University of Southern California Dissertations and Theses

Bacterial nanowires of Shewanella oneidensis MR-1: electron transport mechanism, composition, and role of multiheme cytochromes

(USC Thesis Other)

Bacterial nanowires of Shewanella oneidensis MR-1: electron transport mechanism, composition, and role of multiheme cytochromes

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Bacterial nanowires of Shewanella oneidensis MR-1: electron transport
mechanism, composition, and role of multiheme cytochromes
by
Sahand Pirbadian
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulllment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PHYSICS)
May 2015
Copyright 2015 Sahand Pirbadian
To my lovely wife, Pouye
ii
Acknowledgments
I'd like to use this opportunity to thank my advisor and role model, Moh El-
Naggar, who is both a wonderful person and a great mentor. Moh has inspired
me in so many ways during my four years in his group and for that I will be
forever in his debt. He has introduced me to an exciting subject, and has given me
encouranegment to explore in the lab. He has always been extremely supportive,
patient and positive which has especially helped me when things didn't go as
planned. Moh, I don't think I can thank you enough! I wish long and healthy lives
full of joy for you and your family.
I would like to thank all my labmates, I could never wish for a better group of
people to work with and be around: Tom Yuzvinsky, Edmond (Kar Man) Leung,
Matt (Shuai) Xu, Ben Gross, Ian McFarlane, Yamini Jangir and Hyesuk Byun.
Thanks for always being helpful, supportive and cheerful!
I truly appreciate all the help and guidance that I received from Sarah Barchinger
and John Golbeck during our work together. I hope we can continue our collabo-
rations for years to come.
My special gratitude goes to the USC Center of Excellence in NanoBiophysics
and the USC Center for Electron Microscopy and Microanalysis (CEMMA), and
especially Paul Webster, formerly at USC and currently at Oak Crest Institute of
Science, for preparation of TEM samples and TEM imaging.
iii
Also, I would like to thank my collaborators Liang Shi, Rachida Bouhenni,
Daad Saarini, Bree Reed, Margie Romine, and Yuri Gorby for their help in the
nanowire characterization project reported in this thesis.
I'm grateful to my parents, Parisima Parsaei and Ali Pirbadian, for their uncon-
ditional love and support and for introducing me to science and the wonderful joy
in learning since my early childhood.
And most importantly, I thank my best friend, my life companion, my beautiful
wife Pouye. Meeting you was the best thing that has ever happened to me and
you have lled every day of my life with joy since that day. During my years in
grad school, you have always been supportive, understanding and helpful. Words
cannot express how grateful I am to have you as my wife and I look forward to
spending the rest of my life by your side.
Last but not least, I would like to acknowledge the funding agencies that helped
make this thesis happen: the Air force Oce of Scientic Research (Young Inves-
tigator Research Program Grant FA9550-10-1-0144 to Moh El-Naggar) and the
US Department of Energy (Grant DE-FG02-13ER16415 to Moh El-Naggar). I am
also grateful to the graduate school at USC for their generous Provost Fellowship
during my last two years in grad school.
iv
Contents
Dedication ii
Acknowledgments iii
Abstract vii
1 Introduction 1
2 Background 17
2.1 Energetics of life and oxidative phosphorylation . . . . . . . . . . . 17
2.2 Anaerobic respiration and dissimilatory metal reduction . . . . . . . 19
2.3 Outer membrane cytochromes . . . . . . . . . . . . . . . . . . . . . 21
2.4 MtrF and its crystal structure . . . . . . . . . . . . . . . . . . . . . 21
2.5 Atomic force and scanning tunneling microscopy . . . . . . . . . . . 25
2.6 Pathways of extracellular electron transport (EET) . . . . . . . . . 25
2.7 Bacterial nanowires and their conductive properties . . . . . . . . . 27
2.8 The theory of non-adiabatic electron transfer . . . . . . . . . . . . . 29
2.9 Fluorescence Microscopy . . . . . . . . . . . . . . . . . . . . . . . . 32
3 Multistep Hopping Model of Electron Transport in Bacterial Redox
Chains 35
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2 Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.2 Multistep Hopping . . . . . . . . . . . . . . . . . . . . . . . 40
3.3 Results and comparison to experiments . . . . . . . . . . . . . . . . 43
3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4 In vivo Imaging of Bacterial Nanowires of Shewanella oneidensis
MR-1 and Their Composition and Structure 53
4.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.2 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.2.1 Cell growth conditions . . . . . . . . . . . . . . . . . . . . . 54
v
4.2.2 Perfusion chamber platform . . . . . . . . . . . . . . . . . . 55
4.2.3 RedoxSensor Green assay . . . . . . . . . . . . . . . . . . . 62
4.2.4 Verication of RedoxSensor Green as a probe for active res-
piration in Shewanella . . . . . . . . . . . . . . . . . . . . . 63
4.2.5 Cytoplasmic and periplasmic green
uorescent protein (GFP)
imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.2.6 Chemostat growth and qPCR analysis of the transition from
electron-donor to electron-acceptor limitation . . . . . . . . 68
4.2.7 Strains and plasmids generated for this study . . . . . . . . 71
4.3 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.3.1 In vivo imaging of nanowire formation . . . . . . . . . . . . 72
4.3.2 The production of S. oneidensis MR-1 bacterial nanowires
is correlated with an increase in cellular reductase activity . 75
4.3.3 S.oneidensis MR-1 nanowires are outer membrane and periplas-
mic extensions . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
5 Cytochrome Localization and Nanoscale Characterization of Bac-
terial Nanowires 83
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
5.2 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.2.1 Immuno
uorescence with MtrC or OmcA antibody . . . . . 85
5.2.2 Preparation of MtrF monolayers on Au(111) . . . . . . . . . 86
5.2.3 Scanning tunneling microscopy of MtrF monolayers . . . . . 87
5.2.4 Atomic force microscopy following live perfusion
ow imaging 88
5.2.5 Transmission electron microscopy following nanowire forma-
tion in the in vivo imaging platform . . . . . . . . . . . . . . 89
5.3 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.3.1 Localization of the decaheme cytochromes MtrC and OmcA
along nanowires . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.3.2 Proof of concept for single-molecule electron transfer mea-
surements in MtrF . . . . . . . . . . . . . . . . . . . . . . . 93
5.3.3 Nanoscale characterization of the intermediate steps in nanowire
formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.4 Energy cost calculations of nanowire production . . . . . . . . . . . 99
5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6 Conclusion 103
Bibliography 108
vi
Abstract
In this thesis, we discuss three topics concerning extracellular electron transfer in
the Dissimilatory Metal Reducing Bacterium (DMRB) Shewanella oneidensis MR-
1. One proposed strategy to accomplish extracellular charge transfer in Shewanella
involves forming a conductive pathway to electrodes by incorporating redox com-
ponents on outer cell membranes and along extracellular appendages known as
bacterial nanowires within biolms. In the rst part of this dissertation, to describe
extracellular charge transfer in microbial redox chains, we employed a model based
on incoherent hopping between sites in the chain and an interfacial treatment of
electrochemical interactions with the surrounding electrodes. Based on this model,
we calculated the current-voltage (I-V) characteristics and found the results to be
in good agreement with I-V measurements across and along individual microbial
nanowires produced by the bacterium S. oneidensis MR-1. Based on our analysis,
we propose that multistep hopping in redox chains constitutes a viable strategy
for extracellular charge transfer in microbial biolms.
In the second part, we report the rst in vivo observations of the formation
and respiratory impact of nanowires in the model metal-reducing microbe S. onei-
densis MR-1. Live
uorescence measurements, immunolabeling, and quantitative
gene expression analysis point toS.oneidensis MR-1 nanowires as extensions of the
vii
outer membrane and periplasm that include the multiheme cytochromes respon-
sible for EET, rather than pilin-based structures as previously thought. These
membrane extensions are associated with outer membrane vesicles, structures ubiq-
uitous in Gram-negative bacteria, and are consistent with bacterial nanowires that
mediate long-range EET by our proposed multistep redox hopping mechanism.
Redox-functionalized membrane and vesicular extensions may represent a general
microbial strategy for electron transport and energy distribution.
In addition, to elucidate the membranous nature of Shewanella nanowires,
we imaged these laments using Transmission Electron Microscopy. The TEM
images reported in this thesis also provide the most accurate estimates of bacterial
nanowire dimensions to date. Future TEM and cryo-TEM imaging can estab-
lish the specic alignment and conguration of outer membrane cytochromes that
facilitate electron transport along bacterial nanowires.
In the third part of this thesis, we focus on the molecular conductance of MtrF,
the rst decaheme outer membrane cytochrome with a solved crystal structure.
Decaheme outer membrane cytochromes of Shewanella play a crucial role in all
the suggested pathways of extracellular electron transfer. An understanding of the
electron transfer properties in MtrF will therefore impact all aspects of extracellular
electron transfer research. In this thesis, using puried MtrF, we form monolayers
of the protein on atomically
at gold substrates and address the dry monolayer
with a Scanning Tunneling Microscope (STM) tip. This technique can be used in
the future to examine the conductivity of individual MtrF molecules within the
monolayer in the form of I-V curves. This methodology will allow experimental
comparison with recently developed simulations of MtrF conductance.
viii
Chapter 1
Introduction
Electron transfer plays an important role in our everyday lives, most visibly in all
solid state devices that we regularly use. But what may not be clear to many is
that electron transfer is one of the most fundamental and crucial phenomena in
sustaining life [1, 2]. In fact it is the
ow of electrons that drives the energetics
of life. In respiratory organisms, higher energy electrons start from an electron
donor and move along several steps into and through a living cell and reach their
destination, an electron acceptor. But what does life gain from escorting electrons
from one molecule to another? The same way energy is harvested from water
in a turbine in a dam,
owing electrons through living cells can give energy to
organisms to sustain their lively functions [3].
In respiration, the electron
ow in a cell is channeled through a series of
molecules in the cellular inner membrane (or in the case of eukaryotes the mito-
chondrial inner membrane). A proton gradient is then created across the inner
membrane, and is used to generate Adenosine Triphosphate (ATP), the energy
currency of life [3]. ATP can later be used for various functions including biosyn-
thesis and transportation of molecules in and out of cells. Although alternative
strategies such as substrate level phosphorylation (SLP) can be the primary source
of ATP production in some organisms under certain conditions (including in She-
wanella [4]), electron transfer to electron acceptors must still continue to ensure
that redox reactions in the cell are balanced.
1
While most organisms use some form of electron transport from an electron
donor to an electron acceptor to store energy in the form of ATP, the specics of the
pathways for this process and the electron donors/acceptors used vary in nature.
Aerobic organisms transfer the electrons generated by oxidizing organic molecules
(their electron donor) to molecular oxygen available in soluble form inside the
cell. Therefore aerobic organisms including humans require molecular oxygen to
produce ATP and survive. Anaerobic organisms, on the other hand, are able to
use alternative electron acceptors other than molecular oxygen to sustain the
ow
of electrons. Examples of anaerobic organisms are the bacteria that thrived on
the Earth before the atmosphere had enough molecular oxygen to sustain aerobic
life. In fact species of microbes exist today that have the same ability, i.e. to
survive under anaerobic conditions. But the question is what do these microbes
use instead of molecular oxygen as their electron acceptor?
Long before accumulation of molecular oxygen in the Earth atmosphere, iron
and other metals would have served as terminal electron acceptors for microorgan-
isms [5]. But to be able to reduce insoluble inorganic minerals such as iron oxide
based rocks, a cell requires the machinery to transport its respiratory electrons
outside the cellular outer membrane and perform extracellular electron transfer
(EET). Dissimilatory metal-reducing bacteria (DMRBs) are the group of microbes
existing on Earth today that have the capability of EET [6, 7].
The rst DMRBs, Geobacter and Shewanella (Fig. 1.1), were discovered in the
late 1980s [8{10], thus beginning the search into the mechanisms of extracellular
electron transfer and understanding the long-range electron transport required to
move electrons to cellular exterior for reduction of solid phase metal oxides. It
was later discovered that multi-heme cytochromes (such as the one in Fig. 1.2)
play a crucial role in enabling this long-range transport [11{14]. These proteins
2
Figure 1.1: Image of Shewanella oneidensis MR-1 cells on Fe
2
O
3
. Image credit:
Alice Dohnalkova, Pacic Northwest National Lab, Richland, WA.
incorporate multiple electron carrier heme sites within their structure (Fig. 1.2).
In fact periplasmic and outer membrane cytochromes form a complex along which
electrons can move and cross the cellular membrane [14, 15]. Due to the central
role of these multiheme cytochromes in extracellular electron transport, they have
been the subject of numerous studies in the past two decades [12, 14{19].
Extracellular electron transfer has gathered a large amount of interest in the
past few decades due to its applications in renewable energy conversion [20, 21],
bioremediation [22, 23], and waste water treatment [24]. DMRBs are not only
capable of reducing metal oxide rocks in the environment, but they are able to
transfer electrons to synthetic solid state electrodes [25]. By providing enough
3
Figure 1.2: Crystal structure of MtrF, a decaheme outer membrane cytochrome
of S. oneidensis MR-1. The hemes in this structure are positioned according to
a `staggered-cross' pattern. The small (<1nm) distance between adjacent hemes
enables long range electron transfer in this outer membrane cytochrome [17].
soluble electron donor and tuning the redox potential of the electrode, the
ow
of respiratory electrons and therefore the cellular metabolism and growth can be
controlled. As a result, a steady
ow of electrons can be extracted and converted to
electricity from a community of cells in microbial fuel cells (MFCs) [20] (Fig. 1.3).
Another application of EET is microbial electrosynthesis in which bacteria perform
the above process in reverse: they gain electrons generated by renewable energy
sources such as solar cells to reduce soluble chemicals, resulting in the synthesis of
biofuels [26].
DMRBs are also used for bioremediation by reducing soluble toxic chemicals
that become insoluble upon reduction. Examples of such toxins are chromium or
4
Figure 1.3: A) Photo and B) schematic of a microbial fuel cell (MFC). MFCs
are are able to convert renewable energy from organic matter (or possibly organic
contaminants) using dissimilatory metal-reducing bacteria such as Shewanella and
Geobacter. In this dual-compartment MFC, bacteria are introduced in the anode
compartment, they attach to the anode and transfer their respiratory electrons
to it. These electrons are transferred to the cathode compartment, where oxygen
is reduced to produce water. The separating membrane between the two com-
partments allows transfer of protons from the anode to the cathode compartment
which completes the cell reaction. From [20].
uranium, soluble contaminants in water which become insoluble upon reduction
by bacteria and therefore can be easily removed from water [22]. In addition, since
waste water has a high organic content, DMRBs can be used for its treatment and
for simultaeous production of electricity [24].
Research on extracellular electron transport may also have implications for
astrobiology and the search for extraterrestrial life. Some of the minerals that
MR-1 is known to use as electron acceptor are very abundant on other planets,
e.g. iron oxides on the surface of Mars [27]. Thus a better understanding of the
mechanisms of extracellular electron transport in DMRBs may familiarize us with
possible extraterrestrial metabolisms on other planets including Mars.
The amazing level of sophistication in the design and manufacturing of the
semiconductor devices we use every day is partially due to our deep quantitative
5
understanding of movements and
ows of electrons in these devices. Similarly,
quantitative knowledge of electron transfer processes in biology is essential in our
ability to enhance and tap the energy conversion potential of living cells. The
non-adiabatic electron transfer theory has typically been the starting point for
any physical model of biological electron transfer in the past half century [28].
This semi-classical theory describes how electrons are transferred from a donor
site to an acceptor site in solution and was later modied to also include quantum
mechanical eects [2, 29]. This theory has been tested for systems in which the
distance between the donor and acceptor is xed, resulting in timetables for rates
of electron transfer in various environments.
However, until about two decades ago investigation of biological electron trans-
fer was limited to ET reactions occurring at nanometer length scales. The recent
emergence of EET measurements in DMRBs has indicated the possibility of elec-
tron transfer phenomena in microns-long structures [30{33]. Therefore, the addi-
tional complexity of these larger scale systems calls for more intricate quantitative
models that are still based on the fundamental theory of electron transfer, i.e.
Marcus theory. For instance, rather than having a single step electron hopping
event between a donor and an acceptor site, one may need to account for tens or
hundreds of hopping events to model a single EET process [34].
Understanding the ET properties of individual multiheme cytochromes can
serve as an important stepping-stone to the study of EET as a whole. For this
reason, there has been a large amount of interest in physical modeling of EET in
DMRBs in recent years [34{36]. Outer membrane cytochromes are the subject of
many of these studies, especially after the publication of the rst solved crystal
structure of an outer membrane cytochrome [17, 19, 37{40].
6
Figure 1.4: Pathways of extracellular electron transfer in S. oneidensis MR-1.
The direct pathways include electron transfer via cell surface contact and through
bacterial nanowires, whereas electron shuttling by
avins represents the indirect
pathway.
Localization of cytochromes in the outer membrane of DMRBs facilitates direct
electron transfer from the cellular inner membrane to the solid phase minerals
directly in contact with the cell exterior. However, DMRBs have also devised
strategies to reach insoluble electron acceptors located microns away from their
surface. One method, specically in the case of Shewanella, is to use electron
shuttles, molecules that are reduced at the cell surface by the outer membrane
cytochromes, diuse in solution to reach the terminal electron acceptor and get
oxidized at the metal oxide surface [41{43]. This process is suggested to repeat
7
Figure 1.5: Scanning electron microscopy image of bacterial nanowires from She-
wanella oneidensis MR-1. Image credit: Kar Man Leung, USC.
in cycles, transferring electrons at a fast enough rate to support the respiration
rate of cells. However, for this mechanism to be feasible there needs to be a large
concentration of shuttle molecules and relatively fast diusion (i.e. large diusion
coecient) in solution.
Bacterial nanowires represent another EET pathway (Fig. 1.5), in addition to
cell surface contact and electron shuttling. The discovery of bacterial nanowires
has opened a new chapter in extracellular electron transfer research [30, 31]. These
laments were rst observed in Geobacter sulfurreducens and Shewanella oneiden-
sis MR-1 in 2005 and 2006 respectively. Bacterial nanowires of both organisms
were shown to be conductive across their width under non-physiological conditions
[30, 31]. In addition, Shewanella nanowires (Fig. 1.5) were shown to conduct elec-
trons at a rate of 10
9
e/s at 100 mV applied potential over a length of up to 2m
8
(Shewanella cells are about 2m in length), which is extremely long compared to
typical biological electron transfer length scales, i.e. nanometers [32].
It was shown early on that Geobacter nanowires are type IV pili [30], i.e.
protein laments that are typically used for attachment and twitching motility in
bacteria. Deletion of the genes encoding the pilin subunits of these pili produced
strains that attached normally to iron oxide surfaces but were not able to reduce
these minerals [30]. Therefore it was concluded that the role of Geobacter pili in
reducing minerals is not in the form of cell attachment as previously believed, but
rather pili are transferring electrons to the oxide surface. However, the true identity
and composition of Shewanella nanowires was not clear from the beginning. The
most widely accepted hypothesis was that Shewanella nanowires are also type IV
pili [44].
The mechanism of electron transport in nanowires ofGeobacter andShewanella
has been a topic of controversy since their discovery [34{36, 44{46]. It has been
proposed that Geobacter nanowires conduct electrons through `metallic-like' con-
ductivity that is facilitated by overlapping of -orbitals of the aromatic amino
acids within the pilus structure [47]. On the other hand, Shewanella nanowires
are proposed to incorporate outer membrane cytochromes along their length that
facilitate the
ow of respiratory electrons from the cell to the terminal electron
acceptor by allowing hopping of electrons between adjacent hemes within each
cytochrome and between neighboring cytochromes [32, 34{36]. This multistep
hopping hypothesis was supported by the fact that the Shewanella mutant lack-
ing genes encoding the outer membrane cytochromes MtrC and OmcA produced
non-conductive nanowires [32].
One of the main goals of this dissertation is to present a quantitative evalua-
tion of the two proposed ET mechanisms in bacterial nanowires. In chapter 3, the
9
experimental electron mobility values in bacterial nanowires are compared with the
expected mobility in either of the proposed mechanisms, i.e. `metallic-like' con-
ductivity and multistep hopping. Based on this analysis and considering previous
measurements indicating the importance of cytochromes in nanowire conductivity,
we nd that multistep hopping seems far more likely to be responsible for the
observed ET rates in bacterial nanowires of Shewanella.
Also in chapter 3, we describe the multistep hopping model which is based on
the proposed involvement of outer membrane cytochromes in Shewanella nanowire
conductivity. In this model a chain of redox sites, e.g. hemes, is suggested to facili-
tate long-range electron transport by hopping between adjacent redox centers. We
have combined the Marcus theory of ET to predict the rates of the non-adiabatic
ET between the redox sites and the electrochemical form of Marcus theory to
model the interfacial ET rates between the redox sites and the experimental elec-
trodes addressing the nanowires. A master equation that includes all these rates
is then presented and solved to give the predicted current-voltage behavior of the
nanowire system. In the end, we nd the predicted current-voltage curves in good
agreement with previous experimental data.
Other than the mechanism of electron transport, fundamental questions about
bacterial nanowires still remain unanswered. Specically, the composition and
structure of Shewanella nanowires is unknown. These laments are generally
assumed to be type IV pili [44], perhaps due to their similar proposed function
as Geobacter pili, i.e. transporting electrons. However, it is important to note
that the composition of Shewanella nanowires had never been directly studied.
Researchers have attempted to evaluate the impact of deletion of type IV pili
genes in electricity production of Shewanella cultures in microbial fuel cells and
similar large scale systems [20]. These experiments however fail to isolate the
10
impact of type IV pili in cell attachment and biolm formation from their role in
electron transfer between cells and electrodes as possible structural components of
bacterial nanowires.
There are also other important questions regarding bacterial nanowires: What
is the physiological impact of nanowire production on cells? In particular, how does
producing nanowires enhance a cells ability to survive under electron acceptor-
limited conditions? What is its eect on cellular respiration?
One common weakness in all previous measurements on nanowire conductivity
is that they were performed under non-physiological conditions, i.e. samples are
typically xed with aldehydes and dried while on a surface [30{32, 47]. Both the
xation and the dehydration steps could potentially alter the nanowire structure,
the alignment and conguration of the OM cytochromes, and therefore change
the conductive properties of the nanowire under study compared to its native,
hydrated state. This lack of in vivo work has been due to the challenges that arise
monitoring and sustaining growth of nanowires in solution. However, in this thesis,
we nd that in vivo imaging of bacterial nanowires is in fact the common basis in
answering all of the questions posed above.
In this thesis, we introduce a new experimental approach to investigate She-
wanella nanowires. In chapter 4 of this thesis, we report, for the rst time, the
development of a technique for imaging bacterial nanowires in vivo. We used

