0.06) µF
cm-2 slightly higher than the tBLM construct. The small difference between these
capacitance values might be assigned to an altered effective dielectric constant and
thickness of the tHLM due to the lack of the 6MH islands and the water reservoir. The
results of both membrane systems were reproduced several times and are a proof of
the robustness of the system.
Table 4.3. Results from the EIS analysis obtained by the fitting of the spectra to the equivalent circuit
described for the full WK3SH monolayer and the tHLM system.
RSolvent
Ω
RSpacer
KΩ cm2
CSpacer
µF cm-2
Rbilayer
KΩ cm2
Cbilayer
µF cm-2
Q
µF s α-1
cm-2
α
Csystem
µF cm-2
tHLM
218.1
2.24
2.17
729.08
3.24
7.56
0.81
0.70
81
4.2.1.2. SEIRA spectroscopy of POPC membrane systems
The structure of the POPC tBLM system was characterised by SEIRA difference
spectroscopy. The resulting SEIRA spectrum of the mixed SAM formation is shown in
figure 4.17, using the spectrum of 1-PrOH as reference. The negative bands correspond
to the 1-PrOH removed from the surface, and the positive bands are ascribed to the
adsorption of the SAM molecules on the Au surface. The bands at 2960, 2933 and
2875 cm-1 have contribution of the ν (CHn) stretching vibrational modes of both
molecules. The broad band at 3284 cm-1 corresponds to the ν (OH) stretching from
6MH. The region below 1500 cm-1 contains the (CHn) bending, scissoring and wagging
modes, and the ν (C-O) stretching. The SAM spectrum reveals the most important
advantage of using the WK3SH tethered molecule; that is, contrary to other cholesterol
tether molecules, such as CPEO3 [11], WK3SH lacks of bands in the amide region (1800
to 1500 cm-1) [29]. Thus, the WK3SH:6MH SAM represents an ideal template for the study
of biomolecules. A comprehensive description of the WK3SH:6MH SAM and tBLM can
be found in Wiebalck et al. 2016 [29].
The completion of the assembly of the tBLM system was achieved after the addition
and spontaneous spreading of the POPC vesicles onto the SAM. The SEIRA difference
spectrum after removing the excess of lipids in solution is shown in Figure 4.18, using the
SAM spectrum as reference. The spectrum of the lipids shows a negative broad band at
ca. 1650 cm-1, which represents the water removed from the surface of the SAM. This
Figure 4.16. Impedance spectra represented in Cole-Cole plots for full WK3SH (turquoise) monolayer and
tHLM system after addition of the lipid POPC lipid vesicles (purple).
82
band is assigned to the δ (OH) bending modes of water. The ν (CHn) stretching modes
corresponding to the aliphatic chains (hydrophobic tail) of the lipids are observed in
the region between 3006 cm-1 and 2854 cm-1. The band at 1737 cm-1 represents an
overlap of at least two modes, the non-hydrogen bonded and hydrogen-bonded ester
carbonyl stretching (ester groups in the head of the phospholipid) at 1743 cm-1 and
1733 cm-1, respectively. The latter can offer information about the bilayer-water
interface. In the lower region from 1500 cm-1 to 1200 cm-1 the bands are assigned to the
(CHn) bending, scissoring and wagging. Even tough most lipids present similar IR
spectrum, they can be distinguished by slight changes in the ν (CHn) stretching region
and the wagging and bending bands. These vibrations give information about the
physical state of the fatty acyl chains. In this case, the position of the νas (CH2) and
νs (CH2) stretching vibrations at 2926 cm-1 and 2854 cm-1, respectively, indicates a liquid
crystalline phase of the acyl chains, which for gel phase would appear at lower
frequencies [19]. The phosphate group presents bands between 1240 cm-1 and
1000 cm-1 with a characteristic strong ν (PO-2) asymmetric stretching vibration at
1234 cm-1. Table 4.4 summarizes the molecular vibrational modes corresponding to the
POPC phospholipids. These results confirmed the proper formation of the POPC tBLM
system on the surface of the Au electrode. The POPC lipids spectrum was highly
reproducible in band shape and intensity. Since POPC lipids are present in most of the
living membranes, this tBLM can be considered as a standard membrane-mimetic
system providing a robust and stable construct.
Figure 4.17. SEIRA difference spectrum of the mixed WK3SH-6MH SAM incubated in 1-propanol overnight at
4° C, the solvent spectrum was used as reference.
83
Table 4.4. Band assignment for the POPC lipid vibrational modes.
Band position
(cm-1)
Description
3006
CH asymmetric stretching
2956
CH3 asymmetric stretching
2926
CH2 asymmetric stretching
2895
Head-CH2 symmetric stretching
2873
CH3 symmetric stretching
2854
CH2 symmetric stretching
1737
1743 C=O Ester carbonyl stretching
1730 C=O---H Carboxylic acid-carbonyl
stretching
1490
Head-CH3 asymmetric bending
1465
CH2 symmetric bending
1380
CH3 symmetric bending
1228
asymmetric stretching
1090
symmetric stretching
1070
C-O-P stretching
Figure 4.18. SEIRA difference spectrum of the POPC lipid after 2 hours incubation in 100mM NaCl 20mM BTP
buffer at pH7.4, using WK3SH-6MH SAM as reference. Measured at 25° C.
84
The SEIRA difference spectra of the POPC tHLM showed similar spectral features as for
tBLM with only slight changes in relative intensities (data not shown). Since the lipids are
the same as in the tBLM, the assignment of the IR modes shown in table 4.4 is also
applicable for the tHLM system.
4.2.1.3. Membrane incorporation of enniatin B
This study presents for the first time a direct spectroscopic observation of EB peptide
incorporation into membrane models monitored by SEIRA spectroscopy, since previous
IR studies of EB were collected only in organic solvents. To achieve the results shown in
figure 4.20 the independent incubation of EB in presence of Na+, K+ and Cs+ ions was
investigated by SEIRA spectroscopy. In all three cases the characteristic spectrum of EB
was observed [61] and thus, it suggests a successful incorporation into the POPC
membrane system. To assign the characteristic bands of EB peptide, the incubation
spectrum with Na+ ion (grey) shown in figure 4.19 was chosen as an example. The
bands at 1747 cm-1 and 1660 cm-1 correspond to the ester ν (CO) stretching mode of
the D-hydroxyisovaleric acid and to the amide I of the N-methyl-L-valine units,
respectively. It should be noted that the present amide I normal modes differ from the
conventional amide I composition in proteins or peptides due the absence of the
H-atom at nitrogen. Therefore, the amide I normal modes of EB are composed of mainly
the ν (CO) stretching and a minor contribution of the amide (CN) bond vibration. The
intensity of the peptide bands changed when removing the excess of peptide in
Figure 4.19. SEIRA difference spectra of EB:Na+ complex incubation in 100mM NaCl 20mM BTP buffer at pH 7.4
(grey) and after washing the peptide excess with buffer (black). Both spectra were measured at 25° C and
used the tBLM as reference spectrum.
85
solution after the incubation (black spectrum). The CO band decreased from 1.23 to
0.81 mOD, and the amide I from 1.24 to 0.95 mOD indicating that EB weakly bound to
the membrane surface was removed. The bands in the 2966 cm-1 to 2852 cm-1 region
correspond to ν (CHn) stretching modes mostly from the peptide, as apparent from the
shifted position and altered relative intensities in comparison to lipids. While for lipids the
νas (CH2) at 2926 cm-1 presents the most prominent peak due to the large number of
CH2 groups in the acyl chains, the IR spectrum EB is dominated by the νas (CH3) at 2966
cm-1 due to the isopropyl residues. The broadening of these bands indicates that EB is
not located purely within the membrane core, but potentially distributed also to a
considerable amount at the membrane head group region. In the lower frequency
region there are the (CH) bending and rocking modes are detected at 1469 cm-1,
1416 cm-1, 1391 cm-1 and 1375 cm-1. Since these vibrational modes have shown to be
unchanged for EB in presence of the different ions, the analysis of the ion-incubation
spectra is focused on the region from 1800 to 1500 cm-1. The resulting spectra are shown
in figure 4.20A. In this way, the incubation of EB with Na+ (black spectrum) described
above can be compared to spectra in Cs+ and K+ buffers. The incubation with Cs+ ions
(blue spectrum) depicts bands in the same position as in the Na+ spectra, i.e. at
1747 cm-1 and 1660 cm-1, but with slightly higher intensities. In the case of the K+
incubation, in addition to an increase in the intensity an up-shift of 2 cm-1 for the CO
band and 6 cm-1 was observed for the amide I, resulting in band positions at 1749 and
1666 cm-1 (see Table 4.5). Interestingly, in the case of Cs+ and K+, the changes after
removing the excess of peptide are less pronounced than in the case of Na+. The fact
that EB was incubated in presence of each type of cation independently from the
others allows the assumption that each spectrum corresponds to the respective
complexed species, denoted as EB:ion. Furthermore, since EB represents a dynamic
system interacting with the membrane core, head group region or bulk solution, an
equilibrium between free EB and differently composed EB:ion complexes has to be
considered when assigning the spectral features to certain EB:ion complexes. Different
conditions can influence this equilibrium, like the affinity to the ion and the
concentrations as discuss below. Nonetheless, one can conclude that these results
demonstrate the incorporation and interaction of EB:ion complexes with the POPC tBLM
membrane system.
Water interference in the SEIRA spectra of enniatin B incubation
The peptide shows broad amide I bands for all three ions, which suggests contribution
of the bending mode from water (ca. 1650 cm-1). Even though SEIRA is rather insensitive
to contributions from the bulk due to the steep decay of the surface enhancement,
such minor contributions may affect this spectral region. To analyse the impact of the
water absorption band, the water contribution was removed by subtracting a water
reference spectrum from the entire batch of spectra of incubation experiments (> 50
spectra for each experiment) using a similar approach to Palacky et al. (2011) [110] to
86
87
avoid artefacts due to manual subtraction of single spectra (this spectral correction
was performed in collaboration with Dr. Jacek Kozuch). Accordingly, the water
spectrum was subtracted from the principal components (determined by singular value
decomposition; only those components that showed spectral features above the noise
level were considered as principal components) of the SEIRA spectra by matching the
3400 cm-1 absorption of water and thus removing efficiently the water contribution at
1650 cm-1.
The results from the water subtraction are shown in figure 4.20B. Table 4.5 shows the
positions and intensities, as well as the band intensity ratio for the three incubation
experiments as raw data and after water removal. The position of the CO and amide I
bands are maintained in the water-free spectra of all three ions, so that the spectral
positions in the raw spectra can be used as well for the assignment to the EB:ion
complexes. Accordingly, the peak positions shown in table 4.5 can be tentatively
assigned to the EB:ion complexes formed in membranes with yet unknown
stoichiometry. As mentioned above, these positions may not reflect one species in
each case but possibly mixtures between free EB and different EB:ion complexes in
different locations in the membrane, which will be addressed later in this thesis. In
contrast to Ovchinnikov et al. (CO and amide absorptions at 1746-1744 cm-1 and 1666-
1665 cm-1 for every EB:ion complex), the present study shows a distinct difference of the
EB:K+ complex with respect to Na+ and Cs+. This suggests that in organic solvents
different species are formed than in membranes or that, based on the coincidence of
the spectral positions, a K+ impurity in the previous experiments let to a main
contribution of EB:K+ in all cases. However, in the present work, the similarity of the
spectra in Na+ and Cs+ buffers is against the expectation of a Hofmeister-series-like
trend (Na+ > K+ > Cs+, or vice versa) and allows the assumption that a similar ion-free
species is formed in both cases under steady-state conditions. This would be in line with
an increased selectivity of EB for an ion complex formation with K+.
Table 4.5. SEIRA data collected from the incubation of EB with the tBLM system at different ion-buffer.
Raw spectra
-H2O spectra
CO
(cm-1)
Amide I
(cm-1)
CO/amide I
CO
(cm-1)
Amide I
(cm-1)
CO/amide I
EBNa+
1747
1660
1747
1660
0.81
0.95
0.85
0.89
0.84
1.06
EBCs+
1747
1660
1747
1660
0.85
1.10
0.77
0.93
0.87
1.06
EBK+
1749
1666
1749
1666
0.72
1.02
0.71
0.73
0.76
0.96
88
The band ratios of all three complexes changed significantly after the water removal,
which reflects the influence of water in the spectra (see table 4.5). It is important to
mention at this point that in SEIRA the intensity and ratio of the bands can give
information about the distance and the orientation of the vibrational modes in respect
to the surface [14] [86]. Since the membrane is anchored to the surface and shows only
minor fluctuations at constant electrode potentials, the absolute intensity of the IR
bands of EB are directly dependent on its concentration in the membrane. Thus,
determining the distance or location of the EB complexes in respect to the surface of
the Au film requires the exact knowledge of the number of EB molecules. Therefore, the
absolute intensities in the SEIRA spectra will be used as an indicator of the affinity of EB
to bind to membranes in the presence of the chosen ions. On the other hand, the
orientation can be estimated (or compared) due to the selection rule that implies a
major enhancement for the vibrational modes perpendicular to the Au surface. As
mentioned above, an ester CO and N-methylated amide groups in its backbone
characterise the particular composition of EB. This means that instead of the amide I
and II (total dipole moment perpendicular to each other) as for normal amino acid
composition, EB spectra shows a CO and amide I band. Moreover, in EB backbone all
CO bonds of the esters groups and of the amides point in opposite directions. Due to
this and the C3 symmetry-like structure of EB and its complexes the main component of
the effective transition dipole moment of the CO and amide vibrations is oriented
along the axis of the EB ring. Therefore, the ratio of both intensities will not provide direct
information about the orientation of EB within the membrane in respect to the surface
as use for proteins. However, it may represent a specific value for each EB species due
to characteristic angles of the CO bonds in respect to the ring axes.
