Characterization of the Membrane-Bound Topology of the Colicin E1
Channel Domain
by
Derek Ho
A Thesis
Presented to
The University of Guelph
In partial fulfilment of requirements
for the degree of
Doctor of Science
in
Molecular & Cellular Biology
Guelph, Ontario, Canada
© Derek Ho, July, 2014
ABSTRACT
CHARACTERIZATION OF THE MEMBRANE-BOUND TOPOLOGY OF THE
COLICIN E1 CHANNEL DOMAIN
Derek Ho
University of Guelph, 2014
Advisor:
Professor: A. R. Merrill
Colicins are antimicrobial proteins produced by Escherichia coli that target
susceptible bacteria in response to stressful conditions including nutrient depletion, DNA
damage, over-crowding, and anaerobiosis. The C-terminal channel-forming domain of
colicin E1 forms a lethal ion channel which depolarizes the cytoplasmic membrane of
target bacterial cells. Prior to channel formation, the channel peptide first binds to the
lipid bilayer, followed by protein unfolding and helix elongation. Finally, the channel
domain adopts an insertion-competent state in which it inserts into the membrane to form
the pre-channel state. The channel then opens in response to a trans-negative membrane
potential and facilitates the escape of various ions from the host cells, such as Na+, K+,
and H+, leading to host cell death.
The objective of my thesis work was to characterize the membrane topology of
the colicin E1 channel peptide using state-of-the-art fluoresence methods. Previously,
researchers in our laboratory used cysteine-scanning mutagenesis to determine the
surface topology and secondary structure of Helices 1 - 5 within the colicin E1 channel
peptide in the pre-channel state. In this thesis work, I demonstrated that helices 6, 7 and
10 are three, distinct amphipathic α-helices which adopt a tilted topology on the
membrane surface that correlates well with various methods such as harmonic analysis,
2
fluorescence anisotropy, fluorescence emission maximum, and bilayer penetration depth.
In addition, a total of twelve distances (Å) between three residues on Helix 1 (D347,
S354, E361) and three Trp residues (W424, W460, W495) were estimated using
fluorescence resonance energy transfer (FRET) to probe the 3-D orientation of Helix 1
relative to the rest of the channel domain in the membrane-embedded state. Furthermore,
I employed the method of Schultz and co-workers to genetically encode a fluorescent
amino
acid
(coumarin-derivative)
as
a
FRET
donor
to
DABMI
(4-
(dimethylamino)phenylazophenyl-4-maleimide) as the acceptor. This work provided six
more interhelical distances between Helices 1 – 6 of the channel peptide in the prechannel state. Overall, this work has provided a number of constraints towards the
development of an improved colicin E1 model (umbrella model) that accounts for all of
the FRET data for the pre-channel state.
3
iv
ACKNOWLEDGEMENT
I would like to take this opportunity to express my appreciation to those who
helped me throughout my study. First of all, I want to thank my supervisor, Prof. Rod
Merrill, for providing me with the opportunity to complete my Doctoral degree. It was
my pleasure to be one of his students working in the lab as he provided me with
tremendous amount of valuable advice, skill and experience that will be important for my
future career. Secondly, I also want to thank my advisory committee members, Prof.
George Harauz, Prof. Steffan Graether and Prof. Leonid S. Brown, for their wise
guidance and assistance throughout the completion of my research.
Without their
valuable advice and suggestion, my research project would not have been nearly as
successful.
I also want to take this opportunity to thank all the Merrill lab members, both
current and past members, for all of their help and support throughout my studies. I want
to thank Zhikui Wei, former graduate student in the Merrill’s lab, for his training and
assistance during the early stages of the research project. I also want to thank Gerry
Prentice and Dawn White for their constant help and technical assistance throughout my
studies. Finally, I would like to give special thanks to my parents, Mr. Wilson Ho and
Mrs. Betty Li, for their wonderful love and support that enable me to complete this work.
4
v
TABLE OF CONTENTS
ACKNOWLEDGEMENT………………………………………....-ivTABLE OF CONTENTS……………………………………….….-vLIST OF TABLES………………………………………………..-viiiLIST OF FIGURES………………………………………………..-ixGLOSSARY OF ABBREVIATIONS……………………………..-xiCHAPTER 1
1.1
1.1.1
1.1.2
1.2
1.2.1
1.2.2
1.2.3
1.2.4
1.2.5
1.3
1.3.1
1.4
Introduction……………………………………...-1-
Colicin Biology……………………………………………………………….…-1Colicins…………………………………………………………………………..-1Colicins Structure and Import Mechanism………………………………………-6Colicin E1…………………………………………………………………….…-8Colicin E1 Structure……………………………………………………………..-9Colicin E1 Membrane Binding………………………………………………...-11The Ion-Channel Formation by Colicin E1…………………………………….-14Closed Channel Structure of Colicin E1…………….…………………………-17Colicins Applications…………………………………………………………..-19Fluorescence Spectroscopy………………………...…………..……………..-20Fluorescence resonance energy transfer (FRET)……………………………..,.-20Research Objectives…………………………………………………………..-30-
CHAPTER 2
Membrane Topology of the Colicin E1 Channel
Using Genetically Encoded Fluorescence…………………….…..-322.1
2.2
2.3
2.3.1
2.3.2
2.3.3
2.3.4
2.3.5
2.3.6
2.3.7
Abstract……………………………………………………………………..…-33Introduction…………………………………………………………………...-34Materials and Methods………...……………………………………………..-36Preparation of Colicin E1 Single-Cys and Single-Stop Codon Variants………-36Synthesis of Coumarin Fluorescent Amino Acid………………………………-36Expression and Purification of Coumarin Incorporated Colicin E1 Variants.…-37Selective Labeling of Single-Cys Variants with DABMI……………………...-38Preparation of Large Unilamellar Vesicles (LUVs)……………………………-39Emission Spectra Measurements…………………………………………...…..-39Absorbance Spectra Measurements……………………………………………-40-
5
vi
2.3.8
2.3.9
2.3.10
2.3.11
2.4
2.4.1
2.4.2
2.4.3
2.4.4
2.4.5
2.4.6
2.4.7
2.4.8
2.4.9
2.4.10
2.4.11
2.5
Fluorescence Quantum Yield Measurements…………………………………..-40Fluorescence Lifetime Measurement…………………………………………..-40Fluorescence Lifetime Data Analysis……………………………………….…-41Modeling the Membrane-Bound Channel Domain………………………….…-41Results……………………………………………………………………...….-43Colicin E1 Variants………………………………………………………….…-43Coumarin Fluorescent Amino Acid Synthesis and Its tRNA/Aminoacyl-tRNA
Synthetase Pair System………………………………………………………...-45Protein Expression and Purification……………………………………………-48Folded Integrity of the Coumarin-Tethered Colicin E1 Variants……………...-51Spectroscopic Measurement of Coumarin and DABMI-Cys Tethered
Adducts…………………………………………………………………………-51Determination of the Förster Distance (R0)………………………………….…-55Fluorescence Lifetime Decay Measurements………………………………….-57FRET Efficiencies within the Channel Domain………………………………..-57FRET-Derived Distances within the Channel Domain………………………...-59Comparison of the X-ray Coordinates with FRET Data…………………….…-59Closed-State Model of Colicin E1……………………………………………..-59Discussion……………………………………………………………………...-64-
CHAPTER 3
Harmonic Analysis of the Fluorescence Response
of Bimane Adducts of Colicin E1 at Helices 6, 7, and 10…....…..-70Abstract……………………………………………………………….……….-71Introduction………………………………………………………….………..-71Materials and Methods……………………………………………….………-74Materials………………………………………………………………………..-74Mutagenesis, Protein Purification, and Monobromobimane Labeling...………-74Preparation of Large Unilamellar Vesicles (LUVs)……………………………-756-Methoxy-N-(3-sulfopropyl) quinolinium (SPQ) Assay for in Vitro Channel
Activity…………………………………………...…………………………….-753.3.5 Steady-State Intrinsic Tryptophan Fluorescence ………………………………-763.3.6 Bimane Fluorescence Emission Spectra ……...……………………………….-773.3.7 Steady-State Bimane Fluorescence Anisotropy.……………………………….-773.3.8 Dual Quenching Anslysis………………..……………………………………..-783.3.9 Prediction of Secondary Structure from Fluorescence Parameters...…………..-793.3.10 Data Fitting……………..………………………………………………………-803.3.11 In Silico Analysis of the Soluble Structure…………………………………….-833.4 Results…………………………………………………………………………..-843.4.1 Mutagenesis, Protein Purification and mBBr Labeling…....…………………..-843.4.2 Structural and Functional Analysis of Cys Variants…………………………...-893.4.3 Bimane Fluorescence Emission Maxima of H6, H7, and H10 Cys Variants…..-923.4.4 Bimane Fluorescence Anistropy & Probe Mobility of H6, H7, and H10 Cys
Variants………………………………………………………………………...-933.4.5 Dual Quenching Analysis of the Membrane-bound Depth of H6, H7 and H10
Cys Variants………………………………………………………………..…..-953.4.6 Harmonic Analysis………………………….………………………………….-973.1
3.2
3.3
3.3.1
3.3.2
3.3.3
3.3.4
6
vii
3.5 Discussion……………………………………………………………………….-102-
CHAPTER 4
Resolving the 3-D Spatial Orientation of Helix I of
the Colicin E1 Channel Domain by FRET...…………………....-108Abstract…...………………………………………………………………….-108Introduction…………...……………………………………………………...-109Materials and Methods….…………………………………………………...-111Preparation of Single-Trp, Single-Cys Colicin E1 Variants………………….-111Protein Purification and Monobromobimane Labeling.……………………....-112Preparation of Large Unilamellar Vesicles…………………………………...-113Emission Spectra Measurement………………………………………………-113Absorbance Spectra Measurement……………………………………..……..-113Fluorescence Lifetime Measurement…………………………………………-114Fluorescence Lifetime Data Analysis………………………………………...-114Calculation of FRET Efficiencies and Apparent Distances…………………..-115Modeling Methodology……………………………………………………….-116Results…...……………………………………………………………………-118Preparation of Colicin E1 Single-Trp and Single-Cys Variants……………...-118Protein Expression and Purification…………………………………………..-121Spectroscopic Measurement of Trp and mBBr-Cys Tethered Adducts………-121Calculation of the FRET Distances…………………………………………...-124Time-Resolved Fluorescence Measurements for the Soluble State of ColE1..-130Time-Resolved Fluorescence Measurements for the Membrane-Component State
of Colicin E1………………………………………………………………….-1314.4.7 The 3D model for the Membrane-Bound State of Colicin E1………………..-1364.4.8 Detailed Features of the New Model………………………………………….-1394.4.9 Validation of the New Model for the Colicin E1 Closed-State……………….-1464.5
Discussion……………………...……………………………………………..-1494.1
4.2
4.3
4.3.1
4.3.2
4.3.3
4.3.4
4.3.5
4.3.6
4.3.7
4.3.8
4.3.9
4.4
4.4.1
4.4.2
4.4.3
4.4.4
4.4.5
4.4.6
CHAPTER 5
Summary and Conclusions..………………….-153-
References………………………………………………………...-156-
7
viii
LIST OF TABLES
Table 1-1: Classification of colicins based on their Tol and Ton-dependent import
mechanism….……………………………………………………………………………-7Table 2-1: Spectral Parameters and Distances for Fluorescence Energy Transfer between
Coumarin Donors and Cys-DABMI Acceptors of the Soluble Colicin E1 Channel
Domain…………………………………………………………………………………-56Table 2-2: Spectral Parameters and Distances for Fluorescence Energy Transfer between
Coumarin Donors and Cys-DABMI Acceptors of the Colicin E1 Channel Domain in the
Membrane-Bound State…………………………………………….………….……….-56Table 2-3: Comparison of FRET-Derived Distance Constraints (R) and Cα – Cα
Distances from the Calculated Model of the Membrane-Bound Closed State of the
Colicin E1 Channel Domain……………………………………….…………………..-61Table 2-4: Membrane Binding Energies (∆Gtransf), Membrane Penetration Depths (D),
and Tilt Angles Relative to the Membrane Normal (τ) of Individual α-Helices and αHairpins of Colicin E1 Channel Domain Calculated Using PPM 2.0………..………...-61Table 4-1. Fluorescence lifetime data analysis of the three Tryptophan donors in the
presence and absence of the bimane acceptors in the soluble state……………..…....-125Table 4-2. Spectral parameters and distances for fluorescence energy transfer between
Tryptophan donors and Cys-Bimane acceptors in the soluble state…………………..-126Table 4-3. Fluorescence lifetime data analysis of the three Tryptophan donors in the
presence and absence of the bimane acceptors in the membrane-bound state………..-132Table 4-4. Spectral parameters and distances for fluorescence energy transfer between
Tryptophan donors and Cys-Bimane acceptors in the membrane-bound
state……………………………………………………………………………….…...-133Table 4-5. Lifetime data analysis of the fluorescent decay for the Tryptophan donors
alone and donor-acceptor pair………………………………………………………...-147Table 4-6. Calculation of the FRET-efficiency and FRET-distances based on both the
standard and selective FRET calculation…………………………………...………...-147-
8
ix
LIST OF FIGURES
Figure 1-1. Schematic representation of colicins classes based on their various modes of
actions and transported proteins used for cell entry……………………………...……...-4Figure 1-2. Organization of the colicin operons……………………………………..…-5Figure 1-3. Crystal structures of the pore-forming domain of colicins A, N, B, Ia and
E1………………………………………………………………………………………-10Figure 1-4. Schematic representation of the two models for the closed channel state of
colicin E1………………………………………………………………………..……...-13Figure 1-5. Schematic representations of the transmembrane segments of the open
colicin channel for colicin Ia and A…………………………………...……………….-16Figure 1-6. FRET efficiency versus the Förster distance (Å)………………….……..-23Figure 1-7. Spectral overlap between CFP and YFP………………………….……...-26Figure 2-1. Schematic representation of the colicin E1 channel domain……………...-44Figure 2-2. Synthesis of coumarin fluorescent amino acid (L-(7-hydroxycoumarin-4yl)ethylglycine…………………………...………….………….………………………-47Figure 2-3. Optimization and expression test of the coumarin-containing colicin E1
channel domain…………………………………………………………………..….....-50Figure 2-4. Circular dichroism (CD) analysis of the coumarin-containing colicin E1
channel domain………………………………………………………………………...-53Figure 2-5. Fluorescence and absorbance spectra of both coumarin and DABMI-Cys
tethered adducts within the channel domain in the presence and absence of
membrane…………………………………………………………………..…………..-54Figure 2-6. Apparent distances and energy transfer efficiencies between the various
donor (coumarin) and acceptor (DABMI) pairs within colicin E1………………....….-58Figure 2-7.
Membrane-bound closed-state model of the colicin E1 channel
domain………………………………………………………………….……………....-69Figure 3-1. Preparation of colicin E1 Cys variants…………………..…...…………..-85Figure 3-2. Labeling of Cys ColE1 variants with mBBr……………...………………-88-
9
x
Figure 3-3. Structural and functional tests of Cys ColE1 variants..……...…………...-91Figure 3-4. Fluorescence emission maximum and probe mobility of the bimane-labeled
Cys variants of the colicin E1 channel domain.……………………..…………………-94Figure 3-5. Plot of the quenching ratio…………...……………………...…………....-96Figure 3-6.
Harmonic analysis of the SC for the soluble state of the colicin E1
CD……………………………………………………………………………………...-99Figure 3-7. Harmonic analysis of the SC for colicin E1 for the membrane-bound
state….…………………………………………………………………….…...……...-101Figure 3-8. Models of colicin E1 in the membrane-bound state…...……...………...-107Figure 4-1. Schematic representation of the colicin E1 channel domain……...….....-120Figure 4-2. Fluorescence and absorbance spectra of both Tryptophan and Bimane-Cys
tethered adducts within col E1 in the presence and absence of membrane…………...-123Figure 4-3. Distances between Trp and selected positions on helix 1...……..….…..-129Figure 4-4. Distances between the Trp and selected positions on H1 for the membranebound colicin channel domain……………………………………...…………………-135Figure 4-5. New 3-D model of the membrane-bound colicin E1……………...…….-138Figure 4-6. Overlap of the two models of the membrane-bound colE1……...….…..-140Figure 4-7. Model of the embedding and unfolding of colicin E1…………...…...…-142Figure 4-8. Segmental unwinding on the previous model……..…………...………..-145Figure 4-9. Possible orientation of H1………………..……..………...…………….-152-
10
xi
GLOSSARY OF ABBREVIATIONS
10-DN
10-doxylnonadecane
AMP
Ampicillin
Bimane-Cys
N-acetyl-cysteine conjugated with bimane
CD
Circular dichroism
CSM
Cysteine-scanning mutagenesis
DMG
Dimethylglutaric acid
DOPC
1,2-dioleoyl-sn-glycero-3-phosphocholine
DOPG
1,2-dioleoyl-sn-glycero-3-[phosphoric-(1-glycerol)]
DTT
Dithiothreitol
E
Energy transfer efficiency
F0
Fluorescence intensity in the absence of quencher
FD
Fluorescence intensity in the absence of acceptor
FDA
Fluorescence intensity of the donor with the acceptor
FPLC
Fast protein liquid chromatography
FRET
Forster resonance energy transfer
FTIR
Fourier transform infrared spectroscopy
G
Instrument factor
HPSA
Helical periodicity surface analysis
LMV
Lipid multilayer vesicle
LUV
Large unilamellar vesicles
mBBr
Monobromobimane
11
xii
NATA
N-acetyl-tryptophanamide
PDB
Protein data bank
PFA
Pore-forming ability
QF
Fluorescence quantum yield
Q-ratio
Ratio of quenching
r
Anisotropy
R
Distance between donor and acceptor
R0
Fӧrster distance
SA
Solvent accessibility
SASA
Solvent accessible surface area
SDS-PAGE
Sodium dodecyl sulfate polyacrylamide gel electrophoresis
P
Periodicity
J
Spectral overlap integral
kT
Rate constant of energy transfer
PBS
Phosphate-buffered saline
λem,max
Fluorescence wavelength emission maximum
SPQ
6-methoxy-N-(3-sulfopropyl) quinolinium
WT
Wild-type
KI
Potassium iodide
rpt
Residue per turn
P190H6
Colicin E1 190 residue channel domain with an N-terminal
6 histidine tag
DABMI
4-(dimethylamino)phenylazophenyl-4’-maleimide
SC
Spectral Centroid
12
Chapter 1
1.1
Introduction
Colicin Biology
1.1.1 Colicins
Colicins are antimicrobial proteins produced by Escherichia coli that target
susceptible bacteria in response to stressful conditions, including nutrient depletion, DNA
damage, overcrowding and anaerobiosis (1). The first colicin was identified by Gratia in
1925 as a heat-labile product present in the cultures of E. coli V (43) and the name colicin
was coined by Gratia and Fredericq in 1946, who demonstrated its killing activity against
related species (44). The first experimental studies on colicins were performed by Jacob
et al. in 1952 (45) who demonstrated that colicins are only lethal against related species
but not active against the producing bacteria due to the presence of specific immunity
proteins. It was not until 1963 that scientists were able to classify the various colicins
based on three different modes of lethal action: (i) the formation of a depolarizing ion
channel in the cytoplasmic membrane, (ii) the inhibition of protein and peptidoglycan
synthesis, and (iii) the degradation of nucleic acids (24).
In order for colicins to exert their lethal action, an import mechanism must exist
for entry of colicins into host cells. This served as the basis for classifying colicins into
two groups based on the transport route that was used for host cell entry. Group A
colicins are translocated by the Tol system and include colicins A, E1 to E9, K, L, N, S4,
U and Y. Group A colicins requires the aid of TolA, TolQ and TolR transport proteins
for host cell entry (Fig 1-1). In contrast, group B colicins uses the TonB system and
include colicins B, D, H, Ia, Ib, M, 5 and 10, which generally require TonB, ExbB, ExbD
13
1
and various other transport proteins for cell import (22). In general, colicins that use the
Tol system kill host cells by either the formation of a depolarizing ion channel in the
cytoplasmic membrane or degradation of nucleic acids. In contrast, colicins that use the
TonB system tend to form either a depolarizing ion channel or inhibit protein and
peptidoglycan synthesis in order to kill the host cell as shown in Figure 1-1.
Colicins are produced by strains of E. coli that carry the pCol plasmid and there
are two classes of pCol: type I and type II (Figure 1-2). Type I plasmids mainly encode
group A colicins while type II pCol are larger plasmids that encode group B colicins (22).
In terms of genetic organization of the pCol plasmids, the first gene that encodes colicin
is under the regulation of the LexA promotor whereas the following genes encode either
the immunity or the lysis protein which is under the regulation of its own constitutive
operon that allows a constant production. However, not all colicin plasmids consist of all
three genes as shown in Figure 1-2. One reason for the constitutive production of its
immunity protein is to ensure the absence of free colicins that could potentially kill the
producing cell whereas the lysis protein allows the release of colicins into the medium
(46).
In order for the colicin-producing bacteria to protect against auto-infection, some
of them produce a specific immunity protein co-expressed with colicins. The immunity
protein is located at the cytoplasmic membrane of the producing organism and it forms a
tight complex with the pore-forming domain of colicins that inhibit channel function
(46). As illustrated in Figure 1-2, the transcription of the colicin operon is largely
regulated by the SOS promotor, which is normally repressed by the LexA protein.
However, when the SOS response is triggered by either mutagenic or carcinogenic agents
2 14
that invoke stress conditions, the RecA protein becomes activated in response to the
autocleavage of LexA that allows the transcription of the colicin operon (22).
15
3
Figure 1-1. Schematic representation of colicin classes based on their various modes of
action and transport proteins used for cell entry (22). The name of each colicin is
indicated by the arrow. The double layer rectangle represents the outer and inner
membrane of the host cell. The shaded boxes represent the transport proteins used by the
specific colicins for cell entry. The vertical dotted line divides colicins into either TonB
or Tol-dependent groups. The horizontal dotted line further divides colicins into two
groups based on their modes of action: enzymatic or pore-forming.
16
4
Figure 1-2. Organization of colicin operons (22). The genes are represented by the
black arrowheads. Vertical arrows indicate either promotor (P) or terminator (T) region.
Cxa represents the genes for the specific colicin while cxi and cxl represents the
immunity gene and lysis protein gene, respectively.
17
5
1.1.2 Colicins Structure & Import Mechanism
The general structural organization of colicins involves three domains with the
translocation (T) domain located at the N-terminus of the protein, followed by the
receptor-binding (R) domain in the middle and the cytotoxic (C) channel-forming domain
located at the C-terminal end (20). In fact, the toxic action of colicins involves three
major steps and the participation of each functional domain of the proteins. Generally,
the first step involves R-domain binding to an outer membrane receptor protein. This
induces unfolding of the colicin T-domain that initiates translocation of the protein across
the outer membrane through an import channel protein located at the outer membrane.
This allows the colicins to translocate into the periplasm through specific interaction
between the T-domain and the appropriate import channel proteins.
This translocation process often involves the mediation of various other factors
located within the inner membrane. Finally, the colicin C-domain exerts its cytotoxicity
by one of the three mechanisms: (i) formation of a depolarizing ion channel in the
cytoplasmic membrane; (ii) the inhibition of protein and peptidoglycan synthesis, and
(iii) the degradation of nucleic acids (24). In general, colicins from different groups
(group A or B), depending on their specific lethal action, utilize various outer membrane
receptor-binding proteins as well as their cognate import channels as shown in Table 1-1.
However, there is only a limited set of receptor-binding and channel proteins that is
shared among the various types of colicins.
18
6
TABLE 1-1. Classification of colicins based on their Tol and Ton-dependent import mechanism (22).
translocation proteins required for each specific colicin entry into host cells are indicated.
Both receptor and
TABLE #1 (Group A / B; receptor; import; cytotoxicity)
a
The group A and group B colicins (Tol- and Ton-dependent colicins, respectively), their outer membrane receptors, and the colicin cytotoxic activities are
shown.
b
Cell envelope proteins required for injection of phage DNA.
1.2
Colicin E1
Colicin E1 is a member of the group A colicins that uses the Tol-dependent
system for host cell entry. It belongs to the ion channel-forming group of colicins; this
group also includes colicins A, B, Ia, Ib, N and K (Figure 1-1). It is encoded by the type
I pCol plasmid that transcribes both the colE1 immunity and lysis proteins.
The
immunity protein for colicin E1, Imm, is a 13-kDa integral membrane protein composed
of three-transmembrane helices (46).
It is known that the colE1 immunity protein
interacts with the C-terminal channel domain of colicin E1 in order to protect the
producing cells against infection. Previous mutagenesis studies demonstrated that residue
448 at the N terminus of helix VI and residues 470, 472, and 474 at the C terminus of
helix VII of the channel domain are the critical residues responsible for its immunity
protein recognition (46).
The colicin E1 transport mechanism uses BtuB as an outer membrane receptorbinding protein that interacts with and recognizes the R-domain.