uorescence microscopy to image Shewanella cells that were maintained inside a
microliter-volume perfusion chamber. Monitoring growth of bacterial nanowires
using this method presented a signicant challenge because all previous studies on
Shewanella nanowires used bioreactors with controlled concentration of oxygen to
produce the samples [31, 32]. However, in the new technique, nanowire producing
conditions were obtained inside the perfusion chamber that enabled us to watch
11
the live growth of these laments. The specics of the developed experimental
platform and methods used are explained in chapter 4.
The in vivo imaging platform also allowed us to examine the physiological
impact of nanowire production on cells using a
uorescent indicator of cellular
reductase activity [48, 49]. As nanowires were produced, the
uorescence intensity
of this respiratory indicator increased, showing elevated respiration rates correlat-
ing with nanowire production.
To address the composition of Shewanella nanowires, protein-specic and lipid-
specic
uorescent stains were used to label proteins and lipids on bacterial
nanowires in solution. To our surprise, the lipid stain completely stained all bac-
terial nanowires, indicating the presence of lipid in these laments. To further elu-
cidate the structure, we localized
uorescent proteins inside the cellular periplasm
which also localized inside Shewanella nanowires. This observation demonstrated
that Shewanella nanowires are in fact extensions of the cellular outer membrane.
In vivo imaging of nanowires can also be used to investigate the possible mech-
anism of electron transport in nanowires. As proposed in the multistep hopping
model and also based on previous conductivity measurements, the presence of
outer membrane cytochromes facilitates the long range electron transport along a
nanowire [32, 34]. In this thesis, using the in vivo imaging platform in combina-
tion with immuno
uorescence microscopy we demonstrate the presence of MtrC
and OmcA, the main OM cytochromes of Shewanella, along the laments. In
chapter 5, we describe how nanowires imaged in vivo during their growth can later
be stained with antibodies raised against the specic OM cytochromes, allowing
the visualization of these proteins along the same nanowires.
Given the presence of OM cytochromes along bacterial nanowires, it is bene-
cial to examine the electron transfer properties of individual OM cytochromes.
12
This information can serve as the stepping stone for building a large scale physical
model that accurately describes the long-range electron transport characteristics
of the entire nanowire system. MtrF, the cytochrome studied in this thesis, was
chosen because it is the rst outer membrane cytochrome with a published crystal
structure [17]. In chapter 5, we report the development of an approach for investi-
gation of the electron transfer properties of MtrF at the single-molecule level that is
based on scanning tunneling microscopy (STM). Recombinant MtrF molecules are
adsorbed on an atomically
at gold (Au(111)) substrate to form a uniform mono-
layer. Individual proteins are then approached by an STM tip, while a potential is
applied between the tip and the gold substrate. The transferred current through
MtrF can be measured as a function of the applied tip-substrate potential and
compared with the results from recent simulations of MtrF conductance [40].
Moreover, additional details about the alignment and conguration of OM
cytochromes on nanowires can be obtained by high resolution imaging using atomic
force microscopy (AFM) and transmission electron microscopy (TEM). These
observations can guide the assumptions regarding the conguration and spacing
of the redox sites (hemes) and cytochromes, important factors in the multistep
hopping model. In addition, the exact morphology and dimension of bacterial
nanowires is more accurately measured in AFM and TEM imaging. In chapter 5
of this thesis, we describe AFM and TEM imaging of bacterial nanowires produced
in the perfusion imaging platform. The preliminary TEM images reveal the mem-
branous nature of nanowires while, along with the AFM images, indicating their
various morphologies ranging from vesicle chains to smooth laments.
In the context of MFCs, the biotic-abiotic interaction between the bacteria and
the electrode in the MFC is extremely important in the eciency and quality of
13
renewable energy conversion [50]. Bacterial nanowires constitute one of the mecha-
nisms proposed for electron transfer between bacteria and extracellular solid phase
electron acceptors such as electrodes in a MFC [50]. Using bacterial nanowires, not
only the bacteria adjacent to the electrode can contact the surface to transfer elec-
trons, but other cells remote from the surface can transfer electrons to the cells that
are in contact with the electrode. Thus in a biolm, which is normally formed on
such electrodes in MFCs, bacteria can breathe through one another using bacterial
nanowires and ultimately transfer the electrons extracted by the large community
of cells in the biolm to the electrode [50, 51]. Therefore understanding mecha-
nisms of extracellular electron transfer in bacteria and the properties of bacterial
nanowires can help us improve electron transfer at the interface of bacteria and
electrodes and is crucial in developing more ecient microbial fuel cells as sources
of cheap renewable energy.
In addition to the impact in microbial fuel cell research, bacterial nanowires
might play an important role in biolm formation and maintenance. First, they
may establish contact between cells by connecting cells together to possibly build
the structural backbone of the biolm. Second, due to their conductive properties,
bacterial nanowires might have the ability to facilitate respiration of bacteria in
a biolm by transferring electrons between cells which leads to sustained cellular
respiration and energy conservation [52]. Examples of such biolms occur in oral
pathology, in disorders such as osteonecrosis of the jaw, where biolms are dicult
to treat and are resistant to antibiotics [52, 53]. In this case, the ability to destabi-
lize the structural and metabolic integrity of the biolm can be essential in treating
the disorder. Notably, bacterial nanowires have been shown to exist and play a
role in the stability of these biolms [52, 53]. Therefore a better understanding of
14
bacterial nanowires and their properties can lead to development of treatments for
biolm-related disorders such as osteonecrosis of the jaw.
This thesis includes 6 chapters. Chapter 2 introduces the bioenergetic context,
the mechanisms of extracellular electron transfer, the non-adiabatic electron trans-
fer theory, and describes the specic
uorescence microscopy schemes used in this
thesis. In this chapter, we also review some aspects of the solved crystal structure
of the outer membrane cytochrome MtrF.
Chapter 3 features a quantitative evaluation of the proposed mechanisms of
electron transport in bacterial nanowires based on the experimental evidence. Then
we introduce the multistep hopping model of electron transport and compare its
results with the conductivity measurements on bacterial nanowires. This chap-
ter also motivates the in vivo cytochrome-localization experiments described in
chapter 5.
Chapter 4 introduces the in vivo imaging of bacterial nanowires. We demon-
strate that the production of nanowires correlates with increased cellular reduc-
tase activity. We also show, using
uorescent dyes and proteins, that Shewanella
nanowires are outer membrane extensions. We describe our gene expression analy-
sis of Shewanella nanowire samples in chemostats where we analyze the changes in
expression of various genes including type IV pili genes as cells produce nanowires.
This analysis serves as additional evidence supporting our ndings on bacterial
nanowire composition.
Chapter 5 addresses the localization of outer membrane cytochromes MtrC and
OmcA on Shewanella nanowires as well as the electron transfer properties of the
outer membrane cytochrome MtrF. Antibodies raised against OmcA and MtrC
in combination with
uorescent secondary antibodies are used to demonstrate the
presence of these proteins along nanowires. Samples described in this chapter are
15
prepared in the in vivo imaging platform described in chapter 4, where the target
nanowires imaged during their growth are later checked for OM cytochrome pres-
ence. The ndings reported in this chapter point to the OM cytochromes as pos-
sible facilitators of ET along nanowires, as was assumed in the multistep hopping
model in chapter 3. Based on data from the gene expression analysis on chemo-
stat samples introduced in chapter 4, we also report an increase in expression of
OM cytochrome genes upon production of nanowires. These genes include omcA,
mtrC andmtrA. In addition we describe the development of an approach for inves-
tigating the electron transfer characteristics of individual molecules of the outer
membrane cytochrome MtrF. Puried MtrF is used to assemble a uniform mono-
layer of the cytochrome on atomically
at gold substrate. Single MtrF molecules
can then be addressed with an STM tip in order to measure their conductivity,
resulting in current-voltage curves. In addition, using live
uorescence microscopy
in combination with TEM and AFM imaging, we demonstrate the dierent stages
of nanowire growth and the occasional transformation of nanowires from vesicle
chains to smoother laments during early growth. The results reported here fur-
ther clarify the morphologies of bacterial nanowires and indicate the membranous
nature of these laments. This chapter also points to TEM as a promising tech-
nique to investigate the conguration, alignment and spacing of outer membrane
cytochromes along bacterial nanowires.
Finally, chapter 6 concludes this thesis. We review the work presented in this
thesis and look at possible future work that can be built on the studies reported
here to further advance our knowledge of extracellular electron transfer.
16
Chapter 2
Background
2.1 Energetics of life and oxidative phosphoryla-
tion
Living organisms are highly ordered systems in contrast to their environments. In
order for life to maintain this imbalance of order, free energy has to be continu-
ously transferred from the environment to the system [54]. A limited number of
strategies are used by various forms of life to accomplish this vital task including
respiration, fermentation and photosynthesis. An essential process in these strate-
gies is electron transfer (ET) within or between biomolecules in biomembranes.
The ET reactions lead to the formation of an electrochemical gradient across the
cell or mitochondrial membrane that provides energy for formation of phosphate
bonds in adenosine triphosphate (ATP). The dissociation of this bond, which leads
to the conversion of ATP to adenosine diphosphate (ADP) is a universal source of
free energy in biology [55].
In respiration, electrons derived from a high energy source, an electron donor
(ED),
ow through the electron transport chain (ETC) and are discarded to the
environment, i.e. a low energy electron acceptor (EA) [3]. The energy dispar-
ity between high and low energy electrons is harnessed by the cell in the form of
produced ATP molecules. A variety of electron donors and acceptors are used in
17
Figure 2.1: Schematic of the electron transport chain (ETC) in respiratory organ-
isms. Electron transfer through the components of the ETC drives the pumping of
protons across the inner (mitochondrial) membrane. The resulting proton motive
force then powers the production of ATP, the universal currency of free energy in
biology.
respiratory metabolisms, depending on their energy levels and availability. Exam-
ples of common electron donors and acceptors are carbohydrates and molecular
oxygen (for aerobic organisms), respectively.
Electron transport through the electron transfer chain involves a number of
carrier sites and proton pumps that are placed inside the inner membrane of the
cell (or the mitochondrial inner membrane) as well as the ATP producing enzyme
ATP synthase (Fig. 2.1) [56]. As electrons enter and move along the chain
between the carrier sites, protons are pumped from the inside of the cell to the
outside, building up a proton concentration gradient across the inner membrane
[3]. Subsequently the ATP synthase acts as a channel through which these protons
can move back to the cell interior. The energy from the proton gradient dissipation
is harnessed by ATP synthase to high energy phosphate bonds in ATP molecules
18
through transforming ADP to ATP. Each ATP molecule stores about 500 meV
of energy which can be released through ATP hydrolysis [56]. The synthesized
ATPs are later used in various functions as source of energy, e.g. to synthesize
proteins, RNA and DNA or in active transportation of molecules within the cell
in eukaryotic cells.
Electron transfer from ED to EA is the driving force for cellular respiration and
is continuously required for steady production of ATP [3]. The redox potential
of the EA relative to the ED determines how favored the EA is to be used by the
organism. A more negative potential for ED and a more positive redox potential
for EA are favorable for the cell (Fig. 2.2) [57].
2.2 Anaerobic respiration and dissimilatory
metal reduction
Oxygen is the most energetically favored electron acceptor (Fig. 2.2 2), and it
is widely used by various species. When oxygen is used as electron acceptor, the
respiration is called aerobic as opposed to anaerobic respiration in the case of other
electron acceptors. Metals ions are one group of electron acceptors that are more
available than oxygen in some environments on earth, for instance meters beneath
the ground where oxygen is scarce [22]. Some of the organisms evolved under
oxygen-limited conditions use metal ions to dispose of their respiratory electrons.
The group of prokaryotes that couple metal reduction to cellular respiration is
called dissimilatory metal reducing bacteria or DMRBs. As opposed to oxygen
and other soluble electron acceptors, solid metal oxides can be used by DMRBs
as EA outside the cell [22]. To accomplish this there is an entire apparatus in
19
Figure 2.2: Redox tower containing a list of electron donors and acceptors and the
redox potentials associated with their redox reactions. For electron donors a more
negative potential (top of the tower) and for electron acceptors a more positive
potential (bottom of the tower) is favorable for use by cells. Oxygen has one of
the most positive redox potentials in the list; therefore it is commonly used as the
EA by various organisms. Organisms are categorized (red arrows) according to
the electron donor/acceptor combination they consume. Aerobes, for instance, use
organic matter as electron donor and oxygen as the electron acceptor in respiration.
DMRBs to transfer electrons to the cell exterior and extracellularly reduce the EA
[14].
Shewanella oneidensis MR-1, a facultative anaerobe, is one of the well-studied
DMRBs that was isolated [8] from Lake Oneida in New York. This strain was
initially reported to couple its anaerobic growth to reduction of manganese oxide,
but was later conrmed to be much more versatile in terms of electron acceptor
reduction and hence serves as a great model organism for investigating extracellular
respiration [6, 8]. The connection between metabolic respiration and extracellular
oxide reduction has been shown using electron transport inhibitors [9]. These
20
chemicals disrupt the function of carrier sites in the electron transport chain and
thus stop the
ow of respiratory electrons. When S. oneidensis MR-1 was exposed
to these inhibitors, the oxide reduction stopped; indicating the coupling between
electron transport and reduction of oxides [9].
2.3 Outer membrane cytochromes
All the proposed mechanisms for extracellular electron transfer (EET) in S. onei-
densis MR-1 involve outer-membrane (OM) cytochromes [58]. Cytochromes, in
general, are proteins that incorporate specic iron-containing prosthetic groups
referred to as hemes. Presence of iron atoms in the cytochrome structure enables
it to catalyze redox reactions by carrying the electrons involved in the reaction [19].
In the case of extracellular respiration, outer-membrane cytochromes are shown to
be present on the outer surface of the outer-membrane of the cell [11, 13], enabling
the transfer of electrons from the cell to the terminal electron acceptor. In addi-
tion to that, other cytochromes have been found to be responsible for carrying the
electrons from the inner membrane (the electron transport chain) to the periplasm,
through a porin in the outer membrane, and to the exterior of the cell [14, 59]
(Fig. 2.3). As part of this chain, `porin-cytochrome' modules are found to trans-
port electrons across the outer membrane [18, 59] (Fig. 2.4). Overall, this array
of cytochromes help conduct electrons necessary for ATP production to the outer
lea
et of the outer-membrane (Fig. 2.3).
2.4 MtrF and its crystal structure
Recently the crystal structure of MtrF, one of 42 putative OM cytochromes in She-
wanella has been published [17], raising the possibility of more careful and detailed
21
Figure 2.3: The extracellular electron transport pathway in Shewanella. Respira-
tory electrons are transferred to the cytochrome CymA at the inner membrane and
subsequently to the porin-cytochrome module at the outer membrane. Reprinted
with permission from [14], copyright John Wiley & Sons Inc.
Figure 2.4: The porin-cytochrome module consists of a decaheme cytochrome
embedded from the periplasmic side into a porin in the outer membrane, the porin,
and an outer membrane cytochrome. From [59].
22
Figure 2.5: Position of heme groups in the staggered cross-like structure of the
decaheme outer membrane cytochrome MtrF. The spacing between adjacent hemes
is less than 1 nm. From [17], Copyright National Academy of Sciences USA.
calculations regarding electron transport in MtrF, its homologs and larger extra-
cellular structures incorporating OM cytochromes. The molecule has an oblate
ellipsoid shape with dimensions of 85 70 30

A. This 3.2-

A resolution crystal
structure shows the relative position of all the 10 hemes within a decaheme OM
cytochrome for the rst time [17] (Fig. 2.5). The heme localization pattern in
MtrF resembles a staggered cross with a 65-

A octaheme chain along the length
that is crossed by a 45-

A tetraheme chain in the middle (Fig. 2.5).
Interestingly, the spacings between all adjacent hemes in MtrF are smaller
than 1 nm (Fig. 2.5), suggesting the possibility of ET through the entire molecule
by multistep electron hopping between neighboring hemes [34{36]. Heme 10 is
proposed to be the input site of electrons from the cell, while heme 5, evidenced
23
by its high exposure to the solvent, would act as the exit point of electrons onto
the insoluble mineral [17]. This would mean that electrons move along the longer
chain (from heme 10 to heme 5) to reach the mineral.
In addition to the electron transfer path to the insoluble electron acceptor, there
are two beta-barrel domains containing extended Greek key split-barrel structures
which are common in
avin binding domains [17]. In addition, researchers have
shown that OM cytochromes of Shewanella reduce
avins, molecules proposed to
act as electron shuttles in EET [60]. Although it has not been directly demon-
strated, these ndings point at the possibility that
avins are reduced by binding
to MtrF at these beta-barrel domains. The bound
avins would then exchange
electrons with hemes 2 and 7, the closest hemes to the
avin-binding sites [17].
Therefore, the electron transport from heme 10 to either heme 2 or 7 would be
another scenario of multistep ET that results in reduction of shuttles. This alter-
native pathway is especially benecial if the heme 10-to-heme 5 ET pathway is
blocked, possibly due to lack of direct contact between the cytochrome (at heme
5) and the insoluble mineral.
To investigate the conductive properties of individual MtrF molecules, the pro-
tein must be placed between two conductors that act as electrodes. For this reason,
in the conductivity experiments described in this thesis, we use recombinant MtrF
that has a surface exposed tetra-cysteine tag at its C-terminus (Fig. 2.6), enabling
the cytochrome to covalently bind to Au(III) [40, 61]. Upon binding of MtrF to
gold, the gold substrate can be used as an electrode that is in contact with the
target molecule. The gold-adsorbed MtrF is then approached by the atomically
sharp and conductive tip of the Scanning Tunneling Microscope (STM) from above
to perform the desired conductivity measurements (Fig. 2.7) [40, 61]. This is
24
done by applying a bias voltage between the gold substrate (bottom) and the con-
ductive tip (top) and measuring the current passing through the molecule, between
the two electrodes.
2.5 Atomic force and scanning tunneling
microscopy
Atomic force microscopy (AFM) [62] and scanning tunneling microscopy (STM)
[63] both belong to the family of probe-based microscopy techniques and are used
for nanoscale characterization of samples that are typically deposited on a surface.
In both techniques, a sharp tip approaches the sample from above while the sample-
tip spacing is measured through movements of the tip (in AFM) or the magnitude
of the tunneling current between the tip and the substrate (in STM). Using a
piezoelectric stage that accurately controls the height of the substrate, a feedback
loop maintains the tip-sample distance constant while the tip laterally sweeps the
sample. The obtained map of the stage height as a function of the lateral tip
position gives the topography of the sample. Unlike AFM, STM can also be used
to examine electron transfer properties of the sample by measuring the tunneling
current as a function of the applied tip-substrate potential.
2.6 Pathways of extracellular electron transport
(EET)
Extracellular electron transfer pathways can be categorized into two broad groups:
direct and indirect mechanisms [58]. Indirect mechanisms involve small molecules
such as
avins that either, as hypothesized, diusively shuttle respiratory electrons
25
Figure 2.6: Crystal structure of the decaheme outer membrane cytochrome MtrF
along with the position of the protein C-terminus. A tetra-cysteine tag added at
the C-terminus allows covalent binding of the protein to a gold substrate. From
[17], Copyright National Academy of Sciences USA.
from the cell surface to the solid-phase electron acceptor [41, 42] or chelate metal
ions inside the cell to interact with the cellular oxidoreductases.
In direct mechanisms, electrons are transferred via a number of multiheme
cytochromes directly to the mineral surface. This includes ET through direct con-
tact between cell and solid surface or through micrometer-long conductive laments
known as bacterial nanowires [30, 31]. Another case of direct extracellular elec-
tron transfer occurs in interspecies electron transfer through natural conductive
minerals [64], where a cell from one species capable of oxidizing a specic electron
donor transfers its respiratory electrons through conductive minerals to another
cell of a dierent species that reduces a specic electron acceptor.
26
Figure 2.7: Schematic showing an MtrF molecule adsorbed on an atomically
at
gold substrate with a scanning tunneling microscope (STM) tip approaching it
from above. The STM tip and the gold substrate act as two electrodes in the
MtrF conductance measurements.
2.7 Bacterial nanowires and their conductive
properties
Bacterial nanowires produced by S. oneidensis MR-1 were rst discovered in 2006
[31]. Cell cultures were grown continuously in a chemostat with excess oxygen
as electron acceptor, under ED-limitation conditions. At some point, the oxygen
input rate into the system is reduced so much that the conditions are changed
from ED-limitation to EA-limitation in the reactor. Samples were taken from the
oxygen-limited culture, were imaged with a scanning electron microscope (SEM)
and were conrmed to contain microbial appendages. Further analysis of these
27
samples using scanning tunneling microscopy showed that these appendages are
conductive and therefore can be referred to as nanowires. Also, as another indica-
tion of conductivity of these nanowires, they were shown to be essential in biological
reduction of solid metallic electron acceptors by S. oneidensis MR-1. But further
experiments were yet to be done to determine the exact transport mechanism and
the conductive properties of bacterial nanowires.
The rst quantitative measurement on the conductivity of bacterial nanowires
was performed in 2008 [65]. Using conductive atomic force microscopy a voltage
was applied across the width of a single nanowire and the resulting current was
measured. The I-V curve from that experiment showed nA currents being carried
by the nanowire across its width (Fig. 2.8A). Two years later another experiment
demonstrated that bacterial nanowires are in fact able to transfer electrons longi-
tudinally (m length scales) [32]. Chemostat samples containing nanowires were
deposited on a silicon oxide surface with pre-fabricated gold lines. Using SEM,
each nanowire candidate was found on the non-conductive silicon oxide between
two gold lines. Then by using platinum deposition, two micron-scale electrodes
were deposited to connect each end of the nanowire to one of the gold lines. Later
a voltage was applied between the two gold lines connected to the nanowire and
the current was measured, producing the I-V curve in Fig. 2.8B.
One important feature of both of these results is the magnitude of the current
at around 1 V. This number is orders of magnitude larger than the cellular respi-
ration electron transfer rate of 10
6
electrons per second [32, 66], indicating that
a single nanowire is capable of supporting an individual cell in terms of extracel-
lular respiration. This is consistent with the proposed role of bacterial nanowires,
namely to transfer a cell's respiratory electrons to the terminal electron acceptor.
28
Figure 2.8: I-V curves from conductance measurements on bacterial nanowires of
Shewanella. A) Transverse measurement across the 10-nm width of a nanowire. B)
Longitudinal measurement along the 2000-nm length of a nanowire. The transverse
and longitudinal curves appear to be dierent, raising questions about the physical
electron transport mechanism behind nanowire conductance. A reprinted from
[65], with permission from Elsevier. B reprinted from [32], Copyright National
Academy of Sciences USA.
By looking at the I-V curves from the longitudinal and transverse conductance
measurements, one may ask why the two curves appear to be so dierent. It is
worth noting that despite the importance of theoretical modeling of these conduc-
tance data in answering such questions and revealing the mechanism of electron
transport in bacterial nanowires, these measurements were not modeled in any way
prior to the work reported in this thesis.
2.8 The theory of non-adiabatic electron transfer
An initial step to investigate transport in both individual cytochromes and bacte-
rial nanowires is to specify the electron transfer (ET) rate between two neighboring
29
redox sites, which are in this case two hemes. The powerful theory of non-adiabatic
ET which originated from the theoretical ndings of Rudolph Marcus in 1956 [28]
describes the electron transfer rate in ET reactions in terms of two factors [2].
The rst factor which was the focus of Marcus' earlier work concerns the eect of
nuclear congurations of the donor-solvent-acceptor system on the ET rate that
are collectively re
ected in a single nuclear reorganization parameter (, or the
reorganization energy). Marcus introduced an exponential term involving to
describe the eect of the nuclear congurations on the overall ET rate:
k
et
/ exp
"