4.2.2. Structural study of ennitain B ion complexes in
membrane systems
4.2.2.1. Enniatin B ion-exchange in membrane systems
Continuing with the hypothesis that EB can form complexes with monovalent cations
while inducing ion-transport through the membrane, the effect of the three alkali metal
ions Na+, Cs+ and K+ was investigated by SEIRA spectroscopy in a sequential manner
using the tBLM system. The decision of using K+ as the last one is based on the
observation that EB might have a higher affinity to this ion like the well-characterised
depsipeptide Beauvericin [61]. The resulting spectra of the sequential ion exchange of
EB are shown in Figure 4.21A. The EB spectrum in presence of Na+ ion was the starting
point showing the same band positions as shown for the incubation, CO at 1747 cm-1
and amide I at 1660 cm-1. After exchanging the buffer to Cs+ there was no shift of the
bands, but the intensity of the CO and amide I bands increased from 0.81 to 1.00 mOD
and from 0.95 to 1.33 mOD, respectively. In the case of the EB:K+ complex, the
spectrum showed major changes reflected by the shifts of both bands, and a
89
significant increase in intensity. The amide I band with 1.40 mOD appeared at
1666 cm-1, and the CO band shifted by 2 cm-1 to 1749 cm-1 (1.14 mOD). The shifts of the
bands are the same as observed in the independent incubation (vide supra), and it
can be considered as an important marker to identify the formation of a complex of
different stoichiometry. These experiments support the interpretation that EB presents a
similar major conformational state of the backbone with Na+ and Cs+, i.e. a major
fraction of ion-free EB or/and C=O ligands oriented in similar directions, which in both
cases differs from the conformation adopted in presence of K+ ion. Again, the findings
in previous studies in literature differ from the results presented here. The conditions in
the tBLM system offer a different environment for the peptide than in other studies. The
interaction of the EB with the lipid membrane and with the solution-interface may have
a great impact in the stoichiometry and conformation adopted by the ring peptide.
Ion-exchange in monolayer membrane systems
The same experiment described above was performed using a POPC lipid monolayer,
defined as tHLM system. The main goal of this analysis is to evaluate the necessity of a
lipid bilayer for EB for comparison with the tBLM system. The resulting ion-exchange
spectra are shown in figure 4.21B, following the same colour-code for each ion as in the
tBLM spectra. It was possible to reproduce the spectral positions and the amide I
band-shift when exchanging to the ion solution with K+. Interestingly, there is no shift of
the CO band, which appears at 1747 cm-1 for all three ions. In general, the three
spectra in tHLM are of higher intensity than in the tBLM experiment. Moreover, the right
shoulder in the amide I band for Na+ and Cs+ ions indicates a probable contribution of
different species in the spectra. Overall, these findings show that for EB the formation of
the ion-complex and/or membrane-interaction are independent from the membrane
thickness, ergo EB can be incorporated into mono and bilayer lipid system.
It is important to consider that in the tHLM system there is no aqueous reservoir between
the SAM and lipid monolayer. Therefore, the complex may have reduced vertical
movement within the membrane, located mostly in the top leaflet of the tHLM
(surface-buffer interface) instead.
Ion-complex reversibility
To evaluate the ion-selectivity and reversibility of EB complex formation, the solution was
exchanged back to Na+ after the K+ complex. The results are presented in figure 4.22.
The lower intensity and back shift of the CO and amide I bands when changing from K+
to Na+ reflects a low selectivity of EB between this two ions. Additionally, the affinity of
the peptide can be analysed directly due to the exchange of one solution to another
one without altering the rest of parameters of the system. From the quality of the
spectra and the reproducible behaviour of EB with the different ions, it is possible to
90
Figure 4.21. SEIRA difference spectra of the ion-exchange performed with the tBLM (A) and tHLM (B) systems.
All spectra were measured at 25° C and used the tBLM or tHLM as reference spectrum. For both systems:
EB:Na+ complex is black, EB:Cs+ complex is blue and EB:K+ complex is red.
91
assume a higher affinity of EB to K+ than to the other two ions. Experiments performed at
a 10 times lower EB concentration supports this statement, since the absolute intensity
of K+ spectrum increased notably compared to other ions. The fact that the amide I
band in the Na+ spectrum appears at 1662 cm-1 instead of 1660 cm-1 (Na+ incubation
position) reveals the presence of residual EB:K+ species, which is another evidence for
the higher affinity of EB to K+ ion. The very same effects were observed in the
experiment performed using the tHLM system (data not shown). Nonetheless, the partial
reversibility of the spectral changes of EB in presence of K+ and Na+ may indicate a
preference for K+ over Na+, but with a much lower selectivity as explained by Pressman
(1976) and Kamyar et al (2004). Both studies reported that the affinity of EB towards K+
ion is significantly lower than in the case of Valinomycin and therefore, EB shows
reduced K+/Na+ discrimination.
4.2.2.2. IR spectroscopy of enniatin B
IR spectroscopy was performed for the solid and solution state of the EB to structurally
analyse the peptide outside the tBLM system. The result of the EB solid-state
diamond-ATR-IR spectrum of the lyophilized peptide is shown in figure 4.23. The EB
solid-state spectrum shows similar spectral contributions as the EB SEIRA spectrum in
figure 4.19, with comparable band shape and positions for the ν (CHn) stretching modes
at 2961, 2930 and 2873 cm-1, as well as the bands in the lower region. The backbone
bands appear at 1735 cm-1 (CO) and at 1663 cm-1 (amide I). The solid-state sample
lacks the molecular interactions between the peptide and the ions, and of both with
Figure 4.22. SEIRA difference spectra for the ion-exchange reversibility of EB:Na+ (black) complex after EB:K+
(red). All spectra measured at 25° C.
92
the solution. Therefore, this result allows identifying the characteristic vibrational modes
of the EB peptide. The similarity of the spectra from the two techniques also confirms the
presence of EB in the tBLM system analysed by SEIRA.
The FT-IR transmission spectra of the EB in ion-free solution and of the respective
complexes with Na+, K+ and Cs+ ions are shown in figure 4.24. The aim of this analysis
was to evaluate the peptide behaviour aqueous conditions in absence of the
membrane. The second derivative of the spectra was used for the determination of the
peak positions of the bands, and in all three cases the CO band appears at ca. 1746
cm-1 (except in presence of Cs+, see below), which is in line with literature and supports
the complexation of ions using the amide O-atoms. The most significant change was
observed between the ion-free EB and the K+ complex with an up-shift of the amide I
band from 1656 cm-1 to 1665 cm-1, while the CO remains at a similar position. The
positions for the spectrum in presence of K+ ions are in line with the SEIRA spectra and
thus confirm the assignment of these spectral positions to the EB:K+ complex. The very
broad bands together with the similar spectral positions in absence of ions and with Na+
suggest that in both cases the same species is present. Thus, this indicates that in Na+
presence most of EB remains in its uncomplexed form. These results support the
interpretation of higher affinity of the peptide for the K+ ion as described above. For K+
ion, the peak positions reveal similarities to the spectra detected in tBLM, and suggest
that both in solution and tBLMs EB forms a EB:K+ complex as a major component. In the
case of Na+ ion, the main fraction contributing to the spectra appears to be the
ion-free EB. The difference in the absolute positions of peaks in tBLMs and in solution can
be explained based on the polarity of the different media.
Figure 4.23. ATR-diamond IR spectrum of the dried EB peptide measured at 25° C.
93
In the present study, the intention was to measure the peptide in conditions as close as
possible to the SEIRA experiments. For this reason, the IR transmission spectra of the EB in
solution were recorded using a mixture of the corresponding buffer with a low
percentage of EtOH. Most of the studies described in literature, use organic solvents like
acetonitrile (ACN) due to the hydrophobicity of the peptide. The work used as
reference for enniatins characterization is the work offered by Ovchinnikov et al. in 1974
[61], which presents a detailed analysis of this peptides with focus on EB and its
complexes. Table 4.6 shows the comparison of the two studies. The authors analysed
the crystalline samples of EB peptide and the dissolved complexes with different ions in
ACN by IR spectroscopy. With attention on the backbone-bands region of the solid
state, the CO is shown at 1744 cm-1 and the amide I at 1670 cm-1 with a shoulder at ca.
1657 cm-1. Unfortunately, these authors used a different method for the investigation of
the solid peptide (KBr pellets of crystalline EB), which complicates a direct comparison
of the spectra. On the other hand, the band positions of the complexes in ACN show
more similarities to the spectra measured in IR transmission and SEIRA. The Na+ complex
shows a broad amide I band, and there is an up-shift of this band for the K+ complex,
even though is only 1 cm-1 in ACN compared to the 8 cm-1 in buffer-EtOH. As mentioned
before, this supports the previous interpretation that EB did not form distinct complexes
with the chosen ions in ACN. The changes for the ester CO mode were minimal as well
as in our study.
Figure 4.24. FT-IR transmission difference spectra of the EB peptide alone (green), in presence of Na+ (black), in
presence of Cs+ (blue), and in presence of K+ (red). All spectra were measured at 25° C, in an ion-buffer and
ethanol mixed solution, and using the solution spectra as background.
94
Table 4.6. IR spectroscopy data comparison of the backbone vibrational modes of EB in this project vs
Ovchinnikov 1974.
SEIRA
(Buffer)
IR transmission
(Buffer-EtOH)
Ovchinnikov 1974
(ACN)
ν CO
cm-1
ν amide I
cm-1
ν CO
cm-1
ν amide I
cm-1
ν CO
cm-1
ν amide I
cm-1
EB
-
-
1746
1656
1744
1666
EB:Na+
1747
1660
1747
1656
1746
1665
EB:K+
1748
1666
1746
1665
1744
1666
EB:Cs+
1747
1660
1740
1659
1645
1666
4.2.2.3. Theoretical calculations of enniatin B ion complexes
Using DFT calculation
5
it was attempted to provide a direct assignment of the SEIRA
spectra to certain EB:ion species with a specific stoichiometry. Initially, only the dihedral
angles of the ring backbone were fixed to 180° during geometry optimization. After this
initial optimization, the geometry was optimized without constraints. The starting point
was the crystal structure described in [77], which provides the positions of the atoms of
the EB peptide. These conditions were first applied to EB:Na+ 1:1, EB:Na+ 2:1, EB:K+ 1:1,
and EB:K+ 2:1. Figure 4.25 shows the most favourable structures obtained from the DFT
analysis of EB and its ion complexes. In the case of the EB:Na+ complex, the 1:1
stoichiometry was most favourable, since in the 2:1 complex one of the rings dissociates
after several optimization cycles and the calculation does not converge. This is in
contrary to literature where a ratio 2:1 was suggested for Na+ ion [76]. For the EB:K+
complex, a 2:1 coordination resulted the most suitable structure, which is in line with the
proposed ratio in previous studies [61] [76]. In the 1:1 arrangement, the K+ ion gets
displaced to yield a structure in which solely the amide CO groups coordinate the ion in
an ―open-faced-sandwich-like‖ manner. This structure would not screen the charge of
K+ ion and thus, is not appropriate for accommodation of the ion in the membrane. A
similar process is observed for the 1:1 EB:Cs+ complex, which can be excluded as a
possible ratio in membrane. However, in literature it was suggested that Cs+ forms either
2:1 or 3:2 complexes. In the 3:2 stoichiometry, Cs+ can theoretically be complexed via
different combinations of amides and esters of two EB molecules as shows the
representation in figure 4.26.
5
Dr. Jacek Kozuch provided the DFT calculation for the enniatin B:ion complexes structures, and the respective
calculated spectra.
95
Figure 4.25. 3D representation of the most stable structures resulted from the DFT calculations of EB (top) and
the respective ion complexes with Na+ (second top), K+ (second from bottom) and Cs+ (bottom). The figures
on the left are the side images, and the right figures are the top images of the corresponding structures.
Atoms are depicted as spheres and the covalent bounds as grey lines. The central ion is shown in purple
colour, oxygen atoms in red, nitrogen atoms in blue, carbon atoms in grey, and hydrogen atoms in white.
EB peptide
1:1 EB:Na+ Complex
2:1 EB:K+ Complex
2:1 EB:Cs+ Complex
96
Among these possibilities, the 3:2-AAEA-complex was the most probable 3:2-structure in
a hydrophobic environment (n-pentadecane = membrane core), with a difference to
the other two combinations by 72.7 kJ/mol or 88.2 kJ/mol (see table 4.7). Taking into
account a theoretical reaction 4.1, a energy difference of ΔG = -11.4 kJ/mol is
obtained demonstrating that the 2:1 complex is predicted to be the preferable
complex with Cs+ within the hydrophobic core of a membrane. These findings for Cs+
ion complex disagree to the 1:1 stoichiometry (―open-faced-sandwich-like‖ complex as
the 1:1 with K+ explained above) suggested in the DFT study of EB:Cs+ by Makrlìk et al.
(2014).
4.1
Figure 4.26. Schematic representation of the amide/ester ligand combination for the 3:2 conformation of the
EB:Cs+ complex. The arrows indicate CO ligand-ion coordination. The amide ligands are depicted in black,
while the ester ligands are in red. The cream-colour spheres represented the Cs+ ion. The codes indicating
the ligand combinations are shown bellow each representation.
97
Table 4.7. Sum of electronic and thermal Free Energies for the calculated complexes of EB with Cs+ ion and
EB alone. The unit ―H‖ stand for Hartree/Particle, where 1H = 2.6254995 MJ/mol.
