Interactions between energy transducer TonB and...
Transcript of Interactions between energy transducer TonB and...
Interactions between energy
transducer TonB and ferric
hydroxamate transport proteins from
Escherichia coli
by
David M. Carter
Department of Microbiology and Immunology
McGill University, Montreal
August 2009
A thesis submitted to McGill University in partial fulfillment of the requirements
of the degree of Doctor of Philosophy
© David M. Carter, 2009
ii
Abstract
Ph.D. David M. Carter Department of Microbiology and Immunology
The ferric hydroxamate uptake (Fhu) system of Escherichia coli transports
ferric hydroxamate-type siderophores. This system comprises outer membrane
(OM) receptor FhuA, periplasmic binding protein FhuD, and ABC permease
FhuB/C. In addition, transport through FhuA requires energy input provided by
the cytoplasmic membrane (CM)-embedded TonB/ExbB/ExbD multi-protein
complex.
This thesis focuses on identification and characterization of protein–
protein interactions that facilitate siderophore transport. Phage display
technology predicted protein–protein interactions involved in TonB-dependent
transport; peptide motifs predicted to bind TonB were identified from phage
panning experiments using purified TonB. Peptide sequences that were displayed
on TonB-interacting phage were similar to sequences of periplasm-exposed
regions on FhuA and therefore predicted FhuA regions that TonB might bind to.
Binding to these regions was confirmed by ELISA; predicted TonB-binding
sequences were fused to maltose-binding protein (MBP) and binding to TonB was
confirmed by immunoreactivity towards monoclonal antibodies directed against
MBP.
TonB was also found to bind FhuD. Peptide sequences displayed on
TonB-binding phage identified regions within FhuD to which TonB was predicted
to bind. Furthermore, phage panning experiments against purified FhuD
predicted complementary regions on TonB that FhuD was predicted to bind.
iii
Binding between TonB and FhuD was confirmed in vitro. Accordingly,
biophysical methods confirmed that TonB and FhuD formed a 1:1 siderophore-
independent complex with an affinity (KD) ranging from 20-200 nM.
Further analyses demonstrated that regions proximal to TonB’s C-
terminus were essential for interaction with FhuD. Binding was characterized
between FhuD and three periplasmic TonB derivatives: a derivative possessing
residues 33–239, a derivative with a deletion of TonB’s central proline-rich
region, and a derivative possessing residues 103–239. Surface plasmon resonance
technology confirmed that all derivatives bound FhuD in concentration-dependent
manners with similar low nanomolar affinities. TonB-derived oligopeptides that
were predicted to bind FhuD were computationally docked to a FhuD crystal
structure. Docking solutions suggested that, when bound to TonB in vivo, FhuD’s
siderophore binding site would orient towards the OM where it could bind
siderophore as it emerges from FhuA’s lumen during transport. These findings
increase our knowledge of TonB-dependent transport by delineating regions of
interaction between protein partners of a siderophore transport system.
iv
Résumé
Ph.D. David M. Carter Département de microbiologie et d’immunologie
Le transport des sidérophores de type hydroxamique de la bactérie
Escherichia coli est effectué par le système d’acquisition du fer hydroxamique
(Fhu) qui comprend FhuA, le récepteur de la membrane externe, FhuD, une
protéine périplasmique de liaison et FhuB/C, une perméase de type ABC. De
plus, pour effectuer le transport, FhuA requiert de l’énergie qui lui est fournie par
le complexe de protéines TonB/ExbB/ExbD situé dans la membrane
cytoplasmique.
Cette thèse se concentre sur l’identification et la caractérisation des
intéractions protéine–protéine impliquées dans le transport des sidérophores. La
technique de phage display, en déterminant des motifs peptidiques qui se lient à
la protéine TonB, a permis d’identifier des intéractions protéine–protéine
impliquées dans le transport TonB-dépendant. Les séquences peptidiques de
phage qui se sont liées à la protéine TonB correspondent à des régions de FhuA
exposées au périplasme, ce qui permet d’avancer que les deux protéines entrent en
contact en ces endroits. Pour confirmer ces résultats, les séquences peptidiques
présumées se lier à TonB ont été fusionnées à la protéine de liaison du maltose
(MBP) pour être testées contre TonB par ELISA. L’immunoréaction de ces
constructions avec des anticorps monoclonaux contre MBP ont permis de
confirmer ces intéractions.
TonB se lie aussi à FhuD. La séquence peptidique des phages qui ont liés
TonB correspond à certaines régions de la protéine FhuD. De plus, l’utilisation de
phage display contre FhuD a permis de déterminer une autre série de séquences
peptidiques qui correspondent à la protéine TonB. Ensemble, ces deux séries de
v
séquences peptidiques identifient des sites d’intéractions complémentaires entre
FhuD et TonB. Des méthodes biophysiques ont permis de confirmer l’intéraction
entre TonB et FhuD in vitro. Celles-ci montrent la formation d’un complexe
ayant une stœchiométrie de liaison de 1:1 ayant une affinité (KD) se situant entre
20 et 200 mM et qui est indépendante de la présence d’un sidérophore.
Des analyses plus poussées ont permis de démontrer que la région la plus
proche de l’extrémité carboxy-terminale de TonB était essentielle pour
l’intéraction avec FhuD. Les caractéristiques de liaison entre FhuD et trois
différentes constructions de TonB ont été étudiées: une construction incluant les
acides aminés 33 à 239, une construction dénuée de la partie centrale riche en
proline et une construction comprenant les acides aminés 103 à 239. La technique
de résonance plasmonique de surface (SPR) a permis de confirmer l’intéraction
entre FhuD et les trois différentes constructions comme étant dépendante de leur
concentration et ayant un niveau d’affinité similaire d’ordre nanomolaire.
L’intéraction entre FhuD et les séquences de peptides de TonB prédites a
été étudié par simulation informatique d’ancrage avec la structure cristalline de
FhuD. Les résultats obtenus suggèrent que lorsque FhuD se lie à TonB in vivo, il
se positionne de façon à ce que le site de liaison du sidérophore s’oriente vers la
membrane externe, d’où il pourrait se lier avec le sidérophore dès que celui-ci
quitte la cavité de FhuA.
Ces résultats améliorent notre compréhension du transport TonB-dépendant en
cernant plus précisément les régions d’intéraction entre les différentes protéines
impliquées dans l’acquisition des sidérophores.
vi
Acknowledgements
First, I thank my supervisor, Dr. James Coulton, for providing the
opportunity to study in a stimulating and supportive lab environment. His
encouragement as an active participant in any activity has always extended far
beyond the pursuit of science. While I can confidently say that I have learned
many techniques that will enhance my career as a scientist, I have also learned
other valuable lessons. My abilities to critically think, write and verbally
communicate have all been finely crafted during my training in the Coulton lab.
For these experiences and many more I am grateful.
I am also thankful to members of my Advisory Committee: Dr. Greg De
Crescenzo, Dr. Hervé LeMoual, and Dr. Allan Matte for their helpful advice and
critical evaluation of my academic progress.
I thank members of the Coulton lab, both past and present, who have been
constant sources of encouragement as well as sources of the often needed
diversion. Our conversations about science have always reassured my career
choice, while our social antics have made me feel all that more connected to a
supporting and caring community. To me, members of the Coulton lab have
always been more than just my associates; they are my friends for life.
My wife, Leila, has also been a constant source of support. I will always
be thankful that our shared passions of music, art and fine dining as well as our
mutual desire for adventure never grow old. I am also thankful that our shared
senses of humor are keeping us so young at heart.
I wish to thank the rest of my family, especially my parents, Orangie and
Percy, who deserve far more thanks and appreciation than there is space to write
in this thesis. With unconditional support, they have always been by my side and
have guided me through some of the most difficult decisions I’ve had to make. I
thank them for supporting all of these decisions.
Finally, I wish to thank all of my friends in Montreal, both those who live
here now and those who once did. I especially thank Jacek Stolcman, Rebecca
McTavish and Amy Gowertz for introducing me to everyone here in Montreal and
for making me realize that I have a family here in this wonderful city.
vii
Contributions to original knowledge
1. Use of phage display technology to predict TonB-binding surfaces on the
periplasmic surfaces of FhuA, FepA, FecA and BtuB.
2. Demonstration that FhuA-derived, TonB-binding peptide sequences bound to
TonB in vitro. In doing so, this was the first literature report on the use of phage
display to predict bacterial protein–protein interactions.
3. Use of phage display technology to predict that TonB and FhuD would interact
and localized complementary regions of binding between these two proteins.
4. Demonstration that TonB and FhuD interact in vitro and that complementary
regions of interaction between these proteins localize to where they were
predicted.
5. Determination of the stoichiometry and affinity of TonB–FhuD interactions.
6. Demonstration that TonB, FhuA and FhuD can form a ternary complex.
7. Identification of essential regions of interaction between TonB and FhuD.
8. Prediction of the mode of binding between TonB and FhuD.
9. Generation of TonB–FhuD and TonB–FhuA–FhuD computational models.
viii
Contributions of Authors
Chapter 2: Carter DM, Gagnon JN, Damlaj M, Mandava S, Makowski L,
Rodi DJ, Pawelek PD and Coulton JW. (2006) Phage display reveals multiple
contact sites between FhuA, an outer membrane receptor of Escherichia coli, and
TonB. Journal of Molecular Biology 357(1): 236-251.
David Carter: Purified TonB. Amplified TonB-affinity-selected phage.
Purified and sequenced TonB-affinity-selected phage DNA. Performed global
analyses on TonB affinity-selected peptides. Performed ELISA with purified
TonB, MBP fusion proteins and FhuA. Wrote manuscript.
Jean-Nicolas Gagnon: Panned phage libraries against TonB. Amplified TonB-
affinity selected phage. Purified and sequenced TonB-affinity-selected phage
DNA.
Moussab Damlaj: Panned phage libraries against TonB. Amplified TonB-
affinity selected phage. Purified and sequenced TonB-affinity-selected phage
DNA. Cloned and purified MBP fusions.
Suneeta Mandava, Dr. Lee Makowski and Dr. Diane Rodi: Assisted in RELIC
analyses.
ix
Dr. Peter Pawelek: Performed RELIC analyses and bioinformatic analyses on
TonB-affinity-selected peptides. Performed computational docking experiments.
Wrote relevant sections of manuscript.
Chapter 3: Carter DM, Miousse IR, Gagnon JN, Martinez É, Clements A,
Lee J, Hancock MA, Gagnon H, Pawelek P and Coulton JW. (2006)
Interactions between TonB from Escherichia coli and the periplasmic protein
FhuD. Journal of Biological Chemistry 281(46):35413-35424.
David Carter: Amplified TonB-affinity-selected phage. Purified and sequenced
TonB-affinity-selected phage DNA. Performed RELIC analyses on TonB
affinity-selected peptides. Assisted with TonB and FhuD purifications.
Generated and purified FhuD T181C mutant. Fluorescently labeled FhuD.
Performed fluorescence titrations with TonB and FhuD. Generated computational
TonB–FhuA–FhuD ternary complex model. Wrote manuscript.
Isabelle Racine-Miousse: Panned phage libraries against FhuD. Amplified
FhuD-affinity selected phage. Purified and sequenced FhuD-affinity-selected
phage DNA.
Jean-Nicolas Gagnon: Panned phage libraries against TonB. Amplified TonB-
affinity selected phage. Purified and sequenced TonB-affinity-selected phage
DNA.
x
Éric Martinez: Performed preliminary TonB–FhuD interaction experiments
Abigail Clements: Assisted with material preparation for TonB–FhuA–FhuD
multicomponent SPR analyses.
Ryan Jongchan Lee: Assisted with TonB and FhuD purifications.
Dr. Mark Hancock: Performed and analyzed SPR experiments. Wrote relevant
sections of manuscript.
Dr. Hubert Gagnon: Performed and analyzed DLS experiments. Wrote relevant
sections of manuscript.
Dr. Peter Pawelek: Performed RELIC analyses and bioinformatic analyses on
TonB-affinity-selected peptides. Wrote relevant sections of manuscript.
xi
Chapter 4: Carter DM, Deme JC, Hancock MA and Coulton JW. C-terminal
region of TonB positions periplasmic binding protein FhuD for siderophore
transport in Escherichia coli. Submitted to Protein Science.
David Carter: Generated TonB 103–239 derivative. Purified TonB 103–239
and FhuD. Analyzed AUC data. Collected and analyzed fluorescence data.
Performed computational modeling experiments. Performed bioinformatic
analyses. Wrote manuscript.
Justin Deme: Purified TonB Δ66–100. Analyzed AUC data.
Dr. Mark Hancock: Performed and analyzed SPR experiments. Wrote relevant
sections of manuscript.
xii
Table of contents
Abstract...................................................................................................................ii
Résumé...................................................................................................................iv
Acknowledgements...............................................................................................vi
Contributions to original knowledge..................................................................vii
Contributions of authors....................................................................................viii
List of figures........................................................................................................xx
List of tables....………………………………………………………………..xxiii
Chapter 1: Literature review and thesis objectives
1.0 Role of iron in the bacterial life cycle.............................................................2
1.0.1 Toxicity of iron................................................................................................2
1.0.2 Limitations of iron bioavailability..................................................................3
1.0.3 Bacterial iron sources.....................................................................................4
1.1 Gram-negative cell envelope...........................................................................4
1.1.1 Outer membrane.............................................................................................5
1.1.2 Lipopolysaccharide.........................................................................................5
1.1.3 Inner leaflet of outer membrane......................................................................6
1.1.4 Outer membrane proteins...............................................................................6
1.1.5 Periplasm........................................................................................................8
1.1.6 Peptidoglycan..................................................................................................8
1.1.7 Lipoproteins..................................................................................................10
1.1.8 Periplasmic proteins.....................................................................................10
xiii
1.1.9 Cytoplasmic membrane.................................................................................11
1.2 Nutrient transport across the cell envelope.................................................11
1.2.1 Passive transport...........................................................................................12
1.2.2 Facilitated transport.....................................................................................12
1.2.3 Active transport.............................................................................................14
1.3 General bacterial strategies for iron acquisition.........................................15
1.3.1 Ferrous iron transport..................................................................................15
1.3.2 Ferric iron transport.....................................................................................16
1.4 Siderophores...................................................................................................17
1.4.1 Siderophore biosynthesis..............................................................................20
1.4.2 Siderophore secretion...................................................................................20
1.5 TonB-dependent transporters.......................................................................21
1.5.1 Structures of TonB-dependent transporters..................................................21
1.5.2 Ton boxes......................................................................................................23
1.5.3 Transcriptional regulatory domains.............................................................24
1.5.4 Siderophore-induced conformational changes.............................................25
1.6 The TonB–ExbB–ExbD complex..................................................................29
1.6.1 TonB: Energy transducer..............................................................................30
1.6.2 TonB proline-rich region..............................................................................31
1.6.3 TonB C-terminal regions form a compact structure.....................................32
1.6.4 Oligomeric state of TonB..............................................................................32
1.7 Protein–protein interactions involving TonB..............................................34
1.7.1 TonB–ExbB–ExbD interactions....................................................................34
xiv
1.7.2 TonB–OM receptor interactions...................................................................36
1.7.3 TonB–cell envelope protein interactions......................................................39
1.8 Mechanisms of TonB–dependent energy transduction..............................40
1.8.1 The shuttle model..........................................................................................40
1.8.2 The rotation model........................................................................................42
1.8.3 The mechanical pulling model......................................................................44
1.8.4 Conformational rearrangement of TBDT cork domains...............................46
1.9 Periplasmic siderophore transport...............................................................47
1.9.1 Ligand-induced periplasmic binding protein conformational changes........48
1.10 Cytoplasmic membrane permeases............................................................50
1.10.1 Permease–periplasmic binding protein interactions..................................51
1.10.2 Permeases: transport mechanism...............................................................52
1.10.3 Intracellular fate of iron.............................................................................53
1.11 Introduction to techniques used in this thesis...........................................54
1.11.1 Phage display..............................................................................................54
1.11.2 Surface plasmon resonance........................................................................57
1.11.3 Dynamic light scattering.............................................................................58
1.11.4 Analytical ultracentrifugation.....................................................................59
1.12 Rationale and thesis objectives...................................................................60
Preface to chapter 2.............................................................................................62
xv
Chapter 2: Phage display reveals multiple contact sites
between FhuA, an outer membrane receptor of
Escherichia coli, and TonB
2.0 Summary.........................................................................................................64
2.1 Introduction....................................................................................................65
2.2 Materials and methods..................................................................................70
2.2.1 Bacterial strains and media..........................................................................70
2.2.2 Chemicals and reagents................................................................................70
2.2.3 Protein purification.......................................................................................71
2.2.4 Phage M13 titre............................................................................................71
2.2.5 Panning procedures......................................................................................72
2.2.6 Isolation of phage M13 clones, DNA isolation and sequencing...................73
2.2.7 Global analysis of affinity-selected peptides................................................73
2.2.8 Cloning of peptide-coding DNA sequences into pMal-pIII vector...............74
2.2.9 Peptide-MBP expression...............................................................................74
2.2.10 Enzyme linked immunosorbent assay (ELISA)............................................75
2.3 Results.............................................................................................................76
2.3.1 Isolation of affinity-selected peptides by phage panning..............................76
2.3.2 Global analysis of affinity-selected peptides................................................77
2.3.3 Identification of TonB-binding sites on the periplasmic surface of FhuA....80
2.3.4 Identification of potential TonB-binding sites in structurally conserved OM
receptors.................................................................................................................85
xvi
2.3.5 Interactions of TonB and [FhuA peptide-MBP] fusion proteins in vitro......91
2.4 Discussion........................................................................................................93
2.5 Acknowledgements.......................................................................................101
Preface to chapter 3...........................................................................................102
Chapter 3: Interactions between TonB from Escherichia
coli and the periplasmic protein FhuD
3.0 Summary.......................................................................................................104
3.1 Introduction..................................................................................................105
3.2 Materials and methods................................................................................109
3.2.1 Bacterial strains, phage libraries, and media.............................................109
3.2.2 Chemicals and reagents..............................................................................109
3.2.3 Protein purification.....................................................................................110
3.2.4 Phage display..............................................................................................111
3.2.5 Dynamic light scattering.............................................................................111
3.2.6 Fluorescence spectroscopy.........................................................................113
3.2.7 Surface plasmon resonance (SPR)..............................................................115
3.3 Results...........................................................................................................117
3.3.1 Identification of TonB-binding sites on FhuD by phage display................117
3.3.2 Identification of FhuD-binding sites on TonB by phage display................118
3.3.3 Detection of a TonB–FhuD complex by dynamic light scattering..............126
3.3.4 Detection of a TonB–FhuD complex by fluorescence spectroscopy...........127
xvii
3.3.5 Detection of a TonB–FhuD complex by surface plasmon resonance.........132
3.4 Discussion......................................................................................................135
3.5 Acknowledgements.......................................................................................141
Preface to chapter 4...........................................................................................143
Chapter 4: C-terminal region of TonB positions
periplasmic binding protein FhuD for siderophore
transport in Escherichia coli
4.0 Summary.......................................................................................................145
4.1 Introduction..................................................................................................146
4.2 Materials and methods................................................................................149
4.2.1 Bacterial strains and plasmids....................................................................149
4.2.2 Cloning of TonB 103–239...........................................................................149
4.2.3 Protein expression.......................................................................................149
4.2.4 Protein purifications...................................................................................150
4.2.5 Analytical ultracentrifugation.....................................................................151
4.2.6 Fluorescence spectroscopy.........................................................................151
4.2.7 Surface plasmon resonance........................................................................152
4.2.8 Computational docking...............................................................................153
4.3 Results...........................................................................................................154
4.3.1 TonB derivatives are elongated monomers with similar elements of tertiary
structure...............................................................................................................154
xviii
4.3.2 TonB derivatives bind FhuD with equal affinities......................................157
4.3.3 Computational models predict the orientation of FhuD when bound to
TonB.....................................................................................................................158
4.4 Discussion......................................................................................................161
4.5 Acknowledgments........................................................................................168
Preface to chapter 5...........................................................................................169
Chapter 5: Preliminary crystallization of the TonB–FhuD
complex
5.1 Introduction..................................................................................................171
5.2 Materials and methods................................................................................172
5.2.1 Bacterial strains and plasmids....................................................................172
5.2.2 Protein expression and purification............................................................172
5.2.3 Removal of FhuD His-tag...........................................................................172
5.2.4 TonB–FhuD–Fcn complex formation.........................................................173
5.2.5 TonB–FhuD–Fcn crystallization screening................................................174
5.2.6 Assessing TonB degradation.......................................................................174
5.2.7 TonB–FhuD cross-linking...........................................................................175
5.3 Results...........................................................................................................175
5.3.1 Protein preparations and processing..........................................................175
5.3.2 TonB–FhuD–Fcn complex formation.........................................................176
5.3.2 High-throughput crystallization screening.................................................176
xix
5.3.3 TonB exhibits time-dependent degradation................................................179
5.3.4 TonB–FhuD formaldehyde-cross-linking...................................................181
5.4 Discussion......................................................................................................182
5.5 Acknowledgements.......................................................................................184
Chapter 6: Conclusions and future work
6.0 Thesis objectives within the context of TonB-dependent transport........186
6.1 Directions for future research.....................................................................188
6.1.1 Demonstration of TonB–FhuD interactions in vivo....................................188
6.1.2 Refinement of TonB–FhuD interaction localizations..................................189
6.1.3 TonB–FhuD crystallization.........................................................................190
6.1.4 Phage display predictions of TonB-interacting proteins............................190
6.1.5 Determination of whether TonB regulates binding of siderophore to
FhuD....................................................................................................................191
6.1.6 Elucidation of siderophore binding sites by phage display........................191
References.....................................................................................................193
xx
List of Figures
Chapter 1
Figure 1.1. Gram-negative bacterial cell envelope.................................................7
Figure 1.2. E. coli OM nutrient transporters.........................................................13
Figure 1.3. Siderophores transported by E. coli....................................................19
Figure 1.4. Structures of TonB-dependent transporters........................................23
Figure 1.5. FhuA exhibits periplasmic conformational changes upon binding
ferrichrome.............................................................................................................26
Figure 1.6. The TonB–ExbB–ExbD complex.......................................................31
Figure 1.7. Structures of C-terminal, E. coli TonB derivatives............................33
Figure 1.8. Crystal structures of TonB bound to FhuA and BtuB........................38
Figure 1.9. Shuttle model of TonB-dependent energy transduction.....................41
Figure 1.10. Rotation model of TonB-dependent energy transduction.................43
Figure 1.11. Mechanical pulling model of TonB-dependent energy
transduction............................................................................................................45
Figure 1.12. Periplasmic binding proteins............................................................48
Figure 1.13. Phage panning...................................................................................56
Chapter 2
Figure 2.1. Alignments of affinity-selected peptides to FhuA as identified by
RELIC/MATCH.....................................................................................................83
xxi
Figure 2.2. RELIC/HETEROalign similarity scores mapped to the periplasm-
exposed surface of FhuA (PDB code 2FCP).........................................................86
Figure 2.3. Alignments of affinity-selected peptides to the Ton box regions and
periplasm-exposed turns of BtuB, FecA and FepA as identified by
RELIC/MATCH.....................................................................................................89
Figure 2.4. RELIC/HETEROalign similarity scores mapped to the periplasm-
exposed surfaces of BtuB (PDB code 1NQE), FecA (PDB code 1KMO), and
FepA (PDB code 1FEP).........................................................................................90
Figure 2.5. Interactions of TonB and [FhuA-MBP] fusion proteins in vitro........92
Figure 2.6. TonB-binding surfaces on the periplasmic face of FhuA...................98
Chapter 3
Figure 3.1. Alignments of TonB affinity-selected peptides to FhuD as identified
by RELIC/MATCH..............................................................................................121
Figure 3.2. TonB-binding regions identified by phage display mapped to FhuD
(PDB code 1EFD)................................................................................................122
Figure 3.3. Alignments of FhuD affinity-selected peptides to TonB..................124
Figure 3.4. FhuD-binding region identified by phage display mapped to TonB
(PDB code 1XX3)................................................................................................125
Figure 3.5. Binding of Fcn to FhuD and to FhuD T181C...................................129
Figure 3.6. Binding of TonB to AEDANS-labeled FhuD T181C and to MDCC-
labeled FhuD T181C............................................................................................131
xxii
Figure 3.7. Real-time kinetics of TonB–FhuD binding interaction detected by
SPR.......................................................................................................................134
Figure 3.8. Multicomponent SPR analysis to detect ternary complex formation
between FhuA–TonB–FhuD................................................................................135
Figure 3.9. Model of a FhuA-TonB-FhuD ternary complex...............................140
Chapter 4
Figure 4.1. Schematic representations of TonB derivatives from this study......155
Figure 4.2. Single cycle kinetic analysis of FhuD binding to TonB derivatives
using label-free, real-time SPR............................................................................157
Figure 4.3. TonB region II peptide docked to the surface of FhuD....................160
Figure 4.4. Sequence conservation of FhuD from various pathogenic Gram-
negative bacteria..................................................................................................166
Chapter 5
Figure 5.1. Protein purification and processing..................................................177
Figure 5.2. Complexation of TonB 103–239–FhuD–Fcn...................................177
Figure 5.3. Crystallization screen of the TonB 103–239–FhuD–Fcn
complex................................................................................................................178
Figure 5.4. Degradation of TonB 103–239.........................................................180
Figure 5.5. Formaldehyde cross-linking of TonB–FhuD complex.....................181
xxiii
List of Tables
Chapter 2
Table 2.1. RELIC/MATCH identification of TonB-affinity-selected Ph.D.-12
peptides corresponding to FhuA sequences...........................................................84
Table 2.2. RELIC/MATCH Identification of TonB-affinity-selected Ph.D.-C7C
peptides corresponding to FhuA sequences...........................................................84
Table 2.3. RELIC/MATCH Identification of TonB-affinity-selected Ph.D.-12
peptides corresponding to BtuB, FecA, and FepA sequences...............................87
Table 2.4. RELIC/MATCH Identification of TonB-affinity-selected Ph.D.-C7C
peptides corresponding to BtuB, FecA, and FepA sequences...............................87
Chapter 3
Table 3.1. RELIC/MATCH identification of TonB-affinity-selected Ph.D.-C7C
peptides corresponding to FhuD sequences.........................................................120
Table 3.2. RELIC/MATCH identification of TonB-affinity-selected Ph.D.-12
peptides corresponding to FhuD sequences.........................................................120
Table 3.3. RELIC/MATCH identification of FhuD-affinity-selected Ph.D.-C7C
peptides corresponding to TonB sequences.........................................................123
Table 3.4. RELIC/MATCH identification of FhuD-affinity-selected Ph.D.-12
peptides corresponding to TonB sequences.........................................................123
Table 3.5. DLS analysis of TonB, FhuD and MBP-switch fusion......................126
xxiv
Table 3.6. Summary of ligand binding parameters fit to a single site saturation
ligand binding model...........................................................................................130
Chapter 4
Table 4.1. Kinetics and affinity of TonB–FhuD interactions according to “1:1
titration” model....................................................................................................158
Chapter 1
Literature review and thesis objectives
2
1.0 Role of iron in the bacterial life cycle
This literature review focuses on mechanisms of siderophore-mediated
iron uptake in Gram-negative bacteria, with emphasis placed on the model
organism Escherichia coli. Iron represents a vital element that participates in
many cellular metabolic processes. Most bacteria require iron for growth and
division except for certain lactobacilli species, which replace iron with the
elements manganese and cobalt (1). Cellular processes that require iron include
nucleotide and protein synthesis, respiration, regulation of gene expression,
xenobiotic degradation and maintenance of oxidative homeostasis (2).
Accordingly, iron is a co-factor of many proteins and enzymes including
ribonucleotide reductases, oxidases, cytochromes, peroxidases, and aconitases (2).
Iron’s unique electrochemical properties make it an ideal participant in
processes that require electron transfer. As a first row transition metal, iron
generally exists in one of two oxidation states (3): the ferric (Fe3+
) state and the
ferrous (Fe2+
) state. Under physiological conditions, oxidation states can inter-
convert (3), resulting in an ability to modulate iron’s coordination number. For
this reason, the activities of proteins that utilize iron as a co-factor can be
modulated so as to regulate vital metabolic processes.
1.0.1 Toxicity of iron
Despite iron’s importance as a mediator of cellular function, it also
represents a potentially toxic element due to its propensity to react with oxygen
and reactive oxygen species (ROS). Under aerobic conditions, cellular respiration
3
produces considerable amounts of ROS such as superoxide and H2O2. Ferrous
iron can react with these species through a series of steps known as Fenton
reactions to generate further ROS such as hydroxyl radicals (4). Through
oxidative mechanisms, hydroxyl radicals damage biological macromolecules such
as proteins, nucleic acids and lipids. For this reason, control of oxidative
homeostasis is vitally important to bacteria. By incorporating into enzymes, iron
acts as a sensor for the cellular redox potential; iron-dependent enzyme activities
become modulated as a function of the co-factor’s oxidation state (3). Through
activities of iron-dependent enzymes, such as superoxde dismutase and catalase,
the toxic effects of ROS are mitigated.
1.0.2 Limitations of iron bioavailability
Despite being the fourth most abundant element on Earth (3), iron is
essentially inaccessible to biological systems. Earth’s aerobic and aqueous
environment stabilize iron’s ferric state (5). Ferric iron combines with hydroxides
to form insoluble ferric hydroxides; the concentration of soluble ferric iron is
reduced to levels below 10-18
M (6). For bacteria that colonize host organisms,
bioavailability is reduced even further by sequestration strategies of the innate
immune system. In this niche, bioavailable iron is present at a concentration of
approximately 10-24
M, a concentration well below the micromolar amounts
required for a single generation of a bacterium’s life cycle (7).
4
1.0.3 Bacterial iron sources
Bacterial iron sources are governed by the niche that a particular
bacterium occupies. Environmental-dwelling bacteria obtain iron through soil or
water basin sources. Host-colonizing bacteria obtain iron directly from cellular
sources, such as erythrocytes and macrophages, or from protein sources, such as
hemoglobin, transferrin and lactoferrin. Regardless of the source, bacterial iron
acquisition requires passage of the nutrient across biological barriers. For Gram-
negative bacteria, this requires transport of iron across two distinct membranes
that constitute the cell envelope. Since these barriers discourage transport, a
detailed description of the Gram-negative cell envelope is warranted.
1.1 Gram-negative cell envelope
The Gram-negative cell envelope is a structure that is both beneficial and a
liability. Its benefits include functioning as a protective barrier against toxic
compounds, such as solvents, bile salts or antibiotics (8). In addition, the cell
envelope enables attachment to environmental substrates or to host cells, and
enables the bacterium to evade host immune responses. However, the cell
envelope also presents a diffusion barrier that prevents nutrients greater than 700
Da from diffusing into the cytoplasm (9). The composition of the Gram-negative
cell envelope (Figure 1.1) comprises three main components: the outer membrane
(OM), the periplasm (which includes the peptidoglycan layer), and the
cytoplasmic membrane (CM).
5
1.1.1 Outer membrane
The Gram-negative bacterial OM is an asymmetric structure; exposed to
the extracellular space is a leaflet comprised of lipopolysaccharide (LPS).
Exposed to the periplasmic space is a phospholipid leaflet. Embedded between
these leaflets are proteins that bestow the bacterium with many adaptive
capabilities. Each of these features is now discussed in greater detail.
1.1.2 Lipopolysaccharide
Embedded within the OM outer leaflet are LPS molecules. Each LPS
molecule is a tripartite structure comprising a proximal lipid A moiety, a central
core oligosaccharide and distal O-antigen repeats (10). The lipid A moiety is
essential in forming the integrity of the OM. It possesses phosphorylated
hydrophilic glucosyl-amine head groups and saturated fatty acid tails. Lipid A’s
head groups bestow a net negative charge to the Gram-negative OM, while the
fatty acid tails partition together to form a hydrophobic plane in the outer leaflet.
Connected to lipid A head groups are central, hydrophilic connective core
oligosaccharides, which provide a protective barrier against charged toxins such
as antimicrobial peptides, bile salts and dyes. The central oligosaccharide
comprises two elements: inner and outer cores. The inner core links directly to
lipid A, while the outer core extends into the extracellular space and links to the
distal O-antigen polysaccharides.
Extending beyond the core oligosaccharide into the extracellular medium
are O-antigen repeats. O-antigen repeats are required for virulence, since loss of
6
these regions often attenuates infectivity (11). These repeats form a polymer of
sugars, which display great variation in chemical composition and number.
Chemical compositions vary depending on a given bacterial species and can
include neutral sugars, amino sugars and uronic acids. In addition, the number of
repeats is also species-dependent. Even within strains of a given species, the
number ranges anywhere from zero to over forty repeats (12).
1.1.3 Inner leaflet of outer membrane
The OM inner leaflet consists of phospholipids. Primarily, inner leaflet
phospholipids are similar in composition to those of the CM, but vary somewhat
depending on bacterial species and strains. In E. coli, the inner leaflet
composition prefers enrichment of phosphatidylethanolamine lipids (13). In
addition, smaller amounts of phosphatidylglycerol and di-phosphatidylglycerol
are found (12).
1.1.4 Outer membrane proteins
In addition to LPS and phospholipid, the OM comprises approximately
50% protein (8). These OM proteins (OMPs) form the basis of selective
permeability, a hallmark of OM function. Diverse properties are ascribed to
OMPs, some of which are described in greater detail later. In general, OMPs are
classified either as porins that allow non-specific diffusion of small solutes into
7
Figure 1.1. Gram-negative bacterial cell envelope. Illustrated is a schematic of the
Gram-negative bacterial cell envelope. The asymmetric outer membrane comprises two
lipid leaflets: an outer lipopolysaccharide leaflet and an inner phospholipid leaflet. The
remaining OM elements comprise outer membrane proteins (OMPs). OMPs function as
structural proteins (OmpA, Lpp and Pal), or they function as selective permeability
barriers. Porins, such as OmpF, passively uptake water-soluble solutes (preferentially
cations), whereas OMPs, LamB and FadL, facilitate diffusion of maltosaccharides and
long-chain fatty acids, respectively. OMPs such as the transporter, FhuA, actively uptake
ferric-siderophores. Other OMPs such as TolC facilitate efflux. The periplasm is an
aqueous compartment comprised of a peptidoglycan layer and protein. Peptidoglycan is
bound to the OM by OmpA, Lpp and Pal. Periplasmic proteins, such as maltose binding
protein and FhuD, shuttle maltosaccharides and ferric-siderophores, across the periplasm
respectively. The CM is comprised of a phospholipid bi-layer with embedded
transmembrane proteins. CM proteins have major functions such as respiration and
transport. Proteins such as those comprising electron transport complexes actively pump
protons across the CM and generate the PMF. Proteins such as MscL bi-directionally
transport cations across the CM. Active transporters, such as MalFG/K2 and FhuB/C,
hydrolyze cytoplasmic ATP and transport maltosaccharides and ferric-siderophores into
the cytoplasm, respectively. The TonB–ExbB–ExbD complex harnesses energy of the
PMF and transduces it to OM-embedded TonB-dependent transporters such as FhuA.
Other CM proteins such as EntS are responsible for secretion of siderophores from the
cytoplasm to the periplasm.
8
the periplasm, as enzymes with various catalytic activities, as cell surface
appendages, or as structural elements that maintain OM integrity. OMPs can
traverse both leaflets of the OM as integral membrane proteins or associate with
one leaflet of the OM by a covalently attached lipid as peripheral membrane
proteins (Figure 1.1).
1.1.5 Periplasm
The periplasm occupies space between the OM and CM (Figure 1.1). This
compartment houses key structural components of the cell envelope including the
cell wall (peptidoglycan) and numerous proteins. Proteins found in this
compartment exhibit many diverse functions (discussed in sections 1.1.8 and 1.9).
The volume occupied by this space varies according to environmental factors such
as osmolarity and may be species-specific. Under physiological conditions, the E.
coli periplasm is estimated to have a volume that occupies between 8-20% of total
cell volume and a width between OM and CM ranging from 130 to 250 Å (14).
1.1.6 Peptidoglycan
The cell wall structure provides a rigidity that gives bacteria their
characteristic shapes (Figure 1.1). It also prevents cytoplasmic contents from
rupturing under conditions of low osmolality (15). Structurally, peptidoglycan is
a heteropolymer comprising roughly equal amounts of polysaccharides that are
linked to peptides. The basic repeating peptidoglycan unit is shared amongst most
Gram-negative bacteria and comprises two linked N-acetylglucosamine (NAG)
9
and N-acetylmuramic acid (NAM) sugar residues that are derivatized with
oligopeptides (16). Each NAG-NAM repeat is linked continuously to another
NAG-NAM repeat to form a linear backbone. Projecting away from this
backbone are oligopeptides, which form cross-links with adjacent peptidoglycan
strands to yield the peptidoglycan scaffold.
The peptidoglycan scaffold is a vast sieve-like three-dimensional structure
that encompasses the bacterium. An NMR structure of a NAG-NAM
pentapeptide revealed the structural basis of these properties (17). From the
pentapeptide structure, a three-dimensional model was built; NAG-NAM residues
formed a linear backbone from which the pentapeptide linkers extended. Sieve-
like characteristics were formed by pores between the model’s adjacent cross-
linked peptidoglycan strands. Pore diameters were estimated from this model and
ranged from 70 to 100 Å, wide enough to accommodate passage of nutrients and
even proteins.
The orientation of pore openings with respect to the OM plane is a debated
subject. Currently, two conflicting reports have provided evidence for the
direction of pore orientations. While the NMR study suggested that pores orient
perpendicular to the OM plane (17), an electron cryotomography study suggested
parallel orientations (18). Despite these differences, perpendicular orientations
are attractive since the pore diameters are large enough to accommodate proteins,
especially those either targeted to the OM or those with OM-proximal activities.
10
1.1.7 Lipoproteins
The peptidoglycan layer is secured to the OM by interactions with various
proteins. Three well-characterized E. coli peptidoglycan binding proteins include
OmpA, lipoprotein (Lpp) and Pal (Figure 1.1). OmpA is a ubiquitous structural
OMP that possesses two domains. Its N-terminal transmembrane domain folds
into a monomeric, 8-stranded β-barrel, while its C-terminal periplasmic domain
folds into a peptidoglycan-binding motif (19). Lpp is the most abundant protein
in E. coli with an estimated 700,000 copies per bacterium (20). It forms a
trimeric, coiled-coil structure that bridges its N- and C-termini (21). It
intercalates within the OM inner leaflet by three lipids covalently attached to its
N-terminus, and covalently attaches to peptidoglycan by means of its C-terminal
lysine residue (22). Like OmpA and Lpp, Pal attaches to the inner leaflet of the
OM by its N-terminus, while its C-terminus interacts with peptidoglycan (23).
Together, OmpA, Lpp, and Pal secure the peptidoglycan layer in close proximity
to the OM inner leaflet.
1.1.8 Periplasmic proteins
Many proteins exist within the periplasm. Some are enzymes, while
others bind nutrients (Figure 1.1). In addition, some OM- and CM-embedded
proteins possess significant periplasmic domains. Periplasmic enzymes regulate a
large variety of physiological processes including OM biogenesis, peptidoglycan
assembly, protein folding, disulfide formation, surface appendage assembly and
protein secretion. Periplasmic nutrient-binding proteins represent a large and
11
diverse class of proteins and are discussed in section 1.9. The abundance of
periplasmic proteins, taken together with a limited periplasmic volume and
peptidoglycan sieving properties, suggests that the periplasm is a highly viscous
and gel-like compartment (14). This property results in drastically reduced
diffusion-facilitated nutrient transport, an important theme in the context of
bacterial iron acquisition.
1.1.9 Cytoplasmic membrane
The CM separates cytoplasmic contents from the rest of the bacterium. It
comprises roughly equal amounts of phospholipid and protein content.
Phospholipid compositions vary between species, but generally comprises 70-
80% phosphatidylethanolamine, 15-25 % phosphatidylglycerol and 5-10%
cardiolipin (24). In addition, small amounts of metabolic by-products can
incorporate into the membrane. Many proteins embed within the CM and
participate in physiological processes including cell wall synthesis, protein and
small molecule secretion, nutrient uptake and respiration, of which the latter
activity generates CM-associated proton motive force (Figure 1.1).
1.2 Nutrient transport across the cell envelope
The OM of Gram-negative bacteria provides an efficient barrier against
toxic compounds. However, it also prevents diffusion of larger nutrients,
including iron in the form most often acquired by bacteria. Therefore, bacteria
must utilize a limited number of mechanisms for passage of nutrients across the
12
OM. Mechanisms include passive transport, facilitated transport, and active
transport. Each mechanism is now discussed in greater detail.