(G +)
2
4k
B
T
#
(2.1)
where G is the free energy change as a result of the ET, is the reorganization
energy,k
B
is the Boltzmann constant andT is the temperature. A counter-intuitive
prediction resulting from Marcus' exponential expression was the Marcus inverted
region, where increasing the driving force of the reaction reduces the overall reac-
tion rate. This prediction was later veried by experiments [67, 68].
Later in 1974, Hopeld introduced the eect of the electronic coupling between
the donor and the acceptor in the form of electron tunneling through a square
barrier [29]. This eect was included in the overall ET rate as a pre-exponential
term behind the Marcus exponential expression which is equal to the electron
tunneling rate when the nuclei are in the transition state from reactant to product
in the ET reaction. Combining the electron coupling and the nuclear conguration
eects gives the non-adiabatic ET rate as [2]:
k
et
=
2
h
jH
DA
j
2
p
4k
B
T
exp
"

(G +)
2
4k
B
T
#
(2.2)
30
where is the reorganization energy of the system, k
B
is the Boltzmann con-
stant,T is the temperature, G is the free energy change as a result of the electron
transfer, andjH
DA
j is the coupling matrix element which contains the exponential
tunneling factor. In the context of cytochrome ET, is the reorganization energy
of the cytochrome and G is the electric potential dierence between the donor
and the acceptor as a result of the applied voltage to the system.
In experiments regarding the conductive properties of OM cytochromes and
nanowires, the potential is applied using two electrodes which make contact with
the system. Therefore, other than the protein-protein ET kinetics, the interaction
between the electrode and the cytochrome has to be taken into account [69].
For this reason, we also consider the density of states in the metallic electrode
which is a function of the potential applied to the electrode. So similar to the
protein-protein case, the protein-electrode ET rate is given by [69]:
k
et
=C
electrode
Z
1
1
exp
"

x
+e(EE
0
)
k
B
T

2
k
B
T
4

#
1 + exp(x)
dx (2.3)
where C
electrode
=
2
h
q
k
B
T
4
jHj
2
, E
0
is the potential of the protein, E is the
potential of the electrode, is the reorganization energy of the protein,jHj is the
coupling matrix element, and is the density of states in the electrode [69]. There-
fore this modied version of the non-adiabatic ET theory describes the protein-
electrode ET kinetics. In the end, a complete model attempting to describe the
experimental data on OM cytochrome/nanowire conductivity must address both
the protein-protein and protein-electrode interaction properties using the above
ET rate equations.
31
2.9 Fluorescence Microscopy
Here we describe the various
uorophores used in combination with
uorescence
microscopy in this dissertation to characterize the composition, structure and com-
ponents of bacterial nanowires and outer membrane vesicles.
To visualize proteins in bacterial extracellular structures, we use the
uorescent
dye NanoOrange. This dye is non
uorescent in aqueous solution, and undergoes a
signicant increase in
uorescence upon binding to hydrophobic regions of proteins
[70]. NanoOrange was previously used to visualize nanometers-thick lamentous
protein structures such as bacterial
agella [71]. In this work we use NanoOrange
for staining the bacterial structures of interest, i.e. bacterial nanowires and outer
membrane vesicles, due to its eectiveness in staining similar bacterial structures.
To test the extent of membrane involvement in nanowire formation, we labeled
S. oneidensis MR-1 cells with the membrane stain FM 4-64FX. This styryl dye
is membrane-selective as a result of a lipophilic tail that inserts into the lipid
bilayer and a positively charged head that is anchored at the membrane surface
[72, 73]. The amphiphilic nature of this molecule hinders it from freely crossing
the membrane into the cellular interior except through the endocytic pathway,
as extensively characterized in eukaryotic cells [73, 74]. FM 4-64 has also been
widely used in bacterial cells and shown to specically label membranes but not
extracellular protein laments such as
agella [75, 76], except in a few bacterial
species where
agella are coated in membrane sheaths [77].
In addition, we used a
uorescent redox-sensing dye, RedoxSensor Green
(RSG), to demonstrate the impact of nanowire production on live cells of S.
oneidensis MR-1. RSG is an indicator of bacterial reductase activity and it serves
as a reporter for changes in electron transport chain function [49]. This dye per-
meates through both cellular membranes and interacts with components of the
32
cellular electron transport chain, producing a stable green
uorescence emission
upon reduction. Furthermore, bacterial cells undergo a signicant decrease in
RSG
uorescence upon addition of inhibitors of dierent electron transport chain
components [49]. RSG was previously used as a reporter of bacterial metabolic
activity in environmental samples [48] and as an indicator of the respiratory activ-
ity of Geobacter sulfurreducens cells [78]. In this work, we demonstrate, for the
rst time, the capacity of RSG as a marker for respiratory activity in Shewanella
(chapter 4) and subsequently use RSG to show the increase in cellular respiration
rate upon production of bacterial nanowires.
In this work, we use indirect immuno
uorescence, a popular technique for visu-
alizing specic biomolecules, to observe localization of the OM cytochromes MtrC
and OmcA along bacterial nanowires [79]. In this technique, the structures are
stained with rabbit-raised polyclonal primary antibodies that are specic to MtrC
and OmcA (the antigens) and bind to the corresponding epitopes on these pro-
teins via their fragment antigen-binding (Fab) domains [79] (Fig. 2.9). A goat
anti-rabbit secondary antibody is subsequently added that binds the cytochrome-
bound primary antibodies at their fragment crystallizable region (Fc) [79] (Fig.
2.9). Since the Fc domain is constant between antibodies from the same species
(in this case rabbit), a single anti-rabbit secondary antibody is used to bind the
Fc domains of both MtrC and OmcA-specic rabbit-raised primary antibodies.
The secondary antibody used also carries multiple conjugated FITC (Fluorescein
isothiocyanate)
uorophores (Fig. 2.9), in turn allowing visualization of MtrC
and OmcA along bacterial nanowires.
33
Figure 2.9: Schematic of the indirect immuno
uorescence technique. After binding
of the primary antibody to the antigen, the Fab region of the secondary antibody
binds to the Fc region of the primary antibody. The
uorophore conjugated to
the secondary antibody enables visualization of the antigen through
uorescence
microscopy. Reprinted by permission from Macmillan Publishers Ltd: Journal of
Investigative Dermatology [79], copyright 2013.
34
Chapter 3
Multistep Hopping Model of
Electron Transport in Bacterial
Redox Chains
3.1 Introduction
Reduction-oxidation (redox) reactions and electron transport are essential to the
energy conversion pathways of living cells [80]. Respiratory organisms generate
ATP molecules - life's universal energy currency - by harnessing the free energy
of electron transport from electron donors (fuels) to electron acceptors (oxidants)
through biological redox chains. In contrast to most eukaryotes, which are limited
to relatively few carbon compounds as electron donors and oxygen as the predom-
inant electron acceptor, prokaryotes have evolved into versatile energy scavengers.
Microbes can wield an astounding number of metabolic pathways to extract energy
from diverse organic and inorganic electron donors and acceptors, which has sig-
nicant consequences for global biogeochemical cycles [8, 22, 81].
For short distances, such as between respiratory chain redox sites in mitochon-
drial or microbial membranes separated by < 2 nm, electron tunneling is known
to play a critical role in facilitating electron transfer [80]. Recently, microbial
electron transport across dramatically longer distances has been reported, ranging
35
from nanometers to micrometers (cell lengths) and even centimeters [58]. A few
strategies have been proposed to mediate this long-distance electron transport in
various microbial systems: soluble redox mediators (e.g.,
avins) that diusively
shuttle electrons, conductive extracellular laments known as bacterial nanowires,
bacterial biolms incorporating nanowires or outer membrane cytochromes, and
multicellular bacterial cables that couple distant redox processes in marine sedi-
ments [30{33, 42, 47, 82, 83]. Functionally, bacterial nanowires are thought to oer
an extracellular electron transport (EET) pathway linking metal-reducing bacteria,
including Shewanella and Geobacter species, to the external solid-phase iron and
manganese minerals that can serve as terminal electron acceptors for respiration.
Furthermore, bacterial nanowires of Shewanella were shown to have electron
transport rates up to 10
9
s
1
at 100mV of applied bias and a measured resistivity
on the order of 1
cm; sucient to keep up with the typical specic respiration
rates of these microbes [32]. At the same time, mutants lacking specic multiheme
cytochromes (MtrC and OmcA) produce non-conductive laments under identical
conditions, suggesting that extracellular redox sites are necessary for long-range
EET via microbial nanowires [32].
In addition to the fundamental implications for respiration, EET is an especially
attractive model system because it has naturally evolved to couple to inorganic
systems, giving us a unique opportunity to harness biological energy conversion
strategies at electrodes for electricity generation (microbial fuel cells) and produc-
tion of high-value electrofuels (microbial electrosynthesis) [26].
This chapter focuses on the theoretical basis of how biotic components such
as microbial nanowires and associated multiheme c cytochromes form conductive
36
biolms facilitating direct microbe-to-electrode EET. We build on two recent the-
oretical studies proposing multistep hopping in redox chains as the physical mech-
anism of this long-range charge transfer in Geobacter biolms and Shewanella
nanowires [35, 36]. The proposed model is based on an incoherent multistep hop-
ping mechanism between redox sites, and an interfacial treatment of non-adiabatic
(Marcus theory) electron transfer rate equations to account for the electrochemi-
cal interactions with measurement electrodes [69]. Using this model, we compute
current-voltage (IV ) curves consistent with both transverse and longitudinal
experimental measurements of Shewanella nanowires (published data [32, 65] and
new higher-bias data reported here) as well as inter-site spacings from the recently
determined crystal structure of Shewanella cytochromes [17].
3.2 Modeling
3.2.1 Background
Previous charge transport measurements in microbial nanowires and biolms were
performed over length scales far exceeding the size of individual cells (>1 m)
[32, 35, 47], and are therefore beyond the scope of direct single-step tunneling
mechanisms relevant over much shorter distances (<2 nm) [80]. For this reason,
the direct EET systems have been interpreted in light of two mechanisms: (i)
fully coherent band conduction much like metals and semiconductors [47], and
(ii) incoherent multistep hopping between charge localizing sites, such as redox
cofactors [35, 36]. We start by considering the length and time scales involved in
these physically distinct ideas, and consequently their applicability to microbial
EET.
37
As discussed by Polizzi et al. [36], band conduction requires that the scattering
time of the carriers,T
s
, and the width of the energy band,W , satisfy the condition
T
s
W h with h being the Planck constant. The width of the energy band can
be dened as W = 4jH
DA
j, wherejH
DA
j is the charge transfer integral between
the localizing sites in the system. At the same time, the charge mobility in band
theory is
BT
= 2er
2
jH
DA
jT
s
= h
2
, where e is the charge and r is the distance
between localizing sites [84]. Combining these expressions results in a fundamental
requirement that
BT
er
2
=2 h. In other words, there is a minimum mobility for
the band theory picture. Even using a small separation between neighboring sites,
0:35 nm (consistent with -stacking in conducting polymers), this requires that
the mobility be far greater than 1 cm
2
.V.s
1
[36]. However, from our previous
measurements of Shewanella nanowires [32], the conductivity is measured to be 1
S.cm
1
and the mobility can be estimated to be lower than 10
3
cm
2
.V.s
1
[36].
While enough to sustain microbial respiration [32], this is clearly well below the
band theory limit. For this reason, we contend that fully coherent band conduction
is not a viable model for microbial nanowire conductivity and therefore exclude
the band picture in the model and experimental measurements described below.
Other aspects of the applicability of band theory to describe previously reported
experimental measurements in Geobacter nanowires [47] (also with mobilities far
below the band limit described above) have been debated elsewhere [45].
In the hopping picture, for an electron hopping process between two wells, the
vibrational relaxation rate should be larger than the hopping rate for the hopping
step to be independent of the preceding and succeeding hopping, i:e:k
rel
k
hop
.
This has recently been interpreted by Troisi [85] as a `speed limit', imposing a
maximum charge mobility for sequential hopping. Furthermore, this limit can be
38
conveniently estimated from spectroscopic measurements in organic solids since
the vibrational relaxation rate is given byk
rel
= 2c, wherec is the speed of light
and is the Raman line broadening [85, 86]. Using standard values for molecular
materials, k
rel
typically exceeds 10
11
s
1
[85]. On the other hand k
hop
can be
estimated from the non-adiabatic rate equation for electron transfer between two
sites [2]:
k
hop
=
2
h
jH
DA
j
2
p
4k
B
T
exp
"

(G +)
2
4k
B
T
#
(3.1)
where is the reorganization energy of the system, k
B
is the Boltzmann con-
stant, T is the temperature, and G is the free energy change as a result of the
electron transfer. This expression is frequently encountered in the simplied phe-
nomenological form [36]:
k
hop
(s
1
) = 10
13
exp
"
R
(G +)
2
4k
B
T
#
(3.2)
where is the tunneling decay factor ( 1

A
1
), andR is the eective tunneling
distance between two neighboring sites (the dierence between the nearest neighbor
hopping distance and the distance at van der Waals contact, the latter taken
to be 0:35 nm). At the maximum hopping rate for ecient biological electron
transfer, =G, and taking R = 0:65 nm, a value consistent with typical
inter-cofactor distances in a Shewanella multiheme cytochrome [17], this results
in k
hop
= 1:5 10
10
s
1
. In summary, this rate can fall below the relaxation rate,
allowing for independent hopping steps from electrode to electrode through a chain
of redox sites.
39
Figure 3.1: (a) A redox chain bridging two measurement electrodes. Forward and
backward charge transfer rates within the redox chain are determined by k
hop;f
and k
hop;b
respectively, while the interactions with electrodes are determined by
the electrochemical transfer rates k
red
and k
ox
at the left and right contacts. The
redox sites pictured in the schematic represent hemes such as those found in the
multiheme cytochromes of the dissimilatory metal-reducing bacterium Shewanella
oneidensis MR-1. (b) Scanning electron microscopy image of an experimental
platform, where two Pt electrodes address an individual microbial nanowire from
a single S. oneidensis MR-1 cell (rod shaped cell to the right). [34] - Reproduced
by permission of the PCCP Owner Societies.
3.2.2 Multistep Hopping
The multistep hopping chain model is schematically illustrated in Fig. 3.1(a),
whereL redox sites bridge two measurement electrodes. This picture is motivated
by the experimental approaches recently used to measure the current response of
one dimensional structures such as microbial nanowires when a voltageV is applied
between the electrodes (Fig. 3.1(b)). The forward and backward electron transfer
rates between redox sites within the chain are given by k
hop;f
and k
hop;b
, both of
which follow the expression for k
hop
above but with the corresponding G values
ofeV=L and eV=L, respectively. This electron hopping step between sites i and
i + 1 (1iL 1) is represented by the following reaction [35, 85]:
(i)
reduced
+ (i + 1)
oxidized
k
hop;f
*
)
k
hop;b
(i)
oxidized
+ (i + 1)
reduced
(3.3)
40
which in turns gives the electron
ux between two neighboring redox sites as [35]:
J =k
hop;f
P (i)[1P (i + 1)]k
hop;b
P (i + 1)[1P (i)] (3.4)
where P (i) denotes the occupation probability of site i (reduced probability) and
1P (i) denotes the vacation probability of sitei (oxidized probability). The het-
erogeneous transfer from/to the electrodes (electrochemical oxidation and reduc-
tion of the rst or L'th site) is determined using the electrochemical form of
the non-adiabatic electron transfer rate equation [69], by considering the overlap
between the electronic density of states in the metallic electrode, (Fermi-Dirac
distribution), and the Gaussian oxidation or reduction peaks of the neighboring
redox site, situated + and with respect to the redox potential, as illustrated
in Fig. 3.2. The rates k
red;
and k
ox;
, representing electron transfer from and to
an electrode (to and from site where = 1 or L), are given by [36, 69]:
k
red;
=C
electrode
Z
1
1
exp
"

x
+e(EE
0
)
k
B
T

2
k
B
T
4

#
1 + exp(x)
dx (3.5)
k
ox;
=C
electrode
Z
1
1
exp
"

x
e(EE
0
)
k
B
T

2
k
B
T
4

#
1 + exp(x)
dx (3.6)
where C
electrode
=
2
h
q
k
B
T
4
jHj
2
and (E E
0
)

is the dierence between the
applied electrode potential (E) and the potential of its neighboring redox site (E
0
)
at either the left or right contact. The latter can also be represented as a local
voltage drop. For example, at the left electrode, (EE
0
)
1
= V where is
a fraction of the overall applied voltage V between the two electrodes. These
41
Figure 3.2: A schematic of the electrochemical interaction between the redox
species and a neighboring electrode. Applying a voltage V changes the relation
of the metallic states with respect to the oxidation- reduction probability peaks
of the redox site. (a) When V = 0, the Fermi energy of the metal electrode (eE)
is equivalent to the redox energy (eE
0
), leading the oxidation and reduction rates
to balance, and therefore the net charge transfer is zero. (b) Applying a voltage
shifts the energy level of the electrode with respect to the redox level, favoring
reduction or oxidation (V < 0 pictured, leading to increased reduction rate). [34]
- Reproduced by permission of the PCCP Owner Societies.
heterogeneous transfer rates allow the calculation of the charge
ux at the left and
right electrodes [87]:
J
1
=k
red;1
[1P (1)]k
ox;1
P (1) (3.7)
J
L
=k
ox;L
P (L)k
red;L
[1P (L)] (3.8)
Two symmetric contact electrodes will lead to the same magnitude of local
voltage drop at opposite electrodes, i:e: (EE
0
)
1
=(EE
0
)
L
, which in turn
42
translates tok
red;1
=k
ox;L
andk
ox;1
=k
red;L
. In addition, realizing that the charge

ux from one electrode to the other is equal:
J
1
=k
red;1
[1P (1)]k
ox;1
P (1) =J
L
=k
red;1
P (L)k
ox;1
[1P (L)])P (L) = 1P (1)
(3.9)
Finally, realizing that the
ux within the chain is equivalent to the
ux at the
electrodes, J = J
1
= J
L
, combining equations 3.4 and 3.9 gives the occupation
probability throughout the chain as:
P (i + 1) =
k
hop;f
P (i) +k
ox;1
P (1)k
red;1
[1P (1)]
k
hop;b
+ (k
hop;f
k
hop;b
)P (i)
; 1iL 1 (3.10)
Next, we apply this analysis to calculate the probability proles and correspond-
ing current-voltage (IV ) curves to compare with transverse and longitudinal
transport measurements of microbial nanowires from the bacterium S. oneidensis
MR-1.
3.3 Results and comparison to experiments
For a specicV (applied bias),L (number of sites),R (eective tunneling distance),
(reorganization energy), (fraction dening the contact voltage drop), C
electrode
(pre-exponential of the heterogeneous transfer rates k
red
andk
ox
), andk
B
T (ther-
mal energy), the model outlined above allows the calculation of the occupation
probability prole throughout the redox chain, and consequently the overall cur-
rent response to applied voltage. At each voltage step, the last sites occupation
probabilityP (L) can be calculated as a function of the rst sites occupation prob-
abilityP (1), by solving equation 3.10 recursively. Combining this relation with the
43
symmetric constraint from equation 3.9 results in a unique value forP (1), which is
dependent on the transfer ratesk
hop;f
,k
hop;b
,k
ox
, andk
red
(note all these rates are
functions of voltage). Next, the entire probability prole P (i) can be calculated
from equation 3.10. We performed this calculation in MATLAB, using parameters
consistent with existing transverse (smallL) and longitudinal (largeL)IV mea-
surements ofShewanella nanowires. The results described here assume the value of
k
B
T at room temperature, and take 1 nm as the typical distance between the redox
sites (i:e:R = 0:65 nm), consistent with the recently measured inter-heme spacings
in MtrF, a multiheme cytochrome from Shewanella [17]. We obtained excellent
ts to the experimental data by assuming an eective heterogeneous transfer rate
C
electrode
smaller than the hopping rate between sitesk
hop
. For each experiment, we
calculate the probability prole and the conventional (positive) current (I =J)
as a function of applied bias V . Finally, we t our calculations to IV measure-
ments, and comment on the parameters , , and C
electrode
that give rise to good
agreement with these experiments.
Fig. 3.3 shows the calculated probability prole for a redox chain consist-
ing of 2000 sites as the voltage is swept from 0 to 1 V. As the voltage increases
the occupation probability rapidly changes from a linear and relatively
at pro-
le to a sharp sigmoid with high occupation probability near one electrode and
a high vacation probability near the opposing electrode. With the probabil-
ity prole in hand, the corresponding IV behavior is simply calculated from
I =J
1
=k
ox;1
P (1)k
red;1
[1P (1)], and the result is plotted in Fig. 3.4. When
C
electrode
k
hop
, P (1) 1 under positive bias, except near zero applied voltage
(and similarly P (1) 0 for negative non-small bias). Under these conditions, the
current essentially follows the form of the heterogeneous electron transfer terms
44
Figure 3.3: The occupation probability prole for a chain composed of L = 2000
redox sites as the applied voltage between the surrounding electrodes is stepped
up in 0:1 V increments (corresponding to the experiment in El-Naggar et al. [32]
and the calculated IV proles of Fig. 3.4 and 3.5(b)). Calculation parameters:
L = 2000, = 0:4 eV, = 0:015, R = 0:65 nm, = 1

A
1
, k
B
T = 0:025 eV. [34]
- Reproduced by permission of the PCCP Owner Societies.
Figure 3.4: Calculated current as a function of the local electrode voltage drop for
theL = 2000 redox chain (using equation 3.7 and the occupation probability prole
of Fig. 3.3). The current reaches a constant value when the overlap between the
electrode's Fermi function and the redox peaks saturates at high applied voltage.
Calculation parameters: L = 2000, = 0:4 eV, = 0:015, R = 0:65 nm, = 1