Such interaction
induces unfolding of its T-domain that thereby initiates binding and translocation across
the TolC import channel into the periplasm. As shown in Table 1-1, this translocation
process is also mediated by the TolA and TolQ inner membrane proteins. Finally, the Cdomain of colicin E1 forms an ion channel in the host cell inner membrane. As a result,
it creates a pore in the inner membrane that allows the release of various ions from the
host cells, such as Na+, K+, and H+, and subsequently cell death ensues.
8
20
1.2.1 Colicin E1 Structure
The crystal structure of the colicin E1 channel domain was determined to a
resolution of 2.5 Å (11). Although the crystal structure of the full-length colicin E1 has
never been determined, it likely has a similar structure to the solved full-length colicin Ia
structure (39). In general, the structure of the colicin E1 soluble channel domain revealed
10 α-helices that form an extremely stable, water-soluble protein. This protein comprises
a hydrophobic segment, helices 8 and 9, which act as the hydrophobic core of the soluble
protein (19). The channel domain structure has been described as a helical sandwich that
is folded into three layers: Layer A, the outer layer composed of helices 1, 2 and 10;
Layer B, the inner core layer including helices 5, 8 and 9; and Layer C, the outer layer
composed of helices 3, 4, 6 and 7 (11).
Previous studies demonstrated that the channel domain function and activity is
independent of both the receptor-binding (R) and the translocation (T) domains.
Presently, there are a total of five different crystal structures of pore-forming colicin
channel domains including colicin E1, A, Ia, N, and B. All of the structures are similar in
size, the number of α-helices and the hydrophobic helical hairpin segments as shown in
Figure 3 (22). However, no colicin membrane-bound structure has been determined by
either crystallography or NMR due to technical difficulties.
21
9
Figure 1-3. Crystal structures of the pore-forming domain of colicins A, N, B, Ia and E1
(22). Each type of colicin is shown in two different orientations (top and side view). The
numbers of helices are shown for the colicin A structure. The helices colored in red
represents the two hydrophobic helices and the remaining helices are colored in blue.
22
10
1.2.2 Colicin E1 Membrane Binding
The colicin E1 channel domain binds the lipid membrane with high affinity even
in the absence of a membrane potential; however it does not form an ion-channel unless a
trans-negative potential is present. Previous fluorescence studies suggested that upon
translocation across the host cell outer membrane, colicin E1 adopts an insertioncompetent state which allows the hydrophobic core (helices 8 and 9) to penetrate the
target membrane. Thus, the protein unfolds, binds, and spontaneously inserts into the
membrane to form a closed channel in a series of kinetically defined steps (20).
Subsequently, the channel opens in the presence of a trans-negative membrane potential
and the two channel states exist in equilibrium (21).
Upon interaction with the lipid bilayer and prior to the imposition of a transnegative potential, the channel domain forms an intermediate state structure known as the
closed-channel model. In fact, two well-known structural models have been proposed for
the closed channel state, which are the penknife and umbrella models (Fig. 1-4) (22).
The penknife model suggested that both helices 1 and 2 move away from the rest of the
channel domain that sinks deep into the membrane to form a penknife-like shape. This
model was supported by disulfide bond engineering experiments through the introduction
of cysteine residues that link adjacent helices together by the formation of disulfide
bonds. Results suggested that a single disulfide bond prevents the insertion of the
channel domain by linking helices 1 and 2 with the rest of the structure (47).
In contrast, the umbrella model suggests that only hydrophobic helices, 8 and 9,
are inserted into the hydrophobic milieu of the membrane, and the remaining 8 helices
23
11
spread out onto the membrane surface to form an umbrella-like structure. In fact, the
umbrella model was strongly supported by time-resolved fluorescence resonance energy
transfer (FRET) studies on colicin A which showed that the distances between the
hydrophobic hairpin segment (helices 8 and 9) and the rest of the channel domain
increased by 10 to 15 Å in the membrane-bound state (48). In addition, previous studies
in our laboratory demonstrated that helices 1 – 5 of the colicin E1 channel domain are
amphipathic α-helices that retain their α-helical structure and rest upon the membrane
surface in the lipid-bound state (24-26). However, the exact orientation of the remaining
helices, their depth of bilayer penetration as well as the details of the lipid and protein
contacts are still poorly understood.
24
12
Figure 1-4. Schematic representation of the two models for the closed channel state of
colicin E1 (22). A. Penknife model based on disulfide bonds engineering experiments
on colicin A. B. Umbrella model based on fluorescence resonance energy transfer
(FRET) experiments performed on colicin A and E1.
25
13
1.2.3 The Ion-Channel Formation by Colicin E1
All colicin channels are voltage dependent as positive voltage induces opening of
the channel while negative voltage closes the channel. Previous experimental evidence
suggested that all colicin channels are monomeric and only allow the passage of
monovalent ions, including K+, Na+ and Cl- (49). The formation of the open state was
shown to be influenced by a number of factors including the thickness of the bilayer
membrane (acyl chain length) as well as the spontaneous curvature of the lipid. Although
the exact structure and helical orientation of the open channel has not been determined,
studies on colicins E1 and A demonstrated that several helices, including the first three
helices of the channel domain, can be eliminated without preventing the formation of the
open channel. This was confirmed by mapping the exposure of biotin tags introduced at
various residues to streptavidin.
It was clear from these experiments that helices 2 to 5 do not adopt a
transmembrane orientation in the open state, possibly driven completely across the
membrane by the introduced positive voltage. This leaves the open channel with only
four transmembrane segments to span the membrane which include helices 1 and 6 as
well as the two hydrophobic hairpin segments (helices 8 and 9), the latter adopting a
transmembrane orientation at zero potential (19). As illustrated in Figure 1-5, the open
channel state of both colicin A and Ia was shown to have a helical arrangement and this
further suggested a similar open channel structure for colicin E1 (22). Previous studies
showed that tetraethyl ammonium (TEA), a cation of about 8 Å diameter, can easily
permeate through the open channel (51). Therefore, it was estimated that the diameter of
26
14
the colicin E1 open channel is at least 8 Å. In fact, tetra-cis-2-hydroxyethyl ammonium,
which is even larger than TEA, was able to readily pass through both the colicin A and
colicin Ia open channels (50).
Although the open channel allows the passage of various cellular ions out of the
cells, including K+, Na+ and Cl-, that eventually leads to cell death, this can be rescued
through interaction with the immunity protein. A major role of the immunity protein is to
protect the colicin-producing bacteria against invasion through interaction with the poreforming domain. As described earlier, mutagenesis studies demonstrated that the colicin
E1 immunity protein interacts with residue 448 at the N terminus of helix 6 and residues
470, 472 and 474 at the C terminus of helix 7 of the channel domain (46). According to
Figure 1-5, both helices 6 and 7 of colicin A and Ia form part of the transmembrane
helical structure of the open channel and it is likely that colicin E1 adopts a similar
structure in its open channel state. Therefore, it is possible that the colicin E1 immunity
protein confers resistance against its own colicin by interaction with two of the newly
adopted transmembrane helices, helices 6 and 7, in the open channel state in order to
block the pore.
27
15
Figure 1-5. Schematic representations of the transmembrane segments of the open
colicin channel for colicin Ia and A (22). A. The proposed open channel state for colicin
Ia. Helices 2 to 5 are translocated across the membrane upon imposition of a transnegative potential. B. Proposed open channel state for colicin A. The colicin A channel
resembles that of colicin Ia except helices 2 to 5 are not translocated completely across
the membrane.
28
16
1.2.4 Closed Channel Structure of Colicin E1
Although evidence exists to support both models for the closed channel structure,
more recent evidence favours the umbrella model over the penknife model. In fact, the
strongest support of the penknife model came from the disulfide bond engineering
experiment and the FRET inter-helical distance measurements that were performed on
colicin A. Four cysteine residues were introduced within helices 1, 2, 3, and 10 of colicin
A and disulfide bond formation between helices 2 and 10 prevented membrane insertion
of colicin A (47). In addition, FRET studies on the inter-helical distances in the soluble
and membrane-bound states confirmed that, in the absence of a membrane potential,
helices 1 and 2 splay outward onto the membrane surface and move away from the rest of
the structure (48). However, these studies on colicin A do not provide direct and specific
evidence to support the penknife model of colicin E1.
In support of the umbrella model, various experiments on colicin E1 revealed
that only helices 8 and 9 adopt a transmembrane orientation whereas the remaining 8
helices are spread out onto the surface of the membrane.
Previous site-directed
mutagenesis studies showed that the hydrophobic domain spans the membrane bilayer
twice in a helical hairpin loop (52) whereas NMR studies confirmed that transmembrane
helices account for approximately 20-25% of the channel polypeptide, which is
equivalent to 38 – 48 residues of the 190 residue polypeptide (53). Therefore, these
criteria fit well with helices 8 and 9 existing as transmembrane helices. In addition,
fluorescence quenching experiments demonstrated that residue 492, located within the
loop region between helices 8 and 9, was deeply embedded within the lipid bilayer.
29
17
Therefore, it was concluded that helices 8 and 9 do not adopt a fully transmembrane state.
In fact, this was clearly supported by a recent study that investigated the lipiddestabilizing properties of helices 8 and 9, which suggested that the hydrophobic
segments behave like tilted peptides (19). In support of the parallel orientation relative to
the membrane bilayer for helices 1 – 7 and 10, our lab used cysteine-scanning
mutagenesis in combination with site-directed fluorescence labelling to show that helices
1 – 7 and 10 are amphipathic α-helices that reside at the interfacial region of the
membrane bilayer. In addition, previous protease accessibility experiments indicated that
the region between K420 – K461 is a surface-exposed segment that is not likely inserted
into the membrane in the absence of a membrane potential (54).
30
18
1.2.5 Colicin Applications
With the increase in antibiotic-resistance arising in many bacterial species as well
as the increase in antibiotic usage, bacteriocins such as colicins have potential as
antibiotic alternatives. Previous studies demonstrated that feeding bacteriocin-producing
strains to livestock can prevent the growth of certain pathogenic bacteria (55). Colicinproducing E. coli strains such as the Nissle 1917 have been successfully used to treat calf
diarrhea as well as inflammatory bowel syndrome and Crohn’s disease (56).
In addition, colicins were also found to inhibit proliferation of tumor cell lines as
they appear to be more toxic to tumor cells than to normal cells; however, the tumor
targeting mechanism is still unknown (56).
The application of colicin bacteriocins
presents new opportunities for the development of both antibiotic alternatives and antitumor drugs.
However, this will heavily rely on further research not only on the
identification of novel bacteriocins, but also on understanding both the structure and the
functional mechanism behind the toxin itself.
31
19
1.3
Fluorescence Spectroscopy
1.3.1 Fluorescence Resonance Energy Transfer (FRET)
The elucidation of membrane protein structure often presents great challenges to
scientists due to the failure of both NMR and X-ray crystallography to solve this problem.
One promising technique that shows great success in mapping structural changes of
membrane proteins is Förster resonance energy transfer (FRET). The basic idea of FRET
is to measure the energy transfer efficiency between two different fluorophores in order
to derive the separation distance. Therefore, the introduction of fluorophores at various
locations of a protein would enable scientists to precisely probe the structural changes
that occur over time. In addition, FRET has been a popular technique for measuring
protein-protein interactions between two fluorophore labelled proteins.
For instance, FRET has been used for the visualization of various protein
interactions including oligomerization and binding of ligands/cargo to membrane protein
receptors (58). Besides, FRET has also been applied to the development of various
biosensors that require the incorporation of fluorophores at specific sites in proteins. In
fact, such sensors have already been developed for the intracellular assay of calcium ions,
cAMP activity as well as protease activity (59, 60). One objective of this study is to
apply FRET to estimate the molecular distances between various sites within the colicin
E1 channel domain in order to build a low-resolution 3-D model of its membrane-bound
state that cannot be solved by either NMR or X-ray crystallography.
Historically, resonance energy transfer theory was first developed by Theodor
Förster in 1946 based on his first paper on the study of photosynthesis (61). Although
32
20
FRET is an old-fashioned technique that was invented more than half a century ago, it
has a wide variety of applications in science and it is still a common technique for
monitoring protein conformational changes. Theoretically, FRET involves the presence
of two chromophores: a donor and an acceptor, where an energy transfer occurs from the
donor molecule to the acceptor molecule by a non-radiative long-range dipole-dipole
coupling mechanism if the two chromophores are sufficiently close together (58).
The term non-radiative implies that the energy transfer process does not involve
the emission of a photon from the donor followed by absorption by an acceptor molecule.
In order for the energy transfer process to occur, the distances separating the two
chromophores must be less than the wavelength of the excitation light. In addition, it
requires sufficient overlap between the emission spectrum of the donor and the
absorption spectrum of the acceptor. The distances separating the two chromophores
must be less than 10 nm and FRET can only provide an accurate measurement when the
distances are in the range of 10 – 100 Å which is well-suited for the study of most protein
interactions within the biological systems (62, 63).
According to the theory of ‘resonance energy transfer’, the FRET efficiency
varies as the sixth power of the distance separating the two chromophores. As shown in
Figure 1-6, the Förster distance (R0) is the distance between the donor and acceptor when
the energy transfer efficiency is 50%. For distances either higher or lower than the
Förster distance, the energy transfer efficiency changes sharply to either maximal or zero.
Therefore, it is important that the working distance separating the two chromophores is
close to the characteristic Förster distance for the donor-acceptor pair.
The FRET
calculations indicate that there is a complex relationship between energy transfer
33
21
efficiency and separation distance that involves various parameters. Firstly, as illustrated
in Figure 1-6, the energy transfer efficiency is dependent upon the distances separating
the two chromophores which can be described in the following equation:
[Eq. 1]
where E represents the efficiency of energy transfer, R represents the distance separating
the two chromophores, and R0 is the Förster distance (R0) which is the distance between
the two chromophores when the energy transfer efficiency is 50%. In order to calculate
the distance separating the donor and acceptor, it is important to determine both the
efficiency of energy transfer (E) as well as the Förster distance (R0).
2234
Figure 1-6. FRET efficiency versus the Förster distance (Å). The FRET efficiency
(EFRET) varies with the sixth power of distance between the donor and acceptor. When R
= R0, the energy transfer is 50% efficient between the donor and the acceptor. The
changes of E as a function of R/R0 is most pronounced within the region where 0.5< R/R0
<1.5, therefore the effective range of FRET distances is 0.5 R0 < R < 1.5 R0 (58).
2335
The energy transfer efficiency (E) can be obtained experimentally by either
measuring the fluorescence intensity (steady-state) or the lifetime (time-resolved) of the
donor in the presence and absence of the acceptor as described by the relationship below:
[Eq. 2]
where FDA and FD represents the fluorescence intensity of the donor in the presence and
absence of the acceptor, respectively; and τDA and τD represents the lifetime of the donor
in the presence and absence of the acceptor, respectively. Notably, it was demonstrated
that the use of lifetime decay provides a more reliable efficiency measurement.
According to equation 3, the Förster distance (R0) can be calculated based on the
following relationship:
[Eq. 3]
where κ2 is the orientation factor, QD is the quantum yield of the donor in the absence of
acceptor, η is the refractive index of the medium and J is the spectral overlap between the
donor and acceptor which is given by the following equation:
[Eq. 4]
36
24
where FD is the normalized fluorescence intensity of the donor, A is the molar extinction
coefficient of the acceptor, and λ is the wavelength. The spectral overlap integral is based
on the overlap between the emission spectrum of the donor and the absorption spectrum
of the acceptor as illustrated in Figure 1-7. As described earlier, quantum yield is
measured simply by calculating the ratio of photons emitted through fluorescence relative
to photons absorbed.
37
25
Figure 1-7. Spectral overlap between cyan fluorescent protein (CFY) and yellow
fluorescent protein (YFP). The spectral overlap between donor and acceptor is
determined as the overlap region between the emission spectrum of the donor (CFP, blue
line) and the absorbance spectrum of the acceptor (YFP, yellow line). The spectral
overlap is represented by the shaded grey region (58).
38
26
One of the parameters within the Förster equation that requires further explanation
is the orientation factor, κ2.
Basically, κ2 represents the angle between the two
fluorophores similar to the positions of radio antenna. When the two fluorophores are
parallel to each other, the FRET efficiency will be higher than if they are perpendicular
(58). Therefore, the value of the orientation factor plays a role in the calculation of the
Förster distance. However, κ2 is usually assumed to be 2/3, based on the average value of
all possible angles. Since all the parameters within the Förster equation only affect its
calculation by the sixth power, small changes of both the quantum yield and spectral
overlap integral calculation would not contribute any significant error towards the Förster
distance calculation. For instance, doubling of the donor quantum yield value would only
result in a 12.5% change in the Förster distance. Therefore, it is the measured lifetime
value of the donor determined in both the presence and absence of the acceptor that has
the most significant impact towards the FRET distance calculation.
In terms of the reliability of the FRET technique, there are certain limitations that
should be considered which could significantly affect the accuracy of its measurement.
As mentioned earlier, the Förster distance is dependent upon the orientation factor, κ2,
which can be equal to zero if the two fluorophores are precisely aligned at certain angles.
For instance, if the transition dipole moments of the two fluorophores are facing in an
opposite direction to each other, no FRET signal would be observed even if the
fluorophores are relatively close to one another.
Another technical issue related to FRET is the labelling stoichiometry of the
fluorophore.
Ideally, we would observe 100% labelling efficiency if every protein
molecule receives one fluorophore molecule. However, this is not always the case in
39
27
practice and labelling efficiency usually ranges between 50 – 100% (64). Labelling
efficiency of higher than 100% would indicate the presence of excess fluorophore
associated with the protein either due to covalent or non-covalent attachment and this
should never be used for FRET measurement. Similarly, labelling efficiency of less than
50% should also be avoided as it indicates that more than 50% of the protein molecules
contain no label. As a result, this would not reflect accurate measurement of the donor
lifetime in the presence of the acceptor which would significantly affect the FRET
distance calculation. Similarly, the stoichiometry between the donor and acceptor would
also affect the FRET measurement in a similar manner.
Despite the various technical limitations of the FRET technique, this method has
been successful in probing the structural changes of various colicins including colicins A
and E1 (48).
Several investigations on the colicin A membrane-bound structure
performed by Lakey et al. used FRET to probe the structural changes of various locations
of the protein. Five residues were replaced with cysteine for I-AEDANS labelling while
the three Trp residues were used as the donors in order to probe the changes of the FRET
distances at the five different locations. Results revealed that all five distances increased
upon membrane association which suggested that helices 1 and 2 splay out onto the
membrane surface while the helical hairpin remains closely packed against the rest of the
structure (48).
A previous study in our lab involved the application of FRET to derive structural
changes of colicin E1 upon activation to its insertion-competent state. Eleven single Trp
mutants were prepared by introducing them along the entire protein sequence at aromatic
residues. FRET efficiencies were measured from each Trp residue (donor) to Cys-505
40
28
labelled with I-AEDANS as the acceptor fluorophore. Results revealed that higher FRET
efficiencies were observed the closer the Trp donor was to the N-terminus of the protein
in the pH-activated state. This strongly supports the idea that the formation of the
insertion-competent state for colicin E1 requires helices 1 and 2 unravelling from the
body of the protein.
Finally, a recent FRET study on the colicin E1 membrane-bound structure by
Cramer et al. revealed that the eight amphipathic α-helices of the channel domain form a
quasi-circular arrangement on the membrane surrounding the hydrophobic helical hairpin
(65). The experiment involved creating six single Trp mutants scattered throughout the
protein sequence and using Cys-509 labelled with I-AEDANS as the acceptor
fluorophore. Time-resolved FRET efficiencies were measured to probe the structural
changes that occur upon membrane association. To further test this model, we used
FRET to produce a higher resolution model of the membrane-bound channel domain.
Herein, we used three single Trp mutants (donors) and measured FRET
efficiencies to three randomly chosen sites within helix 1 (Cys-bimane acceptor). Based
on triangulation theory, the 3-D position and orientation of each helix can be precisely
mapped on the membrane surface.
In combination with our 2-D membrane-bound
topology data for helix 1, the goal was to precisely map the 3-D position of helix 1
relative to the channel domain in its membrane-bound state.
41
29
1.4
Research Objectives
Although the X-ray structures of various colicins in the soluble state have been
determined, no membrane-bound structures has been solved.
Despite significant
improvements in both the fields of X-ray crystallography and NMR, the structure
determination of membrane proteins has been difficult. In an attempt to elucidate the
membrane-bound closed channel structure of colicin E1, various techniques were used
including: circular dichroism (CD), Fourier transform infrared spectroscopy (FTIR) and
Fӧrster resonance energy transfer (FRET). These techniques not only provided changes
in helical and beta sheet content between the soluble and membrane-bound states, but it
also helps monitor changes in inter-residue distances. In addition, with the use of sitedirected mutagenesis, fluorescence quenching, site-directed labelling as well as statistical
analysis, our laboratory was able to develop a 2-D membrane-bound model of the colicin
E1 closed channel structure.
Previous studies in our laboratory resulted in a 2-D membrane-bound topology
model of helices 1 – 5 of the colicin E1 closed channel structure by using cysteinescanning mutagenesis (CSM) in combination with site-directed fluorescence labelling.
This method involved scanning the protein structure in a residue-by-residue-fashion (24,
25).
Herein, we continue to use this approach to investigate the membrane-bound
topology of helices 6, 7 and 10 (Val-447 – Gly-475 and Ile-508 – Ile-522) as presented in
Chapter 3. Using site-directed fluorescence labelling combined with our novel helical
periodicity analysis method, we precisely mapped the 2-D membrane-bound topology of
helices 6, 7 and 10 and also determined their helical boundaries.
42
30
A second objective of this work was to measure the inter-helical distances in the
membrane-bound closed channel. The use of FRET in this approach should allow us to
build a low-resolution 3-D model of the closed channel structure. In an attempt to derive
the 3-D orientation of helix 1 relative to the channel domain, three residues (D347, S354
and E361) were chosen for cysteine replacement followed by bimane labelling (Chapter
4). The distances in each donor-acceptor pair (one of three naturally occurring Trp
residues – Trp 424, 460 or 495) and acceptor (Cys-bimane) were measured by FRET.
Therefore, a total of nine distances were measured for helix 1 in the channel peptide for
both the soluble and membrane-bound states.
In addition, our lab has successfully employed the genetically encoded fluorescent
amino acid system developed by Schultz and co-workers to incorporate coumarin into the
colicin E1 channel domain to act as an intrinsic FRET donor to the DABMI-labeled Cys
residues (acceptor) as presented in Chapter 2. Although labor intensive, this approach
allowed us to determine six additional interhelical distances within helices 1 – 6 that
represents an important step towards the construction of a low resolution 3-D model of
the colicin E1 channel domain in the absence of either NMR or X-ray structural data.
43
31
CHAPTER 2
MEMBRANE TOPOLOGY OF THE COLICIN E1
CHANNEL USING GENETICALLY ENCODED
FLUORESCENCE
This work has been published:
Ho, Derek., Lugo, Miguel, R., Lomize, Andrei, L., Pogozheva, Irina, D., Singh, Suneel,
P., Schwan, Adrian, L. and Merrill, A. R. (2011) Biochemistry. 50: 4830-4842
Contributors
Ho, Derek
Lugo, Miguel
Lomize, Andrei
Pogozheva, Irina
Singh, Suneel
Schwan, Adrian
Merrill, Rod
Statement of Contribution
Conducted all experiments that includes mutants preparation,
coumarin synthesis, labelling, FRET measurements, data analysis
and writing the manuscript
Aided with lifetime data analysis, FRET distances calculation,
computer modeling and wrote part of the manuscript
Aided with computer modeling of the membrane-bound channel
domain
Aided with coumarin synthesis and lab equipments
Supervision and funding of the project
44
32
2.1
Abstract
The membrane topology of the colicin E1 channel domain was studied by
fluorescence resonance energy transfer (FRET).
The FRET involved a genetically
encoded fluorescent amino acid (coumarin) as the donor and a selectively labeled
cysteine
residue
tethered
with
DABMI
(4-(dimethylamino)phenylazophenyl-4’-
maleimide) as the FRET acceptor. The fluorescent coumarin residue was incorporated
into the protein via an orthogonal tRNA/aminoacyl-tRNA synthetase pair that allowed
selective incorporation into any site within the colicin channel domain. Each variant
harbored a stop (TAG) mutation for coumarin incorporation and a cysteine (TGT)
mutation for DABMI attachment. Six interhelical distances within helices 1 – 6 were
determined using FRET analysis for both the soluble and membrane-bound states. The
FRET data showed large changes in the interhelical distances among helices 3 – 6 upon
membrane association providing new insight into the membrane-bound structure of the
channel domain.
In general, the coumarin-DABMI FRET interhelical efficiencies
decreased upon membrane binding, supporting the umbrella model for the colicin
channel. A tentative model for the closed state of the channel domain was developed
based on current and previously published FRET data. The model suggests circular
arrangement of helices 1 – 7 in a clockwise direction from the extracellular side and
membrane interfacial association of helices 1, 6, 7, and 10 around the central
transmembrane hairpin formed by helices 8 and 9.
3345
2.2
Introduction
The C-terminal channel-forming domain of colicin E1 forms a lethal ion channel
which depolarizes the cytoplasmic membrane of target bacterial cells (3).
Prior to
channel formation, the channel peptide first binds to the lipid bilayers, followed by
protein unfolding and helix elongation (4 - 6, 11). Finally, the channel domain adopts an
insertion-competent state in which it inserts into membrane to form the prechannel state
(7). The channel then opens in response to a trans-negative membrane potential and
facilitates the escape of various ions from the host cells, such as Na+, K+, and H+, leading
to host cell death (8).
The crystal structure of the soluble channel domain (2.5 Å) (9, 12) shows 10 αhelices that form an extremely stable, water-soluble globular protein (13, 14).
Interestingly, the protein consists of a hydrophobic α-helical hairpin, helices 8 and 9,
which act as the non-polar core of the protein and becomes transmembrane upon
membrane association (15). These two helices are critical to colicin pore formation by
forming a membrane-spanning hairpin that anchors the channel within the bilayer (20, 16,
52). The amphipathic α-helices of the channel peptide were shown to surround the
hydrophobic core of the channel (17, 24 – 26,). In the membrane environment, the
channel peptide forms a structure previously described as an umbrella in which only the
hydrophobic helices 8 and 9 are inserted into the hydrophobic core of the membrane with
the amphipathic helices splayed out onto the membrane surface (18, 19). The umbrella
model has received strong experimental support from time-resolved FRET studies of
colicin E1 (23, 27, 28). However, the precise orientation of the helices as well as the
details of the lipid and protein contacts are still poorly understood.
46
34
It has been shown that the eight amphipathic α-helices on the membrane surface
adopt a two-dimensional arrangement with an area of 4200 Å2, an increase of more than
3-fold of the cross-sectional area of the soluble channel domain (29). Similarly, FRET
data for colicin A revealed that interhelical distances generally increase upon membrane
association (30-32). Our previous study involved FRET analysis where Cys-505 was
labeled with I-AEDANS as acceptor and 11 Trp donor residues were randomly situated
throughout the channel domain. These results revealed that higher relative changes in
FRET efficiencies were observed the closer the Trp donor was to the N-terminus of the
protein (23).
In support of the umbrella model, the Cramer group adopted a similar approach
by using FRET to probe the relative distance of each helix to Cys-509 (29). Although the
data could be accounted for by the formation of a quasi-circular arrangement of the eight
amphipathic α-helices laying on the membrane surface, a number of other models with
various two-dimensional configurations of the helices are also possible. To further test
the proposed quasi-circular arrangement model, we used the system developed by Schultz
and co-workers to incorporate coumarin into the colicin E1 channel domain to act as an
intrinsic FRET donor (33) to DABMI-labeled Cys residues (acceptor). Although labour
intensive, this approach can facilitate the construction of a low-resolution 3-D model of
the closed channel in the absence of both NMR and X-ray structural data.
47
35
2.3
Materials and Methods
2.3.1 Preparation of Colicin E1 Single-Cys and Single-Stop Codon Variants
All colicin E1 channel domain mutants were prepared by site-directed
mutagenesis as previously described (26). A total of 8 single-Cys and single-stop codon
mutants were prepared using the P190H6 construct. All plasmids were purified using the
High Pure Plasmid isolation kit from Roche Diagnostics (Laval, PQ, Canada), and all
mutation sites were confirmed by DNA sequencing (University of Guelph). Each protein
variant consists of one single-Cys and one single-stop codon mutation within each
adjacent helix from helices 1 – 6: (F355C/K369stop, helix 1 and 2; K369stop/L390C,
helix 2 and 3; L390stop/A411C, helix 3 and 4; A411stop/K426C, helix 4 and 5;
K426stop/V441C, helix 5 and 5-loop; V441stop/L452C, helix 5-loop and helix 6 and the
two internal controls are K369stop/A371C and K369stop/K379C, helix 2 and 2).
2.3.2 Synthesis of Coumarin Fluorescent Amino Acid
Step 1: Synthesis of (2S)-2-benzyloxycarbonylamino-5-oxo-heptanedioic acid 1benzyl ester 7-ethyl ester. Carbonyldiimidazole (1.92 g, 11.9 mmol, 1.1 equiv) was
added to a solution of N-carbobenzyloxy-L-glutamic acid α-benzyl ester (Z-Glu-Obzl,
4.00 g, 10.8 mmol, 1.0 equiv) in dry THF (50 mL) (mixture A) while monoethyl
malonate (14.2 g, 108 mmol, 10.0 equiv) was added to magnesium ethoxide (6.16 g, 53.9
mmol, 5.0 equiv) in dry THF (100 mL) (mixture B). Both mixtures were stirred for 4 h at
room temperature before mixture A was added slowly to the mixture B, and the stirring
continued at room temperature overnight. The solvent was removed in vacuo, and the
48
36
product was dissolved in 20 mL of diethyl ether and 20 mL of 0.5 M HCl. The product
was extracted with diethyl ether (3 x 30 mL) and washed with 20 mL of 10% NaHCO3
and 20 mL of brine. The organic layer was treated with anhydrous MgSO4, filtered, and
concentrated in vacuo to afford a milky colored solid. LC-MS (ESI) M/Z calculated for
C24H27NO7 (M + H+) is 442.18; observed 442.07.
Step 2: Synthesis of L-(7-hydroxycoumarin-4-yl)ethylglycine.
The product,
benzyloxy-carbonylamino-5-oxoheptanedioic acid 1-benzyl ester 7-ethyl ester (2.00 g,
4.52 mmol, 1.0 equiv), was added slowly to resorcinol (5.00 g, 45.2 mmol, 10.0 equiv) in
methanesulfonic acid (10.9 g, 113 mmol, 25.0 equiv) on ice. The mixture was stirred at
room temperature for 4 h. Five volumes of cold ether were added to mixture, which was
then cooled at – 30 °C for 5 min. The mixture was warmed to room temperature while
stirring until a yellow / red solid formed as clumps. The precipitate was washed with
cold ether to afford the desired product, which was dissolved in water and used in this
study without further purification. LC-MS (ESI) M/Z calculated for C13H14NO5 (M +
H+) is 264.08 Da; observed 264.14 Da.
2.3.3 Expression and Purification of Coumarin Incorporated Colicin E1 Variants
Both the P190H6 (amp resistance) and pEVOL-CouA (chloramphenicol
resistance) plasmids (33) were used to cotransform BL-21 E. coli and were plated onto
2xYT media agar containing both ampicillin and chloramphenicol acetyl-transferase
selection markers. As a control, BL-21 cells that were transformed with a single plasmid
did not grow in the presence of both selection markers. A single colony from the
49
37
cotransformation was grown overnight at 37 °C (18 h) in 250 mL of 2xYT media
supplemented with ampicillin, chloramphenicol, 0.02% arabinose, and 5 mM coumarin
amino acid. As a control, another colony was grown under identical conditions, except in
the absence of the unnatural amino acid.
Successful coumarin incorporation was
confirmed by testing for colicin E1 channel domain expression using SDS-PAGE.
Variants with proper expression levels were purified as previously described (24).
Protein purity was assessed by SDS-PAGE gel analysis, and protein concentration was
determined by absorbance spectroscopy at A280, using the extinction coefficient (ε) of
29910 M-1cm-1 for the WT channel domain. Notably, the extinction coefficient was
modified to 39610 M-1cm-1 due to the presence of coumarin within the variants.
2.3.4 Selective Labeling of Single-Cys Variants with DABMI
All 8 single-coumarin, single-Cys variants were selectively labeled with DABMI
(4-(dimethylamino)phenylazophenyl-4’-maleimide) to form a DABMI-Cys peptide
adduct. The variants were equilibrated in 80 μM DTT in pH 8.1 buffer for 30 min before
DABMI dye was introduced at a 20:1 molar ratio (probe: protein) and incubated for 1 h.
The mixture was separated by passing it through a BioRad 10DG 10 mL chromatography
column monitoring the eluant by UV/Vis absorbance (250 – 650 nm). The labeling
process was the same as previously described (24). Labeling efficiency was calculated
by determining the concentration of protein relative to the concentration of the dye. The
following extinction coefficients were used for DABMI – Cys: ε460 = 24800 M-1cm-1 and
50
38
ε280 = 8400 M-1cm-1, and protein concentration was calculated from the total Abs 280 nm
taking into account the contribution from DABMI – Cys at 280 nm.
2.3.5 Preparation of Large Unilamellar Vesicles (LUVs)
LUVs were prepared by an extrusion method from 1,2-dioleoyl-sn-glycero-3phosphocholine and 1,2-dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] vesicles at a
60:40 molar ratio (Avanti Polar Lipids, Alabaster, AL) as described earlier. Colicin was
bound to liposomes to form a pre-channel (closed state) by adding 4 μM colicin channel
domain (final conc) in DMG buffer (20 mM dimethyl glutarate, 130 mM NaCl, pH 4.0)
to a solution of LUVs (800 μM, final concentration) as described previously (25, 35).
2.3.6 Emission Spectra Measurements
The fluorescence emission of the incorporated coumarin in the absence of
acceptor was measured at an excitation wavelength of 365 nm, and emission was scanned
between 375 and 650 nm (1 nm steps) and 0.2 s integration time.
Excitation and
emission band-passes were set at 4 and 6 nm, respectively. Samples were measured in
both the presence and absence of LUVs.
2.3.7 Absorbance Spectra Measurements
51
39
The absorbance of the DABMI – Cys labeled variants was measured as previously
described (24), except the wavevlength was scanned between 250 and 650 nm for the
purpose of overlap integral (J) calculations. Absorbances were measured in Halma ultramicroabsorbance cuvettes (light path 10 mm) using a Cary 300 spectrophotometer.
2.3.8 Fluorescence Quantum Yield Measurements
The fluorescence quantum yields of the protein-incorporated coumarin in the
absence of acceptor were measured under identical conditions as described above, except
the excitation light was vertically polarized and the emission was measured at the magic
angle (54.7°). Quinine sulfate in 0.1 N H2SO4 was used as a reference standard, and the
quantum yield was calculated as previously described (24).
2.3.9 Fluorescence Lifetime Measurement
A PTI Laser strobe model C-72 lifetime fluorimeter was used for the timeresolved fluorescence measurements. Instrument response function (IRF) was measured
using a 0.0005% scatter solution. Samples were excited at 365 nm using a pulsed
nitrogen dye-laser operated at 10 Hz. Emission was collected at 450 nm with start and
end delay set at 58 and 120 ns, respectively. Measurements were carried out with 400
channels, 50 ns, integration time, and 15 shots averaged together at 25 °C.
2.3.10 Fluorescence Lifetime Data Analysis
52
40
The data analysis was performed using a 1-to-4 exponential fitting program that
involves the deconvolution of the fluorescence decay. Deviations of the best fits were
characterized by the reduced χ2 statistical analysis.
In addition, residual graphics,
autocorrelation curves, and Durbin – Watson statistics were also used to assess the
quality of each fit. The average fluorescence lifetime (τ) was calculated from the
relationship τ = Σαi τi / Σαi and because Σαi = 1 (normalized pre-exponential values),
then τ = Σαi τi.
2.3.11 Modeling the Membrane-Bound Channel Domain
An all-atom model of the membrane-bound domain was generated based on 14
FRET-derived distance constraints obtained in the present work and previously published
by Cramer’s group (29). The modeling included two stages. First, all individual αhelices (H1 to H10) and potential membrane-bound α-hairpins (H1 – H2, H6 – H7, and
H8 – H9) were taken from the crystal structure of the soluble form (PDB 2i88 file) and
spatially arranged in the membrane by minimizing their transfer energy from water to the
lipid bilayer using the PPM 2.0 method (37, 38). The method is based on the anisotropic
solvent representation of the lipid bilayer described by complex profiles of dielectric
constant and hydrogen-bonding parameters along the membrane normal.
The
calculations were carried out for the native protein and mutants with fluorescent labels
(bimane, DABMI, AEDANS, and coumarin) attached to cysteines or genetically
incorporated at positions 347, 355, 365, 396, 411, 426, 441, and 509.
53
41
Side-chain
conformers were optimized as previously described (40). In the second stage, all helices
and α-hairpins were brought to the common membrane coordinate frame and manually
arranged to fit the set of distance constraints using the molecular modeling module of
QUANTA. The helices were considered as rigid bodies with short covalent connections
and fixed positions with respect to the membrane plane.
54
42
2.4
Results
2.4.1 Colicin E1 Variants
Based upon the earlier quasi-circular arrangement model for membrane-bound
colicin E1 (29), a number of 2-D configurations of the channel domain on the surface of
the bilayer are possible. This is primarily due to the lack of constraints on the model,
since there were only five FRET distances (vectors) used in its construction. To reduce
the number of possible models, we applied a non-natural amino acid incorporation
strategy (33) to genetically encode coumarin fluorescent residues into the channel domain
to act as the FRET donors. This method involved extrinsically labeling a Cys residue
with DABMI (nonfluorescent acceptor) and then measuring FRET efficiencies between
the coumarin (donor fluorophore) and DABMI. The first step involved the preparation of
6 single-stop codon and single-Cys variants to estimate interhelical distances among
helices 1 through 6 (Figure 2-1). The colicin E1 channel domain sequence and the
location of mutation sites within the sequence and the 3-D structure are shown in Figure
8B, C. In this study, the lone Cys residue (Cys-505) was replaced with Ala-505 to avoid
the possibility of double Cys labeling.
Previous studies confirmed that the C505A
mutation does not perturb either the secondary or tertiary structure of the channel domain
(42).
55
43
Figure 2-1. Schematic representation of the colicin E1 channel domain. (A) Quasicircular arrangement model of the colicin E1 channel domain viewed from the top of the
membrane. The 5 FRET distances relative to C509 residue ( – ) are from a previous
report (29). The 6 interhelical distances between the stars are the distances measured in
this study. (B) The primary sequence and secondary structure of the channel-forming
domain of colicin E1 (P190H6). Residues squared in black were subjected to either a stop
or Cys codon replacement for the purpose of this study. The natural Cys-505 was
replaced with Ala-505. (C) The ribbon topology diagram of the 2.5 Å crystal structure of
the P190 peptide (PDB: 2I88). The overall structure consists of 10 α-helices where
helices 8 and 9 are hydrophobic helices (shown in dark) that serve as a membraneanchoring helical hairpin in the membrane-associated state. The location of the residues
subjected to either cysteine or stop codon replacements are shown in sticks and are
labeled.
56
44
2.4.2 Coumarin Fluorescent Amino Acid Synthesis and its tRNA/Aminoacyl-tRNA
Synthetase Pair System
Coumarin was chosen as the FRET donor due to its relatively high quantum yield,
low photobleaching, high stability, and available synthesis (33). Herein, we report a
modified synthetic method with lower cost and greater reproducibility than a previous
report (41).
The coumarin derivative (L-(7-hydroxycoumarin-4-yl)ethylglycine) was
synthesized by first converting N-α-Cbz-L-glutamic acid α-benzyl ester into its side-chain
β-ester as shown in Figure 2-2 (step 1).
The second step involved resorcinol
condensation with a β-ketoester in the presence of a Brønsted and Lewis acid
(methanesulfonic acid) using the von Pechmann reaction pathway to produce the amino
acid form (68-72). Both absorbance and emission spectra of the final product confirmed
the amino acid derivative (Fig 2-2B). NMR (not shown) and mass spectrometry analysis
also supported the correct structure and molecular mass of the coumarin fluorescent
amino acid (Fig 2-2C). To enable the incorporation of the coumarin into specific sites
within the channel domain, it required an orthogonal tRNA/aminoacyl-tRNA synthetase
pair that facilitated the selective introduction of the amino acid into the protein in
response to the amber stop codon, TAG. In this study, we utilized the pEVOL-CouA
plasmid obtained from the Schultz group (73). The vector replicates under a p15A origin
of replication, it carries the chloramphenicol acetyltransferase (CAT) marker, and it is
arabinose inducible. The system features a variant Methanococcus jannaschii tyrosyl
amber suppressor tRNA (MjtRNATyrCUA)/tyrosyl-tRNA synthetase (MjTyrRS) pair that
was uniquely evolved to recognize the coumarin amino acid in response to the TAG
codon. The new pEVOL system was reported to show increased plasmid stability and
4557
affords higher yields of variants using both constitutive and inducible promotors to drive
the transcription of two copies of the M. jannaschii aaRS gene (73).
4658
Figure 2-2. Synthesis of coumarin fluorescent amino acid (L-(7-hydroxycoumarin-4yl)ethylglycine. (A) Conversion of N-α-Cbz-L-glutamic acid α-benzyl ester into the sidechain β-keto ester form, followed by reaction with resorcinol in methanesulfonic acid to
produce the final coumarin fluorescent amino acid. (B) Absorption (▲) and emission (■)
spectra of the purified coumarin fluorescent amino acid. Fluorescence emission intensity
was normalized to the absorbance values. (C) Mass spectrometry analysis of the purified
coumarin fluorescent amino acid. LC-MS (ESI) M/Z calculated for C13H14NO5 (M + H+)
is 264.08 Da; observed 264.14 Da.
59
47
2.4.3 Protein Expression and Purification
To optimize colicin expression for coumarin incorporation, a number of
parameters were tested as shown in Figure 2-3. Both the induction and incubation time
did not affect colicin channel domain expression (Fig. 2-3A), and we were able to
successfully purify the coumarin-containing variants (K369Cou/L390C example in
Figure 2-3A (v)).
Not surprisingly, the percentage of arabinose and the coumarin
concentration were the most critical parameters in driving the expression of the channel
domain. Induction with 0.02% arabinose in the presence of 5 mM coumarin amino acid
were optimal for protein expression as shown in Figure 2-3A (ii, iii), as found previously
(73). The BL-21 E. coli strain was cotransformed with the various P190H6 mutant
plasmids along with the pEVOL-CouA plasmid. Cotransformed colonies were grown in
250 mL of 2xYT media supplemented with ampicillin, chloramphenicol, 0.02%
arabinose, and 5 mM coumarin amino acid and incubated at 37 °C for 18 h. Channel
domain expression was tested by SDS-PAGE, and it was found that only 5 of the 6
variants showed appreciable expression (Fig 2-3A (iv)). Since the F355stop/K369C
variant showed no expression, its donor/acceptor pair location was swapped to produce
the F355C/K369stop variant. Remarkably, the swapping of the donor/acceptor location
restored the channel domain expression to an acceptable level (Fig 2-3A (iv, last lane)).
To assess the coumarin incorporation efficiency into the channel domain, the purified
K369Cou/L390C variant was subjected to ESI – mass spectrometry analysis. The protein
mass was 21 944.2 Da, which nicely correlated with the calculated mass of 21 944.8 Da
(Fig 2-3B). Furthermore, LC-MS / MS analysis confirmed the presence of the coumarin
at residue 379 and a Cys residue at position 390 (data not shown). All 8 variants
60
48
including the two controls were successfully purified (> 95% pure) using immobilized
metal-affinity chromatography (IMAC), as previously described (24).
61
49
Figure 2-3. Optimization of expression test of the coumarin-containing colicin E1
channel domain. (A) (i) Expression levels of colicin E1 with 5 mM coumarin amino acid
and 0.02% arabinose in the growth medium followed by induction at different time points
with different incubation time. (ii) Identical conditions to (i) except arabinose
concentration was varied from 0% to 2% and induced at 0 h. (iii) Identical conditions to
(i) except the concentration of crude coumarin amino acid was varied from 0 to 5 mM
and 0.02% arabinose at 0 h. (iv) Expression test of the 6 single-Cys single-stop codon
variants in 5 mM coumarin amino acid and was induced with 0.02% arabinose at 0 h. (v)
Comparison between purified and nonpurified colicin E1 mutant (K369Cou/L390C). (B)
Mass spectrometry analysis of the K369Cou/L390C variant. Mass spectrometry was
performed on a Qtof mass spectrometer (Micromass) equipped with a Z-spray source and
run in positive ion nanospray mode.
62
50
2.4.4 Folded Integrity of the Coumarin-Tethered Colicin E1 Variants
To assess the structural integrity of the channel domain coumarin variants, Trp
fluorescence emission maximum values (Trp λem,max) were acquired for the WT and all
variants.
The WT channel domain gave a Trp λem,max value near 323 nm in good
agreement with previous data (26). The Trp λem,max values of the variants were similar to
the WT protein. In addition, CD analysis of the variants confirmed that the secondary
structure was also similar to the WT protein (Figure 2-4) . Thus, the incorporation of the
coumarin residue into the channel domain did not significantly perturb the folded
integrity of the variants.
2.4.5 Spectroscopic Measurement of Coumarin and DABMI-Cys Tethered
Adducts
To evaluate the chemical structure and properties of the incorporated coumarin,
both absorbance and fluorescence emission spectra were collected as shown in Figure 25A. The absorption and fluorescence emission spectra of the K369Cou/L390C variant
clearly showed the presence of the coumarin residue within the protein. All 8 variants
including controls were subjected to DABMI-Cys labeling, and the representative
absorbance spectra for the K369Cou/L390C variant are shown in the presence and
absence of covalently tethered DABMI (Fig 2-5B) with its structure shown in Fig 2-5D.
The hallmark peak at 460 nm confirms the successful generation of the DABMI-Cys
tethered adducts. Based on the given extinction coefficients for DABMI, ε460 = 24 800
M-1 cm-1 and ε280 = 8400 M-1 cm-1 (74), the labeling efficiencies for all 8 variants ranged
63
51
between 71% and 108%. The labeling efficiency likely varied due to differences in the
solvent accessibility of the target Cys site within the channel domain.
64
52
Figure 2-4. Circular dichroism (CD) Analysis of the coumarin-containing colicin E1
channel domain. The helical content of all proteins were assessed using a Jasco J-815
circular dichroism (CD) spectrometer. Samples were prepared in 50 mM sodium
phosphate buffer at pH 7.0 and spectra were acquired with a 0.5 mm path-length cuvette
with scanning from 190 – 250 nm. Each spectrum was the average of 8 spectra.
65
53
D
Figure 2-5. Fluorescence and absorbance spectra of both coumarin and DABMI-Cys
tethered adducts within the channel domain in the presence and absence of membrane.
(A) Absorption () and emission () spectra of the coumarin-incorporated
K369Cou/L390C variant. Absorbance values were normalized to the fluorescence
emission intensity peak for coumarin. (B) Absorption spectra of the purified
K369Cou/L390C variant with () and without () DABMI-Cys labeling. (C) Spectral
overlap between the coumarin fluorescence emission spectrum (square symbols) and
DABMI absorbance spectrum ( – ) of the purified K369Cou/L390C variant. Coumarin
emission was measured in the presence () and absence () of membrane to illustrate
the changes in the spectral overlap integral. Absorbance units were normalized to the
fluorescence emission intensity.
(D) Structure of the quencher, 4dimethylaminophenylazophenyl-4’maleimide (DABMI), with molecular weight of
320.35 Da and absorption at 460 nm. It is a non-fluorescent thiol-reactive acceptor for
use in fluorescence resonance energy transfer (FRET) applications.
66
54
2.4.6 Determination of the Förster Distance (R0)
To determine the distance separating the donor / acceptor chromophores within
the channel domain, the initial step involves calculation of the Förster distance (R0)
which is defined as the distance when the energy transfer efficiency is 50%. According
to eq 3, this calculation requires knowledge of the quantum yield (QF) of the donor
(coumarin) in the absence of acceptor as well as the spectral overlap integral (J). The
coumarin emission and the DABMI absorption spectra of K369Cou/L390C are shown in
Figure 2-5C. Subtle changes were observed for the coumarin emission spectrum upon
membrane association. In general, the intensity of coumarin emission decreased upon the
addition of LUVs. In contrast, the coumarin absorbance spectrum remained unchanged
upon membrane binding.
To determine the fluorescence quantum yield (QF) of the coumarin donor, quinine
sulfate was used as a quantum standard. According to Table 2-1 and 2-2, the quantum
yield values of coumarin at various sites within the colicin channel domain ranged from
0.62 to 0.81 (an average of 0.67). These values fall within the expected range for
coumarin (33). In general, the quantum yield of the coumarin donors decreased in the
lipid-bound state (Table 2-2). Furthermore, the Förster distance (R0, Table 2-1 and 2-2)
ranged between 13.4 and 26.8 Å for the soluble state and 13.2 and 26.9 Å for the
membrane-bound state with an average Förster distance near 22.0 Å for the soluble state
and 21.9 Å for the lipid-bound state. Overall, no significant changes were observed
between the two states. Importantly, the Förster distance is ideal for this donor / acceptor
pair for estimation of the inter-residue distances within the channel domain.
67
55
Table 2-1. Spectral parameters and distances for fluorescence energy transfer
between coumarin donors and Cys-DABMI acceptors of the soluble Colicin E1
channel domain.
Table 2-2. Spectral parameters and distances for fluorescence energy transfer
between coumarin donors and Cys-DABMI acceptors of the Colicin E1 channel
domain in the membrane-bound state.
68
56
2.4.7 Fluorescence Lifetime Decay Measurements
Accurate FRET measurements require the careful determination of the donor
fluorescence lifetime (free coumarin τ  5.4 ns). The coumarin fluorescence lifetime
when it was tethered to the channel domain showed an average value of 5.98 ns with a
range of 5.88 – 6.12 ns for the soluble state (Table 2-1 and 2-2). In the presence of the
DABMI acceptor, the coumarin fluorescence lifetime was reduced depending on the
distance between the donor / acceptor pair, as expected (Table 2-1 and 2-2).
2.4.8 FRET Efficiencies within the Channel Domain
The FRET efficiencies between the coumarin donor and DABMI acceptor for
colicin channel domain intrahelical residues ranged from 59.6% (K369 to F355) to 25.6%
(K426 to V441) for the soluble channel domain and from 1.5% (K427 to V441) to 44.8%
(K369 to A371) for the membrane-bound protein (Table 2-1 and 2-2; Figure 2-6A). In
general, the FRET efficiencies were smaller in the lipid-bound state compared to the
soluble state reflecting the expansion of the channel domain upon binding to the
membrane surface. The changes in the FRET efficiencies upon membrane association
were all negative in value and ranged from – 5.0 to – 29.1% (ΔE, Table 2-2 and Figure 26A). These changes imply that the relative spacing between helices of the channel
domain increased upon membrane binding.
69
57
Figure 2-6. Apparent distances and energy transfer efficiencies between the various
donor (coumarin) and acceptor (DABMI) pairs within colicin E1. (A) The energy
transfer efficiencies for the various donor and acceptor pairs were determined by lifetime
analysis of the coumarin donors in the presence and absence of the DABMI acceptor.
Changes of the energy transfer efficiency were compared between the soluble () and
membrane-bound states (). The mean score, taken from the average of all data, for the
soluble state (M = 0.36 SD = 0.003, N = 6) was significantly larger than the scores for
membrane-bound state (M = 0.17, SD = 0.015, N = 6) using the two-sample t test for
unequal variances, t (5) = 5.47, p < 0.003. (B) Apparent distances between the various
donor and acceptor pairs were measured by FRET in both the soluble () and
membrane-bound states (). The mean score for the soluble state (M = 33.66, SD =
26.32, N = 6) using the two-sample t test for unequal variances, t (5) = 2.91, p < 0.03.
The numbers in parentheses correspond to the donor and acceptor sites (residue numbers
within the colicin channel domain) for the FRET efficiency measurements. (C)
Comparison between the soluble FRET distances () and the estimated distances based
on the X-ray crystal structure (PDB: 2i88) of the soluble state (). The α-carbon atom
was used as a reference point for all crystal structure distance measurements.
70
58
2.4.9 FRET-Derived Distances within the Channel Domain
On the basis of eq 1, the FRET distances (R) were calculated as shown in Table 21 and 2-2. The inter-residue distances as calculated from the FRET efficiencies ranged
from 12.5 Å (K369 to A371) to 29.9 Å (A411 to K426) for the soluble channel domain
and from 13.7 Å (K369 to A371) to 40.9 Å (K426 to V441) for the membrane-bound
structure (Table 2-1 and 2-2; Figure 2-6B).
All of the distances increased for the
measured sites upon membrane binding and the increases ranged from 0.6 Å (K369 to
K379) to 17.5 Å (K426 to V441) (ΔR, Table 2-2).
2.4.10 Comparison of the X-ray Coordinates with FRET Data
To assess both the accuracy and reliability of the data, the FRET distances for the
soluble channel domain were plotted against the distances obtained from the X-ray
crystal structure (9, 12), as shown in Figure 2-6C. Although there is a strong correlation
in the distance pattern between the two data sets, nonetheless, the FRET distances were
generally greater than the X-ray distances. One possible explanation may be due to the
shorter distances introduced by the packed crystal structure as compared to the soluble
form of the protein in solution.
2.4.11 Closed-State Model of Colicin E1
We have previously demonstrated that this colicin – lipsosome system used in our
present study represents a functionally relevant colicin prechannel state (closed) that upon
imposition of a membrane potential forms an active ion channel (25, 35). On the basis of
the FRET data in this study, it was found that distances between helices 1 and 2 and
71
59
between helices 2 and 3 did not change much upon membrane association (Figure 2-6B).
In contrast, the largest changes in helical spacing upon membrane binding occurred
between the helix 5 loop and helices 5 and 6. Also, significant increases in distances
were observed for helix 3 to 4 and helix 4 to 5 (Figure 2-6B). To reconcile our new data
with the earlier report by Lindeberg and co-workers (29), we generated a new model of
the membrane-bound state using 14 FRET-derived constraints (Table 2-3). In the first
stage of our modeling procedure, spatial positions in the membrane were calculated for
individual α-helices and several α-hairpins, whose structures were taken from the soluble
x-ray structure of the protein (Table 2-3). When associated with the lipid bilayer, most
helices are known to remain the same as in the crystal structure (24-26, 42), while H3
may be elongated by a few residues.
72
60
Table 2-3. Comparison of FRET-derived distance constraints (R) and Cα – Cα
distances from the calculated model of the membrane-bound closed state of the
Colicin E1 channel domain.
Table 2-4. Membrane binding energies ( Gtransf), membrane penetration depths
(D), and tilt angles relative to the membrane normal (τ) of individual α-helices and
α-hairpins of Colicin E1 channel domain calculated using PPM 2.0.
73
61
According to the calculations, individual amphipathic helices (from H1 to H7, and
H10) are located at the membrane surface with nonpolar residues inserted in the
hydrocarbon core below the boundary formed by carbonyl groups of phospholipids
(Table 2-4).
The surface helices have tilt angles of 80° - 90° with respect to the
membrane normal. H1 appears to be the most tilted helix with the N-terminus inserted at
the depth of ~ 8 Å, in agreement with fluorescence spectroscopy studies (26). Transfer
energies of surface helices are – 3.5 to – 10.5 Kcal/mol. H2 and H10 demonstrated the
weakest membrane binding (transfer energies of – 4.5 and – 3.5 kcal/mol, respectively),
while H1 showed the largest binding energy for the individual surface-located helix
(ΔGtransf = - 10.5 kcal/mol). In contrast, the more hydrophobic H8 and H9 helices likely
adopt a transmembrane orientation (transfer energies of – 12.5 and – 14 kcal/mol and tilt
angles of 27° and 54°, respectively). The incorporation of fluorescent probes into the
interfacial sites only slightly affected the arrangement of helices in the membrane (the
largest effect was seen for H1 labeled at F355C).
We also considered the possibility of the formation of stable membrane-bound αhairpins by H1 – H2, H6 – H7, and H8 – H9 pairs found in the crystal structure of the
soluble channel domain. Significant binding energies of H6 – H7 and H8 – H9 pairs ( 9.8 and – 15.9 kcal/mol, respectively) indicate that these helices could interact with
membrane as α-hairpins. In contrast, energy transfer between the H1 – H2 pair was small
and less than for individual H1 and H2 helices, indicating that H1 and H2 may dissociate
and rearrange in the membrane-bound state, unlike H6 – H7 and H8 – H9 pairs.
Membrane-penetration depth (D ~ 19 Å) and tilt angle (τ = 14°) of the α-hairpin
H8 – H9 suggest that it likely traverses the membrane, leading to the local membrane
74
62
thinning.
The comparison of the binding energy of the H8 – H9 α-hairpin in
transmembrane (- 15.9 kcal/mol) and surface (- 8.8 kcal/mol) orientations further justifies
its transmembrane arrangement. The parameters of α-hairpin H6 – H7 (D ~ 4 Å, τ = 88°)
indicate its surface location. Hence, our results suggest the existence of two stable
membrane-bound α-hairpins, H6 – H7 and H8 – H9, which are stabilized by short turns
and the classic “knobs into holes” packing of interacting helices with an inter-helical
angle of ~ - 160°, as observed in the crystal structure of the water-soluble channel
domain.
In the next modeling stage, all helices and α-hairpins H6 – H7 and H8 – H9 were
brought to the common membrane coordinate system and arranged to fit the set of
distance constraints (Table 2-3). The modeling is simplified by the presence of short
connections between almost all helices.
Hence, a distance between residues from
adjacent helices translates to an inter-helical angle, which defines the mutual arrangement
of two helices at the membrane surface. The close distance (~ 20 Å) between residues
from H9 (W495) and H1 (347, 355), H7 (W460) and between H1 (355) and H2 (369)
define the proximity of surface helices H1, H7 to H9 and circular helix arrangement with
clockwise direction as viewed from the extra-cellular surface (Figure 2-7). This model
also suggests aggregation of predicted α-hairpin H6 – H7 and helices H1, H10 around the
transmembrane α-hairpin H8 – H9.
The proposed model satisfies most of the FRET-derived distances, except those
with H5a (residues 424 and 426). However, this region of H5 is highly distorted in the
crystal structure and might be unfolded upon membrane binding, which would increase
the distances involving residues 424 and 426. Another discrepancy was observed for the
75
63
509 – 356 distance (Table 2-4). The smaller distance between Cα-atoms in the model (by
12 Å) may be observed because fluorescent probes attached to these residues are oriented
in opposite directions. A similar situation with large FRET distances was experimentally
observed for labels located in H2 at proximal residues 369 and 371 oriented in opposite
directions (distance ~ 13 Å, Table 2-1).
2.5
Discussion
NMR has contributed to our understanding of the membrane-bound structure of
colicin; (15, 57, 75); unfortunately to date, neither NMR nor X-ray crystallography
methods have been able to provide a membrane-bound structure of colicin E1. FRET
appears as the only viable alternative to provide sufficient data for a low-resolution 3-D
model of the membrane-bound protein. Although it is a tedious and labour-intensive
process that would ideally require hundreds of variants, it can be achieved based upon the
results of the present study. According to a previous model, the channel domain forms an
umbrella-like shape with helices 8 and 9 forming a transmembrane helical anchor with its
amphipathic helices laying parallel on the membrane surface in a quasi-circular
arrangement surrounding helices 8 and 9 (29).
Herein, we used a new FRET-based approach involving the engineering of the
non-natural fluorescent amino acid, coumarin, into the channel domain as the
fluorescence donor. Consequently, in combination with an extrinsically labeled cysteine
as the fluorescent acceptor, both inter- and intrahelical distances can more readily be
obtained. In this study, we report the success of applying this technique to generate
coumarin-derived colicin E1 variants. Mass spectrometry analysis confirmed the success
76
64
of both the synthesis of coumarin fluorescent amino acid as well as its successful
incorporation into the colicin E1 channel domain.
We propose that the generation of an intrinsic coumarin fluorescence donor
greatly reduces the FRET-measured error associated with the technique. This is largely
due to the elimination of nonspecific labeling when two covalent attachment sites serve
as the donor / acceptor pairs. We chose DABMI as the FRET acceptor due to its small
size, high quantum yield, its large spectral overlap integral with coumarin (donor)
emission, and as it has no intrinsic fluorescence. The DABMI compound exhibits a rigid
covalent bonding angle with a Cys residue, and it is generally free to rotate when tethered
to a protein surface. Consequently, it may help minimize the error associated with the
orientation factor (κ2) (76).
It was fortunate that most variants in this study showed reasonably good
expression. However, as shown in Figure 2-3A, no colicin expression was observed for
F355Cou/K369C variant, but upon reversal of the donor / acceptor pair location, protein
expression was recovered. Herein, it should be noted that the coumarin incorporation
system (33) efficiency not only depends on the site and the nature of the protein, but also
depends on by the mRNA context. In fact, we were unable to obtain reliable FRET data
for a distance estimate between helices 6 and 7 because of the difficulty in finding a
suitable incorporation site for the coumarin within helix 7. Unfortunately, no single rule
can be applied to govern the incorporation efficiency for each residue.
However,
previous experiments suggested that expressing variants in richer media with higher
fluorescent amino acid concentration does improve incorporation efficiency (73).
77
65
In this study, a high correlation was found between the FRET distances for the
soluble channel domain and the distances estimated from the X-ray crystal structure.
Although minor variations were present, most of the deviation could be explained by the
size and flexibility of the donor and acceptor chromophores. In general, the FRET data in
this study are in good agreement with previously reported FRET data (23, 29). Our data
further extend those of previous reports and clearly show that the distances separating
helices 1 – 3 within the colicin channel domain remain relatively unchanged upon
membrane binding, whereas the distances separating helices 3 – 6 significantly increase
for the lipid-bound state. In particular, the helix 5 – 6 loop region was shown to be
highly mobile, and the FRET data suggest that it opens up on the membrane surface upon
channel domain insertion into the bilayer.
The fact that both control variants,
K369Cou/A371C and K369Cou/K379C, showed little or no distance changes between
the soluble and membrane-bound states suggests that helix 2 within the channel domain
remains unchanged in both the soluble and lipid-bound states, which corroborates our
previous findings (26).
Spacing changes among helices 3 – 6 suggest that major
structural rearrangements occur in this region of the channel domain upon membrane
association. Previously, solid-state NMR experiments by Hong and colleagues (77)
demonstrated that the membrane-bound pre-channel state of colicin Ia is considerably
more mobile than the water-soluble structure, and this observation correlates nicely with
the need for enhanced protein mobility to form the open channel.
Our FRET
measurements herein represent only an average measurement of this dynamic structure,
and there is likely more than one substate / conformation for the prechannel structure.
Our goal is to study the dynamic aspects of this structure and to map the relative mobility
78
66
of the prechannel state using time-resolved fluorescence anisotropy measurements.
However, despite the increased mobility of the channel domain upon membrane binding,
we have noted that our FRET data distance distribution function is relatively narrow as
are the distributions of the lifetime decay curves for the donor lifetimes. This indicates
that there is still real, constrained tertiary structure in the colicin membrane-bound
prechannel state and that it has not lost its core structure. Our previous results using
hydrophobic periodicity analysis (24-26, 42) clearly showed that the secondary structure
(helical content) of the prechannel state increases among helices 1 – 5 upon membrane
binding.
Although a wide array of two-dimensional configurations is possible for the
closed-state of the channel domain on the membrane surface, the FRET-derived data in
this study help eliminate a number of possibilities, including the previously published
quasi-circular model of Cramer (29). In summary, our proposed model (Figure 2-7) also
demonstrates the circular arrangement of the helices in the membrane-bound state. New
features of the current model are (a) the clockwise direction of helix arrangement as
viewed from the extracellular surface, (b) the presence of surface H6 – H7 and
transmembrane H8 – H9 α-hairpins, and (c) the formation of a cluster formed by helices
H1, H2, H7 – H6, and H10 that pack around the transmembrane hairpin, H8 – H9, which
acts as a nucleation center. Importantly, our working model represents an important step
forward toward our goal of determining the membrane-bound structure of the closed state
of colicin E1. This model accounts for the previous data (29) as well as our new FRET
data and helps explain why FRET donors near the amino terminus show larger changes in
energy transfer efficiency as compared with the central region of the channel domain
79
67
(23). To further refine this model, our goal is to continue this approach to provide a
greater number of constraints and to improve its overall resolution. In the future, we also
plan to use a similar FRET technique to study the topology of the open channel state in a
helix-by-helix fashion.
However, this would require the presence of a membrane
potential using supported planar bilayers on gold electrodes that are still currently under
development.
80
68
Figure 2-7. Membrane-bound closed state model of the colicin E1 channel domain, top
(A) and side (B) views. Helices are shown by ribbons, multi-colored, hydrocarbon core
boundaries corresponding to the location of lipid carbonyl groups are indicated by gray
dots. Cα atoms of labeled residues used in FRET experiments are shown by numbered
black dots. Distances between labeled residues in the model are indicated by gray
numbers near the black dashes.
81
69
CHAPTER 3
HARMONIC ANALYSIS OF THE FLUORESCENCE
RESPONSE OF BIMANE ADDUCTS OF COLICIN E1 AT
HELICES 6, 7 AND 10
This work has been published:
Ho, Derek., Lugo, Miguel. and Merrill, A. R. (2013) The Journal of Biological Chemistry. 288: 5136-5148
Contributors
Ho, Derek
Lugo, Miguel
Merrill, Rod
Statement of Contribution
Conducted all experiments that includes mutants preparation,
fluorescence labelling and measurements, data analysis and
writing the manuscript
Aided with harmonic data analysis, computer modeling and wrote
part of the manuscript
Supervision and funding of the project
82
70
3.1
Abstract
The pre-channel state of helices 6, 7, and 10 (Val447 – Gly475 and Ile508 – Ile522) of
colicin E1 was investigated by a site-directed fluorescence labeling technique. A total of
44 Cysteine variants were purified and covalently labeled with monobromobimane
fluorescent probe. A variety of fluorescence properties of the bimane fluorophore were
measured for both the soluble and membrane-bound states of the channel peptide,
including the fluorescence emission maximum, fluorescence anisotropy, and membrane
bilayer penetration depth. Using site-directed fluorescence labeling combined with our
novel helical periodicity analysis method, the data revealed that helices 6, 7, and 10 are
separate amphipathic α-helices with a calculated periodicity of T = 3.34 ± 0.08 for helix
6, T = 3.56 ± 0.03 for helix 7, and T = 2.99 ± 0.12 for helix 10 in the soluble state. In the
membrane-bound state, the helical periodicity was determined to be T = 3.00 ± 0.15 for
helix 6, T = 3.68 ± 0.03 for helix 7, and T = 3.47 ± 0.04 for helix 10. Dual fluorescence
quencher analysis showed that both helices 6 and 7 adopt a tilted topology that correlates
well with the analysis based on the fluorescence anisotropy profile. These data provide
further support for the umbrella model of the colicin E1 channel domain.
3.2
Introduction
Colicins are toxic proteins produced by certain strains of Escherichia coli to
provide a survival advantage in a “selfish gene” system (22, 79), and they are often used
in response to metabolic challenges, including DNA damage, catabolite repression, and
nutrient depletion (80). Colicins are a large bacteriocin family that targets susceptible E.
coli and similar bacteria, which do not possess the protective immunity protein (81), by
83
71
acting at a number of levels, including (i) membrane depolarization by ion-conducting
channels (82), (ii) inhibition of protein (83) or peptidoglycan synthesis (84), and (iii)
DNA degradation (85).
Colicins have become a model for study of bacterial protein import (86, 87),
protein folding (29, 88), and membrane insertion (89, 90), and pore formation (91, 92).
The colicin polypeptide can be functionally divided as follows: receptor binding,
translocation, and catalytic / channel domains (11). Colicin E1 has a channel domain that
forms a depolarizing ion channel causing cell death in a host-infected bacterial cell (93).
In order for colicin E1 to enter a target bacterium, the receptor-binding domain must first
bind to the BtuB outer membrane receptor (vitamin B12 receptor) (94). The binding of
the BtuB receptor induces unfolding of the translocation domain, which initiates
migration of the entire protein through the TolC channel and facilitates entry into the
periplasm. This translocation process is also mediated by both the TolA and TolQ inner
membrane proteins. Finally, the channel domain adopts an insertion-competent state in
which it spontaneously inserts into the inner membrane to form the closed channel (17).
The channel then opens in the presence of a trans-negative membrane potential that
allows the escape of various ions from the host cells, such as Na+, K+, and H+, and
subsequently cell death ensues (8).
The crystal structure of the soluble channel domain (95, 96) is composed of 10 αhelices that form an extremely stable, water-soluble globular protein.
The channel
domain is a helical sandwich that is folded into three layers: layer A, the outer layer
composed of H1, H2, and H10; layer B, the inner core layer, including H5, H8, and H9;
layer C, an outer layer composed of H3, H4, H6, and H7 (11). Interestingly, this protein
84
72
also consists of a hydrophobic α-helical hairpin, H8 and H9, which acts as the nonpolar
core of the protein. These two helices are critical to colicin pore formation because they
create a membrane-spanning hairpin upon bilayer association (19). Previous fluorescence
studies suggested that upon translocation across the host cell outer membrane, colicin
adopts an insertion-competent state, which allows the hydrophobic core (H8 and H9) to
penetrate the target membrane. Thus, the protein unfolds, binds, and spontaneously
inserts into the membrane to form the closed channel in a series of kinetically defined
steps (97).
Subsequently, the channel opens in the presence of a trans-negative
membrane potential, and the two channel states exist in rapid equilibrium (21).
Two well known structural models have been proposed for the closed channel
state, which are the penknife and umbrella model. The penknife model was based on
disulfide bond engineering experiments, which suggested that H1 and H2 move away
from the body of the protein, with the remaining helices being deeply buried into the lipid
bilayer (22). In contrast, the umbrella model suggests that only hydrophobic helices, H8
and H9, are inserted into the hydrophobic milieu of the membrane, whereas the remaining
eight helices spread out onto the membrane surface to form an umbrella-like structure. In
fact, the umbrella model was strongly supported by time-resolved fluorescence resonance
energy transfer (FRET) studies on colicin E1 (23). However, the exact orientation of the
helices, their depth of bilayer penetration, and the details of the lipid and protein contacts
still remain unknown. Therefore, the objective of this study was to determine the threedimensional orientation of each helix relative to the lipid membrane in the pre-channel
state.
85
73
Recently, we reported the membrane topology of amphipathic α-helices, H1-H5,
of colicin E1 in its pre-channel state, and it was found that all five N-terminal helices
retain their α-helical structure, with H3 and H5 elongating upon membrane association.
Herein we continue our investigation of the membrane-bound topology of the remaining
helices within the pre-channel state of colicin E1.
Using site-directed fluorescence
labeling combined with our novel helical periodicity analysis method, we found that H6
and H7 are two separate and distinct amphipathic α-helices in the membrane-associated,
pre-channel state and that H6 becomes overwound as a 310 helix upon membrane binding.
In contrast, H10 is a 310 helix in the soluble colicin protein but opens up into a standard
α-helix upon membrane binding.
3.3
Materials and Methods
3.3.1 Materials
All chemicals, unless otherwise stated, were purchased from Sigma. All steadystate fluoresecence measurements were conducted with a PTI-Alphascan-2-spectrofluorimeter (Photon Technologies Inc.) equipped with a thermostatted cell holder, and
data were reported as the mean ± S.D. and measurements were performed at least in
triplicate.
3.3.2 Mutagenesis, Protein Purification, and Monobromobimane Labeling
Each residue from Val447 to Gly475 and from Ile508 to Ile522 of P190H6 was
individually replaced with a cysteine using the Stratagene (La Jolla, Ca) QuikChangeTM
86
74
mutagenesis kit. Plasmid DNA was purified using the High Pure Plasmid Isolation kit
from Roche Applied Science. Both the wild-type (WT) P190H6 and Cys variants were
prepared from transformed lexA- E. coli IT3661 cells as described previously (87).
Protein purity was assessed by SDS-PAGE, and protein concentration was determined by
spectroscopy at A280, using an extinction coefficient (ε) of 29,910 M
–1
cm
–1
(98).
Purified Cys variants were labeled with monobromobimane (mBBr) with a molecular
weight of 271.11 g mol-1 (Molecular Probes Inc.) at a 20:1 molar ratio (probe/protein),
and the labeling efficiency was determined as described previously (24).
3.3.3 Preparation of Large Unilamellar Vesicles (LUVs)
LUVs were prepared from 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC)
and 1,2-dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] (DOPG) vesicles at a 60:40
molar ratio (Avanti Polar Lipids). Lipids were prepared and quantified as previously
described (24) except the buffer used to suspend vesicles consisted of 10 mM NaCl (pH
4.0). Asolectin (Fluka) was purified according to the method of Schendel and Reid (99),
and vesicles were prepared as described previously (35). Phospholipid concentration was
determined using the microBartlett assay (24).
3.3.4 6-Methoxy-N-(3-sulfopropyl)quinolinium Assay for in Vitro Channel
Activity
This assay was performed using a Cary Eclipse spectrofluorimeter (Varian
Instruments). Asolectin vesicles were loaded with 16 mM SPQ in 10 mM DMG, 100
87
75
mM KCl, 1 mM CaCl2 buffer, pH 5.0, using a freeze-thaw method (35), and the vesicles
were diluted to 0.3 mg/mL in 20 mM DMG, 100 mM NaNO3 buffer, at the desired pH.
The vesicle fluorescence was monitored continuously for 2 min at 20 °C with constant
stirring, after which protein was added (2 ng/mL, final concentration). Fluorescence
measurements were taken using excitation and emission wavelengths of 347 and 445 nm,
respectively.
The spectral bandwidth for both wavelengths was set to 5 nm.
The
extravesicular buffer was 100 mM NaNO3, 10 mM DMG, pH 6.0, and the encapsulation
buffer was 100 mM KCl, 10 mM DMG, 1 mM CaCl2, at pH 5.0. The total amount of
encapsulated Cl- was released by the addition of Triton X-100 (0.1%, final concentration)
to the sample. The fluorescence changes upon the addition of the protein were reported
as the percentage of maximum SPQ fluorescence (% Fmax): % Fmax (t) = (F – Fb) / (FT –
Fb) x 100%, where Fb is the initial residual fluorescence of the dye-loaded vesicles and FT
is the maximal fluorescence intensity after detergent lysis of the vesicles.
3.3.5 Steady-State Intrinsic Tryptophan Fluorescence
The folded integrity of all proteins was examined using the intrinsic Trp
fluorescence as previously described (24). Both the native and bimane-labeled variants
as well as the WT P190H6 were diluted to 4 μM in PBS (50 mM NaH2PO4, 50 mM
Na2HPO4, and 100 mM NaCl, pH 7.0).
Intrinsic fluorescence was generated by
excitation of Trp residues at 295 nm (2 nm excitation slit width), and emission was
detected from 305 to 450 nm (4 nm slit width). The resulting traces were corrected for
the buffer and wavelength-dependent bias of the emission components of the
88
76
spectrofluorimeter before calculation of the λ emission maximum (λem,max) from the first
derivative of the smoothed spectra.
3.3.6 Bimane Fluorescence Emission Spectra
The steady-state bimane fluorescence emission spectra of all Cys variants were
measured as described previously (24). All bimane labeled variants were diluted to 4 μM
in DMB buffer (20 mM DMG and 130 mM NaCl (pH 4.0)) in the presence or absence of
excess LUVs. The data were corrected for the buffer and wavelength-dependent bias of
the equipment before calculation of the λem,max from the first derivative of the smoothed
spectra.
3.3.7 Steady-State Bimane Fluorescence Anisotropy
The steady-state fluorescence anisotropy (r) measurements were made using “Tformat” detection by simultaneously comparing the intensities of the vertically (IVV) and
horizontally (IVH) polarized emitted light as described previously (24). Using the IVV and
IVH fluorescence intensities, the anisotropy (r) was calculated as follow:
r = ( IVV – GIVH ) / ( IVV + 2GIVH )
[Eq. 5]
The ‘G’ instrumental factor, measured as IHV / IVV was determined from the intensities of
the vertically (IVV) and horizontally (IHV) polarized emitted light from the horizontally
polarized excitation light. All bimane labelled mutant proteins were diluted to 8 μM in
DMG buffer (20 mM DMG and 130 mM NaCl, pH 4.0) in the presence or absence of
89
77
excess LUVs. The excitation wavelength was set at 381 nm (4 nm slit-width), and
emission was collected at 470 nm (10 nm slit-width) with a signal integration time of 30
seconds. A solvent blank (DMG buffer or LUVs in DMG buffer) was subtracted from
each intensity reading prior to the calculation of the anisotropy value. Probe mobility
values were determined based on the inverse of the measured anisotropy values.
3.3.8 Dual Quenching Analysis
The objective of the dual quenching assay was to determine the relative
penetration depth of each colicin residue within the lipid membrane. The concept behind
this assay is to utilize two different types of quenchers, an aqueous and membrane
embedded quencher, to measure the extent of bimane fluorescence quenching for each
residue. The depth measurement was determined by measuring the quenching of bimane
in the presence of either iodide (KI) (aqueous quencher) or 10-doxylnona-decane (10DN) (membrane embedded quencher) as described previously (32). The relative ratio of
quenching (Q-ratio) caused by the two quenchers provides a reliable measure of the
penetration depth of each residue into the lipid membrane according to the following
equation:
Q-ratio = [(F0 / F10-DN) – 1] / [(F0 / FKI) – 1]
[Eq. 6]
where F0 is the fluorescence intensity of the sample without the addition of either
quencher, and FKI and F10-DN are the fluorescence intensities with the addition of KI and
10-DN, respectively.
90
78
To measure iodide quenching (FKI ), the sample fluorescence was measured in
ratio mode using semi-micro quartz cuvettes (0.5 cm x 0.5 cm) containing 100 μM LUVs
and 7.5 μg of protein or LUVs only (background). FKI was determined after the addition
of a 50 μL aliquot of a 1.7 M KI in 0.85 mM Na2S2O3 stock solution. To measure 10-DN
quenching, LUVs were prepared as above except that the LUVs were doped with 10 mol
% 10-DN. All samples were incubated at 24°C for 30 minutes before the measurement of
initial fluorescence. For all measurements, the excitation wavelength was set at 375 nm
with the emission intensity observed at 467 nm (2.5 and 5 nm for excitation and emission
bandpasses, respectively).
3.3.9 Prediction of Secondary Structure from Fluorescence Parameters
The secondary structure elements were predicted from the observed fluorescence
parameters using a method adopted from Cornette et al. (101). The electronic center of a
probe attached to the protein fluctuates with a trajectory confined in a cone of defined
angle and will probably be experiencing an average dielectric constant, refractive index,
electric field, dipole moment, etc., according to the average local probe environment
during the lifetime of its excited state. For a sequentially labeled helix, any fluorescent
property exhibited by the probe will follow an amplitude and angular frequency described
by a harmonic wave function.
Now, considering both radial and longitudinal
heterogeneity in the microenvironment along the length of the helix, we propose the
following empirical function,
Y(n) = Y(n0) + A(n)  sin [2π (n–n0) - φ/ T] + m  (n–n0)
91
79
[Eq. 7]
which corresponds to an amplitude-modulated, A(n), harmonic wave form of a generic
fluorescent property Y, with period T (in residues per turn (r.p.t)) and phase φ. This, in
terms of the position of the residue n, for a given initial residue n0 (horizontal offset)
would calculate the Y(n0) value (vertical offset).
The amplitude of the sinusoidal function is determined by the exponential profile,
A(n) = Am  ( 1 – exp – (n – n0) / τ )
With maximum amplitude, Am , and length constant, τ .
[Eq. 8]
The fluorescent property
considered for this analysis was the spectral centroid (SC) of the emission spectrum in
wavelength mode, defined by the equation,
Y = SC = (ʃ F(λ) d λ) / (ʃ d λ)
[Eq. 9]
which reflects the center of mass of the whole emission band. The relevant parameters in
Equation 7 and 8 (T, Am , and m) were obtained by weighted nonlinear least squares
regression of the experimental data from Equation 9, using the graphical and statistical
package OriginPro 8.0 (OriginLab Corp.).
3.3.10 Data Fitting
The setting of the initial parameters for the nonlinear fitting of Equation 7 and 8,
based on the calculated data with Equation 9, was performed as follows. (i) The initial
residue n0 to consider (horizontal offset) was based on the helical boundary reported in
the crystal/soluble structure. (ii) The parameter Y (vertical offset) corresponded to the
spectral center of mass at n0, SC (n0). It is an experimental value and was considered
fixed for the most of the fitting. (iii) The setting of the slope m, was obtained according
92
80
to the linear trend observed for the envelope of the harmonic function.
(iv) The
maximum amplitude Am was usually a fixed parameter, set to half of the maximum 
value observed. (v) The parameters φ and τ are correlated with each other; thus, in
general, the phase was set fixed to either φ = 0 or π/2, depending on whether the SC for
the second residue (n0 + 1) is higher or lower than the first residue (n0), respectively.
For the data fitting, the weight of each value was kept uniform except for the last
residue, which was set to a value of 1/100 of the regular weighting. The rationale was to
intentionally reduce the contribution of the last residue in the fitted result but then to
evaluate the magnitude of the residuals for the basis of the assessment of the helical
boundaries.
Soluble Form – For helix 6, the beginning of the helix was set at n0 = 447.
Inspecting the data, the phase was set to φ = π/2, and the slope was set to m = 0. Thus,
minimizing the χ2 of the fitting, the period converged at T = 3.34 ± 0.08 (R2 = 0.76) and
Am = 8.26, with residuals distributed uniformly with relative low dispersion (± 0.5%).
For helix 7, the initial residue was set at n0 = 461 and phase φ = 0. The data for residue
463 were excluded from the analysis because it appeared as a clear outlier of the regular
harmonic function while residue 461 – 462 fits well with the analysis. Thus, fixing the
slope at m = - 0.41 according to the trend observed, the period converged to T = 3.56 ±
0.03 (R2 = 0.93) and Am = 9.96, regardless of the initial values of the free-floating
parameters T, τ , and Am. Helix 10 was considered to initiate at n0 = 514, and, according
to the plot, the phase was set to φ = 0. Again, with m < 0 and with free-floating T, τ, and
Am, the period converged to T = 2.95 ± 0.03 (R2 = 0.81), regardless of the initial values of
the floating parameter, although the maximum amplitude Am reported was significantly
93
81
elevated (Am = 21.4). As an alternative fitting strategy, considering the data at residue
519 as an outlier (actually, by assigning it as a relative weight of 1/100), with freefloating m and Am, the fitting yielded T = 2.99 ± 0.12 (R2 = 0.85) and m = 0 ± 0.19 (a zero
slope), but now with Am = 11.45 ± 1.95. Nevertheless, in both cases, the estimated
periods were closer to each other.
Membrane-associated Form – In the membrane-bound state, the harmonics
behaved in a simpler fashion that forced Equation 8 to reduce to A(n) = Am . For helix 6,
from an initial inspection of the output plot, the parameters were set to m = 0.15 and Am =
10. Thus, starting with the residue n0 = 448, the fitting converged to T = 3.00 ± 0.15 (R2
= 0.90) and φ = 0, with a low residual distribution of ± 1.4%. Helix 7 exhibited a slope of
m = 0. Thus, setting the initial parameters to Am = 10 and φ = π/4, both for fixed or freefloating, and starting at n0 = 461, the fitting converged to T = 3.68 ± 0.03 (R2 = 0.93) and
Am = 10.6 (for the free-floating case). For helix 10, the beginning of the helix was set at
n0 = 513, the phase was set at φ = π/2, and the slope was set at m = 1.67. The fitting
converged to T = 3.47 ± 0.04 (R2 = 0.95) and Am = 11.04 ± 0.9, with residuals lower than
± 1.0%.
Accordingly, for all three helices in both the soluble and membrane states, the
maximum amplitudes were in the range of Am ~ 10 ± 2.0 nm. This number may represent
a constant value that reflects the maximum variation of the observed variable (center of
mass) between the totally buried and the fully exposed probe, under the experimental
conditions of the measurements.
94
82
3.3.11 In Silico Analysis of the Soluble Structure
The modeler and simulator suite, Molecular Operative System, MOE 2011.10
(Computing Chemical Group, Inc.) was used for all the in silico analyses, and the x-ray
crystal structure of the soluble colicin E1 channel domain (Protein Data Bank entry 2I88)
was used as the template structure. This structure was protonated by the MOE module,
Protonated 3D (102), at pH 6.0, 300 K, and 0.1 M ionic strength and was energyminimized using the OPLS-AA (optimized potential for liquid simulation for all atoms)
force field in a generalized Born/integral volume implicit solvent model approach (103)
(with Coulombic electrostatic of ϵint = 2 and ϵext = 80), in order to eliminate low energy
clashes.
The Solvent Accessible Surface Area (SASA) (in Å) of each residue in the ColE1
channel peptide was calculated after rolling a spherical probe with 1.4-Å radus over the
residue surface. This value was standardized by the total exposed SASA, calculated in
the context of the tripeptide Gly-X-Gly (104), with X the residue in consideration. The
local period, Ti, to evaluate the helical content of the backbone was assessed by
measuring the angle of rotation, θI, because, for a perfect helix, nitrogen lies on a cylinder
with θ  100°, resulting in a constant period of T  360/100  3.6 r.p.t. (101).
95
83
3.4
Results
3.4.1 Mutagenesis, Protein Purification and mBBr Labeling
The colicin E1 structure consists of 10 α-helices, where H8 and H9 (colored
black) are the hydrophobic helices that form the core of the soluble protein (Fig 3-1A).
The spatial relationship of H6, H7, and H10 (shaded in gray) relative to the entire
channel peptide is shown. There is only one natural Cys residue (Cys505) deeply buried
within the hydrophobic core of the protein. Normally, it does not react with any thiolspecific reagents under non-denaturing conditions. The sequence under investigation in
this study was composed of H6, H7, and H10 and includes Val447 – Gly475 and Ile508 –
Ile522 (Fig 3-1B). The sequences underlined in black were subjected to cysteine-scanning
mutagenesis.
A total of 43 Cys variants were prepared, and mutation sites were
confirmed by DNA sequencing.
96
84
Figure 3-1. Preparation of colicin E1 Cys variants. (A) Ribbon topology of the 2.5 Å
crystal structure of the P190 peptide (Protein Data Bank entry 2I88). The overall
structure consists of 10 α-helices, where H8 and H9 (shown in black) are the hydrophobic
helices that serve as a membrane-anchoring helical hairpin in the membrane-associated
state. Gray residues (shown in sticks) were subjected to replacement with Cys codon.
(B) The primary sequence and secondary structure of the channel-forming domain of the
colicin E1 (P190H6). Residues underlined with a boldface line were subjected to Cys
substitution.
97
85
All Cys variants exhibited WT expression levels except for F464C, which showed
almost no expression. This indicates that Phe464 might play an important role in the
folded structure of the ColE1 channel domain (CD).
As a result, this variant was
eliminated from this study. All Cys variants were purified as described previously (105),
and purity was assessed by SDS-PAGE analysis showing >95% homogeneity.
In order to report on the local environment of the residues, each Cys variants was
subjected to mBBr labeling as described previously (42). Monobromobimane was used
for labeling in this particular study because it is a well characterized, relatively small,
non-perturbing fluorophore that is essentially non-fluorescent until conjugated to the
protein of interest through disulfide linkage with a cysteine side chain (Fig 3-2A).
Therefore, it is an ideal fluorophore for the purpose of spectroscopic studies.
The mBBr labeling efficiency, which represents the fraction of the variants being
labeled (no label, 0%; complete labeling, 100%) is shown in Fig 3-2B. Notably, only 8
of the 43 Cys variants exhibited labeling efficiency of 40% or less. In order to improve
the labeling efficiency, one option is to slightly denature the variant with 4 M urea prior
to labeling followed by removing the urea to refold the variant. However, the only
problem with this method is that Cys-505 may react with the thiol-specific probe during
this procedure. To resolve the issue, an additioinal eight variants were prepared using the
C505A mutation to facilitate the preparation of single-Cys variants. It was previously
reported that the C505A mutation does not perturb the folded structure and does not
impair channel activity (26). Labeling efficiency was significantly improved with the
incorporation of the C505A mutation with the above approach except for the I454C
98
86
variant (Fig 3-2B). In general, labeling efficiency of most variants ranged between 60
and 100%.
99
87
Figure 3-2. Labeling of Cys ColE1 variants with mBBr. (A) Left, energy-minimized
structure of the colicin E1 adduct 467BIM, labeled in silico based on the x-ray crystal
structure. The helices were colored according to the position in the domain: from blue
(N-terminal) to red (C-terminal). Right, chemical structure of mBBr. (B) Labeling
efficiency measurement of the 43 Cys ColE1 variants under normal condition (hollow
bar). The 8 variants subjected to C505A mutation under denaturing condition with
different labelling efficiency was shown in dark bars. Every other residue (odd numbers)
is shown on the abscissa. Missing data for F464C are due to low expression level of this
variant.
100
88
3.4.2 Structural and Functional Analysis of Cys Variants
Prior to spectroscopic measurements of the variants, it is important to assess both
the structural and functional integrity of the proteins. Herein, the Trp fluorescence
emission maximum values (Trp λem,max) were measured to provide a measure of the
average local environment of the Trp residues and, hence, the folded integrity of the
proteins. A red shift in the Trp λem,max values would be expected for mutation-induced
alteration of the folded integrity. As shown in Fig 3-3A, the mutants showed a Trp
λem,max values similar to the WT protein except for a few variants, including T457C,
W460C, A472C, G475C, L513C, and L520C. According to the crystal structure, these
residues are all located at the boundaries of H6, H7, and H10. It is likely that these
residues provide critical non-covalent interactions between individual helices that help to
maintain proper folding of the channel protein. Alternate hypotheses might be that (i) the
original residues are contributing to the hydrophobic environment, or (ii) the replacement
Cys residues are interacting with Trp emitters. Nonetheless, Cys substitution at these
sites induces only a minor perturbation to the folding integrity of the protein.
In order to assess the effect of Cys replacement on the channel activity of ColE1
channel domain, an SPQ activity assay was performed to measure the in vitro channel
passage efficiency of the protein as described previously (23). As shown in Fig. 3-3B,
most Cys variants showed channel activity ranging from 80 to 120% of the WT level,
with the lowest value (A472C) at 75% of the WT. Thus, correlating with the data shown
in Fig 3-3A, it would seem that Ala472, Thr457, and Leu513 may be important residues for
maintaining channel domain (CD) folded integrity, and thus Cys replacement causes a
101
89
minor reduction in channel activity. In summary, Cys substitution of these residues did
not significantly impair channel activity of the variants. Also, modification of the Cys
side chain with bimane did not impair channel function (Fig 3-3B).
102
90
Figure 3-3. Structural and functional tests of Cys ColE1 variants. (A) Trp λem,max (nm)
values were measured and compared against the WT ColE1 to assess the folding integrity
of the variants. (B) The channel-forming activity of the variants with bimane
modification (dark bars; representative variants shown) and without bimane (hollow
bars) was assessed by measuring the flow rate of SPQ (fluorescent dye) through the
channel in LUVs. Data are missing for F464C due to the low expression level of this
variant. Error bars represent +/- S.D. of samples in triplicate.
103
91
3.4.3 Bimane Fluorescence Emission Maxima of H6, H7, and H10 Cys Variants
In order to determine the relative local environment of the bimane fluorophore
within the protein in terms of its polarity, the bimane λem,max was measured for each Cys
variant. In theory, more deeply buried residues are expected to have a lower bimane
λem,max (blue-shifted) than solvent-exposed residues (red-shifted) in the membrane-bound
state. To calibrate the bimane λem,max values, a standard polarity curve was generated
previously in our laboratory (24) using a bimane-N-acetyl-Cys model compound in a
series of dioxane/water mixtures with known dielectric constant (є) values. Previously
determined parameters (λem,max < 455 nm = buried; 455 nm < λem,max < 470 nm =
moderately accessible; λem,max > 470 nm = solvent accessible) were used to calibrate the
bimane local environment.
The bimane λem,max data suggest that H6, H7, and H10 are amphipathic in nature,
and the polar faces of these helices contain residues that are highly solvent-accessible
(Fig 3-4A). Importantly, the data correlate well with the crystal structure; residues with
high bimane λem,max values are also the hydrophilic residues that are exposed to the
solvent side and vice versa. By visual inspection, the data suggest that both helices 6 and
7 are α-helices with three alpha cycles, whereas H10 is an α-helix with two alpha cycles.
Interestingly, there was no significant difference in the helical lengths between the
soluble and lipid-bound states, which suggested that H6, H7, and H10 retain their helical
pattern in the lipid-bound state (Fig 3-4A).
104
92
3.4.4 Bimane Fluorescence Anisotropy and Probe Mobility of H6, H7, and H10
Cys Variants
In order to determine the bimane fluorescence anisotropy of the individual residue
side chains within H6, H7, and H10, this can be calculated by measuring the inverse of
the probe mobility. In theory, surface-exposed residues are expected to have lower
anisotropy (higher probe mobility) than buried residues that are facing the interior of the
protein or membrane bilayer due to differences in the viscosities of the membrane bilayer
compared with the aqueous medium. As shown in Fig 3-4B, most Cys variants showed
lower probe mobility (1/r) values in the membrane-bound state, and this supports the idea
that the side chain mobility is often reduced upon lipid association.
Based on the probe mobility data (Fig 3-4B), it appears that the residues within
H10 in the soluble protein have higher mobility than the residues in either H6 or H7,
which is not surprising because it is the last helix of the channel protein that results in
more freedom. Notably, residues within H6 appear to have higher mobility than H7 in
the soluble ColE1 channel domain (CD). The bimane λem,max and probe mobility data are
well correlated in terms of both the helical periodicity and pattern. Residues that are
highly solvent-accessible tend to have higher mobility, and vice versa.
105
93
Figure 3-4. Fluorescence emission maximum and probe mobility of the bimane-labeled
Cys variants of the colicin E1 channel domain. (A) The bimane λem,max values of the Cys
variants in the soluble () and membrane-associated state () were measured. (B)
Probe mobility (1/r) was calculated as the inverse of the observed fluorescence
anisotropy (r). Absolute probe mobility (1/r) of the bimane-labeled Cys variants in both
thesoluble () and membrane-associated state () were measured. Average and S.D.
(error bars) values for at least triplicate measurements are shown.
106
94
3.4.5 Dual Quenching Analysis of the Membrane-bound Depth of H6, H7, and
H10 Cys Variants
In order to determine the relative membrane penetration depth of each residue
within these three helices in their lipid-bound state, a dual fluorescence quenching
method was used as described previously (106). Two types of quencher species were
employed in this assay (KI, aqueous quencher; 10-DN, membrane-embedded quencher).
In theory, surface-exposed residues in the lipid-bound state are expected to be more
quenched by KI, resulting in higher values of (F0/FKI – 1), where F0 and FKI represent the
fluorescence intensity of bimane in the absence and presence of KI, respectively. In
contrast, residues that are buried in the lipid-bound state are expected to be more
quenched by 10-DN, resulting in higher values of ((F0/F10-DN ) – 1), as shown in Fig 3-5.
In order to calculate the relative membrane penetration depth, the quenching ratio (Qratio; ((F0/F10-DN) – 1/ (F0/FKI) – 1) was calculated. It should be noted that H8 and H9
form the hydrophobic hairpin within the membrane; therefore, it is not surprising that H7
adopts a tilted topology with a more buried C-terminus because it is connected to the
anchor domain.
107
95
Figure 3-5. Plot of the quenching ratio. The Q-ratio represents the ratio of quenching by
10-DN to that by KI (see “Experimental Procedures” and Equation 6), and it correlates to
the penetration depth of the residues within the bilayer. Average and S.D. (error bars)
values from triplicate measurements are shown.
108
96
3.4.6 Harmonic Analysis
The initial fluorescence emission results support the amphipathic character of H6,
H7, and H10. However, a more robust quantitative analysis is required for helical
characterization. In this sense, a harmonic analysis was performed using Equations 7 9. The strategy employed for the fitting was to start with an initial value of T = 3.6 r.p.t.
(regular α-helix) and to include a range of residues in the analysis according to the
crystal structure (for details, see below).
Soluble Form (Fig 3-6A) – For H6, the fitting for residues 447-457 gave a helical
periodicity of T = 3.34 ± 0.08 r.p.t. According to the residue dispersion, residue 457 was
included as part of H6. For H7, the fitting was residues 461 – 475 converged to T = 3.56
± 0.03 r.p.t., with residues 474 and 475 clearly being not a part of the periodicity. H10,
when fitted over residues 514 – 522, gave a smaller periodicity of T = 2.99 ± 0.12 r.p.t.,
and clearly residues 522 did not follow the harmonic function. For all three helices, the
maximum amplitudes used/reported were Am = 10 ± 1.5 nm, which is consistent with the
same range of values either for exposed or buried probe locations.
On the other hand, the helical content of H6, H7, and H10 was quantified by the
average angle of rotation and the derived periodicity (Fig 3-6B) for all but the end
residues belonging to the helices. Effectively, the backbone of H6 presented a shorter
average period (T = 3.48 r.p.t.) than for an ideal α-helix, with irregular dispersion
around the medium value and a particularly low outlier seen for residue 448. On the
contrary, for H7, the average value is nearly optimal (T = 3.58 r.p.t.) and quite regular
109
97
except for residues 473 and 474. Curiously, H10 shows typical helix properties for the
backbone but shows extreme values for both terminal residues (514 and 521).
110
98
Figure 3-6. (A) Harmonic analysis of the SC for the soluble state of the colicin E1
channel domain. Connected symbols, calculated SC (Equation 9) for the emission
spectrum of the bimane adduct of the channel domain as a function of residue position.
Solid lines, the best fit according to the functions in Equation 7 and 8. The signal at
position 519 (boxed symbol) was considered an outlier for the fitting. (B) The local
period of the helical segments according to the measured angle of rotation, as defined
under “Experimental procedures,” for the soluble structure. The solid horizontal line
was set at T = 3.6 for reference. (C) Correlation SASA-mobility for the soluble form.
Solid symbols and solid line, calculated SASA for the original side chains, standardized
according to the value for the residue totally exposed and then normalized by the highest
value in the helix. Empty symbols and dashed line, normalized mobility of the bimane
adduct, normalized by the highest value in the helix.
111
99
Anisotropy of the ColE1 CD Bimane Adducts – Protein adducts were analyzed
according to the mobility (M = 1/r), using the same harmonic function as described
above, with limited success either for the soluble or the membrane-bound form. Despite
the amplitude signal differences between high and low values, there was no simple
periodic waveform detected.
However, certain trends were evident in the overall
pattern. For instance, the maximum mobility for H6 in the soluble form followed an
exponential waveform. In the case of the soluble H7, the behavior is opposite to that
observed for H6. Nevertheless, the mobility observed by the bimane moiety is in good
agreement with the calculated solvent accessibility for the original residue depicted in
the correlation shown in Fig 3-6C for the standardized signals.
Membrane-associated Form (Fig 3-7A)
- In the membrane-bound state, the
fitting clearly converged to T = 3.00 ± 0.15 r.p.t. for helix 6, and according to the
distribution of residuals, neither residue 447 nor 457 should be part of the helix. For H7,
the fitting yielded T = 3.70 ± 0.1 r.p.t., and the analysis of the residuals led to extension
of H7 to residue 473. In the case of H10, working in the range from residue 513 to 522,
the fitting converged to T = 3.47 ± 0.04 r.p.t., and the helix can be extended to residue
522.
112
100
Figure 3-7. (A) Harmonic analysis of the SC for colicin E1 for the membrane-bound
state. Connected symbols, calculated SC (Equation 9) for the emission spectrum of the
bimane adduct of colicin E1 at the indicated residue. Solid lines, best fit according to the
functions in Equations 7 and 8. (B) Schematic diagram illustrating the predicted
penetration depth based on the Q-ratio data and the harmonic analysis. The penetration
depth was calibrated using dielectric constants measured from the bimane λem,max data.
113
101
3.5
Discussion
In this study, we have successfully scanned the membrane-bound topology of H6,
H7, and H10 and showed that they are also amphipathic α-helices that are located on the
membrane surface which provide further support for the umbrella model of colicin E1 in
the pre-channel membrane-bound state. The calculated helical periodicity in this study
revealed that H6, H7, and H10 are three separate amphipathic α-helices. These data help
to refute earlier models, which proposed that both H6 and H7 adopt a transmembrane
orientation to assist with channel formation in the open channel state (95). In fact, H1 –
H5 of the colicin E1 channel domain were previously demonstrated to exist as
amphipathic α-helices upon membrane association by a similar method (19, 21, 23, 97).
In this study, all of the Cys ColE1 variants were highly purified, and the bimane
labeling efficiency was close to 80% for most variants (Fig 3-2). Both structural and
functional tests revealed that the Cys substitution did not negatively impact either the
folded integrity or channel activity of the ColE1 channel domain (CD) (Fig 3-3). In this
study, both the Q-ratio and probe mobility data strongly indicate that H7 adopts a tilted
topology on the membrane surface with a more buried C terminus. This appears to
accommodate both H8 and H9 to adopt an anchor structure in the pre-channel state,
which lends further credence to the umbrella model of the pre-channel state for colicin E1
(Fig 3-8). In addition, because H6 exhibits a more buried N terminus in this study, these
results also correlate well with data from our previous study, which revealed that the H5
loop region is deeply embedded within the membrane bilayer in the pre-channel state
(23).
114
102
As mentioned above, the use of the SC (Fig 3-6A and 3-7A) offered more uniform
and consistent results because this variable summarized the overall spectra, including the
position of the emission maximum, the bandwidth at half-maximum, and any skewness,
being a more robust reporter of the environment around the probe than any independent
property by itself. Effectively, a plot of the wavelength at the maximum (Fig 3-4A)
reported for the bimane probe was noisier and less periodic than the SC. Furthermore, the
calculation of SC revealed consistent differences between the soluble and membraneassociated states in regard to the bimane fluorescent properties.
In addition, the current harmonic function employed herein offered a more
powerful fitting capability than previously employed, with parameters that hold physical
meaning.
For an ideal mixed solvent system (i.e. membrane-water interface), a
homogenous value for any physical parameter (i.e. the average polarity, P) in the
direction parallel to the interface (xy-plane) or for a vertical gradient in either direction
from z = 0 (z-gradient) is expected. Thus, for a helix embedded at the interface for this
system, the angular distribution of the fluorescence should be independent of the position
of the probe in the helix, unless the principal inertial helix axis presents a tilt angle with
respect to the plane of the membrane. The parameter m in the harmonic function can
account for this topology. Now, for a soluble protein, it is expected that the environment
around a particular helix will be more heterogeneous because there is expected to be
much more variability in the composition and distance according to the sequence and
structure of the protein.
According to the solution structure of the ColE1 channel domain, H6 shows some
variability, whereas H7 is nearly a perfect α-helix (Fig 3-6, A and B). Incidentally, both
115
103
helices were similarly predicted by the harmonic analysis. In contrast, H10 appears as an
overwound helix, as predicted by the harmonic analysis with T  3.0 r.p.t. (i.e. a 310 –
helix). The origin of this discrepancy might be addressed by studying the particular
location and interaction of the probe using molecular modeling approaches.
It is
noteworthy that only H7 required both parameters (m and τ) to account for the scattering
of the fitted data. This is consistent with its location in the folded structure. This helix
goes from the outside to the inside of the protein core, consistent with the negative slope
of the SC.
For the membrane-associated state of the ColE1 channel domain, there is no high
resolution structural model to correlate with the output from the harmonic analysis.
Nevertheless, the current thinking is that the surface location for H6, H7, and H10, and
consequently the bimane probe will experience a more homogenous environment than in
the globular soluble form in regard to the angular and longitudinal dimensions. This
statement was consistent with the fact that all of the helices lacked the damping behavior
(τ  0, A(n) = Am), and, contrary to the general trend for the soluble protein, the
parameter m could account for longitudinal variation. Thus, according to this analysis,
the inertial axis of H6 and H10 seem to be tilted with a positive angle (leaving the
membrane), whereas H7 is horizontal with the plane of the membrane. Upon membrane
binding, H6 exhibited 310 –helix character (T = 3.0 r.p.t.), whereas H7 and H10 exhibited
more relaxed structures than in the soluble form. The anomalies for H6 are also evident
in its shorter length by two residues. On the contrary, H7 and H10 apparently are
extended by one residue each.
116
104
The harmonic analysis for the probe mobility is complex and might require the
combination of multiple sinusoidal components (one of them may be the periodicity of
the helix, T) to account for the data scatter. The caveat here is the lack of physical
meaning ascribed to such new fitting components. However, the overall change of this
variable must be associated with the topology of the helix. Indeed, the increased mobility
is consistent with the probe environment, for example when the probe approaches the H6H7 loop (C terminus of H6) and, conversely, decreased mobility when the probe
approaches the N terminus of H7, exhibiting in both cases a complementary funnel-like
shape. The low mobility shown by the linear trends in both cases probably describes the
restricted environment for the probe inside the core of the protein. In contrast, when the
probe is facing the external surface or is found within the loop region, the mobility is
much higher. Thus, the degree of side chain solvent exposure for the original residues
correlates well with the mobility of the bimane probe (Fig 3-7A). However, some
discrepancies that do exist reflect site-dependent interactions that might limit the range of
rotation of the probe. This is the case for Ser449. On the contrary, the null calculated
exposure of Val473 and Ala474 predict restricted mobility. The relatively high mobility for
these residues provides indirect evidence of local structural changes reported by the probe
in relation to the original side chains.
In general, the fluorescence data in this study correlated well with the crystal
structure of the soluble ColE1 peptide (11, 92), as deduced from the overall comparison
between the harmonic analysis of the signal and the periodicity of the backbone. These
data also prove invaluable toward the construction of an improved membrane topology
model of the channel domain in the pre-channel state (Fig 3-8). Our revised model
117
105
further demonstrates the circular arrangement of the helices in the membrane-bound state
of ColE1. However, one cautionary note is that the three-dimensional orientation of the
channel is still unclear because a wide range of feasible three-dimensional configurations
is possible on the membrane surface. To deduce the various possible three-dimensional
orientation models, our laboratory will apply a rigorous and high density FRET analysis
approach to map the interhelical distances within the ColE1 CD. This approach uses
three Trp donors with the Cys-bimane as the acceptor fluorophore, and this method
should provide new data for the construction of a low resolution three-dimensional model
of the pre-channel state of the ColE1 CD (35). This method will also provide information
toward the possible oligomeric state of the colicin E1 channel in the membrane because it
was previously demonstrated that more than one molecule may participate in the channel
structure of colicin Ia (100).
118
106
Figure 3-8. Models of colicin E1 in the membrane-bound state. The models were
adapted from Ref. 105, where the tilt angles and depth of H6 and H7 were modified while
maintaining the membrane context of the connecting coil regions. The dots represent the
raw data for lipid carbonyl groups of the outer leaflet of the plasma membrane. (A) upper
view. (B) and (C) Side views with a 90° rotation difference between B and C.
119
107
Chapter 4 Resolving the 3-D Spatial Orientation
of Helix 1 of Colicin E1 Channel Domain by
FRET
4.1
Abstract
The three naturally occurring Trp residues (W424, W460, W495) located within
the colicin E1 channel domain presents a powerful triangulation tool that can, in theory,
probes the 3-D spatial orientation of the entire channel domain.
In this study, the
membrane topology of helix 1 of the channel domain was spatially mapped by FRET
using the three Trp residues as the triangulation tool. Initially, three residues (D347,
S354, E361) within Helix 1 were selected for the construction of nine single-Trp singleCys variants to probe for the 3-D orientation of Helix 1 relative to the three Trp residues
in both the soluble and lipid-bound states. Additionally, three single-Trp mutants were
prepared for the purpose of measuring the inter-molecular distances between the three
Trp residues by FRET in order to validate the triangulation frame foundation of the 3-D
model. In theory, this approach can be used to probe the 3-D spatial orientation of the
remaining helices. This powerful approach would allow the construction of a reliable 3D membrane topology structure of the entire channel domain in the absence of both Xray crystallography and NMR data.
In this study, the FRET distances for the soluble
state were shown to have a high correlation with the X-ray data. In general, the selfconsistency of the model in terms of matching the experimental and modelled distances
makes the new closed-state model a robust proposal that is compatible with the
architecture of the soluble-state model.
120
108
4.2
Introduction
Colicins are antimicrobial proteins produced by Escherichia coli that target
susceptible bacteria in response to stressful conditions including: nutrient depletion, DNA
damage, overcrowding and anaerobiosis (1). Colicins can be grouped into three major
categories based on their routes of lethal action: (i) the formation of a depolarizing ion
channel in the cytoplasmic membrane; (ii) the inhibition of protein and peptidoglycan
synthesis, and (iii) the degradation of nucleic acids (22). Colicin E1 is a member of the
ion-channel forming group of colicins which also includes colicins A, B, Ia, Ib, K, and N
(2).
The C-terminal channel-forming domain of colicin E1 forms a lethal ion channel
which depolarizes the cytoplasmic membrane of target bacterial cells (22). Before its
insertion into the membrane, the colicin E1 channel peptide undergoes a series of
structural changes by first binding to the lipid bilayer, followed by protein unfolding and
helix elongation (5, 11). Finally, the channel domain adopts an insertion-competent
conformation in which it spontaneously inserts into the membrane to form the prechannel state (7). The channel then opens in response to a trans-negative membrane
potential which facilitates the escape of various ions from the host cells, such as Na+, K+,
and H+, eventually leading to host cell death (8).
The crystal structure of the soluble channel domain (2.5 Å) (9, 12) is comprised of
10 α-helices that form an extremely stable, water-soluble globular protein (13, 14).
Interestingly, this protein contains a hydrophobic α-helical hairpin, helices 8 and 9, which
acts as the non-polar core of the protein and becomes transmembrane upon membrane
association (15). These two helices are critical to colicin pore formation as they create a
121
109
membrane-spanning hairpin that anchors the channel within the bilayer (20, 36, 52). The
remaining 8 α-helices of the channel peptide were shown to be amphipathic elements that
surround the hydrophobic core of the channel (24 - 26). In the membrane environment,
the channel peptide forms a structure that has been described as an umbrella model in
which only the hydrophobic helices, 8 and 9, are inserted into the hydrophobic core of the
membrane, with the amphipathic helices splayed out onto the membrane surface (18, 19).
The umbrella model has received strong experimental support from time-resolved FRET
studies of colicin E1 (23, 27, 28). However, the exact orientation of the helices as well as
the details of the lipid and protein contacts are still poorly understood.
It has been shown that the eight amphipathic α-helices on the membrane surface
adopt a two-dimensional arrangement with an area of 4200 Å2, an increase of more than
threefold of the cross-sectional area of the soluble channel domain (29). Similarly, FRET
data for colicin A revealed that distances generally increased upon membrane association
(47, 48). One of our previous studies involved FRET analysis where Cys-505 was
labeled with I-AEDANS as an acceptor and eleven Trp donor residues were randomly
situated throughout the channel domain. The results revealed that higher relative change
in FRET efficiencies were observed the closer the Trp donor was to the N-terminus of the
protein (23).
In support of the umbrella model, Cramer et al. adopted a similar approach by
using FRET to probe the relative distance of each helix relative to Cys-509 (29).
Although the data could be accounted for by the formation of a quasi-circular
arrangement of the eight amphipathic α-helices laying on the membrane surface, a
number of other models with various two-dimensional configurations of the helices are
122
110
also possible. To further test the proposed quasi-circular arrangement model, our lab
employed the system developed by Schultz and coworkers to incorporate coumarin into
the colicin E1 channel domain to act as an intrinsic FRET donor (41) with DABMIlabelled Cys residues as the acceptor. This approach allowed us to build a circular
arrangement model with helices 1-7 arranged in a clockwise direction from the
extracellular side around the central transmembrane hairpin formed by helices 8 and 9.
To further refine our original proposed model, the new approach in this study
involved construction of a triangular foundation frame based on the inter-molecular
distances measured between the three endogenous Trp residues by FRET. This was used
to probe the 3-D orientation of helix 1 by FRET in both the soluble and lipid-bound
states, similar to the satellite GPS concept except it was performed at the molecular level.
Although labour intensive, this new approach can facilitate the construction of a lowresolution 3-D model of the closed colicin E1 channel in the absence of high-resolution
structural data.
4.3
Materials and Methods
4.3.1 Preparation of Single-Trp, Single-Cys Colicin E1 Variants
All colicin E1 single Trp, single Cys variants were prepared by site-directed
mutagenesis as previously described (24). Both the endogenous Trp residues and the
three Helix 1 residues (D347, S354, E361) were replaced with either a Phe or Cys
residue. A total of 12 single-Trp single-Cys variants were prepared using the P190H6
construct with a C505A mutation. All plasmids were purified using the High Pure
123
111
PlasmidTM isolation kit from Roche Diagnostics (Laval, PQ, Canada) and all mutation
sites were confirmed by DNA sequencing (Univ. of Guelph).
Three of the twelve
variants were prepared for the inter-molecular distance measurements between the three
intrinsic Trp residues as follow: (W424C/W495F; W495C/W424F; W424C/W460F).
The remaining nine variants were prepared for the FRET measurements between Helix 1
and the three intrinsic Trp residues as follow:
(D347C/W460F/W495F;
D347C/W424F/W495F;
S354C/W460F/W495F;
S354C/W424F/W495F;
E361C/W460F/W495F;
E361C/W424F/W495F;
D347C/W424F/W460F;
S354C/W424F/W460F;
E361C/W424F/W460F).
4.3.2 Protein Purification and Monobromobimane Labeling
Both the wild-type (WT) P190H6 and all the single-Trp single-Cys mutant
proteins were prepared from transformed lex A- Escherichia coli IT3661 cells and
purified using the Immobilized Metal-Affinity Chromatography (IMAC) as previously
described (25). Protein purities were assessed by SDS-PAGE and protein concentrations
were determined by spectroscopy at A280, using an extinction coefficient (ε) of 29910 M1
cm-1 (ε = 17210 M-1cm-1 for single Trp mutant protein) (24). Purified single-Trp single-
Cys mutant proteins were labelled with the small fluorophore monobromobimane (mBBr,
271.11 g/mol) (Molecular Probes, Eugene, OR) at a 20:1 molar ratio (probe:protein), and
the labelling efficiency was determined as previously described (24).
4.3.3 Preparation of Large Unilamellar Vesicles
124
112
LUVs were prepared from 1,2-dioleoyl-sn-glycero-3-phosphocholine and 1,2dioleoyl-sn-glyerco-3-[phospho-rac-(1-glycerol)] vesicles at a 60:40 molar ratio (Avanti
Polar Lipids). Lipids were prepared and quantified as previously described (24), except
the buffer used to suspend vesicles consisted of 10 mM DMG and 100 mM NaCl (pH
4.0). Asolectin (Fluka) was purified according to the method of Schendel and Reid (24),
and vesicles were prepared as described previously (24). Phospholipid concentration was
determined using the microBartlett assay (24).
4.3.4 Emission Spectra Measurement
The fluorescence emission of the Trp donors in the absence of bimane acceptor
was measured at an excitation wavelength of 295 nm and emission was scanned between
305 – 460 nm at 1 nm step size and 0.2 s integration time. Excitation and emission
bandpasses were set at 4 and 8 nm, respectively. Protein samples were measured at 20
μM at room temperature in both the presence and absence of LUVs.
4.3.5 Absorbance Spectra Measurement
The absorbance spectra of the mBBr-Cys labelled mutant proteins were acquired
as previously described (24) except the wavelength was scanned between 250 – 650 nm
for the purpose of the overlap integral (J) calculations. Absorbances were measured in
125
113
Helma ultramicro absorbance cuvettes (light path 10 mm) using a Cary 300
spectrophotometer.
4.3.6 Fluorescence Lifetime Measurement
A PTI Laser strobe model C-72 lifetime fluorimeter was used for the timeresolved fluorescence measurements. Instrument response function (IRF) was measured
using a 0.0005% scatterer solution. Samples were excited at 365 nm using a pulsed
nitrogen dye-laser operated at 10 Hz. Emission was collected at 450 nm with start and
end delay set at 58 and 120 ns, respectively. Measurements were carried out with 400
channels, 50 ns, integration time, and 15 shots averaged together at 25 °C.
4.3.7 Fluorescence Lifetime Data Analysis
The data analysis was performed using a 1-to-4 exponential fitting program that
involves the deconvolution of the fluorescence decay. Deviations of the best fits were
characterized by the reduced χ2 statistical analysis.
In addition, residual graphics,
autocorrelation curves, and Durbin – Watson statistics were also used to assess the
quality of each fit. The average fluorescence lifetime (τ) was calculated from the
relationship τ = Σαi τi / Σαi and because Σαi = 1 (normalized pre-exponential values),
then τ = Σαi τi.
126
114
4.3.8 Calculation of FRET Efficiencies and Apparent Distances
FRET efficiency (E) between the donor and acceptor chromophores relates to the
inverse sixth power of the distance between the chromophores given by the equation
E = (R06) / (R06 + R6)
[Eq. 1]
where E represents the efficiency of energy transfer, R represents the FRET distance
separating the two chromophores, and R0 is the Förster distance, which is the distance
when the energy transfer efficiency is 50%. The energy transfer efficiency (E) can be
obtained experimentally by either measuring the fluorescence intensity (steady state) or
the lifetime (time resolved) of the donor in the presence and absence of the acceptor as
described by the equation:
E = [ 1 – (FDA) / (FD)] = [ 1 – (τ DA) / (τ D)]
[Eq. 2]
where FDA and FD represent the fluorescence intensity of the donor in the presence and
absence of the acceptor, respectively, and τ DA and τ D represent the lifetime of the donor
in the presence and absence of the acceptor, respectively. The Förster distance (R0) can
be calculated based on the following relationship:
R0 = 9.8 x 103 (J κ2 QD η-4)1/6 Å
[Eq. 3]
where κ2 is the orientation factor (assumed to be 2/3), QD is the quantum yield of the
donor in the absence of acceptor, η is the refractive index of the medium (taken as 1.4),
and J is the spectral overlap integral between the donor and acceptor, which is given by
the following equation:
J = [ʃ FD (λ) εA (λ) λ4 δ λ] / [ʃ FD (λ) δ λ]
127
115
[Eq. 4]
where FD is the fluorescence intensity in the presence of the donor only, εA is the molar
extinction coefficient of the acceptor, and λ is the wavelength. The spectral overlap
integral (J) was calculated using a computer program designed by Dr. U. Oehler
(University of Guelph) by inputting both the donor (Trp) emission and the acceptor
(bimane) absorbance spectra.
4.3.9 Modeling Methodology
The 2D-rearrangement of the membrane-bound channel domain presented here
was obtained by implementing an improved methodology by a combination of geometric
and energetic criteria, performed iteratively, to reach a final structural topology that
fulfills the following references: (i) the reported embedding geometry of the helices /
hairpins, (ii) the right protein geometry (dihedral, bond angles, etc), (iii) low or null atom
clashes, and (iv) all sets of FRET-distances.
To begin with, the various probes used in the studies (bimane, AEDANS,
coumarin, etc) were synthesized in-silico for the different adducts of colicin E1, using the
newest Amber12: EHT force-field implemented in MOE.2012. Thus, rather than using
the inter Cα-distances for the mapping, a ‘center’ was assigned to the probes in this
model. Based on this method, the effective coordinate of the probes can be significantly
distant from the anchor point of the side-chain at the Cα-atom. That is the case for the IAEDANS or DABMI dyes, which are elongated molecules with a longer axis of
approximately 10 to 14 Å, respectively, in their extended conformation.
The
improvement with the inclusion of the actual dimensions and ‘center’ of the probes was
128
116
evident in the strong match between the calculated center-to-center distances and the
distances obtained by FRET for the soluble-form.
A first stage of the mapping was only based on geometrical criteria. In this sense,
each helix or hairpin of the labelled protein was taken independently as a solid-body and
positioned onto / into the membrane. Thus, the hairpin H8/9 was positioned overlaying
exactly the location and orientation that was exhibited in the previous model. However,
the rest of the individual α-helices (H3 to H5, and H10) and the hairpins (H2/3 and H6/7)
were manually positioned and / or rotated in the XY-plane in order to simultaneously
match (roughly, within a variation of ± 8 Å) the complete set of FRET-distances.
In the next modelling stage, the mapping was conducted iteratively with the
following steps: (i) fixing the coordinates of the main-chain atoms of all the α-helices; (ii)
energy minimization for loop reorganization in order to achieve the correct side-chains
geometry and to reduce atom clashes that are restricted by the FRET distances constraint;
(iii) assessment of the loop backbone geometry, the bonded (strain) energy of the sideschains and the restraint energy of the FRET pairs as well as the non-bonded interactions
(mainly H-bonds and Van der Waals energy); (iv) unfixing the coordinates of the mainchain atoms of all the α-helices, except the hairpin H8/9; and (v) relaxation of any
remaining atom clashes and restraint energies by a manual rigid-body XY-rotation and /
or XY-translation of the required structural elements (α-helices, α-hairpins and loops).
Finally, repeat all of the above steps (steps i to v) until the data reaches the previously
mentioned references. In order to optimize the side-chains dihedrals and loop main chain
conformation, the final refining step involved conducting a conformational search using
the LowModeMD methodology.
129
117
4.4
RESULTS
4.4.1 Preparation of Colicin E1 Single-Trp and Single-Cys Variants
Two groups of single-Trp and single-Cys colicin E1 variants were prepared for
the FRET studies. The first group comprised three distinct single-Trp and single-Cys
mutants specifically designed for the FRET distances measurement between the three
intrinsic Trp residues (Fig 4-1). In each mutant, two of the three intrinsic Trp residues
were replaced with either a Phe or Cys residue to act as an acceptor labelling site. The
second group comprises nine distinct single-Trp and single-Cys variants for the purpose
of FRET distance measurements between helix 1 and the three intrinsic Trp residue
positions. Three residues (D347, S354, E361) within helix 1 were selected for cysteine
replacement. In each mutant, two of the three intrinsic Trp residues were replaced with
Phe residues while one of the three selected helix 1 residues was replaced with cysteine to
act as an acceptor labelling site. In this study, the lone endogenous Cys residue (Cys505) in all 12 variants was replaced with Ala-505 to avoid the possibility of non-specific
labelling. Previous studies confirmed that the C505A substitution does not perturb either
the secondary or tertiary structure of the channel domain (24).
130
118
131
119
Figure 4-1. Schematic representation of the colicin E1 channel domain. (A) Cartoon
representation of the 12 intra-molecular distances measured by FRET in this study. The
three natural Trp residues (W424, W460 and W495) are represented by yellow stars
while the three selected for Cys substitution (D347, S354 and E361) in Helix 1 are
represented by stars shown in white. (B) The primary sequence and secondary structure
of the channel-forming domain of colicin E1 (P190H6). Residues squared in black were
subjected to either a Cys or Phe codon replacement for the purpose of this study. The
natural Cys-505 was replaced with Ala-505 in all mutants. (C) The ribbon topology
diagram of the 2.5Å crystal structure of the P190 peptide (PDB: 2I88). The location of
the residues subjected to either cysteine or phenylalanine codon replacements are shown
in sticks and are labelled.
132
120
4.4.2 Protein Expression and Purification
All single-Trp and single-Cys variants were expressed and purified as previously
described (25). The purity of the proteins was assessed by SDS-PAGE analysis showing
> 95% homogeneity. All single-Trp and single-Cys variants showed WT expression
levels except for W460C/W495F/C505A with virtually close to no colicin expression.
The result might suggest that Trp-460 is a critical residue for maintaining the overall
structural integrity of the channel domain.
To resolve the problem, a new variant
(W424C/W495F/C505A) was prepared by switching the donor/acceptor position. This
new variant was shown to restore the WT expression level and allow the same (W424W460) molecular distance to be measured by FRET. It should be noted that the other
variant (W424C/W460F/C505A) with Phe mutation at Trp-460 does not have an
expression issue. This further suggested that it might be the aromatic side-chain at
residue 460 that is critical to the overall structural integrity of the channel peptide.
4.4.3 Spectroscopic Measurement of Trp and mBBr-Cys Tethered Adducts
To assess the protein conformation of the bimane labelled and non-labelled
colicin E1 variants, both the absorbance and fluorescence emission spectra were collected
as shown in Figure 4-2A. Subtle changes were observed for the emission spectrum upon
membrane association. In general, the intensity of Trp emission decreased upon the
addition of LUVs.
In contrast, the absorbance spectrum remained unchanged upon
membrane-binding. The Förster Distance (R0) was calculated based on the overlap
133
121
integral between the emission and absorbance spectra. The absorbance spectra in Figure
4-2A demonstrated the successful formation of the mBBr-Cys tethered adduct with the
presence of the absorbance peak at 380 nm. Based on the given extinction coefficients
for mBBr: mBBr = 5000 M-1cm-1 at 380 nm and P190H = 28589 M-1cm-1 at 280 nm, the
labelling efficiency for all 12 colicin E1 variants were calculated and ranged between 54
– 100%. The mBBr labelling efficiency, which represents the fraction of the variant
being labelled (no label, 0%; complete labelling, 100%) is shown in Figure 4-2B. The
labelling efficiency likely varied depending on the solvent accessibility of the Cys site
within the channel domain. For instance, the reported labelling efficiency of E361 site
was significantly lower compared to D347 and S354 sites.
134
122
Figure 4-2. Fluorescence and absorbance spectra of both tryptophan and bimane Cystethered adducts within ColE1. (A) Absorption (‒‒) and emission (‒‒) spectra of
Trp-424 donor and D347C-mBBr labelled acceptor mutant protein. Absorbance values
were adjusted to accommodate the fluorescence emission intensity. (B) Labelling
efficiency of all the bimane labelled colicin E1 mutant proteins (no label, 0%; complete
labelling, 100%).
135
123
4.4.4 Calculation of the FRET Distances
The standard FRET estimate of the distance between the average position of two
probes, the donor and the acceptor, requires the calculation of the Förster distance by
means of the overlap integral and the quantum yield of the donor, as is defined in Eq. [3].
The parameter, J, for each labelled variant, was obtained with Eq. [4] from the
normalized emission spectrum of the donor and the absorption spectrum of the acceptor
as previously shown for the Trp-424 / D347-BIM. The parameter, QD, in each case was
calculated from the radiative (or natural) lifetime of each emitter (τ n, Table 4-1) and the
corresponding observed average lifetimes, (<τ> D, Table 4-2). Based on the data, the
quantum yield of Trp-424 increased in comparison to the wild-type due to its extended
lifetimes, while for the Trp-495 / D347C variant the quantum yield was reduced
significantly (from 0.37 to 0.26 for the wild-type and the variant, respectively).
Thus, the calculated R0 (here called the global R0) ranges between 26 and 32 Å,
with the calculated efficiency (Eq. [2]) and FRET distances tabulated in Table 4-2 (Eq.
[1]). These calculated FRET distances range from 24 Å for the highest transfer efficiency
between the Trp-424 / E361-BIM pair, to 48.2 Å for the lowest transfer efficiency for the
Trp-460 / D347-BIM pair. When the distances are compared with the known Cα-Cα
distances from the crystal structure (Table 4-2), the calculated values are significantly
higher with average deviation of 11.1 Å which is rather large considering the size of both
probes and assuming that all probes are facing outward. Obviously, there must be an
artifact in the calculation and / or in the implicit assumptions that must be accounted for
in order to achieve more reasonable distances.
136
124
Table 4-1. Fluorescence lifetime data analysis of the three Tryptophan donors in
both the presence and absence of the bimane acceptors in the soluble state.
137
125
Table 4-2. Spectral parameters and distances for fluorescence energy transfer
between tryptophan donors and Cys-Bimane acceptors of the Colicin E1 channel
domain in the soluble state.
138
126
The efficiencies are generally higher than the ones from Eq. [2], which implies
that the actual distances are closer to the R0 of the corresponding FRET pair. However,
this latter parameter must be calculated according to the quantum yield of the respective
‘selected’ component of the donor. Thus, based on the estimated fractional contribution
of the longest component, f3, the Q3 from QD was calculated (Table 4-2), along with the
corresponding R0 for each FRET pair using Eq. [3]. Overall, the ‘selective’ R 0 values are
lower than the ‘standard’ counterparts, which means that the energy transfer from the
‘longest’ Trp conformer demands closer distances for an effective interaction than the
global calculated values using the full donor quantum yield. Thus, the lower R0’s belong
to the higher efficiencies for this selective FRET calculation, and yield significantly
lower FRET distances in comparison to the global ones, as shown in Table 4-2. The
selective FRET distances are in reasonable agreement with the Cα-Cα distances with an
average deviation of 5.8 Å.
However, a more realistic comparison involves taking into account the actual
dimension of the probes and their average orientation in the protein. In this sense, the
soluble crystal structure of colicin E1 was modified and labelled in-silico to produce the
same experimental bimane adducts. This involved using the novel Amber12-EHT forcefield in MOE (see Materials and Methods) to achieve the right geometry by energy
minimizing the side-chains of the bimane and placing the neighbor residues closer into an
implicit solvent environment. Table 4-2 and Figure 4-3 show the center-to-center moiety
distances, dc-c, between the side-chain moieties of both the Trp and bimane probes. The
match between the distances from the structural model and the selective FRET calculated
distances is remarkable, with an average deviation of only 1.62 Å.
139
127
This small average deviation is much lower than the dimension of the probes
themselves which can be attributed to the error in the assignment of the arbitrary
reference atoms using as ‘electronic center’.
Since the calculated selective FRET
distances harmonize with the ones obtained by a computational but independent modeling
assessment, it appears to sufficiently validate the underlying assumptions in the former
calculation.
140
128
Figure 4-3. Distances between the TRP and selected positions on H1 for the soluble
Colicin E1. Distances between the TRP donors (W424, W460 and W495) and the bimane
moieties (acceptor) at locations 347, 354 and 361 on helix H1 are shown, calculated by
two FRET approaches: global (filled bars) and selective (hollow bars), or measured
(striped bars) using an in-silico mutation and labelling of the Xray crystal structure. The
solid lines show the general trend for each donor by connecting the measured distances.
141
129
4.4.5 Time-Resolved Fluorescence Measurements for the Soluble State of ColE1
For FRET distances, it is well known that the approach using lifetime
measurements offers advantages over the conventional steady-state measurements. In
this regard, the fluorescence decay for the soluble form of colicin E1 was monitored as
indicated in Materials and Methods, for the three single Trp emitters for each unlabelled
and bimane-labelled Cys-variant, in buffer at pH 7.0. The convoluted fluorescence
decays were analyzed by two fitting approaches: (a) a discrete lifetime analysis by using
from 1 to 3 exponential components and (b) a lifetime distribution analysis by using the
Exponential Series Method with 200 components logarithmically-spaced in the range
from 0.01 to 6 ns. Both methods reached a reduced χ2 in the range of 0.9 – 1.2 and a
Durbin-Watson factor larger than 1.75, while matching the center of the distributions
with the values reported by the discrete analysis, thus both methods provides reliable
parameters for further interpretation.
142
130
4.4.6 Time-Resolved Fluorescence Measurements for the Membrane-Component
State of Colicin E1
It has been reported that in the presence of a membrane in an acidic medium (pH
4.0), the Trp donor of colicin E1 gives a more homogeneous lifetime in comparison with
the soluble form (18), which would suggest a more similar ‘environment’ for them. The
higher quantum yield for Trp-424 and Trp-460 (Table 4-3, bold data), when they are
compared with the soluble form, correlate inversely with the degree of exposure as shown
by quenching with acrylamide (18). Herein, the Trp-424 variants have the most complex
decay in the membrane-bound form compared with the soluble form, exhibiting up to 4
components ranging from 0.13 ns to 7.23 ns for the D347C variant; while the two
variants of Trp-460 (S354C and E361C) present two decays rather than the one for the
soluble state (Table 4-3, shaded rows). This suggested that the Trp decay behaviour of
the variant in the membrane form is more complex than in the soluble form. However,
overall, the number of decay components shows similar trends for the soluble protein,
where the main signal comes from the longest lifetime, while a marginal (on average <
5%) from the shortest one. For the membrane-bound state, the average lifetimes are
shorter than the corresponding soluble-state values type (Table 4-4, shaded columns),
which consequently report much lower quantum yields.
143
131
Table 4-3. Fluorescence lifetime data analysis of the three tryptophan donors in the
presence and absence of the bimane acceptors in the membrane-bound state.
144
132
Table 4-4. Spectral parameters and distances for fluorescence energy transfer
between tryptophan donors and Cys-bimane acceptors in the membrane-bound
state.
145
133
However, in the presence of bimane, the complex behavior of the Trp emission
remains and preserves the number of decay components (Table 4-3, clear rows), although
as expected, with lower average lifetimes (Table 4-4). The bimane moiety seems to have
a larger effect on Trp signal mainly at the position 354, either by quenching the signal to
a larger extent for Trp-424 or by quenching the signal to a lesser degree for Trp-460 and
Trp-495. This is reflected in the ‘global’ energy transfer efficiency and the calculated
FRET distances shown in Table 4-4, where S354-BIM appeared as an outlier with
unrealistic values. As a result, distance plot shows a spike at S354-BIM for all the Trp
emitters (Fig. 4-4, black bars). This behaviour is particularly unusual if one considers the
middle location of S354-BIM at helix 1.
The calculation of the ‘selective’ FRET
distances based on the same assumption and using the same procedure as for the soluble
form, yielded values, less variable than for the global FRET, still with the ‘break’ at S354
(Fig. 4-4, white bars).
146
134
Figure 4-4. Distances between the TRP and selected positions on H1 for the membranebound colicin channel domain. Distances between the TRP donors (W424, W460 and
W495) and the bimane moieties (acceptor) at locations 347, 354 and 361 on helix H1,
calculated by three FRET approaches: (i) using the average lifetime (black bars), (ii) the
“selective” methods with 3,D and <2/3,DA> (white bars), or (iii) the longest component in
each case, 3,D and 3,DA (striped bars). The solid lines show the general trend for each
donor by connecting the distances according to the calculation in (iii).
147
135
4.4.7 The 3D model for the Membrane-Bound State of Colicin E1
In Table 4-4, the Cβ-distances from the model are shown for a number of residues
in the membrane-bound state of colicin E1 (105). Notably, the reported distances do not
correlate to those calculated by the selective FRET method, with an average deviation of
17.9 Å. The differences are obvious for the Trp-460 variants where the closeness of Trp460 to helix 1, based on the model, cannot explain the observed FRET efficiency /
distances. Consequently, a new model must be developed in order to account for all the
data.
The new closed-state model of the channel-forming domain of colicin E1 is
depicted in Figure 4-5 and described briefly as follows: (i) A core is formed by helices 8
and 9 immersed into the membrane bilayer, which is preserved from the previous model.
Thus, these two α-helices will be used as a common reference framework for the purpose
of comparing the two models. (ii) A circular-like arrangement of the peripheral α-helices
around the H8/9 central hairpin, according to the model proposed by the Cramer group
(29). However, a notable difference is the counter-clockwise orientation of the helix 1 to
helix 7 around the central H8/9 core (when viewed from the outside of the membrane), in
contrast to the clockwise direction for the previous model. (iii) The embedded geometry
of each helix in the membrane including H8/9, and the tilt angles and penetration depths,
were taken from the previous model. In other words, the main chain of the α-helices in
the new model are “vertically-positioned” identical to the previous model. (iv) Similar to
the previous model, the structure of the hairpin H6/7 in the new model is preserved from
the soluble structure. However, in the new model, the α-hairpin H1/2 is also preserved
148
136
from the soluble form, in contrast to the previous model where this hairpin is presented in
an open conformation. (v) The new model is more compact than the previous model,
allowing contact among the secondary structure elements, with clusters of α-helices
around the H8/9 anchor instead of helix 1 as implicit in the previous model.
149
137
Figure 4-5. Revised 3D-model of the membrane-bound colicin E1. Colicin E1 (in
ribbons: from N-terminus (blue) to C-terminus (red)), forms the umbrella-like model by
integrating the H8/9 hairpin, while the rest of the -helices are laying on the membrane
surface. (a) Perspective view. The greys dots represent the membrane plane. (b) Top
view with the -helices numbered. In both, the wide green vectors represent the inertial
axis of the H6-H9 helices. The small green-vectors correspond to the axis coordinates.
150
138
4.4.8 Detailed Features of the New Model
In this study, we propose a new all-atoms model of the membrane-bound state of
colicin E1 which reflects the general properties described in the previous model. Indeed,
the geometry of the membrane penetration of H8/9 was calculated to have a centroid at
the center of the bilayer, traversing the membrane almost vertically at the point of lower
hydrophobic thickness (~ 16 Å), with a favorable G of transfer of –15.9 kcal/mol (105).
The transmembrane location / orientation of this hydrophobic α-hairpin was kept in the
new model, and was used as reference framework for the further positioning of the rest of
the channel domain. However, a clear difference between the two models is the opposite
orientation of the H1-H7 segment around the central H8/9 hairpin. The new model
features a counter-clockwise orientation (throughout H1 to H7, viewed from the
extracellular side), resulting in H1-H5 helices facing the opposite side (in relation to the
previous model) to the H8/9 central element (Fig. 4-6).
151
139
Figure 4-6. Overlay of the two models of the membrane-bound colicin E1. Top view of
the previous model (in red ribbons and labels) and the new one (in blue ribbons and label)
for colicin E1, after superposing the H6 to H9 -helices (black labels).
152
140
The new rearrangement results principally from the consideration related to the
mechanism of insertion, unfolding and stability of the membrane-bound form, being
compatible with the FRET-measurements. It has been proposed that the soluble protein
approaches the membrane by positioning the inner surface of the membrane. The relative
perpendicular orientation between the soluble colicin E1 and the membrane would allow
a progressive insertion of the hydrophobic H8/9 centre into the hydrophobic bilayer core
of the membrane. This process would be driven by the electrostatic complementarity of
both surfaces, given the overall positively charged protein surface and the negatively
charged membrane surface.
The H8/9 insertion necessarily implies the disruption of the interaction network
between the peripherical helices (H1-H7), which gradually enhances their contacts
towards membrane elements (phospholipids, sterols, etc). The insertion process requires
an opening of the α-helical ‘shells’ that surround the H8/9 hairpin, by pivoting the
complete segment (H1 to H7 helices) at the level of the short L7/8 loop (Fig. 4-7), thus
exposing the hydrophobic inner face of the amphipathic helices to the membrane.
However, in the soluble colicin E1, the compact H6/7 α-hairpin is embedded within the
larger H3-H5 segmental-hairpin.
This H3-H5 “external” segmental-hairpin would
contact first the membrane surface during the progress of the H8/9 anchoring.
153
141
Figure 4-7. Model of the embedding and unfolding of colicin E1. Hypothetical trajectory
of the sequential upward movement of the hairpin H6/7 (ribbons from blue to red) due to
the insertion of the H8/9 hairpin (vertical red ribbons) into the membrane. The series was
done by pivoting H6/7 at the green C (center of the rotation) around the indicated
vector.
154
142
The upward movement of the B shell, hinging at the L7/8 loop, might be
concerted along the entire H3-H7 segment, or may occur in steps by first raising the H3H5 segmental hairpin, followed by the inner H6/7 α-hairpin, in which case the long L5/6
loop would act as a hinge allowing the independent elevation of the H3-H5 segmentalhairpin. On the other hand, the H1-H2 helices (or the H1/2 α-hairpin), which forms the A
shell, would be pulled up by lifting the B shell or by itself, hinging at the L2/3 loop, due
to its interaction with the membrane surface. In any case, this mechanism of integration
preserves the relative faces among the helices in the soluble form, i.e., residues that face
one another in the soluble protein keep the same relative disposition in the membranebound form.
The final membrane-induced rearrangement of each structural element (either αhelix or α-hairpin) is determined by its stabilization by the surrounding lipid
environment. However, following the membrane binding process, the α-helices and αhairpins would adopt a stable and compact 2D-rearrangement driven by protein-protein
(i.e., helix-helix and loop-helix) and protein-lipid-protein (i.e., lipid-mediated proteinprotein) interactions. The revised model has a 2D-topology compatible with the above
described unfolding and stabilization mechanism, since after the unfolding only a slight
translation and rotation of the surface helices would be required to obtain the final
conformation.
In contrast, the previous model presents a 2D-topology where residues belonging
to the H1-H5 segment are facing in opposite directions, which would require a large
rearrangement of the α-helices by means of a counter-clockwise unwinding of the H1-H5
segment after the docking of helices H1-H7 (Fig. 4-8). This assembly pivots with an axis
155
143
close to the N-terminus of helix H6, to form the configuration in which the H1 helix is
held into a cleft formed by the H6/7 and H8/9 hairpins (Fig. 4-6). This extensive lateral
displacement of the H1-H5 segment would be characterized mainly by protein-lipid
interactions because of the long inter-helical distances. Hence, because of the absence of
a directional driving force to direct the unwinding and the lack of inter-helical
interactions to stabilize the final topology (except for the cluster around H1), the previous
model would yield a spectrum of membrane-bound intermediates, making even the
concept of a membrane-bound state meaningless.
156
144
Figure 4-8. Segmental helical unwinding in the previous model. Depiction of a
hypothetic trajectory needed for the N-segment (H1-H5, only depicting H3-H5), being
unwound (green arrow), around the helices H6-H9 with a pivot (black dot) located at the
L5/6 loop, to achieve the final clockwise configuration proposed in the previous model.
The green circle contains the revised model of colicin showing the counter-clockwise
orientation.
157
145
4.4.9 Validation of the New Model for the Colicin E1 Closed-State
Since the FRET calculation was validated when comparing the FRET-distances
with those obtained from the X-ray structure of the soluble form, it is only necessary to
validate the closed-state model by using an independent set of FRET distances.
However, since the new 2D-rearrangement was obtained by the simultaneous fitting of 18
FRET-distances comprising (i) 5 pairs of secondary-elements, i.e, inter-helical distances,
(ii) the location of 4 helices relative to a common point, 509 AEDANS, and (iii) 3
selected points spread across the structure and three equidistant points on H1, the model
is self-constrained resulting in little possibility for an alternative topology. Moreover,
since the α-helices were considered to be rigid-bodies as defined by their main-chain
atoms, it is highly probable that the structure reflects the actual closed-state of the
protein.
Nevertheless, to compare FRET distances not used as distance constraints during
the modelling, with distances in the final model, the best FRET pairs for this assessment
would come from the combination among the three Trps since they are common nodes on
the network of distance vectors, producing a triangle with respect to H1 positions.
Indeed, this validation was implemented by the construction of the W424/460BIM,
W495/460BIM and W495/424BIM pairs. Table 4-5 shows the lifetime analysis of the
fluorescent decay for the donor alone and donor-acceptor pair for the combinations
mentioned, while Table 4-6 shows the calculation of the FRET efficiency and FRETdistances according to both the standard and selective FRET.
158
146
Table 4-5. Lifetime data analysis of the fluorescent decay for the tryptophan donors
alone and donor-acceptor pair.
Table 4-6. Calculation of the FRET efficiency and FRET distances based on both
the standard and selective FRET calculation for the three tryptophan donors.
159
147
The match between the global FRET distances and those obtained from the
previous model is not good based upon the Cβ-Cβ distances. Not even the correction
with the selective FRET distances can account for those reported by the previous model.
This is particularly interesting for the W460/424BIM pair where both FRET calculations
yielded the same value (around 27.4 Å) which is clearly different than the one reported
by the previous model (40.4 Å). Thus, the new model accounts remarkable well for the
selective FRET distances with an average deviation of just 2.8 Å.
However, this average value is slightly higher than the value obtained upon
averaging all the deviations from the helix 1 Trp data (1.72 Å), which might be explained
by the fact that the inter-TRP FRET distances actually correspond to the TRP/BIM pair,
while those reported by the model are indeed TRP/TRP.
Clearly, both probes are
different in size, number of rotatable bonds and consequently minimized (average)
configurations, along with the fact that the modelled inter-TRP distances are between βcarbons, while the ones with H1 reported in Table 4-4 (previous section), are center-tocenter distances. All these can account for the ~ 1 Å differences between both sets of
deviations.
In summary, the self-consistency of the model in terms of the matching between
experimental and modelled distances of a related but independent set of pairs, makes the
new closed-state model a robust proposal which is compatible with the architecture of the
soluble state, congruent with the umbrella embedding mechanism, and is suitable
precursor for the open channel state.
160
148
4.5
DISCUSSION
Membrane proteins constitute an important class of proteins that are specifically
targeted by medical drugs to perform a variety of functions that are vital to the survival of
organisms. Therefore, the study of the changes in membrane protein structure induced by
membrane binding play an important role towards the successful design of reliable drugs.
Previously, many membrane protein structures were characterized by X-ray
crystallography and NMR spectroscopy. However, neither method has been able to
provide a membrane-bound structure of colicin E1. Although the umbrella model of
colicin E1 has received strong experimental support in the past, the detailed orientation of
the helices as well as the protein and lipid contacts are still poorly understood. Based on
the earlier proposed quasi-circular arrangement model, a number of 2-D configurations of
the channel domain on the surface of the bilayer are possible and this is primarily due to
the lack of constraints in the model.
Herein, we used a new FRET-based approach that involves using the three natural
Trp residues as the triangulation tool to scan for the 3-D orientation of each helix.
Although it is a tedious and labor-intensive process that would require hundreds of
variants, FRET appears as the only viable alternative that can provide sufficient data for
the construction of a low-resolution 3-D model of the membrane-bound protein in the
absence of both NMR and X-ray structural data. In this study, we report the success of
applying this new method to generate a better refined model that incorporates the precise
3-D location of Helix 1 relative to the remaining body of the protein in the membranebound state. In theory, this technique can be continued to scan the 3-D orientation of the
161
149
entire protein on the bilayer surface. In terms of the reliability of the reported data, all the
measured FRET distances in the soluble state correlate well with the distance data
obtained from the X-ray. Additionally, the changes in the energy transfer efficiency
between the donor / acceptor pairs also correlate well with changes in the FRET distances
between the soluble and LUVs state. Although minor variations were present in the
reported data, most of the deviation could be accounted by the orientation and flexibility
of the donor / acceptor chromophores. In general, the FRET data in this study are in good
agreement with previously reported FRET data.
In this study, FRET measurements between the bimane moiety at the N-terminus
(347), central core (354) and C-terminus (361) of H1, and three other locations of the
protein (424, 460 and 495) were measured, in order to obtain information about the
relative 2D-topology of H1 with respect to the H6-H10 helices (the C-segment).
Moreover, since the relative orientation between H1 and H2 is also known by assuming
that the soluble hairpin H1/2 is stabilized in the membrane-bound form (see above for
discussion), the location of H1 would be a key element to define the final membranebound topology of the channel-forming domain. Thus, as it was discussed, due to the
large distances between H1 and W460, H1 appears to be at the opposite side of the H6/7
hairpin with respect to the central H8/9 hairpin. However, since H1 seems to be fairly
“frontal” with respect to W460 and W495 (i.e., the FRET-distances are similar with the
three H1 locations), for a given H1 location (i.e., the XY coordinates of the H1 centroid),
two H1 orientations are possible (Fig. 4-9). This ambiguity could be unraveled taking
into consideration the ‘reasonable’ embedding and unfolding mechanism discussed
above, where the counter-clockwise orientation resembles the α-helices disposition
162
150
around the core H8/9 in the soluble form, resulting in a defined orientation of H1 where
the H1/2 hairpin looks symmetrical but inverted to the hairpin H6/7 with respect to the
H8/9 central core (Fig. 4-5). In this context, the orientation of the H3-H5 α-helices could
also be determined based on the W424-H1 and the interhelical FRET distances.
163
151
Figure 4-9. Possible orientation of helix H1. Superposition of H1 helix according to a
counter-clockwise orientation in the revised model (H1-H10, red ribbons), and the
clockwise (H1-H2, blue ribbons) orientation.
164
152
Chapter 5
Summary and Conclusions
The research in the thesis mainly focuses on the characterization of the membranebound structure of the colicin E1 channel domain in its closed state. Although both NMR
and X-ray crystallography have not yet provided a membrane-bound structure of colicin E1
due to technical difficulties, various approaches were used in the past to derive structural
information on the protein and were shown to be successful. In this thesis, three different
experimental techniques were employed to derive both two and three-dimensional models of
the colicin E1 channel domain in its membrane-bound state. The first approach in this study
involved using CSM in combination with the site-directed fluorescence labelling technique to
obtain the membrane-bound topology of each helix in a residue-by-residue fashion.
However, the limitation of this technique is that it can only provide 2-D structural data on the
protein and not 3-D information. To overcome this limitation, the second approach in this
study involved using the FRET technique in combination with a genetically encoded
fluorescence technique to estimate the relative distances between various sites on different
helices.
As described in Chapter 1, colicin is a well-characterized bacterial toxin that kills its
neighboring cells through various mechanisms. In fact, the host cell entry mechanism of
colicins has been extensively studied since it shares similar import mechanisms with various
toxins from bacterial pathogens. Similarly, it also shares a common pore-forming structure
with other membrane-active proteins including the apoptotic proteins, Bax and tBid (66, 67).
Therefore, the study of the colicin E1 channel domain membrane-bound structure can be used
as a powerful example for the development of novel techniques in solving the membrane-
153
165
bound structure of similar proteins. As presented in this thesis, we were successful in
deriving extensive structural information on the colicin E1 channel peptide in its membranebound state. Although labour-intensive, these techniques show great promise to provide
topological data of membrane proteins that are not amenable to study by other techniques.
Both CSM and site-directed fluorescence labelling are commonly used techniques in
the field of structural biology to derive topological information of membrane-bound proteins.
In this study, this technique allowed us to construct a high-resolution 2-D model for helices 6,
7 and 10 of the colicin E1 channel domain in its closed state. Results revealed that both
helices 6 and 7 adopt an opposite tilted topology on the bilayer in which the C-terminus of
helix 7 is more buried than its N-terminus whereas helix 10 is relatively solvent exposed.
However, the loop region between helices 7 and 8 was shown to be relatively buried within
the bilayer.
In addition, the various measured fluorescence parameters were used to precisely
determine the changes of both the periodicity and the boundary of each helix upon membrane
association.
The harmonic wave function analysis indicated that both periodicity and
frequency are typical for an amphipathic α-helix in the membrane-bound state (3.34 ± 0.08
rpt for helix 6; 3.56 ± 0.03 rpt for helix 7 and 2.99 ± 0.12 for helix 10). In fact, this technique
proved useful in elucidating the 2-D topology of helices 1 – 5 of the channel domain.
The application of FRET in this study allowed us to precisely map the spatial
orientation of helix 1 relative to the rest of the protein. The technique involved using the
three naturally occurring Trp residues within the channel domain as the fluorescence donors
and the bimane-labelled Cys-mutants on each helix as the fluorescence acceptors. A total of
nine distances (Å) between the three residues on helix 1 (D347, S354, E361) and the three
166
154
Trp residues (W424, W460, W495) were obtained by FRET techniques. In this study, three
additional distances between the Trp residues were also measured and acquired by FRET in
order to build a low-resolution 3-D model of helix 1 by applying triangulation theory. All
distances for the soluble protein correlated well with distances obtained from the X-ray
structure and this demonstrated high reliability of the method.
In fact, the use of the FRET technique can facilitate the construction of an
intermediate resolution 3-D model if every possible donor residue was measured in
combination with the three Trp residues.
However, instead of dissecting the spatial
relationship of each helix in a residue-by-residue fashion, future experiments will be designed
to perform FRET measurements at three positions of each helix to build a low-resolution
model in a helix-by-helix approach. As demonstrated by Cramer et al. (65), this technique
can be used to produce a reasonable model that reveals the conformational changes that occur
within the channel domain in its transition from the soluble to the closed state. In order to
eventually study the open state of the channel, we are currently in collaboration with Dr. J.
Lipkowski (Dept. of Chemistry) in the development of supported planar bilayers on gold
electrode surfaces that would allow us to artificially control both the opening and closing of
the colicin E1 channel. In combination with the developed FRET technique, this would
further allow us to investigate the conformational changes of the channel domain between the
closed and open channel states.
167
155
REFERENCES
1. Pugsley, A. P. (1984) The ins and outs of colicins. Part II. Lethal action, immunity and
ecological implications. Microbiol Sci. 1, 203-205
2. Filloux, A., Voulhoux, R., Ize, B., Gerard, F., Ball, G., and Wu, L. F. (2002) Use of
colicin-based genetic tools for studying bacterial protein tansport. Biochimie. 84, 489497
3. Cramer, W. A., Heymann, J. B., Schendel, S. L., Deriy, B. N., Cohen, F. S., Elkins, P.
A., and Stauffacher, C. V. (1995) Structure function of the channel-forming colicins.
Annu. Rev. Biophys. Biomol. Struct. 24, 611-641
4. Zakharov, S. D., Lindeberg, M., and Cramer, W. A. (1999) Kinetic description of
structural changes linked to membrane import of the colicin E1 channel protein.
Biochemistry. 38, 11325-11332
5. Shin, Y. K., Levinthal, C., Levinthal, F., and Hubbell, W. L. (1993) Colicin E1 binding
to membranes: time-resolved studdies of spin-labeled mutants. Science. 259, 960-963
6. Cramer, W. A., Zhang, Y. L., Schendel, S., Merrill, A. R., Song, H. Y., Stauffacher, C.
V., and Cohen, F. S. (1992) Dynamic properties of the colicin E1 ion channel. FEMS
Microbiol. Immunol. 5, 71-81
7. Tian, C., Tetreault, E., Huang, C. K., and Dahms, T. E. (2006) Electrostatic interactions
of colicin E1 with the surface of Escherichia coli total lipid. Biochim. Biophys. Acta.
1758, 693-701
8. Gouaux, E. (1997) The long and short of colicin action: the molecular basis for the
biological activity of channel-forming colicins. Structure. 5, 313-317
9. Elkins, P. A., Song, H. Y., Cramer, W. A., and Stauffacher, C. V. (1994)
Crystallization and characterization of colicin E1 channel-forming polypeptides.
Proteins. 19, 150-157
10. Zakharov, S. D., Kotova, E. A., Antonenko, Y. N., and Cramer, W. A. (2004) On the
role of lipid in colicin pore formation. Biochim. Biophys. Acta 1666, 239-249
11. Elkins, P., Bunker, A., Cramer, W. A., and Stauffacher, C. V. (1997) A mechanism for
toxin insertion into membranes is suggested by the crystal structure of the channelforming domain of colicin E1. Structure. 5, 443-458
12. Parker, M. W., Postma, J. P., Pattus, F., Tucker, A. D., and Tsernoglou, D. (1992)
Refined structure of the pore-forming domain of colicin A at 2.4 Å resolution. J. Mol.
Biol. 224, 639-657
168
156
13. Schendel, S. L., and Cramer, W. A. (1994) On the nature of the unfolded intermediate
in the in vitro transition of the colicin E1 channel domain from the aqueous to the
membrane phase. Protein Sci. 3, 2272-2279
14. Griko, Y. V., Zakharov, S. D., and Cramer, W. A. (2000) Structural stability and
domain organization of colicin E1. J. Mol. Biol. 302, 941-953
15. Kim, Y., Valentine, K., Opella, S. J., Schendel, S. L., and Cramer, W. A. (1998) Solidstate NMR studies of the membrane-bound closed state of the colicin E1 channel
domain in lipid bilayers. Protein Sci. 7, 342-348
16. Lesieur, C., Vecsey-Semjen, B., Abrami, L., Fivaz, M., and Gisou van der, G. F. (1997)
Membrane insertion: The strategies of toxins (review). Mol. Membr. Biol. 14, 45-64
17. Salwinski, L., and Hubbell, W. L. (1999) Structure in the channel forming domain of
colicin E1 bound to membranes: the 402-424 sequence. Protein Sci. 8, 562-572
18. Tory, M. C., and Merrill, A. R. (1999) Adventures in membrane protein topology. A
study of the membrane-bound state of colicin E1. J. Biol. Chem. 274, 24539-24549
19. Lins, L., El, K. K., Charloteaux, B., Flore, C., Stroobant, V., Thomas, A., Dufrene, Y.,
and Brasseur, R. (2007) Lipid-destabilizing properties of the hydrophobic helices H8
and H9 from colicin E1. Mol. Membr. Biol. 24, 419-430
20. Zakharov, S. D. and Cramer, W. A. (2002) Colicin crystal structures: Pathways and
mechanisms for colicin insertion into membranes. Biochim. Biophys. Acta 1565, 333346
21. Slatin, S. L., Qiu, X. Q., Jakes, K. S., and Finkelstein, A. (1994) Identification of a
translocated protein segment in a voltage-dependent channel. Nature 371, 158-161
22. Cascales, E., Buchanan, S. K., Duche, D., Kleanthous, C., Lloubes, R., Postle, K.,
Riley, M., Slatin, S., and Cavard, D. (2007) Colicin biology. Microbiol. Mol. Biol.
Rev. 71, 158-229
23. Steer, B. A., and Merrill, A. R. (1994) The colicin E1 insertion-competent state:
Detection of structural changes using fluorescence resonance energy transfer.
Biochemistry. 33, 1108-1115
24. Musse, A. A., Wang, J., Deleon, G. P., Prentice, G. A., London, E., and Merrill, A. R.
(2006) Scanning the membrane-bound conformation of helix I in the colicin E1
channel domain by site-directed fluorescence labeling. J. Biol. Chem. 281, 885-895
25. Wei, Z., White, D., Wang, J., Musse, A. A., and Merrill, A. R. (2007) Tilted, extended,
and lying in wait: The membrane-bound topology of residues Lys-381 - Ser-405 of the
colicin E1 channel domain. Biochemistry 46, 6074-6085
169
157
26. White, D., Musse, A. A., Wang, J., London, E., and Merrill, A. R. (2006) Toward
elucidating the membrane topology of helix two of the colicin E1 channel domain. J.
Biol. Chem. 281, 32375-32384
27. Zakharov, S. D., Lindeberg, M., Griko, Y., Salamon, Z., Tollin, G., Prendergast, F. G.,
and Cramer, W. A. (1998) Membrane-bound state of the colicin E1 channel domain as
an extended two-dimensional helical array. Proc. Natl. Acad. Sci. U.S.A. 95, 4282-4287
28. Palmer, L. R., and Merrill, A. R. (1994) Mapping the membrane topology of the closed
state of the colicin E1 channel. J. Biol. Chem. 269, 4187-4193
29. Lindeberg, M., Zakharov, S. D., and Cramer, W. A. (2000) Unfolding pathway of the
colicin E1 channel protein on a membrane surface. J. Mol. Biol. 295, 679-692
30. Lakey, J. H., and Slatin, S. L. (2001) Pore-forming colicins and their relatives. Curr.
Top. Microbiol. Immunol. 257, 131-161
31. Duche, D., Izard, J., Gonzalez-Manas, J. M., Parker, M. W., Crest, M., Chartier, M.,
and Baty, D. (1996) Membrane topology of the colicin A pore-forming domain
analyzed by disulfide bond engineering. J. Biol. Chem. 271, 15401-15406
32. Duche, D., Baty, D., Chartier, M., and Latellier, L. (1994) Unfolding of colicin A
during its translocation through the Escherichia coli envelops as demonstrated by
disulfide bond engineering. J. Biol. Chem. 269, 24820-24825
33. Wang, L., Xie, J., and Schultz, P. G. (2006) Expanding the genetic code. Annu. Rev.
Biophys. Biomol. Struct. 35, 225-249
34. Luo, Y., Leszyk, J., Li, B., Gergely, J., and Tao, T. (2000) Proximity relationships
between residue 6 of troponin I and residues in troponin C: further evidence for
extended conformation of troponin C in the troponin complex. Biochemistry. 39,
15306-15315
35. Musse, A. A., and Merrill, A. R. (2003) The molecular basis for the pH-activation
mechanism in the channel-forming bacterial colicin E1. J. Biol. Chem. 278, 2449124499
36. Piston, D. W., and Kremers, G. J. (2007) Fluorescent protein FRET: the good, the bad
and the ugly. Trends Biochem. Sci. 32, 407-414
37. Lomize, A. L., Pogozheva, I. D., and Mosberg, H. I. (2011) Anisotropic Solvent Model
of the Lipid Bilayer. 1. Parameterization of Long-Range Electrostatics and First
Solvation Shell Effects. J. Chem. Inf. Model. 51, 918-929
38. Lomize, A. L., Pogozheva, I. D., and Mosberg, H. I. (2011) Anisotropic Solvent Model
of the Lipid Bilayer. 2. Energetics of Insertion of Small Molecules, Peptides, and
Proteins in Membranes. J. Chem. Inf. Model. 51, 930-946
170
158
39. Abrams, C. K., Jakes, K. S., Finkelstein, A., and Slatin, S. L. (1991) Identification of a
translocated gating charge in a voltage-dependent channel. Colicin E1 channels in
planar phospholipid bilayer membranes. J Gen. Physiol 98, 77-93
40. Lomize, A. L., Pogozheva, I. D., Lomize, M. A., and Mosberg, H. I. (2006) Positioning
of proteins in membranes: a computational approach. Protein Sci. 15, 1318-1333
41. Wang, J., Xie, J., and Schultz, P. G. (2006) A genetically encoded fluorescent amino
acid. J. Am. Chem. Soc. 128, 8738-8739
42. Ho, D., and Merrill, A. R. (2009) Evidence for the amphipathic nature and tilted
topology of helices 4 and 5 in the closed state of the colicin E1 channel. Biochemistry.
48, 1369-1380
43. Gratia, A. (1925) Sur un remarquable exemple d'antagonisme entre deux souches de
colilbacille. C. R. Soc. Biol 93, 1040-1041
44. Gratia, A., and P. Fredericq. (1946) Diversite des Souches Antibiotiques de B-Coli et
Etendue Variable de Leur Champ Daction. C. R. Soc. Biol 140, 1032-1033
45. Jacob, F., L. Siminovitch, and E. Wollman. (1952) Sur la biosynthese d'une colicine et
sur son mode d'action. Ann. Inst. Pasteur 83, 295-315
46. Lindeberg, A., and Cramer, W. A. (2001) Identification of Specific Residues in Colicin
E1 Involved in Immunity Protein Recognition. J. Bacteriology 183, 2132-2136
47. Duche, D., Baty, D., Chartier, M. and Letellier, L. (1994) Unfolding of colicin A
during its translocation through the Escherichia coli envelope as demonstrated by
disulfide bond engineering. J. Biol. Chem 269, 24820-24825
48. Lakey, J. H., Duche, D., Gonzalez-Mauas, J. M., Baty, D. and Pattus, F. (1993)
Fluorescence energy transfer distance measurements. J. Biol. Chem. 230, 1055-1067
49. Slatin, S. L., and Kienker, P. (2003) Colicin channels and protein translocation:
parallels with diphtheria toxin. Pore forming peptides and protein toxins. 99, 102-131
50. Braun, M., and Dittrich, T. (2010) Synthesis of the fluorescent amino acid rac-(7hydroxycoumarin-4-yl)ethylglycine. Beilstein J. Org. Chem. 6, ii:69
51. Bullock, J. O., E. R. Kolen, and J. L. Shear. (1992) Ion selectivity of colicin E1: II.
Permeability to organic cations. J. Membr. Biol. 128, 1-16
52. Song, H. Y., Cohen, F. S. and Cramer, W. A. (1991) Membrane topology of ColE1
gene products: the hydrophobic anchor of the colicin E1 channel is a helical hairpin. J.
Bacteriol. 173, 2927-2934
171
159
53. Kim, Y., Valentine, K., Opella, S. J., Schendel, S. L. and Cramer, W. A. (1998) Solidstate NMR studies of the membrane-bound closed state of the colicin E1 channel
domain in lipid bilayers. Protein Sci. 7, 342-348
54. Zhang, Y. L., and Cramer, W. A. (1992) Constraints imposed by protease accessibility
on the trans-membrane and surface topography of the colicin E1 ion channel. Protein
Sci. 1, 1666-1676
55. Gillor, O., Kirkup, B. C., and Riley, M. A. (2004) Colicins and microcins the next
generation antimicrobials. Adv Appl Microbiol 54, 129-146
56. Diez-Gonzalez, F. (2007) Applications of bacteriocins in livestock. Curr Issues Intest
Microbiol. 8, 15-24
57. Lancaster, L. E., Wintermeyer, W., and Rodnina, M. V. (2000) Structural stability and
domain organization of colicin E1. J. Mol. Biol. 302, 941-953
58. Piston, W. P., and Kremers, G. J. (2007) Fluorescent protein FRET: the good, the bad
and the ugly. TRENDS in Biochem. Sci. 32, 407-414
59. Miyawaki, A. (1997) Fluorescent indicators for Ca2+ based on green fluorescent
proteins and calmodulin. Nature 388, 882-887
60. Nagai, T. (2004) Expanded dynamic range of fluorescent indicators for Ca2+ by
circularly permuted yellow fluorescent proteins. Proc. Natl. Acad. Sci. U. S. A. 101,
10554-10559
61. Clegg, R. M. (2004) The vital contribution of Perrin and Förster. Biophotonics. Int. 11,
42-44
62. Patterson, G. H (2000) Förster distances between green fluorescent protein pairs. Anal.
Biochem. 284, 438-440
63. Wu, P., and Brand, L. (1994) Resonance energy transfer: methods and applications.
Anal. Biochem. 218, 1-13
64. Chen, H. (2006) Measurement of FRET efficiency and ratio of donor to acceptor
concentration in living cells. Biophys. J. 91, L39-L41
65. Griko, Y. V., Zakharov, S. D. and Cramer, W. A. (2000) Structural stability and
domain organization of colicin E1. J. Mol. Biol. 302, 941-953
66. Yan, L., Miao, Q., Sun, Y. and Yang, F. (2003) iBid forms a pore in the liposome
membrane. fEBS Lett. 555, 545-550.
172
160
67. Basanez, G., Sharpe, J. C., Galanis, J., Brandt, T. B., Hardwick, J. M. and Zimmerberg,
J. (2002) Bax-type apoptotic proteins porate pure lipid bilayers through a mechanism
sensitive to intrinsic monolayer. J. Biol. Chem. 277, 49360-49365.
68.
Brun, M. P., Bischoff, L., and Garbay, C. (2004) A very short route to enantiomerically
pure coumarin-bearing fluorescent amino acid. Angew. Chem., Int. Ed. 43, 3432-3436
69.
Lee, J. H., Choi, B. S., Chang, J. H., Lee, H. B., Yoon, J. Y., Lee, J., and Shin, H.
(2007) The decarboxylative Blaise reaction. J. Org. Chem. 72, 10261-10263
70.
Tararov, V. I., Korostylev, A., Konig, G., and Borner, A. (2006) Facile preparation and
purification of mono tert-butyl malonate. Synth. Commun. 36, 187-191
71.
Brooks, D. W., Lu, L. D. L., and Masamune, S. (1979) C-Acylation Under Virtually
Neutral Conditions. Angew. Chem., Int. Ed. Engl. 18, 72-74
72.
Sui, G., Mabrouki, M., Ma, Y., Micic, M., and Leblanc, R. M. (2002) A structural
study of amphiphilic PAMAM (poly(amido amine)) dendrimers in Langmuir and
Langmuir-Blodgett films. J. Colloid Interface Sci. 250, 364-370
73.
Young, T. S., Ahmad, I., Yin, J. A., and Schultz, P. G. (2010) An enhanced system for
unnatural amino acid mutagenesis in E. coli. J. Mol. Biol. 395, 361-374
74.
Yao, X. L., and Hong, M. (2006) Effects of anionic lipid and ion concentrations on the
topology and segmental mobility of colicin Ia channel domain from solid-state NMR.
Biochemistry. 45, 289-295
75.
Lambotte, S., Jasperse, P., and Bechinger, B. (1998) Orientational distribution of
alpha-helices in the colicin B and E1 channel domains: a one and two dimensional 15N
solid-state NMR investigation in uniaxially aligned phospholipid bilayers.
Biochemistry. 37, 16-22
76. Schendel, S. L., and Reed, J. C. (2000) Measuring pore formation by Bcl-2 family
proteins. Methods Enzymol. 322, 274-282
77.
Huster, D., Xiao, L., and Hong, M. (2001) Solid-state NMR investigation of the
dynamics of the soluble and membrane-bound colicin Ia channel-forming domain.
Biochemistry. 40, 7662-7674
78.
Prieto, L., and Lazaridis, T. (2001) Computational studies of colicin insertion into
membranes: the closed state. Proteins. 79, 126-141
79.
Ridley, H., Johnson, C. L., and Lakey, J. H. (2010) Interfacial interactions of poreforming colicins. Adv. Exp. Med. Biol. 677, 81-90
173
161
80. Eraso, J. M., Chidambaram, M., and Weinstock, G. M. (1996) Increased production of
colicin E1 in stationary phase. J. Bacteriol. 178, 1928-1935
81.
Kleanthous, C., and Walker, D. (2001) Immunity proteins. Enzyme inhibitors that
avoid the active site. Trends Biochem. Sci. 26, 624-631
82.
Gould, J. M., and Cramer, W. A. (1977) Studies on the depolarization of the
Escherichia coli cell membrane by colicin E1. J. Biol. Chem. 252, 5491-5497
83.
Masaki, H., and Ogawa, T. (2002) The modes of action of colicins E5 and D, and
related cytotoxic tRNases. Biochimie. 84, 433-438
84.
Harkness, R. E., and Braun, V. (1989) Colicin M inhibits peptidoglycan biosynthesis
by interfering with lipid carrier recycling. J. Biol. Chem. 264, 6177-6182
85.
Cooper, P. C., and James, R. (1984) Two new E colicins, E8 and E9, produced by a
strain of Escherichia coli. J. Gen. Microbiol. 130, 209-215
86.
Lazdunski, C. J., Bouveret, E., Rigal, A., Journet, L., Loubes, R., and Bene-detti, H.
(1998) Colicin import into Escherichia coli cells. J. Bacteriol. 180, 4993-5002
87.
Jakes, K. S., and Cramer, W. A. (2012) Border Crossings. Colicins and Transporters.
Annu. Rev. Genet. 46, 209-231
88.
Berne, S., Sepcic, K., Anderluh, G., Turk, T., Macek, P., and Poklar Ulrih, N. (2005)
Effect of pH on the pore forming activity and conformational stability of ostreolysin, a
lipid raft-binding protein from the edible mushroom Pleurotus ostreatus. Biochemistry.
44, 11137-11147
89.
Ladokhin, A. S., Isas, J. M., Haigler, H. T., and Whites, S. H. (2002) Determining the
membrane topology of proteins: insertion pathway of a transmembrane helix of annexin
12. Biochemistry. 41, 13617-13626
90.
Garcia-Saez, A. J., Mingarro, I., Perez-Paya, E., and Salgado, J. (2004) Membraneinsertion fragements of Bcl-xL, Bax, and Bid. Biochemistry. 43, 10930-10943
91.
Schendel, S. L., Montal, M., and Reed, J. C. (1998) Bcl-2 family proteins as ionchannels. Cell Death Differ. 5, 372-380
92.
Gonzalez, M. R., Bischofberger, M., Pernot, L., van der Goot, F. G., and Freche, B.
(2008) Bacterial pore-forming toxins. The (w)hole story?. Cell Mol. Life Sci. 65, 493507
93.
Stroud, R. M., Reiling, K., Wiener, M., and Freymann, D. (1998) Ion channel-forming
colicins. Curr. Opin. Struct. Biol. 8, 525-533
174
162
94.
Zakharov, S. D., and Cramer, W. A. (2004) On the mechanism and pathway of colicin
import across the E. coli outer membrane. Front. Biosci. 9, 1311-1317
95.
Elkins, P. A., Song, H. Y., Cramer, W. A., and Stauffacher, C. V. (1994)
Crystallization and characterization of colicin E1 channel-forming polypeptides.
Proteins. 19, 150-157
96.
Parker, M. W., Postma, J. P., Pattus, F., Tucker, A. D., and Tsernoglou, D. (1992)
Refined structure of the pore-forming domain of colicin A at 2.4 Å resolution. J. Mol.
Biol. 224, 639-657
97.
Zakharov, S. D., and Cramer, W. A. (2002) Colicin crystal structures. Pathways and
mechanisms for colicin insertion into membranes. Biochim. Biophys. Acta. 1565, 333346
98.
Tory, M. C., and Merrill, A. R. (2002) Determination of membrane protein topology
by red-edge excitation shift analysis. Application to the membrane-bound colicin E1
channel peptide. Biochim. Biophys. Acta. 1564, 435-448
99.
Schendel, S. L., and Reed, J. C. (2000) Measuring pore formation by Bcl-2 family
proteins. Methods Enzymol. 322, 274-282
100.
Zhao, G., and London, E. (2005) Behavior of diphtheria toxin T domain containing
substitutions that block normal membrane insertion at Pro345 and Leu307. Control of
deep membrane insertion and coupling between deep insertion of hydrophobic
subdomains. Biochemistry. 44, 4488-4498
101.
Cornette, J. L., Cease, K. B., Margalit, H., spouge, J. L., Berzofsky, J. A., and DeLisi,
C. (1987) Hydrophobicity scales and computational techniques for detecting
amphipathic structures in proteins. J. Mol. Biol. 195, 659-685
102.
Labute, P. (2009) Protonate3D. Assignment of ionization states and hydrogen
coordinates to macromolecular structures. Proteins. 75, 187-205
103.
Labute, P. (2008) The generalized Born/volume integral implicit solvent model.
Estimation of the free energy of hydration using London dispersion instead of atomic
surface area. J. Comput. Chem. 29, 1693-1698
104.
Connolly, M. L. (1983) Solvent-accessible surfaces of proteins and nucleic acids.
Science. 221, 709-713
105.
Ho, D., Lugo, M. R., Lomize, A. L., Pogozheva, I. D., SIngh, S. P., Schwan, A. L.,
and Merrill, A. R. (2011) Membrane topology of the colicin E1 channel using
genetically encoded fluorescence. Biochemistry. 50, 4830-4842
175
163
106.
Caputo, G. A., and London, E. (2003) Using a novel dual fluorescence quenching
assay for measurement of tryptophan depth with lipid bilayers to determine
hydrophobic α-helix locations within membranes. Biochemistry. 42, 3265-3274
176
164