Calculated structure
Free Energies
EB32Cs_AAEA
- 6436.792218 H
EB32Cs_AEAE
- 6436.764545 H
EB32Cs_AAEE
- 6436.758624 H
EB21Cs
- 4284.600798 H
EB
- 2132.400694 H
The calculated spectra of the most favourable species, i.e. EB, EB:Na+-1:1, EB:K+-2:1,
and EB:Cs+-2:1, are compared in figure 4.27 to the IR spectrum of dried EB and the
SEIRA spectra of EB in the tBLM system in presence of the different ions. The spectra of
the complexes resulted from the constrains-free calculations of the structures in
n-pentadecane as hydrophobic solvent to mimic the membrane core environment
using a polarisable continuum model. For the dried EB structure, DFT spectra were
calculated in n-pentadecane and in water environment as well. To analyse the band
composition of the experimental spectra, a fitting of both peptide bands in the SEIRA
spectra was carried out after removing the water contribution. The red and blue
spectra correspond to the Lorentzian and Gaussian functions, respectively. All four
spectra evaluated here show two components. The band positions for each spectrum
are indicated in table 4.8, along with the DFT calculated frequencies. In principle, one
can assume that the pair of Lorentzian bands refers to a very homogeneous
environment (hydrophobic membrane core), while the Gaussian-shaped bands
describe a heterogeneous environment involving different interactions, i.e. the head
group region where H-bonds, ionic interactions and other specific interactions are
possible. Excluding EB:Na+ 1:1 complex, the relative intensities of the Lorentzians (SEIRA
spectra) resemble very well the CO/amide I ratio obtained from the DFT calculations. In
the case of the IR spectrum of dried EB, it appears to contain a mixture of dried
(Lorentzian, red) and hydrated (Gaussian, blue) peptide. The Lorentzian components of
dried EB at 1713 and 1664 cm-1 look much alike to the DFT spectrum in the hydrophilic
water environment of the polarisable continuum model, which suggest an polar
environment due to the presence of water molecules or a crystalline fraction, in which
dipole moments are oriented in parallel. The similarities between Na+ and Cs+ ions are
reflected in the Lorentzian component of the amide I of the respective SEIRA spectra,
appearing in both cases at 1662 cm-1. In the DFT spectra a different trend is observed
for the EB:ion complexes with an increasing frequency from Na+ over K+ to Cs+. This
contradicts with the trend of the Lorentzian fit and of the overall band positions in the
SEIRA spectra, where positions for Na+ and Cs+ are similar and the EB:K+ complex is
blue-shifted.
98
Table 4.8. Peptide band (CO and amide I) comparison between experimental, component analysis and DFT
spectra. The G and L letters corresponds to Gaussian and Lorentzian functions, respectively. The nP refers to
the conditions within the membrane and H2O to solution-membrane interface.
CO band
(cm-1)
Amide I band
(cm-1)
EB dry
1735
1663
ATR—IR
1735G
1712L
1666L
1656G
Fitting
1715nP
1712H2O
1684nP
1660H2O
DFT
EB:Na+
1747
1660
SEIRA
1746G
1733L
1662L
1648G
Fitting
1698nP
1694H2O
1648nP
1633H2O
DFT
EB:K+
1749
1666
SEIRA
1748G
1737L
1666L
1656G
Fitting
1733nP
1722H2O
1668nP
1650H2O
DFT
EB:Cs+
1747
1660
SEIRA
1747G
1732L
1662L
1648G
Fitting
1729nP
1717H
1669nP
1653H
DFT
Figure 4.27. Band fitting (left spectra) using Lorentzian (red) and Gaussian (blue) functions of the FT-IR
spectrum of dried EB (top) and the SEIRA incubation spectra of EB with Na+, K+ and Cs+ (bottom). The
positions corresponding to the CO and amide I bands are shown in red colour. The DFT calculated spectra
(right side) of the most stable structures for EB (top) and EB complexes with Na+, K+ and Cs+ (bottom) in a
hydrophobic environment (membrane-core) are shown in black. The positions corresponding to the CO and
amide I bands are shown in black colour as well. The grey spectra and positions shown for dried EB
correspond to the peptide in a polar environment.
99
The most important outcome from this multifaceted analysis is the assignment of the 2:1
stoichiometry to the EB:K+ complex, which is supported by SEIRA and DFT results. For Na+
and Cs+ SEIRA and DFT spectra showed less conclusive assignments. The most
favourable structures for Na+ and Cs+ ion complexes of EB in a membrane core-mimic
environment were 1:1 and 2:1, respectively. The assignments presented here for these
ions contradict the suggestions in previous literature.
4.2.3. Mechanism of membrane incorporation and ion binding
To follow the membrane binding kinetics of the peptide and analyse the mechanism of
incorporation, EB was incubated during 70 minutes with the tBLM system for each ion
and monitored by SEIRA. The intensities of the CO and amide I bands are represented
as a function of time in Figure 4.28 (top). The sigmoidal curve indicates an
―auto-catalytic‖ insertion of EB into the lipid bilayer. A suggested two-step binding
process, neglecting the back reactions
6
, is shown in figure 4.29. In the first step,
multimeric species of EB interact with the membrane (PPreAd) and convert into
monomers on the surface (PAd). This step consists of two reactions expressed in reaction
4.2, the association to the membrane, and reaction 4.3, the dissociation of the multimer
species into monomers. The second step of this membrane-binding process is
represented by reactions 4.4, a slow incorporation of one complex into the core of the
membrane, and 4.5, the auto-catalytic reaction where more EB complexes are inserted
into the membrane (PMem).
6
The kinetic equations to fit the binding kinetics spectra of the systems were developed in collaboration with Dr. Jacek
Kozuch.
Figure 4.29. Schematic representation of the reaction for the Peptide-membrane binding kinetics process.
The reaction constant for each step is indicated on the arrows as the Κ of the respective specie, being so:
PreAds the pre-adsorbed multimers; Ads the species adsorbed on the membrane surface; Inc the random
incorporation of species; and Mem the species incorporated in the membrane as a result of the
auto-catalytic process.
100
→
→
→
→
The adsorption of the aggregated species, given by , (reaction 4.2) can be
described by Langmuir-type adsorption. In this way, this reaction is a second-order
reaction considering the peptide in solution (PSol) and the free binding places in the
membrane surface (see eq. 4.6). Once bound to the surface the peptide aggregates
begin to dissociate into monomers. The first term in 4.6 refers to the decrease of the
concentration of the peptide in solution due to the adsorption, and the second term
represents the increase of the concentration due to the dissociation into peptide
monomers. The contrary effect is represented in equation 4.7, since (PPreAd) corresponds
to the multimeric species, which concentration decreases (second term eq. 4.7) with its
decomposition into monomers.
( (
)
)
( (
)
)
The second step treated as an auto-catalytic incorporation (reaction 4.5) requires a
slow random insertion of single EB molecules defined as a stochastic incorporation
(reaction 4.4). The stochastic step is represented in equation 4.8, described as a
Langmuir-like process, which takes into account only the free-spaces in the lipid bilayer.
The concentration of (PAd) depends on three processes. It increases due to the
formation of the monomeric species, decreases due to the stochastic process, and
decreases with auto-catalytic insertion into the membrane core.
( )
( )
4.2
4.3
4.4
4.5
Stochastic
Auto-catalytic
(4.6)
(4.8)
(4.7)
101
Figure 4.28. Kinetic representation of the SEIRA absorbance (mOD) of the CO and amide I bands (purple
dots) against time (min) during the EB:Na+ complex incubation in 100mM NaCl 20mM BTP pH7.4 buffer at
25° C. In purple line there is the fit curve. The green line corresponds to the adsorbed-peptide, the red line to
the peptide in the membrane, and the light blue line to the pre-adsorbed species. The yellow dots represent
the residuals of the fitting.
102
The auto-catalytic incorporation is described as a Prout-Tompkins kinetics, expressed in
the first term of equation 4.9. This stage is concentration-dependent, since it is
necessary for the adsorbed EB on the surface (PAd) to establish contact with the already
incorporated species (PMem) for insertion into the membrane.
( )
( )
The differential equations provided a very good fit to the CO band (experimental data)
of the Na+ complex, shown in bottom of figure 4.28 (purple line). The red line
demonstrates the SEIRA intensity of membrane-incorporated EB, and the light blue and
green refer to the (PPreAd) and (PAd) species, respectively. These results indicate that the
proposed mechanism for the membrane-binding process is consistent with the
experimental data. To achieve a good fit to the data, the previously described
water-correction had to be performed beforehand. The fact that the fit yielded the
observed good result supports the suitability of water-correction in the case of EB.
The same analysis of data was applied to the amide I band of the EB:Na+ complex, as
well to the incubation of the EB:K+ complex. The analysis of the incubation of both
complexes is presented in a comparative manner in figure 4.30 (red lines). The light
green line represents the interfacially adsorbed aggregates (PPreAd); the dark green line
the interfacially (PAd) "monomer/complexes"; and the blue line the incorporated
species (PMem). The two complexes differ significantly in their binding kinetics in which
Na+ shows an ―auto-catalytic‖ process whereas for K+ membrane binding is significantly
accelerated. These results support the suggestion of different stoichiometry for these
two ions, 1:1 for Na+ and 2:1 for K+, since different ion-coordination might lead to
different kinetics. Another interesting observation is that in both cases the band ratio
maintains constant during the incubation time. Compering to the SEIRA results in the
independent incubation of the ion complexes, Na+ complex shows pronounced
changes in band ratio after incubation (0.52 to 0.85) whereas the band ratio for the K+
complex is preserved (0.72 to 0.71). A tentative interpretation might be that this is a
result from the different interaction between EB rings in the interfacially space and those
being incorporated into the membrane core for the Na+ complex compare to the K+
complex. In the case of Cs+ the kinetics could not be fit to the presented model, which
may suggest an even slower incorporation into the tBLM.
The intensity evolution of the peptide bands during the ion-exchange experiment is
presented in figure 4.31 together with the band ratio analysis, both as a function of
time. Interestingly, when exchanging the ions from Na+, to Cs+ and then to K+, there is a
decrease in the intensity of both bands in the beginning followed by an increase over
time. On the other hand, the inverse happens during the evolution of the band ratios,
which show a fast increase accompanied by a quick decrease and stable value in
(4.9)
103
Figure 4.30. Kinetic representation of the SEIRA absorbance (mOD) of the CO and amide I bands (black dots)
against time (min) during the EB:Na+ and EBK+ complex incubation at 25° C. In red line there is the fit curve.
The dark green line corresponds to the adsorbed-peptide, the blue line to the peptide in the membrane,
and the light green line to the pre-adsorbed species. The bottom representations are the band ratio for each
complex.
Figure 4.31. Kinetic representation of the SEIRA absorbance (mOD) of the CO (black dots) and amide I (red
dots) bands against time (min) during the ion-exchange at 25° C (left graph). The graph on the right is a
representation of the band ratio for each complex.
104
both exchange processes. Even though no structural information can be drawn from
these observations, one plausible interpretation for this phenomenon is the migration of
EB molecules to the membrane-solution interface in order to coordinate ions. This fact
reflects the mobility of EB within the membrane and the great capability of the peptide
to facilitate ion transport.
4.2.4. Conclusions
The SEIRA spectroscopy, kinetics and theoretical calculations results presented in this
work provide novel insights into the structural understanding of the EB peptide. The
interpretations made here attempt to bring more light into the controversial
characterization of EB ion complexes, membrane interaction and mechanism of
action. The main conclusions are summarized down below.
The peptide-membrane interaction of EB characterised by SEIRA spectroscopy using
tBLM systems involves different binding processes, which depend on the peptide
concentration and on the ion available. The findings presented here demonstrate the
membrane incorporation of EB complexes into the lipid bilayer. The kinetics of this
interaction is represented by the suggested binding mechanism, where multimer
pre-adsorbed species are bound to the surface, followed by an auto-catalytic
incorporation process. The Na+ ion complex successfully exemplified this mechanism,
which showed to differ from the binding process observed for the K+ ion and thus,
denoting the presence of different species. Additionally, the kinetics observed during
the ion-exchange provided evidence for the EB capability to form ion-transport and
travel through the membrane.
The ion affinity evaluation by the SEIRA spectroscopy showed a higher affinity of EB to K+
ion. However, the ion selectivity was significantly lower than what is suggested for the
well-known Valinomycin or Beauvericin [61]. This was evident in the back reaction from
K+ complex to Na+, denoting an interesting degree of reversibility between the two. The
partial reversibility and affinity might have a great impact in the peptide mechanism of
action, since in in vivo conditions the homeostasis between these two ions is essential
for the cell. Thus, a deregulation in the concentration of these ions between the intra
and extracellular space can lead to membrane permeability, and later, to cell
apoptosis. This is a plausible hypothesis suggested for the antimicrobial activity of EB,
which depends on the peptide concentration and ion availability, as explained in the
kinetics analysis.
Complex stoichiometry has been one of the most debated features of EB. This study
offers a new interpretation on EB ion complex formation. The SEIRA and IR spectroscopy
combined with the DFT calculations revealed that EB is likely to form a 1:1 complex with
Na+ and a 2:1 complex with K+ and Cs+. The DFT results, structure and spectra, showed
the formation of the 2:1 EB: K+ complex in which the C=O ligands from the
N-methyl-L-valine residues contribute predominantly to the coordination of the ion,
105
which is in agreement with the findings by Ovchinnikov et al. for a 2:1 stoichiometry [61].
The results for the Na+ and Cs+ ions were less clear, both showed same peptide bands in
the SEIRA spectra with slight variance in intensity. However, the consistent shifts of the EB
bands in SEIRA when exchanging the buffer solution containing K+, supports the
assumption that these Na+ and Cs+ ions present a similar contribution of the C=O
ligands that is distinctly different to the EB conformation adopted in the K+ complex.
Consequently, taking into consideration the results provided, the present study suggests
the following complex stoichiometry for the three ions evaluated:
These results set a starting point for further investigation of EB complex formation and
mechanism of action interacting with tBLM systems. This work, combined with previous
structural and functional studies of the peptide, opens new possibilities for the
interpretation and prediction of EB complexes stoichiometry.
4.3. Antimicrobial peptide arenicin1
In this section, the POPC/POPG tBLM system was used as membrane-mimetic system to
study the AMP arenicin1 (A1) by SEIRA spectroscopy. This cationic AMP contains
features that are interesting for a spectroscopic evaluation. It is characterised by an
antiparallel β-sheet conformation with nine intrabackbone hydrogen bonds (H-bonds).