A
1
, k
B
T = 0:025 eV. [34] - Reproduced by permission of the PCCP Owner
Societies.
45
Figure 3.5: A comparison of the measured IV characteristics and modeling
results for two dierent experiments [32, 65]. (a) Transverse transport across the
thickness of a microbial nanowire ( 10 nm), using a conductive tip as the top
electrode and a supporting surface as the bottom electrode (calculation parameters:
L = 10, = 0:4 eV, = 0:3, C
electrode
= 1:8 10
10
s
1
, k
B
T = 0:025 eV).
(b) Longitudinal transport along a microbial nanowire (2m) bridging two Pt
electrodes (calculation parameters: L = 2000, = 0:4 eV, = 0:015, C
electrode
=
5 10
11
s
1
,k
B
T = 0:025 eV). Insets show the AFM images corresponding to the
measurements with 250 nm scale bars. [34] - Reproduced by permission of the
PCCP Owner Societies.
equation 3.5 and 3.6, i.e. increasing the voltage increases the current response,
until the overlap between the electrode's Fermi function and the Gaussian redox
peaks saturates for very high voltages, as schematically illustrated in Fig. 3.2 and
discussed elsewhere [36].
We now compare the predictions of this model to measurements of transport in
conductive microbial appendages, performed using two dierent experimental tech-
niques. In the rst experiment [65] (Fig. 3.5(a)) transverse transport is measured
across the width (10 nm, corresponding to L = 10 in the model) of a microbial
46
nanowire using a conductive atomic force microscope (c-AFM) tip as a top elec-
trode and the underlying substrate as a bottom electrode. In the second experiment
[32], longitudinal transport is measured along a nanowire using two electron/ion
beam deposited electrodes separated by 2m (L = 2000). The previously reported
longitudinal measurements revealed linear IV curves for low voltages ( < 1 V),
but the non-linearity predicted by the model appears at higher bias, as can be seen
by the experimental data reported here for the same sample (up to 5 V in Fig.
3.5(b)).
For a similar voltage range, the IV measurements of the transverse and
longitudinal experiments appear distinct [32, 65]. However, the model is in good
agreement with both measurements, as shown on Fig. 3.5. Both experiments
were t with a reorganization energy close to 0:4 eV. While there are no experi-
mental measurements of in Shewanella's cytochromes, 0:4 eV is consistent with
electrochemically determined values of for cytochrome c at electrode surfaces
[88]. It should be noted that measured electrochemically, with a system probing
electron transfer to electrodes, is expected to be smaller than theoretical calcu-
lations in solution without electrodes, since the electrode approach decreases the
outer-sphere reorganization's (
out
) solvent contribution [88, 89]. In addition, the
experiments considered here were performed in air under ambient conditions.
Dierent C
electrode
and parameters are expected for the two dierent exper-
iments, since these parameters re
ect the electronic interaction and fraction of
voltage drop at the measurement contacts. In the transverse experiment, the
nanowire is supported on a
at conductive surface and a metallized AFM tip is
held just above the nanowire (applying a force in the nN range) [65]. The result-
ing transport behavior was t using C
electrode
= 1:8 10
10
s
1
. In the longitudinal
47
experiment, Pt electrodes are directly vacuum deposited onto the nanowire sur-
face, and this intimate electronic contact was re
ected with C
electrode
= 5 10
11
s
1
. It is worth noting that both values are orders of magnitude higher than previ-
ously measured heterogeneous transfer rate constants of Shewanella cytochromes
[16]. The reason for this wide discrepancy is not clear, but it may be attributed
to the very dierent experimental conditions, namely the dry xed environment
of the microbial nanowire measurements, and the likelihood that the system does
not adopt its native conformation under these conditions. By examining , we
estimate that 30% of the overall voltage drop happened at the electrode-wire con-
tact ( = 0:3) in the transverse measurement across 10 nm, compared to 1:5%
( = 0:015) in the measurement along a 2m nanowire. Indeed, it is reasonable to
expect that for a longer redox chain the local voltage drop at the electrode should
be smaller, relative to the total voltage, than in a shorter chain.
Furthermore, we previously measured the contact resistance for a 600-nm long
wire [32], and found that the corresponding local voltage drop at the electrode is
7% of the total applied voltage, i.e. giving an experimentally determined = 0:07
for a 600-nm chain; an intermediate value between the 10 nm and 2m predictions,
further conrming the expected trend. With the t parameters described here (Fig.
3.5 and Table 3.1), it is possible to calculate the voltage threshold corresponding to
the saturation of transport due to the maximum overlap between the metallic Fermi
function and the Gaussian redox peaks (Fig. 3.2 and 3.4). The saturation would
Measurement R (nm) L (eV) C
electrode
(s
1
) (

A
1
) k
B
T (eV)
Fig. 3.5(a) - 10 nm transverse 0:65 10 0:4 0:3 1:8 10
10
1 0:025
Fig. 3.5(b) - 2m longitudinal 0:65 2000 0:4 0:015 5 10
11
1 0:025
Table 3.1: Simulation parameters used to model the transverse and longitudinal
measurements (Fig. 3.5).
48
Figure 3.6: Sensitivity of the calculation to variations in the reorganization energy
while keeping the rest of the parameters the same as Fig. 3.5/Table 3.1. (a)
Transverse transport across the thickness of a microbial nanowire, using a con-
ductive tip as the top electrode and a supporting surface as the bottom electrode
(Calculation parameters: L = 10, R = 6:5, = 0:35 0:45 eV, = 0:3,
C
electrode
= 1:8 10
10
s
1
, k
B
T = 0:025 eV). (b) Longitudinal transport along a
microbial nanowire bridging two Pt electrodes (Calculation parameters: L = 2000,
R = 6:5, = 0:350:45 eV, = 0:015, C
electrode
= 5 10
11
s
1
, k
B
T = 0:025 eV).
[34] - Reproduced by permission of the PCCP Owner Societies.
be expected at 2:5 V and 50 V for the transverse and longitudinal measurements,
respectively. These predictions are useful for designing future experiments to test
and improve the model, provided that the high voltages required do not induce
physical damage in the biological structures under study.
Next, we examine the sensitivity of the calculation to the model parameters.
The assumption made that C
electrode
is smaller than k
hop
places an upper limit on
the eective tunneling distanceR. While this assumption should be easily met for
most reasonable distances that sustain tunneling because of the normally very small
heterogeneous transfer rates [16] (discussed above), using our higher t values for
C
electrode
( 10
10
s
1
), and comparing withk
hop
, suggests an upper limitR 6:9.
49
Figure 3.7: Best ts to the transverse transport measurement (c-AFM) by simul-
taneously adjusting the reorganization energy and the eective heterogeneous
transfer rate C
electrode
. The model t diverges from the experimental data at high
bias for< 0:3 eV and> 1 eV, regardless of the assumed heterogeneous transfer
rate. [34] - Reproduced by permission of the PCCP Owner Societies.
With a typical protein decay coecient of 1

A
1
, this translates to an eective
tunneling distanceR< 0:69 nm (i.e. taking the van der Waals contact distance to
be 0:35 nm, the nearest neighbor hopping distance is less than 1:04 nm). This result
is consistent with the previous nding by Polizzi et al. [36], and is also met by the
experimentally determined heme separation distances in Shewanella's multiheme
cytochromes [17]. As noted above, in this regime the predicted current response
follows the form of the heterogeneous electron transfer terms equation 3.5 and
3.6, which makes the calculated IV behavior very sensitive to and C
electrode
.
Fig. 3.6 illustrates this sensitivity by showing the disagreement between model
50
and experiment for slightly lower and higher reorganization energies, = 0:35
and 0:45 eV respectively, while keeping C
electrode
xed. However, it is possible to
compensate for the variation of the reorganization energy and obtain a good t to
the experimental data by tuningC
electrode
simultaneously, but only within the range
0:3 eV< < 1 eV, outside which we observe disagreement between calculations
and experiments in the high bias region as shown in Fig. 3.7.
Understanding the theoretical basis of extracellular charge transfer in micro-
bial redox chains has signicant implications for elucidating the natural respira-
tory strategies employed by important environmental bacteria. In addition, this
understanding can be harnessed towards dening the limitations and realizing the
untapped potential of emerging technologies (such as biofuel cells) where these
bacteria are employed as electrode-bound catalysts for driving redox reactions. In
light of the multistep hopping model described here (Fig. 3.1(a)), we note that
ecient EET requires a favorable interaction with electrodes (described by the
heterogeneous transfer rates), which may be a limiting factor in real devices, moti-
vating us to investigate new anode materials that maximize the electronic coupling
with redox molecules. Ecient EET also requires small separations between redox
sites within the chain ( 1 nm) [17]; a requirement that seems to be satised
by the heme-to-heme distances of individual outer membrane cytochromes. How-
ever, this hopping picture appears to be sensitive to possible structural defects in
one dimension, and microbes may address this limitation through redundancy and
multiple redox chains spread over cell surfaces and extracellular appendages within
three-dimensional biolms attached to electrodes.
51
3.4 Conclusion
In summary, we presented a model to describe charge transfer in microbial redox
chains between electrodes. After reviewing the applicability of coherent band
conduction and incoherent hopping to the transport measurements of microbial
nanowires, we excluded band conduction as a viable model of charge transport
in these structures because the measured mobilities are far below the minimum
mobility set forth by the coherent mechanism. In contrast, the incoherent hop-
ping mechanism can account for the observed transport rates as well as the form
of the current-voltage curves from transverse and longitudinal measurements of
microbial nanowires. The proposed model is based on an incoherent multistep
hopping mechanism between redox sites, and an interfacial treatment of the elec-
trochemical interactions with the measurement electrodes. Using this model, we
computed the occupation probability prole throughout the redox chain, which
allowed the calculation of the current response to applied voltage between the
electrodes. We found the results to be in good agreement with two previously
reported experiments measuring transport in microbial nanowires produced by the
bacteriumShewanella oneidensis MR-1. Furthermore, the t parameters were con-
sistent with the length scales and electrode contact conditions of each experiment,
as well as the typical reorganization energies expected from c cytochromes known
to be critical for this organism's extracellular charge transfer ability. Our analysis
motivates further experimental and theoretical investigations into the identity and
structural organization of redox components in microbial systems with signicant
environmental and biotechnological implications.
52
Chapter 4
In vivo Imaging of Bacterial
Nanowires of Shewanella
oneidensis MR-1 and Their
Composition and Structure
4.1 Background
A number of fundamental issues surrounding bacterial nanowires remain unre-
solved. Bacterial nanowires have never been directly observed or studied in vivo.
Our direct knowledge of bacterial nanowire conductance is limited to measurements
made under ex situ dry conditions using solid-state techniques optimized for inor-
ganic nanomaterials [30{32, 47], without demonstrating the link between these
conductive structures and the respiratory electron transport chains of the living
cells that display them. Intense debate still surrounds the molecular makeup, iden-
tity of the charge carriers, and interfacial electron transport mechanisms respon-
sible for the high electron mobility of bacterial nanowires. Geobacter nanowires
are thought to be type IV pili, and their conductance is proposed to stem from
a metallic-like band transport mechanism resulting from the stacking of aromatic
amino acids along the subunit PilA [90]. The latter mechanism, however, remains
53
controversial [45, 83]. In contrast, the molecular composition of bacterial nanowires
from Shewanella, the best-characterized facultatively anaerobic metal reducer, has
never been reported. Shewanella nanowire conductance correlates with the abil-
ity to produce outer membrane redox proteins [32], suggesting a multistep redox
hopping mechanism for EET [34, 36].
In this chapter, we address these outstanding fundamental questions by ana-
lyzing the composition and respiratory impact of bacterial nanowires in vivo. We
report an experimental system allowing real-time monitoring of individual bacte-
rial nanowires from living Shewanella oneidensis MR-1 cells and, using
uorescent
redox sensors, we demonstrate that the production of these structures correlates
with cellular reductase activity. Using a combination of gene expression analysis
and live
uorescence measurements, we also nd that the Shewanella nanowires
are membrane- rather than pilin- based, and are associated with outer membrane
vesicles. Our data point to a general strategy wherein bacteria extend their outer
membrane and periplasmic electron transport components, including multiheme
cytochromes, micrometers away from the inner membrane.
4.2 Experimental
4.2.1 Cell growth conditions
All Shewanella strains were grown aerobically in 50 mL of sterile LB broth (Sigma
L3022, 20 g in 1 L of deionized water) in 125-mL
asks from a frozen (80

C)
stock, at 30

C, shaking at 150 rpm up to an OD600 of 2:4 2:8. 15 mL of the
preculture was centrifuged at 4226g for 5 min, the pelleted cells were washed
twice, and nally resuspended in 45 mL of a dened medium consisting of 30 mM
54
Figure 4.1: Perfusion chamber used in the in vivo imaging of nanowires. Chamber
is closed with a glass coverslip and placed on an inverted
uorescence microscope.
Laminar
ow is maintained within the rectangular area at the center of the cham-
ber.
PIPES, 60 mM sodium DL-lactate as an electron donor, 28 mM NH
4
Cl, 1:34 mM
KCl, 4:35 mM NaH
2
PO
4
, 7:5 mM NaOH, 30 mM NaCl, 1 mM MgCl
2
, 1 mM CaCl
2
,
and 0:05 mM ferric nitrilotriacetic acid. In addition, vitamins, amino acids, and
trace mineral stock solutions were used to supplement the medium as described
previously [31]. The medium was adjusted to an initial pH of 7:2.
4.2.2 Perfusion chamber platform
Perfusion chamber
A laminar
ow Vacu-Cell incubation chamber (C&L Instruments, model VC-LFR-
25) was used in all perfusion and live microscopy experiments (Fig. 4.1). The
chamber was constructed from black polyether ether ketone (PEEK, custom order)
and modied with an additional port (0:0805 in), besides the inlet and outlet ports,
that housed an Ag/AgCl reference electrode (not used in the experiments described
in this thesis). An inverted 135-mL serum bottle was used to supply medium to
the perfusion chamber, and a bubble trap (Omnit, model 006BT or 006BT-HF)
55
placed upstream of the chamber (Fig. 4.2 and Fig. 4.3). After every time the
bubble trap and tubings were exposed to the NanoOrange or the RedoxSensor
Green stains, the bubble trap membrane as well as all tubings were changed and
the opened bubble trap was sonicated for 10 min while immersed in ethanol to
remove residual stain. Tygon S3 tubing, ID=1/32 in and OD=3/32 in, from Cole
Parmer Instrument Co was used throughout. The
ow of media from the serum
bottle was established either by peristaltic pumping or by pressurizing the supply
bottle's headspace with air or N
2
using a 6-inch septum needle, depending on
whether aerobic or anaerobic conditions were desired, while the serum bottle was
held slightly higher than the end of the outlet tube of the chamber. The
ow
rate of media through the perfusion chamber was obtained by measuring the time
interval between two consecutive drops at the end of the chamber outlet tube.
Using a vacuum line, the perfusion chamber was sealed against a 4350 mm No.1
coverslip (Gold Seal catalog #: 3329) while
ow of media had started and the
main chamber area was covered with liquid, forming a small (250m deep, 17:4
L volume) laminar
ow rectangular chamber (Fig. 4.1).
The entire perfusion set-up was taped on a Nikon Eclipse Ti-E inverted micro-
scope stage (Fig. 4.2 and Fig. 4.3), equipped with a drift correction unit (Nikon
Perfect Focus System) for maintaining focus at the coverslip-medium interface
during time-lapse imaging. Washed cells were slowly injected into the perfusion
chamber (d = 250m, w = 9; 580m, L = 7; 110m) using a 10-mL syringe
connected to the chamber inlet tube by a 4-way stopcock (Value Plastics, catalog
#: VPB1000110N) and the
ow was stopped to allow the cells to settle on the
coverslip with a surface density of 100 cells in an 83 66m eld of view
while being imaged by the inverted microscope. After inoculation, sterile dened
56
Figure 4.2: Schematic of the perfusion imaging platform. Perfusion chamber is
placed on an inverted microscope stage above the objective while medium
ows
through the chamber from a serum bottle (left) and passes through a bubble trap.
Cells are injected using a syringe connected to the inlet tube with a stopcock.
Waste from the chamber
ows into the waste container (right).
medium (100 mL in an inverted 135mL sealed glass serum bottle) was connected
to the perfusion chamber inlet and passed through at a
ow rate of 51L/s that
remained constant for 3 h.
Oxygen limitation
As previously detailed [31, 91], transition to electron-acceptor limitation was
required to promote the production of bacterial nanowires and membrane vesi-
cles; this was accomplished using one of two methods. In the rst method, the
57
Figure 4.3: Perfusion imaging platform. Perfusion chamber is placed on an inverted
microscope stage above the objective while medium
ows through the chamber
from a serum bottle (left) and passes through a bubble trap. Cells are injected
using a syringe connected to the inlet tube with a stopcock. Waste from the
chamber
ows into the waste container (right).
experiment is started with fully aerobic medium
ow before switching to supply
bottles containing medium made anaerobic by boiling and purging with 100%N
2
(1 L of media boiled in a 2-L
ask and transferred using a pump into serum bottles
that were previously
ushed with N
2
for 15 min, the media in serum bottles were
then bubbled with N
2
for one hour). This method has the advantage of providing
a precise time point for entering acceptor (O
2
) limitation (Fig. 4.10A and Movies
S1 and S4 in [92]). In the second method,O
2
becomes limited close to the coverslip
surface at high cell density, even with aerobic medium, as a result of the laminar

ow (no mixing between adjacent layers) and no-slip condition (zero velocity at
the surface); this is conrmed with the simple calculation outlined below.
58
Figure 4.4: Schematic of laminar
ow of medium inside the perfusion chamber
along. Oxygen becomes limiting close to the coverslip due to the celluar consump-
tion, in turn causing the production of nanowires.
For plane Poiseuille
ow between two parallel plates (w = 9; 580m) separated
by distance d (250m) (Fig. 4.4), the volumetric
ow rate (5L/s in our experi-
ment) is given by Gwd
3
=12, where is the viscosity, allowing us to estimate G,
the pressure gradient in the
ow direction x. In addition, the
uid velocity prole
between the plates is given byGy(dy)=(2), where 0yd. Since the cells are
located at the coverslip (y = 0), we assume that the cells interact with the 1m-
thick layer of media adjacent to the coverslip (no mixing from upper layers) where
the average
ow velocity v is 25m/s. The O
2
concentration will consequently
diminish in the x direction according to C(x) = C
0
x=v, where C
0
is the
inlet concentration ( 8 mg/L at 100% of air saturation) and is the cellular con-
sumption rate which we estimate from previous studies of per cell respiration rates
[32] and our surface cell density to be 0:63 mg/L/s. From this calculation, we nd
59
Figure 4.5: Concentration of molecular oxygen and lactate along the length (L in
Fig. 4.4) of the perfusion chamber close to the coverslip. Oxygen quickly becomes
limiting as medium
ows from left to right inside the chamber.
that the O
2
concentration drops below 10% of air saturation at x = 290m and
reaches 0% atx = 320m (Fig. 4.5). A similar analysis for lactate reveals that the
donor concentration only drops to 96:9% of its inlet value (60 mM) at the outlet
(x = 7; 110m) (Fig. 4.5). In other words, the majority of the surface-bound cells
are acceptor, rather than donor, limited.
Using method 2 described above, nanowires and membrane vesicles were con-
sistently observed after a lag period of 90-120 min from the start of perfusion
ow
60
(Movies S2 and S3 in [92]). We observed higher nanowire production rates and bet-
ter consistency using this method, possibly because the slower transition exposed
the surface-bound cells to a wide range of acceptor concentrations, compared with
the abrupt transition in the rst method. Both methods resulted in identical mem-
brane extensions as observed using membrane
uorescence (FM 4-64FX), protein

uorescence (NanoOrange), and the accompanying increase in reductase activity
(RedoxSensor Green).
Fluorescence microscopy
To visualize the nanowires/vesicles by
uorescence microscopy, either the protein
stain NanoOrange (Life Technologies) (Fig. 4.12A) or the membrane stain FM
4-64FX (Life Technologies) (Fig. 4.12A) was used. For each experimental run,
we added 100L of the NanoOrange reagent or 25g of FM 4-64FX (dissolved
in 200L of deionized water) to the autoclaved serum bottle containing the 100
mL of perfusion medium. Both of these
uorescent dyes allowed staining of cells,
membrane vesicles, and bacterial nanowires.
Fluorescence imaging of NanoOrange and FM 4-64FX was done in the
FITC (Nikon lter set B-2E/C) and TRITC (Nikon lter set G-2E/C) excita-
tion/emission channels, respectively. The exposure time in both cases was 500 ms.
Both channels were used at intensity level 2. The excitation and emission windows
for both channels are described in Table 4.1. NanoOrange
uorescence was visible
in both FITC and TRITC, whereas FM 4-64FX was only bright in TRITC.
Since NanoOrange has signicant emission in the TRITC channel, we were not
able to simultaneously stain structures with FM 4-64FX and NanoOrange. We
began with FM 4-64FX in the medium, and imaged the nanowires produced by
61
either wild-type MR-1 or the pilus-deletion mutant strains (Table 4.2) in TRITC.
We then stopped the medium
ow, manually injected 10 mL of NanoOrange solu-
tion (20L NanoOrange reagent diluted in cell medium) using a syringe connected
to the stopcock, and imaged the same structures again in the FITC channel.
4.2.3 RedoxSensor Green assay
RedoxSensor Green (from BacLight
TM
RedoxSensor
TM
Green Vitality Kit, Life
Technologies, catalog # B34954) was used to assess cellular respiration activity.
Because NanoOrange and RedoxSensor Green (RSG) reagents both
uoresced in
the same channel (FITC), we could not use NanoOrange and RSG simultaneously
in the same experiment. Instead, to test the impact of nanowire production on cel-
lular reductase activity, we used FM 4-64FX and RSG to simultaneously visualize
the nanowires and track the reductase activity, respectively. In these experiments,
0:3M RSG (30L of 1 mM RSG reagent in 100% DMSO) was added to the 100
mL of
ow medium in the supply bottle, along with FM 4-64FX.
To measure the intensity of RSG in cells, brightness and contrast of images were
linearly adjusted, equally for all cells, to capture the entire range of RSG intensity
Filter set name
Excitation
Filter Wave-
lengths
Dichromatic
Mirror Cut-on
Wavelength
Barrier Filter
Wavelengths
B-2E/C - `FITC'
465 495 nm
(bandpass, 480
CWL)
505 nm (longpass,
LP)
515 555 nm
(bandpass, 535
CWL)
G-2E/C -
`TRITC'
528 553 nm
(bandpass, 540
CWL)
565 nm (longpass,
LP)
590 650 nm
(bandpass, 620
CWL)
Table 4.1: Wavelength windows of the
uorescence channels used in the perfusion
chamber experiments.
62
in all cells. The images were then exported to MATLAB where the average pixel
intensity (API, arbitrary units) of every cell was extracted by manual selection
of the cell periphery and calculating the API of the pixels inside the selected
periphery.
4.2.4 Verication of RedoxSensor Green as a probe for
active respiration in Shewanella
The redox sensing functionality of RedoxSensor Green (RSG) was veried in She-
wanella by observing
uorescence change in response to substrate activation (Fig.
4.6), as previously demonstrated for environmental samples [48]. S. oneidensis
Strain Description
Source
MR-1 Shewanella oneidensis, wild-type
[8]
mtrC=omcA MR-1 mtrc, omcA
[31]
pilA MR-1 pilA
This work
pilMQ=mshHQ
MR-1 pilMNOPQ,
mshHIJKLMNEGBACDOPQ
[93]
Plasmid Description
Source
pHGE-PtacTorAGFP
Km
R
, Ptac, TorA
sp
-GFP (periplas-
mic GFP)
[94]
p519ngfp
Km
R
, Plac/Pnpt-2, mob
+
, GFP
(cytoplasmic GFP)
[95, 96]
pPROBE-NT
Km
R
, broad host-range expression
of GFP
[97]
pHydA-YFP
Km
R
,gfp replaced withhydAyfp
in pPROBE-NT
This work
Table 4.2: Strains and plasmids used in this study.
63
Figure 4.6: Verication of RedoxSensor Green functionality by comparing starved
and electron-donor-activated Shewanella oneidensis MR-1 cells. A starved culture
was divided into two subcultures, one of which remained starved while the other
was activated with lactate as an electron donor. (A) Quantitative bar plot of

uorescence intensity of RSG in cells from subculture with electron donor (gray)
and without electron donor (red) at various time points. Control refers to the
initial sample that was starved for 19 hours. For each time point, images were col-
lected from six random elds of view. The bar plot shows the average
uorescence
SEM. (B) Representative re
ection (left) and RSG
uorescence (right) images
of cells from the control sample that was starved for 19 hours (top panel) and
the subculture with added lactate after 7 hours of activation with lactate (bottom
panel). The lactate-activated cells are all visible with
uorescence, in contrast to
the starved cells not visible with
uorescence.
MR-1 cells were taken from LB culture in mid-log phase (OD600 = 1:5), washed
twice and starved by incubation in dened medium without lactate or any other
electron donor, for 19 hours at 30