1.2.1 Passive transport
Passive diffusion is the simplest nutrient uptake mechanism, whereby
small hydrophilic solutes diffuse non-specifically through OM porins. Diffusion
is permitted by differential permeabilities afforded by porin transmembrane
channels. Examples include the cation-selective porins OmpC and OmpF and the
anion-selective porin PhoE. Structural data (Figure 1.2) revealed how porins
permit non-specific diffusion; trimeric β-barrel assemblies insert into and traverse
the OM (25-27). Under physiological conditions, barrel diameters support
passive diffusion of solutes smaller than 700 Da. Diffusion through these pores is
driven by concentration gradients; when concentrations are high enough, nutrients
non-specifically diffuse against their concentration gradient into the periplasm.
Nutrients are then transported across the CM by channels such as the
mechanosensitive channel, MscL (28) (Figure 1.1).
1.2.2 Facilitated transport
Facilitated diffusion is a second mechanism of porin-mediated nutrient
transport. Less concentrated and rare nutrients that cannot diffuse against a
concentration gradient are transported by this mechanism. The maltosaccharide
transporter, LamB, is a canonical facilitated diffusion porin.
13
Figure 1.2. E. coli OM nutrient transporters. Structures are displayed as ribbon representations
and viewed laterally from the plane of the OM (top row) or viewed from the extracellular space
into the transporter lumens (bottom row). For clarity, monomeric lateral views are displayed for
OmpF and LamB. Where present, transporter ligands are displayed as surface representations and
metals are displayed as grey spheres. Left: Passive cation transporter OmpF (PDB code 2J1N);
Middle: facilitated-diffusion maltosaccharide transporter LamB (PDB code 1MPM); Right: active
ferric-siderophore transporter FhuA (PDB code 1QFF).
LamB folds into an 18-stranded β-barrel channel (29) that traverses the OM and,
like OmpF, forms trimers (Figure 1.2). A high affinity sugar binding site is
created by three extracellular loops that fold into and constrict LamB’s lumen
(30). Sugars, such as maltose, bind within this site and are funnelled further into a
lumenal arrangement of residues that facilitate diffusion into the periplasm.
Maltose is then bound by periplasmic maltose-binding protein and delivered to
CM-embedded permease MalFG/K2 (Figure 1.1). ATP hydrolysis facilitates
maltose translocation through MalFG/K2 and into the cytoplasm.
Another OMP that facilitates nutrient diffusion is the OM long chain fatty
acid receptor, FadL. Like the porins described above, FadL forms a
transmembrane β-barrel, which possesses an extracellular, high affinity binding
14
site for long chain fatty acids. However, FadL embeds within the OM as a
monomer. Diffusion of fatty acids into the OM is postulated to occur through a
channel within FadL that extends into the barrel lumen. A portal along one side
of the barrel, and which connects to the lumenal channel, is the postulated access
route to the OM outer leaflet (31). Unlike ions and maltosaccharides, the fate of
fatty acids transported through FadL is less known. Fatty acids might
spontaneously diffuse across the cytoplasm or bind to an unidentified binding
protein. However, little evidence for these mechanisms has been reported.
1.2.3 Active transport
Mechanisms of passive and facilitated diffusion enable uptake of small
nutrients. To facilitate transport of nutrients larger than 700 Da, such as iron
bound to siderophores (discussed in section 1.4), active transport mechanisms are
required. The best-characterized active transporters include CM-embedded
permeases, such as MalFG/K2, which couple active transport to ATP hydrolysis.
However, there is no energy source available to OM transporters. Bacteria have
overcome this limitation by transducing CM-derived energy across the cell
envelope and to the OM. The CM-embedded TonB–ExbB–ExbD multi-protein
complex (Figure 1.1) harnesses and stores energy from the proton motive force
(PMF) and transduces this energy to TonB-dependent iron transporters found in
the OM, such as the ferric-hydroxamate receptor FhuA (Figure 1.2). Mechanisms
of active transport are now discussed in greater detail.
15
1.3 General bacterial strategies for iron acquisition
Iron acquisition is a bacterial paradox. Most require it for survival, yet it
is present in miniscule amounts and its acquisition requires passage through
barriers generated by the cell envelope. To overcome this limitation, bacteria
have evolved specialized iron transport systems. The specific system employed
by a particular bacterium depends on the oxidative state of iron found within that
bacterium’s environment; specialized transport systems exist for uptake of ferrous
iron and for uptake of ferric iron.
1.3.1 Ferrous iron transport
Bacteria that occupy anaerobic or micro-aerobic niches such as the
gastrointestinal tract can obtain soluble ferrous iron by ionic transport systems.
The best-characterized ferrous ion uptake system is the E. coli Feo system, which
is utilized to colonize the gastrointestinal tract (32). A detailed understanding of
how this system operates has yet to be realized. Poorly understood factors include
the mechanisms of ferrous iron transport across the OM and periplasm.
Presumably, ferrous iron translocates into the periplasm through an unidentified
OM porin. An unidentified periplasmic binding protein may then transport
ferrous iron to the CM.
Better understood is the mechanism of ferrous iron transport across the
CM. FeoB, a CM-embedded protein is postulated to actively transport ferrous
ions into the cytoplasm (33). However, unlike the previously described CM
permeases, FeoB displays activity only when coupled to GTP hydrolysis. Other
16
proteins including FeoA and FeoC may facilitate FeoB-mediated transport, yet
their exact roles require further investigation.
1.3.2 Ferric iron transport
Bacteria that obtain ferric iron have evolved different acquisition
strategies. One strategy is to directly acquire iron from host iron-binding proteins.
By capturing proteins such as transferrin, lactoferrin and hemoglobin, bacteria can
exploit rich sources of iron. Specialized TonB-dependent transporters (TBDTs),
such as the Neisseria meningitidis and Neisseria gonorrhoeae transferrin receptor,
TbpA, bind transferrin at the bacterial surface and uptake ferric iron after
stripping it from transferrin (34,35). In addition, a lactoferrin receptor from
Helicobacter pylori has been described (36) that binds ferric-lactoferrin and
uptakes the ferric iron, presumably by a mechanism similar to TbpA.
Synthesis and extracellular secretion of hemophore proteins is a second
bacterial ferric iron acquisition strategy. Hemophores, such as HxuA from
Haemophilus influenza, and HasA from various bacteria including Serratia
marcescens, Pseudomonas aeruginosa, and Yersinia pestis (37) bind to and strip
heme from host proteins hemopexin and hemoglobin, respectively. Having
stripped heme, HxuA and HasA present it to cognate bacterial surface TBDTs.
Heme is subsequently released from hemophore and transported through the
TBDT into the periplasm.
Bacteria also synthesize and secrete siderophores into their extracellular
environment. Siderophores represent a large and diverse class of small organic
17
molecules that possess extremely high affinities for ferric iron, allowing them to
strip and bind iron from minerals and from host iron-binding proteins. Bacterial
ferric-siderophore transport is the focus of this thesis; having bound iron, ferric-
siderophores are transported back into the bacterium using mechanisms described
in the remainder of this review. In general, transport (Figure 1.1) entails capture
of ferric-siderophore by an OM siderophore transporter, such as the ferric-
hydroxamate receptor, FhuA (discussed in section 1.5). By a TonB-dependent
mechanism, siderophores are then actively transported through the receptor and
into the periplasm. Periplasmic binding proteins such as the ferric-hydroxamate-
binding protein, FhuD, then bind the siderophore and deliver it to a CM-
embedded permease, such as the ferric-hydroxamate permease FhuB/C. Given
the diversity of siderophores produced and employed during bacterial iron
acquisition, their properties are now discussed in greater detail.
1.4 Siderophores
Siderophores represent a diverse class of small organic molecules (Mr < 1
kDa) that bind iron with extremely high affinity. Affinities between siderophore
and ferric iron vary; association constants (Ka) have been reported between 1022
to
1052
(38,39). Siderophores are synthesized and secreted by bacteria and fungi
alike and help in the colonization of their environments. Given the importance of
solubilizing and chelating iron for colonization, the abilities to produce and utilize
siderophores are known as virulence factors for many pathogenic bacteria. Over
three hundred siderophores have been characterized, and are classified by the
18
chemistries used to chelate ferric iron. Despite the diversity of scaffolds that
comprise siderophore backbones, most chelate iron with bidentate oxygen
functional groups. Most siderophores belong to one of four different classes:
catecholates (including phenolates), hydroxamates, α-hydroxycarboxylates and
mixed-types (40).
Perhaps the best-characterized siderophore is the catecholate, enterobactin,
the only endogenous siderophore produced by E. coli K-12. The enterobactin
backbone comprises three linked L-serine residues, each derivatized with iron-
chelating di-hydroxybenzoic acid groups. The three serine residues incorporate
into a tri-lactone macrocycle to form a symmetrical scaffold (Figure 1.3).
Phenolate siderophores are similar to catecholates, except for use of a mono-
hydroxybenzoic acid functional group. Examples of phenolate siderophores
include yersiniabactin from Y. pestis and pyochelin from P. aeruginosa.
Produced exclusively by fungi, the hydroxamate-type siderophore
ferrichrome, represents one of the first-characterized microbial iron chelator (6).
Structural scaffolds of hydroxamate siderophores consist of two tripeptides: a
single tripeptide (containing glycine, alanine or serine) linked to a second
tripeptide of derivatized L-ornithine functional groups (Figure 1.3). Iron
chelation occurs within the hydroxamate groups of derivatized ornithines.
Carboxylates represent the third class of siderophores used to chelate iron.
Chelation from this class of siderophores occurs from α-hydroxycarboxylic acid
functional groups. Examples include staphyloferrin A from Staphylococcus
species and achromobactin from Erwinia chrysanthemi. In addition, citrate, a
19
metabolic intermediate, efficiently chelates iron and is transported into bacteria
(Figure 1.3).
Siderophores also incorporate mixed functional groups into their scaffold.
Many mixed siderophores have been described and include such combinations as
citrate-hydroxamate functional groups in aerobactin (Figure 1.3), citrate-
catecholate functional groups in petribactin and catecholate-hydroxamate in
heterobactin B.
Figure 1.3. Siderophores transported by E. coli. A. The catecholate siderophore, enterobactin.
Enterobactin is the only endogenous siderophore produced by E. coli K-12; B. the hydroxamate
siderophore, ferrichrome. R-groups represent sites of species-specific, variable chemical
modifications; C. the carboxylate, citrate. Citrate is a naturally occurring metabolic by-product;
D. the mixed-type siderophore, aerobactin. Aerobactin chelates iron using a combination of
hydroxamate and carboxylate chemistries. Images are adapted from reference 6.
20
1.4.1 Siderophore biosynthesis
Siderophores are synthesized from assemblies of cytoplasmic enzymes
that create bonds using non-ribosomal peptide biosynthesis. Modular enzymes
comprise the siderophore biosynthetic machinery; separate domains catalyze step-
wise assembly of siderophores without need for an RNA template (38). Genes
that encode siderophore biosynthetic enzymes localize in operons with
siderophore-specific locations. For example, in E. coli, genes for enterobactin
biosynthesis cluster within an operon at a single chromosomal locus shared by
genes encoding enterobactin uptake proteins (38). In contrast, genes for synthesis
of vibriobactin from Vibrio cholerae localize within one of its two chromosomes,
but stagger within two different gene clusters, separated by nearly 100 bp (38).
Other bacteria such as the fish pathogen Vibrio anguillarum harbour virulence
plasmids that encode genes for siderophore biosynthesis, in addition to the
cognate siderophore uptake proteins (38).
1.4.2 Siderophore secretion
Once synthesized, siderophores are secreted into the extracellular
environment. Compared to fungi, less is known about secretion of bacterial
siderophores. Enterobactin secretion by E. coli has been studied. The CM-
embedded protein EntS is known to translocate enterobactin across the CM into
the periplasm (40). EntS belongs to the major facilitator superfamily of proteins
that exhibit a broad range of export activities. Once in the periplasm, enterobactin
is postulated (40) to translocate to the extracellular space by transport through
21
OM channel TolC (Figure 1.1). Having translocated through TolC and scavenged
iron with high affinity, ferric siderophores are then transported back into the
bacterium by various TBDTs, discussed below.
1.5 TonB-dependent transporters
Many TBDTs have been discovered. As high affinity receptors, each
binds and transports specific siderophores. TBDTs expressed in E. coli include
the ferric-hydroxamate receptor FhuA, the ferric-enterobactin receptor FepA, and
the ferric-citrate receptor FecA. In addition, TBDTs also transport rare metallo-
nutrients such as the E. coli vitamin B12 (cobalamin) receptor BtuB. Many
TBDTs are also exploited as receptors for entry of various phage and antibacterial
compounds. The remainder of this section focuses on structural knowledge
gained from TBDTs with emphasis placed on the E. coli transporter FhuA.
1.5.1 Structures of TonB-dependent transporters
FhuA is receptor for ferric hydroxamate-type siderophores such as
ferrichrome. In addition, it serves as receptor for bacteriophages T1, T5, Φ80,
UC-1, and the antibacterial compounds colicin M, microcin-25, albomycin and
rifamycin CGP 4832 (41). FhuA is a 79 kDa protein possessing 714 residues that
fold into two unique domains. Residues 1–160 form a globular cork domain that
inserts into a C-terminal, 22-stranded β-barrel domain comprised of residues 161–
714. FhuA was the first TBDT structure (1998) to emerge at atomic-resolution
22
detail (Figures 1.2 and 1.4) and revealed a canonical fold that is conserved
amongst all TBDTs with known structure (42).
FhuA’s structure revealed many striking features. First, unlike smaller
porins that form trimers in the OM, the structure of FhuA revealed a monomer.
Bound to FhuA was a single LPS molecule that indicated how this class of lipids
can interface and surround the outer circumference of an OMP (43). The most
striking observation revealed that FhuA’s globular cork, comprising a four-
stranded β-sheet surrounded by four short α-helices, fits tightly within the barrel
domain (Figure 1.2). No obvious channel was observed within the cork and
barrel, highlighting the need for structural rearrangement in order for siderophore
transport to occur.
Structures of many E. coli TBDTs have since been determined (Figure
1.4), including those of FepA (44), FecA (45), BtuB (46-49), and the Colicin Ia
receptor Cir (50). Structures of TBDTs from P. aeruginosa have also been
solved (Figure 1.4), including the ferric-pyochelin transporter FptA (51) and the
ferric-pyoverdine transporter FpvA (52). Most recently, TBDT structures have
been solved from other organisms including the hemophore receptor HasR (53)
from S. marcescens and the ferric-alcaligin transporter FauA (54) from
Bordetella pertussis (Figure 1.4).
Structural features of TBDTs are similar. All share architectures
comprised of cork domains that fold and insert into β-barrel domains. The β-
barrel domains nearly superimpose; connecting subsequent barrel strands are short
periplasmic turns and longer extracellular loops. Siderophores bind within sites
23
Figure 1.4. Structures of TonB-dependent transporters. Displayed are ribbon representations of
TBDTs. Cork domains are coloured separately from barrel domains. Ligands are represented as
surface models where present.
comprised of each receptor’s extracellular loops and cork apices. Structural
differences arise mainly from the lengths of extracellular loops and from
specialized siderophore binding sites. In all structures, the corks prevent passive-
or facilitated-diffusion of siderophore through receptor lumens. This necessitates
active transport by interaction with TonB; interactions between TonB and TBDTs
facilitate opening of a lumenal pore large enough to allow siderophore
translocation.
1.5.2 Ton boxes
All TBDTs share a common N-terminal feature called the “Ton box” that
functions as a molecular recognition motif for TonB. TonB initially interacts with
siderophore bound-receptors through this motif (discussed in greater detail later).
24
Though not strictly conserved, Ton boxes generally comprise the consensus:
D/ETX1X
2VX
3A, where X
1 and X
2 are hydrophobic amino acids and X
3 is any
amino acid (55). Ton boxes are unresolved in most TBDT structures, but were
resolved as extended conformations in the structures of FepA, BtuB and CirA.
Noteworthy, is that the conformation of BtuB’s Ton box differed when
crystallized in a lipid cubic phase, highlighting both the dynamic nature of the
Ton box, as well as its conformational dependence on chemical environment, a
point re-visited below.
1.5.3 Transcriptional regulatory domains
Transporters such as FecA and FpvA have additional extensions N-
terminal to their Ton boxes that function as transcriptional signalling domains.
Through mechanisms not completely understood, the signalling domains help
regulate transcription of genes in the transporter biosynthetic operon. Structures
of isolated FecA signalling domains have been solved by NMR (56,57) and of the
FpvA signalling domain bound to the transporter by X-ray crystallography (58).
The structures revealed nearly identical folds, comprising two antiparallel β-
sheets flanked by alternating α-helices. The structure of FpvA with its intact
signalling domain (Figure 1.4) exhibited unresolved space near the receptor’s Ton
box. This was interpreted to mean that space occupied by a transporter’s
signalling domain should not spatially occlude a region that TonB is known to
bind.
25
1.5.4 Siderophore-induced conformational changes
The structure of FhuA revealed substantial conformational changes that
occur upon binding the ferric-hydroxamate siderophore, ferrichrome.
Ferrichrome binds to an extracellular surface, within a hydrophobic pocket
formed by FhuA’s three cork apices. In addition, extracellular loop 3 and β-
strands 7 and 9 contribute to the binding site. Ferrichrome makes various
hydrogen bond and van der Waals contacts with conserved FhuA residues.
Binding is strong, with a measured KD of around 200 nM (42).
Upon binding ferrichrome, FhuA exhibits conformational changes that
propagate to the receptor’s periplasmic surface. While the barrel conformation
only slightly changes (RMSD 0.4 Å between apo- and ferrichrome-bound FhuA),
the cork undergoes significant rearrangement. Compared to apo-FhuA, cork
apices translocate approximately 2 Å towards the iron atom when ferrichrome is
bound. The most striking conformational changes localize to FhuA’s periplasmic
face. Upon binding ferrichrome, residues 24–29 (termed the switch helix) extend
and translocate away from a helix-stabilizing hydrophobic groove (Figure 1.5).
Translocation is considerably large; residue Trp-22 displaces approximately 17 Å
from its location in the apo-FhuA state. This structural transition is postulated to
signal a state of ligand occupancy that extends into the periplasm, whereby TonB-
dependent energy transduction begins. However, it is unclear whether this
transition is a crystallization artefact. It has been reported that crystallization
solutes can alter conformations of membrane proteins (59). Furthermore, an
26
Figure 1.5. FhuA exhibits periplasmic conformational changes upon binding ferrichrome.
Detailed ribbon representations of apo-FhuA (PDB code 1QFG, left) and ferrichrome-bound FhuA
(PDB code 1QFF, right) are displayed as viewed from the periplasm, looking into FhuA’s lumen.
FhuA structural elements are coloured as follows: β-barrel domain (blue), cork domain (orange),
N-terminal switch helix residues (green). Ferrichrome is displayed as a red surface representation.
For clarity, extracellular loops are not shown.
electroparamagnetic resonance spectroscopy (EPR) study indicated that in
solution, both apo- and ferrichrome-bound FhuA’s switch helix remains unwound
and in an extended conformation (60).
Shortly after the ferrichrome-bound structure, additional FhuA structures
with various bound ligands were solved. These included the natural siderophore-
antibiotic, albomycin-bound structure (61), the siderophore, phenylferricrocin-
bound structure (61), and the antibiotic-siderophore conjugate, rifamycin CGP
4832-bound structure (62). All ligands occupied regions proximal to the
ferrichrome binding site. Slight differences were observed between binding
modes for the different ligands.
As with ferrichrome, albomycin and phenylferricrocin both elicited the
same periplasmic conformational change, whereby FhuA’s switch helix unwound
and translocated nearly 17 Å away from its original position. However, due to a
slightly different binding mode of rifamycin CGP 4832, the switch helix
27
conformational change was not observed; its binding does not affect the position
of FhuA’s cork apex A. In contrast, binding of the other ligands causes upward
movement of apex A, which propagates distally to the periplasmic face and causes
unwinding of FhuA’s switch helix.
Structures of other ligand-bound TBDTs from E. coli indicated that
binding mechanisms are not strictly conserved. Some similarities to FhuA were
apparent; binding determinants generally involve use of cork apices and
extracellular loops. However, each receptor tends to use a different combination
to fulfill this purpose. In addition, ligand-induced conformational changes differ
slightly for each transporter. For example, the structure of ferric-enterobactin
transporter, FepA, in its apo-state, failed to provide evidence for a switch helix
conformational change (44). The significance of this outcome is not apparent;
bound enterobactin was never resolved, despite having soaked FepA crystals with
the ligand. Assuming that enterobactin at least partially occupied the crystal
lattice, it was concluded that unwinding of a receptor’s switch helix is not a
conserved mechanism for signalling ligand occupancy. Despite uncertainties
associated with the FepA structure, its Ton box was resolved and provided first
evidence that the motif tends to adopt extended and loosely structured
conformations.
Structures of the ferric citrate-bound transporter, FecA, emulate the
ligand-binding features exhibited by FhuA; three cork apices and different
combinations of extracellular loops bind ferric-citrate (45,63). In contrast, FecA’s
siderophore binding site is hydrophilic and unlike FhuA, FecA’s extracellular
28
loops 7 and 8 exhibit large conformational changes upon binding ferric-citrate that
close over the ligand binding site, rendering it solvent-inaccessible. Like FhuA,
apo-FecA possesses an N-terminal switch helix that partially unwinds upon
binding ferric-citrate, suggesting similar ligand-induced signalling mechanisms.
Further supporting this mechanism was the observation that FecA’s Ton box,
partially resolved and extended in the apo-structure, undergoes a structure-to-
disorder transition upon binding ferric citrate. Presumably, the disordered Ton
box is flexible and further extended towards the periplasm where it can interact
with TonB.
Structures of TonB-dependent cobalamin transporter, BtuB, with various
ligands bound revealed further differences between ligand binding modes of the
various TBDTs. A structure of cobalamin-bound BtuB demonstrated that cork
apices and five extracellular loops contact the ligand (46). Compared to apo-
BtuB, cork residues 85–96 shift upward by nearly 6 Å towards bound cobalamin.
BtuB also serves as receptor for colicin E2. A BtuB structure in the presence of
colicin E2’s BtuB-binding domain demonstrated remarkably similar binding
modes between BtuB and the colicin E2-derived ligand as compared to BtuB and
cobalamin (49). Main differences include residues contacting the colicin E2-
derived ligand and induced ordering of loops 5 and 6, which are disordered in
apo- and cobalamin-bound BtuB. BtuB’s mechanism of ligand-induced signalling
is also likely different since, unlike FhuA, FepA and FecA, residues with a switch
helix-like conformation were not present in BtuB. However, the order-to-disorder
transition of ligand signalling may be conserved; upon binding cobalamin, BtuB’s
29
Ton box undergoes a rotation of nearly 180° and becomes slightly more
disordered.
Despite modest differences between ligand binding modes among TBDTs,
all exhibit ligand-induced alterations proximal to their Ton boxes. These
structural alterations are postulated to function as a periplasmic signal of ligand
occupancy that initiates transport. Further support for this mechanism derives
from EPR studies of various TBDTs. When coupled to EPR spectroscopy, the
technique of site-directed spin labelling allows direct measurements of protein
structural dynamics in solution. This approach confirmed that Ton boxes of FecA
and BtuB undergo order-to-disorder transitions upon ligand binding (60).
Unexpectedly, these methods demonstrated that FhuA’s Ton box is always
disordered and that the switch helix conformation observed in crystal structures
may be an artefact of the crystallization conditions. A second EPR study
demonstrated that, in addition to the cobalamin-induced order-to-disorder
transition, ligand binding promotes extension of BtuB’s Ton box by nearly 30 Å
into the periplasm (64). Despite different conclusions between crystallographic
and biophysical studies, all suggest that ligand binding within extracellular loops
of TBDTs propagates a periplasmic signal for recruitment of TonB.
1.6 The TonB–ExbB–ExbD complex
Ligand-bound TBDT structures failed to reveal obvious lumenal
translocation pathways; translocation requires energy input. The CM-embedded
TonB–ExbB–ExbD complex transduces energy to ligand-bound TBDTs. By
30
harnessing chemiosmotic potential of the CM-derived PMF, the TonB–ExbB–
ExbD complex conformationally energizes TonB so that it can transduce stored
energy to ligand-bound TBDTs. The proteins and interactions that comprise the
TonB–ExbB–ExbD complex (Figure 1.6) are now examined in greater detail.
1.6.1 TonB: Energy transducer
TonB transduces stored energy to ligand-bound TBDTs. Originally, TonB
was thought to provide energy solely for iron and cobalt (via cobalamin) uptake.
However, novel TonB-dependent activities have been reported including uptake
of nutrients such as Ni, sucrose and sulphate (65-67). Properties of TonB vary
greatly between bacteria and include differences in length and the number of
expressed isoforms (68). TonB from E. coli is a 239-residue protein with a
molecular weight of approximately 25 kDa (Figure 1.6). It contains three
domains: a single 32-residue N-terminal transmembrane (TM) domain, a central
proline-rich region and a structured C-terminal domain. TonB interacts with
ExbB and ExbD through its TM helix. The remainder is periplasmic. Residues
33 to around 150 are primarily unstructured and include the central proline-rich
domain (residues 66–100) consisting of alternating Lys-Pro, Glu-Pro repeats.
TonB’s C-terminal domain forms a structured region that directly interacts with
TBDTs. These elements are now examined in greater detail.
31
Figure 1.6. The TonB–ExbB–ExbD complex. Schematic representation of the TonB–ExbB–
ExbD complex is illustrated. Transmembrane α-helical domains are represented by cylinders.
Protein N- and C-termini are labelled accordingly. Structured regions of TonB’s and ExbD’s
periplasmic C-terminal domains are represented by ovals. TonB’s central, periplasmic proline-
rich region, is represented by a series of curved blue boxes. Directional flow of protons through
the complex is highlighted in red. Periplasmic and cytoplasmic compartments are labelled
accordingly. Proteins are not drawn to scale. For clarity, only one of each protein is displayed.
1.6.2 TonB proline-rich region
Residues 66–100 of E. coli TonB form a central proline-rich region.
Synthetic oligopeptides corresponding to proline-rich regions from E. coli and S.
typhimurium were characterized by NMR spectroscopy (69,70). Both peptides
were found to be elongated and rigid in solution and were postulated to function
as a linker enabling TonB to interact with ligand-bound OM transporters.
However, the exact role of this region is speculative; in some bacteria the
corresponding region is not as proline-rich as E. coli TonB, and in other species
the region can include up to 283 residues of Pro repeats (68). Furthermore, the
periplasm
cytoplasm
32
region can be deleted without greatly affecting TonB activity in vivo (71). As
discussed below, some studies attribute properties to the region that affect
interactions between TonB and its protein binding partners.
1.6.3 TonB C-terminal regions form a compact structure
C-terminal residues of TonB form a structured domain that is essential for
interactions with OM transporters. Crystal structures of various E. coli
periplasmic TonB derivatives have been crystallized and revealed unique, length-
dependent properties (Figure 1.7). The first C-terminal TonB structures of
derivatives possessing residues 155–239 and residues 162–239 revealed identical,
tightly intertwined dimers (72,73). Compared to shorter derivatives, a slightly
longer derivative that possessed residues 148–239 also crystallized as a dimer, but
much more loosely associated (74). By NMR, an even longer derivative that
possessed residues 103–239 was monomeric in solution (75). In agreement, a
similar TonB derivative from the fish pathogen V. anguilerum that possessed
residues 121–206 was also monomeric by NMR (76).
1.6.4 Oligomeric state of TonB
The in vitro and in vivo oligomeric states of TonB have been thoroughly
investigated. The TonB crystal structures of various periplasmic derivatives
demonstrated a length-dependent tendency to dimerize. Analytical
33
Figure 1.7. Structures of C-terminal, E. coli TonB derivatives. Structures are aligned according to
the NMR structural orientation (right) and displayed as ribbon representations. Dimeric TonB
structures are coloured as follows: Chain A (blue) and Chain B (green) Left: TonB residues 155–
239 form tightly intertwined dimers (PDB code 1IHR); Middle: TonB residues 148–239 form
loosely associated dimers (PDB code 1UO7); Right: TonB residues 152 -239 forms a monomer
(PDB code 1XX3).
ultracentrifugation (AUC) studies later confirmed these findings; derivatives of
TonB that possess residues 153–239 and fewer sediment as dimers, whereas TonB
derivatives possessing residues 143–239 and longer sediment as monomers
(73,77). The findings that longer TonB derivatives are monomeric in vitro,
suggests that in vivo TonB is also monomeric. However, one study that fused the
cytoplasmic domain of ToxR to TonB’s N-terminus concluded that TonB is
dimeric in vivo (78). The TonB fusions supported TonB-dependent functions and
promoted transcription of the cholera gene, an outcome possible only if the ToxR
domains formed dimers (and by analogy, TonB formed dimers). A second study
confirmed this finding by mutating selected aromatic residues in TonB’s C-
terminus to cysteine; TonB dimers spontaneously formed in a PMF-dependent
manner (79). These conflicting outcomes between in vitro and in vivo
characterizations have yet to be reconciled.
34
1.7 Protein–protein interactions involving TonB
Interactions between TonB and its E. coli protein partners are now
described in three parts. Interactions between TonB and ExbB/ExbD are
discussed first. Second, interactions between TonB and TBDTs, a subject of
intense investigation, are discussed in greater detail. Finally, interactions between
TonB and other cell envelope proteins with no obvious physiological roles are
described.
1.7.1 TonB–ExbB–ExbD interactions
Interactions between TonB and CM-embedded ExbB and ExbD proteins
are required for energy transduction across the cell envelope. TonB, ExbB and
ExbD are all membrane proteins that possess varying numbers of TM helices.
Compared to TonB, less is known about the ExbB and ExbD proteins. Both are
TM proteins with contrasting properties. ExbB from E. coli is a 24 kDa protein
comprised of 244 residues that form three TM α-helices connected by loops of
varying length (Figure 1.6). ExbB’s longest loop, which connects TM helices 1
and 2, was demonstrated to project into the cytoplasm (80) where it forms the
bulk of ExbB’s non-membrane-embedded composition.
ExbD is a smaller 16 kDa protein comprised of 141 residues that form a
single TM α-helix. The bulk of ExbD extends into the periplasm (Figure 1.6). A
structure of ExbD’s periplasmic domain from E. coli was solved by NMR (81)
and revealed a highly flexible protein that exhibited structure only at low pH.
35
Unexpectedly, the fold of ExbD’s periplasmic domain was similar to structural
elements of periplasmic siderophore binding proteins FhuD and CeuE.
Together, TonB, ExbB and ExbD form a complex that harnesses the
chemiosmotic potential of the PMF. The TonB–ExbB–ExbD complex must
therefore generate TM channels that facilitate proton translocation. This large
multi-protein assembly is postulated to oligomerize into a complex with a
stoichiometry of 1:7:2 (TonB:ExbB:ExbD) (82). However, aside from this
postulated stoichiometry, less is known about how the proteins assemble in the
CM. A computational model for TM assembly of the complex has been reported
(83). Sequence and domain conservation between ExbB/D and MotA/B proteins
provided a foundation for development of a homologous TonB–ExbB–ExbD
complex. MotA and MotB form a CM-embedded complex that harnesses PMF to
drive rotation of bacterial flagella. Models that matched TM components of
MotA/B with TM components of TonB–ExbB–ExbD enabled TM domain
alignment such that a plausible proton channel was visualized. Importantly,
ionizable residues within the model’s channel and that could promote proton
translocation include conserved ExbB residues. When mutated, these residues
have demonstrable effects on TonB-dependent transport (84). In addition,
mutations of conserved and ionizable TonB residues that line the putative channel
also affect TonB-dependent transport (85,86).
Interactions between TonB and ExbD were recently characterized in vivo.
Both protiens formed PMF-dependent and periplasm-localized, formaldehyde
cross-linked heterodimers, indicating that TonB–ExbD interactions are not solely
36
localized to their TM domains (87). Periplasm-localized regions of ExbD also
formed cross-linked homodimers. However, unlike formation of TonB–ExbD
heterodimers, ExbD homodimers formed, independent of PMF.
1.7.2 TonB–OM receptor interactions
TonB interacts with TBDTs within at least two distinct localizations: Ton
box regions and cork/β-barrel regions. TonB–FhuA interactions were originally
demonstrated in vivo by complementary suppressor mutations between FhuA and
TonB (88); interactions localized binding of TonB to FhuA’s Ton box. Later
studies using formaldehyde cross-linking and biochemical assays confirmed that
TonB bound to FhuA in vitro and to FepA in vivo (89,90) and localized binding of
TonB to FepA’s Ton box (91). In addition, TonB–BtuB Ton box interactions
were extensively studied. By site-directed, disulfide cross-linking and
mutagenesis, the Ton box of BtuB was demonstrated to interact with TonB around
TonB residue 160 in a conformation-dependent manner (92,93). Most recently,
Ton box interactions were studied by NMR (75) and suggested that TonB’s C-
terminal residues form domain-swapped, β-strand interactions upon binding to
Ton boxes.
Global interactions between TonB and FhuA have been the subject of
biophysical characterizations. Stoichiometries and affinities of periplasmic TonB
derivatives bound to FhuA were characterized by AUC and by SPR (77). By
AUC, two periplasmic TonB derivatives possessing either residues 33–239 or C-
terminal residues 155–239 both formed 2:1 TonB–FhuA complexes with FhuA
37
that were enhanced in the presence of siderophore. SPR analyses further revealed
two TonB populations bound to FhuA: a TonB C-terminal, weaker affinity
population and a TonB N-terminal, higher affinity population. When FhuA bound
siderophore, TonB’s C-terminal affinity decreased and its N-terminal affinity
increased.
Deletion of TonB’s proline-rich region reduced the stoichiometry of the
TonB–FhuA complex to 1:1 (94), but did not alter its affinity for FhuA,
suggesting that the region is essential to form the previously observed, 2:1 TonB–
FhuA complex. Further refined SPR analyses (95) demonstrated that TonB–FhuA
interactions are likely to be sequential; a single TonB monomer is recruited to the
periplasmic surface of FhuA, followed by structural rearrangement that recruits a
second monomer.
Crystal structures of periplasmic TonB fragments bound to FhuA (96) and
to BtuB (97) ultimately demonstrated the determinants of binding between these
three proteins (Figure 1.8). Both structures confirmed findings from earlier
studies suggesting that TonB residue 160 interacted with receptor Ton boxes and
further confirmed that this interaction was through β-strand complementation
between TonB’s C-terminus and Ton boxes of FhuA/BtuB. The structures were
remarkably similar and revealed 1:1 TonB–FhuA and TonB–BtuB
stoichiometries. In both structures, TonB occupies approximately one-half of the
periplasmic cork/β-barrel surfaces. Despite the novelty of confirming the modes
of binding between TonB and periplasmic receptor surfaces, the structures did not
38
Figure 1.8. Crystal structures of TonB bound to FhuA and BtuB. Crystal structures of TonB–
FhuA and TonB–BtuB are aligned and displayed as ribbon representations. A. Lateral view of the
TonB–FhuA complex (PDB code 2GRX) as viewed from the OM plane. Protein domains are
coloured as follows: FhuA cork (orange), FhuA β-barrel (blue), TonB (cyan). Bound ferricrocin is
displayed as a red surface representation; B. Lumenal view of TonB–FhuA complex as viewed
from periplasm; C. Lateral view of TonB–BtuB complex as viewed from the OM plane (PDB
code 2GSK). Protein domains are coloured as follows: BtuB cork (green), BtuB β-barrel (orange),
TonB (cyan). Bound cobalamin is displayed as a magenta surface representation; D. Lumenal
view of TonB–BtuB complex as viewed from periplasm.
reveal further perturbations that could explain how energy transduction through
TonB elicits structural rearrangement of receptor corks.
Though not resolved in the TonB–FhuA structure, conformations of FhuA
extracellular loops are postulated to change upon binding siderophore or TonB. A
study combining phage display, SPR and fluorescence spectroscopy demonstrated B
A B
39
conformational changes within FhuA extracellular loops 3, 4 and 5 upon binding
ferricrocin or TonB (98). Since these changes were not observed in apo- and
siderophore-bound FhuA crystal structures, FhuA must exhibit structural
dynamics not accounted for in the crystal structures. These dynamics likely result
in gradual closing of loops over FhuA’s siderophore binding site, similar to the
changes observed for FecA and BtuB upon binding of their respective ligands.
An NMR and infrared spectroscopy study has attempted to characterize
binding between the isolated cork domain of S. typhimurium FepA and E. coli
TonB (56). The FepA cork domain was completely unfolded in solution, yet it
still bound TonB. Interaction did not promote secondary structure formation
within the cork and probably resulted from interactions between FepA’s Ton box
and TonB. However, it is unclear if the lack of TonB–induced cork folding was
due to ortholog incompatibility. From this study, it appears that interactions
between TonB and TBDTs localize primarily to regions accounted for in the
TonB–FhuA and TonB–BtuB crystals structures.
1.7.3 TonB–cell envelope protein interactions
In addition to interacting with TBDTs, TonB has been postulated to
interact with other components of the cell envelope. Formaldehyde cross-linking
studies of fractionated E. coli cell compartments, indicated that TonB formed
complexes with OmpA and Lpp (99). However, the biological significance of
these interactions remains unanswered. Interactions between TonB and Lpp are
particularly interesting considering a recent investigation suggesting that TonB’s
40
C-terminus possesses a structural motif similar to YcfS, an E. coli proline-rich,
peptidoglycan-binding protein (100). The study demonstrated that, like YcfS,
periplasmic TonB derivatives also bind peptidoglycan. Together, these findings
suggest that C-terminal TonB motifs may enable “surveillance” of the OM inner
leaflet through combinations of interactions with OmpA, Lpp and peptidoglycan,
or perhaps all three.
1.8 Mechanisms of TonB–dependent energy transduction
Currently much is known about structural properties of TonB and the
TBDTs to which it transduces energy. However, there is no consensus
understanding of how TonB transfers energy across the cell envelope. It is
recognized that the TonB–ExbB–ExbD complex harnesses PMF to
conformationally charge TonB, but the molecular details of this process are not
elucidated. Three models of TonB energy transduction are now described: the
shuttle model, the rotation model, and the mechanical pulling model.
1.8.1 The shuttle model
In the shuttle model, TonB is postulated to undergo a process of CM-
localized and PMF-dependent energy excitation by interacting with ExbB and
ExbD (Figure 1.9). TonB remains conformationally charged until a TBDT signals
siderophore-occupancy through extension of its Ton box into the periplasm.
TonB then engages the TBDT and is physically released from the CM, whereupon
it fully engages siderophore-bound TBDT at the OM. Energy release then occurs
41
through an unknown mechanism and is transduced to the transporter. This causes
rearrangement of the cork and allows siderophore passage into the periplasm.
After transport, TonB is recycled back to the CM, presumably by interacting with
ExbB and ExbD. Evidence for this model originally derives from cell envelope
fractionation studies. While TonB was always detected in the CM fraction, a
significant proportion (~ 40% of total ) could always be detected in the OM (101).
Deletion of ExbB and ExbD prevented recycling of TonB back to the CM and
promoted a population of TonB that was then only detected in the OM fractions.
A second study provided in vivo evidence to support this hypothesis by
engineering a TonB derivative with a single cysteine introduced at TonB’s N-
terminus. During energy transduction, this N-terminal cysteine became labelled
with the thiol-reactive fluorescent dye Oregon green, indicating that TonB’s N-
terminus had escaped the CM and was accessible to labelling (102).
Figure 1.9. Shuttle model of TonB-dependent energy transduction. A. Siderophore binding to an
OM embedded TBDT signals occupancy by extending the receptor’s Ton box into the periplasm.
Within the CM, the TonB–ExbB–ExbD complex harnesses PMF; B. TonB becomes
conformationally charged; C. TonB escapes the CM and engages ligand-bound receptor. By an
unknown mechanism (denoted with a question mark), TonB elicits conformational
changes within the receptor that enables ligand translocation into the periplasm; D. Discharged
TonB disengages the receptor; E. TonB is recycled back to the CM by ExbB/ExbD and the PMF.