The high percentage of hydrophobic amino acids can explain the interaction with lipids
in the membrane. The (6+) net charge is given by arginine (Arg) amino acids distributed
along the peptide, achieving a high amphipathic surface. The disulphide bridge
between the Cys in position 3 and 20 closes a loop of 18 residues. The A1 peptide has
shown antimicrobial activity against gram-negative bacteria suggested to be a pore
forming peptide, whose activity was increased at 4º C and 37º C [63]. The scarce
literature presenting the peptide-membrane interaction leaves plenty of room for
mechanistic investigations. Here, a series of vibrational spectroscopy experiments
provided complementary structural information to the SEIRA study performed at three
temperatures: 4º C, 25º C and 37º C. The preliminary results of the molecular dynamic
(MD) simulations of A1 using a POPC/POPG membrane in similar conditions to the SEIRA
experiments are presented at the end of the section. As a result, this investigation offers
the foundation for a structure-function study of the peptide-membrane interaction of
A1 under different conditions using the POPC/POPG tBLM system.
4.3.1. IR and Raman spectroscopy of arenicin 1
The IR and Raman experiments were performed to obtain an overview of the peptide
features in solid and solution state without the membrane environment.
1:1 for EB:Na+ 2:1 for EB:K+ 2:1 for EB:Cs+
106
4.3.1.1. ATR-IR spectroscopy of solid arenicin 1
This measurement was done with a solid sample of A1, using a diamond-ATR-IR setup.
The resulting spectrum is shown in figure 4.32. Here, the peptide is static, but distributed
in a preferential direction within the powder. The powder is actually composed of
crystals of A1, which in consequence gives a certain orientation. The bands assigned in
the spectrum represent a β-sheet structure with a turn, as described in literature [19] [36].
It is possible to divide the spectrum in three regions: the high region from 3300 to 2700
cm-1, the amide region from 1800 to 1500 cm-1, and the lower region from 1400 to 500
cm-1. In the high region there is a broad band at 3277 cm-1 representing the ν (NH)
stretching mode, known as well as amide A band, which is part of a Fermi resonance
doublet with the second component at ca. 3100 cm-1 [39]. Between 3000 and 2700 cm-1
there are weak bands corresponding to the symmetric and asymmetric ν (CH)n
stretching modes. It is important to notice the absence of the 2600 cm-1 band of the ν
(SH) stretching mode, which indicates that all sulphur atoms are forming a disulphide
bridge (S-S) closing the loop in the peptide.
The assignments for the bands below 1800 cm-1 are described in table 4.9. The amide
region presents multiple bands due to coupling of the amide I and amide II vibrational
modes. The amide I band appears to be split into two components, a strong band at
ca. 1630 cm-1 and a weaker band at ca. 1695 cm-1. The former having a main
contribution from the out-of-phase ν (C=O) stretching mode [111]. The transition dipole
moment of these bands presents a certain orientation in respect to the direction of
propagation of the chain. The main absorption at 1630 cm-1 corresponds to the
Figure 4.32. ATR-IR spectrum of the solid state of A1 peptide. The assignments of the corresponding vibrational
modes are presented in table 4.9.
107
transition dipole moment perpendicular to the chain direction, and the weaker parallel
to the chain [39]. The amide II band (in-phase ν (C-N) stretching and (N-H) bending
modes) is split as well, with a strong band at ca. 1558 cm-1 and a weak broad band at
1533 cm-1, corresponding to the parallel and perpendicular polarization of the transition
dipole moment, respectively [21]. The sharp band at ca. 1515 cm-1 is assigned to the
strong (C-C) ring vibrational mode from Tyr (tyrosine) [114].
Table 4.9. Band assignment for the A1 vibrational modes in the ATR-IR spectrum in figure 4.32. Band
characteristic indicated in the position as: W, weak; B, broad; S, strong; M, medium [19] [38] [39] [112] [113].
Position
(cm—1)
Description
1695 W
Amide I (minor component)
1668 B
Β-turn
1653
Unordered structure
1631 S
Amide I (major component)
1558
Amide II (major component)
1533
Amide I (minor component)
1515 S
C-C aromatic ring breathing Tyr
1202 S
Indole ring stretching Trp
1183 S
COH deformation (sensitive to H-bonding)
1137 S
NH indole deformation
838
801
Tyr Fermi doublet
722M
Rocking of adjacent in-phase CH2 couple
In the lower region from 1500 cm-1 there are small bands corresponding to the (CH)n
bending and rocking vibrations, (NH) deformation, and other amino acid side-chain
characteristic bands. The band marked at 1205 cm-1 has a left-shoulder, which
indicates an overlap between the amide III ((CN) stretching and (NH) bending, usually
1220 cm-1) and the strong vibration from the (C-C) indole ring from Trp (Tryptophan) at
1203 cm-1 [39]. The low-frequency shoulder of the latter band may overlap with modes
including wagging and deformation coordinates of CH2 and COH groups [39]. The
band at 1137 cm-1 corresponds to a (NH) indole deformation mode. The bands at 838
and 801 cm-1 are assigned to the Fermi doublet of Tyr. The band at 722 cm-1 is assigned
to the in-phase mode coupling of adjacent CH2 groups from the side chains [39].
108
4.3.1.2. FT-Raman spectroscopy of solid arenicin 1
FT-Raman spectroscopy can offer a complementary analysis of the vibrational modes
of the peptide. In that way, modes that are not IR active can be Raman active, and
some modes can be active for both techniques. While the rule for IR active modes
depends on the change on the transition dipole moment with a normal coordinate,
Raman activity requires a change in the polarizability. The Raman spectrum of a solid
sample of A1 is shown in figure 4.33. This spectrum shows the ν (CH)n stretching modes in
the region from 3062 to 2713 cm-1; they are more intense in Raman than in IR due to the
great change in polarizability of these groups. In the Raman spectrum of A1, there is
absence of the SH vibrational mode (2550-2600 cm-1) as shown as well in IR. The
assignment of the bands in the region between 1700 and 500 cm-1, are shown in table
4.10. There are important bands to be mentioned, like the ν (C=O) stretching at
1671 cm-1 and the Fermi doublet at 836 cm-1 and 801 cm-1, both characteristic modes
of Tyr. The Trp characteristic bands appear at 1551 cm-1 (NH deformation) and
1011 cm-1 (indole ring breathing). The (C-S) and (S-S) modes of Cys and the disulphide
bridge at 670 and 504 cm-1, respectively. Combining both analysis, IR and Raman, it
was possible to offer a detailed fingerprint of the A1 peptide composition.
Figure 4.33. FT-Raman spectrum of solid state A1 peptide.
109
Table 4.10. Band assignment for the A1 vibrational modes in the Raman spectrum in figure 4.33. Band
characteristic indicated in the position as: W, weak; B, broad; S, strong; M, medium [114] [115] [116].
Position
(cm-1)
Description
Position
(cm-1)
Description
1671S
Tyr C=O stretching (amide I)
1150
Ring deformation
1616S
Tyr Aromatic ring from
1126
Trp NH deformation indole
1597
CO asymmetric stretching
1075W
Trp C-H scissoring pyrrole ring
1578
COO- asymmetric stretching
1011S
Trp indole ring breathing
(strong Van der Waals interactions)
1551S
Trp NH deformation
960
Trp C-H twisting benzene ring
1517W
Tyr Aromatic ring stretching
878
Trp C-H scissoring indole ring
1460
1360
CH2 deformation
853
Trp C-H bending pyrrole ring ca. 849
Tyr ring breathing
1438-
1424
CH2 scissoring and Trp indole
ring stretching
836B
801
Tyr Fermi doublet
1338
1360
Trp Fermi resonance doublet
(in a hydrophilic environment)
758S
Trp indole ring breathing
1255
Trp C-H rocking of benzene
670W
Cys C-S stretching
1230
Amide III (β-sheet) and Tyr O-C
ring stretching
643S
Tyr ring deformation
1205
Tyr C ring stretching
p-substituted benzene
599
Trp pyrrole ring deformation
1174M
CH deformation indole ring Trp
504W
S-S stretching
4.3.1.3. FT-IR transmission of arenicin1 in solution
A sandwich FT-IR transmission setup was used to measure a saturated solution of A1 in
NaCl-BTP buffer. The resulting spectrum is shown in figure 4.34. The spectrum shows
similar features in the amide region as for the solid state, but with broader bands. As in
the solid state IR, A1 exhibits a β-sheet structure, which reflects the high instability of the
peptide in the different conditions [63]. The assignment of the main bands appearing in
the amide region is shown in table 4.11. The amide II region appears like a big broad
band, but actually it is an overlap of at least three bands at 1561 cm-1, 1532 cm-1, and
the more defined band at ca. 1516 cm-1 for the Tyr ring. Unfortunately, it is more difficult
to discriminate the bands in the lower region in the solution spectrum, which is due to
the lower concentration of A1 as well as the broadening of bands sensitive to
H-bonding. The A1 solution spectrum gives a closer idea of what can be expected in
the SEIRA spectra.
110
Table 4.11. Band assignment for the A1 vibrational modes in the FT-IR transmission spectrum in figure 4.34. The
symbols // and ⊥ indicate a parallel or perpendicular direction, respectively.
β-sheet amide bands
(cm-1)
1692
main Amide I // to sheet direction
1680
β-turn
1622
minor Amide I ⊥ to sheet direction
1561
minor Amide II ⊥ to sheet direction
1532
main Amide II // to sheet direction
Summary
The analysis of the A1 powder by ATR-IR and Raman spectroscopy, offered a
fundamental description of the peptide structure and composition. The information
collected revealed a β-sheet conformation of A1, and vibrational modes of distinctive
amino acid side chains. The bands corresponding to Trp and Tyr amino acids can be
used for the interpretation of intra-strand interactions, for example the 1011 cm-1 band
of Trp in the Raman spectrum representing strong van-der-Waals interactions. On the
other hand, the FT-IR transmission spectrum in solution shows the hydrophilic character
of the peptide reflected in the broadening of the main amide bands due to interaction
with the solution. Overall, this section offered a complete analysis of the A1 peptide
structure as a starting point for the SEIRA study.
Figure 4.34 FT-IR transmission difference spectrum of the saturated solution of A1 in 100 mM NaCl 20 mM BTP
pH 7.4 buffer at room temperature.
111
4.3.2. SEIRA and EIS study of the POPC/POPG membrane
systems
The construction of the tBLM with POPC/POPG lipids was done following the same steps
as for the pure POPC system, aiming a 80:20 ratio of lipid molecules. The main
difference is the mix of POPC with the negatively charged POPG lipids, which required
the addition of a polar solvent to the preparation. The resulting SEIRA difference
spectrum of the 80:20 POPC/POPG tBLM at 25º C is shown in figure 4.35 (black
spectrum). Both lipids have the same aliphatic chains composition (same melting point
transition temperature, - 2º C), which results in equivalent band positions and shape as
for the pure POPC tBLM. Due to the temperature requirements during the study of A1
peptide, the tBLM system was evaluated at three temperatures: 4º C, 25º C and 37º C,
as shown in figure 4.35. The POPC/POPG tBLM system spectrum is conserved in the
regions of interest for the three temperatures, allowing for the analysis of the peptide
bands at each condition. However, it was possible to observe temperature effects on
the silicon prism in the low frequency region below ca. 1500 cm-1(see appendix figure
A.7 for full spectra), where bands appear like a mirror effect, positive at 37º C and
negative at 4ºC.
POPC/POPG tHLM membrane systems
The lipid monolayer was built in the same way as for the POPC tHLM system presented
in section 5.2.1. First, a pure WK3SH SAM was assembled on the Au-film. Afterwards, the
POPC/POPG lipid vesicles were added. The SEIRA difference spectrum of the
Figure 4.35. SEIRA difference spectra of the POPC/POPG tBLM at 4º C (blue), 25º C (black) and 37º C (red),
using the SAM spectra as reference.
112
POPC/POPG tHLM system resembles the one for bilayers, but with slight lower intensities
(data not shown).
EIS of the POPC/POPG tBLM and tHLM systems
The impedance data and its representation were treated in the same way as for the
POPC systems. Figure 4.36 shows the comparison of EIS spectra of both tBLM systems.
The values obtained from the fitting of the Cole-Cole plots of the POPC/POPG systems
are presented in table 4.12. The high percentage of POPC lipids in the mix and the
similarity of the two lipids are reflected by similar capacitance values of the bilayer and
the total system, Cbilayer and Csystem respectively, for both tBLM and tHLM. The
capacitance of the system for the POPC/POPG tBLM was (0.61 ± 0.02) µF cm-2, which is
in the range of the pure POPC tBLM (0.62 ± 0.03) µF cm-2. The tHLM capacitance was
slightly lower than for the pure POPC, (0.66 ± 0.08) µF cm-2 instead of
(0.70 ± 0.06) µF cm-2, but still consistent considering the errors. These results are coherent
since the only difference in the system is the negatively charged POPG lipid.
Table 4.12. Results from the EIS analysis obtained by the fitting of the spectra to the equivalent circuit
described for the POPC/POPG tBLM and tHLM systems.
RSolvent
Ω
Rbilayer
KΩ cm2
Cbilayer
µF cm-2
Rspacer
KΩ cm2
Cspacer
µF cm-2
Q
µF s α-1
cm-2
α
Csystem
µF cm-2
tBLM
337.2
2.00
1.81
325.6
4.81
6.04
0.82
0.61
tHLM
190.4
2.09
3.00
414.1
5.07
9.58
0.78
0.66
Figure 4.36. Impedance spectra represented in a Cole-Cole plot for the POPC/POPG tBLM (light blue) and
the POPC tBLM (red).