C and shaking at 150 rpm in 125-mL
asks.
Following starvation, the cells were split into two subcultures. The rst subcul-
ture was activated with 20 mM lactate, while the second culture remained starved.
Both subcultures were sampled at multiple time points (1, 4, 5.5, and 7 hours after
starvation) and imaged with the same microscope settings (
uorescence channel,
exposure time and intensity) following incubation with 1M RSG reagent at 30

C
64
for 40 min. Six randomly selected elds of view were imaged and analyzed at each
time point.
Fluorescence intensity of cells was analyzed by image processing functions in
MATLAB. The positions and exact periphery of cells were located using re
ec-
tion mode (without
uorescence) and used to create a mask. The mask from the
re
ection image was then t on top of the RSG
uorescence image to remove
pixels lying outside the cells. The average pixel intensity (API) of the masked

uorescence image was calculated and averaged over the six elds of view, for both
starved and activated samples at each time point.
The eect of three electron transport chain (ETC) inhibitors on RSG
uo-
rescence was monitored to further conrm and localize the redox sensing func-
tionality of RSG in Shewanella, as previously reported for other bacteria [49].
Rotenone, antimycin A, and sodium azide were selected because they aect respi-
ration by inhibiting dierent ETC components [98]. Mid-log S. oneidensis MR-1
cells (OD600 = 1:5) from LB cultures were washed twice and incubated in the
dened medium with 1 mM rotenone, 20M antimycin A, 10 mM sodium azide,
or no inhibitor (control) for 15 min at 30

C. The RSG reagent (1M) was added
to each sample and incubated for 30 min at 30

C. Cells from all samples were
placed on a coverslip and imaged by
uorescence microscopy (Fig. 4.7) using the
same settings (
uorescence channel, exposure time and intensity).
4.2.5 Cytoplasmic and periplasmic green
uorescent pro-
tein (GFP) imaging
To express GFP in the cytoplasm, S. oneidensis MR-1 was transformed with plas-
mid p519ngfp [95, 96]. The resulting strain was grown in LB augmented with
65
Figure 4.7: The eect of electron transport chain inhibitors on RedoxSensor Green

uorescence. Shewanella oneidensis MR-1 cells were incubated with no ETC
inhibitor (control), 1 mM rotenone, 20M antimycin A, or 10 mM sodium azide.
Following staining with RedoxSensor Green (RSG), both re
ection and RSG
u-
orescence images were collected. The top panel shows the control sample with-
out any ETC inhibitor. The second panel from the top shows cells inhibited by
rotenone, with cells signicantly darker than the control sample. The third panel,
from the sample inhibited by antimycin A, also shows cells darker than the con-
trol. These images indicate that RedoxSesnor Green is capable of detecting the
drop in respiration activity caused by rotenone and antimycin A inhibiting the
ETC. However, cells inhibited by sodium azide are as bright as the control cells,
indicating that RSG interacts with the ETC at a point downstream of antimycin
A's site of inhibition (cyt bc1 complex), but upstream of azide's site of inhibition
(cyt c oxidase).
66
Figure 4.8: Cells containing periplasmic and cytoplasmic Green Fluorescent Pro-
tein (GFP). (A) Fluorescence image of Shewanella oneidensis MR-1 constitutively
expressing GFP with no signal sequence for export, resulting in a uniform cellular

uorescence pattern throughout the cytoplasm. (B) Shewanella oneidensis MR-1
expressing GFP fused with the twin-arginine translocation (Tat) signal peptide
from the E. coli TorA. Cells were imaged on an agar pad following LB growth
and IPTG induction. Due to the presence of the signal peptide, the fusion pro-
tein is exported to the periplasm, resulting in
uorescence limited primarily to the
periphery of the cell. Scale bars are 5m.
50g/mL kanamycin ahead of use in the perfusion
ow imaging experiment as
described above for the S. oneidensis MR-1. No kanamycin was added to the
washing or perfusion medium.
To localize GFP to the periplasm, S. oneidensis MR-1 was transformed with
plasmid pHGE-PtacTorAGFP (Table 4.2) as described previously [94]. pHGE-
PtacTorAGFP is an IPTG-inducible plasmid expressing GFP fused to the twin-
arginine translocation (Tat) signal peptide from E. coli TorA. Fusion to the Tat
signal peptide enables GFP to be exported to the periplasm [94]. ThisS.oneidensis
periGFP strain was grown in LB with 50g/mL kanamycin, and 0:1 mM IPTG
67
was added to induce the expression of TorA
sp
-GFP at OD600 = 0:4 ahead of use
in the perfusion imaging experiment as described above. Since the Tat system
exports fully folded proteins from the cytoplasm, the cellular
uorescence pattern
of thisS.oneidensis periGFP strain is not necessarily limited to the periplasm only,
and can change over time depending on the induction, export rate, or post-export
cleavage of the signal peptide [94]. Therefore, to assess the successful periplasmic
localization of GFP prior to use in the perfusion experiments (Fig. 4.12B), we
imaged this strain on 3% agar on a slide (no nutrients or further IPTG induction)
which resulted in the characteristic peripheral
uorescence pattern consistent with
periplasmic localization (Fig. 4.8).
4.2.6 Chemostat growth and qPCR analysis of the transi-
tion from electron-donor to electron-acceptor limita-
tion
S. oneidensis MR-1 was grown in chemostat medium (Table 4.3) at 30

C, pH 7.0,
and 20% dissolved oxygen tension (which was adjusted using controlled air andN
2
input) using continuous
ow bioreactors (BioFlo 110; New Brunswick Scientic)
(Fig. 4.9) with a dilution rate of 0:05 h
1
and an operating liquid volume of 1 L,
as previously described [31]. After 48 h of this aerobic growth, a reference sample
was taken (each sample was 10 mL). The dissolved oxygen tension was then
manually dropped to 0% by adjusting the N
2
/air mixture entering the reactor,
and the rst sample after electron acceptor limitation (t = 0) was taken at this
time. O
2
served as the sole terminal electron acceptor throughout the experiment.
Samples were subsequently taken at 15-min intervals for 1 h. At each time
point, 10 mL of cells were harvested in ice-cold 5% citrate-saturated phenol in
68
Figure 4.9: Picture of a chemostat used for analyzing the expression levels of
various genes before and after production of bacterial nanowires by S. oneidensis
MR-1 cells.
69
ethanol to prevent further transcription and protect the RNA. Samples were taken
from three independent biological replicates (dierent chemostats).
The gene expression proling was done as part of a larger collaboration with
Sarah Barchinger and John Golbeck (at Pennsylvania State University, University
Park) [92]. Total RNA was prepared using a hot phenol extraction, as previ-
ously described [99]. Five micrograms of total RNA from each of the time points
described above was used as input for reverse-transcriptase reactions with the
SuperScript III First-Strand Synthesis System, as per the manufacturers protocol
(Life Technologies). Subsequent cDNA was then diluted and used as template in
qPCR experiments with SYBR Select Master Mix (Life Technologies).
Fold change in gene expression relative to the reference sample was calculated
by 2
CT
from at least four reactions of three independent biological replicate
Ingredient FW g/L
Final con-
centration in
media (mM)
PIPES buer 302.4 0.91 3
Sodium hydroxide 40 0.3 7.5
Ammonium chloride 53.49 0.53 9.9
Potassium chloride 74.55 0.1 1.34
Potassium phosphate monobasic 136.09 0.6 4.4
Sodium sulfate 142.04 0.19 1.34
Sodium lactate (60%) 112.06 3.7 19.8
Minerals solution [31] 10 mL
Amino acids solution [31] 10 mL
Vitamins solution [31] (after
autoclaving)
10 mL
MgCl
2
(after autoclaving) 95.211 0.095 1
CaCl
2
(after autoclaving) 110.98 0.11 1
Ferric nitrilotriacetic acid (after
autoclaving)
0.05
Table 4.3: Chemostat medium.
70
samples, using recA for normalization. Similar results were obtained using rpoB
for normalization. The sequences of the primers used for each gene are shown in
Table 4.4.
4.2.7 Strains and plasmids generated for this study
pilA: This strain was provided by our collaborators Rachida Bouhenni and Daad
Saarini (at University of Wisconsin, Milwaukee) [92]. To generate a chromoso-
mal deletion of pilA, 1 kb DNA fragments that
ank this gene were amplied
using the primers pilAF (5'- GCATTGGCATGTCGATGAT), pilAR (5'- ATG-
TAAGCCTGTGGTGGGCATTTTCTCGCTCCAATACAG), pilBF (5'- CCCAC-
CACAGGCTTACAT), and pilBR (5'- TTCGCCCACCATTACCAC). The 1 kb
Primer
Name
Primer Sequence (5'-3')
RecAQF AGCTATAGCCGCTGAAATCG
RecAQR CCTCGACATTGTCATCATCG
RpoBQF AGGCGTATTCTTCGATCACG
RpoBQR AACCATGAACCACGGTAAGG
MtrCQF CTCAAGAGTTTGCGGATGGT
MtrCQR CATGTCGGATTCAACGTGAC
MtrAQF CGGCACTTACCATCACAATG
MtrAQR ATCCCACTTCGACGCATAAG
OmcAQF AACTGTGCATCTTGCCACAC
OmcAQR TCGCCACCTTTATGGATAGC
PilAQF GAAAGGCCTTGCAGGATATG
PilAQR TGCTTCAGCAACATCAGGAG
MshAQF TGCGTCTGCATTACAAGGTC
MshAQR TCTCCCAATGAAAAGGTTGG
PilEQF TGGCCAATCTACAAGAGCAG
PilEQR TCACTGAGTAATTACCGTTTTCG
FimTQF CACGTAACACCGCAATCAAC
FimTQR CTTCAGTGCATTGACCATCG
Table 4.4: qPCR Primers.
71
fragments were used as template to amplify a 2 kb fragment using cross-over PCR.
The resulting 2 kb DNA that lacks pilA was cloned into pER21 and used to gen-
erate chromosomal deletion mutant as described previously [93].
pHydA-YFP: Also as part of our collaboration, this fusion protein was
provided by Samantha Reed and Margaret Romine (at Pacic Northwest
National Lab) [92]. pProbeNT-YFP was constructed by replacing the
EcoRI-NsiI fragment encoding GFP in pProbeNT [97], with the EcoRI-
NsiI fragment encoding YFP from pEYFP-C1 (Clonotech laboratories). A
PCR product encoding hydA gene (SO 3920) and its upstream promoter
was generated by amplication of Shewanella oneidensis MR-1 genomic
DNA with primers 5'-gcgcTCTAGAGTCGACCATGCCGATAATCT-3' and 5'-
gcgcGAGCTCGTTACCCAGCCATGAAGAGC-3'. The PCR product was
digested with SstI and XbaI (sites underlined in primer sequence) and then cloned
into same sites of pProbeNT-YFP, yielding pHydA-YFP. S. oneidensis MR-1 was
transformed with this plasmid and the resulting strain, designated S. oneidensis
HydA-YFP, was grown in LB augmented with 50g/mL kanamycin ahead of use
in the perfusion
ow imaging experiment as described above for the wild-type. No
kanamycin was added to the washing or perfusion medium.
4.3 Results and discussion
4.3.1 In vivo imaging of nanowire formation
Previous reports demonstrated increased production of bacterial nanowires and
associated redox-active membrane vesicles in electron acceptor (O
2
)-limited She-
wanella cultures [31, 32, 91]. To directly observe this response in vivo, we subjected
72
Figure 4.10: In vivo observations of the formation and respiratory impact of bac-
terial nanowires in S. oneidensis MR-1. (A) A leading membrane vesicle and the
subsequent growth of a bacterial nanowire observed with
uorescence from the
protein stain NanoOrange. Extracellular structure formation was rst observed in
the t = 0 frame, captured 20 min after switching from aerobic to anaerobic perfu-
sion. (Scale bars: 5m.) (B) Combined green (RedoxSensor Green) and red (FM
4-64FX)
uorescence images of a single cell before and after (35 min later) the pro-
duction of a bacterial nanowire. Movie S6 in [92] shows the time-lapse movie of B.
(Scale bars: 5m.) (C) The time-dependent RedoxSensor Green
uorescence for
nanowire-producing cells, compared with neighboring cells that did not produce
nanowires (n = 13, mean cellular pixel intensitySE). The nanowires were rst
observed at the t = 0 time point.
Shewanella oneidensis MR-1 to O
2
-limited conditions in a microliter-volume lam-
inar perfusion
ow imaging platform (Fig. 4.2, Fig. 4.3 and Experimental section
above) and monitored the production and growth of extracellular laments from
individual cells with
uorescent microscopy. The cells and attached lamentous
appendages were clearly resolved (Fig. 4.10 and Movies S1-S4 in [92]) at the
73
surface-solution interface using NanoOrange, a merocyanine dye that undergoes
large
uorescence enhancement upon binding to proteins [70, 71]. This dye was
previously used to label bacterial nanowires recovered from chemostat cultures
[31, 100]. In all our experiments, the production of laments coincided with the
formation of separate or attached spherical membrane vesicles, another observa-
tion consistent with previous electron and atomic force microscopy measurements
of Shewanella nanowires [91]. In Fig. 4.10A, a leading membrane vesicle can
be clearly seen emerging from one cell 20 min after switching to anaerobic
ow
conditions, followed by a trailing lament. These proteinaceous vesicle-associated
laments were widespread in all of the S. oneidensis MR-1 cultures tested; the
response was displayed by 65 8% of all cells (statistics obtained by monitoring
6; 466 cells from multiple random elds of view in six separate biological repli-
cates). As a representative example, Movie S5 in [92] shows an 83 66m area
where the majority of cells produced the laments. The length distribution of the
laments is plotted in Fig. 4.11, showing an average length of 2:5m and reaching
up to 9m (100 randomly selected laments from six biological replicates).
The laments described here were the only extracellular structures observed in
our experiments, and possess several features of the conductive bacterial nanowires
previously reported in O
2
-limited chemostat cultures. Specically, the dimensions
of these laments [32, 65, 101], their association with membrane vesicles [91], and
their production duringO
2
limitation [31] led us to conclude that these structures
are the bacterial nanowires whose conductance was previously measured ex situ
under dry conditions [32]. Additionally, when we labeled cells grown inO
2
-limited
chemostat cultures with the same
uorescent dyes, we observed identical structures
with the same composition as the perfusion cultures reported here (see below).
74
Figure 4.11: Distribution of Shewanella oneidensis MR-1 nanowire length. His-
togram of nanowire length (100 randomly selected nanowires from 6 separate bio-
logical replicates). The average is 2:5m.
4.3.2 The production of S. oneidensis MR-1 bacterial
nanowires is correlated with an increase in cellular
reductase activity
To directly measure the physiological impact bacterial nanowire production has
on S. oneidensis, we labeled cells with RedoxSensor Green (RSG) in the perfusion
imaging platform described above. RSG is a
uorogenic dye that yields green
u-
orescence upon interaction with bacterial reductases in the cellular electron trans-
port chain, and has been previously demonstrated to be an indicator of active
respiration in pure cultures and environmental samples [48, 49, 78]. Because the
redox-sensing ability of RSG was not previously characterized in Shewanella, we
rst conrmed that electron donor (lactate)-activated respiration increases RSG
75

uorescence in aerobic cultures relative to starved controls (Fig. 4.6), and that the
addition of specic electron transport inhibitors abolishes RSG
uorescence (Fig.
4.7). S. oneidensis MR-1 cells displayed a signicant increase in RSG
uorescence
concomitant with nanowire production (Fig. 4.10B and C and Movie S6 in [92]),
indicating increased respiratory activity compared with nearby control cells that
did not produce nanowires in the same eld of view under identical perfusion con-
ditions. DMSO, which is respired extracellularly byShewanella [102], was available
as a terminal electron acceptor in all RSG-labeled experiments.
4.3.3 S. oneidensis MR-1 nanowires are outer membrane
and periplasmic extensions
Membrane vesicles have previously been observed to be associated with Shewanella
nanowires [91] (Fig. 4.10A). To test the extent of membrane involvement in
nanowire formation, we labeled S. oneidensis MR-1 cells with the membrane stain
FM 4-64FX. To our surprise, the entire length of the Shewanella nanowires was
clearly stained with this reliable lipid bilayer dye (Fig. 4.10B, red channel, and
Fig. 4.12), indicating that membranes are a substantial component of Shewanella
nanowires, contrary to previous suggestions that these structures are pilin based.
We stained cells producing both nanowires and membrane vesicles with NanoOr-
ange and FM 4-64FX, demonstrating that proteins and lipid colocalized on these
extracellular structures, consistent with being derived from the cell envelope (Fig.
4.12A).
Most known bacterial vesicles are composed primarily of outer membrane and
periplasm. To determine whether Shewanella nanowires contain periplasm, we
expressed either GFP fused to a signal sequence that enables GFP export to the
76
Figure 4.12: Bacterial nanowires contain lipids, proteins, and periplasm. (A)
Shewanella oneidensis MR-1 cells and attached bacterial nanowires, stained with
both NanoOrange (Upper) and FM 4-64FX (Lower), indicating the presence of
proteins and membranes, respectively. (Scale bars: 5m.) (B) Bacterial nanowires
from S. oneidensis MR-1 strains containing GFP only in the cytoplasm (Upper)
or in the periplasm as well (Lower). The green and red channels monitor GFP and
FM 4-64FX
uorescence, respectively. The nanowires display green
uorescence
only when GFP is present in the periplasm (Lower Left). (Scale bars: 2m.)
periplasm (Fig. 4.8) after folding in the cytoplasm [94] or YFP fused to the
S. oneidensis periplasmic [Fe-Fe] hydrogenase large subunit HydA (Experimental
section above). We observed
uorescence along the bacterial nanowires in both
constructs (Fig. 4.12B and Fig. 4.13). However, no
uorescence was detected along
nanowires from a strain expressing cytoplasmic-only GFP (Fig. 4.12B). Taken
together, these results indicate that Shewanella nanowires are outer membrane
extensions containing soluble periplasmic components.
It has previously been proposed, but never demonstrated, that pili are impor-
tant components of Shewanella nanowires. To test whether pili play a role in S.
77
Figure 4.13: Localization of the periplasmic [Fe-Fe] hydrogenase large subunit
HydA along bacterial nanowires. Images are of Shewanella oneidensis MR-1 cells
expressing HydA fused to yellow
uorescent protein (HydA-YFP) under the con-
trol of the native hydA promoter. The perfusion
ow imaging platform was used to
monitor bacterial nanowire production (arrows). The red (left) and green (right)
channels show
uorescence from the membrane stain FM 4-64FX and YFP, respec-
tively. Brightness/contrast adjustment and bilinear interpolation was applied
equally to the entire image on the right to highlight the HydA-YFP
uorescence.
Scale bars are 5m.
oneidensis nanowire production, we harvested RNA [99] from chemostat cultures
where O
2
served as the only electron acceptor, before and at time intervals after
transition from electron donor (lactate) limitation toO
2
limitation. Electron accep-
tor limitation is known to result in increased production of bacterial nanowires, as
previously demonstrated in chemostat cultures [31, 32] (and conrmed by electron
microscopy). We then used qPCR to measure changes in the expression of key
genes necessary for type IV pilus assembly: pilA, encoding the type IV major pilin
subunit; mshA, encoding the mannose sensitive hemagglutinin (msh) pilin major
subunit; andpilE andfimT , encoding type IV minor pilin proteins. Expression of
all these pilin genes either remained constant or decreased after electron acceptor
limitation, when nanowire production was observed (Fig. 4.14A). Furthermore,
mutants lacking the type IV pilin major subunit (pilA) or both the type IV and
78
Figure 4.14: The genetic and molecular basis of bacterial nanowires. (A) Expres-
sion of pilin genes in S. oneidensis MR-1 measured by qPCR. Chemostat cultures
were grown under aerobic conditions (dissolved oxygen tension of 20%) at 30