42
The shuttle model has undergone much scrutiny and criticism. In a recent
study, TonB hybrids were generated, which fused green fluorescent protein (GFP)
to either terminus of TonB (100); fusion proteins were expressed at chromosomal
levels in E. coli. One hybrid, possessing an N-terminal, cytoplasm-localized GFP
fusion, exhibited wild-type iron uptake kinetics and was the only hybrid to exhibit
GFP fluorescence. Since GFP fluorescence never decreased during transport, it
was concluded that TonB’s N-terminus never left the CM. This interpretation is
in accordance with reports claiming that periplasmic GFP does not fluoresce
(103). The same study also determined that TonB fractionates equally between
the OM and CM. Taken together, these results shed doubt on the shuttle model by
suggesting that TonB fractionation patterns are artefacts of sucrose-density
fractionation.
1.8.2 The rotation model
The rotation model combines distant sequence homology between ExbB
and MotA and between ExbD and MotB with features of dimeric TonB crystal
structures. Described previously, ExbB and ExbD share sequence homology and
topological features with MotA and MotB of the bacterial stator complex. Both
ExbB/D and MotA/B harness PMF and in the case of MotA/B, proton flow
through these proteins transduces into mechanical energy that rotates the bacterial
flagellum. The rotation model assumes that when bound to a TBDT, TonB causes
rigid body rotation of the transporter’s structural elements that open a pore for
siderophore translocation into the periplasm. Energy transduction would proceed
43
as TonB–ExbB–ExbD harnesses PMF. The model was originally supported by
observations of a cleft between the tightly intertwined, dimeric interface in the
first TonB crystal structures. The cleft was considered a structural element that
could interact with ligand-bound TBDTs. Presumably, rotation of dimeric TonB
(either from two neighbouring TonB–ExbB–ExbD complexes or from two TonB
monomers associated within a single TonB–ExbB–ExbD complex) would
propagate torsional force that structurally alters the receptor’s cork (Figure 1.10).
This perturbation would then facilitate transport of siderophore into the periplasm.
Figure 1.10. Rotation model of TonB-dependent energy transduction. A. Siderophore binding to
an OM-embedded TBDT signals occupancy by extending Ton box into the periplasm; B. TonB
engages ligand-bound receptor and together with ExbB/ExbD harnesses the PMF. This action
causes rotation of TonB and application of a torsional force that through an unknown mechanism
drives siderophore transport into the periplasm. Although one monomer of TonB is illustrated, the
model can accommodate two monomers of TonB, either within one TonB–ExbB–ExbD complex
or from two neighbouring complexes; C. TonB disengages receptor and returns to its ground-level
energy state.
44
The rotation model supports mechanisms of TonB-dependent transport
that are independent of TonB’s oligomeric state. Rotation of a TonB monomer
bound to a receptor as observed in the TonB–FhuA and TonB–BtuB crystal
structures could be coupled to the postulated 1:7:2 stoichiometry of the CM-
embedded TonB:ExbB:ExbD complex. Similarly, two neighbouring TonB–
ExbB–ExbD complexes could co-localize such that periplasmic regions of TonB
dimerize and bind to a siderophore-bound TBDT. Rotations of each complex
might still promote structural alterations of the TBDT cork such that siderophore
transport could occur.
1.8.3 The mechanical pulling model
Most recently, a particularly attractive model has emerged, suggesting that
TonB transduces energy to siderophore-bound TBDTs in the form of a
mechanical pulling force. Comparative structural analyses of FhuA, FecA, FepA
and BtuB revealed cork/barrel interface features similar to transient protein–
protein complexes (104). These features include interstitial waters that cushion
each receptor’s cork/barrel interface, which might act as a lubricant that facilitates
removal or unfolding of the cork domain, if sufficient energy were input.
Furthermore, arrangements of each receptor’s cork β-strands are oriented
such that only a modest amount of energy input might be required for unfolding.
These hypotheses derive from single molecule unfolding studies; modest forces
applied perpendicular to the plane of strands in a β-sheet are required to break
inter-strand hydrogen bonds and unfold the sheet (105). In contrast, large forces
45
applied parallel to the plane of a β-sheet are required to unfold the same β-sheet.
Strands comprising cork domain β-sheets are oriented perpendicular to the OM
plane. Therefore bound TonB could unfold this sheet through application of a
modest pulling force away from the membrane plane (Figure 1.11).
The mechanical pulling hypothesis was simulated in silico by means of
steered-molecular dynamics. TonB’s most N-terminal residue that was resolved
within the TonB–BtuB crystal structure was subjected to a pulling force
perpendicular to the OM plane (106). As the simulation proceeded, TonB
remained bound to BtuB’s Ton box, and BtuB’s cork began to unfold. After
TonB’s N-terminus had retracted nearly 200 Å away from its starting position,
Figure 1.11. Mechanical pulling model of TonB-dependent energy transduction. A. TonB
engages ligand-bound receptor as described for the rotation model; B. TonB exerts a mechanical
pulling force perpendicular to the membrane plane, causing the receptor cork to unfold;
siderophore translocates into the periplasm; C. TonB disengages receptor as in the rotation model
and returns to its ground-level energy state.
46
a pore large enough to accommodate cobalamin translocation opened. Forces
measured during simulation largely exceed the magnitudes of measured forces
described for other biological mechanisms. However, the simulations
demonstrated the mechanism’s plausibility.
1.8.4 Conformational rearrangement of TBDT cork domains
The mechanisms described above attempt to rationalize how TonB
transduces energy to ligand-bound transporters. The mechanical pulling
mechanism offers insight into how energy transduction might couple to unfolding
of a ligand-bound transporter’s cork domain. However, whether energy
transduction results in complete ejection of the cork from the receptor lumen or if
it causes local unfolding is a controversial topic. Evidence for both possibilities
exists. For example, TonB-dependent transport of the ferric-hydroxamate
siderophore, ferricrocin, was not attenuated when FhuA’s cork was tethered to its
barrel by disulfide bridges (107,108), indicating that FhuA’s cork remains within
the barrel during transport. A local conformational change must open a pore large
enough to allow passage of ferric-siderophore. In a similar study, FepA’s cork
domain was subjected to cysteine mutagenesis and expressed in E. coli (109).
Some, but not all of the cork cysteines were labelled during TonB-dependent
transport. Thus, FepA’s cork may partially unfold during transport. However,
another FepA labelling study indicated that extensive portions of FepA’s cork
become labelled during transport, suggesting that its cork completely ejected from
the barrel (110). These conflicting observations require further investigation.
47
1.9 Periplasmic siderophore transport
Having crossed the outer membrane, siderophores are delivered to CM-
embedded ATP permeases by periplasmic binding proteins (PBP). In E. coli, the
PBP, FhuD, binds and transports ferric-hydroxamate-type siderophores across the
periplasm (Figure 1.12). Other siderophore-binding proteins include ferric-
enterobactin-binding protein FepB, and ferric-citrate-binding protein FecB. In
addition, the PBP BtuF is required for TonB-dependent transport of cobalamin
across the periplasm (Figure 1.12).
FhuD is a 32 kDa bi-lobal protein comprised of 296 residues. Structures
of FhuD with various bound hydroxamate siderophores revealed similar modes of
binding (111,112); siderophores bind in a shallow solvent-exposed pocket situated
between FhuD’s N- and C-terminal lobes (Figure 1.12). Aromatic and conserved
Tyr and Arg residues line the binding site and form hydrogen bonds with bound
siderophores. Whereas all ferric-hydroxamate-like siderophores occupy the same
binding site, the determinants of binding, such as hydrogen bonding patterns and
hydrophobic interactions between FhuD and ligands slightly differ. These
properties bestow FhuD with a broad ligand-binding specificity and accordingly,
it binds siderophores with modest affinities, in the low micromolar range (113).
The structure of BtuF (Figure 1.12) revealed a similar, FhuD-like fold, yet
exhibited a binding site with contrasting features (114). Both FhuD and BtuF
possess N- and C-terminal lobes bridged by a rigid α-helix. Similarly, BtuF binds
cobalamin in a pocket between these lobes. However, BtuF’s binding site is
comparatively hydrophilic, reflecting the cobalamin-specific binding
48
determinants. This specificity bestows BtuF with high affinity (Kd ~15 nM) for
cobalamin (115).
1.9.1 Ligand-induced periplasmic binding protein conformational changes
Some periplasmic binding proteins, such as maltose-binding protein
(MBP), exhibit large conformational changes upon binding ligands (116). The
structure of MBP (Figure 1.12) is similar to siderophore binding proteins; it
contains N- and C-terminal lobes with a maltose-binding site in between (117).
Figure 1.12. Periplasmic binding proteins. Various periplasmic binding proteins are displayed as
ribbon representations. A. Ferric-hydroxamate binding protein, FhuD (PDB code 1EFD). Bound
ferrichrome is displayed as a red surface representation; B. cobalamin-binding protein, BtuF (PDB
code 1N4A). Bound cobalamin is displayed as a magenta surface representation; C. apo-maltose
binding protein (PDB code 1N3X); D. holo-maltose binding protein (PDB code 1ANF). Bound
maltose is displayed as a pink surface representation
49
Comparison of apo-MBP and maltose-bound MBP structures (Figure 1.12) reveal
that MBP’s N- and C-terminal lobes close over bound maltose, excluding it from
solvent (118).
Compared to the structure of ferrichrome-bound FhuD, an unpublished
apo-FhuD structure demonstrated virtually no conformational change (119).
Similarly, an apo-BtuF structure demonstrated only slight reduction of mobility
upon binding cobalamin (120). This conformational change is notably less
pronounced than those associated upon ligand binding of MBP-like PBPs.
Structural elements bridging the lobes of MBP-like PBPs facilitate the large
ligand-induced conformational changes. Three β-strands bridge the lobes of
MBP, whereas rigid α-helices bridge the lobes of siderophore binding proteins. In
MBP, maltose binding exploits the flexibility of this bridge to promote surface
area burial within the ligand binding site. In contrast, rigidity of the α-helical
bridge in siderophore-binding PBPs reduces the degree of surface area burial that
can be afforded when ligands bind.
Despite the conformational rigidity of FhuD and BtuF structures, there is
recent evidence that both proteins exhibit more substantial conformational
changes upon binding ligands. Molecular dynamics simulations of FhuD and
BtuF suggested that there is at least some degree of flexibility in these proteins,
not accounted for in the crystal structures (119,121,122). Furthermore, structures
from two crystal forms of the FhuD-like PBP, FitE, from a clinical E. coli isolate,
also demonstrated a degree of flexibility that was reduced upon ligand binding
A C
C
50
(123). Further investigation is required to determine if siderophore PBPs exhibit
similar ligand binding mechanisms as MBP-like PBPs.
1.10 Cytoplasmic membrane permeases
Siderophore-bound PBPs ultimately deliver their cargo to CM-embedded
permeases that transport siderophore into the cytoplasm. Permeases belong to the
broad and ubiquitous class of ABC transporters found in all kingdoms of life
(124). ABC transporters comprise two functionally distinct domains: a TM gated
channel and a cytoplasmic nucleotide-binding domain (NBD). Domain
stoichiometries vary; FhuB/C, the ferric-hydroxamate siderophore permease,
comprises a single polypeptide unit (FhuB) that forms the TM channel and two
individual cytoplasmic NBDs (FhuC). In contrast, BtuC/D, the cobalamin
transport permease, comprises a dimer of BtuC chains that form the TM channel
and two cytoplasmic NBDs (BtuD). The BtuC/D organization appears to be the
most common, since homologous ferric-citrate and ferric-enterobactin permeases
(FecC/D and FepC/D respectively) share the dimeric TM domain and NBD
organisation (125).
As prototypical metal-chelate permease, the BtuC/D crystal structure
revealed a representative TM domain organization (126) that is likely shared
amongst homologs. Each BtuC monomer forms a CM-embedded, ten TM helical
bundle. Between the BtuC dimer interface is a gated channel that forms the
cobalamain translocation route. Bound to the cytoplasmic surface of BtuC is a
BtuD NBD dimer. Overall, the BtuC/D structure resembles other ABC
51
transporter structures, including the putative metal-chelate-type transporter
HI1470/1 from H. influenzae (127), the molybdate transporter ModB/C from
Archaeoglobus fulgidus (128), and the maltose transporter MalFG/K2 from E. coli
(129).
1.10.1 Permease–periplasmic binding protein interactions
Siderophore transport across the CM initiates once a siderophore-bound
PBP docks to its cognate permease. A structure of BtuF bound to BtuC/D
revealed the determinants of binding between a TonB-dependent transport PBP
and its CM-permease (130). Conserved Arg residues on periplasm-exposed
surface loops of BtuC formed electrostatic interactions with conserved Glu
residues on each lobe of BtuF. Negatively charged residues on the lobes of FhuD
are also conserved and a similar binding mechanism may position FhuD on the
surface of FhuB/C. An investigation that mutated homologous residues of
FhuD2, a FhuD ortholog from S. aureus, provided evidence to support this
mechanism (131). Ferrichrome transport was impaired in bacteria harbouring
these mutants, indicating their importance in vivo. The interface between FhuD
and FhuB was further mapped using synthetic FhuB-derived peptides that bound
to FhuD (132); a FhuB peptide corresponding to predicted periplasmic loop 3
bound to FhuD and inhibited ferrichrome transport in vivo, indicating that the
FhuB region interacts with FhuD during transport.
52
1.10.2 Permeases: transport mechanism
Bacterial ferric-iron acquisition culminates with ferric-siderophore
passage through a permease and into the cytoplasm. Energy for transport across
the CM is provided by cytoplasmic, NBD-mediated ATP hydrolysis. Insight into
permease transport has been provided by structures of various ABC-transporters,
each considered to represent a conformational intermediate within a common
transport mechanism. These structures reveal differently gated transporter
conformations, described as either inward- or outward-facing. In its resting,
nucleotide-free state, the structure of BtuC/D displays an outward-facing
conformation, with its channel opened towards the periplasm (126). In contrast,
the structure of putative metal chelate transporter HI1470/1 revealed an inward-
facing conformation, with its channel open to the cytoplasm (127). The structure
of apo-BtuF bound to BtuC/D revealed an intermediate conformation (130).
Whether a transporter adopts an inward- or outward-facing conformation is
postulated to depend on occupancy of its NBDs (133). Comparisons of different
nucleotide-bound states of other ABC transporters revealed a coupling mechanism
that propagates to the transporters’ TM domains. Upon binding ATP, one NBD
conformationally engages its neighboring NBD. This spatial engagement
propagates through the transporter to yield an outward-facing conformation of its
TM helices. After ATP hydrolysis, ADP is bound and the NBDs disengage. This
conformational change then propagates through the transporter to yield an inward-
facing conformation, and a channel opening directed to the cytoplasm.
53
Concomitant with nucleotide exchange, ligands delivered by PBPs are considered
to switch their localization from the periplasm to the cytoplasm.
1.10.3 Intracellular fate of iron
Having acquired ferric-siderophores in the cytoplasm, iron is ultimately
removed from the siderophore backbone and processed. Given that iron binds to
siderophore backbones with extremely high affinity, its release requires catalytic
removal. Intracellular esterases, such as enterochelin esterase, hydrolyze
siderophore backbones, causing cytoplasmic release of iron (134). Other iron
release mechanisms involve enzymatic reduction of iron to it ferrous state, which
owing to its weaker affinity, is more easily removed from the siderophore
backbone. FhuF, an E. coli Fe-S cluster-containing enzyme was recently
demonstrated to reduce ferric-hydroxamate siderophores, which led to release of
iron into the aqueous phase (135).
Once released into the cytoplasm, iron is processed into a variety of forms.
One fate involves iron sequestration and storage. Proteins such as bacterioferritin
and mini-ferritin act as iron storage compartments. These proteins oligomerize
into large spherical compartments that bind up to 3000 atoms of iron for ferritin,
and up to 500 atoms for mini-ferritin (136). A second fate involves shuttling iron
into Fe-dependent proteins, such as metabolic enzymes. In addition, the
intracellular iron concentration affects iron-regulated gene transcription.
Transcription of iron-regulated operons, such as those encoding siderophore
transport systems, become de-repressed under iron-deplete conditions.
54
Conversely, when iron stores are sufficient, iron-regulated gene transcription is
repressed.
1.11 Introduction to techniques used in this thesis
Genetic and biophysical techniques are described in this thesis that
elucidated protein–protein interactions involved in TonB-dependent siderophore
uptake. A brief introduction to each technique is now discussed to facilitate a
greater understanding of the strategies employed to characterize these interactions.
1.11.1 Phage display
Phage display is a molecular genetic technique that predicts and localizes
binding determinants between two macromolecules. Reported applications
include localization of small molecule binding sites on drug targets (137,138),
identification of antibody epitopes (98), and identification/localization of protein–
protein interaction surfaces (139). The latter aspect of phage display plays a
central role in this thesis and is described in greater detail in chapters 2 and 3.
Phage display capitalizes on the vast knowledge of bacteriophage M13
molecular biology. Combinatorial M13 phage libraries are engineered to display
random peptide fusions at the N-termini of various coat proteins. The libraries
described chapters 2 and 3 display up to three copies each of random peptide
fusions at the N-terminus of the pIII minor coat protein. Combinatorial phage
libraries incorporate up to one billion random sequences, each displayed
separately on a single phage particle. Large numbers of sequence displays ensure
55
that peptide sequence space and peptide conformation space is efficiently sampled
(140).
The process of predicting and localizing protein–protein interactions by
phage display occurs through an iterative process known as phage panning, or
affinity selection (Figure 1.13). The goal of panning is to identify target proteins
with sequences similar to those displayed by phage that specifically interacted
with a selected bait protein. This is accomplished experimentally by incubating
phage libraries with purified bait protein in a well of a microtitre plate. A wash
procedure ensures removal of non-interacting or weakly-interacting phage, while
phage that display bait-interacting peptides are retained. Bait-interacting phage
are then eluted and amplified by infecting an E. coli host strain. Amplified phage
are harvested and represent an enriched population that were selected based upon
their affinity for the bait protein. Amplified phage are generally subjected to
additional rounds of panning to further enrich for a population that displays high
affinity bait-binding peptide sequences.
After subsequent rounds of phage panning, phage DNA is purified and
sequenced; peptide sequences that interacted with the bait protein are identified.
Phage-derived sequences are then compared to target protein sequences by pair-
wise alignments and scored with an alignment matrix. Sequences scoring above a
selected threshold are considered to align significantly with the target protein
sequence. Ideally, many phage-derived sequences align within a region of the
target protein sequence, so as to form a cluster. Clusters of aligned peptides
indicate likely regions of interaction between bait and target proteins.
56
Figure 1.13. Phage panning. The process of phage panning identifies complementary protein
interaction sequences. A. A combinatorial phage library is incubated with a bait protein that coats
a microtitre well; B. weakly- or non-interacting phage are washed away, leaving only bait-
interacting phage; C. bait-interacting phage are eluted; D. eluted phage are amplified by infection
of host bacteria; E. panning with the enriched phage population is repeated; F. phage DNA is
harvested and sequenced after a few rounds of panning; G. peptides derived from bait-interacting
phage are aligned against target protein(s). Bait-interacting peptides that align within a cluster on
the target sequence define a region of probable interaction between bait and target proteins.
Information from phage panning can refine regions of interaction between
proteins known to interact, or can identify novel protein–protein interactions.
Predictions of novel protein–protein interactions are then demonstrated.
Complementary protein–protein interaction surfaces can be predicted if a previous
target protein is purified and subsequently used as bait for a second phage panning
experiment. These activities enable predictions of interacting regions between
two proteins and guide further experimental strategies to refine complementary
interaction surfaces.
57
1.11.2 Surface plasmon resonance
The biophysical technique of SPR is a valuable and sensitive tool to
characterize protein–protein interactions in real-time and without need for
labelling. Information such as kinetic association and dissociation rates can be
measured and used to calculate protein–protein binding constants. Furthermore,
information about interaction stoichiometries can be estimated from fits of data to
appropriately selected models.
In SPR, changes in bulk refractive indices that occur when protein
analytes bind to immobilized protein partners are measured and quantified (141).
Immobilization proceeds by covalently tethering a protein (the ligand) to the
surface of a chemically derivatized SPR sensor chip, which is encased within a
microfluidic flow cell. Various chemistries are used to selectively couple and
orient a protein ligand to the sensor chip. These include amine-coupling
chemistries that randomly orient the ligand, or thiol-coupling chemistries that can
directionally orient the ligand. Once immobilized, a protein analyte is then
injected over the ligand surface and responses are recorded.
Interactions between immobilized protein ligand and analyte cause
accumulation of mass on the sensor chip surface. Mass accumulation is optically
detected as a change in the bulk refractive index. Real-time association between
ligand and analyte are recorded in values of resonance units. As mass
accumulates on the sensor chip surface, there is a concomitant increase in the
amount of resonance units recorded.
58
A second injection of SPR running buffer initiates dissociation of the
ligand–analyte complex. As analytes dissociate from the sensor chip surface, the
accumulated mass decreases and a concomitant decrease in the amount of
resonance units is recorded. Throughout the SPR experiment, changes in bulk
refractive indices that occur in real-time upon association and dissociation of
analyte are recorded and the resulting sensorgrams are fit to an appropriately
selected experimental model.
The choice of model used to fit SPR sensorgrams depends largely on
known information pertaining to the biological system. For example, a model that
describes simple 1:1 binding is selected if two proteins are known to interact in a
1:1 stoichiometry. If other stoichiometries are known, more appropriate models
are selected. When fitting the sensorgrams to a model, rates of association and
dissociation are estimated. These rates are used to calculate the binding constant
for the protein–protein interaction being considered.
1.11.3 Dynamic light scattering
Dynamic light scattering (DLS) is an optical technique characterized by
the unique ways that particles scatter light in solution. In DLS, purified protein is
placed into a sample cell and light is shone upon the sample. Samples might
contain an individual protein or a protein–protein complex. Incident light is
scattered by the protein particles and the scattering intensities are measured over
short time intervals ranging from 100 ns to 30 ms. The magnitude of scattered
light intensity fluctuates with time and arises due to Brownian motion of the
59
suspended protein particles. The rate that scattered light intensities fluctuate
arises from the translational diffusion coefficient of the protein particles, which is
a function of the protein’s hydrodynamic radius and molecular weight (142).
Since light scattering profiles are related to particle molecular weight, they enable
estimations of protein sizes in solution. This is accomplished by mathematical
deconvolution of scattering profiles into a histogram describing the sizes of
components that could so scatter the light. The ability to measure sizes of
proteins in solution makes DLS a useful tool for investigating protein–protein
complexes. If two proteins interact in solution, the molecular weight
corresponding to the complex should be measurable by DLS.
1.11.4 Analytical ultracentrifugation
Analytical ultracentrifugation (AUC) is a hydrodynamic technique that
complements information provided from DLS measurements. Like DLS, AUC
also enables characterizations of molecular weight and stoichiometry of
macromolecular complexes, and provides information on the shape of a given
protein in solution. This is obtained from measurements of protein sedimentation
profiles. In response to an applied centrifugal force, proteins sediment in ways
that depend on hydrodynamic shape and size (143). Experimentally, this is
followed by measuring absorbance of a protein sample as it sediments. The
protein sample is placed inside a sample cell that is housed inside a specialized
centrifuge rotor. Attached to the rotor is an optics system that monitors the
sample cell’s radial absorbance during the sedimentation experiment; upon
60
application of centrifugal force, the protein sediments toward the bottom of the
sample cell. Sedimentation is optically monitored by measuring sample
absorbance at 280 nm along the cell’s length, and proceeds until the protein has
sedimented to the bottom of the sample cell.
Resulting sedimentation profiles are then fit to the Lamm equation, which
describes the sedimentation behaviour of a particle in solution. Fits to the Lamm
equation yield information on the sedimentation coefficient of the protein, which
is directly proportional to its molecular weight (144). Additional parameters that
are fit include the protein’s frictional ratio, a measure of its hydrodynamic shape.
Frictional ratios measure deviation of protein hydrodynamic shape relative to that
of a perfect sphere; globular proteins possesses frictional ratio values of around
1.3, whereas elongated proteins possess values around 1.8 or greater.
1.12 Rationale and thesis objectives
TonB is a central element involved in siderophore transport. Upon
starting this project, a wealth of knowledge existed about how TonB interacts with
OM receptors. However, relatively less was known about other periplasmic TonB
interactions. The goals of this project were first, to predict which regions on the
periplasmic surface of FhuA would bind to TonB. This objective is discussed in
Chapter 2. Our use of phage display accurately predicted periplasm-exposed
regions on FhuA that were later confirmed with our X-ray crystal structure of the
TonB–FhuA complex. In addition to the novelty of that structure, it provided
proof of principle that our phage display strategies were revealing true protein–
61
protein interfaces. Discussed in chapter 3 is the second objective; to predict
additional periplasmic interactions that TonB is capable of making. To our
surprise, the periplasmic binding protein, FhuD, was predicted to bind TonB. We
then demonstrated that TonB and FhuD interacted in vitro. In chapter 4, a third
objective is discussed; to identify essential determinants of binding between TonB
and FhuD. Using biophysical and computational characterizations, we delineated
essential FhuD-binding regions in TonB and identified a plausible mode of
binding that orients FhuD’s siderophore binding site towards the OM, when
bound to TonB. In chapter 5 the final objective of this project is discussed; co-
crystallization of the TonB–FhuD complex. Here, protein preparations and
TonB–FhuD complex formation are discussed. Furthermore, properties of TonB
that are refractory to crystallization were identified. In chapter 6, our
experimental findings are summarized in the context of current models that
describe mechanisms of TonB-dependent transport.
62
Preface to chapter 2
Chapter 2 describes the use of phage display technology to predict TonB-
binding surfaces on the periplasm-exposed regions of TonB-dependent
transporters. Phage libraries were panned against TonB and sequences derived
from TonB-interacting phage were identified. These were compared to primary
sequences of TonB-dependent transporters and scored through pair-wise
alignments. This strategy predicted regions within the transporters that TonB
would bind to. These regions were subsequently visualized on crystal structures
of E. coli transporters FhuA, FepA, FecA and BtuB. Regions within FhuA that
were predicted to bind TonB were cloned within the N-terminus of maltose-
binding protein. Binding of these FhuA-derived peptides to TonB was confirmed
by ELISA.
63
Chapter 2
Phage display reveals multiple contact sites between FhuA,
an outer membrane receptor of Escherichia coli, and TonB
David M. Carter, Jean-Nicolas Gagnon, Moussab Damlaj, Suneeta Mandava,
Lee Makowski, Diane J. Rodi, Peter D. Pawelek and James W. Coulton
Journal of Molecular Biology (2006). 357: 236-251
Copyright © 2005, Elsevier Ltd All rights reserved
64
2.0 Summary
The ferric hydroxamate uptake receptor FhuA from Escherichia coli
transports siderophores across the outer membrane (OM). TonB–ExbB–ExbD
transduces energy from the cytoplasmic membrane to the OM by contacts
between TonB and OM receptors that contain the Ton box, a consensus sequence
near the N-terminus. Although the Ton box is a region of known contact between
OM receptors and TonB, our biophysical studies established that TonB binds to
FhuA through multiple regions of interaction. Panning of phage-displayed
random peptide libraries (Ph.D.-12, Ph.D.-C7C) against TonB identified peptide
sequences that specifically interact with TonB. Analyses of these sequences using
the REceptor LIgand Contacts (RELIC) suite of programs revealed clusters of
multiply-aligned peptides that mapped to FhuA. These clusters localized to a
continuous periplasm-accessible surface: Ton box/switch helix; cork
domain/strand; and periplasmic turn 8. Guided by such matches, synthetic
oligonucleotides corresponding to DNA sequences identical to fhuA were fused to
malE; peptides corresponding to the above regions were displayed at the N-
terminus of E. coli maltose-binding protein (MBP). Purified FhuA peptides fused
to MBP bound specifically to TonB by ELISA. Furthermore, they competed with
ligand-loaded FhuA for binding to TonB. RELIC also identified clusters of
multiply aligned peptides corresponding to the Ton box regions in BtuB, FepA,
and FecA; to periplasmic turn 8 in BtuB and FecA; and to periplasmic turns 1 and
2 in FepA. These experimental outcomes identify specific molecular contacts
65
made between TonB and OM receptors that extend beyond the well-characterized
Ton box.
2.1 Introduction
Iron is essential in bacteria for processes such as respiration, RNA
synthesis and inactivation of reactive oxygen species (145). Under physiological
conditions, free ferric iron, present at extremely low concentrations (10-18
M),
limits bacterial growth (6). To circumvent this problem, Gram-negative bacteria
have evolved specialized iron transport systems for uptake of siderophores, iron
chelators of less than 1000 Da. These transport systems are composed of a high-
affinity outer membrane (OM) receptor, a periplasmic binding protein and an
ATP-dependent cytoplasmic membrane (CM) transporter. The TonB–ExbB–
ExbD complex provides energy required by the OM receptors for unidirectional
transport of siderophores (146). Anchored in the CM, this complex harnesses
energy of the proton motive force. Given that TonB physically associates with
OM receptors in vivo, it is likely that these interactions result in energy
transduction necessary for ligand transport (89,90).
One paradigm of a TonB-dependent OM receptor is FhuA, receptor for the
siderophores ferrichrome and ferricrocin. In addition, FhuA serves as receptor for
phages T1, T5, 80 and UC-1, the siderophore-antibiotic conjugate albomycin,
rifamycin CGP 4832, the bacterial toxin colicin M, and the antimicrobial peptide
MccJ21 (147-149). With the exception of phage T5, uptake of these ligands
requires energy from the TonB–ExbB–ExbD complex for transport. X-ray
66
crystallographic studies of the TonB-dependent E. coli OM receptors FhuA
(42,150), FepA (44), FecA (45), BtuB (46), and the OM receptors FpvA (52) and
FptA (51) from Pseudomonas aeruginosa revealed common structural features.
All six receptors are composed of a C-terminal 22-stranded -barrel domain and
an N-terminal domain termed the cork that occupies the interior of the barrel. In
FhuA, the cork tightly interacts with the interior of the barrel via a network of
hydrogen bonds and salt bridges (42). Recent analyses (104) of available
structures of TonB-dependent OM receptors indicates that the cork-barrel
interface is highly hydrated. Presence of a large number of interfacial water
molecules may reduce the energy required to effect subtle conformational changes
or even significant movements of the cork domain, either of which could facilitate
siderophore transport (151).
The first crystal structure of the C-terminal domain of TonB (residues
155–239) (72) provided unexpected evidence that TonB forms a dimer. The
authors speculated that a region of the dimer interface proximal to residue Asn-
200 forms a binding cleft for OM receptors. Based on their structural data, they
proposed a mechanism whereby the proton motive force provides propeller-like
torsional motion of dimeric TonB, thereby inducing conformational change in
OM receptors. An alternate model postulated that TonB shuttles between the CM
and the OM (102,152). In the shuttle model, TonB may dissociate from the CM
after having been energized by ExbB–ExbD, delivering stored potential energy to
OM receptors and allowing transport of ligands into the periplasm. Two
additional structures from progressively longer C-terminal TonB constructs have
67
since been reported. An X-ray crystallographic structure of TonB 148–239
indicated that this protein was dimeric, although much more loosely packed than
the dimer formed by TonB 155–239 (74). Most recently, an NMR structure of
TonB 103–239 demonstrated a monomer in solution (75).
The oligomeric state of TonB when in contact with OM receptors has
recently been investigated by numerous methods. Our analytical
ultracentrifugation studies (77) of TonB-FhuA interactions identified a 2:1
complex. These findings were corroborated by in vivo studies using a bacterial
two-hybrid system (78). However, Koedding et al. (73) proposed that TonB
dimerization is not necessary since in vivo, TonB constructs that could block
TonB-dependent processes were shown by analytical ultracentrifugation to be
monomeric in solution. Even though Khursigara et al. and Koedding et al. draw
different conclusions from their observations, the two studies may be reconciled
since both groups determined that the C-terminal portion of TonB forms a dimer
in solution; longer TonB constructs (composed of at least residues 145–239) are
monomeric in solution, consistent with the reported NMR structure. Most
recently, Ghosh and Postle (79) demonstrated TonB dimerization at the CM in a
manner requiring both the TonB signal anchor sequence and energy input from
the ExbB–ExbD complex. They found monomeric TonB at the OM, indicating
that dynamics of TonB oligimerization during ligand transport remain only
partially understood.
All TonB-dependent OM receptors share a conserved five to six amino
acid sequence near their amino termini, a region termed the Ton box that is
68
proposed to interact with TonB during energy transduction. Mutations in this
region abolished TonB-dependent activities of these receptors; TonB-independent
receptor-mediated activities remained unaffected (153,154). Gudmundsdottir et
al. (55) demonstrated that rather than strict amino acid sequence, conformation of
the Ton box is important for its specific interaction with TonB; mutations that
were most deleterious to normal functions of BtuB were those that introduced
glycine or proline into the Ton box. Other amino acid substitutions in the same
region had little effect on TonB-dependent activities. Another study showed that
a mutation (Ile14Pro) in the Ton box of FepA resulted in total loss of TonB-
dependent activities (91). In addition, insertion of cysteine residues in the Ton
box of BtuB (92,93) or FecA (155) and insertion of cysteine residues around
amino acid 160 of TonB resulted in disulfide bridge formation between these two
protein partners. Finally, binding of vitamin B12 to BtuB triggers a
conformational change in the Ton box, a conclusion supported by site-directed
spin-labelling (156), biotin maleimide labelling (157) and X-ray crystallography
(46).
Some reports originally proposed that regions other than the Ton box
participate in interactions of OM receptors with TonB because deletions of the
cork domain including the Ton box in FhuA (158) and FepA (159) did not
completely abolish TonB-dependent transport processes. By synthesizing
periplasm-directed TonB fragments, Howard et al. (160) demonstrated reduced
TonB activity in FhuA and corkless FhuA variants alike. Killmann et al. (161)
also observed reduced TonB activity in FhuA and corkless FhuA via interference
69
from synthetic nonapeptides corresponding to regions near glutamine160 of
TonB. To resolve contradictions, Vakharia and Postle (162) as well as Braun et
al. (163) clearly demonstrated that the phenotype observed for the FepA and
FhuA cork deletions was due to interprotein complementation. However, these
results do not exclude the possibility that TonB interacts with regions other than
the cork of OM receptors since FepA mutants that lack a cork domain co-
precipitated with TonB (162) suggesting weak interactions in addition to the Ton
box residues.
TonB has been shown to interact differentially with OM receptors
(78,164). Mutation in the C-terminal portion of TonB resulted in diverse
phenotypes with respect to TonB-specific processes. In both studies, results were
interpreted by proposing distinct interactions of TonB with each OM receptor.
Our recent studies (77,94,95) investigating TonB-FhuA interactions using
surface plasmon resonance (SPR) showed that TonB possesses two affinities for
FhuA. A low-affinity binding site is located in the C-terminal segment of TonB
between residues 155–239; a higher affinity binding site becomes accessible when
residues 33–154 are present. Consistent with the hypothesis of multiple physical
interactions for TonB with OM receptors, our schematic model for mechanism
incorporated two binding sites for TonB.
Using phage display of peptides as an experimental strategy, we now
report identification of novel TonB-binding motifs in addition to the well
characterized Ton box. These motifs map to a continuous surface along the
periplasmic face of FhuA. FhuA sequences corresponding to potential TonB-
70
interacting regions identified by phage display were shown to bind specifically to
TonB in vitro. To our knowledge, this is the first application of phage display in
identifying protein–protein interactions within the bacterial cell envelope. These
findings provide substantive evidence for interactions between TonB and OM
receptors that extend beyond the Ton box.
2.2 Materials and methods
2.2.1 Bacterial strains and media
The phage libraries Ph.D.-12 and Ph.D.-C7C were purchased from New
England Biolabs (NEB). E. coli K-12 ER2738 (NEB) was used for amplification
and titration of phage M13 pools; E. coli DH5α was used for cloning of peptide-
coding DNA sequences into pMal-pIII, a vector (165) obtained from C.J. Noren.
[Peptide-MBP] fusions were expressed in E. coli NM522 (NEB). Cultures were
grown in Luria Bertani (LB) broth supplemented with 0.2% [w/v] glucose for
expression of [peptide-MBP] fusions.
2.2.2 Chemicals and reagents
5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside (X-Gal) and
isopropyl-ß-D-thiogalactopyranoside (IPTG) were purchased from Biovectra
(Charlottetown, PE, Canada). Protein-grade Tween-20 was from Calbiochem.
All antibiotics used in this study were obtained from Sigma-Aldrich.
Oligonucleotides were synthesized at the Sheldon Biotechnology Centre, McGill
71
University. Ni-NTA resin used for purifications of H6.'TonB and FhuA was from
Qiagen and amylose resin for purification of [peptide-MBP] fusions was from
NEB.
2.2.3 Protein purification
The TonB construct referred to as H6.'TonB was described by Moeck and
Letellier (166); the first 32 amino acids corresponding to the transmembrane
anchor were removed and replaced by a 20 residue linker including a
hexahistidine tag. The genetically engineered FhuA protein (167) contains a
hexahistidine tag at position 405 within a surface-exposed loop. Both proteins
were purified as previously reported (77).
2.2.4 Phage M13 titre
Plaque assays were performed to determine phage titres. LB plus
tetracycline (20 µg ml-1
) was inoculated with E. coli ER2738 and grown at 37°C
until mid-logarithmic phase. Agarose top (0.7% agarose [w/v] in LB) was melted
and dispensed in 3 ml aliquots and kept at 45°C. Ten-fold serial dilutions of
phage were prepared. E. coli ER2738 (100 µl) was mixed with 100 µl of phage
dilutions, followed by 10 min incubation. Infected cells were transferred to
agarose top and poured onto pre-warmed LB/X-Gal/IPTG plates. After overnight
incubation at 37°C, blue plaques were counted to determine phage titres.
72
2.2.5 Panning procedures
H6.'TonB was diluted to 100 µg ml-1
in Tris buffered saline (TBS; 50 mM
Tris-HCl, 150 mM NaCl, pH 7.5) and 150 µl was adsorbed to six wells of a 96-
well polystyrene microtitre plate (Nunc Maxisorp). The plate was incubated at
4°C overnight. Wells were blocked by adding 400 µl of blocking buffer (TBS + 5
mg ml-1
BSA + 0.02% [w/v] NaN3) and incubating the plate 2 h at 37°C. They
were washed six times with TBST (TBS + 0.1% Tween-20 [v/v]). Ten µl of
phage library Ph.D.-12 (1.5 × 1011
pfu) or Ph.D.-C7C (1.5 × 1011
pfu) was diluted
to 1 ml in TBST and 120 µl was dispensed in each well. The plate was incubated
for 1 h at room temperature with gentle agitation. Unbound phages were removed
from the wells by washing ten times with TBST. Bound phages were eluted by
100 µl of elution buffer (0.2 M glycine-HCl, pH 2.2, 1 mg ml-1
BSA). After 10
min incubation at room temperature, the eluates were pooled to one tube
containing 90 µl of 1 M Tris-HCl (pH 9.1) for neutralization. Titres of eluted
phages were determined as described above. The phage pool was amplified by
following the manufacturer’s instructions provided with the Ph.D. libraries
(NEB). This panning procedure was repeated for a total of three rounds for the
Ph.D.-12 library and four rounds for the Ph.D.-C7C library. For each subsequent
round the input number of phages was 1.5 × 1011
pfu for the Ph.D.-12 library and
2 × 1011
pfu for the Ph.D.-C7C library. Stringency of selection was increased by
using 0.3% [v/v] Tween-20 in TBS for the second round and 0.5% [v/v] Tween-
20 in TBS for subsequent rounds.
73
2.2.6 Isolation of phage M13 clones, DNA isolation and sequencing
An overnight culture of E. coli ER2738, diluted 1/100 in LB plus
tetracycline, was dispensed in 1 ml aliquots. Phage plaques from titration of the
final round of selection were stabbed using a pipette tip and inoculated into 1 ml
E. coli ER2738. The cultures were incubated 4.5 h at 37°C, then transferred to
microcentrifuge tubes and centrifuged at 10,000 × g for 2 min to pellet bacterial
cells. Supernatants were decanted. NaN3 (0.02%) was added to each stock and
titres were determined by plaque assays. Single-stranded DNA from each clone
was isolated using Spin M13 Kit (Qiagen). Quality of isolated DNA was assessed
by agarose gel electrophoresis. DNA samples were sequenced at the Sheldon
Biotechnology Centre, McGill University and at the Genome Québec Innovation
Centre.
2.2.7 Global analysis of affinity-selected peptides
After isolating and sequencing affinity-selected phage from both the
Ph.D.-12 and Ph.D.-C7C libraries, the deduced peptides were statistically
analyzed using the REceptor LIgand Contacts (RELIC) bioinformatics server
(http://relic.bio.anl.gov) (168). The program POPDIV was used to calculate the
diversity of affinity-selected peptides from both libraries. The program INFO was
used to calculate the information contents associated with affinity-selected
peptides. For all analyses, randomly selected peptides were provided by RELIC.