113
4.3.3. Peptide-membrane interaction study by SEIRA
4.3.3.1. Arenicin1 interaction with POPC/POPG tHLM system.
As initial approach, A1 was incubated at 37º C with the POPC/POPG tHLM system to
test the peptide interaction with a lipid monolayer. The SEIRA difference spectrum is
shown in figure 4.37. For the first time, it was possible to observe the A1 β-sheet structure
in SEIRA spectroscopy using a tHLM system. Interestingly, the amide region of the
spectrum resembles quite well the spectrum obtained by the FT-IR transmission
experiments of the solution sample of A1. The changes in the position of some of the
bands are attributed to the interaction of the peptide with the membrane. The main
absorption of the amide I mode appears at 1630 cm-1 and the weaker absorption at
1691 cm-1. In between these two bands there is a broad band at ca. 1677 cm-1
assigned to the β-turn conformation. The two components corresponding to the amide
II mode are at ca. 1562 cm-1 (strong band) and at 1532 cm-1 (weaker band). The band
at 1516 cm-1 corresponds to the stretching mode of the aromatic ring of Tyr, as shown in
the solid and solution spectra, this band is a marker for this residue. On the high-energy
side of the spectrum, the (CH)n vibrational modes appear as negative bands, which
may indicate changes in the lipid orientation due to the interaction with A1. The region
bellow 1400 cm-1 seems to be conserved compared to the FT-IR transmission spectrum.
These preliminary results show the capability of A1 to interact with a lipid monolayer
and to disturb the system (after 5 hours incubation).
Figure 4.37. SEIRA spectrum of the A1 peptide interacting with POPC/POPG lipid monolayer, using the tHLM
spectrum as reference.
114
4.3.3.2. SEIRA spectroscopy of arenicin1 interaction with POPC/POPG
tBLM system: temperature dependence
As explained in the beginning of the section, this study aimed to evaluate the
behaviour of A1 in presence of the POPC/POPG tBLM at three temperatures, 4º C,
25º C, and 37º C. The SEIRA difference spectra of the independent incubation of A1 at
each temperature, after 3 hours, are presented in figure 4.38. The second derivatives of
the spectra revealed that the bands appear in the same amide band position for the
three temperatures. In none of the cases there is a band for the (SH) mode, which
means all Cys residues are connected closing the peptide loop. Contrary to what was
observed above for the tHLM system, the lipid bands at ca. 1737 cm-1 (doublet for the
COO- of the lipids) and those in the (CH)n region are positive in this case. In both
membrane systems A1 causes changes in the lipid organisation as a result of the
interaction. It is possible to identify lipid changes in the high region due to the
difference in band shape; For the lipids the bands corresponding to the asymmetric
and symmetric stretching mode of the CH2 groups are of higherintensity, while the
peptide (amino acid side-chains) shows higher intensities for the asymmetric and
symmetric modes of the CH3 groups. Although the differences between the three
temperatures were less pronounced than expected, the SEIRA spectra suggest that A1
activity increases with the temperature.
Even if, at first glance, the amide region of the spectra might look different than for the
tHLM, the band position matches to the characteristic bands for the β-sheet structure
Figure 4.38. SEIRA spectra of the A1 peptide binding to the lipid bilayer, tBLM spectrum was used as
reference. The experiment was performed at three different temperatures: 4º C (blue), 25º C (black), and
37º C (red).
115
described in the monolayer evaluation (second derivative assignment). The
interference of water is the likely reason for the variation of the baseline in the amide
region. The water interference along with the lipid band changes indicates the ability of
A1 to disturb the membrane in a deeper manner than for the tHLM system. In other
words, due to the interaction of A1 with the lipid bilayer water leaks through the
membrane, which is reinforced by the water overlapping in the amide region and the
positive band of the OH stretching at ca. 3450 cm-1 (data not shown). As suggested by
Andrä et al. (2008) [63], these alterations can be considered as partial lesions in the
membrane integrity. Additionally, the interaction of A1 with the lipids is practically
spontaneous showing changes in the lipid bands of the SEIRA spectrum after just three
minutes of incubation. This observation indicates that A1 is likely to induce membrane
permeabilisation by a toroidal pore formation [117] or by a carpet model mechanism as
suggested in [63], where accumulated membrane lesions of the affected cells in vitro
resulted in the realising of cytoplasmic material. On the other hand, the results
presented here offer an evaluation of A1 in different states and conditions, and it was
demonstrated that A1 is highly stable and conserves its β-sheet structure in solid, solution
and lipid environment, which disagrees with the observations by Andrä et al. (2008) [63].
Potential dependence study of arenicin1
The aim of the potential-dependence analysis was to evaluate the changes induced
by potentials applied to the system containing A1. The resulting SEIRA difference
spectra at 25º C using the +400 mV spectrum as reference are shown in figure 4.39. The
red spectrum depicts minimal changes in the amide region at +300 mV. However,
when applying lower potentials the changes in the amide I band increase, most
marked at 1630 cm-1. The high frequency region shows practically no change, as well
as in the low region below 1600 cm-1. The changes induced by negative potentials
were not reversible as shown by the spectrum measured after applying +400 mV again.
The changes were visible even after hours under open circuit conditions. It is possible to
conclude that these changes originate from the peptide and not from the background
by comparing the data to the potential control experiment shown in figure 4.40. The
solution beneath the lipids in the tBLM increases by applying negative potentials, as
described in Kozuch, Jacek PhD dissertation [21]. This is observed in the increase of the
intensity of the (OH) stretching and bending modes from water during the application
of negative potentials.
In the next step, the same experiment was performed at 37º C. The SEIRA difference
spectra are shown in figure 4.41, using the +400 mV spectrum as reference. This time
changes were significantly pronounced, not only in the amide region, but also in the
ν (CH)n stretching frequency region. The bands in the high region represent a strong
perturbation of the lipid bilayer arrangement, reflected as well in the raise of the ν (CO)
stretching band at 1737 cm-1. In this case, the changes of the amide I band are
negative indicating an apparent re-orientation of A1 in the membrane. The increasing
ν (OH) stretching band from water adds evidence to the permeabilisation of the
116
Figure 4.40. SEIRA spectra of the tBLM potential control experiment. Potentials were applied from +400 mV
to -400 mV (blue) in 100 mV steps. All difference spectra are calculated using the first +400 mV spectrum as
reference, and measured at 25º C.
117
membrane. As shown at 25º C, the changes induced upon the application of negative
potentials were irreversible and the variations continued after two hours. As overall
observation, applying negative potentials enhanced the membrane perturbation
activity of A1, which was more evident at 37º C than at 25º C.
4.3.4. Molecular dynamics simulation of arenicin1
In order to provide atomistic details regarding the behaviour of A1 in solution together
with the interaction pathway in a lipid environment, all-atom MD simulations of the
peptide (pdb-code 2JSB [63]) employing CHARMM force field were performed. The
investigation was mainly focus on: i) comparative conformational study of the natural
peptide and an analogue without S-S bridge, ii) evaluation of possible dimerization,
and iii) peptide-membrane interaction.
The trajectories showed that both A1 monomers, natural and analogue, adopt an
antiparallel β-sheet with β-turn secondary structure (see figure 4.42). There was not
significant difference between the trajectories of both A1 monomers. Interestingly,
inter-strand H-bonds and non-covalent interactions, like the π-interaction between
Arg16 and Tyr5, are responsible for the overall stabilisation of the structure of both A1
monomers throughout the 300 ns of simulation. This great stability of A1 is in line with the
Figure 4.41. SEIRA spectra of the A1 peptide after applying potentials from +400 mV to -400 mV (blue) in
100 mV steps, using the POPC/POPG tBLM system, measured at 37º C. The +400 mV (green) potential was
applied again as a last step. All difference spectra are calculated using the first +400 mV spectrum as
reference.
118
spectroscopic results obtained from the solid, solution and tBLM experiments presented
here.
The dimerization of the natural A1 (S-S bridge) in solution takes place during the first
60 ns of the MD simulation (see figure 4.43). The main interactions between the two
monomers were observed as π-interactions (at distance < 3 Å) between Trp2-Val15 and
Trp21-Val10 (turn), as well as interactions involving residues Val4, Tyr5, Tyr7 and Arg19 of
one monomer and Val13, Tyr17 and Tyr7 of the other monomer. Dimerization was not
observed for the peptide without the disulphide bridge, although there appeared
random interaction (at distance < 3 Å) towards the end of the simulation, between the
ring of Trp21 and Arg1 (shown in appendix figure A.8).
Figure 5.42. snapshots of 3D representation of the cationic AMP A1 (left) and its analogue without cys S-S
bridge (right). Yellow colour denotes the antiparallel β-sheet conformation while the β-turn is depicted in
turquoise. The N- and C-terminal regions are unstructured represented in white colour. The disulphide bond
between Cys3 and Cys20 is shown in yellow forming a loop in the peptide. The Cys 3 and 20 are depicted in
mauve colour in the analogue peptide.
Figure 5.43. Snapshot of a 3D representation of the cationic AMP A1 dimer in solution. The secondary
structure is shown in similar colour-code as the natural monomer in figure 4.42, but in a more transparent
manner to emphasise the side chain of the aromatic residues involved in the dimerization process. The side
chain of Trp and Tyr are depicted in mauve and violet colour, respectively.
119
Initial
Final
A
B
C
120
According to the initial goal, two natural A1 monomers were used for the theoretical
study in a POPC/POPG 80:20 membrane, mimicking the experimental conditions used
in the SEIRA experiments. The results showed a rapid and strong interaction of A1 with
the membrane in the ns scale. The snapshots from the initial and final stage of the
simulation are shown in figure 4.44. The peptide conserves its antiparallel β-sheet
conformation during the course of the simulation. The images of the final stage show
that the peptide is partially incorporated into the lipid bilayer (top leaflet), and it
prevails in an orientation perpendicular to the normal of the membrane surface.
Moreover, the perturbation of the lipid bilayer structure was produced by the strong
interaction of Arg residues with the negatively charged POPG lipid head groups. This
interaction was accompanied by a disorder of the membrane induced by the
interaction of hydrophobic side-chain groups of A1 with the acyl chains of the lipids
(interacting side-chain groups are shown in mauve colour). In the course of 300 ns of
simulation no dimerization was observed.
To conclude, the results of the trajectories demonstrated the stability of A1 secondary
structure in solution and in presence of the POPC/POPG membrane, as well as the
importance of Arg and aromatic residues for the peptide perturbation of the lipid
bilayer. The important role of Arg residues was also shown by Cho and Lee (2011),
where a modified peptide characterised by reduced net charge exhibited weaker
interaction with the negatively charged membrane and consequently, lower
antimicrobial activity [78]. The perturbation of the membrane provoked by the
interaction of Arg residues with the POPG head groups and subsequent pore formation,
may support a disordered toroidal pore mechanism of action due to the
disorganisation of the lipids, or a carpet model as suggested by Ändra et al (2008) [63].
These mechanisms differ from the proposed toroidal pore formation by dimers of the
arenicin 2 [118]. Nonetheless, this is the beginning of the investigation of A1 peptide with
membranes, and the analysis of further MD simulation scenarios may afford a
comprehensive understanding of the peptide behaviour.
4.3.5. Conclusions.
This work presents a study of the structure and peptide-membrane interaction of the
AMP A1, combining vibrational spectroscopy and membrane model systems. The results
of all the vibrational spectroscopic techniques used here, ATR-IR, FT-Raman, FT-IR
transmission and SEIRA, revealed that A1 adopts a β-sheet with a turn secondary
structure in solid, solution and membrane-like environments. SEIRA experiments showed
that A1 interaction with the lipids is practically spontaneous, which was supported by
the preliminary results from the MD simulations of the system. In this case, the incubation
of A1 at different temperatures (4º C, 25º C and 37º C) showed fewer variations than
expected, though the most pronounced changes in the membrane were observed at
37° C. Additionally, initial results from the potential-dependence evaluation showed
121
that negative potentials below -100 mV provoke irreversible perturbation of the lipid
bilayer, which is significantly increased at 37º C.
The results presented here may be considered as a starting point for further
investigation of A1 in membrane models to evaluate its binding process to artificial
membranes and modes of action. Investigation on different peptide concentrations,
ion strengths of the solution, temperatures, ions and potentials could provide more
insights into A1 mechanism of action, though so far it looks like A1 induces a
disorganization of the lipid bilayer supporting the carpet or disordered toroidal models.
122
5. Conclusions
A nanodisc system was successfully immobilised on a functionalised nanostructured
Au-film for the characterisation of the slow-mutant C128S of the ChR II membrane
protein. The quality of the SAMs was evaluated by EIS and SEIRA spectroscopy. The pure
NTA monolayer showed the most robust results and better reproducibility, contrary to
the defects in the mixed SAM with the 3MP molecule. The ChR II protein embedded in
the nanodisc was bound to the NTA SAM through a His-tag linker in the MSP. Excitingly,
here is shown for the first time the transition of the ChR II C128S mutant from the DA to
the P390 state observed by SEIRA spectroscopy using nanodisc systems. This transition
was induced by illumination of the DA sample with 468 nm (blue) light, showing
changes in the amide I region at (-) 1664 cm-1 and (+) 1653 cm-1 assigned to the
structural changes of the retinal protein. Further transitions were investigated by UV-vis
and FT-IR transmission spectroscopy. In all cases, the changes observed matched the
changes of the ChR II C128S mutant in absence of the nanodisc system, denoting the
conservation of protein structure and function after incorporation into the membrane
system. These results validate the use of nanodisc systems to study membrane proteins,
even though further optimisation of the system, as well as the illumination procedure in
the case of ChR II is still necessary.
Tethered bilayer lipid membrane systems have proved to be an excellent
platform for the study of membrane-active AMPs. Here, two tBLM systems were
described and successfully constructed onto a functionalised Au-film with a mixed SAM
composed of hydrophobic WK3SH tethered molecule and hydrophilic 6MH. Both tBLMs
were characterised by SEIRA and EIS. The optimum ratio of these two SAM molecules
(80:20, WK3SH:6MH) provided the appropriate platform for the formation of lipid bilayer
islands and a reproducible system.
The membrane interaction of the hydrophobic peptide EB and its complex formation
with monovalent ion were studied by SEIRA spectroscopy using POPC tBLM systems. The
results obtained here, represent a comprehensive study of EB by vibrational
spectroscopy, together with exciting preliminary results of EB:ion complexes by DFT
calculations. The FT-IR, SEIRA and DFT results indicated marked changes between the
EB:Na+ and EB:K+ species, which implies the formation of complexes of different
stoichiometry. These findings suggest a 1:1 and 2:1 ratio for the Na+ and K+ complexes,
respectively, which is in conflict with previously proposed ratios for Na+ ion [61]. Another
controversial result is the suggested 2:1 ratio for the EB:Cs+ complex, which neither
matches the predictions by Ovchinnikov nor the proposed 1:1 stoichiometry by Makrlìk
(DFT calculations) [119]. On the other hand, the low Na+/K+ selectivity and ion affinity
are in line with the gradient suggested in literature. Thus, the combined study presented
here can be considered as a starting point for a more detailed investigation of EB and
similar depsipeptides such as Beauvericin and valinomycin.
123
The POPC/POPG tBLM was employed to investigate the cationic AMP A1. Secondary
structure analysis by means of ATR-IR, FT-Raman, FT-IR, SEIRA spectroscopy and also,
preliminary results achieved by MD simulations proved the antiparallel β-sheet
conformation with a turn in solid and solution, as well as in contact with the membrane
system. This work aimed to offer more insight into the A1 behaviour, since the
conservation of its β-sheet conformation when interacting with membranes is on current
debate. The SEIRA spectroscopy results showed a perturbation of the negative charged
membrane by A1 at 4º C, 25º C, and 37º C. This perturbation was increased after
applying negative potentials. These changes were not reversible and appeared to be
aggravated at 37º C. The MD simulation results of the peptide with a POPC/POPG
membrane support the interpretation of the SEIRA measurements, providing additional
atomistic information of A1 behaviour in solution and in the membrane environment.
Though this is the initial stage, this work is of great relevance for a better understanding
of the mode of action of this unique AMP. Future investigations will focus on a
comprehensive analysis of this cationic AMP effects in contact with tBLM systems.
124
Outlook
This work showed the successful construction and application of membrane-mimetic
systems for the study of membrane-interacting biomolecules by a
spectroelectrochemical approach combining SEIRA and EIS spectroscopy. In case of
the nanodisc system, there are a growing number of studies that supports the
application and advances in this field applied to membrane proteins. The promising
results shown here demonstrate the versatility and great potential of tBLM systems for
the study of membrane-active AMPs. Improvements of conditions and a gradual
increase of the complexity towards a more realistic model of cell membranes are some
of the directions of future studies. Additional interest lies in further experiments
concerning EB and A1, and perhaps their respective modified peptides such as the
azido-modified EB for Stark effect analysis by measuring the shift of the strong azide
band ( 2160-2120 cm-1) to estimate complex stoichiometries; or A1 mutants to test the
influence of amino acids such as Arg, Tyr, Trp or Cys that play a key role in the peptide‘s
activity. In the case of A1, further investigation of salt gradient, temperature and
potential dependence of A1 activity, as well as time and peptide concentration are
some suggested experiments. In general, the extension of the study of AMPs with tBLM
systems combined with the spectroelectrochemical approach and theoretical
calculations to other potential targets is an attractive research field for the future.
125
126
Bibliography
[1]
Microbiology Society. (2017) MicrobiologyOnline. [Online].
http://www.microbiologyonline.org.uk
[2]
Manash Pratim Barkataki. (2015, September) Optogenetics. [Online].
http://optogenetics.weebly.com
[3]
Albert L Lehninger, David L Nelson, and Michael M Cox, Lehninger principles of
biochemistry. New York: W.H. Freeman, 2005.
[4]
Nermina Malanovic and Karl Lohner, "Gram-positive bacterial cell envelopes: The
impact on the activity of antimicrobial peptides," Biochimica et Biophysica Acta
(BBA) - Biomembranes, vol. 1858, no. 5, pp. 915-917, May 2016,
https://doi.org/10.1016/j.bbamem.2016.03.005.
[5]
S J Singer and Grath L Nicholson, "The Fluid Mosaic Model of the Structure of Cell
Membranes," Science, vol. 175, no. 4023, pp. 720-731, Frebraury 1972,
https://doi.org/10.1126/science.175.4023.720.
[6]
Jonathan D. Nickels et al., "The in vivo structure of biological membranes and
evidence for lipid domains," PLOS Biology, pp. 1-22, May 2017,
https://doi.org/10.1371/journal.pbio.2002214.
[7]
Paul Mueller, D O Rudin, H T Tien, and W C Wescott, "Reconstitution of cell
membrane structure in vitro and its transformation into an excitable system,"
Nature, pp. 979-980, 1962.
[8]
David C Lee and Denis Chapman, "Infrared Spectroscopic Studies of
Biomembranes and Model Membranes," Bioscience Reports, vol. 6, no. 3, pp. 235-
256, 1986, Review.
[9]
Lars J. C. Jeuken et al., "Phase separation in mixed self-assembled monolayers
and its effect on biomimetic membranes," Sensors and Actuators B Chemical, vol.
124, pp. 501-509, 2007.
[10]
Timothy H. Bayburt, Yelena V. Grinkova, and Stephen G. Sligar, "Self-Assembly of
Discoidal Phospholipid Bilayer Nanoparticles with Membrane Scaffold Proteins,"
NANO Letters, vol. 2, no. 8, pp. 853-856, 2002.
[11]
Jacek Kozuch et al., "Voltage-dependent structural changes of the membrane-
bound anion channel hVDAC1 probed by SEIRA and electrochemical
impedance spectroscopy," Physical Chemistry Chemical Physics, vol. 16, no. 20,
pp. 9546-9555, 2014.
127
[12]
Kenichi Ataka, Sven Timo Stripp, and Joachim Heberle, "Surface-enhanced
infrared absorption spectroscopy (SEIRAS) to probe monolayers of membrane
proteins," Biochimica et Biophysica Acta, vol. 1828, pp. 2283–2293, june 2013.
[13]
Sophie A Weiss, Richard J Bushby, Stephen D Evans, and Lars J. C. Jeuken, "A study
of cytochrome bo3 in a tethered bilayer lipid membrane," Biochimica et
Biophysica Acta (BBA) - Bioenergetics, vol. 1797, no. 12, pp. 1917-1923, December
2010, https://doi.org/10.1016/j.bbabio.2010.01.012.
[14]
Ekaterina Zaitseva, Marcia Saavedra, Sourabh Banerjee, Thomas P Sakmar, and
Reiner Vogel, "SEIRA spectroscopy on a membrane receptor monolayer using
lipoprotein particles as carriers.," Biophysical Journal, vol. 99, no. 7, pp. 2327-2335,
October 2010, doi: 10.1016/j.bpj.2010.06.054.
[15]
Vera Jansen et al., "Controlling fertilization and cAMP signaling in sperm by
optogenetics," e-Life, vol. 4:e05161, no.., pp. 1-15, January 2015, DOI:
10.7554/eLife.05161.
[16]
Curtis W. Meuse, Gediminas Niaura, Mary L. Lew, and Anne L. Plant, "Assessing the
Molecular Structure of Alkanethiol Monolayers in Hybrid Bilayer Membranes with
Vibrational Spectroscopies," Langmuir, vol. 14, no. 7, pp. 1604–1611, Febraury 1998,
DOI: 10.1021/la9700679.
[17]
Diego Millo et al., "Characterization of hybrid bilayer membranes on silver
electrodes as biocompatible SERS substrates to study membrane–protein
interactions," Colloids and Surfaces B: Biointerfaces, vol. 81, no. 1, pp. 212-216, July
2010, https://doi.org/10.1016/j.colsurfb.2010.07.010.
[18]
Joel H. Collier and Phillip B. Messersmith, "Phospholipid strategies in
biomineralization and biomaterials research," Annu. Rev. Mater. Res., vol. 31, pp.
237–63, 2001.
[19]
Lukas K. Tamm and Suren A. Tatulian, "Infrared spectroscopy of proteins and
peptides in lipid bilayers," Quaternary Reviews of Biophysics, vol. 30, no. 4, pp. 365-
429, 1997.
[20]
M. Cortijo, A. Alonso, J. C. Gómez-Fernández, and D. Chapman, "Intrinsic protein-
lipid interactions: Infrared spectroscopic studies of gramicidin A,
bacteriorhodopsin and Ca2+-ATPase in biomembranes and reconstituted
systems," Journal of Molecular Biology, vol. 157, no. 4, pp. 597-618, June 1982,
https://doi.org/10.1016/0022-2836(82)90501-0.
[21]
Jacek Kozuch, Structure-Function Relationships of Membrane Proteins -
Spectroelectrochemical Investigation of Artificial Membranes. Berlin, Germany:
Technischen Universität Berlin, 2013.
128
[22]
Ilia G. Denisov and Stephen G. Sligar, "Nanodiscs in Membrane Biochemistry and
Biophysics," Chemical Reviews, vol. 117, p. 4669−4713, Febraury 2017, DOI:
10.1021/acs.chemrev.6b00690.
[23]
Andrew J. Leitz, Timothy H. Bayburt, Alexander N. Barnakov, Barry A. Springer, and
Stephen G. Sligar, "Functional reconstitution of β2-adrenergic receptors utilizing
self-assembling Nanodisc technology," BioTechniques, vol. 40, no. 5, pp. 601-612,
May 2006, DOI: 10.2144/000112169.
[24]
Zeting Zhang et al., "Ca2 + modulating α-synuclein membrane transient
interactions revealed by solution NMR spectroscopy," Biochimica et Biophysica
Acta (BBA) - Biomembranes, vol. 1838, no. 3, pp. 853–858, March 2014,
https://doi.org/10.1016/j.bbamem.2013.11.016.
[25]
Chanjuan Wan et al., "Insights into the molecular recognition of the granuphilin
C2A domain with PI(4,5)P2," Chemistry and Physics of Lipids, vol. 186, pp. 61-67,
February 2015, https://doi.org/10.1016/j.chemphyslip.2015.01.003.
[26]
Timothy H. Bayburt and Stephen G. Sligar, "Membrane Protein Assembly into
Nanodiscs," FEBS Letters, vol. 584, no. 9, pp. 1721-1727, May 2010.
[27]
Nicolas Bocquet et al., "Real-time monitoring of binding events on a
thermostabilized human A2A receptor embedded in a lipid bilayer by surface
plasmon resonance," Biochimica et Biophysica Acta (BBA) - Biomembranes, vol.
1848, no. 5, pp. 1224-1233, May 2015,
https://doi.org/10.1016/j.bbamem.2015.02.014.
[28]
A. Das, J. Zhao, G. C. Schatz, S. G. Sligar, and R. P. Van Duyne, "Screening of type I
and II drug binding to human cytochrome P450-3A4 in nanodiscs by localized
surface plasmon resonance spectroscopy.," Analytical chemistry, vol. 81, no. 10,
pp. 3754–3759, 2009, DOI: 10.1021/ac802612z.
[29]
Swantje Wiebalck et al., "Monitoring the Transmembrane Proton Gradient
Generated by Cytochrome bo 3 in Tethered Bilayer Lipid Membranes Using SEIRA
Spectroscopy," The Journal of Physical Chemistry B, vol. 120, no. 9, pp. 2249-2256,
2016.
[30]
Jacqueline Knobloch, Daniel K. Suhendro, Julius L. Zieleniecki, Josepph G. Shapter,
and Ingo Köper, "Membrane–drug interactions studied using model membrane
systems," Saudi Journal of Biological Sciences, vol. 22, no. 6, pp. 714-718,
November 2015, https://doi.org/10.1016/j.sjbs.2015.03.007.
129
[31]
Ali Adem Bahar and Dacheng Ren, "Antimicrobial Peptides," Pharmaceuticals, vol.
6, pp. 1543-1575, 2013, doi:10.3390/ph6121543.
[32]
Mariana Ruíz Villareal. (2007, -) Membrane transport protein. [Online].
https://en.wikipedia.org/wiki/membrane_transport_protein
[33]
Víctor A. Lórenz-Fonfría et al., "Transient protonation changes in
channelrhodopsin-2 and their relevance to channel gating," PNAS, vol. 110, no.
14, pp. E1273–E1281, March 2013, DOI: 10.1073/pnas.1219502110.
[34]
Eglof Ritter, Patrick Piwowarski, Peter Hegemann, and Franz J. Bartl,
"Channelrhodopsin C128T Mutant Light-dark Adaptation of," The Journal of
Biological Biophysics, vol. 288, no. 15, pp. 10451–10458, 2013, DOI:
10.1074/jbc.M112.446427.
[35]
Zoe Cournia et al., "Membrane Protein Structure, Function and Dynamics: A
Perspective from Experiments and Theory," The Journal of Membrane Biology, vol.
248, no. 4, pp. 611–640, 2015, doi:10.1007/s00232-015-9802-0.
[36]
Samuel Krimm and Jagdeesh Bandekar, "VIBRATIONAL SPECTROSCOPY AND
CONFORMATION OF PEPTIDES. POLYPEPTIDES. AND PROTEINS," ADVANCES IN
PROTEIN CHEMISTRY, vol. 38, pp. 181-364, 1986.
[37]
Friedrich Siebert and Peter Hildebrandt, Vibrational Spectroscopy in Life Science.:
Wiley-VCH Verlag GmbH & Co. KGaA, 2008.
[38]
S. Yu. Venyaminov and N. N. Kalnin, "Quantitative IR spectrophotometry of
peptide compounds in water (H2O) solution. I. Spectral parameters of amino acid
residue absorption bands," Biopolymers, vol. 30, pp. 1243-1257, 1990.
[39]
Andreas Barth, "Infrared spectroscopy of proteins," Biochimica et Biophysica Acta,
vol. 1767, pp. 1073–1101, 2007, doi:10.1016/j.bbabio.2007.06.004.
[40]
Víctor A. Lórenz-Fonfría and Joachim Heberle, "Channelrhodopsin unchained:
Structure and mechanism of a light-gated cation channel," Biochimica et
Biophysica Acta (BBA) - Bioenergetics, vol. 1837, no. 5, pp. 626-642, 2014,
https://doi.org/10.1016/j.bbabio.2013.10.014.
[41]
Christian Bamann, Ronnie Gueta, Sonja Kleinlogel, Georg Nagel, and Enrst
Bamberg, "Structural guidance of the photocycle of channelrhodopsin-2 by an
interhelical hydrogen bond," biochemistry, vol. 49, no. 2, pp. 267-278, 2010.
[42]
Sara Bruun et al., "Light−Dark Adaptation of Channelrhodopsin Involves
Photoconversion between the all-trans and 13-cis Retinal Isomers," Biochemistry,
vol. 54, pp. 5389-5400, 2015, DOI: 10.1021/acs.biochem.5b00597.
130
[43]
Georg Nagel et al., "Channelrhodopsin-2, a directly light-gated cation-selective
membrane channel," PNAS, vol. 100, no. 24, pp. 13940–13945, 2003.
[44]
Sara Bruun et al., "The chromophore structure of the long-lived intermediate of the
C128T channelrhodopsin-2 variant," FEBS Letters, vol. 585, pp. 3998–4001, 2011,
doi:10.1016/j.febslet.2011.11.007.
[45]
Peter Hegemann and Georg Nagel, "From channelrhodopsins to optogenetics,"
EMBO Molecular Medicine, vol. 5, no. 2, pp. 173–176, Febraury 2013, DOI:
10.1002/emmm.201202387.
[46]
Andre Berndt et al., "Structural foundations of optogenetics: Determinants of
channelrhodopsin ion selectivity," PNAS, vol. 113, no. 4, pp. 822–829, November
2016, doi: 10.1073/pnas.1523341113.
[47]
Katja Stehfest, Eglof Ritter, André Berndt, Franz Bartl, and Peter Hegemann, "The
Branched Photocycle of the Slow-Cycling Channelrhodopsin-2 Mutant C128T," The
Journal of Molecular Biology, vol. 398, pp. 690-702, 2010,
doi:10.1016/j.jmb.2010.03.031.
[48]
André Berndt, Ofer Yizhar, Lisa A. Gunaydin, Peter Hegemann, and Karl Deisseroth,
"Bi-stable neural state switches," nature neuroscience, vol. 12, no. 2, pp. 229-234,
2009, doi:10.1038/nn.2247.
[49]
J. M. Ageitos, A. Sánchez-Pérez, P. Calo-Mata, and T. G. Villa, "Antimicrobial
peptides (AMPs): Ancient compounds that represent novel weapons in the fight
against bacteria," Biochemical Pharmacology, 2016,
http://dx.doi.org/10.1016/j.bcp.2016.09.018.
[50]
Rollin D. Hotchkiss and René J. Dubos, "FRACTIONATION OF THE BACTERICIDAL
AGENT FROM CULTURES OF A SOIL BACILLUS," The Journal of Biological Chemistry,
vol. 132, pp. 791-792, 1940.
[51]
James G. Hirsch, "PHAGOCYTIN: A BACTERICIDAL SUBSTANCE FROM
POLYMORPHONUCLEAR LEUCOCYTES," The Journal of Experimental Medicine, vol.
103, no. 5, pp. 589-611, 1956, DOI: 10.1084/jem.103.5.589.
[52]
René J. Dubos, "Studies on a bactericidal agent extracted from a soil bacillus," The
Journal of Experimental Medicine, vol. 70, no. 1, pp. 1-10, 1939.
[53]
M. Stauss-Grabo, S. Atiye, T. Le, and M. Kretschmar, "Decade-long use of the
antimicrobial peptide combination tyrothricin does not pose a major risk of
acquired resistance with gram-positive bacteria and Candida spp.," Pharmazie,
vol. 69, no. 11, pp. 838-841, November 2014,
http://dx.doi.org/10.1691/ph.2014.4686.
131
[54]
David A. Phoenix, Sarah R. Dennison, and Frederick Harris, Antimicrobial Peptides,
1st ed.: Wiley-VCH Verlag GmbH & Co. KGaA, 2013.
[55]
J. Bradshaw, "Cationic antimicrobial peptides : issues for potential clinical use.,"
BioDrugs Journal, vol. 17, no. 4, pp. 233-240, 2013.
[56]
Kelly L. Brown and Robert E. W. Hancock, "Cationic host defense (antimicrobial)
peptides," Current Opinion in Immunology, vol. 18, no. 1, pp. 24-30, 2006,
https://doi.org/10.1016/j.coi.2005.11.004.
[57]
Xiaowei Zhao, Hongyu Wu, Hairong Lu, Guodong Li, and Qingshan Huang, "LAMP:
A Database Linking Antimicrobial Peptides," PLOS one, vol. 8, no. 6, pp. e66557-
e66557, 2013, https://doi.org/10.1371/journal.pone.0066557.
[58]
Guozhang Zou et al., "Toward understanding the cationicity of defensins. Arg and
Lys versus their noncoded analogs," The Journal of Biological Chemistry, vol. 282,
no. 27, pp. 19653–19665, 2007, DOI: 10.1074/jbc.M611003200.
[59]
Annett Rozek, Carol L. Friedrich, and Robert E. W. Hancock, "Structure of the
Bovine Antimicrobial Peptide Indolicidin Bound to Dodecylphosphocholine and
Sodium Dodecyl Sulfate Micelles," Biochemistry, vol. 39, no. 51, pp. 15765–15774,
2000, DOI: 10.1021/bi000714m.
[60]
Berton C. Pressman, "Biological applications of ionophores," Annual Review of
Biochemistry, vol. 45, pp. 501-530, 1976,
https://doi.org/10.1146/annurev.bi.45.070176.002441.
[61]
Yu A. Ovchinnikov et al., "The enniatin ionophores. Conformation and ion binding
properties," Int. J. Peptide Protein Res. , vol. 6, pp. 465-498, 1974.
[62]
Leonard T. Nguyen, Evan F. Haney, and Hans J. Vogel, "The expanding scope of
antimicrobial peptide structures and their modes of action," Trends in
Biotechnology, vol. 29, no. 9, pp. 464-472, 2011,
https://doi.org/10.1016/j.tibtech.2011.05.001.
[63]
Jörg Andrä et al., "Structure and mode of action of the antimicrobial peptide
arenicin," Biochemistry Journal, vol. 410, pp. 113-122, 2008,
doi:10.1042/BJ20071051.
[64]
Tatiana V. Ovchinnikova et al., "Purification and primary structure of two isoforms
of arenicin, a novel antimicrobial peptide from marine polychaeta Arenicola
marina," FEBS Letters, vol. 577, pp. 209-214, 2004, doi:10.1016/j.febslet.2004.10.012.
[65]
Brandon Findlay, George G. Zanel, and Frank Schweizer, "Cationic Amphiphiles, a
New Generation of Antimicrobials Inspired by the Natural Antimicrobial Peptide
Scaffold," Antimicrobial Agents and Chemotherapy, vol. 45, no. 10, pp. 4049-4058,
2010, doi: 10.1128/AAC.00530-10.
132
[66]
Lennart Richter et al., "Engineering of Aspergillus niger for the production of
secondary metabolites," Fungal Biology and Biotechnology, vol. 1, no. 4, pp. 1-13,
2014.
[67]
E. Gaumann and O. Jaag, "Die physiologischen Grundlagen des parasitogenen
Welkens," Ber. Schweiz. Bot. Ges., vol. 57, pp. 3-34, 1947.
[68]
Ralene R. Mitschler, Ruth Welti, and Steve J. Opton, "A Comparative Study of Lipid
Compositions of Cryptosporidium parvum (Apicomplexa) and Madin-Darby
Bovine Kidney Cells," The Journal of Eukaryotic Microbiology, vol. 41, no. 1, pp. 8-
12, 1994, DOI: 10.1111/j.1550-7408.1994.tb05927.x.
[69]
Michèle German-Fattal, "Fusafungine, an Antimicrobial with Anti-Inflammatory
Properties in Respiratory Tract Infections," Clinical Drug Investigation, vol. 21, no. 9,
pp. 653–670, 2001.
[70]
Arlene A. Sy-Cordero, Cedric J. Pearce, and Nicholas H. Oberlies, "Revisiting the
enniatins: a review of their isolation, biosynthesis, structure determination, and
biological activities," J Antibiot, vol. 65, no. 11, pp. 541-549, 2012,
doi:10.1038/ja.2012.71.
[71]
G. Meca et al., "Antifungal effects of the bioactive compounds enniatins A, A1, B,
B1," Toxicon, vol. 56, no. 3, pp. 480-485, 2010,
https://doi.org/10.1016/j.toxicon.2010.04.013.
[72]
Rita Dornetshuber et al., "Enniatin Exerts p53-Dependent Cytostatic and p53-
Independent Cytotoxic Activities against Human Cancer Cells," Chem. Res.
Toxicol., vol. 20, pp. 465-473, 2007.
[73]
Wim Wätjen et al., "Enniatins A1, B and B1 from an endophytic strain of Fusarium
tricinctum induce apoptotic cell death in H4IIE hepatoma cells accompanied by
inhibition of ERK phosphorylation," Molecular nutrition & Food research, vol. 53, no.
4, pp. 431-440, 2009, DOI: 10.1002/mnfr.200700428.
[74]
K. A. Mereish, R. Solow, D. L. Bunner, and A. B. Fajer, "Interaction of cyclic peptides
and depsipeptides with calmodulin," Pept. Res. , vol. 3, no. 5, p. 233.237, 1990.
[75]
Claudia Behm, Gisela H. Degen, and Wolfram Föllmann, "The Fusarium toxin
enniatin B exerts no genotoxic activity, but pronounced cytotoxicity in vitro,"
Molecular Nutrition & Food research, vol. 53, no. 4, pp. 423-430, 2009, DOI:
10.1002/mnfr.200800183.
[76]
Majidreza Kamyar, Pakiza Rawnduzi, Christian R. Studenik, Katerina Kouri, and
Rosa Lemmens-Gruber, "Investigation of the electrophysiological properties of
enniatins," Archives of Biochemistry and Biophysics, vol. 429, no. 2, pp. 215-223,
2004, https://doi.org/10.1016/j.abb.2004.06.013.
133
[77]
N. E. Zhukhlistova, "X-ray Crystal Structure of the Complex of Enniatin B with KNCS,"
Crystallography Reports, vol. 47, no. 3, pp. 478-487, 2002.
[78]
Jaeyong Cho and Dong Gun Lee, "The characteristic region of arenicin-1 involved
with a bacterial membrane targeting mechanism," Biochemical and Biophysical
Research Communications, vol. 405, pp. 422-427, 2011.
[79]
Oksana G. Tarvkova and Gerald Brezesinski, "Adsorption of the antimicrobial
peptide arenicin and its linear derivative to model membranes – A maximum
insertion pressure study," Chemistry and Physics of Lipids, vol. 167-168, pp. 43-50,
2013.
[80]
Cana Park and Dong Gun Lee, "Fungicidal effect of antimicrobial peptide
arenicin-1," Biochimica et Biophysica Acta, vol. 1788, pp. 1790-1796, 2009.
[81]
Jaeyong Cho and Dong Gun Lee, "The antimicrobial peptide arenicin-1 promotes
generation of reactive oxygen species and induction of apoptosis," Biochimica et
Biophysica Acta, vol. 1810, pp. 1246-1251, 2011.
[82]
Radek Macháň, Martin Hof, Tatsiana Chernovets, Maxim N. Zhmak, and Tatiana V.
Ovchinnikova, "Formation of arenicin-1 microdomains in bilayers and their specific
lipid interaction revealed by Z-scan FCS," Anal. Bioanal. chem. , vol. 399, pp. 3547-
3554, 2011.
[83]
Kunichi Ataka and Joachim Heberle, "Electrochemically Induced Surface-
Enhanced Infrared Difference Absorption (SEIDA) Spectroscopy of a Protein
Monolayer," Journal of the American Chemical Society, vol. 125, no. 13, pp. 4986-
4987, 2003, DOI: 10.1021/ja0346532.
[84]
Erik Goormaghtigh, Vincent Raussens, and Jean-Marie Ruysschaert, "Attenuated
total reflection infrared spectroscopy of proteins and lipids in biological
membranes," Biochimica et Biophysica Acta (BBA) - Reviews on Biomembranes,
vol. 1422, no. 2, pp. 105-185, 1999, https://doi.org/10.1016/S0304-4157(99)00004-0.
[85]
A. Hartstein, J. R. Kirtley, and J. C. Tsang, "Enhancement of the Infrared Absorption
from Molecular Monolayers with Thin Metal Overlayers," Physical Reviews Letters,
vol. 45, no. 3, pp. 201-204, 1980.
[86]
Masatoshi Osawa, "Surface-enhanced infrared absorption spectroscopy," pp. 1-
15, 2002.
[87]
Kenichi Ataka et al., "Oriented Attachment and Membrane Reconstitution of His-
Tagged Cytochrome c Oxidase to a Gold Electrode:  In Situ Monitoring by
Surface-Enhanced Infrared Absorption Spectroscopy," J. Am. Chem. Soc., vol. 126,
no. 49, pp. 16199–16206, July 2004, DOI: 10.1021/ja045951h.
134
[88]
Nattawadee Wisitruangsakul et al., "Redox-linked protein dynamics of
cytochrome c probed by time-resolved surface enhanced infrared absorption
spectroscopy," Phys. chem. chem. phys, vol. 10, no. 34, pp. 5276-5286, 2008.
[89]
Diego Millo, Peter Hildebrandt, Maria-Eirini Pandelia, Wolfgang Lubitz, and Ingo
Zebger, "SEIRA Spectroscopy of the Electrochemical Activation of an Immobilized
[NiFe] Hydrogenase under Turnover and Non-Turnover Conditions," Angewandte
Chemie-International Edition, vol. 50, no. 11, pp. 2632–2634, 2011.
[90]
Masatoshi Osawa, Kenichi Ataka, Katsumasa Yoshii, and Yuji Nishikawa, "Surface-
Enhanced Infrared Spectroscopy: The Origin of the Absorption Enhancement and
Band Selection Rule in the Infrared Spectra of Molecules Adsorbed on Fine Metal
Particles," Applied Spectroscopy, vol. 47, no. 9, pp. 1497-1502, 1993.
[91]
F. Lisdat and D. Schäfer, "The use of electrochemical impedance spectroscopy for
biosensing," Anal Bioanal Chem., vol. 391, no. 5, pp. 1555-1567, 2008.
[92]
James K. R. Kendall et al., "Effect of the Structure of Cholesterol-Based Tethered
Bilayer Lipid Membranes on Ionophore Activity," ChemPhysChem, vol. 11, no. 10,
pp. 2191–2198, May 2010, DOI: 10.1002/cphc.200900917.
[93]
V. F. Lvovich, Impedance Spectroscopy: Applications to Electrochemical and
Dielectric Phenomena.: John Wiley & Sons, Inc., 2012.
[94]
Evgenij Barsoukov and J. Ross Macdonald, Impedance Spectroscopy: Theory,
Experiment, and Applications, 2nd ed., 2005.
[95]
Gintaras Valincius, Tadas Meškauskas, and Feliksas Ivanauskas, "Electrochemical
Impedance Spectroscopy of Tethered Bilayer Membranes," Langmuir, vol. 28, no.
1, pp. 977–990, 2012.
[96]
P. M. Gomadam and J. W. Weidner, "Analysis of electrochemical impedance
spectroscopy in proton exchange membrane fuel cells," International Journal of
Energy research, vol. 29, no. 12, pp. 1133-1151, 2005.
[97]
B. A. Boukamp, "A nonlinear least squares fit for analysis of immittance data of
electrochemical systems," Solid-state ionics, vol. 20, pp. 31-40, 1986.
[98]
S. Ohki, "Properties of Lipid Bilayer Membranes. Membrane Thickness," Journal of
Theoretical Biology, vol. 26, pp. 277-287, 1970.
[99]
H. G. Coster and J. R. Smith, "The molecular organisation of bimolecular lipid
membranes. A study of the low frequency Maxwell-Wagner impedance
dispersion," Biochim Biophys Acta., vol. 373, no. 2, pp. 151-64, 1974.
135
[100]
V. D. Jovic and B. M. Jovic, "EIS and differential capacitance measurements onto
single crystal faces in different solutions: Part I: Ag(111) in 0.01 M NaCl," Journal of
Electroanalytical Chemistry, vol. 541, pp. 1-11, 2003.
[101]
Hiroto Miyake, Shen Ye, and Masatoshi Osawa, "Electroless deposition of gold thin
films on silicon for surface-enhanced infrared spectroelectrochemistry,"
Electrochemistry Communications, vol. 4, no. 12, pp. 973-977, 2002,
https://doi.org/10.1016/S1388-2481(02)00510-6.
[102]
S. Trasatti and O. A. Petrii, "Real surface area measurements in electrochemistry,"
Pure Appl. Chem., vol. 63, no. 5, pp. 711-734, 1991,
http://dx.doi.org/10.1351/pac199163050711.
[103]
Francisco Javier Velázquez Escobar, Vibrational spectroscopy of phytochromes
and phytochrome-related photoreceptors. Berlin, Germany: Technischen
Universität Berlin, 2015.
[104]
Amit Vaish et al., "A generalized strategy for immobilizing uniformly oriented
membrane proteins at solid interfaces ," Chemical Communications, vol. 49, no.
26, pp. 2685-2687, 2013.
[105]
Marcel G. Friedrich et al., "Activity of Membrane Proteins Immobilized on Surfaces
as a Function of Packing Density," The Journal of physchal chemistry B, vol. 112,
no. 10, pp. 3193–3201, 2008.
[106]
Emma M. Ericsson et al., "Site-Specific and Covalent Attachment of His-Tagged
Proteins by Chelation Assisted Photoimmobilization: A Strategy for Microarraying of
Protein Ligands," Langmuir, vol. 29, no. 37, pp. 11687–11694, 2013.
[107]
Jonas Schartner et al., "Universal Method for Protein Immobilization on Chemically
Functionalized Germanium Investigated by ATR-FTIR Difference Spectroscopy,"
Journal of the American Chemical Society, vol. 135, p. 4079−4087, 2013.
[108]
Paul Mueller and Donald O. Rudin, "Development of K+ - Na+ discrimination in
experimental bimolecular lipid membrane by macroyclic antibiotic," Biochemical
and Biophysical research communications, vol. 26, no. 4, pp. 398-404, 1967.
[109]
M. M. Shemyakin et al., "Cyclodepsipeptides as Chemical Tools for Studying Ionic
Transport Through Membranes," The Journal of Memrane Biology, vol. 1, pp. 402-
430, 1969.
[110]
Jan Palacky, Peter Mojzes, and Jirí Bork, "SVD-based method for intensity
normalization, background correction and solvent substraction in Raman
spectroscopy exploiting the properties of water stretching vibrations," The Journal
of Raman Spectroscopy, vol. 42, no. 7, pp. 1528-1539, 2011.
136
[111]
Jan Kubelka and Timothy A. Keiderling, "Differentiation of β-Sheet-Forming
Structures: Ab Initio-Based Simulations of IR Absorption and Vibrational CD for
Model Peptide and Protein beta-Sheets," Journal of the American Chemical
Society , vol. 123, pp. 12048-12058, 2001.
[112]
Yu. N. Chirgadze, 0. V. Fedorov, And N. P. Trushina, "Estimation of Amino Acid
Residue Side-Chain Absorption in the Infrared Spectra of Protein Solutions in Heavy
Water," Biopolymers, vol. 14, pp. 679-694, 1975.
[113]
Nurettin Demirdöven et al., "Two-Dimensional Infrared Spectroscopy of Antiparallel
beta-Sheet Secondary Structure," Journal of the American Chemical Society, vol.
126, pp. 7981-7990, 2004.
[114]
Yong Xie, Dongmao Zhang, Gotam K. Jarori, V. Jo Davisson, and Dor Ben-Amotz,
"The Raman detection of peptide tyrosine phosphorylation," Analytical
Biochemistry, vol. 332, pp. 116–121, 2004.
[115]
A. E. Aliaga et al., "Surface-enhanced Raman scattering study of L-tryptophan,"
Journal of Raman Spectroscopy, vol. 40, pp. 164–169, 2009.
[116]
Guangyong Zhu, Xian Zhu, Qi Fan, and Xueliang Wan, "Raman spectra of amino
acids and their aqueous solutions," Spectrochimica Acta Part A: Molecular and
Biomolecular Spectroscopy, vol. 78, pp. 1187–1195, 2011.
[117]
Evgeniy S. Salnikov et al., "Structure and Alignment of the Membrane-Associated
Antimicrobial Peptide Arenicin by Oriented Solid-State NMR Spectroscopy,"
Biochemistry, vol. 50, pp. 3784-3795, 2011.
[118]
Zakhar O. Shenkarev et al., "Molecular Mechanism of Action of β-Hairpin
Antimicrobial Peptide Arenicin: Oligomeric Structure in Dodecylphosphocholine
Micelles and Pore Formation in Planar Lipid Bilayers," Biochemistry, vol. 50, no. 28,
pp. 6255–6265, 2011.
[119]
Emanuel Makrlík, Stanislav Böhm, Petr Vanura, and Ivan Raich, "Extraction and DFT
study on interaction of the cesium cation with enniatin B," Journal of Molecular
Structure, vol. 1076, pp. 564–567, 2014.
137
138
Acknowledgements
Once a wiseman told me that all the experiences from my past will prepare me to live
in the present. Throughout all these experiences, what has encouraged me to grow
and keep walking were the people I encountered along the way. Therefore, I would
like to thank to each and all of you for forming part of my life.
I would like to express my profound gratitude to Prof. Peter Hildebrandt, for the great
opportunity you have given me. I am forever grateful for your support, encouragement
and guidance during these years, not only in professional but also in personal matters.
I would like to thank to the School of Analytical Science Adlershof (SALSA) for the
founding of my studies, as well as for the international environment, training and
get-together evenings. It was a valued experience to form part of such a great school.
Special thanks to Dr. Kozuch, for introducing me to the spectroscopic world and for
your contagious enthusiasm for science; and to Dr. Velázquez Escobar for the push
when it was needed, the self-value (common sense) reminders, as well as for the
scientific support and critical point of view.
Big thanks to my personal cheerleaders Alejandra De Miguel and Enrico Forbrig for your
unconditional support, the uncountable moments of laughter and unforgettable
moments during these years. Many stories shared and many more to come. The
conferences will miss us!
I thank to María Fernández, Dr. Kielb and Johannes Salewski for the never-ending talks,
‗coffee‘-breaks, and the laughter together.
I would also like to show my appreciation to all the members of the Physical Chemistry /
Biophysical Chemistry and Biomolecular Modelling teams that enriched the day-by-day
work with smiles and good vibes.
Gracias a mi mejor amiga Luky, por tu Amistad incondicional y por recordarme que la
vida es mucho más que trabajo y responsabilidades. Mil gracias por estar ahí siempre
sin importar las distancias. Tus abrazos y sonrisa están siempre conmigo.
Moltes gracies a tu Arnau, per mantenir contacte després d‘aquests anys, per
inspirarme i demostrar que mai es tard per fer el que un vol. Ara et toca a tu!
I also want to thank to Anja, Alma, Nikolaj and Frederik for the universal revelations and
the never boring life together no matter where.
My deepest and special thanks to Rasmus for showing me possibilities in this crazy world
we live in. Thank you for the good food we share, for holding my hand when I felt down
and showing me light when I could not see. More importantly, thanks for the loving,
peaceful, kind and wonderful person you are. I am happy to walk this life by your side.
139
Gracias a mi Bernusqui, estos años no hubiesen sido lo mismo sin vos. Gracias por los
largos paseos, por elegirme a mí para compartir tu vida y por enseñarme sobre el
perdón. Estarás por siempre en mi corazón.
Y un gracias monumental al tesoro más valioso que el universo me ha concedido, mi
familia. El amor, esfuerzo y sacrificios que mis padres han hecho para darles a sus hijos
un futuro mejor, no tiene medida. Los valores con que nos criaron nos han hecho las
increíbles personas que somos hoy. Gracias mamá y papá. Por supuesto, mil gracias a
mi ‗sis‘ y a mi ‗bro‘ por la hermosa relación que tenemos y por las batallas que
combatimos juntos, estoy muy orgullosa de nosotros. Gracias también a mis abuelos y
familiares. Se extrañan los domingos en familia, pero se que vuestro amor está siempre
conmigo. Y un gracias especial a vos abu, se que donde sea que estés ahora cuidas
de mí. Los amo familia.
140
141
Appendix
Figure A.1. Images of the top and side view of the EB structures obtained from the DFT calculation of
the 1:1 complex with Cs+ ion, also called ―open-faced-sandwich‖. Oxygen atoms are in red, nitrogen
in blue, carbon in grey, hydrogen in white, and metal ion in purple.
Figure A.3. Images of the top and side view of the EB structures obtained from the DFT calculation of
the 2:1 complex with Na+ ion. Oxygen atoms are in red, nitrogen in blue, carbon in grey, hydrogen in
white, and metal ion in purple.
142
Figure A.4. Images of the top and side view of the EB structures obtained from the DFT calculation of
the 3:2-AAEA complex with Cs+ ion. Oxygen atoms are in red, nitrogen in blue, carbon in grey,
hydrogen in white, and metal ion in purple.
Figure A.5. Images of the top and side view of the EB structures obtained from the DFT calculation of
the 3:2-AAEE complex with Cs+ ion. Oxygen atoms are in red, nitrogen in blue, carbon in grey,
hydrogen in white, and metal ion in purple.
143
Figure A.6. Images of the top and side view of the EB structures obtained from the DFT calculation of
the 3:2-AEAE complex with Cs+ ion. Oxygen atoms are in red, nitrogen in blue, carbon in grey,
hydrogen in white, and metal ion in purple.
Figure A.7. SEIRA difference spectra of the POPC/POPG tBLM incubation at 4° C (blue), 25° C (black) and
37° C (red). The blue and red spectra depict the mirror temperature effect described in section 4.3.3.2.
144
Figure A.9. Monitoring secondary structure plot of chain A (yellow A1 peptide in figure 4.44,
section4.3.4) of two A1 monomers interacting with the POPC/POPG membrane. Yellow colour
denotes the antiparallel β-sheet conformation, while the β-turn is depicted in green and the coil in
white. On the left side of the graph there is the amino acid sequence of A1 from the N- to the
C-terminal, from top to bottom respectively.
145
Figure A.10. Monitoring secondary structure plot of chain U (yellow A1 peptide in figure 4.44,
section4.3.4) of two A1 monomers interacting with the POPC/POPG membrane. Yellow colour
denotes the antiparallel β-sheet conformation, while the β-turn is depicted in green and the coil in
white. On the left side of the graph there is the amino acid sequence of A1 from the N- to the
C-terminal, from top to bottom respectively.
146
147
148
Statement of Authorship
I hereby declare that I am the sole author of this doctoral thesis and I have not used
any sources other than those listed in the bibliography and identified as references. I
further declare that I have not submitted this thesis to any other institution in order to
obtain a degree.
Barbara Daiana Gonzalez
Berlin, 31.07.2017
149