C for
48 h before the dissolved oxygen tension (DOT) was reduced to 0%. Samples were
harvested from cultures right before reducing the DOT (t< 0, reference sample),
at t = 0, and in 15-min intervals for 1 h. The fold change in gene expression rela-
tive to the reference sample was calculated by 2
CT
from at least four reactions
of three independent chemostat cultures, using recA for normalization. (B) Com-
bined green (RedoxSensor Green) and red (FM 4-64FX)
uorescence images of the
pilA strain, lacking the type IV pilin major subunit PilA, before and after (95
min later) the production of a bacterial nanowire. pilA is capable of producing
bacterial nanowires with a similar respiratory impact as wild-type S. oneidensis
MR-1, evidenced by the increase in reductase activity (green
uorescence) after
nanowire production. (Scale bars: 2m.)
msh pilus biogenesis systems (pilMQ=mshHQ) [93] produced bacterial
nanowires and displayed an increase in reductase activity in the perfusion imaging
platform (Fig. 4.14B and Fig. 4.15) - a response identical to wild-type S. oneiden-
sis MR-1. The chemostat qPCR and perfusion pilus-deletion observations both
support the conclusion that Shewanella nanowires are distinct from pili.
Why were the membrane and periplasmic components of these structures over-
looked in previous studies? One important factor is the diculty of isolating the
appendages and separating them from cells before downstream compositional anal-
ysis (e.g., liquid chromatography-tandem mass spectrometry). In fact, membrane
79
Figure 4.15: Nanowire production by pilMQ=mshHQ and its impact on
cell respiration indicated by RedoxSensor Green
uorescence. Images show com-
bined
uorescence from membrane stain FM 4-64FX (red) and RedoxSensor Green
(green) before (top) and after (bottom) nanowire production. The production of
extracellular structures was correlated with a sudden increase in redox sensing

uorescence, similar to the wild-type strain. Scale bars are 5m.
and periplasmic proteins were previously identied in a study focused on devel-
oping optimal methods for removing the Shewanella laments, although it was
not possible to rule these proteins out as an artifact of the shearing procedure
[103]. The present in vivo microscopy work circumvents some of the previous arti-
fact problems by
uorescently labeling specic cellular components (protein, lipid,
and periplasm) on intact nanowires attached to whole cells. In light of these new
results, we revisited the chemostat samples from our previous conductance study,
and noted that in those samples, SEM images also revealed vesicular morphologies,
80
Figure 4.16: O
2
-limited chemostat cultures contain membranous and vesicle-
associated laments. (A) Scanning electron microscopy of Shewanella oneiden-
sis MR-1 from an O
2
-limited chemostat which was previously used for nanowire
conductance measurements [32]. Scale bar is 500 nm. (B) Fluorescence image
of an O
2
-limited chemostat sample stained by the membrane stain FM 4-64FX.
A representative membranous lament is observed connecting two MR-1 cells.
Brightness/contrast adjustment and bilinear interpolation was applied equally to
the entire inset image to highlight the lament. Scale bar is 2m.
non-uniform cross-sections, and diameters > 5 nm (larger than pili; Fig. 4.16A
and the source-drain devices in [32]). These suggestive links did not become clear
until the present in vivo results. We xed samples fromO
2
-limited chemostat cul-
tures and labeled them with FM 4-64FX. The nanowires from chemostat cultures
contained both protein and membrane (Fig. 4.16B), further demonstrating that
these structures are the same as those observed in the perfusion cultures in vivo.
Though the net patterns of gene expression measured in the planktonic chemostat
cultures may dier from the surface-attached perfusion cultures, it is important to
stress that identical membrane extension phenotypes were observed in both these
experiments where O
2
served as the sole electron acceptor in limiting concentra-
tions.
81
4.4 Conclusion
Bacterial nanowires fromShewanella oneidensis MR-1 were previously shown to be
conductive under nonphysiological conditions. Intense debate still surrounds the
molecular makeup, identity of the charge carriers, and cellular respiratory impact
of bacterial nanowires. In this chapter, using in vivo
uorescence measurements
and quantitative gene expression analysis, we demonstrated that S. oneidensis
MR-1 nanowires are extensions of the outer membrane and periplasm, rather than
pilin-based structures, as previously thought. To study the physiological impact
of nanowire production on Shewanella cells, we also demonstrated that production
of bacterial nanowires correlates with an increase in cellular reductase activity. In
the next chapter, in order to examine the mechanism of electron transport in She-
wanella nanowires, we investigate the involvement of outer membrane cytochromes
along these laments.
82
Chapter 5
Cytochrome Localization and
Nanoscale Characterization of
Bacterial Nanowires
5.1 Introduction
As a metal reducer, Shewanella has evolved an intricate EET pathway to trac
electrons from the inner membrane, through the periplasm, and across the outer
membrane to external electron acceptors, including minerals and electrodes. This
pathway includes the periplasmic decaheme cytochrome MtrA, as well as the outer
membrane decaheme cytochromes MtrC and OmcA, which may interface to soluble
redox shuttles or, via solvent-exposed hemes, directly to the insoluble terminal
acceptors [17].
However, the presence of outer membrane cytochromes along bacterial
nanowires had never been directly shown before this work. Therefore, in light of the
structural nding that Shewanella nanowires are outer membrane and periplasmic
extensions (chapter 4), in this chapter we will examine whether expression of these
periplasmic and outer membrane cytochromes increases during nanowire produc-
tion in S. oneidensis bioreactors that transition from electron donor-limitation to
electron acceptor-limitation.
83
In addition, we will use immuno
uorescence to visualize and indicate the pres-
ence of specic outer membrane cytochromes along bacterial nanowires produced
in the in vivo imaging platform described in chapter 4. After this demonstration
we will examine the electron transport characteristics of single molecules of outer
membrane cytochromes. This basic knowledge can serve as the buiding block for
future large scale models describing the long range transport physics in bacte-
rial nanowires. To this end, we demonstrate the development of an experimental
approach to examine the basic physics of electron transfer in individual molecules
of the MtrC homolog, MtrF, which is the rst outer membrane cytochrome with
a solved crystal structure. Monolayers of MtrF are produced and the mono-
layer properties are investigated using scanning tunneling microscopy (STM). The
approach developed here can be used in future in combination with tunneling
spectroscopy (TS) I-V measurements to examine electron transfer characteristics
of individual MtrF molecules.
During in vivo imaging of bacterial nanowire formation, ocassionally we
observed the initial growth of a vesicle chain that later transitioned into a smoother
lament. In addition, it had been previously shown that bacterial nanowires are
typically associated and attached to outer membrane vesicles in Shewanella [91].
To investigate the involvement of outer membrane vesicles in nanowire forma-
tion, and to gain a more accurate estimate of the morphologies and dimensions
of bacterial nanowires, we performed atomic force microscopy (AFM) as well as
transmission electron microscopy (TEM) on nanowires previously produced in the
in vivo imaging platform described in chapter 4. This technique revealed the inter-
mediate steps in the formation of nanowires and demonstrated the true dimensions
of these laments.
84
5.2 Experimental
5.2.1 Immuno
uorescence with MtrC or OmcA antibody
S. oneidensis MR-1 and omcA=mtrC (Table 4.2 in chapter 4) were used in the
perfusion chamber experiment as described in chapter 4. As soon as the nanowires
were produced and observed through staining by FM 4-64FX, the media
ow was
stopped and the chamber was opened in a petri dish containing 20 mL of sterile
medium such that the coverslip remained hydrated. The sample (coverslip with
attached cells) was cut to retain only the area corresponding to the laminar
ow
region of the chamber and was xed with 4% (vol/vol) formaldehyde solution in
PBS (at pH 7.4) for 1 h at room temperature (RT). After rinsing in PBS (each rinse
corresponds to incubation in fresh PBS for 30 seconds), the sample was incubated
in 0:15% glycine in PBS at RT for 5 min to quench free aldehyde groups and reduce
background
uorescence. The sample was then transferred to a blocking solution
of 1% BSA in PBS for 5 min, and reacted with the diluted polyclonal rabbit-
raised MtrC or OmcA-specic primary antibody [104] at RT for 30 min (MtrC
Ab: 2:6g/mL, OmcA Ab: 1:9g/mL, both in 1% BSA/PBS). The MtrC and
OmcA-specic primary antibodies were provided by our collaborator Liang Shi (at
Pacic Northwest National Lab). After rinsing four times in PBS, the sample was
incubated with anti-rabbit FITC-conjugated secondary antibody (Thermo Scien-
tic Pierce antibodies, catalog no. 31635; 7:5g/mL in 1% BSA/PBS) at RT for
30 min. Finally, the coverslip was rinsed twice in PBS before immuno
uorescence
imaging in the green channel while the side of the coverslip with cells remained
hydrated on the microscope. To perform immuno
uorescence on the same cells and
85
nanowires observed during live imaging (Movies S7 and S8 in [92] and Fig. 5.4),
we modied the coverslips with surface scratches that acted as ducial markers.
5.2.2 Preparation of MtrF monolayers on Au(111)
A 4Cys/V5/6His tag (DDDDKAACCPGCCKGKPIPQPLLGLDSTRTGHH-
HHHH) was added at the C-terminus of MtrF to make the recombinant MtrF
used in this study. This tag allowed purication of the protein, while its cysteine
residues established covalent bonds between the recombinant protein and gold sur-
faces. As was described previously for recombinant MtrC [105], we prepared a
construct with a gene encoding this recombinant MtrF, resulting in the construct
(pLS289) which was then introduced into S. oneidensis MR- 1 to give LS622 [106].
Using the protocol described in [17], we overexpressed LS622 and then puried the
tetra-cysteine tagged MtrF. The puried MtrF (1.7 mg/mL) was divided into 100
L aliquots, stored at -20

C in 20 mM HEPES-buered solution with protease
inhibitor, 10% glycerol, 250 mM NaCl, 1% OGP, and 5 mM-mercaptoethanol at
pH 7.6. Using a fresh buer solution of 20 mM OGP and 50 mM HEPES at pH 7.6,
we exchanged the solution twice by centrifuging at 14,000 rcf at 10

C in Amicon
Ultra-0.5 mL centrifugal lters (10 kDa molecular weight cut-o), as described
previously [61]. We then fourfold diluted the extracted solution containing MtrF
with deionized water.
The diluted solution was carefully deposited on atomically
at Au(111) sub-
strates (Agilent annealed N9805B) (Fig. 5.1). These gold surfaces are made by
Au(111) evaporation onto a 50 to 75 m thick mica substrate that is 1:4 1:1 cm
in size. The gold surface itself is 1:0 1:1 cm and is then
ame-annealed to give
atomically
at terraces that are several hundred nanometers across. The Au(111)
86
Figure 5.1: Deposited MtrF solution on atomically
at Au(111) substrate. The
substrate along with the deposited solution is placed on top of wet Kimwipes and
inside a closed petri dish for 15 hours at 4

C.
substrate along with the solution were incubated for 15 hours at 4

C while sit-
ting on wet Kimwipes inside a closed petri dish. Before STM measurements, the
majority of the solution was removed from the surface using a pipette applied to
the corner, followed by wicking to minimize the remaining solution on the surface.
The substrate and surface-adsorbed MtrF layer were then left at room temperature
and ambient humidity for about 2 hours to dry. The sample was then attached
to an 18-mm diameter mounting disk (Bruker Product #:APSH-0010) by apply-
ing silver paint between the mounting disk and the bottom of the mica substrate.
The electrical connection between the gold and the mounting disk was achieved
by extending the silver paint over the mica surface and to the corner of the gold
surface. The sample was then left to dry for 15 mins at room temperature.
5.2.3 Scanning tunneling microscopy of MtrF monolayers
A Bruker Innova AFM/STM was used in combination with 80/20 Pt/Ir STM tips
(Bruker CLST-PTBO) to perform STM measurements. Initially 500500 nm areas
87
of the dried MtrF sample were scanned while the STM tip was held at a more posi-
tive potential (+1 V) than the substrate with a tunneling current setpoint of 2 nA.
These scans showed `pits' in the MtrF monolayer (Fig. 5.6B), similar to the fea-
tures previously observed in MtrC and OmcA monolayers [61]. Subsequently the
STM tip-substrate potential dierence was reversed and the same area was scanned
while the STM tip was held at -3V with respect to the substrate, with a 2 nA tun-
neling current setpoint. As suggested with MtrC and OmcA monolayers [61, 107],
this second scan `activated' the MtrF monolayer by removing the OGP overlayer
that was covering the MtrF proteins. The tip potential and current setpoint were
then changed back to the initial values (+1 V and 2 nA) and the activated area was
scanned for a third time, revealing individual MtrF molecules forming a uniform
monolayer on the atomically
at terraces of the Au(111) substrate.
5.2.4 Atomic force microscopy following live perfusion
ow
imaging
Following in vivo imaging of vesiculation and nanowire growth under perfusion
ow
conditions, samples were imaged by atomic force microscopy (AFM) to correlate

uorescence and nanoscale topography data (Fig. 5.7). Each sample (coverslip
and attached cells) was xed in 2:5% glutaraldehye overnight at 4

C. The sample
was rinsed four times in deionized water (each rinse was a 30-sec incubation in
DI-water) and left to air dry. The coverslip was then cut and prepared for AFM
imaging. The AFM imaging was performed using a tapping mode tip (Asylum
Research, silicon probe model AC240TS with 2 N/m nominal spring constant)
on a Veeco Innova instrument. Scratch marks previously placed on the coverslips
88
acted as ducial markers, allowing us to perform AFM on the same cells and
nanowires observed during live cell growth.
5.2.5 Transmission electron microscopy following nanowire
formation in the in vivo imaging platform
After production of nanowires by the cells attached to the coverlips in the in vivo
imaging platform was conrmed, the perfusion chamber was opened in a Petri dish
containing 20 mL of 1 Phosphate Buer Saline (PBS) at pH=7.4. The coverslip
was then cut to only retain the area corresponding to the laminar
ow region of
the perfusion chamber (rectangular area in Fig. 4.1). The sample was then left
in a 4% (vol/vol) formaldehyde solution in PBS for 1 hour. The sample was then
washed in 100 mM (pH 7.2) sodium cacodylate buer by soaking 5 times in the
solution. Afterwards, the coverslip was soaked in 1% aqueous osmium tetroxide for
30 min, washed with deionized water and then soaked again for 30 min in reduced
osmium (the same osmium solution used before but containing 0.3% potassium
ferrocyanide, from a 3% aqueous solution).
The coverslip was washed with deionized water and then quickly passed through
increasing concentrations of ethanol (30%, 50%, 70%, 80%, 90% and 100%) and
given two extra changes in 100% ethanol. The sample was then soaked in a 50:50
mix of dry ethanol and epoxy resin (Eponate 12 kit from Ted Pella, with a formu-
lation that made a harder rock). No catalyst was added to the resin mixture for
soaking. After the coverslip was left for an hour in the mixture, the resin amount
was increased to a 1:3 ratio of ethanol to resin (no catalyst used).
After soaking the sample overnight, the old resin:ethanol mix was washed away
and was replaced by fresh resin that contained N,N-Dimethylbenzylamine (BDMA)
89
as a catalyst. Next, the back side of the coverslip was wiped, small gelatin capsules
(Electron Microscopy Sciences, gelatin capsules, catalog #: 70104) were lled and
the excess resin was blotted o from the specimen side. The lled capsules were
then inverted onto the coverslip to embed the specimen in resin. The resin was
polymerized by heating overnight in a 60

C oven. Polymerized resin blocks were
removed from the oven and the glass substrate was removed from the resin. This
was done by breaking the glass and separating the resin blocks. The resin with
glass attached was repeatedly immersed in liquid nitrogen and warmed up, until
the glass fell away from the resin. Anything that was initially on the glass surface
was embedded in the resin, but only at the top face of the block.
To section the bacteria that were initially on the glass coverslip surface, the
block was put into an ultramicrotome (Leica UC 6) and the front of the block was
cut with a diamond knife (Diatome: EMS). The knife was positioned so that it was
not perfectly aligned with the block face, but was oset to only cut a band from
the block on one side. All sections were collected and mounted on Formvar-coated
single slot grids (Electron Microscopy Sciences, catalog#:G205-Cu), contrasted
with 3% aqueous uranyl acetate, washed in a jet of water, and contrasted again
with an aqueous solution of lead citrate (Reynolds recipe [108], but modied by
Venable and Coggeshall [109]). The lead citrate was EM grade which was prepared
fresh as follows: 0.1 g lead citrate was dissolved in 1 mL of 1 N sodium hydroxide
and made up to 10 mL. The grids were contrasted by
oating them, section-side
down on drops of the heavy metal solution and then dipping them in degassed
water 10 sequentially in 4 dierent beakers. The rst beaker (10 mL) contained
1 drop of 1 N sodium hydroxide.
90
The grids were dried and imaged with a Tecnai BioTwin Spirit operating at
120 kV (FEI Company, Hillboro, OR).
5.3 Results and discussion
5.3.1 Localization of the decaheme cytochromes MtrC and
OmcA along nanowires
We measured the expression levels ofmtrA,mtrC,omcA, anddmsE, a periplasmic
decaheme cytochrome required for maximal extracellular respiration of DMSO
[102], during and after the transition toO
2
limitation in chemostat cultures. These
EET components had signicantly increased expression in response to electron
acceptor limitation (Fig. 5.2 and Fig. 5.3).
Using MtrC- and OmcA-specic antibodies [104] following in vivo observation
of the target nanowires in the perfusion imaging platform (Movies S7 and S8
in [92]), we observed MtrC and OmcA localization at the periphery of the cell,
as expected. We also observed clear localization of these cytochromes along the
membrane-stained bacterial nanowires (25 of 35 nanowires labeled with anti-MtrC
and 19 of 22 nanowires labeled with anti-OmcA), whereas no
uorescence was
detected from mtrC=omcA negative control cells or their membrane extensions
(none of 22 nanowires labeled with anti-MtrC and none of 20 nanowires labeled
with anti-OmcA; Fig. 5.4).
Though the conductance of Shewanella nanowires was previously only demon-
strated under nonphysiological conditions [32], the data reported here are consis-
tent with membrane extensions that could function as nanowires to mediate EET
from live cells. Localization of MtrC and OmcA to these membrane extensions
91
Figure 5.2: Proling ofmtrC,mtrA, andomcA expression during transition from
electron donor to electron acceptor limitation. Shewanella oneidensis MR-1 cells
were grown under aerobic conditions (dissolved oxygen tension of 20%) at 30

C
in a chemostat for 48 hours before the dissolved oxygen tension was reduced to
0% (t = 0). Samples were taken from the chemostat right before reducing the DO
tension (t< 0, reference sample), at t = 0 and in 15-minute intervals for an hour.
Following RNA purication and qPCR, the fold change in gene expression relative
to the reference sample was calculated by 2
CT
from at least four reactions of
three independent cultures, using recA for normalization.
provides the most compelling evidence to date, and directly supports the proposed
multistep redox hopping mechanism [34, 36, 83], allowing long-range electron trans-
port along a membrane network of heme cofactors that line Shewanella nanowires
(Fig. 5.5). We have shown that S. oneidensis nanowires contain periplasm (Fig.
4.12B in chapter 4); therefore, it is also possible that periplasmic proteins and
soluble redox cofactors may contribute to electron transport through these exten-
sions.
92
Figure 5.3: Proling of dmsE expression during transition from electron donor
to electron acceptor limitation. Shewanella oneidensis MR-1 cells were grown
under aerobic conditions (dissolved oxygen tension of 20%) at 30

C in a chemostat
for 48 hours before the dissolved oxygen tension was reduced to 0% (t = 0).
Samples were taken from the chemostat right before reducing the DO tension
(t< 0, reference sample), att = 0 and in 15-minute intervals for an hour. Following
RNA purication and qPCR, the fold change in gene expression relative to the
reference sample was calculated by 2
CT
from at least four reactions of three
independent cultures, using recA for normalization.
5.3.2 Proof of concept for single-molecule electron transfer
measurements in MtrF
Considering previous ndings pointing at a long range electron transport in bacte-
rial nanowires facilitated by cytochromes [31, 32], we aimed at developing a method
to examine, at the single-molecule level, the electron transfer properties of outer
membrane cytochromes. This basic understanding is a fundamental step toward
building models that describe the entire system of cytochromes mediating the long
range electron transfer in bacterial nanowires.
93
Figure 5.4: Involvement of outer membrane cytochromes in bacterial nanowires.
Labeling with antibodies against MtrC (Left) or OmcA (Right) and membrane

uorescence (FM 4-64FX) images of wild-type S. oneidensis MR-1 (Upper) com-
pared with the mtrC=omcA control strain (Lower). Nanowire-localized MtrC
and OmcA are observed in the wild-type strain. (Scale bars: 2m.)
Figure 5.5: Proposed structural model for Shewanella nanowires. S. oneiden-
sis MR-1 nanowires are outer membrane (OM) and periplasmic (PP) extensions
including the multiheme cytochromes responsible for extracellular electron trans-
port.
94
Figure 5.6: A) Scanning tunneling microscopy image of an activated MtrF mono-
layer on atomically
at Au(111) terraces. Individual MtrF molecules forming a
uniform protein monolayer can be resolved. B) Scanning tunneling microscopy
image of an MtrF monolayer before activation. The pits seen here are observed as
a result of the detergent overlayer blanketing the protein monolayer.
Here we follow the previously reported procedures used in preparation and
investigation of single molecules of MtrC and OmcA [61]. However, the avail-
ability of the crystal strcuture of MtrF [17], unlike MtrC and OmcA, presents a
unique opportunity for comparison between future experimental results that will
be obtained using this method and the existing theoretical models that are built
upon the knowledge of the MtrF crystal structure.
We report, for the rst time, scanning tunneling microscopy measurements
on monolayers of Shewanella's outer membrane cytochrome MtrF. Puried MtrF
molecules adsorbed on the Au(111) substrate via covalent thiol bonds between the
cysteine residues in the tetra-cysteine tag of the protein and the gold surface. This
strong bond allows electrons to tunnel from the protein to the gold substrate dur-
ing STM and TS measurements. Initial scanning of the MtrF monolayer revealed
features that were previously observed in MtrC and OmcA monolayers and were
referred to as `pits' [61] (Fig. 5.6B). A second scan of the same area with the tip-
substrate bias reversed `activated' the monolayer. During the activation scan, as
95
suggested previously with MtrC and OmcA samples [61, 107], the detergent (Octyl
-D-glucopyranoside or OGP) overlayer blanketing the protein monolayer is dis-
rupted enough to open a window for interaction of the STM tip with cytochromes
adsorbed on the surface. The detergent overlayer forms in solution when the OGP
detergent used to solubilize the cytochromes covers the hydrophobic region of the
proteins adsorbed on the surface. After the activation scan, imaging with the ini-
tial tip-substrate bias revealed the uniform monolayer of cytochromes and allowed
the estimation of single protein dimensions (Fig. 5.6A). Individual cytochromes
within the monolayer can be visualized continuously after a one-time activation.
Based on the obtained STM images, we found the lateral dimensions of MtrF
molecules to be in the 6 to 9 nm range. Although, the expected distance between
heme 5 and 10 in the MtrF crystal strcuture (7 nm, [17]) lies in the mea-
sured range of MtrF size in our experiment, it is possible that the adsorbed MtrF
molecules on the surface were tilted at an agle realtive to the ideal upright position
where the heme 5 to heme 10 axis is normal to the substrate surface. This tilt
can be explained by the fact that the C-terminus of the protein, where the tetra-
cysteine tag is located, lies closer to the MtrF center and relatively distant from
heme 10 [17]. In addition,
exibility of the C-terminal region or the tag as well as
partial denaturation of the protein can explain the possible tilting of the protein.
The platform developed here has enabled recent STM and tunneling spectroscopy
(TS) I-V measurements of individual MtrF molecules [40] that together with the
recently reported kinetic Monte Carlo simulations of MtrF conductivity [40] led to
a deeper understanding of the basic physics behind electron transport in MtrF.
96
5.3.3 Nanoscale characterization of the intermediate steps
in nanowire formation
The extension of outer membrane laments, and their functionalization with elec-
tron transport proteins, may represent a widespread strategy for EET. Virtually all
Gram-negative bacteria produce outer membrane vesicles, and can alter the rate
of production and composition of those vesicles in response to various stress condi-
tions [91, 110]. More recent electron microscopy reports describe membrane vesicle
chains and related membrane tubes (also referred to as periplasmic tubules) that
form cell-cell connections in the social soil bacterium Myxococcus xanthus [111] as
well as the phototrophic consortium Chlorochromatium aggregatum [112]. Con-
sistent with these reports, we observed both a transition from vesicle chains to
smoother laments (Movie S9 in [92]), as well as nanowires connecting separate
Shewanella cells (Movie S4 in [92]).
To gain a clearer picture of the role of membrane vesicles in Shewanella
nanowire formation, we performed atomic force microscopy (AFM) on the same
bacterial nanowires after observing their growth with
uorescent imaging under
perfusion
ow conditions. In addition, we imaged bacterial nanowires produced
in the in vivo imaging platform using transmission electron microscopy (TEM).
Such observations were not possible in steady-state chemostat cultures where the
nanowire growth is not conned to the surface-solution interface. Perfusion was
stopped and the samples were xed quickly after observing the early signs of
nanowire production, allowing us to examine the initial stages of nanowire for-
mation with nanoscale resolution using AFM (Fig. 5.7) and TEM (Fig. 5.8).
Using AFM, we measured the nanowires to be 10 nm in diameter under dry condi-
tions, consistent with previous observations [32, 65, 101] and roughly corresponding
97
Figure 5.7: Correlated atomic force microscopy (AFM) and live-cell membrane