74
2.2.8 Cloning of peptide-coding DNA sequences into pMal-pIII vector
Sequences identical to fhuA were cloned to create N-terminal fusions to
MBP. Oligonucleotides corresponding to the coding and non-coding strands of
the Ton box (6EDTITVTAA
14), Fhu switch (
21AWGPAAT
27), Fhu cork1
(143
SSPGGLL149
), Fhu cork2 (146
GGLLNMVSK154
), Fhu cork3
(153
SKRPTTEPL161
) and Fhu turn (583
AKAALSA589
) were synthesized at Sheldon
Biotechnology Centre. Design of the oligonucleotides included ends compatible
for ligation into EagI/Acc65I-digested pMal-pIII vector plus a linker region
encoding the peptide GGGS that separates the FhuA sequence from the N-
terminus of MBP. Oligonucleotides were denatured at 95oC for 10 min and
annealed by cooling overnight to room temperature. Ligation required incubation
of annealed products with the expression vector pMal-pIII (Zwick et al. (165))
previously digested with EagI and Acc65I. Ligated products were transformed
into E. coli DH5α and the resulting plasmids sequenced (Sheldon Biotechnology
Centre) to verify faithful incorporation of fhuA inserts.
2.2.9 Peptide-MBP expression
E. coli NM522 harbouring the pMal-pIII derivatives for expression of the
peptide sequences were grown in LB broth supplemented with 0.2% glucose to
mid-logarithmic phase. Expression of [peptide-MBP] fusions was induced by
adding IPTG to a final concentration of 0.2 mM. The culture was grown for an
additional 3 h, after which cells were harvested by centrifugation. Periplasmic
75
extract was obtained by cold osmotic shock. Cells were suspended in 500 ml of
30 mM Tris-HCl, 1 mM EDTA, 20% sucrose, pH 8.0 and stirred for 20 min, room
temperature. Cells were pelleted by centrifugation, resuspended in 500 ml of ice-
cold 5 mM MgSO4, and stirred for 30 min at 4°C. Cells were pelleted by
centrifugation. The supernatant (cold osmotic shock fluid) was decanted and
filtered through a 0.45 μm filter to remove residual contaminant cells; 10 ml of
1 M Tris-HCl, pH 7.4 were added to the solution. The cold osmotic shock fluid
was concentrated to 40 ml using a YM10 ultrafiltration membrane (Millipore)
before loading onto an amylose column. For elution from the amylose column,
buffers were supplemented with 10 mM maltose. After [peptide-MBP] fusions
were purified and confirmed by SDS-PAGE, they were subjected to N-terminal
protein sequencing (Sheldon Biotechnology Centre); there were no ambiguities,
mismatches, or cleavages of peptide sequences.
2.2.10 Enzyme linked immunosorbent assay (ELISA)
H6.'TonB (20 pmol) was coated into each well of a Qiagen Ni-NTA
HisSorb 96-well microtitre plate and incubated overnight at 4oC. Use of these
plates increased sensitivity of the assay compared to polystyrene (Nunc Maxisorp)
plates. Binding of TonB was accomplished by attachment of its N-terminal
hexahistidine tag to the No-NTA-coated plate. Plates were blocked by addition of
2.5% casein [w/v] in Tris-buffered saline (TBS) and incubated for 3 h at room
temperature, followed by three washes with TBS. MBP fusion proteins (200
pmol) were added to each well and incubated at room temperature for 1 h with
76
shaking. MBP was probed by addition of anti-MBP antibodies (NEB) followed
by 1 h incubation. Anti-MBP was detected by addition of a secondary goat anti-
mouse IgG coupled to alkaline phosphatase (Cedarlane). Non-specifically bound
proteins were removed after each step by three washes with TBS containing
0.05% Tween-20 (Calbiochem). Plates were developed (37oC) by addition of p-
nitrophenyl phosphate (PNPP, MP Biomedicals). Endpoint absorbance
measurements (405 nm) were taken at 3 h. Raw absorbance data were corrected
by subtraction of a PNPP-containing blank. Competition between ferricrocin-
bound FhuA and MBP constructs for TonB was used as a measure of binding
specificity. For these measurements, equimolar amounts of FhuA and MBP
fusion proteins (200 pmol each) were mixed and added to H6.'TonB-coated wells.
Development and detection of ELISA signals were performed as above.
2.3 Results
2.3.1 Isolation of affinity-selected peptides by phage panning
A TonB construct containing residues 33–239 of E. coli TonB plus an N-
terminal hexahistidine tag (hereafter referred to as H6.'TonB) was assessed by
phage panning for interactions with randomly displayed peptides. Two different
peptide libraries display fusions at the N-terminus of the M13 bacteriophage
minor coat protein, pIII. The Ph.D.-12 library contains a linear dodecamer
separated by a linker; the Ph.D.-C7C library contains a heptapeptide sequence
constrained by a disulfide to form a loop. To enrich for affinity-selected peptides
77
against TonB, biopanning experiments were conducted. Each library was
incubated with H6.'TonB immobilized in wells of a microtitre plate to allow
retention of TonB-binding phages during washes. Affinity-selected phages were
then eluted. To select for unconstrained peptides that could bind to TonB, we
used the Ph.D.-12 library. The selection procedure was repeated for three rounds;
phage titres were determined after each round of selection. Titres of Ph.D.-12-
eluted phage increased approximately 70-fold over three successive rounds of
panning: 2.0 × 105 pfu ml
-1 (round 1) to 1.37 × 10
7 pfu ml
-1 (round 3). Panning
was not extended after the third round due to accumulation of redundant peptide
sequences. To select for constrained peptides that could bind to TonB, we
repeated our panning procedure with the Ph.D.-C7C library. The selection
procedure was repeated for four rounds. Following each round, the phage titre
ranged from 1.3 to 1.7 × 107 pfu ml
-1 and increased to 1.5 × 10
8 pfu ml
-1 from the
fourth round.
After panning the two libraries, 227 Ph.D.-12 phage clones and 271 Ph.D.-
C7C phage clones were isolated; their single-stranded DNA was purified and
sequenced. The peptide displayed by each phage was determined. From the 498
phage clones, we obtained 105 unique Ph.D.-12-derived sequences and 135
unique Ph.D.-C7C-derived sequences.
2.3.2 Global analysis of affinity-selected peptides
The RELIC suite of programs evaluates statistical properties of affinity-
selected peptides (168). RELIC confirmed that our ensembles of phage-derived
78
peptide sequences resulted from affinity to immobilized H6.'TonB. One indicator
of affinity selection is reduction in the diversity of sequences obtained from
iterative rounds of affinity selection (140,168,169). In this context, diversity is a
unit-less measure of the relative proportions of the 20 possible amino acids
observed within peptide sequences of a given population. Numerical methods
based on limited sequence data were developed to assess the diversity of a phage
population. Selection of peptides due to affinity for the immobilized target should
yield a population of peptides whose diversity is less than an equivalent number
of randomly-selected peptides from the parent library. This was evident for the
affinity-selected peptides obtained from biopanning against immobilized
H6.'TonB. RELIC/POPDIV, a program from the RELIC suite, determined that
the 105 unique affinity-selected peptides from the Ph.D.-12 parent library yielded
a diversity value of 0.004 ± 0.002; the 135 unique peptides from the Ph.D.-C7C
parent library yielded a diversity value of 0.066 ± 0.022. In contrast, ensembles
of 100 randomly selected peptides from each of the Ph.D.-12 and Ph.D.-C7C
libraries were shown to have diversities of 0.040 ± 0.016 and 0.079 ± 0.026,
respectively (140). The Ph.D.-12 affinity-selected peptides clearly contain a less
diverse population compared to randomly selected peptides from the same library,
while diversity of the Ph.D.-C7C affinity-selected peptides is not significantly
different from randomly selected peptides of the same library. The values
obtained for the H6.'TonB-affinity-selected peptides are comparable to those
obtained for Taxol-affinity-selected peptides (Ph.D.-12: 0.011 ± 0.007; Ph.D.-
79
C7C: 0.053 ± 0.021), and gamma-ATP-affinity-selected peptides (Ph.D.-12: 0.011
± 0.003; Ph.D.-C7C: 0.065 ± 0.011) (137,168).
Another indicator of affinity selection is a positive shift in the information
content of an ensemble of affinity-selected peptides compared to randomly
selected peptides. Information is defined as –ln(PN) where PN refers to the
probability of observing any given peptide within a parent library by chance alone
(140). Commonly occurring peptides within the parent library have a high
probability of being observed and consequently possess relatively low information
content. Conversely, rare peptides are less likely observed and are associated
with relatively higher information content. Affinity selection yields an ensemble
of peptides in which the distribution of information content is positively shifted
compared to randomly selected peptides. We observed such shifts for our
ensembles of affinity-selected peptides retrieved from biopanning against
immobilized H6.'TonB. Distributions of information content were calculated by
RELIC/INFO for affinity-selected peptides. For both libraries, significant
positive shifts in information content were evident compared to randomly selected
peptides (data not shown). Our affinity-selected peptide sequences therefore
reflect an enrichment of phage-borne peptides in which selection was primarily
due to affinity to immobilized H6.'TonB. Based upon these analyses, it is
unlikely that biases due to non-specific interaction between the phage and
H6.'TonB or factors based upon advantageous phage growth contributed
significantly to the selected peptides.
80
2.3.3 Identification of TonB-binding sites on the periplasmic surface of FhuA
To identify potential TonB-binding sites on the periplasm-exposed surface
of FhuA, we analyzed the ensembles of TonB-binding peptides isolated from the
Ph.D.-12 and Ph.D.-C7C libraries for similarities to the FhuA sequence. Given a
set of affinity-selected peptides obtained by phage display, the RELIC/MATCH
program identifies potential binding regions within the primary sequence of a
target protein that are most similar to affinity-selected peptides (168). Using a
variant of the BLOSUM62 substitution matrix, affinity-selected peptide sequences
are aligned to each residue within the target protein sequence and evaluated for
pairwise similarity within a scoring window. A peptide in a given position that
scores above a cutoff value (13 in the case of a 5-residue scoring window) is
considered to align to the protein at that position (168) and clusters of aligned
peptides may correlate with sequences involved in binding to the molecular target
(137).
The ensembles of 105 Ph.D.-12 peptides and 135 Ph.D.-C7C peptides
isolated from TonB-binding phages were evaluated by RELIC/MATCH for their
ability to align to the primary sequence of FhuA. Analyses focused on surfaces of
FhuA potentially exposed to the periplasmic space: strand regions of the β-barrel
that are located below the aromatic amino acid girdle, periplasm-exposed β-barrel
turns, and surfaces of the cork domain and barrel lumen accessible to the
periplasm. Given these constraints, we observed three regions in which
RELIC/MATCH identified multiple peptides that align to periplasm-exposed
surfaces of FhuA: the Ton box/switch helix (region I); the cork/1 strand
81
(region II); and periplasmic turn 8 (region III) (Figure 2.1). The Ph.D.-12 and
Ph.D.-C7C peptides that aligned to the FhuA Ton box overlapped residues from
the Ton box itself, and overlapped residues immediately C-terminal to the Ton
box that coincide with entry into the FhuA switch helix. Peptides aligned to the
FhuA cork domain could be similarly partitioned into two sub sites: the C-
terminal half of cork domain β strand D, and a contiguous region that comprises a
shallow loop entering into β1 of the barrel domain. In contrast, peptides aligned
to turn 8 localized to a single sub site comprised of residues within the
periplasmic turn (Figure 2.1).
Nine peptides from the Ph.D.-12 library were found to align to FhuA with
RELIC/MATCH scores of 13 or greater within a 5-residue scoring window (Table
2.1). Three peptides aligned with scores greater than or equal to 14 within a 7-
residue scoring window. The three highest-scoring Ph.D.-12 peptides aligned to
the FhuA cork domain. The lowest-scoring Ph.D.-12 peptides aligned to turn 8,
indicating that unconstrained peptides from the Ph.D.-12 library more effectively
mimic extended regions in the FhuA cork and Ton box than a periplasmic turn of
the barrel domain. Of the six Ph.D.-C7C peptides identified (Table 2.2) to align
within these same regions, three aligned with a RELIC/MATCH score of 13 or
greater within a 5-residue scoring window. Three additional peptides aligned
with a score of 13 within a 6-residue scoring window. The highest-scoring Ph.D.-
C7C peptide, ILAALSA, matched turn 8 (583
AKAALSA589
) with a score of 20,
consistent with the constrained nature of the Ph.D.-C7C peptides.
82
Similarity scores for each residue within FhuA scanned against the
ensemble of Ph.D.-12 affinity-selected peptides were determined using the
program RELIC/HETEROalign and displayed along the surface of FhuA (Figure
2.2A). Two regions potentially contributing to TonB-binding sites were
identified. A region shaded in blue-green corresponds to the Ton box/switch
region of FhuA (Figure 2.2D, region I). The region of highest similarity, shaded
in green, is similarly exposed to the periplasm and corresponds to the broad loop
(residues 153–160; Figure 2.2D, region II) connecting the C-terminus of the cork
domain to the β1 strand of the barrel domain. RELIC/HETEROalign similarity
scores corresponding to the Ph.D.-C7C ensemble of peptides were also
superimposed on the periplasmic surface of FhuA (Figure 2.2B). In this case, the
highest-scoring region shaded in red corresponds to periplasmic turn 8 (Figure
2.2D, region III). An additional region of high similarity (green) connected to
region II and consisting of residues from FhuA periplasmic turn 1 was also
observed. When the RELIC/HETEROalign scores from both ensembles were
averaged and displayed on the periplasmic surface of FhuA (Figure 2.2C), we
observed an almost continuous TonB-binding region: originating at turn 8 (Figure
2.2D, region III), spanning the Ton box/switch region (Figure 2.2D, region I), and
extending to the junction of the cork and barrel domains (Figure 2.2D, region II).
83
Figure 2.1. Alignments of affinity-selected peptides to FhuA as identified by RELIC/MATCH.
A. Region I: Ton box/switch helix; B. Region II: cork/1 strand; C. Region III: periplasmic turn 8.
Peptides from the Ph.D.-12 library are aligned below the FhuA sequence; peptides from the Ph.D.-
C7C library are aligned above the FhuA sequence. See Tables 2.1 and 2.2 for peptide match scores
and window sizes. Alignment positions are highlighted according to their pairwise alignment
score: +4, black background and white character; +1, grey background and dark character.
84
Table 2.1. RELIC/MATCH identification of TonB-affinity-selected Ph.D.-12 peptides
corresponding to FhuA sequences.
Peptidea
Peptide match
scoreb
Scoring window
(residues) Region
Alignment
positionc
NLLPTYTPLKLM 18 7 cork 156 HSNLPTKRPTSL 17 5 cork 153 CPNGLLGPCPSL 16 5 cork 145 KVWSLEPPGPAA 15 5 Ton box 22 TQPLGLLPSRHL 15 5 cork 145 IPHHHTLNMESH 15 5 cork 149 KPPSTWPQNALH 15 7 cork 154 LETRTTPPAKSQ 14 7 Ton box 10 KLVDESSTSPLS 13 5 cork 157 AGMATSRSTSPL 13 5 cork 157 KHVPVHSALSVN 13 5 turn 8 585 LEPQINSVGMVR 13 5 turn 8 590
aResidues contained in Scoring Window shown in bold.
bSum of pairwise comparisons of aligned residues within the Scoring Window.
cPosition of first residue of the Scoring Window numbering from the N-terminus of the
mature protein.
Table 2.2. RELIC/MATCH Identification of TonB-affinity-selected Ph.D.-C7C peptides
corresponding to FhuA sequences.
Peptidea
Peptide match
scoreb
Scoring window
(residues) Region
Alignment
positionc
ILAALSA 20 5 turn 8 585 LPTGGLL 15 5 cork 145 LTRSPAA 13 6 Ton box/switch helix 11 KSAFLPW 13 6 Ton box/switch helix 19 TKPLTQQ 13 6 cork 158 LAALKST 13 5 turn 8 585
a Residues contained in Scoring Window shown in bold.
b Sum of pairwise comparisons of aligned residues within the Scoring Window.
c Position of first residue of the Scoring Window numbering from the N-terminus of the mature
protein.
85
2.3.4 Identification of potential TonB-binding sites in structurally conserved OM
receptors
In addition to identifying three regions along the periplasm-exposed
surfaces of the FhuA cork and barrel domains, RELIC/MATCH analyses
identified TonB affinity-selected peptides that multiply align to corresponding
regions in BtuB, FecA, and FepA. From the Ph.D.-12 library, 16 peptides were
identified by RELIC/MATCH to align. Of the seven highest-scoring Ph.D.-12
peptides (score ≥ 15), five aligned to the Ton box regions in BtuB, FecA, and
FepA; two peptides aligned to periplasmic turns (Table 2.3). From the Ph.D.-C7C
library, 17 peptides were identified and aligned to BtuB, FecA, and FepA (Table
2.4). The seven highest-scoring Ph.D.-C7C peptides (score ≥ 15) aligned to
periplasmic turns in the barrel domains of all three proteins, consistent with the
constrained nature of the peptides.
RELIC/MATCH aligned affinity-selected peptides to periplasm-exposed
regions of BtuB that clustered in a region immediately C-terminal to the BtuB
Ton box, as well as at periplasmic turn 8 (Figure 2.3A). The BtuB Ton box
cluster is comprised predominantly of peptides from the Ph.D.-C7C library, as is
the BtuB turn 8 cluster. RELIC/HETEROalign similarity scores mapped to the
periplasm-exposed surface of BtuB clearly identify the BtuB Ton box (Figure
2.4A, region I) and turn 8 (Figure 2.4A, region II) as the highest-scoring regions.
86
Figure 2.2. RELIC/HETEROalign similarity scores mapped to the periplasm-exposed surface of
FhuA (PDB code 2FCP). Molecular surfaces rendered by PyMOL (170). Normalized
RELIC/HETEROalign similarity scores are represented on surfaces as colour ramps (blue=30 <
green =75 < red=120). A. Ph.D.-12 similarity scores; B. Ph.D.-C7C similarity scores; C. average
of Ph.D.-12 and Ph.D.-C7C similarity scores; D. superposition of the FhuA ribbon representation
and the molecular surface mapped in panels A to C. Regions corresponding to RELIC/MATCH
alignment clusters shown in Figure 2.1 are indicated by Roman numerals: (I) Ton box/switch
helix; (II) cork/1 strand; and (III) periplasmic turn 8.
Two clusters of affinity-selected peptides map to similar regions on the
OM receptor FecA. The FecA Ton box cluster of four Ph.D.-12 peptides and two
Ph.D.-C7C peptides occurs in a region immediately N-terminal to the FecA Ton
box (Figure 2.3B). The highest-scoring Ph.D.-12 peptide, containing the
sequence WSLEP within a 5-residue scoring window, was found to align to
87
Table 2.3. RELIC/MATCH Identification of TonB-affinity-selected Ph.D.-12 peptides
corresponding to BtuB, FecA, and FepA sequences.
Peptidea
Peptide match
scoreb
Scoring window
(residues) Protein Region
Alignment
positionc
TMGFTAPRFPHY 15 5 BtuB Ton box 11 KLVDESSTSPLS 13 5 BtuB turn 8 472 KVWSLEPPGPAA 17 5 FecA Ton box 69 AALGTYSTHTPT 15 5 FecA turn 8 596 SSMKVWSLPPAP 15 5 FecA Ton box 71 NLLPTYTPLKLM 14 6 FecA turn 8 594 NNSQKPAPVSPF 13 5 FecA Ton box 72 GNSVNKTWTHDY 13 5 FecA Ton box 66 KHVPVHSALSVN 13 5 FecA Ton box 80 SLKNYPVSWKNT 15 7 FepA Ton box 7 NLLPTYTPLKLM 15 5 FepA Ton box 3 TQPLGLLPSRHL 15 5 FepA turn 1 180 HSNLPTKRPTSL 14 7 FepA Ton box 3 QSPVNHHYHYHI 14 5 FepA Ton box 6 KLVDESSTSPLS 14 6 FepA turn 1 183 SHSNTTQTRPSD 13 5 FepA Ton box 9 a Residues contained in Scoring Window shown in bold.
b Sum of pairwise comparisons of aligned residues within Scoring Window.
c Position of first residue of Scoring Window relative to the N-terminus of the mature protein.
Table 2.4. RELIC/MATCH Identification of TonB-affinity-selected Ph.D.-C7C peptides
corresponding to BtuB, FecA, and FepA sequences.
Peptidea
Peptide match
scoreb
Scoring window
(residues) Protein Region
Alignment
positionc
TKPLTQQ 15 5 BtuB turn 8 472 TGPLPNR 15 5 BtuB turn 8 472 SPRTTPF 13 5 BtuB Ton box 17 MLEKPRL 13 5 BtuB Ton box 15 NQPRGPQ 12 4 BtuB Ton box 16 LTQTPTR 12 4 BtuB turn 8 475 LTQTPTR 15 5 FecA turn 8 596 HATLPPT 15 5 FecA turn 8 596 SWDPAPL 13 4 FecA Ton box 72 TLSPKLH 12 4 FecA Ton box 70 TLSPKLH 12 4 FecA turn 8 596 TGPLPNR 15 5 FepA turn 1 180 SWDPAPL 15 5 FepA turn 2 240 SHFAPHQ 15 5 FepA turn 2 241 SQVPLKS 13 5 FepA turn 2 242 HMSPLGA 12 4 FepA turn 1 181 MHMAPLS 12 4 FepA turn 2 241
a,b,c See footnotes to Table 2.3.
88
69WTLEP
73 with a score of 17 (Table 2.3, line 3). An additional Ph.D.-12 peptide
was found to align to the core of the FecA Ton box (81
ALTV84
) with a score of 13
within a 5-residue scoring window (Table 2.3, line 9). Two Ph.D.-12 and three
Ph.D.-C7C peptides were determined by RELIC/MATCH to align to FecA
periplasmic turn 8. RELIC/HETEROalign similarity scores that mapped to the
periplasmic surface of FecA indicate that the FecA Ton box region (Figure 2.4B,
region I) is not as highly scored as turn 8 (Figure 2.4B, region II). Furthermore,
RELIC/HETEROalign identified a peptide at turn 6 as being another highly
scored region (yellow), suggesting another possible TonB-binding site undetected
by RELIC/MATCH.
In contrast to FhuA, BtuB, and FecA, RELIC/MATCH did not identify a
cluster of affinity-selected peptides which could align at FepA periplasmic turn 8.
As with the other OM receptors, a peptide cluster aligned to a region proximal to
the Ton box of FepA. In this case, the cluster was comprised of five Ph.D.-12
peptides aligned in a region immediately N-terminal to the FepA Ton box.
Clusters were found to align to FepA periplasmic turns 1 and 2. At FepA turn 1,
two Ph.D.-12 peptides and two Ph.D.-C7C peptides aligned; at FepA turn 2, four
Ph.D.-C7C peptides were found to align (Figure 2.3C). RELIC/HETEROalign
similarity scores that mapped to the periplasm-exposed surface of FepA displayed
a high degree of heterogeneity. However, turns 1 and 2 (Figure 2.4C, regions II
and III, respectively) are prominent as highly-scored regions.
89
Figure 2.3. Alignments of affinity-selected peptides to the Ton box regions and periplasm-
exposed turns of BtuB, FecA and FepA as identified by RELIC/MATCH. Peptides from the
Ph.D.-12 library are aligned below the protein sequences and peptides from the Ph.D.-C7C library
are aligned above the protein sequences. See Tables 2.3 and 2.4 for respective peptide match
scores and window sizes. Alignment positions are highlighted according to their pairwise
alignment score: +4, black background and white character; +1, grey background and dark
character. A. matches to BtuB Ton box region and to periplasm-exposed turn 8; B. matches to
FecA Ton box region and periplasm-exposed turn 8; C. matches to FepA Ton box region and
periplasm-exposed turns 1 and 2.
A B
C
90
Figure 2.4. RELIC/HETEROalign similarity scores mapped to the periplasm-exposed surfaces of
BtuB (PDB code 1NQE), FecA (PDB code 1KMO), and FepA (PDB code 1FEP). See legend to
Figure 2.2 for details. The normalized RELIC/HETEROalign similarity scores mapped to the
molecular surfaces shown in A to C are averages of Ph.D-12 and Ph.D.-C7C similarity scores. A.
BtuB; Roman numerals correspond to RELIC/MATCH alignment clusters shown in Figure 2.3A:
(I) BtuB Ton box; (II) BtuB periplasmic turn 8; B. FecA; Roman numerals correspond to
RELIC/MATCH alignment clusters shown in Figure 2.3B: (I) FecA Ton box; (II) FecA
periplasmic turn 8; C. FepA; Roman numerals correspond to RELIC/MATCH alignment clusters
shown in Figure 2.3C: (II) FepA periplasmic turn 1; (III) FepA periplasmic turn 2. The region N-
terminal to FepA Ton box (Figure 2.3C, I) is not visible in 1FEP and therefore not shown.
B A C
91
2.3.5 Interactions of TonB and [FhuA peptide-MBP] fusion proteins in vitro
Oligonucleotides encoding peptides that are identical to FhuA sequences
within the Ton box/switch helix, cork/β1 strand, and periplasmic turn 8 were
synthesized, annealed, and cloned into a pMal-pIII expression vector. The vector
encodes maltose-binding protein (MBP); FhuA peptide sequences were fused to
the N-terminus of MBP. Peptide fusions corresponded to the following regions in
FhuA: [Ton box-MBP] and [Fhu switch-MBP] both representing region I; [Fhu
cork1-MBP], [Fhu cork2-MBP], and [Fhu cork3-MBP] representing region II; and
[Fhu turn-MBP] representing region III. Amino acid sequences for the FhuA
peptides are listed in Materials and methods. The [FhuA peptide-MBP] fusion
proteins were expressed, purified from E. coli periplasmic extracts, and tested for
their ability to bind to recombinant H6.'TonB in vitro by means of ELISA.
Despite DNA sequence fidelity of cloning and repeated attempts of protein
expression, we were unable to obtain [Fhu cork2-MBP]; the remaining five fusion
proteins were purified to homogeneity in abundant amounts (10 mg l-1
of culture).
H6.'TonB was applied to wells of a Ni-NTA coated microtitre plate. By
tethering the N-terminal hexahistidine tag to the Ni-NTA surface, the C-terminus
of H6.'TonB would be oriented outwards, mimicking the in vivo orientation of
TonB. This strategy markedly increased sensitivity of the assay compared to
randomly coated H6.'TonB (data not shown). Binding of each [FhuA peptide-
MBP] to H6.'TonB was detected (Figure 2.5) in levels greater than those afforded
by wild-type MBP. Binding was specific; incubation of a mixture containing
equimolar amounts ferricrocin-bound FhuA and a given [FhuA peptide-MBP]
92
reduced ELISA signals compared to incubations with [FhuA peptide-MBP] alone.
A continuous FhuA binding landscape containing the sum of all potential binding
sites was presented to H6.’TonB; fusion proteins displaying FhuA peptide
sequences were effectively out-competed. The level of competition was similar
among all fusion proteins tested and resulted in minimal binding, comparable to
the wild-type MBP control. Equivalent amounts of H6.'TonB were coated per
well; nearly complete competition of all TonB binding sites was achieved. These
data therefore establish that RELIC-identified regions of FhuA interact in vitro
with TonB when displayed as N-terminal peptide fusions to MBP.
Figure 2.5. Interactions of TonB and [FhuA-MBP] fusion proteins in vitro. H6.'TonB (20 pmol)
was incubated with MBP proteins (200 pmol) bearing FhuA sequences as N-terminal fusions:
MBP (wild-type maltose-binding protein); Ton box (6EDTITVTAA
14); Fhu switch
(21
AWGPAAT27
); Fhu cork1 (143
SSPGGLL149
); Fhu cork3 (153
SKRPTTEPL161
); Fhu turn
(583
AKAALSA589
). Numbering is according to the mature FhuA protein sequence. Fusion proteins
bound to H6.'TonB were detected by development of PNPP at 405 nm after incubation with the
appropriate antibodies (see Materials and methods). Black bars indicate signals observed due to
interactions between MBP-fusions and H6.'TonB after 180 min. White bars indicate signals
observed when equimolar amounts of [FhuA peptide–MBP] fusion and FhuA (200 pmol each)
were simultaneously incubated with H6.′TonB for 180 min.
93
2.4 Discussion
To identify surfaces on bacterial OM receptors in addition to the well-
characterized Ton box that might provide binding sites for TonB, we adopted
experimental strategies of phage display. Peptides were affinity-selected, based
upon their favourable interactions with H6.'TonB. Decreased amino acid
diversity and information contents were used to gauge the statistical merit of our
pool of affinity-selected peptides and indicated enrichment compared to randomly
selected peptides from the parent libraries. Statistical analyses of this pool of
peptides confirmed their affinity towards H6.'TonB; they were not selected by
nonspecific associations.
Using this approach, we identified novel TonB-binding regions on
surfaces of OM receptors that are known to interact with TonB. These regions
localized to periplasm-accessible surfaces. We also identified regions that are
TonB-inaccessible, based upon known crystal structures.
Our use of RELIC/MATCH identified regions on the periplasm-accessible
surfaces of the OM receptors FhuA, BtuB, FecA and FepA with sequence
similarities to the pool of TonB affinity-selected peptides. For the OM receptors
examined, the regions were all similarly located: the Ton box/switch helix region;
the cork domain/1 strand; and one or more periplasmic turns. Because these
three regions do not reveal obvious signature sequences for interactions with
TonB, we conclude that molecular recognition is based upon particular
conformations of peptides within the binding surfaces.
94
Noteworthy is that the Ton box, a region of known interaction between
TonB and FhuA, was not overly represented among our affinity-selected peptides.
The recent report of Peacock et al. (75) examined interactions between a TonB
construct containing residues 103–239 and synthetic Ton box peptides from
various OM receptors. Isothermal titration calorimetry confirmed weak affinities
between TonB and the Ton box peptides; Kd’s were in the micromolar range.
Complemented by findings that TonB recognizes a conformation of the Ton box
rather than a strict amino acid sequence, this outcome suggests that few phage-
displayed peptides adopted conformations that TonB might recognize. Even if
there were some greater population of phages with sequences similar to the Ton
box regions of OM receptors, their affinities towards TonB might have been too
low and thus may have escaped selection.
Periplasmic turns identified on the OM receptors were significantly
represented in our affinity-selected peptide populations. The Ph.D.-C7C affinity-
selected peptides that mapped to these turns possess conformations, which impose
significant selection constraints for interactions with immobilized TonB. Their
identification is meaningful because they probe regions within TonB that are
similarly constrained. Thus, the periplasmic turns identified in the OM receptors
apparently accommodate bound TonB and represent novel sites of contact beyond
the Ton box. Identification of these regions is consistent with previous findings
of Braun et al. (158), Howard et al.(160), Scott et al (159), and Killman et al.
(161), demonstrating that the FhuA -barrel possesses binding sites necessary to
95
support TonB-dependent function. OM receptor turns identified in the present
study therefore represent previously inferred TonB binding sites.
The most striking regions within FhuA that were identified by phage
display as interactors with TonB were those of the cork region. Not all identified
regions within the cork are completely exposed to the periplasm, yet when
selected sequences were displayed as [FhuA peptide-MBP] fusions, they were
shown to interact with TonB in vitro. It is difficult to envision TonB’s association
with regions inside the FhuA cork domain, even if the cork were partially
displaced from the barrel. However such regions might represent points of
contact between TonB and FhuA at different stages during siderophore transport.
Hence, our analyses primarily focused on regions clearly exposed to the periplasm
because these regions would most likely represent initial binding sites for TonB.
Our recent biophysical data (77) illustrated TonB’s association with apo-FhuA in
a 2:1 ratio; addition of siderophore enhanced this interaction. We advocated a
model whereby there is an initial encounter complex between FhuA and a
monomer of TonB, followed by a kinetically limiting TonB rearrangement that
facilitates recruitment of a second TonB monomer. Deletion of cork residues 21–
128 of FhuA caused an alteration in the stoichiometry of the complex, as did
deletion of the TonB proline-rich region. In both cases, only 1:1 TonB-FhuA
complexes were observed; however, SPR analysis indicated that these complexes
formed with high affinity. We interpreted these 1:1 complexes as the encounter
complex; multiple sites of interaction are positioned within this complex. Our
phage display data agree with this interpretation because we observe discrete
96
regions of FhuA that interact with TonB. These results do not exclude aspects of
TonB function that may occur at the CM. The report of Ghosh and Postle (79)
showed that TonB can form dimers that were previously unidentified through in
vitro studies. However, this same report indicated that monomeric TonB
associated with the OM. It is likely that the oligomeric state of TonB is dynamic
throughout the energy transduction cycle; perhaps energy from the proton motive
force unwinds the CM-localized TonB dimer, resulting in a monomeric form of
TonB competent to interact with OM receptors.
Figure 2.6 illustrates a representation of FhuA’s periplasmic surface;
regions that interact with TonB in vitro are highlighted in red. Significantly, the
highlighted regions form a continuous surface to which TonB can bind. For apo-
FhuA (Figure 2.6A), this surface appears segmented; the switch is coiled near turn
8 where its helix conformation is stabilized by hydrophobic interactions. Upon
siderophore binding, the gap between these two surfaces is bridged; the switch
helix unwinds and engages space nearer the cork/1 strand (Figure 2.6B).
Bridging of this space might facilitate stronger interactions of TonB with the cork.
Upon activation via ExbB-ExbD, TonB may then exert additional conformational
changes within the cork that permit passage of the siderophore.
TonB apparently facilitates OM receptor activity in part through its
interaction with the -barrel. Our data are consistent with previous findings (158-
161) because we identified periplasmic turn 8 and the shallow loop leading into
1 strand as binding sites for TonB. Turn 8 forms part of the continuous surface
97
that bridges the Ton box/switch helix region to the proximal cork regions and it
likely represents one of the initial points of contact with the OM receptor.
To further elucidate molecular mechanism, we examined the electrostatic
surfaces of both the periplasmic faces of FhuA (apo- and ligand-loaded) and the
most recent NMR structure of monomeric TonB. The electrostatic surface of apo-
FhuA contains an electronegative surface near the cork/1 strand that we
identified from phage display as being a potential binding surface for TonB.
Binding of siderophore causes unwinding of the switch helix and presents an
additional electronegative surface to which TonB can bind. This additional
surface is nearly identical to the surface that we show to interact with TonB in
vitro.
Through automated docking of monomeric TonB to the periplasmic face
of FhuA, we obtained an unbiased low energy complex that places TonB directly
over the RELIC-identified surfaces (Figure 2.6C). This surface orients TonB such
that its electropositive Arg-171 fits within the electronegative cleft formed by
residues Glu-159, plus Asp-186 and Asp-187 in FhuA’s periplasmic turn 1
(Figure 2.6D). Significantly, this region was identified by HETEROalign from
affinity-selected peptides in both the Ph.D.-12 and Ph.D.-C7C libraries (Figure
2.2C). In this docked conformation, TonB Gln-160 is oriented toward the lumen
of the FhuA barrel within a space near the switch helix and presumably the Ton
box (Figure 2.6 D). We propose that this TonB–FhuA docking solution may
represent an initial encounter complex between one molecule of TonB and one
molecule of FhuA.
98
Figure 2.6. TonB-binding surfaces on the periplasmic face of FhuA. Periplasmic faces (green) of
ligand-free FhuA (panel A, PDB code 2FCP) and ligand-loaded FhuA (panel B, 1FCP) are
represented in the same orientation as in Figure 2.2. Indicated in red are the regions shown to
interact in vitro with TonB. Panel C. Docking of monomeric TonB to FhuA: view of the
periplasmic surface of FhuA (see Figure 2.2D for orientation) shaded in green; residues from the
peptides Fhu switch, Fhu cork 1, Fhu cork 2, Fhu cork 3, and Fhu turn shaded in red. The three-
dimensional NMR model of the TonB C-terminal domain (PDB code 1XX3) was tested by the
computer program AUTODOCK (171) for its ability to dock to the periplasmic surface of FhuA
(PDB code 1FCP). Using Lamarckian Genetic Algorithm (LGA) routines (10 LGA runs in total;
population size: 50; max. number of energy evaluations: 250,000), a cluster of three low-energy
docked TonB conformations was obtained (average docking energy 37.63 kcal mol-1
). The lowest
energy (35.7 kcal mol-1
) member of the cluster of docked TonB 151–239 molecules is shown as a
ribbon representation (yellow). Panel D. View of contacts between docked TonB 151–239
(yellow ribbon) and FhuA (green ribbon). TonB residues and labels are shown in blue; FhuA
residues and labels are shown in red. The N-terminus of the FhuA structure (PDB code 1FCP,
residue 19) is indicated by an asterisk.
99
The recombinant H6.’TonB used in this study is incapable of becoming
energized. However, Howard et al. (160) reported that overexpressed, truncated
TonB constructs exported to the periplasm blocked TonB-dependent processes by
interfering with endogenous TonB function. By associating with OM receptors,
the constructs effectively competed for available TonB binding sites on FhuA and
FecA. They concluded that the interactions were meaningful because the
truncated TonB constructs recognized ligand-loaded receptors despite an inability
to accept energy input from ExbB–ExbD. The interactions were not fortuitous;
they did not occur by non-productive intrinsic affinity between TonB and OM
receptors. Our results are consistent with those of Howard et al., establishing that
we have identified biologically relevant sites of interaction between TonB and
OM receptors extending beyond the Ton box and that include regions of the -
barrel.
In addition to identifying periplasm-accessible TonB binding sites on OM
receptors, we report that our phage display strategies identified distal regions on
OM receptors of potential interaction with TonB. These regions were near the
cork apex and on extracellular loops of FhuA and of other OM receptors (region
II, cork1 designation). Because there are no data demonstrating TonB’s insertion
into the receptor lumen nor interacting with such extracellular regions, we can
only surmise what such interactions might imply. Nevertheless, it is intriguing
that some of our affinity-selected peptides mapped to an apical region of the FhuA
cork domain (Figure 2.1B; 145
PGGLL149
) that structurally overlaps a region on
FepA shown to be important in receptor function (172,173). In FepA, these
100
residues are located at the extracellular surface of the cork and they cluster near
the barrel wall forming the structurally conserved quadrupole or lock region. By
employing different mutagenesis strategies, Barnard et al. (172) and Chakraborty
et al. (173) demonstrated that mutations of these residues affected ligand
transport, but not ligand binding. An intriguing interpretation of their results is
that this region might somehow be involved in interactions with TonB. Our
affinity-selected peptides point to a similar region within FhuA that is on the
extracellular side of the cork and that clusters near the barrel wall, a region that
we designated cork1. MBP constructs bearing FhuA fusions of this region were
able to bind to TonB in vitro. Using different experimental tools, the studies of
Barnard et al., Chakraborty et al. and our phage display direct attention towards
possible interactions with TonB at regions removed from the periplasmic space.
It is tempting to propose that such a sterically occluded region might interact with
TonB. Cork-barrel interfaces are highly hydrated by water molecules acting as
bridges in a hydrogen bonding network between the two domains (104,151). This
network is similar to those found in transient protein–protein interfaces that
engage in gross domain movements or conformational change. Furthermore, their
analysis identified a region of structural conservation among cork domains of
TonB-dependent OM receptors that they referred to as the latch. The latch
(residues 139–141 in FhuA) was suggested to play a role in the transport process,
perhaps involving an unfolding event during transport. The latch is immediately
adjacent to the cork1 region that we identified in our phage display study and
which we showed to interact with TonB. These concordant observations merit
101
continuing experimentation to provide additional insights into the molecular
mechanisms of bacterial OM receptors and the role of TonB.
2.5 Acknowledgements
This research was supported by operating grants to J.W.C. from the
Canadian Institutes of Health Research (CIHR) and to L.M. from National
Institutes of Health. J.-N. Gagnon received a PGS-M scholarship and M.D., a
USRA scholarship from Natural Sciences and Engineering Research Council of
Canada. Canada Foundation for Innovation provided infrastructure for molecular
modelling to the Montreal Integrated Genomics Group for Research on Infectious
Pathogens. Sheldon Biotechnology Centre at McGill University is supported by a
Multi-user Maintenance Grant from CIHR. We appreciate contributions of
experimental materials by A. Clements, N.M. Desy, C. Ng; critical reviews of the
manuscript by M. A. Hancock and G. Marczynski; and editorial support by J.A.
Kashul.
102
Preface to chapter 3
Regions of interaction between TonB and FhuA were predicted by our use
of phage display in chapter 2. Our later crystal structure of TonB bound to FhuA
provided the proof of principle that our strategies are predicting biologically
relevant interactions. In chapter 3, this strategy extends to the prediction and
demonstration that TonB interacts with periplasmic binding protein, FhuD.
Described therein is the use of phage display to predict and localize regions of
interaction between TonB and FhuD. These predictions are then confirmed with
biophysical experiments that revealed the stoichiometry and affinity between
TonB and FhuD. Furthermore, we demonstrated that TonB can interact with the
OM receptor FhuA and FhuD at the same time. We conclude with a model of
how such a ternary complex might facilitate siderophore uptake.
103
Chapter 3
Interactions between TonB from Escherichia coli and
the periplasmic protein FhuD
David M. Carter, Isabelle R. Miousse, Jean-Nicolas Gagnon, Éric Martinez,
Abigail Clements, Jongchan Lee, Mark A. Hancock, Hubert Gagnon,
Peter D. Pawelek and James W. Coulton
Journal of Biological Chemistry (2006). 281: 35413-35424
Copyright 2006, American Chemical Society
104
3.0 Summary
For uptake of ferrichrome into bacterial cells, FhuA, a TonB-dependent
outer membrane receptor of Escherichia coli is required. The periplasmic protein
FhuD binds and transfers ferrichrome to the cytoplasmic membrane-associated
permease FhuB/C. We exploited phage display to map protein–protein
interactions in the E. coli cell envelope that contribute to ferrichrome transport.
By panning random phage libraries against TonB and against FhuD, we identified
interaction surfaces on each of these two proteins. Their interactions were
detected in vitro by dynamic light scattering and indicated a 1:1 TonB–FhuD
complex. FhuD residue Thr-181, located within the siderophore binding site and
mapping to a predicted TonB-interaction surface, was mutated to cysteine. FhuD
T181C was reacted with two thiol-specific fluorescent probes; addition of the
siderophore ferricrocin quenched fluorescence emissions of these conjugates.
Similarly, quenching of fluorescence from both probes confirmed binding of
TonB and established an apparent KD of approximately 300 nM. Prior saturation
of FhuD’s siderophore binding site with ferricrocin did not alter affinity of TonB
for FhuD. Binding, further characterized with surface plasmon resonance,
indicated a higher-affinity complex with KD in the low nanomolar range.
Addition of FhuD to a pre-formed TonB–FhuA complex resulted in formation of
a ternary complex. These observations lead us to propose a novel mechanism in
which TonB acts as a scaffold, directing FhuD to regions within the periplasm
where it is poised to accept and deliver siderophore.
105
3.1 Introduction
Iron, an essential nutrient for almost all bacterial species, is required for
metabolic processes including electron transfer, oxygen activation and
biosynthesis of amino acids and nucleosides (145). However, Fe3+
is scarce in the
extracellular environment. Gram-negative bacteria have evolved transport
processes that utilize siderophores to scavenge extracellular Fe3+
by high-affinity
chelation. Different siderophore receptors are expressed at the bacterial outer
membrane (OM), each with specificity for a particular metal-chelated
siderophore. Transport of receptor-bound siderophores into the periplasm
requires contribution of energy provided by the TonB–ExbB–ExbD complex that
is anchored in the cytoplasmic membrane (CM). TonB spans the periplasm to
make contacts with cognate OM receptors. By harnessing energy produced from
the proton motive force, TonB may propagate conformational changes to OM
siderophore receptors, resulting in release of siderophore into the periplasm.
The ferrichrome transport system consists of four proteins (FhuA, FhuB,
FhuC, and FhuD) expressed by Gram-negative bacteria. The FhuA protein
comprises two domains: an N-terminal globular cork domain is enclosed by a 22-
stranded C-terminal β-barrel domain (42,150). Connections between β-strands in
the barrel domain are such that long loops participating in ferrichrome binding are
exposed to the extracellular environment; short turns are exposed to the
periplasm. Uptake of ferrichrome is a TonB-dependent process, mediated by
contacts between TonB and the OM receptor FhuA. TonB is elongated and has
three domains: an N-terminal domain anchored in the CM, an intermediate
106
domain containing Pro-Glu and Pro-Lys repeats, and a globular C-terminal
domain with a central β-sheet and two α-helices. To date, structural data is only
available for the C-terminal domain (72,74,75). We recently solved the crystal
structure of the 1:1 TonB–FhuA complex (96). The C-terminal domain of TonB
makes extensive contacts with the N-terminal consensus Ton box of FhuA, as
well as residues Ala 26 and Glu 56 of the cork domain and with periplasmic turns
8 and 10. These contacts orient TonB such that it may mediate conformational
disruption of the internal cork domain of FhuA, allowing for passage of
siderophore into the periplasm.
Although recent structural and biophysical data have clarified initial steps
of the siderophore transport cycle involving TonB–receptor interactions, little is
known about the fate of the siderophore once transported into the periplasm.
Specifically, the molecular mechanisms of siderophore transport from periplasm
to cytoplasm a largely uncharacterized. FhuD in the periplasm binds
hydroxamate siderophores ferrichrome, coprogen and aerobactin (174). Loss of
FhuD function in vivo prevented growth of E. coli under iron-limiting conditions
when ferrichrome, coprogen, or aerobactin were used as the sole iron source,
suggesting that FhuD is a necessary component of the hydroxamate siderophore
transport system (174). FhuD was reported to interact with regions of the CM-
embedded permease FhuB. Interactions between FhuD and FhuB have been
demonstrated by cross-linking studies, protease protection assays (113), and
ELISA (132). Interaction of FhuB with FhuD is apparently independent of
siderophore binding by FhuD (113). Taken together, these results suggest that
107
FhuD functions as a carrier protein: ferrichrome released from the OM receptor is
delivered by FhuD to the permease. The integral membrane protein FhuB then
translocates the siderophore into the cytoplasm mediated by ATP hydrolysis of
FhuC (175).
The crystal structure of FhuD in complex with gallichrome, a ferrichrome
analogue, has been reported (111), as well as structures of FhuD in complex with
albomycin, coprogen, and Desferal (112). The fold of this 32 kDa protein is
bilobal: globular N- and C-terminal domains are connected by a long α-helix that
confers rigidity to the protein. The siderophore binding site residing in a shallow
cleft between the two lobes is hydrophobic, having predominantly aromatic
residues. Siderophore binds to FhuD through both hydrophobic and hydrophilic
interactions. Methylene carbon atoms in the siderophore form hydrophobic
interactions with numerous aromatic FhuD residues in the binding cleft.
Hydrogen bonds are formed between hydroxamate groups of the siderophore and
FhuD residues Arg-84 and Tyr-106. A hydrogen bond with the siderophore is
also formed with FhuD residues Asn-215 and Ser-219 through an intermediate
water molecule.
The overall fold of FhuD is similar to that of BtuF (114,120), the
periplasmic cobalamin-binding protein of E. coli. Periplasmic metal-binding
proteins TroA (176) and PsaA (177) are also structurally related to FhuD. These
proteins share a fold distinct from those of classical periplasmic proteins such as
maltose binding protein (178). However unlike maltose binding protein, FhuD
does not exhibit gross conformational rearrangements upon ligand binding. The
108
linker connecting the N-terminal and C-terminal domains in FhuD is a kinked α-
helix that crosses these domains only once. The structure of FhuD and the
hydrophobicity of the siderophore binding site suggest that large-scale opening
and closing of the binding site does not occur upon siderophore binding and
release (178). Molecular dynamics simulations also suggest that FhuD is
conformationally rigid, but that subtle conformational differences in the C-
terminal domain between the apo- and holo-forms may be sufficient for
discrimination by FhuB (119).
What molecular events result in capture of ferrichrome by FhuD following
its TonB-dependent release from FhuA? Given the apparent weak affinity (1 M)
of ferrichrome for FhuD (113), binding is unlikely to be a diffusion-governed
process. Efficiency of siderophore capture would be enhanced by positioning a
binding protein proximal to the lumen of the OM receptor. This organization
would promote direct transfer of ferrichrome from FhuA to FhuD. Here we report
the first biophysical evidence that TonB specifically interacts with FhuD.
Discrete regions of protein–protein interactions on the surfaces of both FhuD and
TonB were identified by phage display. Interactions were confirmed by dynamic
light scattering, fluorescence spectroscopy and surface plasmon resonance. Our
results suggest that siderophore released from FhuA during the transport cycle is
transferred to FhuD via a coordinated transfer mechanism mediated by TonB.
Hence, TonB would act as a periplasm-spanning scaffold, directly connecting
siderophore transport events between the OM and CM.
109
3.2 Materials and methods
3.2.1 Bacterial strains, phage libraries, and media
Random peptide phage libraries Ph.D.-C7C and Ph.D.-12 were purchased
from New England Biolabs (NEB); E. coli ER2738 also from NEB was used for
amplification and titration of phage M13 pools. E. coli ER2566 was used to
express recombinant TonBs (77); E. coli BL21 (DE3) pLysS was used to express
recombinant FhuDs. Plasmid pCMK01 expresses a hexahistidine-tagged TonB
32-239 (25 kDa; hereafter identified as TonB) and pWA01 expresses a
hexahistidine-tagged TonB 32-239 with an engineered cysteine residue at its
amino terminus (hereafter identified as Cys-TonB) (94,95). FhuD was expressed
from pMR21 provided by W. Köster (VIDO, Saskatoon, SK); the N-terminus of
FhuD containing the signal sequence was removed and replaced by a
decahistidine tag (179) (32 kDa; hereafter identified as FhuD). Plasmid pMR21
was commercially mutated to cysteine at Thr181 by Norclone Biotech
Laboratories (London, ON); this protein is hereafter identified as FhuD T181C.
Mutagenesis was confirmed by DNA sequencing at Sheldon Biotechnology
Centre, McGill University (Montreal, QC). All bacteria were cultured in Luria
Bertani (LB) broth containing antibiotics when necessary.
3.2.2 Chemicals and reagents
5-Bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal) and
isopropyl-β-D-thiogalactopyranoside (IPTG) were purchased from Biovectra
110
(Charlottetown, PE). Protein-grade Tween-20 was purchased from Calbiochem.
Antibiotics were purchased from Sigma-Aldrich. Ni-NTA resin used for protein
purifications was purchased from Qiagen. The reducing agent Tris-(2-
carboxyethyl)phosphine (TCEP) and fluorescent dyes 5-((((2-
iodoacetyl)amino)ethyl)amino)naphthaline-1-sulfonic acid (AEDANS) and 7-
diethylamino-3-((((2-maleimidyl)ethyl)amino)carbonyl)coumarin (MDCC) were
purchased from Invitrogen.
3.2.3 Protein purification
TonB and Cys-TonB were purified as described previously (77). To
purify overexpressed FhuD or FhuD T181C, cell pellets were suspended in 50 ml
of buffer A containing 50 mM Tris pH 8.2, 150 mM NaCl and 5 mM imidazole
plus one Complete mini EDTA-free protease inhibitor cocktail tablet (Roche);
0.16 mg/ml lysozyme and 16 µM phenylmethyl sulphonylfluoride (PMSF) were
then added. Cells were shaken at room temperature for 30 min, followed by
addition of 0.04 mg/ml DNase, 0.04 mg/ml RNase, and an additional 16 µM
PMSF. To lyse bacteria, cells were passed twice through an Emulsiflex-C5
(Avestin, Ottawa, ON). Cell lysate was centrifuged (27,000 × g, 4 °C) for 50 min
and filtered through 0.45 m syringe filters. Filtered cell extracts containing
FhuD or FhuD T181C were applied to Ni-NTA resin equilibrated with buffer A.
FhuDs were eluted with 50 mM Tris pH 8.2 containing 125 mM imidazole,
pooled and applied to a POROS HQ 20 anion exchange column (Applied
Biosystems). Bound proteins were washed with 50 mM Tris pH 8.2 containing
111
125 mM imidazole, eluted with 160 mM NaCl and applied inline to a POROS MC
20 column (Applied Biosystems). After extensive washing, proteins were eluted
with 50 mM Tris pH 8.2 and 120 mM imidazole and applied to a second POROS
HQ 20 column, washed as described above, and eluted with 180 mM NaCl in 50
mM Tris pH 8.2. Purified proteins were dialyzed in a 24,000 Mr cut-off dialysis
membrane (SpectraPor) for 16 h at 4 °C in 100 mM Hepes, 150 mM NaCl, pH
7.4. Homogeneity of FhuDs was confirmed by SDS-PAGE and silver staining of
750 ng of total protein. Concentrations of protein were determined by either a
Bradford or BCA assay using bovine serum albumin as standard.
3.2.4 Phage display
Phage panning against TonB as target was described previously (180).
Purified FhuD was diluted to 100 g/ml in TBS (Tris-buffered saline: 50 mM
Tris, pH 7.5, 150 mM NaCl) and 150 l of protein was adsorbed to a polystyrene
microtitre plate (Nunc Maxisorp). Plates coated with immobilized FhuD were
incubated for 16 h at 4 °C followed by blocking (2 h at 37 °C) with TBS
containing 5 mg/ml bovine serum albumin. The unselected phage library (NEB)
was then added. Phage panning, clone isolation, DNA sequencing, and
bioinformatic analyses were performed as described previously (180).
3.2.5 Dynamic light scattering
Light scattering was measured from purified TonB and FhuD dialyzed
twice (18 h, 4 °C) in 100 mM Hepes, pH 7.4 containing 150 mM NaCl. Purified
112
[Fhu switch-MBP] fusion protein (180) (containing FhuA residues
21AWGPAAT
27 fused to the N-terminus of maltose binding protein) and BSA
were dialyzed against the same buffer and used in DLS measurements as positive
and negative controls, respectively. TonB (4.0 M) and FhuD (3.0 M) were
separately analyzed as discrete scattering species. Similarly, BSA (1.5 µM) and
[Fhu switch-MBP] (2.5 µM) were analyzed separately.
For a 1:1 molar ratio of TonB to FhuD, each protein at 1.7 µM was mixed
prior to recording DLS readings. For 1:1 mixtures of TonB with BSA or with
[Fhu switch-MBP], proteins were each at 1.0 µM. Protein mixtures were
incubated for 30 min at room temperature prior to centrifugation and analysis.
Data acquisition was performed in a 12 µl quartz cuvette at 20 °C using a
temperature controlled DynaPro E-50-830 dynamic light scattering instrument
(Protein Solutions, Charlottesville, VA). The scattering signal was measured at a
wavelength of 824.9 nm and an angle of 90°. Data were collected for 7000 s with
a 10 s averaging time, and replicated with two independent protein preparations.
From the Dynamics v.6.3.18 software (Protein Solutions, Charlottesville, VA),
data were filtered (baseline < 1.01 and sum of squares < 500) before exporting to
Sedfit v.9.3(143). Analyses of hydrodynamic radii (Rh) were performed using the
continuous intensity distribution model (143) in Sedfit at a resolution of 100 for
radii between 1 and 50 nm. Buffer densities and viscosities were set to 1.00442
and 0.01065, respectively, as determined by Sednterp v.1.08
(www.jphilo.mailway.com). All values of Rh from the Dynamics software
113
exhibited less than 14% polydispersity, except for the TonB–FhuD mixture and
[Fhu switch-MBP] (21% and 19%, respectively).
For Sedfit analyses of Discrete Non-interacting Species (DNS),
autocorrelation data sets were imported from the Dynamics software package and
fit to a single species field autocorrelation function. Values of s for TonB were
previously determined (77). Using analytical ultracentrifugation, we determined
by sedimentation velocity experiments sedimentation coefficients for FhuD and
the TonB–FhuD complex: 2.27 s and 3.5 s, respectively. From literature reports, s
values for MBP (181) and BSA (182) were obtained. All were constrained in
DNS analyses. Molecular mass values for discrete scattering species, either
uncomplexed TonB, uncomplexed FhuD, or 1:1 heterocomplexes were initially
set to predicted values and then refined by nonlinear regression until rmsd errors
were minimized. In addition to proteins TonB, FhuD, [Fhu switch-MBP] or
complexes formed by these proteins, two scattering species were observed; the
Dynamics program predicted these uncharacterized species to have hydrodynamic
radii of ~1 nm and ~100 nm respectively. Hydrodynamic parameters for these
species were factored into DLS analyses to optimize fits to the autocorrelation
function.
3.2.6 Fluorescence spectroscopy
The fluorescent dye AEDANS was conjugated to FhuD and FhuD T181C
in a reaction buffer of 100 mM Hepes pH 7.4, 150 mM NaCl. Following
reduction of disulfide bonds with a ten-fold molar excess of TCEP, dye was added
114
to a ten-fold molar excess. Conjugation proceeded in the dark with stirring for 4 h
at room temperature. Reactions were quenched by addition of β-mercaptoethanol.
Excess label was removed by exhaustive dialysis against four 1 l changes of 100
mM Hepes, pH 7.4 containing 150 mM NaCl in the dark at 4 °C. After dialysis,
free dye was present at picomolar concentrations. Conjugates were then
centrifuged at 18,000 × g for 30 min at 4 °C. Labeled proteins were stored at 4 °C
in the dark. Conjugation of FhuD T181C with the dye MDCC was performed as
described above except that MDCC was dissolved in DMSO prior to its addition
to protein. Efficiency of labeling (mol dye:mol protein) was calculated from
absorption data using the following tabulated (Invitrogen) molar extinction
coefficients: 5700 M–1
cm–1
at 336 nm for AEDANS and 50,000 M–1
cm–1
at 419
nm for MDCC and from protein concentrations as determined by protein assays.
Fluorescence data were collected with a Varian Cary Eclipse fluorescence
spectrophotometer. Emission spectra were recorded at excitation and emission
wavelengths of 280 nm and 340 nm respectively for intrinsic fluorescence
measurements; at 336 nm and 490 nm respectively for AEDANS-labeled FhuD
and FhuD T181C; and at 419 nm and 466 nm respectively for MDCC-labeled
FhuD T181C. Excitation and emission slits were set between 2.5−5 nm and 5−10
nm respectively. Measurements were taken in triplicate at 20 °C. Data were
corrected for changes in fluorescence intensity attributed to dilution of protein and
the minimal fluorescence contributions of Fcn, TonB and buffer (100 mM Hepes
pH 7.4, 150 mM NaCl).
115
Binding of Fcn to either FhuD (1.5 M) or FhuD T181C (0.5 M) was
monitored by recording the fluorescence emission after additions of Fcn up to a
ten-fold molar excess. For each data point, Fcn was added from a stock solution
and after three minute incubation, the change in fluorescence was recorded.
Titrations of labeled conjugates with either Fcn or TonB were conducted in an
identical manner. Fluorescence quenching was expressed as the percentage
decrease in fluorescence upon ligand addition compared to the theoretical
maximum whereby quenching would result in complete loss of fluorescence.
Data were fit (Sigmaplot) to an equation describing a rectangular hyperbola using
the single binding site model or to a sum of two hyperbolics using the model that
describes two independent binding sites.
3.2.7 Surface plasmon resonance (SPR)
Binding interactions between TonB and FhuD or between Cys-TonB and
FhuD were examined in real-time using BIACORE 2000/3000 instrumentation
with research-grade CM4 sensor chips (BiacoreAB, Uppsala, Sweden).
Experiments were performed in triplicate at 25 oC using filtered (0.2 m) and
degassed HBS-ET (50 mM Hepes pH 7.4, 150 mM NaCl, 3 mM EDTA, 0.05%
(v/v) Tween-20). EDC (1-ethyl-3-(3-dimethylaminopropyl)-carbo diimide), NHS
(N-hydroxysuccinimide), and PDEA (2-(2-pyridinyldithio)ethaneamine) were
from BiacoreAB. Protein-grade detergents (10% Tween-20, 10% Triton X-100,
30% Empigen) were from Calbiochem. All other chemicals were reagent grade
quality.
116
For amine coupling, TonB was immobilized according to a standard
Biacore protocol. For ligand thiol-coupling, 20 l of freshly mixed solution I
(200 mM EDC and 50 mM NHS in water) was injected (5l/min) over the sensor
chip activating carboxymethyl groups to reactive esters. Reactive thiol groups
were then introduced by a 30 l injection of freshly prepared solution II (80 mM
PDEA in 0.1 M sodium borate pH 8.5). Diluted Cys-TonB ligand (3 g/ml in 10
mM sodium acetate pH 4.5) was injected manually until ~120 RU were bound.
Finally, three injections (20 l) of freshly prepared solution III (50 mM L-cysteine
in 0.1 M sodium formate pH 4.3 containing 1 M NaCl) deactivated excess
reactive groups and removed any non-specifically bound ligand. Coupling
efficiencies were typically ~50%. Reference surfaces were prepared in a similar
manner without any ligand addition.
Immobilized TonB and Cys-TonB surfaces were washed overnight at 5
l/min in running buffer. Prior to use, FhuD analyte was dialyzed against HBS-
ET and immobilized TonB or Cys-TonB surfaces were conditioned at 50 l/min
using regeneration scheme A as follows: two 25 l injections each of (i) 0.05%
(v/v) Empigen, 0.5 M NaCl, 50 mM EDTA, 10 mM NaOH in HBS-ET, (ii) 0.05%
(v/v) Triton X-100, 0.5 M NaCl, 50 mM EDTA, 10 mM NaOH in HBS-ET, and
(iii) HBS-ET. For kinetic experiments, FhuD (0.1-1 M in the absence and
presence of a ten-fold molar excess of Fcn) was injected at 50 l/min (120 s
association + 120 s dissociation) over amine-coupled TonB or thiol-coupled Cys-
TonB. Surface performance and mass transfer tests confirmed that the ligand
density and regeneration conditions were appropriate. All acquired data were
117
double-referenced (183) and analyzed globally according to the simple 1:1
binding model (A + B = AB) or to the heterogeneous ligand model in the
BIAevaluation 4.1 software (BiacoreAB). Kinetic estimates represent fits to the
experimental data with χ2 values below 1.
Multi-component SPR analyses between FhuA, TonB and FhuD were
performed. Initially, amine-coupled TonB surfaces (250 RU) or thiol-coupled
Cys-TonB surfaces (100 RU) were prepared. Then, either a TonB–FhuA or a
TonB–FhuD binary complex was formed by injecting each analyte at 50 l/min.
By injecting FhuD over TonB–FhuA complexes or by injecting FhuA over TonB–
FhuD complexes, ternary complex formation was assessed.
3.3 Results
3.3.1 Identification of TonB-binding sites on FhuD by phage display
Following affinity selection versus immobilized TonB (180), 135 unique
disulfide-constrained peptides from the Ph.D.-C7C library and 105 unique linear
peptides from the Ph.D.-12 library were analyzed. These phage-displayed
peptides were scanned for their similarity to the primary sequence of FhuD using
the REceptor LIgand Contacts (RELIC) program RELIC/MATCH (168). Among
these sequences, 8 from the Ph.D.-C7C library (Table 3.1) and 13 from the Ph.D.-
12 library (Table 3.2) were found to share similarities with the primary sequence
of FhuD. The Ph.D.-C7C and Ph.D.-12 sequences were observed (Figure 3.1) to
cluster at four discrete regions along FhuD: loop 2 (region I), helix 2 (region II),
118
loop 8 (region III), and loop 23 (region IV). When these four regions were
mapped (Figure 3.2A) onto the surface of the three-dimensional structure of FhuD
(PDB code: 1EFD), they displayed a binding surface that overlaps the siderophore
binding site (Figure 3.2B). Regions I, II, and IV comprise a continuous binding
surface of approximately 17 Å x 17 Å that is formed at the base of the siderophore
binding pocket. In addition to loop 23, region IV also includes residues from
helix 13 near the C-terminus of FhuD. Although the surface-exposed residues in
region III could form a potential TonB-binding landscape with regions I, II, and
IV, residues 81-88 in loop 3 preclude formation of such a continuum. No FhuD
residues within this interval were identified by our phage display analysis,
suggesting that loop 3 is not TonB-binding. Were there a continuous landscape of
all four regions, it would require a displacement of FhuD’s loop 3 to
accommodate TonB binding. Recent molecular dynamics simulations on the
FhuD structure (119) indicated that the C-terminal domain of FhuD has more
overall mobility than the N-terminal domain. However, within the relatively
static N-terminal domain, loop 3 was observed to be the most mobile region.
3.3.2 Identification of FhuD-binding sites on TonB by phage display
Purified FhuD was immobilized and sequentially panned with the Ph.D.-
C7C library and with the Ph.D.-12 library. Panning yielded 109 unique sequences
from the Ph.D.-C7C library and 38 unique sequences from the Ph.D.-12 library.
Of these sequences, 15 from the Ph.D.-C7C library (Table 3.3) and 10 from the
Ph.D.-12 library (Table 3.4) were found by RELIC/MATCH to be similar to the
119
primary sequence of TonB. When aligned to TonB, these sequences were
observed (Figure 3.3) to cluster at three discrete TonB regions: an N-terminal
domain (region I), an intermediate domain (region II), and the C-terminal domain
(region III). Three-dimensional structural information has been reported
(72,74,75) only for the C-terminal domain of TonB. We mapped (Figure 3.4A)
the cluster from region III to the NMR structure of the TonB C-terminal domain
(PDB code: 1XX3). Region III forms a continuous solvent-exposed binding
surface on TonB (Figure 3.4B), adjacent to FhuA-binding residues that we
recently observed in the TonB–FhuA crystal structure (96). These observations
suggest a potential FhuA–TonB–FhuD ternary complex. Since the three-
dimensional structure of TonB residues in the N-terminal domain and
intermediate domains are unknown, we consider the possibility that residues from
TonB regions I and II could form a continuous FhuD-binding landscape with
TonB region III.
120
Table 3.1. RELIC/MATCH identification of TonB-affinity-selected Ph.D.-C7C peptides
corresponding to FhuD sequences.
Peptidea
Peptide match
scoreb
Scoring
window
(residues)
Region Alignment
positionc
PYGAALH 16 5 loop 2 26 PYGAALH 16 5 loop 23 245 YGGATLL 15 5 loop 23 245 QPAVANT 13 5 loop 2 28 SYLNVMH 13 5 loop 23 249 TGPLPNR 13 4 helix 2 43 NPTPEKR 13 4 loop 8 78 KPSSPPF 12 4 helix 2 40
aResidues contained in Scoring window are shown in bold.
bSum of pairwise comparisons of aligned residues within the Scoring Window.
cPosition of first residue of the Scoring window numbering from the N-terminus of the
mature protein.
Table 3.2. RELIC/MATCH identification of TonB-affinity-selected Ph.D.-12 peptides
corresponding to FhuD sequences.
Peptidea
Peptide match
scoreb
Scoring
window
(residues)
Region Alignment
positionc
LLADTTHHRPWT 17 7 loop 2 30 HWKHPWGAWDTL 16 7 loop 2 26 KVWSLEPPGPAA 15 5 helix 2 41 YSPPSPEPPRIK 15 5 loop 8 78 QDRGILVEPPRM 14 8 helix 2 36 DFDVSFLSARMR 14 8 loop 23 244 KLWELNPPQVRT 14 7 helix 2 37 SPAPTNNYTYRL 14 6 loop 2 30 TQPLGLLPSRHL 14 5 loop 2 22 QTALITIHHSLT 13 6 loop 2 32 YGNSLPPRLGPP 13 5 loop 23 245 LWAKLWVVPERA 12 5 helix 2 36 SANLSWRESWPT 12 5 loop 23 246
a,b,c See footnotes to Table 3.1.
121
Figure 3.1. Alignments of TonB affinity-selected peptides to FhuD as identified by
RELIC/MATCH. A. Region I: FhuD loop 2; B. Region II: FhuD helix 2; C. Region III: FhuD
loop 8; D. Region IV: FhuD loop 23. Peptides from the Ph.D.-12 library are aligned below the
FhuD sequence; peptides from the Ph.D.-C7C library are aligned above the FhuD sequence. See
Tables 3.1 and 3.2 for peptide match scores and window sizes. Alignment positions are
highlighted according to their pairwise alignment score: +4, black background and white
character; +1, grey background and dark character.
122
Figure 3.2. TonB-binding regions identified by phage display mapped to FhuD (PDB code
1EFD). A. Ribbon representation of FhuD (blue) with predicted TonB-binding regions shaded
yellow. Regions corresponding to RELIC/MATCH alignment clusters shown in Figure 3.1 are
indicated by Roman numerals: (I) loop 2; (II) helix 2; (III) loop 8; (IV) loop 23; B. molecular
surface representation of FhuD (blue) with predicted TonB-binding regions shaded yellow. The
bound ligand gallichrome from the 1EFD structure is shown in stick representation and colored by
atoms (carbon: white, nitrogen: blue, oxygen: red).
123
Table 3.3. RELIC/MATCH identification of FhuD-affinity-selected Ph.D.-C7C peptides
corresponding to TonB sequences.
Peptidea
Peptide match
scoreb
Scoring
window
(residues)
Region Alignment
positionc
PAPERPQ 16 6 N-term 39 HASPAHN 15 6 intermediate 121 VISAASQ 15 6 C-term 146 QSFPRQL 14 6 C-term 149 NRPSSWL 14 5 intermediate 119 TAENSSP 13 5 intermediate 124 KTSPAWI 13 5 intermediate 127 MTARTTS 13 5 intermediate 129 ISPAQSS 13 4 N-term 39 PAVPAKA 12 5 N-term 36 HLAPAAR 12 5 intermediate 127 KALMRTS 12 5 C-term 153 KPLFHNT 12 4 intermediate 122 HHWAPTR 12 4 intermediate 126 HNMPAQT 12 3 N-term 38
a,b,c See footnotes to Table 3.1.
Table 3.4. RELIC/MATCH identification of FhuD-affinity-selected Ph.D.-12 peptides
corresponding to TonB sequences.
Peptidea
Peptide match
scoreb
Scoring
window
(residues)
Region Alignment
positionc
LHTPWHLPAPEI 16 4 N-term 32 KSLSRHDHIHHH 15 6 C-term 153 YHSPPHTPPAPL 14 6 intermediate 122 SFVGLVELPQNL 14 5 N-term 31 VSRHQSWHPHDL 14 5 C-term 155 KTLTLPLSNTSK 13 6 intermediate 119 KIMRMPRLMTRN 13 6 C-term 147 LHFPLDYPQALG 13 5 C-term 145 WHSPWSTPPAPS 13 4 N-term 31 LHWPLYTPPASP 12 4 N-term 33
a,b,c See footnotes to Table 3.1.
124
Figure 3.3. Alignments of FhuD affinity-selected peptides to TonB. A. Region I: TonB N-
terminal domain; B. Region II: TonB intermediate domain; C. Region III: TonB C-terminal
domain. Peptides from the Ph.D.-12 library are aligned below the TonB sequence; peptides from
the Ph.D.-C7C library are aligned above the TonB sequence. See Tables 3.3 and 3.4 for peptide
match scores and window sizes. Alignment positions are highlighted according to their pairwise
alignment score: +4, black background and white character; +1, grey background and dark
character.
125
Figure 3.4. FhuD-binding region identified by phage display mapped to TonB (PDB code 1XX3).
A. Ribbon representation of the TonB C-terminal domain (yellow) with predicted FhuD-binding
region III shaded blue. The region corresponding to RELIC/MATCH alignment cluster shown in
Figure 3.3C is indicated by the Roman numeral (III); B. molecular surface representation of TonB
(yellow) with the predicted FhuD-binding region III shaded blue.
126
3.3.3 Detection of a TonB–FhuD complex by dynamic light scattering
To identify and characterize a TonB–FhuD complex, we employed
dynamic light scattering, a technique previously shown (184) to be effective for
analyzing protein–protein complexes. Analysis of hydrodynamic distribution
with the program Sedfit (143) revealed discrete hydrodynamic radii (Table 3.5)
for all proteins, each with rmsd values less than 0.0097. Given its larger frictional
ratio (77), TonB would exhibit a larger hydrodynamic radius (Rh) despite a lower
molecular mass compared to FhuD. Hence, TonB and FhuD exhibited similar Rh
values. The Rh obtained from an equimolar mixture of these proteins indicated
that a 1:1 heterocomplex had formed. As a control experiment, we observed an
increase in Rh when a [Fhu switch-MBP] fusion protein harbouring a previously
characterized (180) TonB-binding peptide was mixed with TonB, indicating
formation of a TonB–[Fhu switch-MBP] heterocomplex. No change in Rh was
observed after mixing TonB with BSA, as compared to the Rh for each individual
protein, indicating no formation of a TonB–BSA complex (data not shown).
aRh, hydrodynamic radius calculated from continuous intensity distribution model in
Sedfit v.9.3. bRmsd, root mean square deviation from best fits to the autocorrelation curve for Rh
analyses. cPercent mass from DNS analyses, molecular mass >1000 kDa accounted for less than
11% and the remainder was comprised of molecular mass <1.2 kDa. dRmsd for
discrete non interacting species analysis.
Table 3.5. DLS analysis of TonB, FhuD and MBP-switch fusion.
127
To estimate the molecular mass of a TonB–FhuD heterocomplex, a
Discrete Non-interacting Species (DNS) analysis was performed using Sedfit.
Results from DNS analyses clearly indicated formation of a 1:1 TonB–FhuD
heterocomplex. Molecular mass values for uncomplexed TonB and uncomplexed
FhuD determined by Sedfit from DLS autocorrelation data agree with their
predicted values (Table 3.5, rows 1 and 2). A 1:1 mixture of TonB and FhuD
resulted in formation of a scattering species with a refined molecular mass of 58
kDa (Table 3.5, row 3), corresponding to a 1:1 TonB–FhuD complex.
Both our phage display outcomes (180) and the X-ray crystallographic
structure (96) of the TonB–FhuA complex demonstrated that residues in the
switch helix region of FhuA interact directly with TonB. DLS analyses of [Fhu
switch-MBP] fusion protein uncomplexed or in complex with TonB yielded
results (Table 3.5, rows 4, 5) in agreement with these previous analyses. The
refined molecular mass for uncomplexed [Fhu switch-MBP] agreed with its
predicted molecular mass. When [Fhu switch-MBP] was mixed with TonB in a
1:1 molar ratio, an abundant scattering species of approximately 61 kDa was
observed, consistent with formation of a 1:1 heterocomplex between TonB and
the fusion protein.
3.3.4 Detection of a TonB–FhuD complex by fluorescence spectroscopy
Guided by the TonB-binding surface on FhuD that was predicted from
phage display, we generated a mutant, FhuD T181C, to which were conjugated
thiol-reactive probes capable of reporting changes in local environment. This
128
mutant was assessed for its binding of ligand and for its binding of TonB.
Rohrbach et al. (113) observed that addition of ferrichrome caused marked
quenching of FhuD’s intrinsic fluorescence and used this observation to quantify
binding of various ligands. We extended this feature to measure the ligand
binding capacity of FhuD T181C. Addition of Fcn to either FhuD or to FhuD
T181C caused substantial decreases in emission maxima. Binding curves were
generated by plotting the percentage decrease in fluorescence as a function of Fcn
added (Figure 3.5A). Fits of these data (Table 3.6) to a single binding site model
yielded similar apparent dissociation constants (KD app) of either 1.2 ± 0.2 M
(FhuD) or 0.6 ± 0.2 M (FhuD T181C), in agreement with the previously reported
(113) KD of 1 M. Our results establish that the T181C mutation does not
compromise ligand binding.
Taking advantage of the environmental sensitivity of the thiol-reactive
probes AEDANS and MDCC, each probe was conjugated to FhuD T181C.
Absorbance data (not shown) indicated that AEDANS labeled FhuD T181C at
approximately a 1:1 ratio, consistent with Cys 181 being solvent accessible and
reactive. MDCC labeled approximately 1 mole of conjugate for every 3 mole of
protein. This outcome likely resulted from labeling conditions. MDCC must be
dissolved in DMSO; addition of this labeling solution to FhuD resulted in slight
precipitation compared to the aqueous labeling conditions used with AEDANS, in
which no precipitation was observed.
129
Figure 3.5. Binding of Fcn to FhuD and to FhuD T181C. A. FhuD (●) and FhuD T181C (■) were titrated with the indicated amounts of Fcn; quenching of intrinsic fluorescence is plotted as a function of Fcn concentration; B. binding of Fcn to AEDANS-labeled FhuD T181C
(●) and to MDCC-labeled FhuD T181C (▼). Proteins were titrated with Fcn and quenching of
probe fluorescence was plotted as a function of Fcn added. Error bars in panels A and B represent
the standard deviation from three independent experiments. Lines through data indicate best fits to
a single binding site model as determined with Sigmaplot.
Upon addition of Fcn to AEDANS-labeled or to MDCC-labeled FhuD
T181C, extrinsic fluorescence was quenched. Titration of these conjugates with
Fcn is represented by the binding curves depicted in Figure 3.5B. Fits of these
data (Table 3.6) to a single binding site model yielded a similar KD app: 0.9 ± 0.2
M for AEDANS-labeled FhuD T181C; 0.31± 0.03 M for MDCC-labeled FhuD
T181C. These binding constants agree with those determined by intrinsic protein
fluorescence and thus validate the utility of our experimental approach.
130
Quenching of each probe’s fluorescence upon addition of Fcn therefore reported
occupancy of the FhuD siderophore binding site.
FhuD T181C labeled with either AEDANS or MDCC demonstrated
marked changes in fluorescence when mixed with TonB. Upon addition of TonB
to labeled conjugates, quenching was observed (Figure 3.6), demonstrating that
presence of TonB altered the probe’s environment. These changes were not
observed when labeled FhuD T181C was mixed with an excess of BSA (data not
shown). At low micromolar amounts of added TonB, saturation was achieved.
Fits of these data to a single binding site model generated values listed in Table
3.6. The KD app for the TonB–FhuD complex was 0.31 ± 0.05 M for AEDANS-
labeled FhuD T181C and 0.27 ± 0.04 M for MDCC-labeled FhuD T181C. Prior
saturation of the FhuD binding site with Fcn did not affect binding of TonB
(Figure 3.6) because similar dissociation constants (Table 3.6) were determined
despite occupancy.
a Given by the equation y = Bmax × [L]/(Kd+[L]), where [L] indicates ligand
concentration; reported uncertainties represent standard errors associated with best
fits to the single binding site model. b Fcn, ferricrocin.
Table 3.6. Summary of ligand binding parameters fit to a single site saturation ligand
binding modela
131
Figure 3.6. Binding of TonB to AEDANS-labeled FhuD T181C and to MDCC-labeled FhuD
T181C. TonB was added to a solution of labeled FhuD (in the absence or presence of Fcn) and
changes in extrinsic fluorescence were recorded. A. Response upon addition of TonB to either
FhuD T181C-AEDANS (●) or Fcn-bound FhuD T181C-AEDANS (▼); B. response upon
addition of TonB to either FhuD T181C-MDCC (●) or Fcn-bound FhuD T181C-MDCC (▼). Lines through data indicate best fits to a single binding site model as determined with Sigmaplot.
Results are representative of three experiments.
Given the spatial separation of putative interaction surfaces on TonB and
FhuD, we attempted fits of our fluorescence data to a model describing two
independent binding sites. When the data from titration of AEDANS-labeled
FhuD T181C with TonB were fit to this model, a larger standard error was
obtained and the resulting KD app values were not meaningful (data not shown).
Conversely, when the data from titration of MDCC-labeled FhuD T181C with
TonB were fit to this model, the standard error improved compared to the single
binding site model. However, the resulting KD app values obtained from this
132
procedure yielded large uncertainties confounding their interpretation. While we
cannot discount the possibility of two binding sites for TonB on FhuD, our results
favour a single binding site.
Alteration of labeling conditions affected conjugation efficiencies. By
labeling FhuD T181C overnight with AEDANS at 4 ˚C, two moles of label
incorporated for every mole of protein. FhuD contains a single endogenous Cys-
237; based upon the X-ray crystal structure, this residue is not expected to be
solvent-exposed. To determine if Cys-237 were reactive to conjugation, we
labeled FhuD using the conditions above; unexpectedly, FhuD was labeled. Cys-
237 localizes outside the phage display-identified TonB-binding surfaces. Given
the reactivity of Cys-237 toward conjugation, we used this conjugate to exclude
the possibility that TonB binds to regions other than those implicated by phage
display. Addition of TonB to FhuD-AEDANS caused an insignificant increase in
the fluorescence emission of AEDANS (data not shown). Given this outcome that
contrasts with FhuD T181C-AEDANS, we consider that Cys-237 is not part of a
TonB-binding environment. Our conclusion is that TonB-binding localizes to
regions on FhuD that were identified by phage display.
3.3.5 Detection of a TonB–FhuD complex by surface plasmon resonance
Our previous use of SPR technology quantified binding of FhuA to
immobilized TonB (77). Given the outcomes of this experimental design, we
adapted its use for the study of TonB–FhuD interactions. Dose-dependent
binding of FhuD to amine- and thiol-coupled TonB surfaces was observed
133
(Figure 3.7). Purified FhuA (0.1-1 M) also bound to immobilized TonB as a
positive control; FhuD binding (0.5 M) was unaltered in the presence of BSA
(0.5 M) as a competitor (data not shown). To improve kinetic analyses, a
homogeneous presentation of a lower density thiol-coupled Cys-TonB surface
was utilized. A high-affinity interaction (KD ~20 nM) between FhuD and TonB
was determined by SPR using a single binding site model. This interaction was
characterized by slow association (ka, ~2 x 104 M
-1s
-1) and slow dissociation (kd,
~4 x 10-4
s-1
) rates. Presence of the siderophore Fcn did not significantly alter the
affinity between FhuD and TonB by SPR (Figure 3.7B).
For reasons analogous to those considered with fluorescence data, we
attempted to fit our SPR data to a model describing multiple binding sites. SPR
data were fit to the heterogeneous ligand model, a model previously used to
distinguish independent binding sites on biological macromolecules (185,186).
Similar to fluorescence, fits of the data to this model quantified two binding sites;
a high affinity site with KD in the low nanomolar range and a weak affinity site
with KD in the low micromolar range. However, uncertainties associated with
these fits confounded their interpretation. Because of this outcome we favour a
single site of interaction between TonB and FhuD.
By having established formation of a TonB–FhuD complex, we performed
multi-component SPR analyses to identify a ternary FhuA–TonB–FhuD complex.
Initially, either a TonB–FhuA complex or a TonB–FhuD complex was formed by
injecting each analyte over amine-coupled TonB surfaces. Once these binary
complexes had formed, the secondary analyte was then injected.
134
Figure 3.7. Real-time kinetics of TonB–FhuD binding interaction detected by SPR. A.
Representative SPR sensogram for FhuD (top to bottom: 1000, 500, 250 and 100 nM) binding to
amine-coupled TonB (250 RU) in the absence of Fcn. B. Representative SPR sensogram for FhuD
(top to bottom: 1000, 500, 250 and 100 nM) binding to thiol-coupled Cys-TonB (48 RU) in the
absence (blue) or presence (red) of a tenfold molar excess of Fcn.
Injection of FhuD over a preformed TonB–FhuA complex indicated formation of
a ternary FhuA–TonB–FhuD complex (Figure 3.8A). Similarly, injection of
FhuA over a previously formed TonB–FhuD complex indicated formation of a
ternary FhuA–TonB–FhuD complex (Figure 3.8B). Ternary complexes formed
independent of the order of analyte addition. Qualitatively, association and
dissociation rates were also independent of the order of analyte addition and
mirrored those rates observed upon formation of each respective binary complex.
135
Figure 3.8. Multicomponent SPR analysis to detect ternary complex formation between FhuA–
TonB–FhuD. A. SPR sensogram indicating (I) baseline for buffer flowing over amine-coupled
TonB (250 RU); (II) increased signal change due to binding of FhuA (1 M); (III) stable FhuA–
TonB complex after a 0.5 M NaCl wash; (IV) increased signal change due to binding of FhuD (1
M); (V) return to baseline after regeneration. B. SPR sensogram indicating (I) baseline for
buffer flowing over amine-coupled TonB (250 RU); (II) increased signal change due to binding of
FhuD (1 M); (III) increased signal change due to binding of FhuA (1 M); (IV) return to baseline
after regeneration.
3.4 Discussion
Translocation of siderophores into the cytoplasm of Gram-negative
bacteria is partially understood, mainly with respect to the interplay between
TonB and TonB-dependent OM receptors. However, molecular events occurring
after translocation of siderophore into the periplasm remain largely unknown.
136
Binding of the translocated iron-bound siderophore to the periplasmic binding
protein may be determined purely by diffusion within the periplasm. However,
this mechanism poorly accounts for the weak affinity exhibited by FhuD toward
its ligands. Sprencel et al. determined (187) a value of approximately 4,000
copies of the ferric enterobactin periplasmic binding protein FepB in
E. coli, a low value compared to other periplasmic binding proteins such as those
involved in sugar or amino acid transport. Köster and Braun (174) demonstrated
that chromosomally-encoded FhuD was undetectable from silver-stained SDS-
PAGE gels of periplasmic extracts. Taken together with the modest KD of 1.0 M
(113), these data suggest that diffusion alone would be insufficient to account for
unidirectional siderophore transport. In addition to siderophore capture,
diffusion-governed docking of siderophore-bound FhuD to the CM permease may
be an inefficient process.
This study highlights novel interactions in the periplasm that are involved
in siderophore uptake by E. coli. By exploiting phage display and adopting three
biophysical strategies, we identified and mapped an interface between two
interacting protein partners. On FhuD, a TonB-binding surface was identified that
partially overlaps the FhuD siderophore binding site. On TonB, three distinct
regions of FhuD-binding surfaces were identified, two of which localized to
regions for which no structural data exists. Given apparent concordance between
phage display(180), and structural biology (96), the regions identified in this study
imply specific interactions between TonB and FhuD.
137
Interactions between TonB and FhuD are multidimensional. By DLS, we
identified a 1:1 TonB–FhuD complex. Fluorescence spectroscopy indicated an
apparent affinity for this complex to lie within the mid-nanomolar range. This
affinity contrasts with the low nanomolar range determined by SPR. Our current
SPR studies for FhuD–TonB are similar to the previous experimental design used
to monitor the TonB–FhuA kinetics in real-time (77). The affinity range
predicted for TonB–FhuD interactions by SPR (low nM) is consistent with the
previously reported range (94) of TonB–FhuA interactions also by SPR. The
differences in reported affinities between fluorescence and SPR may be attributed
to solution-phase versus immobilized systems or to buffer requirements of the
different technologies. Given these differences, our data indicates that the affinity
between TonB and FhuD lies somewhere in the low to mid-nanomolar range.
Significantly, the presence of Fcn did not alter the affinity between TonB
and FhuD; there was no evidence for competition. However, the TonB-binding
surface on FhuD as identified by phage display both overlaps the siderophore
binding site and extends beyond it; competition would not be expected.
Furthermore, there is no evidence to suggest large conformational changes in
FhuD upon binding siderophore (119). In the siderophore-bound state, FhuD
probably maintains a rigid backbone; this rigidity would not influence binding of
TonB. One implication of these findings is that interactions between siderophore
and FhuD are distinct from interactions between TonB and FhuD. It remains to
be determined whether siderophore can still bind to FhuD when complexed with
138
TonB, or whether bound TonB prevents siderophore binding through occlusion of
FhuD’s siderophore binding site.
A regulated mechanism of siderophore transport would involve
coordination of protein–protein interactions that would facilitate direct transfer of
siderophore among protein partners. For such a mechanism, siderophore transfer
would follow a sequence of directed exchanges from OM receptor, to periplasmic
binding protein, to CM permease. Directed transfer of this sort would require a
scaffold whereby protein–protein interactions drive spatial and temporal
localization of the periplasmic binding protein to regions involved in siderophore
translocation activities. The dynamic nature by which TonB cycles or changes its
conformation during energy transduction offers itself as an ideal candidate to
fulfill this role.
Our SPR data indicated formation of a ternary FhuA–TonB–FhuD
complex. This finding, taken together with the recent structural determinations of
TonB bound to FhuA (96) and to BtuB (97) underscores the possibility of such a
complex. It would position a periplasmic binding protein to accept siderophore
immediately after its translocation through a TonB-dependent OM receptor.
Examination of the TonB–FhuA interaction surface provides a means to evaluate
residues involved in the protein–protein interaction. In the TonB–FhuA crystal
structure, residues N-terminal to Arg-158 on TonB were not resolved and may not
participate in the TonB–FhuA interface. The FhuD-interacting residues on TonB
that were predicted by phage display (region III, Figure 3.3C) localize
139
immediately N-terminal to TonB Arg-158. We infer that these residues are poised
to bind FhuD in a FhuA–TonB–FhuD ternary complex.
Given these experimental outcomes, we propose a structural model for a
FhuA–TonB–FhuD complex. Examination of complementary binding regions on
solvent-accessible surfaces for known TonB and FhuD structures reveals no
obvious means by which the two proteins can interact. However, one can
rationally position FhuD to the TonB–FhuA structure using these surfaces as
docking constraints. Figure 3.9 depicts a model for a ternary FhuA–TonB–FhuD
complex that is based on our experimental evidence. In this model, FhuD-binding
regions on TonB are coloured according to the convention adopted in Figure 3.4.
TonB residues 153
PRALS157
corresponding to region III (Figure 3.3C) were
computationally modeled at its N-terminus. Their position illustrates that
interaction with FhuD at this region would result in the apposition of its
siderophore binding site with the FhuA lumen such that it would intersect with the
trajectory of the translocated siderophore. Furthermore, TonB residues interacting
with FhuA in the TonB–FhuA crystal structure are distal to those TonB residues
which contact FhuD. Separation of these binding surfaces may therefore
coordinate transduction of energy to FhuA with directed localization of FhuD
beneath the FhuA lumen.
The precise sequence of these events has yet to be established is. Our data
cannot distinguish whether FhuD remains bound to TonB during the energy
transduction cycle or if binding is a transient event. One possibility is that FhuD
binds and dissociates as a function of the TonB energy transduction cycle,
140
Figure 3.9. Model of a FhuA-TonB-FhuD ternary complex. Stereo image depicting a possible
ternary complex between FhuA, TonB and FhuD. FhuD (PDB code 1EFD) was manually docked
under the TonB-FhuA crystal structure (PDB code 2GRX) using phage display-identified protein–
protein interaction surfaces as docking constraints. Complementary phage display-identified
surfaces are coloured blue on both TonB (yellow, surface representation) and FhuD (salmon,
ribbon representation). The orientation localizes the FhuD siderophore binding site beneath the
lumen of FhuA (green, ribbon representation). For clarity, a molecular surface is projected on
TonB.
perhaps resulting from the β-strand exchange that TonB is known to undergo
(72,74,75,96,97). Prolonged association of FhuD with TonB seems unlikely as
FhuD must ultimately deliver siderophore to FhuB/C. It is intriguing that the
most extreme N-terminal FhuD-binding surface on TonB is at a region close to
probable contact sites with ExbB/ExbD. Such placement may be a means by
which TonB directs transfer of FhuD from the OM to CM as TonB disengages
141
FhuA. This mechanism would ensure that FhuD samples a space proximal to the
CM, thereby enhancing the probability of encountering FhuB.
Previous reports have indicated (114,188) that interactions between the
FhuD homolog BtuF and its cognate transporter BtuC/D are virtually irreversible.
This observation remains to be clarified in light of our experimental evidence,
which for the first time identifies TonB as a binding partner for periplasmic
binding proteins. We advocate a mechanism favouring transient associations of
periplasmic siderophore-binding proteins with TonB and with their cognate CM
transporters at discrete steps during the siderophore transport cycle.
We previously used phage display in concert with biophysical methods to
map unambiguously the network of protein–protein interactions that occur at the
TonB–FhuA interface. These outcomes were recently confirmed by the X-ray
structure of the TonB–FhuA complex. This strategy is now extended to identify
interfaces involved in TonB–FhuD interactions. Such interactions may serve to
coordinate spatial and temporal localization of periplasmic binding proteins to
environments involved in siderophore uptake. Given the diversity of components
among different siderophore transport systems, we propose that in addition to its
role as energy transducer, TonB acts as a unifying element, a scaffold to regulate
the unidirectional flow of iron-bound siderophore from the OM to the CM.
3.5 Acknowledgements
Research was supported by operating grant MOP-62774 to J.W.C. from
the Canadian Institutes of Health Research (CIHR). D.M.C. was recipient of the
142
McGill Graduate Studies Fellowship and an F.C. Harrison Fellowship,
Department of Microbiology and Immunology. Natural Sciences and Engineering
Research Council (Canada) provided an Undergraduate Student Research Award
to I.R.M. and a Post Graduate Scholarship to J.-N.G. É.M. was trainee from the
École supérieure de biotechnologie Strasbourg. A.C. won a Traveling Award for
Research Training from the National Health and Medical Research Council
(Australia). Canada Foundation for Innovation awarded infrastructure to the
Montreal Integrated Genomics Group for Research on Infectious Pathogens.
Sheldon Biotechnology Centre at McGill University is supported by multi-user
maintenance grants from CIHR. We appreciate contributions of experimental
materials by M. Damlaj, J. Deme, J. Gilbert, K. James and C. Ng; and editorial
support by J.A. Kashul.
143
Preface to chapter 4
In chapter 3, we demonstrated that TonB and FhuD form a complex.
Regions within TonB that are essential for this interaction are examined in chapter
4. We characterized the properties of periplasmic TonB derivatives and asked
how these properties affect binding to FhuD. By SPR, we determined that neither
TonB’s N-terminal region nor its proline-rich region is required for interaction
with FhuD. These findings suggested that TonB’s central and C-terminal regions
were essential for interaction with FhuD. We then computationally modeled how
such regions might bind to FhuD. Our model suggested that when bound to
TonB, FhuD’s siderophore binding site could project towards the OM, where its
orientation could facilitate siderophore capture as it emerges during transport.
144
Chapter 4
C-terminal region of TonB positions periplasmic binding protein
FhuD for siderophore transport in Escherichia coli
David M. Carter, Justin C. Deme, Mark A. Hancock and James W. Coulton
Manuscript submitted July, 2009
145
4.0 Summary
The Ferric hydroxamate uptake (Fhu) system of Escherichia coli actively
transports ferric hydroxamate-type siderophores. The main components of the
Fhu system include outer membrane receptor FhuA, periplasmic binding protein
FhuD, cytoplasmic membrane-embedded permease FhuB/C, and the energy
transducing TonB–ExbB–ExbD complex. Recently, we demonstrated that TonB
binds FhuD. Here we extend these analyses by delineating regions within TonB
that are essential for this interaction. By analytical ultracentrifugation and
fluorescence spectroscopy, we characterized properties of three periplasmic TonB
variants: a derivative possessing residues 33–239; a derivative with a deletion
(residues 66–100) of the central proline-rich region; and a truncated derivative
possessing residues 103–239. Our analyses indicate that all derivatives are
elongated monomers, consistent with knowledge that TonB possesses a structured
C-terminal domain and unstructured central and N-terminal regions. By surface
plasmon resonance, all TonB derivatives exhibit similar low nanomolar binding
affinities for FhuD. Hence, essential FhuD-binding determinants localize within
predominantly unstructured regions of TonB. To further characterize the nature
of this interaction, we computationally docked onto the surface of FhuD, an
oligopeptide that corresponds to an unstructured region of TonB and that was
predicted to bind FhuD. These simulations place the peptide against FhuD at
predicted TonB-binding surfaces. Extension of N- and C-termini of the lowest
energy docked peptide indicates that when bound to TonB, FhuD’s siderophore
binding site orients towards the outer membrane.
146
4.1 Introduction
Most bacteria require the essential nutrient iron in order to grow and
divide. However, the bioavailable concentration of iron is bacteriostatic due to its
propensity to form insoluble ferric hydroxides in the environment and due to
sequestration of labile iron pools by host iron-binding proteins (6,7). Given these
limitations, bacteria have evolved high affinity iron uptake systems whereby they
secrete siderophores capable of scavenging iron under iron-deplete conditions.
Having bound iron with high affinity, ferric-siderophores are transported back
into a bacterium by various families of siderophore transport systems (145). The
ferric hydroxamate uptake (Fhu) system of Escherichia coli is responsible for
uptake of hydroxamate siderophores including ferricrocin. The main components
include outer membrane (OM) siderophore receptor FhuA, periplasmic binding
protein (PBP) FhuD, and cytoplasmic membrane (CM)-embedded ATP permease
FhuB/C.
The structure of FhuA (42) revealed a two-domain protein comprising an
N-terminal cork domain that independently folds and inserts into a C-terminal β-
barrel domain. Passive transport of siderophore through FhuA is prevented by
occlusion of its lumen with the cork domain. Transport requires energy-
dependent rearrangement of FhuA’s cork domain so that a pore large enough to
allow translocation can form. The CM-associated TonB–ExbB–ExbD multi-
protein complex provides the energy input required to facilitate this structural
rearrangement. By harnessing the proton motive force, the TonB–ExbB–ExbD
147
complex conformationally activates TonB in a manner that promotes energy
transduction to siderophore-bound FhuA.
Structural information is incomplete for the TonB–ExbB–ExbD complex.
TonB is a modular 239-residue protein embedded within the CM by a 33-residue,
N-terminal transmembrane helix. It interacts with ExbB and ExbD through this
single helix (85) in a TonB:ExbB:ExbD stoichiometry postulated to be 1:7:2. (82)
The remainder of TonB is periplasmic and mostly unstructured. Central to the
periplasmic domain is a proline-rich region consisting of Lys-Pro, Glu-Pro
repeats, considered to adopt an extended structure based on an NMR study of the
isolated residues (70). This may serve to span the periplasm; however, deletion of
this region only marginally reduced TonB-dependent transport (71). The C-
terminal 89 residues of TonB form a structured domain responsible for energy-
independent interactions with TonB-dependent OM receptors (125).
Structural studies demonstrated that TonB’s C-terminal domain is
dynamic. The first crystal structure of a TonB derivative, residues 155–239,
revealed a tightly intertwined dimer (72). Analytical ultracentrifugation (AUC)
studies (73,77) and a second crystallographic study
(73) later confirmed this
finding. A slightly longer TonB derivative possessing residues 148–239 also
crystallized as a dimer (74), albeit of loosely associated monomers.
Unexpectedly, this study also demonstrated that the same derivative was
monomeric by AUC. A longer derivative of TonB possessing residues 103–239
was demonstrated to be monomeric by NMR spectroscopy (75). Most recently,
another structure of a monomeric TonB variant from the fish pathogen Vibrio
148
anguillarum was solved by NMR (76). From these studies it is apparent that
derivatives of TonB possessing at least residues 148–239 prevent dimerization in
solution, probably by altering distal conformations of a C-terminal hinge region
centered on residues 231
KING234
of E. coli TonB.
In addition to binding TonB-dependent receptors, TonB also binds the
PBP FhuD (189) and probably binds other PBPs involved in TonB-dependent
transport. By phage display technology, we identified complementary regions
between TonB and FhuD that were predicted to interact and we then demonstrated
complex formation in vitro. Here we extend these results by delineating regions
within TonB that are essential to forming a complex with FhuD. Through a series
of periplasmic TonB deletions, we investigated the importance of each region
predicted to bind FhuD. First, structural properties of each TonB derivative were
investigated to ascertain whether the length of TonB influenced its overall shape
and tertiary structure and whether these differences could affect binding of FhuD.
Second, SPR was used to examine the affinity between each TonB derivative and
FhuD. Finally, a computational model was generated to predict the way in which
a given TonB-derived, FhuD-binding peptide might bind to FhuD. Our results
indicate that unstructured regions on TonB, proximal to the structured C-terminal
domain, constitute the minimal elements required to bind FhuD and further
suggest that these regions bind FhuD in a manner that orients FhuD’s siderophore
binding site towards the OM.
149
4.2 Materials and methods
4.2.1 Bacterial strains and plasmids
E. coli strains BL21 DE3 (pLysS) and ER2566 were used for protein
expression. Plasmids pWA01 (77), pWA02
(94) and pDMC01 (this study) were
used to express TonBs 33–239, Δ66–100 and 103–239 respectively. Plasmid
pCMK02 (94) was used to express a TonB Δ66–100 variant possessing an N-
terminal cysteine for SPR analysis. Plasmid pMR21, a gift from W Köster
(VIDO, SK), was used to express a decahistidine-tagged version of FhuD.
Plasmid pMal-pIII was used to express maltose binding protein.
4.2.2 Cloning of TonB 103–239
A periplasmic region of TonB corresponding to residues 103–239 was
sub-cloned from pWA01, which contains residues 33–239. Primers
corresponding to the appropriate region were synthesized (Alpha DNA, Montreal)
and the region PCR-amplified. The PCR product was ligated into digested pJD01
(unpublished data, a mutated version of pMR21 with a TEV protease-cleavable
His-tag) to generate pDMC01, a decahistidine-tagged product of TonB 103–239
plus a TEV protease cleavage site.
4.2.3 Protein expression
Bacterial cultures harbouring expression plasmids were grown at 37°C
overnight in LB broth (Fisher Scientific) supplemented with appropriate
antibiotics. Cultures were then diluted 100-fold into 6 × 1 l each of fresh LB
150
broth plus antibiotics and grown at 37°C until mid-log phase. Protein expression
was induced by addition of IPTG (BioVectra, Charlottetown, PE) to final
concentrations of either 0.5 mM (TonB 33–239 and Δ66–100), 0.4 mM (TonB
103–239) or 1 mM (FhuD). Upon induction, cells were grown for either four
hours at 30°C (TonB 33–239 and Δ66–100), one hour at 37°C (TonB 103–239) or
overnight at 37°C (FhuD). Cells were then harvested by centrifugation of 3 l each
from the 6 l culture and the resulting pellets were stored at -20°C until further use.
Maltose binding protein was expressed as previously described (180).
4.2.4 Protein purifications
TonBs were purified by first thawing a frozen cell pellet in 40 ml of buffer
Hni (50 mM Hepes pH 7.5, 30 mM NaCl and 5 mM imidazole (Fluka)). Once
thawed, the cell paste was supplemented with RNAse, DNAse, MgCl2, lysozyme,
PMSF, and a Complete EDTA-free protease inhibitor tablet (Roche). Cells were
then lysed by two passages through an Emulsiflex (Avestin) and clarified by
centrifugation at 27,000 × g for 1 h at 4°C. The supernatant was further clarified
by passage through a 0.45-µm filter and loaded onto a column containing 25 ml of
Ni-NTA superflow (Qiagen) equilibrated in buffer Hni. After a 500 mM salt
wash, contaminants were removed with a step to 50 mM imidazole (Fluka).
TonBs were then eluted with a linear gradient up to 500 mM imidazole over four
column volumes (CV). The Ni-NTA eluate was pooled, supplemented with 3
mM EDTA plus an additional 175 µM PMSF and applied to a 5 ml Source S
strong cation exchange column (GE Health Sciences) equilibrated in buffer Hn
151
(50 mM Hepes pH 7.5 plus 30 mM NaCl). After a wash, TonBs were eluted with
a gradient up to 200 mM NaCl over 10 CV followed by a 1 CV gradient up to 500
mM NaCl and an additional 1 CV gradient up to 1 M NaCl. FhuD was purified as
described previously (189) and maltose binding protein was purified as described
previously (180).
4.2.5 Analytical ultracentrifugation
Prior to AUC data collection, TonBs were dialyzed overnight against a
buffer containing 50 mM Tris pH 7.5 and 150 mM NaCl. Sedimentation velocity
experiments were conducted on a Beckman XL-I analytical ultracentrifuge (77).
Data were fit to a c(s) model using the program Sedfit (143).
4.2.6 Fluorescence spectroscopy
Prior to fluorescence data collection, TonBs were dialyzed overnight
against the same buffer used in AUC experiments. TonBs were then diluted to
3.5 µM in dialysis buffer and centrifuged for 10 min at 4°C. Spectra were
recorded on a Varian Cary Eclipse fluorimeter with the following settings: λex =
280 nm, excitation slit = 2.5 nm, λem = 300–400 nm, emission slit = 5 nm,
averaging time = 0.1 sec, scan rate = 50 nm/min and a detector voltage of 700 V.
A total of 10 scans were collected at 25°C and averaged. Spectra were corrected
by subtraction of background fluorescence due to buffer alone. Protein
concentrations were normalized by a Bradford assay (Bio-Rad) and by
measurement of absorbance at 280 nm on a Varian Cary Bio 1 spectrophotometer.
152
4.2.7 Surface plasmon resonance
Binding interactions between FhuD (~32 kDa) and TonB (~25 kDa TonB
33–239; ~21 kDa TonB Δ66–100; ~18 kDa TonB 103–239) were examined using
label-free, real-time Biacore 3000 instrumentation (GE Healthcare Bio-Sciences
AB). Prior to all SPR experiments, purified protein preparations were dialyzed
against 50 mM Hepes pH 7.4 containing 150 mM NaCl. Experiments were
performed on research-grade CM4 sensor chips at 25°C using filtered (0.2 µm)
and degassed HBS-ET running buffer (50 mM HEPES pH 7.4, 150 mM NaCl, 3
mM EDTA, 0.05% (v/v) Tween 20). Protein-grade detergents (Tween 20,
TritonX-100, Empigen) were from Anatrace; all other chemicals were reagent-
grade quality. As previously described (189), TonB derivatives (3 μg/ml in 10
mM sodium acetate pH 5.5) and corresponding reference surfaces were
immobilized using Biacore Amine or Thiol Coupling Kits as recommended by the
manufacturer. To assess binding specificity and kinetics, PBPs (0–500 nM FhuD
and negative MBP control) were titrated over immobilized surfaces at 30 μl/min
using 'KINJECT' mode (1 min association +/- 15 min dissociation). For binding
assays performed in the presence of siderophore, a 10-fold molar excess of
ferricrocin was added to FhuD and incubated for at least 30 min prior to injection.
In all cases, surfaces were regenerated between titration series at 50 μl/min using
two 30 s pulses of solution I (HBS-ET containing 0.5 M NaCl, 5 mM NaOH,
0.05% (v/v) Triton-X100) and solution II (HBS-ET containing 0.5 M NaCl, 5 mM
NaOH, 0.05% (v/v) Empigen), followed by 'EXTRACLEAN' and 'RINSE'
procedures. SPR data are representative of duplicate injections acquired from
153
three independent trials. Mass transport-independent data were double-referenced
(183) and analyzed according to a “1:1 Titration” model in the BIAevaluation
v4.1 software (190).
4.2.8 Computational docking
Oligopeptides corresponding to TonB residues 123
SPFENTAPARL133
(predicted FhuD-binding region II) and 153
PRALSRNQ159
(predicted FhuD-
binding region III) were used as flexible ligands and were docked to the crystal
structure of FhuD (PDB code 1EFD) (111) with the program AutoDock 4 (191).
The TonB region II peptide was modeled as an α-helix with the program Pymol
v1.1r1 (170). Two glycines were attached to each terminus to act as inert spacers
that should not influence binding specificity during simulation. Prior to docking,
the α-helical peptide was minimized using an AMBER force field implemented in
the program UCSF Chimera v1.3 (192). For docking, the peptide was given 12
torsional degrees of freedom along selected sidechains. Additionally, the
backbone was allowed some torsional freedom to relax during the course of
simulation. The TonB region III ligand was derived from crystal structure PDB
code 1U07 (74). Residues comprising region III were truncated from the PDB
coordinates, capped with glycines and minimized as with the region II peptide.
Selected sidechains and backbone atoms were also given torsional freedom to
relax during the course of simulation. Two separate simulations consisting of 50
runs each were performed. A Lamarckian genetic algorithm (171) was used in
each run to dock the peptides; 25,000,000 energy evaluations were performed
154
with a population size of 300 over 29,000 generations, a mutation rate of 0.02,
and a crossover rate of 0.8.
4.3 Results
4.3.1 TonB derivatives are elongated monomers with similar elements of tertiary
structure
Previously, we demonstrated that a periplasmic derivative of TonB
interacts with FhuD to form a 1:1 siderophore-independent complex (189). To
extend these results, we investigated whether truncations of TonB could affect its
FhuD-binding properties. Three periplasmic TonB variants were characterized
(Figure 4.1): a derivative possessing residues 33–239 (denoted hereafter as TonB
33–239); a derivative with an internal deletion of TonB’s proline-rich region
(TonB Δ66–100); and a third derivative possessing residues 103–239 (TonB 103–
239). Both TonB 33–239 and TonB Δ66–100 retained the three predicted FhuD-
binding regions (regions I–III respectively, Figure 4.1B and C), whereas TonB
103–239 retained the second and third FhuD-binding regions, respectively
(regions II and III, Figure 4.1D). All derivatives were isolated to near
homogeneity (at least 95% purity) as assessed by SDS-PAGE. Similarly, FhuD
was isolated to near homogeneity.
To understand the influence of hydrodynamic shape on formation of a
TonB–FhuD complex, we characterized the frictional ratios of the various TonB
derivatives by sedimentation velocity. Our earlier studies (77,94) determined the
frictional ratios of TonB 33–239 and TonB Δ66–100. Both values (f/fo = 2.39 for
155
TonB 33–239; f/fo = 2.27 for TonB Δ66–100) indicated that these TonB
derivatives are elongated, consistent with knowledge that TonB must span the
periplasm for energy transduction. In this study, we similarly modeled frictional
ratios of each derivative using a continuous distribution c(s) model in Sedfit
(143). In agreement with our previous analyses, TonB 33–239 is elongated and
possesses a frictional ratio of 2.36; RMSD = 0.0041. Similarly, TonB Δ66–100 is
also elongated and possesses a frictional ratio of 2.02; RMSD 0.0046. The
shortest TonB 103–239 exhibits a smaller frictional ratio of 1.80; RMSD =
0.0044, indicating that it is considerably less elongated than the other derivatives,
but still deviates from values typically associated with globular proteins (f/fo
~1.3).
Figure 4.1. Schematic representations of TonB derivatives from this study. A. Full-length TonB.
TM: transmembrane domain, PRR: proline-rich region. Predicted FhuD-binding regions I–III are
indicated in Roman numerals. TonB amino acid numbering is indicated above the schematic. N-
and C-termini are labeled. B. TonB 33–239. Schematic is labeled according to panel A. N-
terminal histidine tag is indicated. C. TonB Δ66–100. Schematic is labeled as in panel B. D.
TonB 103–239. Schematic is labeled as in panel B.
156
Stoichiometries of TonB derivatives were also determined. Our previous
analyses demonstrated that both TonB 33–239 and TonB Δ66–100 sediment as
monomers in solution. Our current analyses agree with these previous reports and
indicate that both TonB 33–239 and TonB Δ66–100 are monomeric in solution.
Consistent with the NMR structure (75) of a similar derivative, TonB 103–239 is
also monomeric in solution. Our data indicate a hydrated molecular weight of
19.0 kDa, in agreement with the theoretical molecular weight (18.4 kDa) of
decahistidine-tagged TonB 103–239 monomer.
To determine whether the derivatives possessed similar elements of
tertiary structure, we used intrinsic fluorescence spectroscopy to examine the
environments surrounding fluorophores in each TonB derivative. Of the ten
aromatic residues within TonB’s primary sequence, six localize within the
structured C-terminal domain. If each derivative were to adopt a similar fold,
fluorescence spectra should be identical. When prepared at equimolar
concentrations, all TonB derivatives possess similar spectra, with overlapping
maxima centered at 335 nm (data not shown). This indicates that all TonB
derivatives possess similar folds within their C-termini. Together with AUC data,
which demonstrates that all TonB derivatives are elongated monomers, we
interpret our fluorescence results to mean that all periplasmic derivatives of
monomeric TonB have unstructured N-terminal regions and a structured C-
terminal region.
157
4.3.2 TonB derivatives bind FhuD with equal affinities
Using the single-cycle kinetics approach (190), binding interactions
between immobilized TonB derivatives and FhuD were examined with label-free,
real-time SPR. FhuD exhibits specific, concentration-dependent binding to all
TonB surfaces (Figure 4.2). As a negative control, there was little or no non-
specific binding to any of the surfaces when another PBP, maltose binding protein
(MBP), was titrated under identical assay conditions. FhuD binding to all
derivatives is unaltered in the presence of siderophore (data not shown) and, when
analyzed according to a “1:1 Titration” model in the BIAevaluation software,
similar kinetic (slow on- and slow off-rates) and affinity (low nanomolar)
constants are observed (Table 4.1). Accounting for differences in the molecular
weight, coupling chemistry, and surface density of each derivative, the observed
binding responses (RU and subsequent affinity calculations) indicate that all three
TonB derivatives bind FhuD in a similar manner.
Figure 4.2. Single cycle kinetic analysis of FhuD binding to TonB derivatives using label-free,
real-time SPR. FhuD was titrated (0–500 nM; 30 µl/min × 1 min association ± 15 min
dissociation) over immobilized surfaces: red, 180 RU amine-coupled TonB 33–239; cyan, 100 RU
thiol-coupled TonB Δ66–100; yellow, 150 RU amine-coupled TonB 103–239. For all TonB
derivatives, MBP (green, 0–500 nM) exhibited little or no non-specific binding. Coloured lines
represent experimental data and black lines represent best fit to “1:1 titration” model.
158
Table 4.1. Kinetics and affinity of TonB–FhuD interactions according to “1:1 titration”
model
†Reported errors for apparent rate constants represent the standard deviation from three
independent SPR trials.
4.3.3 Computational models predict the orientation of FhuD when bound to TonB
For the predicted FhuD-binding regions on TonB, our SPR results
demonstrated that region I was not essential to form a TonB–FhuD complex.
Since the smallest TonB regions that bind to FhuD were likely to be regions II and
III, we computationally modeled how the peptides might bind to FhuD. Two
approaches were employed to test this hypothesis. First, we took advantage of
currently available TonB structural models; regions N-terminal to residue 150 are
considerably flexible and unstructured. Given this knowledge, we modeled the
core of TonB region II, consisting of residues 123
SPFENTAPARL133
, as an α-
helical oligopeptide. After energy minimization, the TonB oligopeptide was
docked to the crystal structure of FhuD using PDB code 1EFD (111) and the
program AutoDock 4 (191). During simulation, the peptide was given torsional
degrees of freedom and allowed to sample space that was limited to the side of
FhuD that our phage display analysis predicted it would bind.
Two separate simulations gave low energy solutions centering the TonB
peptide on the predicted FhuD-binding surface (Figure 4.3A). Results focus on
the two lowest energy solutions observed in each of the two simulations. Both
solutions predict that the peptide will spontaneously bind, with energies ranging
TonB derivative ka × 103 (M
-1 s
-1)† kd × 10
-5 (s
-1)† KD × 10
-9 (M)
†
TonB 33–239 7.1 ± 1.2 5.0 ± 3.1 7.0 ± 1.8
TonB Δ66–100 3.5 ± 0.3 2.6 ± 1.2 7.5 ± 1.1
TonB 103–239 4.2 ± 0.8 5.4 ± 1.7 13 ± 5.9
159
from -2.39 kcal/mol to -1.06 kcal/mol and -2.53 kcal/mol to -1.35 kcal/mol for the
first and second simulation respectively. These values are consistent with
magnitudes of energies involved in salt bridge formation or in hydrogen bond
formation.
While orientations vary for low energy docked peptides, N-termini from
the four lowest energy region II conformers center within 7 Å of FhuD residue
Leu-50. This falls within a cavity on FhuD that exhibits elements of net
electronegativity as well as nonpolarity (Figure 4.3B). The cavity is also
proximal to the first FhuD region that our previous phage display analysis
predicted would serve as a TonB-binding surface. Significantly, lowest energy
TonB region II peptides orient so as to promote biologically constructive
interactions with FhuD. TonB is anchored to the CM by its N-terminal
transmembrane domain, while its C-terminus projects toward the OM. By
extending the peptide backbone trajectory of the C-termini of lowest energy
docked region II peptides, TonB’s C-terminus could project towards the vicinity
of FhuD’s siderophore binding site. Similarly, extension of the N-termini of
lowest energy docked peptides could project TonB’s N-terminus towards the CM.
These results support a model where upon binding to TonB, FhuD’s siderophore
binding site orients towards the OM.
The second docking approach used structural data corresponding to the
most complete structure of TonB (PDB code 1UO7) (74). Three-dimensional
coordinates that correspond to TonB region III (residues 153
PRALSRNQ159
) were
160
Figure 4.3. TonB region II peptide docked to the surface of FhuD. A. Lowest energy TonB-
derived oligopeptide (binding energy -2.53 kcal/mol) bound to the surface of FhuD (blue). The
phage display-predicted TonB-binding surface is coloured red. The peptide N-terminus is marked
with a white arrow; B. electrostatic surface representation of FhuD, contoured from -2 kT (red) to
+2 kT (blue). The predicted TonB-binding surface is outlined in black; C. temperature factor
distribution of FhuD. Temperature factors derived from FhuD PDB file 1EFD are mapped onto
the surface of FhuD. Colours are contoured such that dark blue represents regions of low
uncertainty (B-factors ~13), green represents regions of intermediate uncertainty (B-factors ~23)
and red represents regions of greatest uncertainty (B-factors ~41). The predicted TonB-binding
surface is outlined in black.
161
excised from the crystal structure model and used as a ligand that we then docked
to FhuD. However, docking solutions with this peptide produced only positive
binding energies (data not shown).
4.4 Discussion
Our previous phage display and biophysical analyses demonstrated that, in
addition to binding TonB-dependent OM receptors such as FhuA, TonB also
binds PBPs such as FhuD. Using three periplasmic derivatives, we now identify
TonB regions that are important for binding FhuD. Our hydrodynamic analyses
of the TonB derivatives indicate that all are monomeric and elongated in solution.
These data are consistent with the NMR structure of TonB 103–239, which
revealed a monomer with a structured C-terminal domain plus an unstructured and
presumably extended N-terminal region. A shorter TonB derivative possessing
residues 148–239 was also monomeric in solution (73), yet crystallized as a dimer.
In contrast, shorter derivatives with residues 155–239 and fewer dimerize
in solution (77). Given their propensity to form dimers, we did not attempt to
characterize binding of FhuD to shorter TonB derivatives. We reasoned that
removal of predicted FhuD-binding regions II and III could promote TonB
dimerization. If such a derivative did not bind FhuD, it might be due to
dimerization and not due to loss of predicted FhuD-binding regions. Furthermore,
we did not attempt to model the hydrodynamic properties of TonB–FhuD
complexes. Our c(s) analyses measured global weight-averaged frictional ratios,
162
which would necessitate further deconvolution in order to gain insight into the
frictional ratio of the TonB–FhuD complex.
Our fluorescence results indicate that tertiary structural features are
superimposable amongst the derivatives. These elements probably correspond to
the structured region observed in the TonB 103–239 NMR structure.
Accordingly, the spectra of derivatives examined here probably arise due to
positioning of TonB fluorophores accounted for by that structure. In contrast,
TonB derivatives that dimerize in solution render their fluorophores in different
environments and yield detectable changes in the intrinsic fluorescence spectra
(100,193).
In agreement with our previous multi-cycle analyses in which TonB 33–
239 was both amine- and thiol-coupled for SPR (189), our single-cycle approach
(Figure 4.2) further demonstrates that FhuD interacts with amine-coupled TonB
33–239 with low nanomolar affinity (KD ~7 nM; Table 4.1). Our present study
also confirms that TonB 33–239 and FhuD interactions are consistent with simple
1:1 kinetics (i.e. single-cycle curve fitting with “1:1 titration” model (190); and
that binding is unaltered in the presence of siderophore. Two additional
derivatives were immobilized (TonB Δ66–100, TonB 103–239; thiol- or amine-
coupled based upon construct designs) to further delineate key regions in TonB
that mediate binding to FhuD. TonB Δ66–100 binds FhuD in a specific,
concentration-dependent manner (KD ~7.5 nM), like TonB 33–239. This result
suggests that presence of the proline-rich region in TonB (i.e. spacing between
predicted FhuD-binding regions I and II-III) is not essential for binding to FhuD.
163
TonB 103–239, a derivative that still retains predicted FhuD-binding regions II
and III, also binds FhuD with low nanomolar affinity (KD ~13 nM), like TonB
33–239. This result suggests that TonB region I is not essential for complex
formation with FhuD. Within error, our SPR analyses indicate that all TonB
derivatives bind FhuD with similar affinity and in a siderophore-independent
manner. Overall, these in vitro outcomes demonstrate that residues 33–102 of
periplasmic TonB are not an absolute requirement for TonB–FhuD complex
formation.
All derivatives characterized here are elongated monomers with similar
structural features: an unstructured N-terminal region and a structured C-terminal
region. The precise location where there is crossover into structure depends on
the total length of the TonB derivative. Derivatives possessing residues 148–239
and longer exhibit C-terminal structure anywhere between residues 150–158.
This implies that there are structural differences between the predicted FhuD-
binding regions II and III. Region II is most likely disordered, given its location
within the TonB sequence. Region III is capable of adopting secondary structural
features similar to β-strand-like conformations as observed in the crystal structure
of TonB residues 148–239, the NMR structure of residues 103–239 and of TonB
bound to the transporters FhuA (96) and BtuB (97). However, region III is
disordered in smaller TonB derivatives possessing residues 155–239, indicating
that the conformation of region III depends on the presence of longer N-terminal
stretches.
164
Given the flexibility of TonB, we modeled TonB-derived regions II and III
as oligopeptides with particular conformations and docked them to the more rigid
FhuD. This approach yields reasonable docking solutions for region II peptides
bound to FhuD and provides insight into the nature of TonB–FhuD interactions.
The FhuD surface to which the TonB peptide docked exhibits distinctive features.
The predicted TonB-binding surface delineates a boundary that encompasses a
landscape of varying polarity and charge (Figure 4.3B) and indicates that there are
defined clusters of TonB-binding determinants on FhuD.
The predicted TonB-binding surface also exhibits some degree of
flexibility. A region of uncertainty in the FhuD crystal structure and described by
the model’s temperature factors falls within boundaries of the predicted TonB-
binding surface (Figure 4.3C). This observation is intriguing for two reasons.
First, flexible and unstructured regions of proteins often mediate protein–protein
interactions by allowing degrees of conformational sampling necessary to
promote binding (194). Second, flexibility seems to be structurally conserved,
given that all structures of ligand-bound FhuD as well as vitamin B12-bound BtuF
are associated with high thermal parameters within this region (data not shown).
Flexibility of TonB-dependent PBPs was recently demonstrated by molecular
dynamics studies: both FhuD (122) and BtuF (121,122) undergo considerable
breathing motions, not accounted for in the crystal structures. Taken together,
these findings suggest that flexibility is also a binding determinant between TonB
and FhuD.
165
Noteworthy is that the predicted TonB-binding surface of FhuD is
conserved amongst various pathogenic Gram-negative bacteria. Figure 4.4A
illustrates a sequence alignment of FhuD proteins from selected Gram-negative
bacteria. Conserved regions within this alignment are mapped onto a surface
representation of FhuD in Figure 4.4B. Conservation is visualized on the FhuD
surface through a color contour: red indicates highly conserved residues, green
indicates moderately conserved residues, and dark blue indicates non-conserved
residues. A cluster of highly conserved and moderately conserved residues falls
within the boundary delineated by the predicted TonB-binding surface.
As discussed above, this region coincides within a region of greater
flexibility and also coincides within a region of defined electrostatic and polar
properties. Taken together, these findings provide evidence that predictions borne
from phage display analyses reveal a surface on FhuD with emergent properties
that are characteristic of a protein binding site.
Results from our docking simulations reveal plausible modes for TonB
peptides that bind to FhuD. The fact that many solutions yield negative binding
energies demonstrates that favourable interactions are capable of forming even
when considering a static FhuD surface. Greater flexibility of this surface might
promote stronger interactions that were not modeled in our simulations.
Significantly, our docking solutions orient the region II oligopeptide such that the
siderophore binding site of FhuD can orient towards the OM when bound to
TonB.
166
Figure 4.4. Sequence conservation of FhuD from various pathogenic Gram-negative bacteria. A.
Clustal W sequence alignment of FhuDs. Bacterial species are indicated to the left of each row in
the alignment. White letters boxed in black indicate strictly conserved residues amongst all
species. White letters boxed in dark grey indicate strongly conserved residues. Black letters
boxed in light grey indicate modestly conserved residues; B. FhuD sequence conservation mapped
to the surface of FhuD. Scores from the alignment in panel A are mapped to the surface of FhuD
and are coloured as follows: Dark blue indicates less than 20% conservation of given residues.
Green indicates approximately 50% sequence conservation and red indicates greater than 80%
sequence conservation. The predicted TonB-binding surface is outlined in black.
Despite the successes afforded from docking region II peptides to FhuD,
the same strategies were unable to yield interpretable data when region III was
docked. This outcome is probably due to incompatibility between the
conformation of region III as observed in known structures and the predicted
TonB-binding topology on FhuD. Furthermore, the co-crystal structures of TonB
bound to FhuA and to BtuB revealed that region III forms extensive interactions
167
with the transporters’ Ton boxes. Accordingly, we modeled region III with a
conformation derived from these structures. However, since this conformation
yielded poor docking solutions, we can only speculate on how it could interact
with FhuD.
The distance on TonB that separates FhuD-binding region II from region
III includes nearly 20 residues (TonB residues 133–152). Because much
redundancy exists within this sequence, it probably confers little structural
propensity. However, this spacing might provide a flexible linker, enabling TonB
region III to contact more of the predicted TonB-binding surface, not greatly
populated in our docking simulations. Of the residues comprising region III, three
make hydrogen bond contacts with FhuA and BtuB Ton boxes, while four
residues project in the opposite direction. The first of these residues, TonB
residue Arg-154, could form favourable electrostatic interactions with the rest of
the FhuD surface, even when bound to an OM receptor. It is possible that our
lowest energy docking solution represents a favourable conformation of region II
bound to FhuD. TonB residues 133–152 might then act as a flexible linker that
anchors residue Arg-154 within region III to the remainder of this surface. This
orientation would place FhuD’s siderophore binding site directly beneath a
receptor’s lumen where it could capture siderophore as it emerges during
transport.
Further structural work is required to better understand the way in which
TonB and FhuD interact. Our study indicates that TonB region I is not essential
in forming a TonB–FhuD complex. While we cannot exclude the possibility that
168
region I contributes in some way to complex formation, we speculate that it might
do so transiently at some point during the energy transduction cycle. Absence of
the proline-rich region also has no effect on complex formation. Similarly,
removal of the first 102 residues of TonB does not prevent complex formation.
We predict that even smaller derivatives of TonB that contain only regions II and
III will still form a complex. Such a derivative would be an ideal candidate for
structural studies as it should reduce the mobility of these N-terminal regions
when bound to FhuD. This is critical since TonB–FhuD co-crystallization
strategies may require reduction of the overall flexibility between these two
proteins. Insight provided from these approaches is necessary to reveal ways that
TonB can unify siderophore transport by acting as a protein–protein interaction
scaffold.
4.5 Acknowledgments
This work was supported by operating grants to J.W.C. from the Canadian
Institutes of Health Research (CIHR). Sheldon Biotechnology Centre is
supported by a Research Resource Grant from CIHR. Canada Foundation for
Innovation provided infrastructure for surface plasmon resonance to the Montreal
Integrated Genomics Group for Research on Infectious Pathogens. D.M.C. is
recipient of a fellowship from the Groupe d'étude des protéines membranaires,
Université de Montréal. J.C.D. is recipient of a Canada Graduate Scholarship
from the Natural Sciences and Engineering Research Council of Canada. The
authors thank C. Ng-Thow-Hing, P. Schuck and K. James for help with AUC data
collection, S.G. Paquette for providing maltose-binding protein and R. Huey and
D. Goodsell for helpful advice on conducting the docking experiments.
169
Preface to Chapter 5
Determinants of binding between TonB and FhuD were identified in chapter 4.
Our finding that TonB 103–239 bound to FhuD with similar affinities as TonB 33–239
and TonB Δ66–100, suggested that it could be useful for co-crystallization trials. In
chapter 5, initial attempts of on-going efforts to co-crystallize a TonB–FhuD complex are
described. In addition, properties of TonB that are refractory to crystallization are
identified. The chapter ends with suggestions for future attempts at co-crystallizing the
complex.
170
Chapter 5
Crystallization screening of the TonB–FhuD complex
171
5.1 Introduction
Interactions between the energy transducer TonB, and the periplasmic
binding protein, FhuD were described in chapters 3 and 4. Both full-length
periplasmic TonB, possessing residues 33–239, and a shorter derivative
possessing residues 103–239 displayed concentration-dependent binding to FhuD,
consistent with formation of a 1:1 complex. Since TonB 103–239 possesses the
majority of predicted FhuD-binding regions and was demonstrated to be less
elongated (and presumably less flexible), it was selected as a candidate for
crystallization trials with FhuD.
To understand atomic details of TonB–FhuD interactions, structural
knowledge is desired. X-ray crystallography provides a means to directly
visualize protein–protein interactions. Determinants of binding between TonB
and FhuD can be revealed if a TonB–FhuD complex can be crystallized. A
structure solved from these crystals will directly test hypotheses borne from phage
display, biophysical, and computational analyses.
Protein crystallization remains the bottleneck of structure determination
efforts. High-throughput screening has considerably reduced the efforts required
to identify crystallization conditions. When screening, conditions that promote
protein nucleation and incorporation into a crystal lattice are identified. While
this process may directly identify conditions that generate crystals, it may also
identify promising conditions; refined screening then narrows conditions that
favour crystallization. This chapter describes efforts to purify and crystallize a
TonB–FhuD complex. In addition, the stability of TonB within the TonB–FhuD
172
complex is examined. Finally, attempts to improve crystallization outcomes by
chemically cross-linking TonB and FhuD are described.
5.2 Materials and methods
5.2.1 Bacterial strains and plasmids
Decahistidine-tagged TonB 103–239 was expressed from plasmid
pDMC01, as described in chapter 4. FhuD-TEV (decahistidine-tagged FhuD plus
a TEV protease cleavage site to facilitate tag removal) was expressed from
plasmid pJD01 (chapter 4). Both were expressed in E. coli strain BL21 DE3
(pLysS). Hexahistidine-tagged TEV NIa protease was expressed from plasmid
pMHTd238, a construct from the Protein Structure Initiative Material Repository
(PSI-MR; Harvard University, Boston, MA, USA) (195). TEV was expressed in
E. coli BL21-CodonPlus-RIL (Stratagene), a strain possessing an additional
plasmid for constitutive expression of rare codons.
5.2.2 Protein expression and purification
TonB 103–239 and FhuD-TEV were expressed and purified as outlined in
chapter 4. TEV was expressed and purified as described previously (195).
5.2.3 Removal of FhuD His-tag
FhuD’s decahistidine tag was removed by TEV proteolysis. Prior to
proteolysis, FhuD-TEV was stored in chromatographic running buffer from the
173
final step in its purification: 50 mM Tris pH 8.2 plus 120 mM imidazole (Fluka).
FhuD-TEV in running buffer (~20 mg) was diluted into TEV proteolysis buffer:
50 mM Tris pH 8.0, 15 mM NaCl and glycerol was added to approximately 10%
(v/v). TEV protease (~2 mg) was added to yield a TEV:FhuD ratio of 1:10 (w/w).
Cleavage proceeded overnight at room temperature with gentle stirring. Cleaved
FhuD (hereafter referred to as FhuD) was separated from TEV protease and
uncleaved FhuD by applying the reaction mixture to a Poros MC 20 Ni-chelate
column (Applied Biosystems); cleaved FhuD was collected in the flowthrough
portion. Completion of proteolysis was confirmed by SDS-PAGE and by
Western blotting for loss of immunoreactivity towards anti-His monoclonal
antibodies (Clonetech, Mountain View, CA).
5.2.4 TonB–FhuD–Fcn complex formation
Prior to complex formation, TonB 103–239 and FhuD were dialyzed
against 50 mM Tris pH 7.4, 50 mM NaCl and 10% glycerol. Protein
concentrations were determined by a Bradford Assay (Bio-Rad). To aid the
downstream strategies of structure determination, FhuD’s siderophore binding site
was saturated with the siderophore, ferricrocin (Fcn); a 10-fold molar Fcn excess
was added to FhuD. TonB–FhuD complexes were formed by mixing TonB 103–
239 with ferricrocin-bound FhuD at a 1:1 molar ratio. Complexes were
concentrated to ~10 mg/ml by ultrafiltration with YM-10 membranes (Millipore).
Equimolar complex formation was assessed by visualization of silver-stained
SDS-PAGE gels.
174
5.2.5 TonB–FhuD–Fcn crystallization screening
The TonB–FhuD–Fcn complex was shipped for high-throughput
crystallization screening at the Hauptman-Woodward Medical Research Institute
(Buffalo, NY). Screening was conducted in microbatch format; distributed
among wells of a crystallization tray were 1536 different crystallization cocktail
solutions. Each well contained a different cocktail that comprised various
precipitating agents and buffers. An equal volume (< 1 µl) of the TonB–FhuD–
Fcn complex was then dispensed into each well. Paraffin oil was layered over the
wells and screening proceeded over the course of six weeks. Crystallization
outcomes were monitored weekly over the course of six weeks; magnified images
of the contents in each well were visually examined for crystallization outcomes,
or for promising leads.
5.2.6 Assessing TonB degradation
Proteolytic degradation of TonB 103–239 within the TonB–FhuD complex
and degradation of isolated TonB 103–239 were assessed by SDS-PAGE and
Western blotting. TonB 103–239 within the TonB–FhuD complex was assessed
for degradation in the absence of protease inhibitors one month after the initial
complex formation. Isolated TonB was assessed for degradation over a time
course of one week in the presence and absence of a Complete EDTA-free
protease inhibitor tablet (Roche). Degradation patterns were characterized by the
appearance of smaller molecular weight bands in a silver-stained SDS-PAGE gel
175
and by loss of immunoreactivity towards anti-His monoclonal antibodies in a
Western blot.
5.2.7 TonB–FhuD cross-linking
To improve crystallizability of the TonB–FhuD complex, TonB 103–239
was cross-linked to FhuD. Proteins were dialyzed into a buffer containing 50 mM
Hepes pH 7.5 plus 150 mM NaCl. TonB and FhuD were then mixed at 1:1 molar
ratios and formaldehyde (Pierce) was added to a 10,000-fold molar excess.
Cross-linking proceeded for approximately 72 h by incubation of the reaction
mixture at 20°C without agitation. Extent of cross-linking was assessed by SDS-
PAGE and by Western blotting for immunoreactivity towards anti-His
monoclonal antibodies; samples containing 5 µg of total protein were
concentrated to dryness in a Vacufuge (Eppendorf), re-suspended in SDS-PAGE
sample buffer, and electrophoretically developed. Prior to electrophoresis,
samples were heated for 10 minutes at 60°C. As negative control, MBP was
mixed with TonB 103–239 and cross-linked as outlined above.
5.3 Results
5.3.1 Protein preparations and processing
After purification, protein purities were assessed by silver-stained SDS-
PAGE gels; all proteins purified to apparent electrophoretic homogeneity (Figure
5.1). To improve FhuD’s solubility, its decahistidine tag was removed by the
176
activity of purified TEV protease. Overnight proteolysis was nearly complete and
approximately 80% of input FhuD was recovered in its cleaved form (Figure 5.1).
Furthermore, removal of FhuD’s His-tag was confirmed by loss of
immunoreactivity towards anti-His monoclonal antibodies in a Western blot (data
not shown).
5.3.2 TonB–FhuD–Fcn complex formation
Having purified both proteins to homogeneity, the TonB–FhuD complex
was formed. After saturating FhuD’s siderophore binding site with Fcn, TonB
103–239 was added to yield a complex with 1:1 molar ratio. The complex was
then concentrated to ~10 mg/ml and analyzed by SDS-PAGE to assess the
complex’s stoichiometry. When silver-stained, corresponding SDS-PAGE gels
indicated that TonB 103–239 and FhuD–Fcn were mixed at approximately the
desired molar ratio (Figure 5.2).
5.3.2 High-throughput crystallization screening
By high-throughput screening, we searched for conditions that promoted
crystallization of the TonB–FhuD–Fcn complex. After six weeks of observation,
crystal formation was not apparent. Most conditions exhibited varying degrees of
precipitation (data not shown). Other conditions failed to reveal any change and
the complex remained clarified in solution. However, a few conditions yielded
promising leads that are worthy of further investigation (Figure 5.3).
177
Figure 5.1. Protein purification and processing. Typical TonB 103–239 (left) and FhuD (right)
purifications are depicted as grayscale images of silvers-stained SDS-PAGE gels. Left: TonB
103–239 preparation. Lane 1: molecular weight markers with selected bands identified; lane 2:
purified TonB 103–239. Right: FhuD preparation. Lane 1: molecular weight markers with
selected bands identified; lane 2: decahistidine-tagged FhuD; lane 3: purified TEV protease; Lane
4: cleaved, label-free FhuD.
Figure 5.2. Complexation of TonB 103–239–FhuD–Fcn. FhuD was saturated with a 10-fold
molar excess of Fcn and complexed with TonB 103–239 to yield a 1:1 protein molar ratio. A
typical TonB–FhuD complex is depicted as a grayscale image of a silver-stained SDS-PAGE gel.
Lane 1: molecular weight markers with selected weights illustrated to the left; lane 2: cleaved
FhuD standard; lane 3: TonB 103–239 standard; lanes 4–10: gradient of TonB–FhuD–Fcn
complexes (120 ng total protein in lane 4 up to 6 µg total protein in lane 10).
178
Figure 5.3. Crystallization screen of the TonB 103–239–FhuD–Fcn complex. Indicated in panels
A–D are images from the most promising leads taken after six weeks of incubation. In each panel
a grayscale image is displayed to the left of an equivalent colour image. Crystallization conditions
are as follows: A. 0.1 M Ammonium sulfate, 0.1 M CAPS pH 10, 40% (w/v) PEG 4000; B. 1.9 M
Sodium malonate pH7; C. 0.12 M Sodium phosphate monobasic monohydrate, 0.68 M Potassium
phosphate dibasic pH 7.5; D. 0.1 M Sodium molybdate dihydrate, 0.1 M MES pH 6, 20% (w/v)
PEG 8000.
A
B
C
D
A
B
C
D
179
Cocktails that yielded favourable outcomes exhibited no apparent trend in
composition. The most favourable precipitating agents included complex anions
such as malonate and phosphate and varying lengths of the organic polymer
polyethylene glycol (PEG). Buffer counter ions were predominantly monovalent
cations, such as Na+, and generally exhibited pH values near neutrality. However,
the buffer system compositions varied considerably and included ammonium
sulfate/CAPS, Na+/K
+ phosphate, and sodium molybdate/MES.
Outcomes of the promising conditions varied considerably. One condition
with the precipitant, PEG 4000, yielded needle-like clusters whose growth might
be controlled under favourable conditions (Figure 5.3A) Other promising
conditions yielded dense, brown-coloured particulates, the colour of which may
arise due to presence of Fcn (Figure 5.3B-D). Particulates derived from
incubation with sodium malonate, exhibited crystalline-like dimensional qualities
(Figure 5.3B). However, the small volume of this crystallization drop and the
presence of glycerol, which obscured visual inspection, made a confirmation of
crystal growth impossible.
5.3.3 TonB exhibits time-dependent degradation
Over time, TonB 103–239 suffers proteolytic cleavage leaving less in-tact
to form a complex with FhuD. After approximately two months, an
electrophoretically pure band of TonB 103–239 within the TonB–FhuD complex
(Figure 5.4A) yields a ladder of proteolytic degradation products (Figure 5.4B).
Regions within TonB that are susceptible to degradation were determined by
180
Western blotting for loss of immunoreactivity towards anti-His monoclonal
antibodies. Over the course of one week and in the absence of protease inhibition,
TonB first suffers proteolytic cleavage at its N-terminus followed by loss of C-
terminal residues (Figure 5.4). Bands corresponding to the largest degradation
product are not immunoreactive towards anti-His antibodies, indicating that
TonB’s N-terminus was lost. In contrast, smaller degradation products remain
immunoreactive, indicating loss of TonB’s C-terminal residues. Addition of a
protease inhibitor cocktail improves stability. After two months, degradation was
reduced in the presence of an inhibitor cocktail tablet (data not shown).
Figure 5.4. Degradation of TonB 103–239. The susceptibility of TonB 103–239 to suffer
proteolytic degradation in complex with FhuD and in isolation was assessed by SDS-PAGE and
Western blotting. Grayscale images of silver-stained SDS-PAGE gels and a Western blot are
depicted. A. TonB–FhuD complex developed on the day it was prepared. Lane 1: molecular
weight markers. Selected weights are indicated to the left of the gel; lane 2: TonB–FhuD
complex. B. TonB–FhuD complex after 2 months. Lanes are labelled as in Panel A. C. Western
blot of isolated TonB 103–239. Lane 1: molecular weight markers. Selected weights are
indicated to the left of the blot; lane 2: TonB 103–239 after one week without proteolytic
inhibition; lane 3: silver-stained SDS-PAGE overlay of the same lane 2 material aligned to the
Western to emphasize selected loss of immunoreactivity towards anti-His monoclonal antibodies.
A B C
181
5.3.4 TonB–FhuD formaldehyde-cross-linking
Formaldehyde cross-linking was conducted as a strategy to reduce
flexibility of the TonB–FhuD–Fcn complex. After 72 h, cross-linking was
evident by appearance of higher order bands that were immunoreactive towards
anti-His monoclonal antibodies in a Western blot (Figure 5.5). In the absence of
FhuD, TonB 103–239 formed an ~38 kDa cross-linked species consistent with
formation of a dimer (Figure 5.5B). In the presence of FhuD, an ~32 kDa band
appeared, in addition to the TonB dimer (Figure 5.5B). The 32 kDa band may
arise from trace amounts of uncleaved FhuD. However, it was absent in the FhuD
control and only appeared in the presence of TonB 103–239. Formation of the
TonB–FhuD complex was evident by appearance of an ~50 kDa band in the
Western (Figure 5.5B, lane 4), consistent with the molecular weight of a
Figure 5.5. Formaldehyde cross-linking of TonB–FhuD complex. Completion of cross-linking
was assessed by SDS-PAGE and Western blotting. Grayscale images of silver-stained SDS-
PAGE gels and a Western blot are depicted. A. SDS-PAGE of cross-linked samples. Lane 1:
molecular weight markers. selected weights are labelled to the left of the gel.; Lane 2: TonB 103–
239 standard; Lane 3: FhuD standard; Lane 4: TonB and FhuD mixed at a 1:1 ratio; Lane 5: MBP
standard; Lane 6: TonB and MBP mixed at a 1:1 ratio. B. Western blot of the gel displayed in
panel A. Lane identities are as in panel A. Selected molecular weights are indicated to the left of
the blot. The putative TonB–FhuD complex is marked by an arrow in lane 4.
A B
182
TonB 103–239–FhuD complex (Mr = 48.1 kDa). Cross-linking was specific;
incubation with MBP failed to resolve higher order cross-linked bands in the
Western. However, cross-linking did not proceed to completion, as large amounts
of uncross-linked species were apparent in all conditions tested.
5.4 Discussion
Strategies to co-crystallize a TonB–FhuD complex require a researcher to
address many technical challenges. First, both proteins must be purified to
homogeneity. Our protein preparations display electrophoretic homogeneity and
fulfill this essential criterion. Second, proteins must remain soluble when highly
concentrated. We previously observed that FhuD is only marginally soluble when
concentrated (unpublished observations). We rationalized that removal of FhuD’s
His-tag might improve its solubility. This hypothesis proved correct. We
succeeded to complex TonB 103-239 with FhuD and concentrate the sample
enough to withstand crystallization screening without serious precipitation. After
six weeks many of the crystallization cocktails remained clarified, indicating that
the complex has no real tendency to precipitate. However, concentration of
cleaved FhuD in the absence of TonB 103–239 offers no real advantage compared
to decahistidine-tagged FhuD; both proteins precipitated with time.
A third technical challenge to overcome is protein degradation. TonB is
proteolytically labile; in the absence of protease inhibition TonB completely
degrades within two months. This is problematic for crystallization. Addition of
protease inhibitors somewhat improves the situation. However, it is unclear
183
whether such additives affect crystallization outcomes. Desirable would be to
identify sites of proteolysis within TonB by mass spectrometry. Knowledge of
proteolytic sites could be used to generate stable mutants by site-directed
mutagenesis. This may stabilize TonB by obscuring protease recognition sites.
Stabilized TonB might better incorporate into a three-dimensional crystal lattice.
Discussed in chapter 4, a shorter TonB derivative that possesses only
predicted FhuD-binding regions II and III plus its C-terminal domain might be
better suited for crystallization purposes. For example, a TonB derivative
containing residues 124–239 (TonB 124–239) would possess fewer of the
residues known to be flexible and unstructured (56), yet would still retain
essential FhuD-binding regions. Those regions might be stabilized upon binding
FhuD, yielding a more stable complex.
A TonB 124–239 derivative could also be useful for crystallization
screening of a complex with ExbD. The Coulton lab possesses an ExbD construct
that expresses the entire protein as fusion to glutathione S-transferase. The
construct has been purified to homogeneity and is stable (unpublished results).
The TonB 124–239 derivative would be appropriate for complex formation with
ExbD, as it contains TonB residue 150 that was recently demonstrated to interact
with ExbD residue 92 (87).
Given that TonB 103–239 within the TonB–FhuD complex degraded with
time and is known to be conformationally flexible, we attempted to chemically
cross-link the complex. This approach was successfully used to crystallize an
adrenodoxin–adrenodoxin reductase complex (196). Our data indicated that
184
TonB–FhuD complexes could be covalently trapped; however, cross-linking
never proceeded to completion. This outcome appears to be common for other
bona fide protein–protein complexes such as TonB–FepA complexes (166), and
E. coli HypB–SlyD hydrogenase accessory complexes (197). The amounts of
complex that visibly cross-linked would not yield enough material to facilitate
crystallization screening. If such amounts were purified, any unavoidable and
anticipated sample loss would yield diminishing amounts of complex.
Finally, another strategy that warrants further investigation is to increase
the molar amount of TonB 103–239 over FhuD when preparing complexes for
crystallization. This approach was successful for TonB–FhuA and TonB–BtuB
complex formation; two-fold and six-fold molar TonB excesses over FhuA and
BtuB, respectively, yielded highly diffracting crystals (96,97). Given TonB’s
proteolytic susceptibility, this approach has potential merit.
5.5 Acknowledgements
Advice and assistance with crystallization screening was provided by N.
Croteau. High-throughput screening was conducted by T. Veatch at the
Hauptman-Woodward Institute. Thanks to J. Deme for experimental
contributions.
185
Chapter 6
Conclusions and future work
186
6.0 Thesis objectives within the context of TonB-dependent transport
The objectives of this thesis were to characterize periplasm-localized
interactions between TonB and protein partners of the ferric hydroxamate uptake
system. In chapter 2, interactions between TonB and FhuA were predicted and
localized (189). Our use of phage display was the first literature report to
document predictions of bacterial protein–protein interactions. These predictions
were later confirmed with our TonB–FhuA crystal structure (96). In addition to
the structure’s novelty, it proved that phage display can accurately predict
protein–protein interactions. Therefore, we are confident that findings from
chapter 3, that TonB and FhuD interact (189), represent bona fide biological
interactions. Our characterization of TonB–FhuD interactions elucidated the
stoichiometry and affinity of complex formation. In chapter 4, we elucidated
regions within TonB that are essential for this interaction. We further predict that
TonB can orient FhuD towards the OM when the two proteins are complexed.
We predict that this enhances siderophore uptake by localizing FhuD near to the
inner leaflet of the OM, where it can bind siderophore during transport.
There are many reasons why this mechanism is biologically plausible. By
analogy to FepB (187), there are probably around a few thousand copies of FhuD
per cell under iron-limiting conditions. Conversely, TonB is postulated to be
present at up to 400 copies per cell (82). The affinity that we calculate and the
molar excess of FhuD over TonB would ensure that it remains bound to TonB
during transport. This feature is consistent with the postulated mechanical pulling
model of TonB-dependent energy transduction (106). By retracting towards the
187
CM, TonB bound to FhuA would accomplish at least three productive outcomes.
First, retraction would cause FhuA’s cork to unfold enough to allow siderophore
translocation into the periplasm. Second, by binding and orienting FhuD towards
the OM, TonB would constructively position a siderophore binding protein.
Third, by retracting towards the CM, TonB would position siderophore-bound
FhuD nearer to the CM-embedded FhuB/C permease. This would allow FhuD to
exchange binding partners and engage its cognate permease.
Determinants of binding between TonB and FhuD govern this protein–
protein interaction. We identified features on FhuD that might comprise these
determinants. Until we elucidate the TonB–FhuD crystal structure, these features
remain speculative. However, intriguing are our findings that putative TonB-
binding determinants on FhuD are flexible. Flexibility of this FhuD region and of
TonB might act as a switch that governs binding between these proteins. Within
the context of the mechanical pulling model, TonB may conformationally cycle as
it retracts toward the CM. This could serve as a mechanism to release FhuD,
allowing delivery to FhuB/C. Reports of TonB’s conformational plasticity in vivo
support this hypothesis (79,164). Thus, we support a model whereby TonB acts
as a dynamic scaffold that couples energy transduction to protein trafficking.
Many features of this model remain to be elucidated and are now discussed.
188
6.1 Directions for future research
6.1.1 Demonstration of TonB–FhuD interactions in vivo
Our biophysical characterizations have unequivocally demonstrated that
TonB and FhuD interact in vitro. Desirable would be to corroborate these
findings in vivo. Simple in vivo experimental strategies that could demonstrate
TonB–FhuD interactions include cross-linking and pull-down assays. Both
methods have benefits and drawbacks. Formaldehyde cross-linking demonstrated
TonB–FepA and TonB–ExbD interactions in vivo (87,91). However, the same
method also demonstrated TonB–OmpA and TonB–Lpp interactions, which
arguably are not biologically relevant (99).
Disulfide cross-linking might improve the specificity of interactions,
especially since we have data that demonstrates complementary regions of
interaction between TonB and FhuD. Surface-exposed residues on both proteins
that localize within these complementary regions could be individually mutated to
cysteine. Were the proteins to interact in vivo, their cross-linked species would
easily be detected by Western analysis. However, surface-exposed cysteines
might also promote homodimerization of either protein, which could interfere
with heterodimerization.
A pull-down assay could also potentially demonstrate TonB–FhuD
interactions. The experimental design would be challenging. In principle, either
histidine-tagged TonB or histidine-tagged FhuD could be immobilized on Ni2+
chelate resin as bait. If TonB were selected, it would be most desirable to capture
189
its full-length, transmembrane form. However, full-length, transmembrane TonB
is both difficult to purify and prone to rapid proteolysis (unpublished
observations). Therefore, FhuD would be the logical choice as bait. Cell lysate
from a culture that expresses tagless TonB would then be incubated with
immobilized FhuD. Addition of imidazole should then elute the TonB–FhuD
complex, which could be detected by SDS-PAGE or by Western analysis.
Capture of full-length, transmembrane TonB as prey may also be confounded by
its tendency to rapidly degrade and by the membrane components of the cell
envelope.
6.1.2 Refinement of TonB–FhuD interaction localizations
Chapter 4 described interactions between TonB derivatives and FhuD.
Outcomes of these studies indicated that neither TonB’s proline-rich region, nor
residues 33-102 were essential for interaction with FhuD. Desirable, would be to
identify regions within TonB that are most essential for this interaction. Different
strategies could be employed to identify these regions. Alanine-scanning
mutagenesis could be employed to mutate TonB’s three FhuD-interacting regions
that were identified from phage display. By mutating each region, individually or
in combination, one could infer which regions are essential for interaction with
FhuD by SPR analysis. This information would guide the generation of TonB
derivatives that may be more amenable to crystallization.
190
6.1.3 TonB–FhuD crystallization
Chapter 5 described our initial attempts to crystallize a TonB–FhuD
complex. Our approach favoured use of a truncated derivative of TonB and a
derivative of FhuD without its decahistidine tag. By high-throughput screening
we identified a few promising crystallization leads. These must now be refined in
order to identify conditions that grow co-crystals. The precipitants PEG and
malonate produced the most favourable outcomes. These represent the most
obvious leads to refine through further screening. Refinement of these conditions
should concentrate on screening precipitant concentrations, values of pH and
perhaps inclusion of other additives. Other possibilities were discussed in
Chapter 5.
6.1.4 Phage display predictions of TonB-interacting proteins
Phage panning against purified TonB has successfully predicted
periplasm-localized interactions between TonB and FhuA, and has successfully
predicted interactions between TonB and FhuD. Data obtained from phage
panning against purified TonB can now be used to predict additional periplasm-
localized TonB interactions. This represents an on-going effort in the Coulton lab
whereby TonB affinity-selected peptides are being compared to sequences of Fhu
system proteins: ExbB, ExbD and FhuB/C. Upon completion, these analyses
should predict all regions of interaction between TonB and partner proteins that
localize within the periplasm. This information will assist the construction of an
191
interaction map that can rationalize key features of TonB-dependent energy
transduction.
6.1.5 Determination of whether TonB regulates binding of siderophores to FhuD
Our findings that TonB binds with similar affinities apo-FhuD and
siderophore-bound FhuD demonstrated that occupancy of FhuD’s binding site
does not regulate binding to TonB. However, it is still not known whether TonB
regulates binding of siderophore to FhuD. TonB-dependent regulation is possible
since our phage display analyses predicted that TonB could bind to FhuD in a
region proximal to its siderophore binding site. Whether TonB influences
siderophore binding could be determined by comparing siderophore affinities
between apo-FhuD and TonB-bound FhuD. The most direct way to accomplish
this is by titration of FhuD’s binding site with siderophore in the presence of
TonB. Fluorescence spectroscopy and isothermal titration calorimetry are two
experimental techniques to accomplish this objective. Both would enable
calculations of siderophore binding affinities.
6.1.6 Elucidation of siderophore binding sites by phage display
Our use of phage display focused on the prediction and localization of
protein–protein binding sites. However, its applications extend beyond the realm
of protein–protein interactions. Phage display has also been used to predict small
molecule binding sites on proteins (137). This application could be used to pan
phage libraries against the siderophore ferricrocin, in order to identify
192
siderophore-binding patterns within the Fhu system. The known binding sites
within FhuA and FhuD would serve as excellent controls to assess the ability of
the technique to discriminate siderophore binding sites. This strategy could
potentially identify siderophore binding motifs within FhuA that could correspond
to the translocation pathway. Were such a pathway found, it could be occluded
by introduction of disulfide bonds as was previously done for FhuA (108).
Ferricrocin uptake could then be measured to see if the occlusion reduced
transport. In a similar way, ferricrocin binding sites within other siderophore-
binding proteins, such as FhuB/C could be identified. This information could also
identify ferricrocin binding sites in cytoplasmic proteins involved in siderophore
synthesis or degradation.
193
References
1. Weinberg, E. D. 1997. The Lactobacillus anomaly: total iron abstinence.
Perspect. Biol. Med. 40:578-583.
2. Lieu, P. T., M. Heiskala, P. A. Peterson, and Y. Yang. 2001. The roles
of iron in health and disease. Mol. Aspects Med. 22:1-87.
3. Outten, F. W. and E. C. Theil. 2009. Iron-based Redox Switches in
Biology. Antioxid. Redox. Signal. 11:1029-1046.
4. Halliwell, B. and J. M. Gutteridge. 1984. Oxygen toxicity, oxygen
radicals, transition metals and disease. Biochem. J. 219:1-14.
5. Boukhalfa, H. and A. L. Crumbliss. 2002. Chemical aspects of
siderophore mediated iron transport. Biometals 15:325-339.
6. Raymond, K. N. and E. A. Dertz. 2004. Biochemical and physical
properties of siderophores, p. 3-17. In Jorge H.Crosa, Alexandra R.Mey,
and Shelley M.Payne (ed.), Iron Transport in Bacteria. ASM Press,
Washington D.C.
7. Raymond, K. N., E. A. Dertz, and S. S. Kim. 2003. Enterobactin: an
archetype for microbial iron transport. Proc. Natl. Acad. Sci. USA
100:3584-3588.
8. Lin, J., S. Huang, and Q. Zhang. 2002. Outer membrane proteins: key
players for bacterial adaptation in host niches. Microbes. Infect. 4:325-
331.
9. Hancock, R. E. and H. Nikaido. 1978. Outer membranes of gram-
negative bacteria. XIX. Isolation from Pseudomonas aeruginosa PAO1
and use in reconstitution and definition of the permeability barrier. J.
Bacteriol. 136:381-390.
10. Nikaido, H. 2003. Molecular basis of bacterial outer membrane
permeability revisited. Microbiol. Mol. Biol. Rev. 67:593-656.
11. Lerouge, I. and J. Vanderleyden. 2002. O-antigen structural variation:
mechanisms and possible roles in animal/plant-microbe interactions.
FEMS Microbiol. Rev. 26:17-47.
12. Lugtenberg, B. and A. L. Van. 1983. Molecular architecture and
functioning of the outer membrane of Escherichia coli and other gram-
negative bacteria. Biochim. Biophys. Acta 737:51-115.
194
13. Nikaido, H. 1996. Outer Membrane, p. 29-47. In Escherichia coli and
Salmonella Cellular and Molecular Biology. ASM Press, Washington,
D.C.
14. Oliver, D. B. 1996. Periplasm, p. 88-103. In Escherichia coli and
Salmonella. ASM Press, Washington, D.C.
15. Parkt, J. T. 1996. The Murein Sacculus, p. 48-57. In Escherichia coli and
Salmonella: Cellular and Molecular Biology. ASM Press, Washington,
D.C.
16. Vollmer, W. and U. Bertsche. 2008. Murein (peptidoglycan) structure,
architecture and biosynthesis in Escherichia coli. Biochim. Biophys. Acta
1778:1714-1734.
17. Meroueh, S. O., K. Z. Bencze, D. Hesek, M. Lee, J. F. Fisher, T. L.
Stemmler, and S. Mobashery. 2006. Three-dimensional structure of the
bacterial cell wall peptidoglycan. Proc. Natl. Acad. Sci. USA 103:4404-
4409.
18. Gan, L., S. Chen, and G. J. Jensen. 2008. Molecular organization of
Gram-negative peptidoglycan. Proc. Natl. Acad. Sci. USA 105:18953-
18957.
19. Pautsch, A. and G. E. Schulz. 1998. Structure of the outer membrane
protein A transmembrane domain. Nat. Struct. Biol. 5:1013-1017.
20. Ni, Y., J. Reye, and R. R. Chen. 2007. Lpp deletion as a permeabilization
method. Biotechnol. Bioeng. 97:1347-1356.
21. Shu, W., J. Liu, H. Ji, and M. Lu. 2000. Core structure of the outer
membrane lipoprotein from Escherichia coli at 1.9 Å resolution. J. Mol.
Biol. 299:1101-1112.
22. Piovant, M. and C. Lazdunski. 1982. Biosynthesis of Escherichia coli
Braun's lipoprotein precursors in vitro and binding to membrane vesicles.
Eur. J. Biochem. 125:623-629.
23. Cascales, E., A. Bernadac, M. Gavioli, J. C. Lazzaroni, and R.
Lloubes. 2002. Pal lipoprotein of Escherichia coli plays a major role in
outer membrane integrity. J. Bacteriol. 184:754-759.
24. Kadner, R. J. 1996. Cytoplasmic Membrane, p. 58-87. In Escherichia
coli and Salmonella Cellular and Molecular Biology. ASM Press,
Washington D.C.
195
25. Basle, A., G. Rummel, P. Storici, J. P. Rosenbusch, and T. Schirmer.
2006. Crystal structure of osmoporin OmpC from E. coli at 2.0 Å. J. Mol.
Biol. 362:933-942.
26. Cowan, S. W., T. Schirmer, G. Rummel, M. Steiert, R. Ghosh, R. A.
Pauptit, J. N. Jansonius, and J. P. Rosenbusch. 1992. Crystal structures
explain functional properties of two E. coli porins. Nature 358:727-733.
27. Yamashita, E., M. V. Zhalnina, S. D. Zakharov, O. Sharma, and W.
A. Cramer. 2008. Crystal structures of the OmpF porin: function in a
colicin translocon. EMBO J. 27:2171-2180.
28. Perozo, E. and D. C. Rees. 2003. Structure and mechanism in prokaryotic
mechanosensitive channels. Curr. Opin. Struct. Biol. 13:432-442.
29. Dutzler, R., Y. F. Wang, P. Rizkallah, J. P. Rosenbusch, and T.
Schirmer. 1996. Crystal structures of various maltooligosaccharides
bound to maltoporin reveal a specific sugar translocation pathway.
Structure. 4:127-134.
30. Charbit, A. 2003. Maltodextrin transport through LamB. Front Biosci.
8:265-274.
31. Hearn, E. M., D. R. Patel, B. W. Lepore, M. Indic, and B. van den
Berg. 2009. Transmembrane passage of hydrophobic compounds through
a protein channel wall. Nature 458:367-370.
32. Stojiljkovic, I., M. Cobeljic, and K. Hantke. 1993. Escherichia coli K-
12 ferrous iron uptake mutants are impaired in their ability to colonize the
mouse intestine. FEMS Microbiol. Lett. 108:111-115.
33. Cartron, M. L., S. Maddocks, P. Gillingham, C. J. Craven, and S. C.
Andrews. 2006. Feo--transport of ferrous iron into bacteria. Biometals
19:143-157.
34. Blanton, K. J., G. D. Biswas, J. Tsai, J. Adams, D. W. Dyer, S. M.
Davis, G. G. Koch, P. K. Sen, and P. F. Sparling. 1990. Genetic
evidence that Neisseria gonorrhoeae produces specific receptors for
transferrin and lactoferrin. J. Bacteriol. 172:5225-5235.
35. Legrain, M., V. Mazarin, S. W. Irwin, B. Bouchon, M. J. Quentin-
Millet, E. Jacobs, and A. B. Schryvers. 1993. Cloning and
characterization of Neisseria meningitidis genes encoding the transferrin-
binding proteins Tbp1 and Tbp2. Gene 130:73-80.
36. Dhaenens, L., F. Szczebara, and M. O. Husson. 1997. Identification,
characterization, and immunogenicity of the lactoferrin-binding protein
from Helicobacter pylori. Infect. Immun. 65:514-518.
196
37. Cescau, S., H. Cwerman, S. Letoffe, P. Delepelaire, C. Wandersman,
and F. Biville. 2007. Heme acquisition by hemophores. Biometals 20:603-
613.
38. Crosa, J. H. and C. T. Walsh. 2002. Genetics and assembly line
enzymology of siderophore biosynthesis in bacteria. Microbiol. Mol. Biol.
Rev. 66:223-249.
39. Jones, A. M., S. E. Lindow, and M. C. Wildermuth. 2007. Salicylic
acid, yersiniabactin, and pyoverdin production by the model
phytopathogen Pseudomonas syringae pv. tomato DC3000: synthesis,
regulation, and impact on tomato and Arabidopsis host plants. J. Bacteriol.
189:6773-6786.
40. Miethke, M. and M. A. Marahiel. 2007. Siderophore-based iron
acquisition and pathogen control. Microbiol. Mol. Biol. Rev. 71:413-451.
41. Braun, V. 2009. FhuA (TonA), the career of a protein. J. Bacteriol.
191:3431-3436.
42. Ferguson, A. D., E. Hofmann, J. W. Coulton, K. Diederichs, and W.
Welte. 1998. Siderophore-mediated iron transport: crystal structure of
FhuA with bound lipopolysaccharide. Science 282:2215-2220.
43. Ferguson, A. D., W. Welte, E. Hofmann, B. Lindner, O. Holst, J. W.
Coulton, and K. Diederichs. 2000. A conserved structural motif for
lipopolysaccharide recognition by procaryotic and eucaryotic proteins.
Structure. 8:585-592.
44. Buchanan, S. K., B. S. Smith, L. Venkatramani, D. Xia, L. Esser, M.
Palnitkar, R. Chakraborty, D. van der Helm, and J. Deisenhofer.
1999. Crystal structure of the outer membrane active transporter FepA
from Escherichia coli. Nat. Struct. Biol. 6:56-63.
45. Ferguson, A. D., R. Chakraborty, B. S. Smith, L. Esser, D. van der
Helm, and J. Deisenhofer. 2002. Structural basis of gating by the outer
membrane transporter FecA. Science 295:1715-1719.
46. Chimento, D. P., A. K. Mohanty, R. J. Kadner, and M. C. Wiener.
2003. Substrate-induced transmembrane signaling in the cobalamin
transporter BtuB. Nat. Struct. Biol. 10:394-401.
47. Sharma, O., E. Yamashita, M. V. Zhalnina, S. D. Zakharov, K. A.
Datsenko, B. L. Wanner, and W. A. Cramer. 2007. Structure of the
complex of the colicin E2 R-domain and its BtuB receptor. The outer
membrane colicin translocon. J. Biol. Chem. 282:23163-23170.
197
48. Cherezov, V., E. Yamashita, W. Liu, M. Zhalnina, W. A. Cramer, and
M. Caffrey. 2006. In meso structure of the cobalamin transporter, BtuB, at
1.95 Å resolution. J. Mol. Biol. 364:716-734.
49. Kurisu, G., S. D. Zakharov, M. V. Zhalnina, S. Bano, V. Y. Eroukova,
T. I. Rokitskaya, Y. N. Antonenko, M. C. Wiener, and W. A. Cramer.
2003. The structure of BtuB with bound colicin E3 R-domain implies a
translocon. Nat. Struct. Biol. 10:948-954.
50. Buchanan, S. K., P. Lukacik, S. Grizot, R. Ghirlando, M. M. Ali, T. J.
Barnard, K. S. Jakes, P. K. Kienker, and L. Esser. 2007. Structure of
colicin I receptor bound to the R-domain of colicin Ia: implications for
protein import. EMBO J. 26:2594-2604.
51. Cobessi, D., H. Celia, and F. Pattus. 2005. Crystal structure at high
resolution of ferric-pyochelin and its membrane receptor FptA from
Pseudomonas aeruginosa. J. Mol. Biol. 352:893-904.
52. Cobessi, D., H. Celia, N. Folschweiller, I. J. Schalk, M. A. Abdallah,
and F. Pattus. 2005. The crystal structure of the pyoverdine outer
membrane receptor FpvA from Pseudomonas aeruginosa at 3.6 angstroms
resolution. J. Mol. Biol. 347:121-134.
53. Krieg, S., F. Huche, K. Diederichs, N. Izadi-Pruneyre, A. Lecroisey, C.
Wandersman, P. Delepelaire, and W. Welte. 2009. Heme uptake across
the outer membrane as revealed by crystal structures of the receptor-
hemophore complex. Proc. Natl. Acad. Sci. USA 106:1045-1050.
54. Brillet, K., A. Meksem, E. Lauber, C. Reimmann, and D. Cobessi.
2009. Use of an in-house approach to study the three-dimensional
structures of various outer membrane proteins: structure of the alcaligin
outer membrane transporter FauA from Bordetella pertussis. Acta
Crystallogr. D. Biol. Crystallogr. 65:326-331.
55. Gudmundsdottir, A., P. E. Bell, M. D. Lundrigan, C. Bradbeer, and R.
J. Kadner. 1989. Point mutations in a conserved region (TonB box) of
Escherichia coli outer membrane protein BtuB affect vitamin B12
transport. J. Bacteriol. 171:6526-6533.
56. Peacock, R. S., V. V. Andrushchenko, A. R. Demcoe, M. Gehmlich, L.
S. Lu, A. G. Herrero, and H. J. Vogel. 2006. Characterization of TonB
interactions with the FepA cork domain and FecA N-terminal signaling
domain. Biometals 19:127-142.
57. Ferguson, A. D., C. A. Amezcua, N. M. Halabi, Y. Chelliah, M. K.
Rosen, R. Ranganathan, and J. Deisenhofer. 2007. Signal transduction
pathway of TonB-dependent transporters. Proc. Natl. Acad. Sci. USA
104:513-518.
198
58. Brillet, K., L. Journet, H. Celia, L. Paulus, A. Stahl, F. Pattus, and D.
Cobessi. 2007. A beta strand lock exchange for signal transduction in
TonB-dependent transducers on the basis of a common structural motif.
Structure. 15:1383-1391.
59. Fanucci, G. E., J. Y. Lee, and D. S. Cafiso. 2003. Spectroscopic
evidence that osmolytes used in crystallization buffers inhibit a
conformation change in a membrane protein. Biochemistry 42:13106-
13112.
60. Kim, M., G. E. Fanucci, and D. S. Cafiso. 2007. Substrate-dependent
transmembrane signaling in TonB-dependent transporters is not
conserved. Proc. Natl. Acad. Sci. USA 104:11975-11980.
61. Ferguson, A. D., V. Braun, H. P. Fiedler, J. W. Coulton, K.
Diederichs, and W. Welte. 2000. Crystal structure of the antibiotic
albomycin in complex with the outer membrane transporter FhuA. Protein
Sci. 9:956-963.
62. Ferguson, A. D., J. Ködding, G. Walker, C. Bos, J. W. Coulton, K.
Diederichs, V. Braun, and W. Welte. 2001. Active transport of an
antibiotic rifamycin derivative by the outer-membrane protein FhuA.
Structure. 9:707-716.
63. Yue, W. W., S. Grizot, and S. K. Buchanan. 2003. Structural evidence
for iron-free citrate and ferric citrate binding to the TonB-dependent outer
membrane transporter FecA. J. Mol. Biol. 332:353-368.
64. Xu, Q., J. F. Ellena, M. Kim, and D. S. Cafiso. 2006. Substrate-
dependent unfolding of the energy coupling motif of a membrane transport
protein determined by double electron-electron resonance. Biochemistry
45:10847-10854.
65. Blanvillain, S., D. Meyer, A. Boulanger, M. Lautier, C. Guynet, N.
Denancé, J. Vasse, E. Lauber, and M. Arlat. 2007. Plant carbohydrate
scavenging through tonb-dependent receptors: a feature shared by
phytopathogenic and aquatic bacteria. PLoS. One. 2:e224.
66. Schauer, K., B. Gouget, M. Carrière, A. Labigne, and H. de Reuse.
2007. Novel nickel transport mechanism across the bacterial outer
membrane energized by the TonB/ExbB/ExbD machinery. Mol.
Microbiol. 63:1054-1068.
67. Tralau, T., S. Vuilleumier, C. Thibault, B. J. Campbell, C. A. Hart,
and M. A. Kertesz. 2007. Transcriptomic analysis of the sulfate
starvation response of Pseudomonas aeruginosa. J. Bacteriol. 189:6743-
6750.
199
68. Chu, B. C., R. S. Peacock, and H. J. Vogel. 2007. Bioinformatic analysis
of the TonB protein family. Biometals 20:467-483.
69. Evans, J. S., B. A. Levine, I. P. Trayer, C. J. Dorman, and C. F.
Higgins. 1986. Sequence-imposed structural constraints in the TonB
protein of E. coli. FEBS Lett. 208:211-216.
70. Brewer, S., M. Tolley, I. P. Trayer, G. C. Barr, C. J. Dorman, K.
Hannavy, C. F. Higgins, J. S. Evans, B. A. Levine, and M. R.
Wormald. 1990. Structure and function of X-Pro dipeptide repeats in the
TonB proteins of Salmonella typhimurium and Escherichia coli. J. Mol.
Biol. 216:883-895.
71. Larsen, R. A., G. E. Wood, and K. Postle. 1994. The conserved proline-
rich motif is not essential for energy transduction by Escherichia coli
TonB protein. Mol. Microbiol. 12:857.
72. Chang, C., A. Mooser, A. Plückthun, and A. Wlodawer. 2001. Crystal
structure of the dimeric C-terminal domain of TonB reveals a novel fold.
J. Biol. Chem. 276:27535-27540.
73. Koedding, J., P. Howard, L. Kaufmann, P. Polzer, A. Lustig, and W.
Welte. 2004. Dimerization of TonB is not essential for its binding to the
outer membrane siderophore receptor FhuA of Escherichia coli. J. Biol.
Chem. 279:9978-9986.
74. Ködding, J. F., F. Killig, P. Polzer, S. P. Howard, K. Diederichs, and
W. Welte. 2005. Crystal structure of a 92-residue C-terminal fragment of
TonB from Escherichia coli reveals significant conformational changes
compared to structures of smaller TonB fragments. J. Biol. Chem.
280:3022-3028.
75. Peacock, R. S., A. M. Weljie, S. P. Howard, F. D. Price, and H. J.
Vogel. 2005. The solution structure of the C-terminal domain of TonB and
interaction studies with TonB box peptides. J. Mol. Biol. 345:1185-1197.
76. Lopez, C. S., R. S. Peacock, J. H. Crosa, and H. J. Vogel. 2009.
Molecular characterization of the TonB2 protein from the fish pathogen
Vibrio anguillarum. Biochem. J. 418:49-59.
77. Khursigara, C. M., G. De Crescenzo, P. D. Pawelek, and J. W.
Coulton. 2004. Enhanced binding of TonB to a ligand-loaded outer
membrane receptor: role of the oligomeric state of TonB in formation of a
functional FhuA•TonB complex. J. Biol. Chem. 279:7405-7412.
78. Sauter, A., S. P. Howard, and V. Braun. 2003. In vivo evidence for
TonB dimerization. J. Bacteriol. 185:5747-5754.
200
79. Ghosh, J. and K. Postle. 2005. Disulphide trapping of an in vivo energy-
dependent conformation of Escherichia coli TonB protein. Mol.
Microbiol. 55:276-288.
80. Kampfenkel, K. and V. Braun. 1993. Topology of the ExbB protein in
the cytoplasmic membrane of Escherichia coli. J. Biol. Chem. 268:6050-
6057.
81. Garcia-Herrero, A., R. S. Peacock, S. P. Howard, and H. J. Vogel.
2007. The solution structure of the periplasmic domain of the TonB
system ExbD protein reveals an unexpected structural homology with
siderophore-binding proteins. Mol. Microbiol. 66:872-889.
82. Higgs, P. I., R. A. Larsen, and K. Postle. 2002. Quantification of known
components of the Escherichia coli TonB energy transduction system:
TonB, ExbB, ExbD and FepA. Mol. Microbiol. 44:271-281.
83. Zhai, Y. F., W. Heijne, and M. H. Saier, Jr. 2003. Molecular modeling
of the bacterial outer membrane receptor energizer, ExbBD/TonB, based
on homology with the flagellar motor, MotAB. Biochim. Biophys. Acta
1614:201-210.
84. Braun, V. and C. Herrmann. 2004. Point mutations in transmembrane
helices 2 and 3 of ExbB and TolQ affect their activities in Escherichia coli
K-12. J. Bacteriol. 186:4402-4406.
85. Larsen, R. A. and K. Postle. 2001. Conserved residues Ser(16) and
His(20) and their relative positioning are essential for TonB activity,
cross-linking of TonB with ExbB, and the ability of TonB to respond to
proton motive force. J. Biol. Chem. 276:8111-8117.
86. Larsen, R. A., G. E. Deckert, K. A. Kastead, S. Devanathan, K. L.
Keller, and K. Postle. 2007. His(20) provides the sole functionally
significant side chain in the essential TonB transmembrane domain. J.
Bacteriol. 189:2825-2833.
87. Ollis, A. A., M. Manning, K. G. Held, and K. Postle. 2009. Cytoplasmic
membrane proton motive force energizes periplasmic interactions between
ExbD and TonB. Mol. Microbiol. 73:466-481.
88. Gunter, K. and V. Braun. 1990. In vivo evidence for FhuA outer
membrane receptor interaction with the TonB inner membrane protein of
Escherichia coli. FEBS Lett. 274:85-88.
89. Moeck, G. S., J. W. Coulton, and K. Postle. 1997. Cell envelope
signaling in Escherichia coli. Ligand binding to the ferrichrome-iron
receptor fhua promotes interaction with the energy-transducing protein
TonB. J. Biol. Chem. 272:28391-28397.
201
90. Skare, J. T., B. M. Ahmer, C. L. Seachord, R. P. Darveau, and K.
Postle. 1993. Energy transduction between membranes. TonB, a
cytoplasmic membrane protein, can be chemically cross-linked in vivo to
the outer membrane receptor FepA. J. Biol. Chem. 268:16302-16308.
91. Larsen, R. A., D. Foster-Hartnett, M. A. McIntosh, and K. Postle.
1997. Regions of Escherichia coli TonB and FepA proteins essential for in
vivo physical interactions. J. Bacteriol. 179:3213-3221.
92. Cadieux, N., C. Bradbeer, and R. J. Kadner. 2000. Sequence changes in
the ton box region of BtuB affect its transport activities and interaction
with TonB protein. J. Bacteriol. 182:5954-5961.
93. Cadieux, N. and R. J. Kadner. 1999. Site-directed disulfide bonding
reveals an interaction site between energy-coupling protein TonB and
BtuB, the outer membrane cobalamin transporter. Proc. Natl. Acad. Sci.
USA 96:10673-10678.
94. Khursigara, C. M., G. De Crescenzo, P. D. Pawelek, and J. W.
Coulton. 2005. Deletion of the proline-rich region of TonB disrupts
formation of a 2:1 complex with FhuA, an outer membrane receptor of
Escherichia coli. Protein Sci. 14:1266-1273.
95. Khursigara, C. M., G. De Crescenzo, P. D. Pawelek, and J. W.
Coulton. 2005. Kinetic analyses reveal multiple steps in forming TonB-
FhuA complexes from Escherichia coli. Biochemistry 44:3441-3453.
96. Pawelek, P. D., N. Croteau, C. Ng-Thow-Hing, C. M. Khursigara, N.
Moiseeva, M. Allaire, and J. W. Coulton. 2006. Structure of TonB in
complex with FhuA, E. coli outer membrane receptor. Science 312:1399-
1402.
97. Shultis, D. D., M. D. Purdy, C. N. Banchs, and M. C. Wiener. 2006.
Outer membrane active transport: structure of the BtuB:TonB complex.
Science 312:1396-1399.
98. James, K. J., M. A. Hancock, V. Moreau, F. Molina, and J. W.
Coulton. 2008. TonB induces conformational changes in surface-exposed
loops of FhuA, outer membrane receptor of Escherichia coli. Protein Sci.
17:1679-1688.
99. Higgs, P. I., T. E. Letain, K. K. Merriam, N. S. Burke, H. Park, C.
Kang, and K. Postle. 2002. TonB interacts with nonreceptor proteins in
the outer membrane of Escherichia coli. J. Bacteriol. 184:1640-1648.
100. Kaserer, W. A., X. Jiang, Q. Xiao, D. C. Scott, M. Bauler, D.
Copeland, S. M. Newton, and P. E. Klebba. 2008. Insight from TonB
202
hybrid proteins into the mechanism of iron transport through the outer
membrane. J. Bacteriol. 190:4001-4016.
101. Letain, T. E. and K. Postle. 1997. TonB protein appears to transduce
energy by shuttling between the cytoplasmic membrane and the outer
membrane in Escherichia coli. Mol. Microbiol. 24:271-283.
102. Larsen, R. A., T. E. Letain, and K. Postle. 2003. In vivo evidence of
TonB shuttling between the cytoplasmic and outer membrane in
Escherichia coli. Mol. Microbiol. 49:211-218.
103. Feilmeier, B. J., G. Iseminger, D. Schroeder, H. Webber, and G. J.
Phillips. 2000. Green fluorescent protein functions as a reporter for
protein localization in Escherichia coli. J. Bacteriol. 182:4068-4076.
104. Chimento, D. P., R. J. Kadner, and M. C. Wiener. 2005. Comparative
structural analysis of TonB-dependent outer membrane transporters:
implications for the transport cycle. Proteins 59:240-251.
105. Brockwell, D. J., E. Paci, R. C. Zinober, G. S. Beddard, P. D. Olmsted,
D. A. Smith, R. N. Perham, and S. E. Radford. 2003. Pulling geometry
defines the mechanical resistance of a beta-sheet protein. Nat. Struct. Biol.
10:731-737.
106. Gumbart, J., M. C. Wiener, and E. Tajkhorshid. 2007. Mechanics of
force propagation in TonB-dependent outer membrane transport. Biophys.
J. 93:496-504.
107. Chakraborty, R., E. Storey, and D. van der Helm. 2007. Molecular
mechanism of ferricsiderophore passage through the outer membrane
receptor proteins of Escherichia coli. Biometals 20:263-274.
108. Eisenhauer, H. A., S. Shames, P. D. Pawelek, and J. W. Coulton. 2005.
Siderophore transport through Escherichia coli outer membrane receptor
FhuA with disulfide-tethered cork and barrel domains. J. Biol. Chem.
280:30574-30580.
109. Devanathan, S. and K. Postle. 2007. Studies on colicin B translocation:
FepA is gated by TonB. Mol. Microbiol. 65:441-453.
110. Ma, L., W. Kaserer, R. Annamalai, D. C. Scott, B. Jin, X. Jiang, Q.
Xiao, H. Maymani, L. M. Massis, L. C. Ferreira, S. M. Newton, and P.
E. Klebba. 2007. Evidence of ball-and-chain transport of ferric
enterobactin through FepA. J. Biol. Chem. 282:397-406.
111. Clarke, T. E., S. Y. Ku, D. R. Dougan, H. J. Vogel, and L. W. Tari.
2000. The structure of the ferric siderophore binding protein FhuD
complexed with gallichrome. Nat. Struct. Biol. 7:287-291.
203
112. Clarke, T. E., V. Braun, G. Winkelmann, L. W. Tari, and H. J. Vogel.
2002. X-ray crystallographic structures of the Escherichia coli periplasmic
protein FhuD bound to hydroxamate-type siderophores and the antibiotic
albomycin. J. Biol. Chem. 277:13966-13972.
113. Rohrbach, M. R., V. Braun, and W. Köster. 1995. Ferrichrome
transport in Escherichia coli K-12: altered substrate specificity of mutated
periplasmic FhuD and interaction of FhuD with the integral membrane
protein FhuB. J. Bacteriol. 177:7186-7193.
114. Borths, E. L., K. P. Locher, A. T. Lee, and D. C. Rees. 2002. The
structure of Escherichia coli BtuF and binding to its cognate ATP binding
cassette transporter. Proc. Natl. Acad. Sci. U. S. A 99:16642-16647.
115. Cadieux, N., C. Bradbeer, E. Reeger-Schneider, W. Köster, A. K.
Mohanty, M. C. Wiener, and R. J. Kadner. 2002. Identification of the
periplasmic cobalamin-binding protein BtuF of Escherichia coli. J.
Bacteriol. 184:706-717.
116. Boos, W. and H. Shuman. 1998. Maltose/maltodextrin system of
Escherichia coli: transport, metabolism, and regulation. Microbiol. Mol.
Biol. Rev. 62:204-229.
117. Quiocho, F. A., J. C. Spurlino, and L. E. Rodseth. 1997. Extensive
features of tight oligosaccharide binding revealed in high-resolution
structures of the maltodextrin transport/chemosensory receptor. Structure.
5:997-1015.
118. Telmer, P. G. and B. H. Shilton. 2003. Insights into the conformational
equilibria of maltose-binding protein by analysis of high affinity mutants.
J. Biol. Chem. 278:34555-34567.
119. Krewulak, K. D., C. M. Shepherd, and H. J. Vogel. 2005. Molecular
dynamics simulations of the periplasmic ferric-hydroxamate binding
protein FhuD. Biometals 18:375-386.
120. Karpowich, N. K., H. H. Huang, P. C. Smith, and J. F. Hunt. 2003.
Crystal structures of the BtuF periplasmic-binding protein for vitamin B12
suggest a functionally important reduction in protein mobility upon ligand
binding. J. Biol. Chem. 278:8429-8434.
121. Liu, M., T. Sun, J. Hu, W. Chen, and C. Wang. 2008. Study on the
mechanism of the BtuF periplasmic-binding protein for vitamin B12.
Biophys. Chem. 135:19-24.
122. Kandt, C., Z. Xu, and D. P. Tieleman. 2006. Opening and closing
motions in the periplasmic vitamin B12 binding protein BtuF.
Biochemistry 45:13284-13292.
204
123. Shi, R., A. Proteau, J. Wagner, Q. Cui, E. O. Purisima, A. Matte, and
M. Cygler. 2009. Trapping open and closed forms of FitE: a group III
periplasmic binding protein. Proteins 75:598-609.
124. Procko, E., M. L. O'Mara, W. F. Bennett, D. P. Tieleman, and R.
Gaudet. 2009. The mechanism of ABC transporters: general lessons from
structural and functional studies of an antigenic peptide transporter.
FASEB J. 23:1287-1302.
125. Krewulak, K. D. and H. J. Vogel. 2008. Structural biology of bacterial
iron uptake. Biochim. Biophys. Acta 1778:1781-1804.
126. Locher, K. P., A. T. Lee, and D. C. Rees. 2002. The E. coli BtuCD
structure: a framework for ABC transporter architecture and mechanism.
Science 296:1091-1098.
127. Pinkett, H. W., A. T. Lee, P. Lum, K. P. Locher, and D. C. Rees. 2007.
An inward-facing conformation of a putative metal-chelate-type ABC
transporter. Science 315:373-377.
128. Hollenstein, K., D. C. Frei, and K. P. Locher. 2007. Structure of an
ABC transporter in complex with its binding protein. Nature 446:213-216.
129. Oldham, M. L., D. Khare, F. A. Quiocho, A. L. Davidson, and J. Chen.
2007. Crystal structure of a catalytic intermediate of the maltose
transporter. Nature 450:515-521.
130. Hvorup, R. N., B. A. Goetz, M. Niederer, K. Hollenstein, E. Perozo,
and K. P. Locher. 2007. Asymmetry in the structure of the ABC
transporter-binding protein complex BtuCD-BtuF. Science 317:1387-
1390.
131. Sebulsky, M. T., B. H. Shilton, C. D. Speziali, and D. E. Heinrichs.
2003. The role of FhuD2 in iron(III)-hydroxamate transport in
Staphylococcus aureus. Demonstration that FhuD2 binds iron(III)-
hydroxamates but with minimal conformational change and implication of
mutations on transport. J. Biol. Chem. 278:49890-49900.
132. Mademidis, A., H. Killmann, W. Kraas, I. Flechsler, G. Jung, and V.
Braun. 1997. ATP-dependent ferric hydroxamate transport system in
Escherichia coli: periplasmic FhuD interacts with a periplasmic and with a
transmembrane/cytoplasmic region of the integral membrane protein
FhuB, as revealed by competitive peptide mapping. Mol. Microbiol.
26:1109-1123.
133. Locher, K. P. 2009. Structure and mechanism of ATP-binding cassette
transporters. Philos. Trans. R. Soc. Lond B Biol. Sci. 364:239-245.
205
134. Greenwood, K. T. and R. K. Luke. 1978. Enzymatic hydrolysis of
enterochelin and its iron complex in Escherichia Coli K-12. Properties of
enterochelin esterase. Biochim. Biophys. Acta 525:209-218.
135. Matzanke, B. F., S. Anemuller, V. Schunemann, A. X. Trautwein, and
K. Hantke. 2004. FhuF, part of a siderophore-reductase system.
Biochemistry 43:1386-1392.
136. Expert, D., A. Boughammoura, and T. Franza. 2008. Siderophore-
controlled iron assimilation in the enterobacterium Erwinia chrysanthemi:
evidence for the involvement of bacterioferritin and the Suf iron-sulfur
cluster assembly machinery. J. Biol. Chem. 283:36564-36572.
137. Rodi, D. J., R. W. Janes, H. J. Sanganee, R. A. Holton, B. A. Wallace,
and L. Makowski. 1999. Screening of a library of phage-displayed
peptides identifies human bcl-2 as a taxol-binding protein. J. Mol. Biol.
285:197-203.
138. Rodi, D. J., G. E. Agoston, R. Manon, R. Lapcevich, S. J. Green, and
L. Makowski. 2001. Identification of small molecule binding sites within
proteins using phage display technology. Comb. Chem. High Throughput.
Screen. 4:553-572.
139. Burch, L. R., M. Scott, E. Pohler, D. Meek, and T. Hupp. 2004. Phage-
peptide display identifies the interferon-responsive, death-activated protein
kinase family as a novel modifier of MDM2 and p21WAF1. J. Mol. Biol.
337:115-128.
140. Rodi, D. J., A. S. Soares, and L. Makowski. 2002. Quantitative
assessment of peptide sequence diversity in M13 combinatorial peptide
phage display libraries. J. Mol. Biol. 322:1039-1052.
141. Piliarik, M., H. Vaisocherova, and J. Homola. 2009. Surface plasmon
resonance biosensing. Methods Mol. Biol. 503:65-88.
142. Philo, J. S. 2006. Is any measurement method optimal for all aggregate
sizes and types? AAPS. J. 8:E564-E571.
143. Schuck, P. 2000. Size-distribution analysis of macromolecules by
sedimentation velocity ultracentrifugation and Lamm equation modeling.
Biophys. J. 78:1606-1619.
144. Schuck, P. 1998. Sedimentation analysis of noninteracting and self-
associating solutes using numerical solutions to the Lamm equation.
Biophys. J. 75:1503-1512.
145. Wandersman, C. and P. Delepelaire. 2004. Bacterial iron sources: from
siderophores to hemophores. Annu. Rev. Microbiol. 58:611-647.
206
146. Braun, V. and M. Braun. 2002. Iron transport and signaling in
Escherichia coli. FEBS Lett. 529:78-85.
147. Braun, V., K. Hantke, and W. Köster. 1998. Bacterial Iron Transport:
Mechanisms, Genetics, and Regulation., p. 68-145. In Metal ions in
biological systems. Marcel Dekker Inc., New York.
148. Ferguson, A. D., J. W. Coulton, K. Diederichs, and W. Welte. 2001.
The ferric hydroxamate uptake receptor FhuA and related TonB-
dependent transporters in the outer membrane of gram-negative bacteria.,
p. 834-849. In Handbook of Metalloproteins. John Wiley & Sons Ltd.,
Chichester.
149. Moeck, G. S. and J. W. Coulton. 1998. TonB-dependent iron acquisition:
mechanisms of siderophore-mediated active transport. Mol. Microbiol.
28:675-681.
150. Locher, K. P., B. Rees, R. Koebnik, A. Mitschler, L. Moulinier, J. P.
Rosenbusch, and D. Moras. 1998. Transmembrane signaling across the
ligand-gated FhuA receptor: crystal structures of free and ferrichrome-
bound states reveal allosteric changes. Cell 95:771-778.
151. Faraldo-Gómez, J. D., G. R. Smith, and M. S. Sansom. 2003.
Molecular dynamics simulations of the bacterial outer membrane protein
FhuA: a comparative study of the ferrichrome-free and bound states.
Biophys. J. 85:1406-1420.
152. Postle, K. and R. J. Kadner. 2003. Touch and go: tying TonB to
transport. Mol. Microbiol. 49:869-882.
153. Kadner, R. J. 1990. Vitamin B12 transport in Escherichia coli: energy
coupling between membranes. Mol. Microbiol. 4:2027-2033.
154. Schöffler, H. and V. Braun. 1989. Transport across the outer membrane
of Escherichia coli K-12 via the FhuA receptor is regulated by the TonB
protein of the cytoplasmic membrane. Mol. Gen. Genet. 217:378-383.
155. Ogierman, M. and V. Braun. 2003. Interactions between the outer
membrane ferric citrate transporter FecA and TonB: studies of the FecA
TonB box. J. Bacteriol. 185:1870-1885.
156. Merianos, H. J., N. Cadieux, C. H. Lin, R. J. Kadner, and D. S. Cafiso.
2000. Substrate-induced exposure of an energy-coupling motif of a
membrane transporter. Nat. Struct. Biol. 7:205-209.
157. Cadieux, N., P. G. Phan, D. S. Cafiso, and R. J. Kadner. 2003.
Differential substrate-induced signaling through the TonB-dependent
transporter BtuB. Proc. Natl. Acad. Sci. USA 100:10688-10693.
207
158. Braun, M., H. Killmann, and V. Braun. 1999. The -barrel domain of
FhuA5-160 is sufficient for TonB-dependent FhuA activities of
Escherichia coli. Mol. Microbiol. 33:1037-1049.
159. Scott, D. C., Z. Cao, Z. Qi, M. Bauler, J. D. Igo, S. M. Newton, and P.
E. Klebba. 2001. Exchangeability of N termini in the ligand-gated porins
of Escherichia coli. J. Biol. Chem. 276:13025-13033.
160. Howard, S. P., C. Herrmann, C. W. Stratilo, and V. Braun. 2001. In
vivo synthesis of the periplasmic domain of TonB inhibits transport
through the FecA and FhuA iron siderophore transporters of Escherichia
coli. J. Bacteriol. 183:5885-5895.
161. Killmann, H., C. Herrmann, A. Torun, G. Jung, and V. Braun. 2002.
TonB of Escherichia coli activates FhuA through interaction with the -
barrel. Microbiology 148:3497-3509.
162. Vakharia, H. L. and K. Postle. 2002. FepA with globular domain
deletions lacks activity. J. Bacteriol. 184:5508-5512.
163. Braun, M., F. Endriss, H. Killmann, and V. Braun. 2003. In vivo
reconstitution of the FhuA transport protein of Escherichia coli K-12. J.
Bacteriol. 185:5508-5518.
164. Ghosh, J. and K. Postle. 2004. Evidence for dynamic clustering of
carboxy-terminal aromatic amino acids in TonB-dependent energy
transduction. Mol. Microbiol. 51:203-213.
165. Zwick, M. B., L. L. Bonnycastle, K. A. Noren, S. Venturini, E. Leong,
C. F. Barbas, III, C. J. Noren, and J. K. Scott. 1998. The maltose-
binding protein as a scaffold for monovalent display of peptides derived
from phage libraries. Anal. Biochem. 264:87-97.
166. Moeck, G. S. and L. Letellier. 2001. Characterization of in vitro
interactions between a truncated TonB protein from Escherichia coli and
the outer membrane receptors FhuA and FepA. J. Bacteriol. 183:2755-
2764.
167. Moeck, G. S., P. Tawa, H. Xiang, A. A. Ismail, J. L. Turnbull, and J.
W. Coulton. 1996. Ligand-induced conformational change in the
ferrichrome-iron receptor of Escherichia coli K-12. Mol. Microbiol.
22:459-471.
168. Mandava, S., L. Makowski, S. Devarapalli, J. Uzubell, and D. J. Rodi.
2004. RELIC--a bioinformatics server for combinatorial peptide analysis
and identification of protein-ligand interaction sites. Proteomics. 4:1439-
1460.
208
169. Makowski, L. and A. Soares. 2003. Estimating the diversity of peptide
populations from limited sequence data. Bioinformatics. 19:483-489.
170. Delano W.L. 2008. The Pymol Molecular Graphics System, DeLano
Scientific LLC, Palo Alto, CA, USA.
171. Morris, G. M., D. S. Goodsell, R. S. Halliday, R. Huey, W. E. Hart, R.
K. Belew, and A. J. Olson. 1998. Automated docking using a Lamarckian
genetic algorithm and an empirical binding free energy function. J.
Comput. Chem. 19:1639-1662.
172. Barnard, T. J., M. E. Watson, Jr., and M. A. McIntosh. 2001.
Mutations in the Escherichia coli receptor FepA reveal residues involved
in ligand binding and transport. Mol. Microbiol. 41:527-536.
173. Chakraborty, R., E. A. Lemke, Z. Cao, P. E. Klebba, and D. van der
Helm. 2003. Identification and mutational studies of conserved amino
acids in the outer membrane receptor protein, FepA, which affect transport
but not binding of ferric-enterobactin in Escherichia coli. Biometals
16:507-518.
174. Köster, W. and V. Braun. 1990. Iron (III) hydroxamate transport into
Escherichia coli. Substrate binding to the periplasmic FhuD protein. J.
Biol. Chem. 265:21407-21410.
175. Schultz-Hauser, G., W. Köster, H. Schwarz, and V. Braun. 1992.
Iron(III) hydroxamate transport in Escherichia coli K-12: FhuB-mediated
membrane association of the FhuC protein and negative complementation
of fhuC mutants. J. Bacteriol. 174:2305-2311.
176. Lee, Y. H., R. K. Deka, M. V. Norgard, J. D. Radolf, and C. A.
Hasemann. 1999. Treponema pallidum TroA is a periplasmic zinc-
binding protein with a helical backbone. Nat. Struct. Biol. 6:628-633.
177. Lawrence, M. C., P. A. Pilling, V. C. Epa, A. M. Berry, A. D.
Ogunniyi, and J. C. Paton. 1998. The crystal structure of pneumococcal
surface antigen PsaA reveals a metal-binding site and a novel structure for
a putative ABC-type binding protein. Structure. 6:1553-1561.
178. Krewulak, K. D., R. S. Peacock, and H. J. Vogel. 2004. Periplasmic
Binding Proteins Involved in Bacterial Iron Uptake, p. 113-129. In Iron
Transport in Bacteria. ASM Press, Washington DC.
179. Rohrback, M. R., S. Paul, and W. Köster. 1995. In vivo reconstitution of
an active siderophore transport system by a binding protein derivative
lacking a signal sequence. Mol. Gen. Genet. 248:33-42.
209
180. Carter, D. M., J. N. Gagnon, M. Damlaj, S. Mandava, L. Makowski,
D. J. Rodi, P. D. Pawelek, and J. W. Coulton. 2006. Phage display
reveals multiple contact sites between FhuA, an outer membrane receptor
of Escherichia coli, and TonB. J. Mol. Biol. 357:236-251.
181. Yang, Y. R. and H. K. Schachman. 1996. A bifunctional fusion protein
containing the maltose-binding polypeptide and the catalytic chain of
aspartate transcarbamoylase: assembly, oligomers, and domains. Biophys.
Chem. 59:289-297.
182. Lebowitz, J., M. S. Lewis, and P. Schuck. 2002. Modern analytical
ultracentrifugation in protein science: a tutorial review. Protein Sci.
11:2067-2079.
183. Myszka, D. G. 1999. Improving biosensor analysis. J. Mol. Recognit.
12:279-284.
184. Lamb, A. L., A. S. Torres, T. V. O'Halloran, and A. C. Rosenzweig.
2000. Heterodimer formation between superoxide dismutase and its
copper chaperone. Biochemistry 39:14720-14727.
185. Mader, C., C. Huber, D. Moll, U. B. Sleytr, and M. Sára. 2004.
Interaction of the crystalline bacterial cell surface layer protein SbsB and
the secondary cell wall polymer of Geobacillus stearothermophilus PV72
assessed by real-time surface plasmon resonance biosensor technology. J.
Bacteriol. 186:1758-1768.
186. Nakajima, H., L. Cocquerel, N. Kiyokawa, J. Fujimoto, and S. Levy.
2005. Kinetics of HCV envelope proteins' interaction with CD81 large
extracellular loop. Biochem. Biophys. Res. Commun. 328:1091-1100.
187. Sprencel, C., Z. Cao, Z. Qi, D. C. Scott, M. A. Montague, N. Ivanoff, J.
Xu, K. M. Raymond, S. M. Newton, and P. E. Klebba. 2000. Binding of
ferric enterobactin by the Escherichia coli periplasmic protein FepB. J.
Bacteriol. 182:5359-5364.
188. Borths, E. L., B. Poolman, R. N. Hvorup, K. P. Locher, and D. C.
Rees. 2005. In vitro functional characterization of BtuCD-F, the
Escherichia coli ABC transporter for vitamin B12 uptake. Biochemistry
44:16301-16309.
189. Carter, D. M., I. R. Miousse, J. N. Gagnon, E. Martinez, A. Clements,
J. Lee, M. A. Hancock, H. Gagnon, P. D. Pawelek, and J. W. Coulton.
2006. Interactions between TonB from Escherichia coli and the
periplasmic protein FhuD. J. Biol. Chem. 281:35413-35424.
210
190. Karlsson, R., P. S. Katsamba, H. Nordin, E. Pol, and D. G. Myszka.
2006. Analyzing a kinetic titration series using affinity biosensors. Anal.
Biochem. 349:136-147.
191. Morris, G. M., R. Huey, W. Lindstrom, M. F. Sanner, R. K. Belew, D.
S. Goodsell, and A. J. Olson. 2009. AutoDock4 and AutoDockTools4:
Automated docking with selective receptor flexibility. J. Comput. Chem.
Epub ahead of print
192. Pettersen, E. F., T. D. Goddard, C. C. Huang, G. S. Couch, D. M.
Greenblatt, E. C. Meng, and T. E. Ferrin. 2004. UCSF Chimera--a
visualization system for exploratory research and analysis. J. Comput.
Chem. 25:1605-1612.
193. Ködding, J. F. 2004. Biochemical and X-ray crystallographic studies on
the energy transducing protein TonB and the TonB-dependent siderophore
receptor FhuA from Escherichia coli. Ph.D. thesis. University of
Konstanz.
194. Dyson, H. J. and P. E. Wright. 2005. Intrinsically unstructured proteins
and their functions. Nat. Rev. Mol. Cell Biol. 6:197-208.
195. Blommel, P. G. and B. G. Fox. 2007. A combined approach to improving
large-scale production of tobacco etch virus protease. Protein Expr. Purif.
55:53-68.
196. Lapko, A., A. Müller, O. Heese, K. Ruckpaul, and U. Heinemann.
1997. Preparation and crystallization of a cross-linked complex of bovine
adrenodoxin and adrenodoxin reductase. Proteins 28:289-292.
197. Leach, M. R., J. W. Zhang, and D. B. Zamble. 2007. The role of
complex formation between the Escherichia coli hydrogenase accessory
factors HypB and SlyD. J. Biol. Chem. 282:16177-16186.