uorescence of bacterial nanowires. (A-C) Tapping AFM phase images of S. onei-
densis MR-1 cells after producing bacterial nanowires in the perfusion
ow system.
The sample is xed and air-dried before AFM imaging. (Scale bars: 2m.) (Insets)
In vivo
uorescence images of the same cells/nanowires at the surface/solution
interface in the perfusion platform. The cells and the nanowires are stained by
the membrane stain FM 4-64FX. (Scale bars: 1m.) The morphologies observed
range from vesicle chains (A) to partially smooth laments incorporating vesicles
(B), which is consistent with SEM imaging of chemostat samples (Fig. 4.16A),
and, nally, continuous laments (C). See also Movie S9 in [92], which captures
the transition from a vesicle chain to a continuous bacterial nanowire.
to two lipid bilayers. In addition, we were able to resolve dierent morphologies
corresponding to dierent stages of nanowire formation (Fig. 5.7 and Fig. 5.8),
consistent with the chain-to-lament transition in Movie S9 in [92]. The mor-
phologies observed ranged from vesicle chains (Fig. 5.7A and Fig. 5.8A and B)
to partially smooth laments incorporating vesicles (Fig. 5.7B, also consistent
with SEM imaging in Fig. 4.16A in Chapter 4), and nally continuous laments
(Fig. 5.7C and Fig. 5.8C and D). In addition to possibly mediating EET up
to micrometers away from the inner membrane, the vesiculation and extension
of outer membranes into the quasi one-dimensional morphologies observed here
increase the surface area-to-volume ratio of cells. This shape change can present a
signicant advantage, increasing the likelihood cells will encounter the solid-phase
minerals that serve as electron acceptors for respiration.
98
5.4 Energy cost calculations of nanowire produc-
tion
To better understand the energetics involved in the process of nanowire produc-
tion by a cell, we calculated the free energy necessary to produce a 1-m long
nanowire which includes the free energy of bending the lipid bilayer (cellular outer
membrane) into a 1-m smooth cylinder as well as the energy required to produce
the outer membrane cytochromes on the nanowire. The contribution of the former
part can be calculated using the expression for the bending energy of lipid bilayer
[56]:
E
bend
=
K
b
2
Z
(C
1
+C
2
)
2
dA (5.1)
whereK
b
is the bending rigidity of the lipid bilayer and is typically estimated to be
around 20K
B
T (K
B
being the Boltzmann constant andT temperature in Kelvin)
[56], C
1
and C
2
are the two principle curvatures at each point on the surface, and
R
dA is taken over the entire surface. Assuming a radius of 10 nm for the lament,
calculation of this integral over the surface of a 1-m cylinder-shaped nanowire
gives 6000K
B
T (or energy of hydrolysis of 300 ATP molecules [56]) as the total
requirement for the bending energy. Previous measurements in mid-log cultures of
E. coli have shown an ATP concentration of 10
6
ATPs per cell [113]. Assuming
a similar value for ATP concentration in Shewanella means that the the above
membrane bending energy represents less than
1
1000
th of the cellular ATP budget.
Next, we calculate the cost of production and assembly of the OM cytochromes
on nanowires. The average length of MtrC and OmcA is 700 amino acids, and
the average energy cost of production and adding an amino acid to the peptide
99
chain is 5.2 ATPs per amino acid [56]. Therefore, each OM cytochrome costs about
3600 ATP molecules to make. Assuming a single linear chain of OM cytochromes
along a 1-m nanowire (the minimum number of cytochromes to sustain multistep
hopping), where each cytochrome is approximately 10 nm long, we can conclude
that the nanowire contains a minimum of about 100 OM cytochromes. Multiplying
that number by the cost per cytochrome gives an overall cytochrome cost of 3:6
10
5
ATPs, which is 3 orders of magnitude larger than the membrane bending cost
to produce the nanowire, and comparable to the cellular ATP content ( 10
6
ATPs
per cell [113]).
5.5 Conclusion
We demonstrated the presence of outer membrane cytochromes MtrC and OmcA
along bacterial nanowires, which points to the involvement of cytochromes in
nanowire conductivity and to multistep hopping as its underlying mechanism. In
addition, we developed an approach to investigate the electronic properties of indi-
vidual molecules of MtrF, an MtrC homolog, which can be used in the future to
build a more comprehensive model of nanowire conductance based on cytochrome-
mediated electron transport.
Given the ubiquity of membrane vesicles and related extensions in Gram-
negative bacteria, the localization of electron transport proteins along membrane
extensions in a manner consistent with bacterial nanowires that could mediate
extracellular electron transport, and the nding of nanowire-based cell-cell con-
nectivity, our results raise the intriguing possibility of redox-functionalized mem-
brane extensions as a general microbial strategy for EET and cell-cell signaling.
Our study motivates further experimental and theoretical work to build a detailed
100
understanding of the full biomolecular makeup, electron transport physics, and
physiological impact of bacterial nanowires.
101
Figure 5.8: Transmission electron microscopy (TEM) images of S. oneidensis MR-
1 bacterial nanowires produced in the perfusion imaging platform. A and B show
nanowires in the form of vesicle chains and C and D demonstrate smoother l-
aments. The transition from vesicle chains to smooth laments is also observed
both in live
uorescence microscopy and in atomic force microscopy of nanowires.
102
Chapter 6
Conclusion
Dissimilatory metal reduction in bacteria has been the topic of many studies in
the past two decades. The interest in this topic is mainly rooted in the desire for
renewable energy recovery using bacteria as well as to gain a better understanding
of the global elemental cycles aected by microbes [6, 51]. One of the strategies
used by the dissimilatory metal reducing bacteria such as Shewanella oneidensis
MR-1 to couple their respiration to the reduction of extracellular electron acceptors
involves bacterial nanowires [31]. The conductance of bacterial nanowires has been
characterized before, although under non-physiological conditions [32, 65]. How-
ever, the physical mechanism behind this long-range transport, the compositional
and structural properties, and the physiological relevance of nanowires remained
elusive until this work. Motivated by these questions, in this thesis we addressed
three topics regarding bacterial nanowires.
The rst topic concerns the physical mechanism of electron transport along
bacterial nanowires which has been the subject of intense debate in the past few
years [45]. Some studies have suggested that cytochromes, with their heme groups
acting as redox sites, are responsible for mediating long range electron transport in
nanowires and biolms [35, 36], whereas others have described Geobacter nanowire
electron transport to stem from `metallic-like' conductivity mediated by stacking
of aromatic amino acids along nanowires [47].
103
In chapter 3, to elucidate the physical nature of electron transport in She-
wanella nanowires, we developed a theoretical model that aims at describing the
results from two separate conductance studies on bacterial nanowires ofS. oneiden-
sis MR-1 [32, 65]. In addition, it was previously shown that deletion of two outer
membrane cytochromes, MtrC and OmcA, fromS. oneidensis MR-1 results in non-
conductive nanowires. Informed by this observation, our proposed model features
a chain of redox sites along nanowires, i.e. heme groups within outer membrane
cytochromes, that mediate the long-range electron transfer along nanowires via
a multistep hopping mechanism. As opposed to `metallic-like' conductivity, the
proposed ET mechanism in nanowires of Geobacter, multistep hopping involves
electrons localizing on each site after every hopping event between two adjacent
redox sites along the chain. We found the outcome of this model to match well with
the measured conductance (current-voltageIV ) curves ofShewanella nanowires.
Future models informed by the exact structure of outer membrane cytochromes
as well as the redox potential of their hemes can motivate further experimental
investigations into the electron transfer properties of bacterial nanowires. While
the overall data is most consistent with a multistep hopping mechanism, questions
still remain about the high electronic couplings measured in experiments compared
to the theoretical predictions. This discrepancy motivates the use of new experi-
mental approaches such as spectroscopic measurements in the future to shed more
light on the experimental electronic coupling values.
The second topic revolves around fundamental properties of bacterial nanowires
that are still unknown, including their composition, structure, and physiological
impact. This lack of basic understanding about nanowires was partially caused by
the fact that all previous experimental investigations of bacterial nanowires were
104
performed under non-physiological conditions. In addition, S. oneidensis MR-1
nanowires were widely assumed, but never shown, to be type IV pili, proteinaceous
laments that are typically used for cell attachment and twitching motility. This
assumption was based on the fact that Geobacter nanowires were previously shown
to be type IV pili.
In chapter 4, we addressed these issues by developing a platform that, for the
rst time, allowed us to image in vivo growth of nanowires from live Shewanella
cells. Through staining the structures with
uorescent dyes and localizing the
Green Fluorescent Protein (GFP) to the cellular cytoplasm and periplasm, we
showed that bacterial nanowires are extenions of the outer membrane that contain
periplasm. In addition, using a
uorescent redox activity sensor, we measured an
increase in the cellular respiration rate upon production of nanowires, hinting at
the physiological impact of these structures.
Our study motivates future work in which, under physiological conditions (i.e.
in live cells), electron transfer rates through nanowires that connect cells to elec-
trodes are electrochemically measured. This experiment can quantitatively char-
acterize the physiological impact of nanowires on live Shewanella cells. In light
of our nding that nanowires are outer membrane extensions, it is possible that
these structures are also used for molecular transport between cells through the
extended periplasm contained within the nanowire structure. Future studies can
shed light on this issue, potentially by monitoring in vivo transport of
uorescent
proteins within nanowires that connect live cells to one another and revealing the
exchange of periplasmic content between connected cells.
The third topic of this thesis is about visualization of cytochromes along
nanowires as well as nanoscale characterization of these laments. The role of
105
cytochromes in Geobacter and Shewanella nanowire conductivity has been the
topic of controversy in recent years [45], mainly due to the alternative proposition
of a `metallic-like' transport mechanism in Geobacter nanowires [47] and the nd-
ing that the spacing between cytochromes bound to Geobacter nanowires is too
large to sustain cytochrome-mediated electron transfer along nanowires [114, 115].
However, in Shewanella, the deletion of two outer membrane cytochromes MtrC
and OmcA from the genome was shown to result in non-conductive nanowires [32].
In chapter 5, to investigate the involvement of cytochromes in nanowire conduc-
tance, we stained nanowires produced in our in vivo imaging platform with OmcA
and MtrC-specic antibodies. Fluorescence microscopy revealed the presence of
these two cytochromes along the entire length of bacterial nanowires. Further-
more, nanowires produced from a Shewanella strain lacking MtrC and OmcA were
not stained by the antibodies. These results point at the possible role of outer
membrane cytochromes in nanowire conductivity and are consistent with the pro-
posed multistep hopping mechanism of long-range electron transfer in Shewanella
nanowires [34]. In addition, in this chapter we demonstrated the development of
a method for investigation of electron transfer characteristics of outer membrane
cytochromes at the single-molecule level. This information can serve as the build-
ing block of a large scale physical model that describes nanowire conductance.
In future work, it would be necessary to examine the exact conguration and
spacing between cytochromes along bacterial nanowires. This information will
be benecial in understanding the physics of long-range electron transport in
nanowires as well as in building more accurate theoretical models describing this
phenomenon.
106
Shewanella nanowires were previously reported to be smooth, quasi-one-
dimensional structures that are about 10 nm in diameter [31, 32]. In light of our
nding that nanowires are outer membrane extensions, we examined the dimension
and various morphologies of these structures using atomic force microscopy (AFM)
and transmission electron microscopy (TEM). Using live
uorescence microscopy,
we report the observation of a transition from vesicle chain to smooth laments
during nanowire formation. The dierent stages of this transition were also cap-
tured in the AFM and TEM images.
An intriguing question that remains to be addressed by future studies is about
the formation mechanism of bacterial nanowires. The fact that some nanowires
initially appear as vesicle chains points at the important role of vesicles in nanowire
formation. However, more work is required to elucidate how exactly nanowires are
produced by Shewanella cells, both at the genetic regulation and biophysical level.
107
Bibliography
[1] Harry B Gray and Jay R Winkler. Electron
ow through metallopro-
teins. Biochimica et Biophysica Acta (BBA)-Bioenergetics, 1797(9):1563{
1572, 2010.
[2] R Al Marcus and Norman Sutin. Electron transfers in chemistry and biology.
Biochimica et Biophysica Acta (BBA)-Reviews on Bioenergetics, 811(3):265{
322, 1985.
[3] Peter Mitchell. Coupling of phosphorylation to electron and hydrogen trans-
fer by a chemi-osmotic type of mechanism. Nature, 191(4784):144{148, 1961.
[4] Kristopher A Hunt, Jerey M Flynn, Bel en Naranjo, Indraneel D Shikhare,
and Jerey A Gralnick. Substrate-level phosphorylation is the primary source
of energy conservation during anaerobic respiration of Shewanella oneidensis
strain MR-1. Journal of Bacteriology, 192(13):3345{3351, 2010.
[5] Karrie A Weber, Laurie A Achenbach, and John D Coates. Microorganisms
pumping iron: anaerobic microbial iron oxidation and reduction. Nature
Reviews Microbiology, 4(10):752{764, 2006.
[6] Kenneth H Nealson and Daad Saarini. Iron and manganese in anaerobic
respiration: environmental signicance, physiology, and regulation. Annual
Reviews in Microbiology, 48(1):311{343, 1994.
[7] Derek R Lovley. Dissimilatory metal reduction. Annual Reviews in Micro-
biology, 47(1):263{290, 1993.
[8] Charles R Myers and Kenneth H Nealson. Bacterial manganese reduction
and growth with manganese oxide as the sole electron acceptor. Science, 240,
1988.
[9] Charles R Myers and Kenneth H Nealson. Respiration-linked proton translo-
cation coupled to anaerobic reduction of manganese (IV) and iron (III) in
Shewanella putrefaciens MR-1. Journal of Bacteriology, 172(11):6232{6238,
1990.
108
[10] Derek R Lovley and Elizabeth JP Phillips. Novel mode of microbial energy
metabolism: organic carbon oxidation coupled to dissimilatory reduction of
iron or manganese. Applied and Environmental Microbiology, 54(6):1472{
1480, 1988.
[11] Charles R Myers and Judith M Myers. Localization of cytochromes to the
outer membrane of anaerobically grownShewanellaputrefaciens MR-1. Jour-
nal of Bacteriology, 174(11):3429{3438, 1992.
[12] Judith M Myers and Charles R Myers. Role for outer membrane cytochromes
OmcA and OmcB of Shewanella putrefaciens MR-1 in reduction of man-
ganese dioxide. Applied and Environmental Microbiology, 67(1):260{269,
2001.
[13] CR Myers and JM Myers. Cell surface exposure of the outer membrane
cytochromes of Shewanella oneidensis MR-1. Letters in Applied Microbiol-
ogy, 37(3):254{258, 2003.
[14] Liang Shi, David J Richardson, Zheming Wang, Sebastien N Kerisit, Kevin M
Rosso, John M Zachara, and James K Fredrickson. The roles of outer mem-
brane cytochromes of Shewanella and Geobacter in extracellular electron
transfer. Environmental Microbiology Reports, 1(4):220{227, 2009.
[15] Robert S Hartshorne, Catherine L Reardon, Daniel Ross, Jochen Nuester,
Thomas A Clarke, Andrew J Gates, Paul C Mills, Jim K Fredrickson, John M
Zachara, Liang Shi, et al. Characterization of an electron conduit between
bacteria and the extracellular environment. Proceedings of the National
Academy of Sciences, 106(52):22169{22174, 2009.
[16] Robert S Hartshorne, Brian N Jepson, Tom A Clarke, Sarah J Field, Jim
Fredrickson, John Zachara, Liang Shi, Julea N Butt, and David J Richard-
son. Characterization of Shewanella oneidensis MtrC: a cell-surface deca-
heme cytochrome involved in respiratory electron transport to extracellular
electron acceptors. JBIC Journal of Biological Inorganic Chemistry, 12(7):
1083{1094, 2007.
[17] Thomas A Clarke, Marcus J Edwards, Andrew J Gates, Andrea Hall, Gaye F
White, Justin Bradley, Catherine L Reardon, Liang Shi, Alexander S Beliaev,
Matthew J Marshall, et al. Structure of a bacterial cell surface decaheme
electron conduit. Proceedings of the National Academy of Sciences, 108(23):
9384{9389, 2011.
[18] Gaye F White, Zhi Shi, Liang Shi, Zheming Wang, Alice C Dohnalkova,
Matthew J Marshall, James K Fredrickson, John M Zachara, Julea N Butt,
109
David J Richardson, et al. Rapid electron exchange between surface-exposed
bacterial cytochromes and Fe (III) minerals. Proceedings of the National
Academy of Sciences, 110(16):6346{6351, 2013.
[19] Marian Breuer, Kevin M Rosso, Jochen Blumberger, and Julea N Butt.
Multi-haem cytochromes in Shewanella oneidensis MR-1: structures, func-
tions and opportunities. Journal of The Royal Society Interface, 12(102):
20141117, 2015.
[20] Orianna Bretschger, Anna Obraztsova, Carter A Sturm, In Seop Chang,
Yuri A Gorby, Samantha B Reed, David E Culley, Catherine L Reardon,
Soumitra Barua, Margaret F Romine, et al. Current production and metal
oxide reduction by Shewanella oneidensis MR-1 wild type and mutants.
Applied and Environmental Microbiology, 73(21):7003{7012, 2007.
[21] Leonard M Tender, Clare E Reimers, Hilmar A Stecher, Dawn E Holmes,
Daniel R Bond, Daniel A Lowy, Kanoelani Pilobello, Stephanie J Fertig, and
Derek R Lovley. Harnessing microbially generated power on the sea
oor.
Nature Biotechnology, 20(8):821{825, 2002.
[22] Kenneth H Nealson, Andrea Belz, and Brent McKee. Breathing metals as a
way of life: geobiology in action. Antonie Van Leeuwenhoek, 81(1-4):215{222,
2002.
[23] Derek R Lovley, Elizabeth JP Phillips, Yuri A Gorby, and Edward R Landa.
Microbial reduction of uranium. Nature, 350(6317):413{416, 1991.
[24] Hong Liu, Ramanathan Ramnarayanan, and Bruce E Logan. Production
of electricity during wastewater treatment using a single chamber microbial
fuel cell. Environmental Science & Technology, 38(7):2281{2285, 2004.
[25] Daniel R Bond, Dawn E Holmes, Leonard M Tender, and Derek R Lovley.
Electrode-reducing microorganisms that harvest energy from marine sedi-
ments. Science, 295(5554):483{485, 2002.
[26] Korneel Rabaey and Ren e A Rozendal. Microbial electrosynthesisrevisiting
the electrical route for microbial production. Nature Reviews Microbiology,
8(10):706{716, 2010.
[27] Kenneth H Nealson and B Lea Cox. Microbial metal-ion reduction and
mars: extraterrestrial expectations? Current opinion in microbiology, 5(3):
296{300, 2002.
[28] Rudolph A Marcus. On the theory of oxidation-reduction reactions involving
electron transfer. I. The Journal of Chemical Physics, 24(5):966{978, 1956.
110
[29] JJ Hopeld. Electron transfer between biological molecules by thermally
activated tunneling. Proceedings of the National Academy of Sciences, 71(9):
3640{3644, 1974.
[30] Gemma Reguera, Kevin D McCarthy, Teena Mehta, Julie S Nicoll, Mark T
Tuominen, and Derek R Lovley. Extracellular electron transfer via microbial
nanowires. Nature, 435(7045):1098{1101, 2005.
[31] Yuri A Gorby, Svetlana Yanina, Jerey S McLean, Kevin M Rosso, Dianne
Moyles, Alice Dohnalkova, Terry J Beveridge, In Seop Chang, Byung Hong
Kim, Kyung Shik Kim, et al. Electrically conductive bacterial nanowires
produced by Shewanella oneidensis strain MR-1 and other microorganisms.
ProceedingsoftheNationalAcademyofSciences, 103(30):11358{11363, 2006.
[32] Mohamed Y El-Naggar, Greg Wanger, Kar Man Leung, Thomas D Yuzvin-
sky, Gordon Southam, Jun Yang, Woon Ming Lau, Kenneth H Nealson,
and Yuri A Gorby. Electrical transport along bacterial nanowires from She-
wanella oneidensis MR-1. Proceedings of the National Academy of Sciences,
107(42):18127{18131, 2010.
[33] Christian Pfeer, Steen Larsen, Jie Song, Mingdong Dong, Flemming
Besenbacher, Rikke Louise Meyer, Kasper Urup Kjeldsen, Lars Schreiber,
Yuri A Gorby, Mohamed Y El-Naggar, et al. Filamentous bacteria transport
electrons over centimetre distances. Nature, 491(7423):218{221, 2012.
[34] Sahand Pirbadian and Mohamed Y El-Naggar. Multistep hopping and extra-
cellular charge transfer in microbial redox chains. Physical Chemistry Chem-
ical Physics, 14(40):13802{13808, 2012.
[35] Sarah M Strycharz-Glaven, Rachel M Snider, Anthony Guiseppi-Elie, and
Leonard M Tender. On the electrical conductivity of microbial nanowires
and biolms. Energy & Environmental Science, 4(11):4366{4379, 2011.
[36] Nicholas F Polizzi, Spiros S Skourtis, and David N Beratan. Physical con-
straints on charge transport through bacterial nanowires. Faraday discus-
sions, 155:43{61, 2012.
[37] Marian Breuer, Piotr Zarzycki, Jochen Blumberger, and Kevin M Rosso.
Thermodynamics of electron
ow in the bacterial deca-heme cytochrome
MtrF. Journal of the American Chemical Society, 134(24):9868{9871, 2012.
[38] Marian Breuer, Piotr Zarzycki, Liang Shi, ThomasA Clarke, MarcusJ
Edwards, JuleaN Butt, DavidJ Richardson, JamesK Fredrickson, JohnM
Zachara, Jochen Blumberger, et al. Molecular structure and free energy
111
landscape for electron transport in the decahaem cytochrome MtrF. Bio-
chemical Society Transactions, 40(6):1198, 2012.
[39] Marian Breuer, Kevin M Rosso, and Jochen Blumberger. Electron
ow in
multiheme bacterial cytochromes is a balancing act between heme electronic
interaction and redox potentials. Proceedings of the National Academy of
Sciences, 111(2):611{616, 2014.
[40] Hye Suk Byun, Sahand Pirbadian, Aiichiro Nakano, Liang Shi, and
Mohamed Y El-Naggar. Kinetic Monte Carlo simulations and molecular con-
ductance measurements of the bacterial decaheme cytochrome MtrF. Chem-
ElectroChem, 1(11):1932{1939, 2014.
[41] Dianne K Newman and Roberto Kolter. A role for excreted quinones in
extracellular electron transfer. Nature, 405(6782):94{97, 2000.
[42] Enrico Marsili, Daniel B Baron, Indraneel D Shikhare, Dan Coursolle, Jef-
frey A Gralnick, and Daniel R Bond. Shewanella secretes
avins that medi-
ate extracellular electron transfer. Proceedings of the National Academy of
Sciences, 105(10):3968{3973, 2008.
[43] Nicholas J Kotloski and Jerey A Gralnick. Flavin electron shuttles dominate
extracellular electron transfer by Shewanella oneidensis. MBio, 4(1):e00553{
12, 2013.
[44] Thomas Boesen and Lars Peter Nielsen. Molecular dissection of bacterial
nanowires. Mbio, 4(3):e00270{13, 2013.
[45] Sarah M Strycharz-Glaven and Leonard M Tender. Reply to the com-
ment on on electrical conductivity of microbial nanowires and biolmsby NS
Malvankar, MT Tuominen and DR Lovley, Energy Environ. Sci., 2012, 5,
doi: 10.1039/c2ee02613a. Energy & Environmental Science, 5(3):6250{6255,
2012.
[46] Nikhil S Malvankar and Derek R Lovley. Microbial nanowires for bioenergy
applications. Current Opinion in Biotechnology, 27:88{95, 2014.
[47] Nikhil S Malvankar, Madeline Vargas, Kelly P Nevin, Ashley E Franks, Ching
Leang, Byoung-Chan Kim, Kengo Inoue, T unde Mester, Sean F Covalla,
Jessica P Johnson, et al. Tunable metallic-like conductivity in microbial
nanowire networks. Nature Nanotechnology, 6(9):573{579, 2011.
[48] Marina G Kalyuzhnaya, Mary E Lidstrom, and Ludmila Chistoserdova. Real-
time detection of actively metabolizing microbes by redox sensing as applied
to methylotroph populations in Lake Washington. The ISME journal, 2(7):
696{706, 2008.
112
[49] D Gray, RS Yue, Ching-Ying Chueng, and William Godfrey. Bacterial vital-
ity detected by a novel
uorogenic redox dye using
ow cytometry. In Amer-
ican Society of Microbiology Meeting: Washington, DC, USA, 2005.
[50] Bruce E Logan. Exoelectrogenic bacteria that power microbial fuel cells.
Nature Reviews Microbiology, 7(5):375{381, 2009.
[51] Bruce E Logan and John M Regan. Microbial fuel cells-challenges and appli-
cations. Environmental Science & Technology, 40(17):5172{5180, 2006.
[52] Parish P Sedghizadeh, Satish KS Kumar, Amita Gorur, Christoph
Schaudinn, Charles F Shuler, and J William Costerton. Microbial biolms
in osteomyelitis of the jaw and osteonecrosis of the jaw secondary to bisphos-
phonate therapy. The Journal of the American Dental Association, 140(10):
1259{1265, 2009.
[53] Greg Wanger, Yuri Gorby, Mohamed Y El-Naggar, Thomas D Yuzvinsky,
Christoph Schaudinn, Amita Gorur, and Parish P Sedghizadeh. Electrically
conductive bacterial nanowires in bisphosphonate-related osteonecrosis of the
jaw biolms. Oral surgery, oral medicine, oral pathology and oral radiology,
115(1):71{78, 2013.
[54] Erwin Schrodinger and Lewin. What is life? University Press, 1967.
[55] Koscak Maruyama. The discovery of adenosine triphosphate and the estab-
lishment of its structure. Journal of the History of Biology, 24(1):145{154,
1991.
[56] Rob Phillips, Jane Kondev, Julie Theriot, and Hernan Garcia. Physical
biology of the cell. Garland Science, 2012.
[57] Rudolf K Thauer, Kurt Jungermann, and Karl Decker. Energy conservation
in chemotrophic anaerobic bacteria. Bacteriological Reviews, 41(1):100, 1977.
[58] Mohamed Y El-Naggar and Steven E Finkel. Live wires. The Scientist, 27
(5):38{43, 2013.
[59] David J Richardson, Julea N Butt, Jim K Fredrickson, John M Zachara,
Liang Shi, Marcus J Edwards, Gaye White, Nanakow Baiden, Andrew J
Gates, Sophie J Marritt, et al. The porin{cytochrome model for microbe-to-
mineral electron transfer. Molecular Microbiology, 85(2):201{212, 2012.
[60] Dan Coursolle, Daniel B Baron, Daniel R Bond, and Jerey A Gralnick.
The Mtr respiratory pathway is essential for reducing
avins and electrodes
in Shewanella oneidensis. Journal of bacteriology, 192(2):467{474, 2010.
113
[61] Nicholas S Wigginton, Kevin M Rosso, Brian H Lower, Liang Shi, and
Michael F Hochella. Electron tunneling properties of outer-membrane deca-
heme cytochromes from Shewanella oneidensis. Geochimica et cosmochimica
acta, 71(3):543{555, 2007.
[62] Gerd Binnig, Calvin F Quate, and Ch Gerber. Atomic force microscope.
Physical Review Letters, 56(9):930, 1986.
[63] Gerd Binnig and Heinrich Rohrer. Scanning tunneling microscopy. IBM
Journal of research and development, 44(1-2):279{293, 2000.
[64] Souichiro Kato, Kazuhito Hashimoto, and Kazuya Watanabe. Microbial
interspecies electron transfer via electric currents through conductive miner-
als. Proceedings of the National Academy of Sciences, 109(25):10042{10046,
2012.
[65] Mohamed Y El-Naggar, Yuri A Gorby, Wei Xia, and Kenneth H Nealson.
The molecular density of states in bacterial nanowires. Biophysical Journal,
95(1):L10{L12, 2008.
[66] H Liu, GJ Newton, R Nakamura, K Hashimoto, and S Nakanishi. Electro-
chemical characterization of a single electricity-producing bacterial cell of
Shewanella by using optical tweezers. Angewandte Chemie (International
ed. in English), 49(37):6596{6599, 2010.
[67] Lucius S Fox, Mariusz Kozik, Jay R Winkler, and Harry B Gray. Gaus-
sian free-energy dependence of electron-transfer rates in iridium complexes.
Science, 247(4946):1069{1071, 1990.
[68] Michael R Wasielewski, Mark P Niemczyk, Walter A Svec, and E Bradley
Pewitt. Dependence of rate constants for photoinduced charge separation
and dark charge recombination on the free energy of reaction in restricted-
distance porphyrin-quinone molecules. Journal of the American Chemical
Society, 107(4):1080{1082, 1985.
[69] Christopher ED Chidsey. Free energy and temperature dependence of elec-
tron transfer at the metal-electrolyte interface. Science, 251(4996):919{922,
1991.
[70] Laurie J Jones, Richard P Haugland, and Victoria L Singer. Development
and characterization of the nanoorange R
protein quantitation assay: A

uorescence-based assay of proteins in solution. BioTechniques, 34(4):850{
861, 2003.
114
[71] Hans-Peter Grossart, Grieg F Steward, Josena Martinez, and Farooq Azam.
A simple, rapid method for demonstrating bacterial
agella. Applied and
Environmental Microbiology, 66(8):3632{3636, 2000.
[72] William J Betz, Fei Mao, and Corey B Smith. Imaging exocytosis and
endocytosis. Current opinion in neurobiology, 6(3):365{371, 1996.
[73] S Bolte, C Talbot, Y Boutte, O Catrice, ND Read, and B Satiat-Jeunemaitre.
FM-dyes as experimental probes for dissecting vesicle tracking in living
plant cells. Journal of Microscopy, 214(2):159{173, 2004.
[74] LR Gring. FRET analysis of transmembrane
ipping of FM4{64 in plant
cells: is FM4{64 a robust marker for endocytosis? Journal of microscopy,
231(2):291{298, 2008.
[75] Valentina Rosu and Kelly T Hughes. 28-dependent transcription in
Salmonella enterica is independent of
agellar shearing. Journal of Bac-
teriology, 188(14):5196{5203, 2006.
[76] Kris M Blair, Linda Turner, Jared T Winkelman, Howard C Berg, and
Daniel B Kearns. A molecular clutch disables
agella in the Bacillus subtilis
biolm. Science, 320(5883):1636{1638, 2008.
[77] Andrew C Lowenthal, Marla Hill, Laura K Sycuro, Khalid Mehmood, Nina R
Salama, and Karen M Ottemann. Functional analysis of the Helicobacter
pylori
agellar switch proteins. Journal of bacteriology, 191(23):7147{7156,
2009.
[78] Dena L Cologgi, Sanela Lampa-Pastirk, Allison M Speers, Shelly D Kelly,
and Gemma Reguera. Extracellular reduction of uranium via Geobacter con-
ductive pili as a protective cellular mechanism. Proceedings of the National
Academy of Sciences, 108(37):15248{15252, 2011.
[79] Ian D Odell and Deborah Cook. Immuno
uorescence techniques. Journal of
Investigative Dermatology, 133(1):e4, 2013.
[80] Harry B Gray and Jay R Winkler. Electron tunneling through proteins.
Quarterly Reviews of Biophysics, 36(03):341{372, 2003.
[81] George A Kowalchuk, Susan E Jones, and Linda L Blackall. Microbes orches-
trate life on earth. The ISME Journal, 2(8):795{796, 2008.
[82] Harald Von Canstein, Jun Ogawa, Sakayu Shimizu, and Jonathan R Lloyd.
Secretion of
avins by Shewanella species and their role in extracellular elec-
tron transfer. Applied and environmental microbiology, 74(3):615{623, 2008.
115
[83] Rachel M Snider, Sarah M Strycharz-Glaven, Stanislav D Tsoi, Jerey S
Erickson, and Leonard M Tender. Long-range electron transport inGeobacter
sulfurreducens biolms is redox gradient-driven. Proceedings of the National
Academy of Sciences, 109(38):15467{15472, 2012.
[84] Ferdinand C Grozema and Laurens DA Siebbeles. Mechanism of charge
transport in self-organizing organic materials. InternationalReviewsinPhys-
ical Chemistry, 27(1):87{138, 2008.
[85] Alessandro Troisi. The speed limit for sequential charge hopping in molecular
materials. Organic Electronics, 12(12):1988{1991, 2011.
[86] James C Bellowsa and Paras N Prasadb. Dephasing times and linewidths
of optical transitions in molecular crystals. Temperature dependence of line
shapes, linewidths, and frequencies of Raman active phonons in naphthalene.
The Journal of Chemical Physics, 70(4):1864{1871, 1979.
[87] Aleksandr M Kuznetsov, Peter Sommer-Larsen, and Jens Ulstrup. Reso-
nance and environmental
uctuation eects in STM currents through large
adsorbed molecules. Surface science, 275(1):52{64, 1992.
[88] Carlo Augusto Bortolotti, Magdalena E Siwko, Elena Castellini, Antonio
Ranieri, Marco Sola, and Stefano Corni. The reorganization energy in
cytochrome c is controlled by the accessibility of the heme to the solvent.
The Journal of Physical Chemistry Letters, 2(14):1761{1765, 2011.
[89] Stefano Corni. The reorganization energy of azurin in bulk solution and in
the electrochemical scanning tunneling microscopy setup. The Journal of
Physical Chemistry B, 109(8):3423{3430, 2005.
[90] Madeline Vargas, Nikhil S Malvankar, Pier-Luc Tremblay, Ching Leang,
Jessica A Smith, Pranav Patel, Oona Synoeyenbos-West, Kelly P Nevin,
and Derek R Lovley. Aromatic amino acids required for pili conductivity
and long-range extracellular electron transport in Geobacter sulfurreducens.
MBio, 4(2):e00105{13, 2013.
[91] Y Gorby, J McLean, A Korenevsky, K Rosso, Mohamed Y El-Naggar, and
Terrance J Beveridge. Redox-reactive membrane vesicles produced by She-
wanella. Geobiology, 6(3):232{241, 2008.
[92] Sahand Pirbadian, Sarah E Barchinger, Kar Man Leung, Hye Suk Byun,
Yamini Jangir, Rachida A Bouhenni, Samantha B Reed, Margaret F Romine,
Daad A Saarini, Liang Shi, et al. Shewanella oneidensis MR-1 nanowires
are outer membrane and periplasmic extensions of the extracellular electron
116
transport components. Proceedings of the National Academy of Sciences, 111
(35):12883{12888, 2014.
[93] Rachida A Bouhenni, Gary J Vora, Justin C Binger, Sheetal Shirodkar,
Ken Brockman, Ricky Ray, Peter Wu, Brandy J Johnson, Eulandria M Bid-
dle, Matthew J Marshall, et al. The role of Shewanella oneidensis MR-1
outer surface structures in extracellular electron transfer. Electroanalysis, 22
(7-8):856{864, 2010.
[94] Qixia Luo, Yangyang Dong, Haijiang Chen, and Haichun Gao. Mislocaliza-
tion of rieske protein PetA predominantly accounts for the aerobic growth
defect of Tat mutants inShewanella oneidensis. PloS one, 8(4):e62064, 2013.
[95] Ann G Matthysse, Serina Stretton, Catherine Dandie, Nicholas C McClure,
and Amanda E Goodman. Construction of GFP vectors for use in Gram-
negative bacteria other than Escherichia coli. FEMS microbiology letters,
145(1):87{94, 1996.
[96] Jerey S McLean, Greg Wanger, Yuri A Gorby, Martin Wainstein, Je
McQuaid, Shunichi Ishii, Orianna Bretschger, Haluk Beyenal, and Ken-
neth H Nealson. Quantication of electron transfer rates to a solid phase
electron acceptor through the stages of biolm formation from single cells
to multicellular communities. Environmental Science & Technology, 44(7):
2721{2727, 2010.
[97] William G Miller, Johan HJ Leveau, and Steven E Lindow. Improved gfp
andinaZ broad-host-range promoter-probe vectors. MolecularPlant-Microbe
Interactions, 13(11):1243{1250, 2000.
[98] In Seop Chang, Hyunsoo Moon, Orianna Bretschger, Jae Kyung Jang, Ho Il
Park, Kenneth H Nealson, and Byung Hong Kim. Electrochemically active
bacteria (EAB) and mediator-less microbial fuel cells. JMicrobiolBiotechnol,
16(2):163{177, 2006.
[99] VA Rhodius and CA Gross. Using DNA microarrays to assay part function.
Methods in enzymology, 497:75{113, 2010.
[100] Kar Man Leung, Greg Wanger, Qiuquan Guo, Yuri Gorby, Gordon Southam,
Woon Ming Lau, and Jun Yang. Bacterial nanowires: conductive as silicon,
soft as polymer. Soft Matter, 7(14):6617{6621, 2011.
[101] Kar Man Leung, Greg Wanger, Mohamed Y El-Naggar, Yuri Gorby, Gordon
Southam, Woon Ming Lau, and Jun Yang. Shewanella oneidensis MR-1 bac-
terial nanowires exhibit p-type, tunable electronic behavior. Nano Letters,
13(6):2407{2411, 2013.
117
[102] Jerey A Gralnick, Hojatollah Vali, Douglas P Lies, and Dianne K New-
man. Extracellular respiration of dimethyl sulfoxide by Shewanella oneiden-
sis strain MR-1. Proceedings of the National Academy of Sciences of the
United States of America, 103(12):4669{4674, 2006.
[103] Erinn C Howard, Emily R Petersen, Lisa A Fitzgerald, Paul E Sheehan, and
Bradley R Ringeisen. Optimal method for eciently removing extracellular
nanolaments from Shewanella oneidensis MR-1. Journal of microbiological
methods, 87(3):320{324, 2011.
[104] Catherine L Reardon, AC Dohnalkova, Ponnusamy Nachimuthu, David W
Kennedy, DA Saarini, Bruce W Arey, Liang Shi, Zheming Wang, D Moore,
Jerey S Mclean, et al. Role of outer-membrane cytochromes MtrC and
OmcA in the biomineralization of ferrihydrite by Shewanella oneidensis MR-
1. Geobiology, 8(1):56{68, 2010.
[105] Liang Shi, Baowei Chen, Zheming Wang, Dwayne A Elias, M Uljana Mayer,
Yuri A Gorby, Shuison Ni, Brian H Lower, David W Kennedy, David S Wun-
schel, et al. Isolation of a high-anity functional protein complex between
OmcA and MtrC: two outer membrane decaheme c-type cytochromes of She-
wanella oneidensis MR-1. Journal of Bacteriology, 188(13):4705{4714, 2006.
[106] Liang Shi, Jiann-Trzwo Lin, Lye M Markillie, Thomas C Squier, and Brian S
Hooker. Overexpression of multi-heme c-type cytochromes. Biotechniques,
38:297{299, 2005.
[107] Nicholas S Wigginton, Kevin M Rosso, and Michael F Hochella. Mechanisms
of electron transfer in two decaheme cytochromes from a metal-reducing
bacterium.TheJournalofPhysicalChemistryB, 111(44):12857{12864, 2007.
[108] Edward S Reynolds. The use of lead citrate at high ph as an electron-opaque
stain in electron microscopy. The Journal of Cell Biology, 17(1):208{212,
1963.
[109] John H Venable and Richard Coggeshall. A simplied lead citrate stain for
use in electron microscopy. The Journal of Cell Biology, 25(2):407{408, 1965.
[110] Adam Kulp and Meta J Kuehn. Biological functions and biogenesis of
secreted bacterial outer membrane vesicles. Annual Review of Microbiology,
64:163, 2010.
[111] Jonathan P Remis, Dongguang Wei, Amita Gorur, Marcin Zemla, Jessica
Haraga, Simon Allen, H Ewa Witkowska, J William Costerton, James E
118
Berleman, and Manfred Auer. Bacterial social networks: structure and com-
position of Myxococcus xanthus outer membrane vesicle chains. Environmen-
tal Microbiology, 16(2):598{610, 2014.
[112] Gerhard Wanner, Kajetan Vogl, and J org Overmann. Ultrastructural char-
acterization of the prokaryotic symbiosis in Chlorochromatium aggregatum.
Journal of Bacteriology, 190(10):3721{3730, 2008.
[113] Michael H Buckstein, Jian He, and Harvey Rubin. Characterization of
nucleotide pools as a function of physiological state in Escherichia coli. Jour-
nal of bacteriology, 190(2):718{726, 2008.
[114] Ching Leang, Xinlei Qian, T unde Mester, and Derek R Lovley. Alignment of
the c-type cytochrome OmcS along pili of Geobacter sulfurreducens. Applied
and Environmental Microbiology, 76(12):4080{4084, 2010.
[115] Nikhil S Malvankar, Mark T Tuominen, and Derek R Lovley. Lack of
cytochrome involvement in long-range electron transport through conductive
biolms and nanowires ofGeobacter sulfurreducens. Energy & Environmental
Science, 5(9):8651{8659, 2012.
119

Abstract (if available)

Abstract In this thesis, we discuss three topics concerning extracellular electron transfer in the Dissimilatory Metal Reducing Bacterium (DMRB) Shewanella oneidensis MR-1. One proposed strategy to accomplish extracellular charge transfer in Shewanella involves forming a conductive pathway to electrodes by incorporating redox components on outer cell membranes and along extracellular appendages known as bacterial nanowires within biofilms. In the first part of this dissertation, to describe extracellular charge transfer in microbial redox chains, we employed a model based on incoherent hopping between sites in the chain and an interfacial treatment of electrochemical interactions with the surrounding electrodes. Based on this model, we calculated the current-voltage (I-V) characteristics and found the results to be in good agreement with I-V measurements across and along individual microbial nanowires produced by the bacterium S. oneidensis MR-1. Based on our analysis, we propose that multistep hopping in redox chains constitutes a viable strategy for extracellular charge transfer in microbial biofilms. ❧ In the second part, we report the first in vivo observations of the formation and respiratory impact of nanowires in the model metal-reducing microbe S. oneidensis MR-1. Live fluorescence measurements, immunolabeling, and quantitative gene expression analysis point to S. oneidensis MR-1 nanowires as extensions of the outer membrane and periplasm that include the multiheme cytochromes responsible for EET, rather than pilin-based structures as previously thought. These membrane extensions are associated with outer membrane vesicles, structures ubiquitous in Gram-negative bacteria, and are consistent with bacterial nanowires that mediate long-range EET by our proposed multistep redox hopping mechanism. Redox-functionalized membrane and vesicular extensions may represent a general microbial strategy for electron transport and energy distribution. ❧ In addition, to elucidate the membranous nature of Shewanella nanowires, we imaged these filaments using Transmission Electron Microscopy. The TEM images reported in this thesis also provide the most accurate estimates of bacterial nanowire dimensions to date. Future TEM and cryo-TEM imaging can establish the specific alignment and configuration of outer membrane cytochromes that facilitate electron transport along bacterial nanowires. ❧ In the third part of this thesis, we focus on the molecular conductance of MtrF, the first decaheme outer membrane cytochrome with a solved crystal structure. Decaheme outer membrane cytochromes of Shewanella play a crucial role in all the suggested pathways of extracellular electron transfer. An understanding of the electron transfer properties in MtrF will therefore impact all aspects of extracellular electron transfer research. In this thesis, using purified MtrF, we form monolayers of the protein on atomically flat gold substrates and address the dry monolayer with a Scanning Tunneling Microscope (STM) tip. This technique can be used in the future to examine the conductivity of individual MtrF molecules within the monolayer in the form of I-V curves. This methodology will allow experimental comparison with recently developed simulations of MtrF conductance.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

From single molecules to bacterial nanowires: functional and dynamic imaging of the extracellular electron transfer network in Shewanella oneidensis MR-1

PDF

Kinetic Monte Carlo simulations for electron transport in redox proteins: from single cytochromes to redox networks

PDF

Electronic, electrochemical, and spintronic characterization of bacterial electron transport

PDF

From cables to biofilms: electronic and electrochemical characterization of electroactive microbial systems

PDF

Electrochemical studies of outward and inward extracellular electron transfer by microorganisms from diverse environments

PDF

From fuel cells to single cells: electrochemical measurements of direct electron transfer at microbial-electrode interfaces

PDF

Electrochemical studies of subsurface microorganisms

PDF

Identification of a new bacterial sensing mechanism: characterization of bacterial insoluble electron acceptor sensing

PDF

Mathematical modeling in bacterial communication and optogenetic systems

PDF

Electrochemical investigations and imaging tools for understanding extracellular electron transfer in phylogenetically diverse bacteria

PDF

Survival and evolution of Shewanella oneidensis MR-1: applications for microbial fuel cells

Asset Metadata

Creator Pirbadian, Sahand (author)

Core Title Bacterial nanowires of Shewanella oneidensis MR-1: electron transport mechanism, composition, and role of multiheme cytochromes

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Physics

Publication Date 04/08/2015

Defense Date 03/11/2015

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag bacterial nanowires,bioelectronics,dissimilatory metal reducing bacteria,extracellular electron transfer,membrane cytochromes,membrane extensions,microbial fuel cells,OAI-PMH Harvest,respiration,Shewanella,vesicle chain

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor El-Naggar, Mohamed Y. (committee chair), Finkel, Steven E. (committee member), Haas, Stephan W. (committee member), Haselwandter, Christoph (committee member), Kresin, Vitaly V. (committee member)

Creator Email sahand.pirbadian@gmail.com,spirbadi@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-543883

Unique identifier UC11297554

Identifier etd-PirbadianS-3259.pdf (filename),usctheses-c3-543883 (legacy record id)

Legacy Identifier etd-PirbadianS-3259.pdf

Dmrecord 543883

Document Type Dissertation

Format application/pdf (imt)

Rights Pirbadian, Sahand

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA