. . .2 Kuﬂaﬁcﬁhi. ~ .
u.
.3.

.
r. La". .14.... a. .
gamma. .3
nﬁﬁn ".552...
.2: xbxlmx:.lnt

1%.? A...

 

u

,3

. .12.. 1......

inl‘...

«w

 

h.»

 

. 11.36.: JII.
«L n. .{is I!
3:4? 3... } ..il .3?!
33.2.:éish. I... turn,
glniﬁsvvv:s !
‘0...
a!

Kalil! it“

idvémmukﬁ

Hahn-twin“.
Lirolrlhu
ll

15... i.
by“?

El

 

O 7 O LIBRARY
' Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

METAL REDUCTION BY
DESULFITOBACTERIUM HAFNIENSE DOB-2

----a--¢-->-u-c-o-n-—- n-

presented by

- -n—a-.-o---~-.-u

Christina Lydia Harzman

has been accepted towards fulﬁllment
of the requirements for the

PhD. degree in Microbiology and Molecular
Genetics

 

 

i

l / \ ' , L r
” I “ ’4/ (will ‘1 L-

M or Professors Signatu 5

92¢. 61200?“ '

Date

oa--u---.-n-n-.-----o-o-o-.-n---o-I-n-o--I- - - -

MSU is an Aﬁ'innative Action/Equal Opportunity Employer

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KIProj/AowPresIClRC/Dateomjndd

METAL REDUCTION BY
DES ULFI T OBA C TERI UM HAFNIENSE DCB-2

By

Christina Lydia Harzman

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the Degree of

DOCTOR OF PHILOSOPHY
Microbiology and Molecular Genetics

2009

ABSTRACT
METAL REDUCTION BY DESULFITOBA CTERIUM HAFNIENSE DCB—2
By

Christina Lydia Harzman

I evaluated the ability of the Gram-positive dehalogenator Desulﬁtobacterium hafniense
strain DCB-2 to reduce metals and to determine which would support grth as
respiratory electron acceptors. Nine metals [As(V), Cd(II), Co(HI), Cr(VI), Cu(II),
Fe(III), Ni(II), Se(VI and IV), and U(VI)] were used to test growth. Slowed growth was
found with 1 mM Cu(II), 1 mM Ni(II), 2.5 mM Se(VI), and 0.5 mM U(VI) and above.
While grth was observed with all metals, use of metals as a sole terminal electron
acceptor was found only with As(V), Cu(II), and Fe(III), although biomass yield
indicated some beneﬁt ﬁ'om Se(V I) and U(VI) as terminal electron acceptors. Valence
state analyses detected reduction of As(V to 111), F e(III to II) (provided both as soluble
ferric citrate and solid ferric oxide), Se(VI to O and IV to O), and U(VI to IV). While
Cu(II) and Co(III) showed the characteristic colors of reduction, changes in their valence
state were below detection limits. Extracellular vesicles (containing RNA surrounded by
both membrane and cell wall) were produced in response to Se(VI) and Se(IV) by
budding from cells. The vesicles but not cells contained reduced Se and appeared to be a
protection mechanism from Se toxicity. Both red and black crystalline Se(O) were
formed. Microarrays containing probes for D. hafniense were used to investigate changes

in gene expression when it was grown on nitrate, Fe(III), Se(VI), or U(VI) in comparison

to growth by pyruvate fermentation. Gene expression observed during nitrate reduction

can be characterized by two highly induced gene groups annotated for assimilatory

sulfate reduction and N2 ﬁxation-like ﬁtnction by the Nifl operon. The most highly up-

regulated genes during Se(VI) reduction were of unknown function. Comparing Fe(III)
reduction and U(VI) reduction revealed many genes induced in common, especially those
involved with coenzyme metabolism and energy metabolism, e.g. DMSO reductase and
NADH dehydrogenase. Genes induced with U(VI) reduction suggested responses to
uranium toxicity by rapid replication of DNA and an increase in motility through
induction of ﬂagellar genes and chemotaxis. Fe(III) reduction revealed the induction of
genes that could produce of large amounts of extracellular material, which was supported
by studies that showed Fe(III)-grown cells produced more bioﬁhn on silica and plastic
coated beads than cells fermenting pyruvate. Genome analysis and expression studies
suggested that this organism may grow as an acetogen, since the acetyl-CoA pathway
(Wood-Ljungdahl pathway) was up-regulated during Fe(III) and U(VI) reduction.

Growth studies demonstrated acetogenic growth with 30:30:40 gas combinations of

H2:C0:N2, H2:C02:N2 and CO:C02:N2 as the only carbon and energy sources,

consistent with using the acetyl-CoA pathway. This pathway is shared with Moorella
thermoacetica, the closest sequenced acetogenic relative. To test similar physiologies, D.
hafniense and M. thermoacetica were compared for grth and metal reduction. M.
thermoacetica reduced Se(VI to 0) and As(V to III), but no growth or reduction occurred

with soluble Fe(III).

To my parents, Leonard and Barbara

iv

ACKNOWLEDGEMENTS

The author would like to thank her advisors, Professor Terence Marsh and Professor
James Tiedje, for support, encouragement and enumerable discussions related to this
work. Thanks also go to the other members of the dissertation committee, Professor
Julius Jackson for many conversations regarding this research, as well as Professor
Robert Britton and Professor David Long. The author also extends her gratitude to Dr.

Sang-Hoon Kim and Dr. Shelly Kelly for their collaborations.

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................... ix
LIST OF FIGURES ......................................................................................................... x
CHAPTER I
INTRODUCTION ........................................................................................................... 1
OVERVIEW ............................................................................................................... 1
BACKGROUND ......................................................................................................... 5
Metals ...................................................................................................................... 5
Bacterial mediated metal reduction .......................................................................... 5
Highly studied systems of bacterial dissimilatory metal reduction ............................ 7
Desulﬁtobacteria .................................................................................................... 10
REFERENCES .......................................................................................................... 14
CHAPTER 11
Survey of Metal Electron Acceptor Use by Desulﬁtobacterium hafniense DCB-2 ......... 18
ABSTRACT .............................................................................................................. 18
INTRODUCTION ..................................................................................................... 19
METHODS ............................................................................................................... 21
Culture conditions ................................................................................................. 21
Time course measurements .................................................................................... 22
Yield analyses ....................................................................................................... 23
Calculations ........................................................................................................... 24
Metal valence state determination .......................................................................... 25
RESULTS ................................................................................................................. 26
Fermentative growth with different metals ............................................................. 26
Respiration with different metal electron acceptors ................................................ 28
Reduction of metals ............................................................................................... 31
Mineralization of metals by cells ........................................................................... 32
Biomass production with different metals .............................................................. 34
Grth yield of cells with metals ........................................................................... 36
DISCUSSION ........................................................................................................... 38
ACKNOWLEDGEMENTS ....................................................................................... 42
REFERENCES .......................................................................................................... 44
CHAPTER IH
Gene Expression of Desulﬁtobacterium hafniense DCB—2 Exposed to Nitrate, Fe(III),
Se(VI), and U(VI) and Proof of Its Acetogenic Growth ................................................. 46
ABSTRACT .............................................................................................................. 46
INTRODUCTION ..................................................................................................... 47
METHODS ............................................................................................................... 49
Bacterial genome sequence .................................................................................... 49

vi

Culture conditions ................................................................................................. 49

RNA extraction, cDNA synthesis and labeling ....................................................... 50
Microarray characteristics ...................................................................................... 51
Microarray hybridization and analysis ................................................................... 52
Experimental determination of acetogenic ability ................................................... 53
Culture purity ........................................................................................................ 54
RESULTS and DISCUSSION ................................................................................... 55
Pyruvate-fermenting cell metabolism ..................................................................... 55
Dissimilatory Fe(III)-reducing cell metabolism ...................................................... 59
U(VI)-reducing cell metabolism ........................................................................... 63
Nitrate-reducing cell metabolism ........................................................................... 66
Se(VI)-reducing cell metabolism ........................................................................... 69
Nitrogenase activity ............................................................................................... 72
Determination of an active acetyl-CoA reduction pathway ..................................... 73
Gene expression corresponding to cell morphology differences ............................. 76
Conclusions ........................................................................................................... 77
ACKNOWLEDGEMENTS ....................................................................................... 78
REFERENCES .......................................................................................................... 79
CHAPTER IV
Exterior Membrane-Bound Selenium Vesicles Produced by Desulﬁtobacterium hafniense
DCB-2 in Response to Selenate Exposure ...................................................................... 82
ABSTRACT .............................................................................................................. 82
INTRODUCTION ..................................................................................................... 83
METHODS ............................................................................................................... 85
Culture conditions ................................................................................................. 85
Growth .................................................................................................................. 86
Microscopy ............................................................................................................ 86
XANES and XRD analyses .................................................................................... 89
RESULTS ................................................................................................................. 91
Growth in the presence of Se ................................................................................. 91
Sphere formation ................................................................................................... 93
Production of two elemental Se precipitates ........................................................... 94
Chronology of spherical, membrane-bound Se vesicles .......................................... 97
DISCUSSION ......................................................................................................... 101
ACKNOWLEDGEMENTS ..................................................................................... 106
REFERENCES ........................................................................................................ 107
CHAPTER V
A Comparison of As(V), Fe(III) and Se(VI) Reduction by the Acetogens
Desulﬁtobacterium hafniense DCB-2 and Moorella thermoacetica .............................. 110
ABSTRACT ............................................................................................................ 1 10
INTRODUCTION ................................................................................................... 1 10
METHODS ............................................................................................................. 1 l 1
Bacterial strains ................................................................................................... l 11
Experimental determination of growth ................................................................. 112

vii

Measurements of growth with metals ................................................................... 112

RESULTS ............................................................................................................... 1 13
DISCUSSION ......................................................................................................... l 16
ACKNOWLEDGEMENTS ..................................................................................... 1 17
REFERENCES ........................................................................................................ l 18
CHAPTER VI
SUMMARY AND CONCLUSIONS ........................................................................... 120
APPENDIX A
Bioﬁlm-Forming Capacities of Desulﬁtobacterium hafniense DCB-2 .......................... 128
APPENDIX B
Minimal Medium and Culturing Methods for Desulﬁtobacterium hafniense DCB-2133
APPENDIX C
Protocols for Selenium Vesicles Isolation from Se-Reducing Cultures of
Desulﬁtobacterium hafniense DCB-2 with RNA and Protein Analyses ........................ 150

viii

LIST OF TABLES

Table Page
Chapter I

Table 1.1. Concentrations of metals used in this study and at NABIR FRC in Oak Ridge,

TN ..................................................................................................... 12
Chapter II

Table 2.1. Mineralization of different electron acceptors ..................................... 33

Table 2.2. Effect of electron acceptor concentration and carbon sources on grth

”Chin”... .................................................... 36

Table 2.3. Effect of different electron acceptor concentrations on grth yield .......... 37

Table 2.4. Carbon balance of electron donors (lactate and pyruvate) and yield of cells as

C (mMoles) ........................................................................................... 38
Chapter 111

Table 3.1. Effect of acetogenic conditions on cell yields of M. thermoacetica and

D. hafniense .......................................................................................... 74
Chapter V

Table 5.1. Generation times of M. thermoacetica and D. hafniense with metals. . . . . ....1 14

Table 5.2. Se(VI) mineralization ability of M. thermoacetica and D. hafniense ......... 115

r

LIST OF FIGURES

Images in this dissertation are presented in color.

Figure Page
Chapter I

Figure 1.1. Three methods of dissimilatory metal reduction by bacteria (metal shown

here is Fe(III) as an example). (A) Direct contact. (B) Extrusion of metal-chelators. (C)

Use of electron shuttle type molecules ............................................................ 9

Chapter 11
Figure 2.1. Organic acid concentrations in cultures grown fermenting pyruvate in the
presence of different metals. Best ﬁt curves and the standard error are shown. All metal
concentrations are 1 mM, except Cu(II) at 0.75 mM and U(VI) at 0.5 mM. Cultures were
monitored until stationary phase was achieved. (A) Pyruvate consumption over time.

(B) Acetate production over time ................................................................. 27

Figure 2.2. Generation times of D. hafniense grown with pyruvate under fermentation
conditions in the presence of the indicated metals. The generation time with pyruvate
fermentation alone (no metal present) is represented as 0.01 mM with each metal.

Standard error is shown for all measurements as vertical lines ............................... 28

Figure 2.3. Organic acid concentrations of cultures incubated with different metal
electron acceptors (respiration conditions). Cultures were monitored until stationary
phase was achieved. Best ﬁt curves and the standard error are shown. (A) Lactate

consumption over time. (B) Acetate production over time .................................... 30

Figure 2.4. Generation times of D. hafniense during respiration with lactate and different
electron acceptors. Standard error is shown for all measurements as vertical

lines .................................................................................................... 3 1

Chapter III
Figure 3.1. Schematic diagram of the key metabolism pathways during fermentation with
pyruvate. Menaquinone is abbreviated as “MK”, cytochrome as “cyt”, and
phosphoenolpyruvate as “PEP” ................................................................... 58

Figure 3.2. Schematic diagram of the key metabolism pathways during dissimilatory
Fe(III) reduction with lactate. Menaquinone is abbreviated as “MK”, cytochrome as
“cyt”, and phosphoenolpyruvate as “PEP”. Green arrows indicate genes similarly
expressed as pyruvate fermentation. Red arrows are genes up-regulated (in comparison

to pyruvate fermentation) in response to dissimilatory Fe(III) reduction .................... 62

Figure 3.3. Schematic diagram of the key metabolism pathways during dissimilatory
U(VI) reduction with pyruvate. Menaquinone is abbreviated as “MK”, and cytochrome
as “cyt”. Green arrows indicate genes similarly expressed as pyruvate fermentation. Red
arrows are genes up-regulated (in comparison to pyruvate fermentation) in response to

dissimilatory U(VI) reduction ..................................................................... 65

Figure 3.4. Schematic diagram of the key metabolism pathways during nitrate reduction
with lactate. ’ Menaquinone is abbreviated as “MK”, cytochrome as “cyt”, adenylyl
sulfate as “APS”, and phospho-adenylyl sulfate as “PAPS”. Green arrows indicate genes
similarly expressed as pyruvate fermentation. Red arrows are genes up-regulated (in

comparison to pyruvate fermentation) in response to nitrate reduction ..................... 68

Figure 3.5. Schematic diagram of the key metabolism pathways during Se(VI) reduction
with pyruvate. Menaquinone is abbreviated as “MK”, and cytochrome as “cyt”. Solid
green arrows indicate genes similarly expressed as pyruvate fermentation. Red arrows
are genes up-regulated (in comparison to pyruvate fermentation) in response to Se(VI)
reduction The vesicle is external in D. hafniense, though for simplicity depicted here as

internal ................................................................................................ 71

xi

3

Figure 3.6. Induced expression of the genes involved in nitrate reduction and nitrogen
ﬁxation (pathways shown at left). Relative induction is shown in grey scales (light to
dark; low to high induction). No evidence was found which supports gene expression of
nitrate and nitrite reductase genes in pyruvate-fermenting cells. Therefore, relevant cells

are shown in white .................................................................................. 72

Figure 3.7. SEM of D. hafniense DCB-2 exposed to different electron acceptors. (A)
Curved rod morphology during pyruvate fermentation. (B) Curved rod morphology
during nitrate respiration. (C) Elongated rod exhibited during U(VI) exposure. (D)
Elongated rod with vesicles during grth with Se(VI). (E) Streptobacilli made of
shorter rods during Fe(III) respiration. (F) SEM of F e(III) biomass. White arrow points
to cell in G. (G) Backscatter image of F e(III) biomass from E. White arrow points to cell
in F .................................................................................................... 76

Chapter IV
Figure 4.1. Growth of D. hafniense in sodium selenate medium. (A) Effect of sodium
selenate concentration on grth rate: (A) the zero concentration is displayed as 0.01 to
accommodate the log scale. (B) Growth with 1 mM sodium selenate measured by optical
density, pyruvate consumption and acetate production. Standard error is shown for all

measurements as vertical lines ..................................................................... 92

Figure 4.2. Morphological changes of D. hafniense cells when grown under pyruvate
fermentation conditions. (A) SEM of a cell grown on pyruvate with no selenium. (B-F)
Cells grown with 1 mM selenate. (B) SEM of elongated cells with buds or spheres at 24
h. (C) SEM of cells showing an increase in sphere number and development at 72 h.
White arrows in C and E indicate spheres veriﬁed to be attached to cells via visible
ﬁlament connections. (D) SEM of spheres both detached and cell-associated at 72 h. (E)
SEM at 72 h of ‘beads on a string’ (ﬁlament with Se spheres). (F) TEM at 72 h of cells
surrounded by spheres (very dark circles). Black arrows point to Se spheres where we
found Se and lipid-binding osmium predominant around the outer edge. This was
conﬁrmed by EDS ................................................................................... 93

xii

Figure 4.3. (A) Se K-edge XANES spectra from sterile selenium medium control
(negative control), D. hafniense cultures that formed black and red precipitates and the Sc
metal reference made at the APS at Argonne National Laboratory. The measured spectra
for red precipitate are shown as symbols, and the LCF is shown as a line. (B) XRD
spectra of red and black selenium precipitates from D. hafniense grown in 1 mM sodium

selenate compared to the XRD pattern of metallic elemental selenium ..................... 96

Figure 4.4. Representative SEM EDS and TEM EDS mapping of D. hafniense cells
grown in 1 mM sodium selenate. (A) SEM of Se exposed cell displaying buds and
vesicles. (B) SEM backseatter image of A. Arrows indicate regions of the image
scanned by EDS. Arrow 1 points to the cell, while arrow 2 points to an associated
vesicle. (C) EDS of cell in B at portions indicated by arrows 1 and 2. (D) TEM of
hexagonal crystals (or) and surrounding vesicles (v). (E) TEM backseatter image of
hexagonal crystals. (F) TEM of vesicles and cells. Line depicts the TEM EDS line scan
through four vesicles and two cells from points 3 to 4. (G) TEM EDS line scan of G
illustrating changes in carbon, osmium, and selenium from points 3 to 4 on the line.

Locations of the vesicles (v) and cells (c) seen in (F) are represented on the line ......... 99

Appendix A
Figure A.1. Bright ﬁeld microscopic images (100x magniﬁcation) of beads. (A) Cross-
section of a Dupont activated carbon bead showing the thin white plastic surface coating.
(B). Side view of a Siran glass bead illustrating its rough surface. (C) Top view of a

Siran glass bead showing its deeply pitted surface ............................................ 129

Figure A.2. Bioﬁhn cultures with Dupont and Siran beads after 25 medium transfers.
(A) Undisturbed (left) and shaken (right) cultures on Siran beads in DCB-l with
pyruvate. (B). Undisturbed (left) and shaken (right) cultures on Dupont beads in DCB-l
with pyruvate. (C) Culture grown on Dupont beads prior to medium transfer (left) and
following transfer (right) of fresh ferric citrate medium. (D) Culture grown on Siran
beads prior to medium transfer (left) and following transfer of fresh ferric citrate medium.

(E) Above view of the bioﬁhn produced by a culture grown on Siran beads in ferric

xiii

citrate medium ..................................................................................... 130

Appendix B
Figure B.1. Anaerobic inoculation process. (A) The materials necessary for culturing
into sterile medium. Only DCB-l medium is pictured with an anaerobic gas bottle and a
bottle dropper of ethanol. The vitamins and carbon source stock bottles are not shown.
(B) The position of the hand just prior to removal of the needle from the anaerobic gas
bottle is shown. The index ﬁnger is on the syringe plunger ready to expel gas to prevent
oxygen ﬁ'om coming in the needle. (C) The expelling of the needle contents into a serum
bottle. (D) The mixing and measuring of the liquid from a serum bottle for subsequent

transfer. The serum bottle is held upside down to obtain the liquid ........................ 145

Figure B.2. Black butyl rubber stoppers with crimped aluminum seals on a specially
designed 3-port 1 l ﬂask. These 3-port ﬂasks were used to hold maximally 500 ml

anaerobically ....................................................................................... 148

xiv

Chapter I

INTRODUCTION

OVERVIEW
Metals have played an intimate role in human history from everyday use to technological
innovations, arts, communications, and weaponry; however, humans are not the only
masters of metallurgy (Nriagu 1996). Long before humans evolved, microbes had found
their niche interacting with metals through redox reactions including both oxidative and
dissimilatory reduction to gain energy for growth. Presently, there are two dissimilatory
metal-reducing model organisms, Geobacter sulfurreducens and Shewanella oniedensis
MR-l that have been widely studied; however, both are Gram-negative Proteobacteria.
Given the central role of the cell membrane in metal reduction, having a model Gram-
positive dissimilatory metal-reducing organism is important to further scientiﬁc
understanding of how energy capture is realized in organisms with a very different cell
envelope. This study explores the metal interactions and dissimilatory metal-reducing
ability of the Gram-positive bacterium, Desulﬁtobacterz'um hafniense DCB-2, and its

potential role in bioremediation.

The purpose of this research was to investigate the potential of using D. hafniense as the
ﬁrst Gram-positive model organism for dissimilatory metal reduction. The guiding

hypothesis for this research was that D. hafniense is capable of reducing metals for

energy gain through dissimilatory, including respiratory, metal reduction. Additionally,
D. hafniense is capable of reducing the same metals as have been found to be reduced by

model, Gram-negative, metal-reducing organisms.

The main objective of this research was to determine the range of metal-reducing abilities

of Desulﬁtobacterium hafniense DCB-2. The speciﬁc objectives were:

0 to determine the spectrum of metals reduced and which metals support grth by

respiratory coupling of metal reduction to energy production.

- to determine whether there are differences in the cell morphology caused by
reduction of different metals, and if speciﬁc adaptations occurred to aid metal

reduction.

0 to identify genes and major pathways involved with Fe(III), Se(VI), and U(VI)

reduction.

Nine metals were selected for study to represent a broad spectrum of the periodic table
with a potential for biological reduction (Chapter 2). Metals surveyed were
geochemically diverse and included a pnicto gen (i.e. arsenic), a chalco gen (i.e. selenium),
an actinide (i.e. uranium), and transition metals (F aure 1991). The nine metals were: the
(I group metals Cd(II), Co(III), Cr(VI), Cu(II), Fe(III), and Ni(II); the p group metals

As(V) and Se(VI); and the f group metal U(VI). These are economically important

metals of concern to human health and to the US. Department of Energy for

bioremediation (US. Geological Survey 2005).

I found that D. hafniense reduced As(V), Fe(III), Se(VI), and U(VI), but color change
reactions also indicated reduction of Co(III), Cr(V I), and Cu(II) (Chapter 2). Growth
with all metals occurred when pyruvate fermentation conditions were provided, but
growth did not occur at the same rate. This is consistent with evidence for differing rates
of metal reduction by Shewanella and Geobacter (Liu 2002). Studies on D. hafniense
and other Desulﬁtobacterz'um species with As, Fe, Mn, S, and Se had previously reported
growth (Christiansen 1996, Venkateswaran 1999, Finneran 2002, Niggemyer 2001, and
Shelobolina 2003), but grth via respiratory metal reduction had not been veriﬁed.
Being able to gain energy from these metal reductions would provide an opportunity for

D. hafniense to grow speciﬁcally on the metal as well as gain energy from fermentation.

I observed morphological differences of cells reducing F e(III), Se(VI), and U(VI). These
cell morphologies were compared to the shape of cells grown by nitrate respiration and
pyruvate fermentation (Chapters 3 and 4). The cell morphology observed under
dissimilatory metal reduction of U(VI) was an elongation of the cells to 5 mn. This is
twice the length of a typical pyruvate-fermenting cell. Dissimilatory metal reduction of
Fe(III) produced streptobacilli-type chain growth of short cells. Moreover, Fe was
detected in the large amounts of extracellular material produced during Fe(III) reduction
(Chapters 2, 3, 4, and Appendix A). Se(VI) reduction yielded elongated cells with

extracellular vesicles containing Se. While spheres of Se had been reported previously

(Kessi 1999), I conclusively demonstrated the Se to be in membrane and cell wall-bound
vesicles. Negligible amounts of selenium were located in cells and the surrounding
matrix; however, the matrix did contain hexagonal crystals of selenium. This is also the
ﬁrst study to demonstrate two different colors of elemental Se precipitate being produced
by a single bacterium. The possibility for stages of development in the Se vesicles led to

the formulation of a model of Se vesicle development (Chapter 4).

Identifying the genes involved in metal reduction by D. hafniense is important as many
genes for metal reduction have not yet been identiﬁed despite the history of study on the
model dissimilatory metal-reducing bacteria G. sulfurreducens and S. oneidensis MR-l
(Fredrickson 2005). This limitation is in large part due to the fact that most studies
involved a gene-by-gene approach (Fredrickson 2005). Only recently have S. oneidensis
and G. sulfurreducens studies transitioned into the whole genome approach, allowing the
functional determination of hundreds of genes at once (Beliaev 2005, Bencheikh—Latmani
2005, Gao 2004, and Methé 2004). Moreover, studies of genes involved in metal
reduction by Gram-positive organisms are minimal. We found large differences in gene
expression amongst pyruvate fermentation, nitrate reduction, and Fe(III) reduction
(Chapter 3). U(VI) reduction resulted in similar up-regulated genes as F e(III) reduction,
while in contrast, Se(VI) had many similarities with nitrate reduction. Gene expression
supported trends observed in morphology and grth yield studies. To illustrate the
ﬁmctionality of highly expressed genes, the ability of D. hafniense to produce bioﬁlm
and grow acetogenically were also analyzed (Chapters 3 and Appendix A). Bioﬁlm

production was found under all grth conditions presented with varying yields

(Appendix A). Growth of D. hafniense as an acetogen was conﬁrmed with combinations

of H2, CO, and C02; thus, greatly expanding the breadth of its known metabolic

capabilities (Chapter 3). A phylogenetically related organism, Moorella thermoacetica,
was also tested for reduction of metals under anaerobic growth with lactate, and found to

reduce As(V) and Se(VI), but not soluble Fe(III) (Chapter 5).

BACKGROUND
Metals. The terms “metal” and “heavy metal” have been used in a variety of ﬁelds to
mean different things (e. g. politics, food, health, science, and engineering). For example,
these terms can be applied generally to mean elements that are capable of
bioaccumulating in plant or animal tissues (Duffus 2002). Elements may also be
classiﬁed as metals due to causing toxicity effects in organisms, provoking environmental
concern, or possessing a speciﬁc gravity of at least 4.5 (Duffus 2002). My use of
“metals” requires that elements possess a speciﬁc gravity greater than 4.5 and fall under
the environmentally signiﬁcant deﬁnition of bioaccumulation, toxicity, and human
concern. By this deﬁnition, both the metalloids As and Se can be included in the term

“metals”.

Bacterial mediated metal reduction. Metals may change mobilization potentials (e.g.
solubility) by passive or active means. Passive metal mobilization depends upon the
physicochemical characteristics surrounding the metal that affects its solubility, such as
pH, temperature, water activity, and neighboring materials (physical environment).

Passive mobilization does not include environmental inﬂuences caused by organisms.

5

Active metal mobilization requires direct involvement of an organism to mobilize a
metal. Active mobilization by bacteria may involve passive or active interactions with
metals. Passive interactions of bacteria with metals may include pH, cell wall

interactions, ion exchange, or indirect removal associated with exopolysaccharides and

H28 production (Lloyd 2003). Active interactions may include enzymatic reactions,

chelation, and metabolism (Lloyd 2003). This metabolism may either oxidize or reduce

the metal.

Active metal reduction by bacteria occurs for two reasons. The ﬁrst is to prevent toxic
conditions in the environment surrounding a cell (without energy gain for the cell from
the reaction). The second reason is the uptake and transformation of metals to beneﬁt the
cell, which occurs via assimilatory and/or dissimilatory metal reduction. Assimilatory
metal reduction incorporates metals into a cell, such as to form metaloenzymes.
Dissimilatory metal reduction is the use of metals as electron acceptors for energy
generation to drive metabolism without the incorporation of the metal. This also includes
the use of metals as an electron sink for metabolism. This will prevent a build up of

radicals in a cell allowing metabolism to continue.

Three types of dissimilatory metal reduction are known to occur (Luu 2003 and Nevin
2002) (Figure 1.1). The ﬁrst requires direct contact between the metal and the cell
membrane for metal reduction to occur (Figure 1.1A). The second is the production and
extrusion of metal-chelating molecules from a cell (Figure 1.13). These chelated (metal-

containing) complexes are then taken up by a cell and the metal is subsequently reduced

6

 

within the cell. The third method employs electron shuttle-type molecules (e. g. humic
acids) to travel to and indirectly reduce metals (Figure 1.1C). This is the only method
that does not involve direct reduction of a metal by a cell. Instead, these electron-shuttle
molecules, which are capable of undergoing repetitive oxidation and reduction reactions,
provide an intermediate electron acceptor for a cell. These electron shuttle molecules
may be environmentally-derived or produced by cells, but they function as an electron

intermediate between a metal and a cell.

Highly studied systems of bacterial dissimilatory metal reduction. Studies on
dissimilatory metal reduction have centered around two Gram-negatives, Shewanella and
Geobacter. Each genus has more than one member whose genome has been fully
sequenced. The focus of research with Shewanella has been on S. oneidensis MR—l; for
Geobacter it is G. sulfurreducens. Though both are model organisms for dissimilatory
metal reduction, most studies have only tested if metals were reduced and not if energy
gain occurred. S. oneidensis MR-l (a y-Proteobacteria) was isolated from sediment at
Oneida Lake, New York in 1931 (Venkateswaran 1999). This facultative anaerobe has a
polar ﬂagella and reduces a range of compounds, including: fumarate, iodate, nitrate,
nitrite, oxygen, sulﬁte, elemental sulﬁrr, thiosulfate, As(V), Co(III) EDTA, Cr(VI),
Fe(III), Mn(IV), Mn(III), Se(VI), Tc(VII), and U(VI) (Venkateswaran 1999). G.
sulfurreducens (a 8-Proteobacteria) was isolated from an enrichment of iron-reducing
bacteria taken ﬁ'om hydrocarbon-contaminated soil near Norman, Oklahoma (Caccavo
1994). This anaerobic rod is non-motile except when expressing ﬂagella in the presence

of metal oxides as a chemotactic response to the solid metal (Childers 2002 and Lin

2004). It is capable of reducing elemental sulfur, As(V), Au(III), Cr(VI), F e(III), Hg(II),
Mn(VI), Se(V I), Tc(VII), and U(VI) (Methé 2003). Neither bacterium ferments sugars or

forms spores.

An array of metals are reduced by both S. oneidensis and G. sulfurreducens; however,
they differ in their distinct methods of dissimilatory metal reduction of Fe(III). G.
sulfurreducens requires direct contact with Fe metals in order to effectively reduce them
(Lovley 1993). This is due to the electron shuttle machinery being directly mounted on
the cell membrane. Although S. oneidensis does also reduce metals via direct membrane
contact, it does not require direct contact for Fe reduction as it is able to make use of
external electron shuttle-type molecules (e. g. humics) as well as produce electron shuttle-
type molecules (e. g. quinones) (Bencharit 2005, Kim 2005, Leang 2003, Lloyd 2003, Luu

2003, Magnuson 2001 and Mehta 2005).

   
 
  

Fe(ll) A

Oxidized Product

Reduced Substrate

  
   
 

  

Chelator A DMR

Bacteria
E Fe(ll)

Oxidized Product

 

      

Fe(III)-
chelated
complex Reduced Substrate
Electron C
Shuttle
e Oxidized
- Molecule Product
/
Fe(ll)
Reduced
Electron Substrate
Shuttle
Molecule

Figure 1.1. Three methods of dissimilatory metal reduction by bacteria (metal shown here is Fe(III) as

an example). (A) Direct contact. (B) Extrusion of metal-chelators. (C) Use of electron shuttle type

molecules.

The genome sequence of S. oneidensis MR-l was completed in 2002 by The Institute for
Genomic Research (TIGR). The chromosome is 4.97 Mb with a 162 kb plasmid and has
45.9% G-C content (Heidelberg 2002). The G. sulfurreducens genome sequence was
completed in 2003 by TIGR and contains a 3.81 Mb chromosome with 60.9% G-C
content (Methé 2003). There are 4931 predicted protein-encoding genes in S. oneidensis
and 3466 in G. sulfurreducens (Heidelberg 2002 and Methé 2003). Upon obtaining the
full genomic sequence for S. oneidensis and G. sulfurreducens, the genomes were
examined for unexplored ﬁmctions (Fredrickson 2005 and Methé 2003). From these
analyses, a higher than expected number of c-type cytochromes were found for both
bacteria (42 in S. oneidensis and 90 in G. sulfurreducens) (Fredrickson 2005).
Cytochromes are heme-containing proteins that function as electron carriers in the cell,
and are important in shuttling electrons during metal reduction (Beliaev 2001 and Butler

2004).

Desulﬁtobacteria. The Desulﬁtobacterium genus was established in 1994 when
Desulﬁtobacterz‘um dehalogenans was named after being discovered for its ability to
reductively dechlorinate chlorophenols (Utkin 1994). Since then, 17 representatives of
_ the genus (composed of six species) have been isolated (Villemur 2006). All members
are anaerobic, Gram-positive, slightly curved rods that are motile by one or more ﬂagella

(Villemur 2006). Taxonomically falling with the low G-C Clostridia, they reside in the

10

family Peptococcaceae, order Clostridiales. All desulﬁtobacteria are capable of

fermentative growth using pyruvate to form primarily acetate and C02. Desulﬁtobacteria

respire with lactate (electron donor) using nitrate, thiosulfate or sulﬁte as an electron
acceptor to produce ammonia and sulﬁde. Only D. chlororespirans C023 is unable to

respire nitrate (Sanford 1996).

Desulﬁtobacteria are quite ubiquitous in the environment. They have been found in
methanogenic lake sediment (D. dehalogenans), tetrachloroethene contaminated soil (D.
hafniense Y51), compost soil (D. chlororespirans C023), clay soil (D. hafniense G-2),
and human feces (D. hafniense DP-7) (Utkin 1994, Suyama 2001, Sanford 1996,
Shelobolina 2003, and van de Pas 2001). A study of soil samples from 44 sites
throughout the province of Quebec, Canada found Desulﬁtobacterium species to reside in
over 70% of the samples (Lanthier 2001). The Desulﬁtobacterium genus has been
documented in Fe(III)-enrichment cultures from rice paddy soil and subsurface sediment
as well as in contaminated military base locations (Kostka 2002). One such military site
is the aquifer at area 6, Dover Air Force Base, Delaware, which is contaminated with
chlorinated solvents such as vinyl chloride and cis-dichloroethene (Davis 2002).
Desulﬁtobacterz’um was also found at the Field Research Center (FRC) at Oak Ridge
National Laboratory, Oak Ridge, TN, which is under the jurisdiction of the US.
Department of Energy (Brooks 2001). It is heavily contaminated with uranium,
chromium, technetium, chlorinated solvents such as tetrachloroethylene, and nitric acid

(Table 1.1) (Brooks 2001).

ll

Table 1.1. Concentrations of metals used in this study and at
NABIR FRC in Oak Ridge, TN

 

 

Aqueous Sludge Study
Phase (ppm or pg/g) Concentration
Metal Abbreviation (mM) dry weight (mM)
Iron Fe 1208 92,031 50 mM
Uranium U 316.57 620 0.5 mM
Nickel Ni 128 98.8 1 mM
Chromium Cr 60 163.9 1 mM
Copper Cu 44 145.3 0.75 mM
Cobalt Co 1.4 3.3 1 mM
Arsenic As 0.115 32.5 1 mM
Selenium Se 0.033 0.02 1 mM
Cadmium Cd - <06 1 mM

 

Source: Brooks 2001.

Two species of Desulﬁtobacterium have been isolated ﬁ'om metal contaminated sites. D.
metallireducens was cultured from a uranium-contaminated aquifer, which is able to
reduce ferric citrate, manganese(IV) oxide, and elemental sulfur (F inneran 2002).
However, unlike other members of the genus, D. metallireducens does not form spores.
Desulﬁtobacterium sp. strain GBFH was enriched from arsenic contaminated lake
sediment, and found capable of reducing the metals As(V), Fe(III), Mn(VI), and Se(VI)
(Niggemyer 2001). After being isolated ﬁom clay soil, D. hafniense G-2 was found to

reduce Fe(III to H) (Shelobolina 2003).

Since its isolation from municipal sludge in 1992, studies on Desulﬁtobacterium
hafniense DCB-2 have largely focused on the spectrum and mechanism of its

dehalogenation ability, such as dechlorination of chlorophenols and chloroethenes

l2

. I‘T—‘i

(Christiansen 1996 and Madsen 1992). It is an anaerobic, Gram-positive, spore-forming
rod, motile by one or two terminal ﬂagella and an optimum growth temperature of 37°C
(Christiansen 1996). Besides having been discovered to generate electricity, it is capable
of grth by fermentation and respiration with nitrate, nitrite, sulﬁte, elemental sulfur,
thiosulfate, Fe(III), and Mn(IV) (Christiansen 1996, Madsen 1992, and Milliken 2007).
Two of my laboratory ﬁndings were the ability of D. hafniense to grow acetogenically
and to produce copious amounts of bioﬁlm under conditions of both pyruvate

fermentation and dissimilatory Fe(III) reduction (Chapter 5 and Appendix A).

From sequencing done by the Joint Genome Institute (JGI), D. hafniense was found to
contain a 5.2 Mb chromosome of 47.5% G-C content with 4715 candidate protein-
encoding genes. It is one of only two members to be sequenced ﬁ'om this genus (D.
hafniense Y-51 has also been sequenced). Scanning the genome for novel functional
gene annotations uncovered catalase and superoxide dismutase genes, 43 annotated
cytochromes, as well as genes for arsenic reduction, chemotaxis, and bacteriophage.
Several genes for membrane transport systems for Co, Cu, Fe, Mn, and Ni were also

found.

13

REFERENCES

Beliaev, A. S., D. M. Klingeman, J. A. Klappenbach, L. Wu, M. F. Romine, J. M.
Tiedje, K. H. Nealson, J. K. Fredrickson, and J. Zhou. 2005. Global transcriptome

analysis of Shewanella onez‘densis MR-l exposed to different terminal electron acceptors.
J. Bacteriol. 187:7138-7145.

Beliaev, A. S., D. A. Saﬂ‘arini, J. L. McLaughlin, and D. Hunnicutt. 2001. MtrC, an
outer membrane decahaem c cytochrome required for metal reduction in Shewanella
putrefaciens MR-l. Mol. Microbiol. 39:722-730.

Bencharit, S. and M. J. Ward. 2005. Chemotactic responses to metals and anaerobic
electron acceptors in Shewanella oneidensis MR-l. J. Bacteriol. 18725049-5053.

Bencheikh-Latmani, R., S. Middleton Williams, L. Haucke, C. S. Criddle, L. Wu, J.
Zhou, and B. M. Tebo. 2005. Global transcriptional proﬁling of Shewanella oneidensis
MR-l during Cr(VI) and U(VI) reduction. Appl. Environ. Microbiol. 71:7453-7460.

Brooks, S. C. 2001. Waste characteristics of the former S-3 ponds and outline of uranium
chemistry relevant to NABIR Field Research Center studies. NBIR Field Research
Center, Oak Ridge, TN.

Butler, J. E., F. Kaufmann, M. V. Coppi, C. Nunez, and D. R. Lovley. 2004. MacA, a
diheme c-type cytochrome involved in F e(III) reduction by Geobacter sulfurreducens. J.
Bacteriol. 186:4042-4045.

Caccavo, Jr., R, D. J. Lonergan, D. R. Lovley, M. Davis, J. F. Stoly, and M. J.
Mclnerney. 1994. Geobacter sulfurreducens sp. nov., a hydrogen- and acetate- oxidizing
dissimilatory metal-reducing microorganism. Appl. Environ. Microbiol. 60:3752-3759.

Childers, S. E., S. Ciufo, and D. R. Lovley. 2002. Geobacter metallireducens accesses
insoluble F e(III) oxide by chemotaxis. Nature. 416:767-769.

Christiansen, N. and F. K. Ahring. 1996. Desulﬁtobacterium hafniense sp. nov.,
anaerobic, reductively dechlorinating bacterium Int. J. Syst. Bacteriol. 46:442-448.

Davis, J. W., J. M. Odom, K. A. Deweerd, D. A. Stahl, S. S. Fishbain, R. J. West, G.
M. Klecka, and J. G. DeCarolis. 2002. Natural attenuation of chlorinated solvents at

Area 6, Dover Air Force Base: characterization of microbial community structure. J.
Contain. Hydro]. 57:41-59.

Duffus, J. H. 2002. “Heavy metal” - a meaningless term? (IUPAC Technical Report).
Pure App. Chem. 74:793-807.

Faure, G. Principles and Applications of Inorganic Geochemistry, MacMillan Publishing
Co.: New York, 1991.

14

Finneran, K. T., H. M. Forbush, C. V. Van Praagh, and D. R. Lovley. 2002.
Desulﬁtobacterium metallireducens sp. nov., an anaerobic bacterium that couples growth

to the reduction of metals and humic acids as well as chlorinated compounds. Int. J. Syst.
Evol. Microbiol. 52: 1929-1935.

Fredrickson, J. K. and M. F. Romine. 2005. Genome-assisted analysis of dissimilatory
metal-reducing bacteria. Curr. Opin. Microbio 1. 1621-6.

Gao, H., Y. Wang, X. Liu, T. Yan, L. Wu, E. Alm, A. Arkin, D. K. Thompson, and J.
Zhou. 2004. Global transcriptome analysis of the heat shock response of Shewanella
oneidensis. J. Bacteriol. 186:7796-7803.

Heidelberg, J. F., I. T. Paulsen, K. E. Nelson, E. J. Gaidos, W. C. Nelson, T. D. Read,
J. A. Eisen, R. Seshadri, N. Ward, B. Methé, R. A. Clayton, T. Meyer, A. Tsapin, J.
Scott, M. Beanan, L. Brinkac, S. Daugherty, R. T. DeBoy, R. J. Dodson, A. S.
Durkin, D. H. Haft, J. F. Kolonay, R. Madupu, J. D. Peterson, L. A. Umayam, 0.
White, A. M. Wolf, J. Vamathevan, J. Weidman, M. Impraim, K. Lee, K. Berry, C.
Lee, J. Mueller, H. Khouri, J. Gill, T. R. Utterback, L. A. McDonald, T. V.
Feldblyum, H. O. Smith, J. C. Venter, K. H. Nealson, and C. M. Fraser. 2002.
Genome sequence of the dissimilatory metal ion-reducing bacterium Shewanella
oneidensis. Nature Biotech. 20:1 118-1123.

Kessi, J., M. Ramuz, E. Wehrli, M. Spycher, and R. Bachofen. 1999. Reduction of
selenite and detoxiﬁcation of elemental selenium by the phototrophic bacterium
Rhodospirillum rubrum. Appl. Environ. Microbiol. 654734-4740.

Kim, B. C., C. Leang, Y. H. Ding, R. H. Glaven, M. V. Coppi, and D. R. Lovley.
2005. OmcF, a putative c-type monoheme outer membrane cytochrome required for the

expression of other outer membrane cytochromes in Geobacter sulfurreducens. J.
Bacteriol. 187:4505-4513.

Kostka, J. E., D. D. Dalton, H. Skelton, S. Dollhopf', and J. W. Stucki. 2002. Growth
of iron(III)-reducing bacteria on clay minerals as the sole electron acceptor and

comparison of growth yields on a variety of oxidized iron forms. Appl. Environ.
Microbiol. 68:6256-6262.

Lanthier, M., R. Villermur, F. Lépine, J. Bisaillon, and R. Beaudet. 2001. Geographic
distribution of Desulﬁtobacterium frappieri PCP-l and Desulﬁtobacterium spp. in soils
ﬁ'om the province of Quebec, Canada. FEMS Microbiol Ecol. 36: 185-191.

Leang, C., M. V. Coppi, and D. R. Lovley. 2003. Och, a c-type polyheme
cytochrome, involved in Fe(III) reduction in Geobacter sulfurreducens. J. Bacteriol.
18522096-2103.

Lin, W. C., M. V. Coppi, and D. R. Lovley. 2004. Geobacter sulfurreducens can grow
with oxygen as a terminal electron acceptor. Appl. Environ. Microbiol. 70:2525-2528.

15

Liu, C. Y. A. Gorby., J. M. Zachara, J. K. Fredrickson, and C. F. Brown. 2002.
Reduction kinetics of Fe(III), Co(III), U(VI), Cr(VI), and Tc(VII) in cultures of
dissimilatory metal-reducing bacteria. Biotech. Bioengin. 80:637-649.

Lloyd, J. R. 2003. Microbial reduction of metals and radionuclides. FEMS. Microbiol.
Rev. 27:411-425.

Lloyd, J. R., C. Leang, A. L. Hodges Myerson, M. V. Coppi, S. Ciufo, B. Methé, S. J.
Sandler, and D. R. Lovley. 2003. Biochemical and genetic characterization of PpcA, a
periplasmic c-type cytochrome in Geobacter sulfurreducens. Biochem. J. 369( 1):153-
161.

Lovley, D. R. 1993. Dissimilatory metal reduction. Annu. Rev. Micriobiol 47:263-290.

Luu, Y.-S. and J. A. Ramsay. 2003. Review: microbial mechanisms of accessing
insoluble Fe(III) as an energy source. World J. Microbiol Biotech. 19:215-225.

Madsen, T. and D. Licht. 1992. Isolation and Characterization of an anaerobic
chlorophenol-transforming bacterium. Appl. Environ. Micro. 58:2874-2878.

Magnuson, T. S., N. Isoyama, A. L. Hodges-Myerson, G. Davidson, M. J. Maroney,
G. G. Geesey, and D. R. Lovley. 2001. Isolation, characterization and gene sequence
analysis of a membrane-associated 89 kDa F e(III)-reducing cytochrome c from
Geobacter sulfurreducens. Biochem. J. 359: 147- 152.

Mehta, T., M. V. Coppi, S. E. Childers, and D. R. Lovley. 2005. Outer membrane
protein c-type cytochromes required for Fe(III) for Fe(III) and Mn(IV) oxide reduction in
Geobacter sulfurreducens. Appl. Environ. Microbiol. 71 :8634-8641.

Methé, B. A., K. Webster, K. P. Nevin, J. E, Butler, and D. R, Lovley. 2004. DNA
microarray analysis of nitrogen ﬁxation and Fe(III) reduction in Geobacter
sulfurreducens. Appl. Environ. Microbiol. 71 :2530-2538.

Methé, B. A., K. E. Nelson, J. A. Eisen, I. T. Paulsen, W. Nelson, J. F. Heidelberg, D.
Wu, M. Wu, N. Ward, M. J. Beanan, R. J. Dodson, R. Madupu, L. M. Brinkac, S. C.
Daugherty, R. T. DeBoy, A. S. Durkin, M. Gwinn, J. F. Kolonay, S. A. Sullivan, D.
H. Haft, J. Selengut, T. M. Davidsen, N. Zaf'ar, 0. White, B. Tran, C. Romero, H. A
Forberger, J. Weidman, H. Khouri, T. V. Feldblyum, T. R. Utterback, S. E. Van
Aken, D. R. Lovley, and C. M. Fraser. 2003. Genome of Geobacter sulfurreducens:
metal reduction in subsurface environments. Science. 302:1967-1969.

Milliken, C. E. and H. D. May. 2007. Sustained generation of electricity by the spore-

forming, Gram-positive, Desulﬁtobacterium hafniense strain DCB2. Appl. Microbiol.
Biotechnol. 73:1180-1189.

l6

Nevin, K. P., and D. R. Lovley. 2002. Mechanisms for accessing insoluble Fe(III) oxide
during dissimilatory Fe(III) reduction by Geothrixfermentans. Appl. Environ. Microbiol.
68(5):2294-2299.

Niggemyer, A., 8. Spring, E. Stackebrandt, and R. F. Rosenzweig. 2001. Isolation and
characterization of a novel As(V)-reducing bacterium: implications for arsenic
mobilization and the genus Desulﬁtobacterium. Appl. Environ. Microbiol. 67:5568-5580.

Nriagu, J. O. 1996. A history of global metal pollution. Science. 272:223-224.

Posey, J. E. and F. C. Gherardini. 2000. Lank of role for iron in the Lyme disease
pathogen. Science. 288: 1651-1653.

Sanford, R. A., J. R. Cole, F. E. Loffler, and J. M. Tiedje. 1996. Characterization of
Desulﬁtobacterium chlororespirans sp. nov., which grows by coupling the oxidation of
lactate to the reductive dechlorination of 3-chloro-4-hydroxybenzoate. Appl. Environ.
Microbiol. 62:3800-3808.

Shelobolina, E. S., C. G. VanPraagh, and D. R. Lovley. 2003. Use of ferric and ferrous
iron containing minerals for respiration by Desulﬁtobacterium frappieri. Geomicrobiol. J.
20:143-156.

Suyama, A., R. Iwakiri, K. Kai, T. Tokunaga, N. Sera, and K. Furukawa. 2001.
Isolation and characterization of Desulﬁtobacterium sp. strain Y51 capable of efﬁcient

dehalogenation of tetrachloroethene and polychloroethanes. Bioschi. Biotechnol.
Biochem. 65:1474-1481.

Utkin, 1., C. Woese, and J. Wiegel. 1994. Isolation and characterization of
Desulﬁtobacterium dehalogenans gen. nov., sp. nov., an anaerobic bacterium which
reductively dechlorinates chlorophenolic compounds. Int. J. Syst. Bacteriol. 44:612-619.

U.S. Geological Survey, Mineral Commodities Summaries, January 2005.

van de Pas, B. A., H. J. Harmsen, G. C., Raangs, W. M. de Vos, G. Schraa, and A. J.
Stams. 2001. A Desulﬁtobacterium strain isolated from human feces that does not
dechlorinate chloroethenes or chlorophenols. Arch. Microbiol. 17 5:389-394.

Venkateswaran, K., D. P. Moser, M. E. Dollhopf, D. P. Lies, D. A. Saﬂ'arini, B. J.
MacGregor, D. B. Ringelberg, D. C. White, M. Nishijima, H. Sano, J. Burghardt, E.
Stackebrandt, and K. H. Nealson. 1999. Polyphasic taxonomy of the genus Shewanella
and description of Shewanella oneidensis sp. nov. Int. J. Syst. Bacteriol. 49:705-724.

Villemur, R., M. Lanthier, R. Beaudet, and F. Lépine. 2006. The Desulﬁtobacterium
genus. FEMS Microbiol. Rev. 30:706-733.

17

Chapter 11

Survey of Metal Electron Acceptor Use by Desulﬁtobacten‘um hafniense DCB-Z

ABSTRACT
Metal reduction has largely been studied in Gram-negative bacteria where reduction
pathways are localized in the cell membrane. Here, we report the metal reduction
capabilities of a Gram-positive dehalogenator, Desulﬁtobacterium hafniense strain
DCB-2. Cultures were grown in liquid media with nine different metals as potential
electron acceptors. Growth was unaffected with Cr(VI), Cd(II) and < 1 mM of Cu(II) and
Ni(II). Growth rate was enhanced with 0.05 mM Co(III) and all tested concentrations of
As(V) as well as concentrations of S 1 mM Se(VI). Growth was depressed in the
presence of higher concentrations of Cu(II), Ni(II), Se(VI), and all concentrations of

U(VI). Growing cultures reduced soluble Fe(III), U(VI), As(V), and Se(VI) to insoluble

products and solubilized solid Fe(III). The growth yield per e- transferred was highest

with Fe(III) followed by U(VI), Se(VI), and As(V), all with cell yields above
fermentation alone. The carbon balance of cultures grown with Fe(III) showed much
higher carbon in cells and soluble extracellular products than provided ﬁ'om the substrate
carbon. D. hafniense most likely is using the citrate chelate used to supply the Fe(III)
along with some carbon ﬁxation from the acetyl-CoA pathway as the source of the extra
carbon. The results show D. hafniense can gain energy for growth, i.e. respiration, with

reduction of Fe(III), nitrate, and As(V), some energy beneﬁt from U(VI) and Se(VI)

l8

reduction, reduction of Cr(VI), Co(III) and Cu(II) but no signiﬁcant energy beneﬁt, and

no reduction of Cd(II) and Ni(II).

INTRODUCTION
Metals are naturally abundant within the environment, but they can attain toxic
concentrations as a result of mining, manufacturing, waste disposal, and other human
activities, resulting in potential hazards to human health (J arup 2003). Maximizing
bioremediation of environments contaminated with metals may be enhanced by
understanding the bacteria performing the desired task. Bacterial interactions with metals
have involved both in vitro and in situ studies (Lovley 1993). Controlled conditions of in
vitro studies are critical to provide information on basic requirements for the desired

reduction under in situ conditions.

Determining the spectrum of metal redox ability of a bacterium provides information of
how the organism interacts with metals and allows inferences to be made with other
related organisms. Bacteria appear to reduce metals for three reasons: to prevent metal
toxicity, to assimilate and incorporate a metal into cells for metabolic use, and to obtain
energy through dissimilatory metal reduction (Lovley 1993 and Methé 2004). Metal
reduction by bacteria has been reported for arsenic, chromium, cobalt, copper, iron,
selenium, uranium and other metals (Lloyd 2003, Stolz 2002). Many studies of metal
reduction by cultivated bacteria have used media supplemented with yeast extract and
other complex nutrients, complicating the interpretation of the impact of metals on

bacterial growth. Few metal reduction studies have examined in a comparative study one

19

bacterium’s ability to reduce a spectrum of metals in one publication, and these that have
were performed with Gram-negative organisms. As many of these reduction pathways
have been found to localize in the membranes of Gram-negative bacteria, here we have
chosen to study metal reduction in Gram-positive Desulﬁtobacterium hafniense DCB-2

(Liu 2002, Lovley 2002 and 1993).

D. hafniense possesses many qualities that should prove beneﬁcial to bioremediation,
such as the ability to form bioﬁlrns and possessing a terminal ﬂagella for responding to
chemotactic responses it detects (Appendix A). Having been isolated as a dehalogenator
respiring chlorophenol, D. hafniense has been traditionally studied for its ability to
dehalogenate (Madsen 1992). It is a spore-former able to survive harsh conditions such
as acidiﬁcation that often accompanies metal contamination, and was shown to grow with
and reduce Fe(III) and Mn(IV) (Christiansen 1996). The presence of an As(V) reduction
pathway and ability to reduce (As) was discovered by Niggemyer (2001), but result
interpretation of this study is complicated by the inclusion of yeast extract in the medium.
Moreover, the reduction products and the effect on cell grth were not reported in both
studies. D. hafniense has been found at the Field Research Site, Oak Ridge National
Laboratory (Oak Ridge, TN), which is heavily contaminated with U and Cr (Kostka
2002). The ﬁll] range of contaminating metals that this bacterium is capable of reducing
has not been determined. In this report, we explore the ability of D. hafniense in minimal
media to reduce nine different metals, including whether they enhance or inhibit growth,
their valence state and ligand binding. The metals are: As(V), Cd(II), Cr(VI), Co(III),

Cu(II), Fe(III), Ni(II), Se(VI), and U(VI). This provides the ﬁrst broad view of

20

interactions a Gram-positive bacterium has with a spectrum of metals.

METHODS
Culture conditions. For facilitated electron acceptor and metal toxicity analyses, D.
hafniense was cultured in a modiﬁed DCB-l medium containing 20 mM sodium pyruvate
(pyruvate) (Ldfﬂer 1996 and Chapter 4). For dissimilatory metal reduction analyses, a
modiﬁed carbonate-buffered freshwater (CBF) medium containing 20 mM sodium lactate

(lactate) was used with metals provided as the sole electron acceptor. Modiﬁed CBF

medium consisted of (per liter): 0.25 g NH4C1, 0.1 g KCl, 10 ml mineral mix, 0.6 g

NaHzPO4 X H20, 2.5 g NaHCO3 (to buffer medium to pH 7), and 0.25 mg resazurin

(Lovley 1988). For CBF medium, salts and trace elements were mixed, heated and

cooled under COzzNz (20:80) (Miller 1974). This medium was then dispensed and

autoclaved for 20 min. Modiﬁed DCB-l medium was cooled under a gas mixture of
COZ;N2 (5:95). Using strict anaerobic technique, 1x Wolin vitamins (Wolin 1963), the

carbon source, and a metal electron acceptor from separate sterilized stock solutions were
added prior to inoculation (Tschech 1984, Widdel 1983). The following metal salts were
each tested separately in DCB-l and CBF media: sodium arsenate, sodium selenate,
sodium selenite, cupric chloride, cupric sulfate, ferric citrate (CBF only), ferric oxide
(CBF only), nicke1(II) chloride, cobalt(III) chloride, cadmium(II) chloride, and uranyl
acetate. Ferric citrate medium was made using CBF medium with 50 mM ferric citrate
being added prior to salts, lactate and vitamins as described above (Lovely 1996). The

controls were medium with metal without bacteria, medium without metals inoculated

21

with a live culture, medium with metals and heat-killed bacteria inoculated as a cell
pellet, and a heat-killed culture to which metals were then added. None of the controls

displayed evidence of metal reduction in any measurement taken.

Time course measurements. Following prescreening for growth in the metal condition,
a minimum of 20 biological replicates were measured per condition over time.
Inoculated cultures were incubated at 37°C in 1 l anaerobic bottles with side-arm tubes
attached to allow for optical density measurements without removing culture. Optical
density at 600 nm was obtained by mixing and inverting bottles such that culture in the

side-arm tubes could be read using a Spectronic 20D spectrophotometer (Milton Roy,
Rochester, NY). Samples were aseptically removed in an anaerobic glove box of H2:N2
(4:96) at the start of the experiment and every 12 h thereafter for organic acid analysis as
monitored by HPLC, colony counts, color change over time, and microscopy (Chapter 4).
Samples from each time point were also ﬁ‘ozen at -20°C for further analyses. Samples
were no longer taken once the culture reached stationary phase. Time point samples for
HPLC were ﬁltered through a 0.45 pm ﬁlter prior to run on the HPLC to measure acetate,

pyruvate, and lactate as described in Chapter 4. Time point samples of all cultures were

inoculated in triplicate onto R2A supplemented with 10 mM pyruvate in an anaerobic

glove box of szNz (4:96). The plates were incubated at 25°C for 10-12 days prior to

counting colonies.

The state of metal reduction was initially inferred (and later conﬁrmed with XANES

22

data) based on color changes of the cultures. Color change reactions of metal reduction
are: clear to red/black for Se, clear to yellow for As, burnt orange to brown to olive green
to yellow for Fe, yellow to clear to green/gray to dark brown for U, blue to green to
yellow to brown for Cu, purple to pink to clear to brown for Co, clear to bright yellow for

Cd, and green to clear to brown for Ni.

Morphology was observed using light microscopy with methylene blue or Gram stain,
scanning electron microscopy (SEM), SEM energy dispersive spectroscopy (SEM EDS),
transmission electron microscopy (TEM), and TEM energy dispersive spectroscopy
(TEM EDS) (Chapter 4). Culture purity was determined throughout the study by
ﬂuorescent in situ hybridization (FISH) with Cy5-labeled, D. hafniense-speciﬁc rDNA

probe (Lanthier 2001) and DAPI staining (Bloem 1995).

Yield analyses. TOC analysis was performed with all samples on the last time point of
the growth curves to determine the mass and carbon balance of the metal reducing
reactions. Medium controls without cells were also processed as were additional controls
without cells but with metal electron acceptor added. Of the controls with metal electron
acceptor added, only the Fe(III) citrate CBF medium control without cells had
background levels of organic carbon, and thus is the only metal medium control reported
here. Additional replicate cultures in Fe(III) citrate had the initial pellet and supernatant

processed separately to aid in biomass analysis.

Prior to submission for TOC analysis, three replicate ﬁozen culture samples were thawed

23

and pelleted by spinning at 2000 g for 30 min. The supernatant was decanted and mixed
with two volumes of 95% ethanol and spun at 2000 g for 30 min to precipitate small
soluble extracellular polymeric substances. The two pellets from each sample from the
two centrifugations were combined, ﬂash frozen in liquid nitrogen and freeze-dried at
-45°C. Tin 5 x 9 mm capsules (Costech, CA) were ﬁlled with homogenized freeze-dried
culture, compacted, sealed, and shipped for TOC analysis at the UC Davis Stable Isotope

Facility (Davis, CA).

ICP-MS was used to determine the solution phase concentrations of As(V), Co(III),
Cu(II), Fe(III), Ni(II), Se(VI), and U(VI) (each with ﬁve biological replicates) using a
PerkinElmer Optima 4300DV inductively coupled plasma optical emission spectrometer

(ICP-OES) using the following parameters: air as the shear gas, argon plasma phase at a

ﬂow rate of 17 mL minJ, 0.2 mL min-1 auxiliary gas ﬂow, 0.6 mL min.1 nebulizer gas

ﬂow, and 1.5 mL min.1 sample aspiration rate. Frozen samples were thawed and spiked

with Y as an internal standard, and run in triplicate. Using a CCD detector, metal
concentrations were read at 193.696 nm for As, 228.615 nm for Co, 327.394 nm for Cu,
238.201 nm for Fe, 231.604 nm for Ni, 196.026 nm for Se, 385.962 nm for U, and

371.029 nm for Y.

Calculations. Generation time was calculated using cell count data. The slopes of
colony counts over time were determined for each sample replicate. Two points in the

steepest portion used to calculate doubling time. These doubling times for sample

24

replicates were then averaged to determine the generation time and standard error for
each medium. Conﬁdence intervals of 95% can be determined using: average i 1.96 X

standard error.

The rate of organic acids consumption (e.g. pyruvate and lactate) and production (e. g.
acetate) were calculated using the slopes of the organic acid concentration over time for
each sample replicate. The slope of the line between two points in the steepest portion of
the “log phase” slope was calculated. These two point slopes were then averaged to
calculate the rate of consumption or production of the organic acids for each sample

condition.

Carbon balances were determined based on 20 mM pyruvate or lactate substrate and

measured organic acids and biomass produced. The biomass data was obtained by TOC

analysis, which was converted to mmol C assuming the biomass is C5H7OzN.

Metal valence state determination. The reduction status of the metal in the biomass

samples was examined using x-ray absorption near edge structure (XANES) analysis.

Samples were prepared in an anaerobic glove box of N22H2 (95:5) to prevent metal
oxidation (Chapter 4). The K-edge XANES of all metal measurements were made at
beamline locations MRCAT lO-ID (Segre 2000) and PNC-CAT 20-BM at the Advanced

Photon Source at Argonne National Laboratory (Argonne, IL). At MRCAT, the insertion

device was tapered, and the Si(III) double crystal monochromator was slew scanned.

25

Several EXAF S measurements in transmission mode were averaged for each sample.

The incident ionization chamber was ﬁlled with N2, and the transmission ionization

chamber was ﬁlled with Aer2 (50:50) for metals. A Rh mirror was used to remove

harmonic x-rays. The x-ray beam incident on the sample was 1 mm X 1 mm. The
beamline parameters were similar at PNC-CAT 20-BM to MRCAT, where the Si(III)
monochromator was detuned 15% to enhance harmonic rejection, and scans were
collected in step scanning mode in ﬂuorescence using a thirteen element Ge solid state
detector. The same elemental metal standard was used for each run of the same element.
The XANES spectra were processed using Athena (Ravel 2005) and IFEFFIT software
(Newville 2001). The spectra were aligned to each other and the reference spectra.

Multiple scans from the same sample were merged to increase the data quality.

RESULTS
Fermentative growth with different metals. D. hafniense grew in DCB-l fermenting
pyruvate to acetate in the presence of all eight metals. Greater consumption of pyruvate
occurred in the presence of all metals compared to no metal present, except for 1 mM
Se(VI) (Figure 2.1A). The fastest consumption of pyruvate occurred with As(V). The
acetate production balanced pyruvate consumption for all conditions except with U(VI)
(Figure 2.1B). With 0.5 mM U(VI), less acetate was produced over time. As the U(VI)
was added as uranyl acetate, this amount of acetate was subtracted. The fastest and

highest carbon consumption and production curves were in the presence of 1 mM As(V).

26

A

 

‘A no metal1
:+As(V)
3.0mm
‘.Cowo
i-CMW
imen

lOSeND
ioUND l

 

i
l
l
l
l
l
l

Pyruvate concentration (mM)

 

 

 

16 - 3

Warsaw
12 ——————————————————————————— i.+Asnn l
.cmm ‘
. Co(lll) i
l—cmm i
ime0 1
i o Se(VI) i
l . U(VI) i

Acetate concentration (mM)

 

 

 

Time (h)

Figure 2.1. Organic acid concentrations in cultures grown fermenting pyruvate in the presence of
different metals. Best ﬁt curves and the standard error are shown. All metal concentrations are 1 mM,
except Cu(II) at 0.75 mM and U(VI) at 0.5 mM. Cultures were monitored until stationary phase was

achieved. (A) Pyruvate consumption over time. (B) Acetate production over time.

27

 

l0.01 mM l0.05 mM l0.1 mMTTlTaZS li DOTSTmM ET0F/51mM‘
l1mM l2.5mM I5mM l10mM l25mM

 

 

4o
35 ——————————————————————— a”.-- ”*7.
30 ——————————————— , eeeeeeeee 7 LL, m e-
25 E-, _______________________ “LL- -L _-
2o ......................... ,,L,., e
15 ____________ . ............. LL_ L-
10 -__L-_-LL..--_--L.-L _ _L_ L.

 

Generation time (h)

 

 

As(V) Cd(Il) Cr(Vl) Co(lll) Cu(ll) Ni(ll) Se(VI) U(VI)

Figure 2.2. Generation times of D. hafniense grown with pyruvate under fermentation conditions in the
presence of the indicated metals. The generation time with pyruvate fermentation alone (no metal present)

is represented as 0.01 mM with each metal. Standard error is shown for all measurements as vertical lines.

The time required for D. hafniense to double was 8.9 h when grown with pyruvate alone
(Figure 2.2 -shown as 0.01). This generation time was unaffected in the presence of
Cd(II), Cr(VI), or Co(HI), except with 0.05 mM Co(III) it was faster. At all
concentrations, As(V) provided for a faster generation time, 6 to 7.8 h. Generation time
was unaffected with increasing concentrations of Cu(II) or Ni(II) up to 1 mM, after which
the generation time increased signiﬁcantly to 12 and 36 h respectively. This may be due
to concentration dependent metal toxicity effects on the cells. Concentrations of Se(VI)
at 0.1 mM decreased generation time, but concentrations of 2.5 mM Se(VI) and above
caused a signiﬁcantly increased generation time to 13 h or more. Longer generation

times than pyruvate alone were also found in the presence of all concentrations of U(VI)

28

tested.

Respiration with different metal electron acceptors. Lactate can not be fermented by
D. hafniense, but it can serve as an electron donor for respiration that produces acetate
(Christiansen 1996). As cultures incubated with lactate alone have no change in the
amount of lactate (20 mM) consumed or acetate produced over time, cultures with a non-
metal electron acceptor (nitrate) and lactate were monitored as a respiration control to
which Fe(III), As(V) or Cu(II) with lactate were compared (Fig 2.3A). Growth was
visible in all these conditions. Lactate consumption rates were fastest in the medium with
Fe(III); however, acetate production rates were similar for both Fe(III) and nitrate over
time. Lactate consumption decreased in the presence of As(V) (122.5) and Cu(II) (1 :3.6)
compared to nitrate. Acetate production mirrored lactate consumption for nitrate, Fe(III)
and Cu(II). An increased lag in acetate production was present with As(V) (Figure 2.3B).
These trends were reﬂected in the generation time with each metal (Figure 2.4). F e(III)
and nitrate provided similar generation times at 4.5 and 4.6 h respectively. A marked

increase in generation time occurred with As(V) (7 h) and Cu(II) (27 h).

29

 

- Cu(ll) +As(V) . Fe(III) 1 nitrate’ A

 

 

N
O

.5
U!

 

 

Lactate concentration (mM)
8

 

 

 

 

 

5 _____________
0
0
Time (h)
i -Cu(||) +As(V) - Fe(lll) x nitrate B
20 .- E I
15 —————————————— f E i; ————————————————

E
.E.
c
3
E
8 10 -.
r:
c
0
t
<

 

 

 

  

200 250 300 350 400
Time (h)

Figure 2.3. Organic acid concentrations of cultures incubated with different metal electron acceptors
(respiration conditions). Cultures were monitored until stationary phase was achieved. Best ﬁt curves and

the standard error are shown. (A) Lactate consumption over time. (B) Acetate production over time.

30

 

25-————~—e———- e- H-,, gee- -—
E
o 20-____-_ ___i- 9------“ -- a-
E
'a
515————————~-—- ,, -_,- ,-,,, _~
‘3
a; 1o.L_L, -—_e—_-i giggi-iegii- _a
1:
t3

5-

o-

 

 

 

50 mM Fe(III) 10 mM nitrate 1 mMAs(V) 0.75 mM Cu(II)

Figure 2.4. Generation times of D. hafniense during respiration with lactate and different electron

acceptors. Standard error is shown for all measurements as vertical lines.

Reduction of metals. Change in the oxidation state of metals after growth in DCB-l and
CBF media was determined by XAS (data not shown). While reduction of Cd(II),
Co(III), Cu(II) and Ni(II) was not found with either medium, the ligand bound to the
metal was altered, producing color changes of precipitate [Ni(II), green turned clear;
Co(III), purple to clear; Cd(II), clear to yellow; Cu(II), blue to green to clear]. Although
A reduction was below detection limits with XAS, the change in color observed in cultures
with Co(III) is characteristic of conversion to Co(II) in previous studies (Roh 2002).
Reduction in both media of U(VI to IV), As(V to 111), and Se(VI to 0) was detected by
XAS. Solid Fe(III) oxide and soluble Fe(III) citrate were also reduced [Fe(III to 11)] by
cells in CBF medium with lactate. Reduction of Cr(VI) was not assayed by XAS because

reduction of the metal was observed by color and occurred within minutes of addition to

31

cultures, a character noted in other studies (Roh 2002).

Mineralization of metals by cells. Mineralization (removal of metals from solution)
was determined by ICP-MS in the presence of live and heat-killed cells for both media,
reported as means of ﬁve biological replicates (Table 2.1). When presented with a solid
Fe(III) oxide, live cultures solubilized Fe, likely to more easily make use of it as an
electron acceptor. An increasing mineralization of soluble Fe(III) citrate was found as
further reduction occurred simultaneously with the color changing process of Fe(III to H)
reduction by cells (Table 2.1). Passive metal interactions were not observed for cultures
with Fe. Cells did not appear to inﬂuence mineralization for Cu or C0 regardless of
medium as all changes were within the margin of error (Table 2.1). Cultures with Ni(II),
regardless of medium produced passive mineralization of Ni by cells; similar results were
found for live and heat-killed cultures. Live cultures demonstrated substantial
mineralization with As and Se with little passive effect by heat-killed cultures. Similar to
Se, U was virtually eliminated from solution by active cell cultures; however, passive

interactions by heat-killed cultures may account for nearly half of the loss of soluble U.

32

f"

Table 2.1. Mineralization of different electron acceptors.

 

 

EA concentration EA in % EA in solution
Electron acceptor Medium (mM) solution (ppm) versus nc
As (V) nc DOB-1 1 70.6 100
As (V) hk DCB-1 1 66.2 93.9
As (V) cells DOB-1 1 26.2 37.1
As (V) nc CBF 1 68.6 100
As (V) cells CBF 1 34.7 50.6
Co (III) nc 008-1 1 18 100
C0 (III) cells 008-1 1 17.4 96.7
Cu (ll) chloride nc CBF 0.75 3.25 100
Cu (II) chloride cells CBF 0.75 3.38 104
Cu (ll) sulfate nc CBF 0.75 7.34 100
Cu (ll) sulfate cells CBF 0.75 8.16 111
Fe (III) oxide nc CBF 50 0.874 100
Fe (III) oxide cells CBF 50 32 3660
Fe (III) citrate nc CBF 50 2790 100
Fe (III) citrate cells (brown) CBF 50 968 34.7
Fe (III) citrate cells (black) CBF 50 823 29.5
Fe (III) citrate cells (olive) CBF 50 788 28.2
Fe (III) citrate cells (yellow) CBF 50 760 27.2
Ni (ll) nc DCB-1 1 24.4 100
Ni (II) hk DOB-1 1 18.2 74.6
Ni (II) cells 008-1 1 16.9 69.3
Ni (II) nc CBF 1 7.5 100
Ni (II) cells CBF 1 4.92 65.6
Se (VI) nc 008-1 1 78.1 100
Se (VI) hk DOB-1 1 76.5 97.9
Se (VI) cells DOB-1 1 0.18 0.23
U (VI) (acetate) nc DCB-1 0.5 92 100
U (VI) (acetate) hk DOB-1 0.5 52.80 57.4
U (VI) (acetatg) cells DCB-1 0.5 0.41 0.446

 

nc = negative control (metal present without cells)

hk = heat-killed cells
cells = live cultures

33

Biomass production with different metals. Total organic carbon (TOC) was reported
as a mean of 18 measurements for all culture conditions in order to determine the effect
of metals on biomass yield (Table 2.2). Fermentation of pyruvate alone produced
equivalent amounts of biomass as in the presence of Cd(II), indicating that Cd had no
inﬂuence on growth of cells. Increased biomass was found with both Cu(II) and Co(III);
however, Ni(II) dramatically decreased biomass production. Increasing amounts of
Se(VI) in the culture resulted in increased biomass production by cells. Similar yields
were found with addition of either 1 mM Se(VI) or 1 mM Se(IV). The presence of 1 mM
As(V) nearly doubled biomass production by cells. D. hafniense was previously
determined to be unable to ferment acetate (Christiansen 1996). Here, we report growth
with acetate alone, which yields 72% less biomass than pyruvate fermentation (Table
2.2). This is further exempliﬁed by an increase in biomass yield with pyruvate
fermentation supplemented with acetate compared to pyruvate. To determine if the
acetate in cultures with U(VI) was potentially enhancing growth, uranyl acetate was
provided as the sole carbon source. Signiﬁcant biomass production was found with
uranyl acetate alone (8.36 mg/L), which was 40% and 13.5% the amount of biomass

produced when uranyl acetate was with lactate and pyruvate respectively (Table 2.2).

Unable to ferment lactate, background biomass levels of culture inocula could be
determined in CBF medium with lactate alone (Table 2.2) (Christiansen 1996).
Signiﬁcant biomass production was found in the presence of all metals tested: Fe(III),
As(V), Co(III), Cu(II), and U(VI) (Table 2.2). Dramatic increases in biomass were seen

with 1 mM Cu(II), 1 mM As(V), 50 mM Fe(III), and 0.5 mM and higher U(VI). Only

34

Fe(III) citrate CBF medium without cells produced complexes with ethanol. This
background was quite high at 12.9% of the total biomass produced when Fe(III) citrate
medium was inoculated. The amount of biomass produced by growth with Fe(III) was
1330 mg/ L, which is 38 fold higher than the next highest biomass in CBF medium [10

mM U(VI)]. However, only 17% :t1% of that biomass is from the pelleted sample.

Growth yield of cells with metals. Growth yield was determined with six metals
[U(VI), Fe(III), Ni(II), Co(III), As(V), and Se(VI)] (Table 2.3). Fe(III) was the most

efﬁciently mineralized metal for the amount of carbon produced with an 8-fold higher

growth yield per e-. Growth yields per e- were similar for pyruvate with Se(VI) and

U(VI) in DCB-l, and As(V) in CBF with lactate. These were 20% or higher yields than

with pyruvate fermentation alone, which was 0.97 M of carbon per e-. In contrast,

cultures in DCB-l with pyruvate and As(V) or Co(III) had 20% less grth yield than
pyruvate alone (Table 2.3). The measured difference in cultures with As(V) and pyruvate

or lactate appears to indicate a change in metal mineralization efﬁciency and biomass

dependent not only on the number of e- available, but also on the electron donor source.

35

PI

Table 2.2. Effect of electron acceptor concentration

and carbon sources on biomass yield.

 

 

 

EA concentration Carbon source Biomass produced
Electron Acceptor (mM) Medium (20 mM) mg/ L
- - DCB-1 pyruvate 1 5.1
As (V) 1 003-1 pyruvate 28.3
Cd (ll) 1 DOB-1 pyruvate 15.1
Co (Ill) 1 DOB-1 pyruvate 19.4
Cu(l I) CI 1 DCB-1 pyruvate 17.1
Ni (ll) 1 DOB-1 pyruvate 6.74
Se (V I) 0.5 DCB—1 pyruvate 16.4
Se (VI) 1 003-1 pyruvate 18.8
Se (V I) 5 008-1 pyruvate 22.2
Se (VI) 10 003-1 pyruvate 37.1
Se (IV) 1 008-1 pyruvate 18.5
- - CBF lactate 0.295
As (V) 1 CBF lactate 32.1
Co (Ill) 1 CBF lactate 2.25
Cu(ll) CI 0.5 CBF lactate 3.75
Cu(ll) CI 1 CBF lactate 22.9
Fe (III) (citrate) nc* 50 CBF lactate 198
Fe (III) (citrate) 50 CBF lactate 1 530
- - DCB-1 pyruvate 1 5.1
- - DOB-1 pyruvate + acetate 18.3
- - DCB-1 acetate 4.22
U (VI) (acetate) 0.5 DOB-1 - 8.36
U (VI) (acetate) 0.05 DCB-1 pyruvate 21.2
U (VI) (acetate) 0.5 DOB-1 pyruvate 61.7
- - CBF lactate 0.295
U (VI) (acetate) 0.25 CBF lactate 4.65
U (VI) (acetate) 0.5 CBF lactate 21.2
U (VI) (acetate) 1 CBF lactate 21.8
U (VI) (acetate) 5 CBF lactate 40.6
U (VI) (acetate) 1 0 CBF lactate 35.3
no = no cells

* control condition to account for C complexes formed with Fe(III) citrate and ethanol

36

. .mmanﬁ 03200 on 2 005.5200 :00... 2020288 no; 20$ 5390 0o *0» .0>O..
.20... .8 025680 0000 was $5 62.600 8:050 con: 098.088 0:03.900... «EB. 82008 0.00.0 2.00“. w< Eco 00.2.30 3 000235 30:53 05 m_ £5...

 

 

00. v 0 r .0 50+ v.0 0.0? 90.. n. _.0 $192: 03.0 3.0 0.0 7000 A2880“ ca 3
v0; 00d P. 3 z. N v0? 00? 0.02:3 0N0 9.0 P 7000 22 00
and 00.0 0.0 _..0 N P F 0000 3.0 0.02:3 0.00 0.09 _. 7000 a: _z
:09 N00 «0.. 0.0.. 0.0.0? 05.? .000. 0:200. NNN 00K 00 “.00 0.0.30 2.: 0“.
and 00.9 03.0 N0 0.9 N: v.9 0.0258 N00 v.5 P 7000 2.0 00
vmé R0 03‘ 0.0 08 00.9 use 2800. 0.00 :0 F ".00 E 0<
050 59 ed v.0? 0.0? 000. F 0.0m 0.02:3 :0 N.0N F 7000 00 we.
:20 2253 000390 .x. =2 .5 22.5 ozgeae 4 \ 0 mm 22.... CE .x. 0.38 22.5 82005. EM: 35003.
000395 +1 \0 o 7.00.55 counssmcoo o co_§uo.a cozaEacoo c0028 c. .508 0003005 09:00 6:50 c2560 anEu (m 2980m—
20.» £390 203 £390 Baa. (m 0500.4. 00500 5900 33:50 000520 .5900 225.8 (m 5 <0 .003.

 

0.03 5320 cc 20:00:09.8 35080 c9820 ecu 0:25:30 0.3.30 E2050 .0 «00.5 .0N 030...

37

Table 2.4. Carbon balance of electron donors (lactate and pyruvate) and yield of cells as C (mMoles)

 

 

 

Yield of C (mMoles)
C from
Electron Substrate/ substrate Cell Additional
acceptor electron donor equivalent Lactate Pyruvate Acetate biomass Carbon
no metal pyruvate 60 0 49.5 9.29 1.26 -0.05
Fe(III) lactate 60" 2.22 0 38.2 111 -94.4
Se(VI) pyruvate 60 0 49.2 9.28 1.56 -0.04

 

*Citrate is also a possible weak substrate, though not measured in a our studies.

The calculated carbon balances show that 99.9% of all carbon can be accounted for
during pyruvate fermentation without metals and pyruvate fermentation with Se(VI)
reduction (Table 2.4). Table 2.4 data are derived from means of eighteen biological
replicates for organic acids and 5 biological replicates for cell biomass per condition
The bulk of carbon remains as pyruvate in these conditions. However, grth in CBF
medium with lactate and Fe(III) citrate demonstrated a severe lack of carbon substrate to

account for the carbon yield (Table 2.4).

DISCUSSION
The ability of D. hafniense to remove soluble metal electron acceptors was determined in
cultures fermenting pyruvate or with lactate as the substrate which could require
respiration of the metal for growth. Growth in the presence of all metals was achieved;
however, depressed generation times were found with 1 mM Cu(II) and 1 mM Ni(II),
which is illustrated in the lowered biomass yields and grth rates with Ni(II). This is
surprising as weak respiration of Cu(II) with lactate was observed as well as
characteristic color changes though Cu(II) was not detected to be reduced by XAS.

These longer generation times may be due to concentration dependent metal toxicity

38

effects on cells. In contrast, Co(III) addition to pyruvate fermenting cultures produced
more biomass and a faster generation time than pyruvate fermentation alone. Weak
biomass production was also detected during respiration with Co(III). While reduction of
Co(HI) EDTA has been reported previously (Gorby 1998), reduction of Co(III) chloride
was not detected by XAS in this study. Cd(II) was not reduced by pyruvate-fermenting
cultures nor did the metal appear to signiﬁcantly impact growth; this is similar to

previously observed results with M. thermoacetica (Cunningham 1993).

Metals that were reduced included U(VI), Se(VI), As(V), and Fe(III). Despite reducing
Se(VI to 0) and U(VI to IV), 2.5 mM and above Se(VI), and all concentrations of U(VI)
signiﬁcantly slowed generation times of pyruvate fermenting cultures. Disproportionate
production of extracellular organic carbon or rapid, incomplete cell division may explain

the larger biomass yields accompanying the slowed generation times.

A larger amount of biomass was produced and mineralization efﬁciency was enhanced
consistently in cultures respiring metals with lactate. The largest biomass production by
38 fold was measured during F e(III) citrate respiration. Previous reports have also found
large productions of biomass in response to soluble Fe(III) reduction, such as with the
thermophilic, Fe-oxidizing bacteria (Bridge 1998). However, the provided lactate
substrate is not enough carbon to account for the large carbon yield. This carbon can
easily be accounted for if only a third of the citrate ﬁ'om Fe(III) citrate is used as a carbon
source. However, as no previous reports found citrate as a carbon source for D. hafniense

despite its use in medium (Christiansen 1996), citrate is not likely a strong carbon source.

39

We also found evidence of grth with acetate, a substrate also not reported to suggest
growth (Table 2.2). As the amount of acetate is so high in the Fe(III) citrate condition, it

may serve as a carbon source. Recent work demonstrating acetogenic ability by

D. hafniense has been observed in cultures containing 30% C02 (Chapter 3). In this

study, 20% C02 was present in respiration conditions, so C02 ﬁxation is a potential

source for carbon substrate. But the biomass contribution of C02 ﬁxation is not

sufﬁcient to cover the negative net carbon yield, as we found the biomass produced by
acetogenically growing D. hafniense to be over 300 fold less than what we found here
with Fe(III) citrate and lactate (Chapter 3). Further studies would be necessary to
determine that it is a combination of carbon resources (lactate, citrate, carbon dioxide,
and possibly an increasing acetate pool) that greatly enable biomass production by D.
hafniense with Fe(III). While the grth yield from citrate provided as Fe(III) citrate
should also be measured to determine its role, a combination of carbon resources likely

account for the carbon balance discrepancy found.

Large production of soluble extracellular polymeric substances was also found with
Fe(III) citrate cultures. High amounts of soluble extracellular organic carbon production
have also been reported to occur during halorespiration of trichloroethene (Ayala-del-
Rio, 2002). Contributing over 70% of the biomass in D. hafniense with Fe(III) citrate,
this soluble fraction may facilitate Fe(III) reduction; however, ﬁirther tests will need to be

done to determine the role of such signiﬁcant soluble biomass production.

40

 

Enhanced growth yield and a faster generation time were observed in the presence of
As(V). This study conﬁrms that previously discovered As(V) reduction genes in D.
hafniense are functional with the ability to reduce As(V to H1) and use it as a sole
terminal electron acceptor (Niggemyer 2001). Genes for the full reduction of As(V to 0)
in D. hafniense are present and the reaction has a higher energy yield; however, here we
found a preference for only partial reduction. This is important for bioremediation as

As(III) is more mobile than As(V), unless bound to sulﬁde (Laverman 1995).

Clear evidence exists of the ability of D. hafniense to grow using metal electron
acceptors. Lactate alone can not be fermented, but in the presence of metals it was
readily respired to acetate, showing that the metals must be serving as the terminal
electron acceptors. Even during fermentation with pyruvate faster grth rates and
enhanced biomass production were seen with most metals. The carbon balances during

pyruvate fermentation and pyruvate fermentation during Se(VI) reduction account for

virtually all carbon. Assuming the biomass as C5‘H702N, the equation for the carbon

balance of pyruvate fermentation without metal can be written as (things not measure in

brackets):

20 pyruvate + [0.25 NH3] + [0.05 C02] + [0.61 H20]

-) 16.5 pyruvate + 4.65 acetate + 0.25 biomass + [0.13 ﬁg]

This is the ﬁrst large study on a Gram-positive species exposed to a broad array of metal

41

"I

electron acceptors. D. hafniense is capable of obtaining energy through fermentation and
respiration, including both halorespiration and metalorespiration (Madsen 1992 and
Christiansen 1996). Electron acceptors for respiration include: sulﬁte (not sulfate),
thiosulﬁte, nitrate, nitrite, arsenate, selenate, selenite, F e(III), U(VI), and halogenated
compounds, such as chlorophenols and tetrachloroethene (Villemur 2006). This
organism is also able to tolerate high concentrations of potentially toxic metals before
exhibiting declining growth. Determining an array of usable metal electron acceptors for
an organism is of importance to ﬁtrthering knowledge of the organism’s role not only in
the environment in which it was found but for applications, such as in bioremediation
through mineralization and reduction of compounds toxic to plants and animals. The
ability of D. hafniense to grow in the presence of Cr(VI) and reduce U(VI) with large
biomass production present the possibility of its niche at Oak Ridge National Laboratory
(Oak Ridge, TN) as a desired metal-reducer. This study also not only provides a new
organism to study the various pathways of metal reduction, but this organism is Gram-
positive. Many of these pathways have been extensively studied in Gram-negative
organisms, and many proteins have been found to localize to the cell membrane (Lovley
1993 and Schrdder 1997). Due to the difference in outer membrane, it is possible that
having a Gram-positive model organism for metal reduction may lead to the discovery of

different redox proteins and pathways involved in metal reduction.

ACKNOWLEDGEMENTS
Funding for this study was provided by the U.S. Department of Energy, Ofﬁce of

Science, Environmental Remediation Science and Genomes to Life Programs. Research

42

at the Advanced Photon Source at Argonne National Laboratory (Argonne, IL) was
supported by the U.S. Department of Energy, Ofﬁce of Science, Ofﬁce of Basic Energy
Sciences, under Contract No DE-AC02-06CH11357. The assistance of Dr. Shelly Kelly
was greatly appreciated during beamline studies and performing XAS analyses at
Argonne National Laboratory. The equipment operation and collection of ICP-MS data
was performed by Dr. Ed O’Loughlin of the Molecular Environmental Science Group in

the Biosciences Division at Argonne National Laboratory for which we are grateful.

43

REFERENCES

Ayala-del-Rio, H. L. 2002. Long-terrn effects of phenol and phenol plus trichloroethene
application on microbial communities in aerobic sequencing batch reactors. Ph.D.
dissertation. Michigan State University, East Lansing.

Bloem, J. 1995. Fluorescent staining of microbes for total Counts. In: Akkermans A.D.L.,
Elsas J .D., Bruijn F.J. (eds), Molecular Microbial Ecology Manual. Kluwer Academic
Press, Dordrecht, The Netherlands, pp. 1-12.

Bridge, T. and D. B. Johnson. 1998. Reduction of soluble iron and reductive dissolution
of ferric iron-containing minerals by moderately thermophilic iron-oxidizing bacteria.
Appl. Environ. Microbio l. 64:2181-21 86.

Christiansen, N. and F. K. Ahring. 1996. Desulﬁtobacterium hafniense sp. nov.,
anaerobic, reductively dechlorinating bacterium. Int. J. Syst. Bacteriol. 46:442-448.

Cunningham, D. P. and L. L. Lundie Jr. 1993. Precipitation of cadmium by
Clostridium thermoaceticum. Appl. Environ. Microbiol. 59:7-14.

Gorby, Y., F. Caccavo, Jr., and H. Bolton, Jr. 1998. Microbial Reduction of

cobaltmEDTA- in the presence and absence of manganese(VI) oxide. Environ. Sci.
Technol. 32:244-250.

J arup, L. 2003. Hazards of heavy metal contamination. Brit. Med. Bull. 68: 167-182.

Kostka, J. E., D. D. Dalton, H. Skelton, S. Dollhopf, and J. W. Stucki. 2002. Growth
of iron(III)-reducing bacteria on clay minerals as the sole electron acceptor and

comparison of growth yields on a variety of oxidized iron forms. Appl. Environ.
Microbiol. 68:6256-6262.

Lanthier, M., R. Villemur, F. Lepine, J-G. Bisaillon, and R. Beaudet. 2001.
Geographic distribution of Desulﬁtobacterium frappieri PCP-1 and Desulﬁtobacterium
spp. in soils from the province of Quebec, Canada. FEMS Micro. Ecol. 36:185-191.

Laverman, A. M., J. S. Blum, J. K. Schaefer, E. J. P. Phillips, D. R. Lovley, and R. S.
Oremland. 1995. Growth of strain SES-3 with arsenate and other diverse electron
acceptors. Appl. Environ. Microbiol. 61:3556-3561.

Liu, C. Y., A. Gorby, J. M. Zachara, J. K. Fredrickson, and C. F. Brown. 2002.
Reduction kinetics of Fe(III), Co(III), U(VI), Cr(VI), and Tc(VII) in cultures of
dissimilatory metal-reducing bacteria. Biotech. Bioeng. 80:637-649.

Léfﬂer, F. E., R. A. Sanford, and J. M. Tiedje. 1996. Initial characterization of a
reductive dehalogenase from Desulﬁtobacterium chlororespirans C023. Appl. Environ.
Microbiol. 62:4982-4985.

44

Lovley, D. R. 2002. Dissimilatory metal reduction: from early life to bioremediation.
Amer. Soc. Microbiol. News. 68:231-237.

Lovley, D. R. 1993. Dissimilatory metal reduction. Annu. Rev. Microbiol. 47:263-290.

Lovley, D. R. and E. J. P. Phillips. 1988. Novel mode of microbial energy metabolism:

organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl.
Environ. Microbiol. 54:1472-1480.

Madsen, T. and D. Licht. 1992. Isolation and characterization of an anaerobic
chlorophenol—transforming bacterium. Appl. Environ. Microbiol. 58:2874-2878.

Methé, B. and C. M. Fraser. 2004. Roll with the ﬂow: microbial masters of redox
chemistry. Trends in Microbiol. 12:439-441.

Miller, T. L. and W. J. Wolin. 1974. A serum bottle modiﬁcation of the Hungate
technique for cultivating obligate anaerobes. Appl. Microbiol. 27:985-987.

Niggemyer, A., S. Spring, E. Stackebrandt, and R. F. Rosenzweig. 2001. Isolation and
characterization of a novel As(V)-reducing bacterium: implications for arsenic
mobilization and the genus Desulﬁtobacterium. Appl. Environ. Microbiol. 67:5568-5580.

Ravel, B. and M. Newville. 2005. ATHENA, ARTEMIS, HEPHAESTUS: data analysis
for X-ray absorption spectroscopy using IFEFFIT. J. Synch. Rad. 12:537-541.

Roh, Y., S. V. Liu, G. Li, H. Huang, T. J. Phelps, and J. Zhou. 2002. Isolation and
characterization of metal-reducing Mennoanaerobacter strains ﬁom deep subsurface
environments of the Piceance Basin, Colorado. Appl Environ. Microbiol. 68:6013-6020.

Schrtlder, 1., S. Rech, T. Kralft, and J. M. Macy. 1997. Puriﬁcation and
characterization of the selenate reductase ﬁom Thauera selenatis. J. Biol. Chem.
272:23765-23768.

Segre, C. U., N. E. Leyarovska, L. D, W. M. Lavender, P. W. Plag, A. S. King, A. J.
Kropf, B. A. Bunker, K. M. Kemner, P. Dutta, R. S. Druan, and J. Kaduk. 2000. The
MRCAT insertion device beamline at the advanced photon source. Synch. Rad. Inst.
CP521:419-422.

Stolz, J. F., P. Basu, and R. S. Oremland. 2002. Microbial transformation of elements:
the case of arsenic and selenium. Int. Microbiol. 5:201-207.

Villemur, R., M. Lanthier, R. Beaudet, and F. Lépine. 2006. The Desulﬁtobacterium
genus. FEMS Microbiol. Rev. 30:706-733.

Wolin, E. A., R. S. Wolfe, and M. J. Wolin. 1963. Viologen dye inhibition of methane
formation by Methanobacz'llus omelianskii. J. Bacteriol. 87:993-998.

45

Chapter III

Gene Expression of Desulﬁtobacterium hafniense DCB-2 Exposed to Nitrate, Fe(III),

Se(VI), and U(VI) and Proof of Its Acetogenic Growth

ABSTRACT
The ability of the Gram-positive, dehalogenator Desulﬁtobacterium hafniense DCB-2 to
reduce a broad spectrum of metals has only recently been discovered. Despite extensive
research, pathways responsible for metal reduction remain poorly understood. Gram-
negatives have been the primary focus of these gene studies, with results indicating much
of the metal-reduction is membrane associated. The genome sequence of D. hafniense
has recently been completed lending itself to gene expression studies with microarrays.
To better understand metal-reduction by D. hafniense, analysis of the genome,

transcriptome and physiology were combined. Metal-reducing conditions of Fe(III),

Se(VI) and U(VI) were compared to pyruvate fermentation and N03- respiration. Results

show that D. hafniense can greatly shift its metabolism in response to these metals.
Interestingly, nitrate reducing-conditions did not induce the annotated nitrate reduction
pathway in D. hafniense above levels during pyruvate fermentation. Instead, despite the
large amount of available nitrate, the annotated nitrogen ﬁxation pathway was induced.
Three general patterns of expression appear when similarities in gene expression and the

number of electrons available for reduction were compared. These similarities were:

46

Fe(III to II) and U(VI to IV); N03- (to NH4) and Se(VI to 0); and pyruvate fermentation

with a unique pattern of expression. Morphological observations reﬂect gene expression,
such as the short rods in chains surrounded by copious amounts of extracellular biomass
exhibited during Fe(III) reduction. The data also supports the presence of an active
acetyl-CoA reduction pathway (Wood-Ljungdahl pathway) in D. hafniense DCB-2. The
three patterns of expression differ in their induction levels of the acetyl-CoA reduction
pathway with Fe and U inducing the pathway most. This is likely due to the necessity of
cells to rid themselves of excess electrons, and the decreased electron accepting capacity

with Fe and U compared to the other groups. Studies in vitro with D. hafniense in

mixtures of H2, C02, and CO conﬁrm it can grow as an acetogen.

INTRODUCTION
The use of bacteria to immobilize or degrade toxic compounds, such as metals,
radionuclides and chlorinated solvents, through bioremediation has been proposed and
put into practice at several U.S. Department of Energy sites (Gorby 1994, Rilely 1992,
Rooney-Varga 1999). At these sites, genomic studies through phylogenetic and
functional genome arrays in situ have compiled an immense amount of data, but it is
difﬁcult to decipher which organisms (other than the target study organism) are
performing the desired redox tasks for remediation within the given environment. Only a
few representative bacteria have been studied for their ability to reduce desired

contaminants, making optimization of a desired bioremediation difﬁcult.

47

 

Controlled in vitro studies to determine reduction capabilities of different bacteria and
their gene expression is one method to solve this problem. Gene expression studies of a
greater diversity of bacterial species allows for better comparisons to in situ conditions
and provides more opportunity for detailed monitoring and optimization of the desired
bioremediation. Genomic studies of whole-genome microarray expression proﬁles on a
single bacterial species grown with different electron acceptors, such as metals, have
focused largely on Gram-negative organisms. Much of their redox machinery has been
found to be membrane associated, thus a study of the gene expression proﬁle of a Grarn-

positive bacterium is warranted (Lovley 1993a).

Desulﬁtobacterium hafniense DCB-2 is a recently sequenced anaerobic Gram-positive
dehalogenator also capable of metal reduction (Christiansen 1996, Madsen 1992,
Chapter 2). Using molecular analyses, D. hafniense has been found in both pristine
environments and environments highly contaminated with uranium, chromium, and
nitrate (Kostka 2002, Lanthier 2001). Current studies have found that besides iron and
manganese, this bacterial species is capable of reducing arsenic, selenium, and uranium

(Madsen 1992, Niggemyer 2001, Chapter 2). Here we performed expression studies with

D. hafniense in the presence of different potential electron acceptors, including NO3-,
Fe(III), Se(VI), and U(VI). Toxicity and redox pathways of the anaerobic organism were
examined for each electron acceptor. Following expression analysis, gene studies

conﬁrmed the presence of an active acetyl-CoA reduction pathway (Wood-Ljungdahl

pathway).

48

 

METHODS
Bacterial genome sequence. D. hafniense strain DCB-2 was originally isolated from
municipal sludge in Copenhagen, Denmark, using a chlorophenol enrichment (Madsen
1992). Its genome has been sequenced by the U.S. Department of Energy’s Joint
Genome Institute and annotated by the automated pipeline operated by Oak Ridge
National Laboratory’s Computational Genomics Group, but at the time of microarray
development for this paper the genome was at 10x coverage (August 11, 2005). The

latest sequence and annotation analysis for D. hafniense DCB-2 may be found at

http://genome.j'gi-psforgZdraft microbes/desha/desha.home.html.

 

Culture conditions. For all studies, D. hafniense was inoculated into medium following
autoclave sterilization and addition of IX Wolin vitamins (Wolin 1963), the test
compound, and a carbon source. Iron reduction studies used modiﬁed ferric citrate
medium containing 50 mM ferric citrate (Lovely 1988, Chapter 2) with 20 mM sodium
lactate (lactate) added. Modiﬁed DCB-l medium (L6fﬂer 1996, Chapter 4) was used for
selenium and uranium studies with 1 mM sodium selenate, and 500 uM uranyl acetate,
added with vitamins and a sole carbon source of 20 mM sodium pyruvate (pyruvate).
Carbonate-buffered freshwater medium (CBF) (Lovely 1988, Chapter 2) was made with
10 mM nitrate, Wolin vitamins, and 20 mM lactate as the sole carbon source. As a
positive control, D. hafniense was also grown in modiﬁed DCB-l medium with only 20
mM pyruvate and Wolin vitamins to which all cultures were compared during

hybridization and microarray analyses.

49

Cultures were incubated at 37°C to mid-logarithmic phase as monitored by HPLC and
optical density, correlating data to a previously derived standard curve of grth that
monitored total colony forming units, HPLC, color change, and optical density (Chapters
2 and 4). While both black and red precipitates were produced by D. hafniense in the
presence of sodium selenate (Chapter 4), only red precipitate-producing cultures were
examined in this study. Scanning electron microscopy (SEM) and transmission electron
microscopy (TEM) were used to observe differences in cell morphology among culture

conditions with a minimum of 20,000 cells viewed per condition (Chapter 4).

RNA extraction, cDNA synthesis and labeling. In triplicate, cell cultures were
harvested at mid-log phase by centriﬁrgation under anaerobic conditions with the pellet
suspended in RNA Protect (QIAGEN, Valencia, CA) at room temperature with a 45 min
lysozyme digestion. RNA was extracted from the cell pellet using an RNeasy RNA
extraction kit (QIAGEN) according to the manufacturer's instructions with the DNase
digestion step included. RNA quality was tested by running a 1x FA gel containing

0.74% formaldehyde and 1.2% agarose, and RNA concentration was determined using a

UV-spectrophotometer at OD260.

Direct amino-allyl labeling was performed as according to a protocol of The Institute for
Genomic Research (TIGR) http://www.tigr.org/tdb/micromy/protocol_sTIGR.ﬂitml.
Total RNA (5 pg) was denatured by incubation at 70°C for 10 min in the presence of 0.4
mM Random Hexamer Primers (Invitrogen, Carlsbad, CA) and snap-cooled 10 min on

ice. cDNA was synthesized at 42°C overnight using the RNA-primers mix, 3 ul 0.1 M

50

dithiothreitol, 1.2 ul 25x deoxynucleoside triphosphates (3:2 ratio of amino-ally] dUTP to
dTTP), 6 ul 58 First Strand synthesis buffer, and 2 pl Superscript II reverse transcriptase
(Invitrogen). Amino-ally] dUTP was purchased from Ambion (Austin, TX). RNA was
hydrolyzed through addition of 10 ul 1 M NaOH, 10 ul 0.5 M EDTA and incubation at
65°C for 15 min. The reaction was neutralized with 10 ul of 1 M HCl. Puriﬁcation of
cDNA was performed following amino-ally] label incorporation and chemical dye
coupling using the Qiaquick PCR puriﬁcation system (QIAGEN) according to the TIGR
protocol. Dyes were incorporated at 25°C for 1.5 h with 4.5 ul amino-allyl labeled
cDNA in 0.1 M sodium carbonate (pH 9) and 4.5 ul reactive Cy5 and Cy3 N-
hydroxysuccinamide ester ﬂuorophores (Amersham, Piscataway, NJ) in dimethyl
sulfoxide (GE Biosciences, Piscataway, NJ). Amino-ally] labeled cDNA quantity and
ﬂuorophore incorporation efﬁciency were determined using UV-visible

spectrophotometry.

Microarray characteristics. Glass slide arrays were made by the Institute for
Environmental Genomics (IEG) at Oklahoma State University (Dr. Jizhong Zhou,
Norman, OK) from the D. hafniense DCB-2 draft sequence with 10" coverage (August
11, 2005) at which time D. hafniense was believed to have a 5.2 Mb chromosome. The
probe design was performed by IEG (He 2005, Li 2005) with on average one probe per
predicted open reading frame. A total of 4,667 probes were present in duplicate for a
total of 9,984 spots present on the array including positive and negative controls.

Negative controls included 10 human and 10 Arabidopsis probes. Positive controls

consisted of probes to detect six different quantities (5~300 ng ill-1) of genomic DNA

51

spotted four times in duplicate per slide.

Microarray hybridization and analysis. Microarray slides were incubated at 49°C for
1 h in prehybridization solution containing 5x SSC buffer, 0.1% sodium dodecyl sulfate
(SDS), and 0.1 mg/ml bovine serum albumin. A series of room temperature washing
steps followed, including: two washes in 0.1x SSC buffer for 5 min and two washes in
sterile water for 30 sec. Slides were then dried by centrifugation at 1800 rpm for 2 min
using a Sorvall RT 6000D centrifuge with an H1000B hanging basket rotor. LifterSlips®
(22 by 60 mm; Erie Scientiﬁc, Portsmouth, NH) were prepared by washing once in water
and once in 95% ethanol and drying with ﬁltered forced air. LifterSlips® were placed

over the microarray on the slide.

Equivalent concentrations of Cy dye labeled cDNA from the metal sample and the
alternate Cy dye labeled positive control (pyruvate fermentation without additional
electron acceptor) were then mixed, dried in a SpeedVac, and resuspended in 50 ul of
hybridization buffer (5x SSC buffer, 0.1% SDS, 25% formamide, 0.1 mg/ml salmon
testis DNA [Sigma-Aldrich]). This mixture was heated to 95°C for 10 min, spin-cooled,
reheated to 70°C, and applied to the array using capillary action. Microarray slides were
then sealed in a hybridization chamber (CMT Hybridization Chamber; Corning

Incorporated, Corning, NY) and incubated for 18 h in a 49°C shaking water bath (30

rpm).

Microarray hybridizations were performed with a dye swap and 3 biological replicates for

52

a total of 6 slides per comparison. LifterSlips® were removed from hybridized
microarray slides carefully in a wash of 1" SSC buffer and 0.1% SDS at 49°C. Slides
were then washed in a shaking water bath (50 rpm) for 5 min in 1x SSC buffer and 0.1%
SDS at 49°C, twice for 5 min in 0.18 SSC buffer and 0.1% SDS at room temperature,
four times for 1 min in 0.1x SSC buffer at room temperature, and once for 10 s in 0.01)‘

SSC buffer. Washed slides were spun dry at 1800 rpm for 2 min.

Slides were immediately scanned on an Axon 40003 laser scanner (Axon Laboratories,
Molecular Devices, Sunnyvale, CA) using histogram veriﬁcation in GenPix Pro 5.1
(Axon Laboratories) for raw data production followed by gene identiﬁer alignment.
Correction for background was made by background subtraction of the median value.
Median signal intensities for each channel (Cy5 and Cy3) were imported into GeneSpring
6.0 (Silicon Genetics, Redwood City, CA) and normalized by dye swapping and (per chip
and per gene) using Lowess intensity-dependent normalization. Only spots with more
than 80% of pixels above background with P-values less than 0.01, and 2 standard
deviations in either Cy5 or Cy3 channel were considered differentially expressed genes
and used for further analysis. Our analyses ﬁrrther focused on operons containing genes

of similar expression to determine important metabolic pathways.

Experimental determination of acetogenic ability. When testing D. hafniense DCB-2
for the ability to grow as an acetogen, Moorella thermoacetica served as the positive

acetogen control. D. hafniense and M. thermoacetz'ca were originally grown in modiﬁed

DCB-l medium fermenting pyruvate with a head space of C02:N2 (5:95) (Brumm 1988,

53

Loffler 1996, Chapter 4). Cultures were incubated at 37°C for D. hafniense and 55°C for

M. thermoacetica. These fermenting cultures served as the initial inocula for testing for

C02 ﬁxation. The medium was 20 m1 of modiﬁed DCB-l medium with ﬁve different
gas mixtures [C02:N2 (40:60), CO:N2 (40:60), CO:C02:N2 (30:30:40), H2:C02N2

(30:30:40), and H2:C02;:N2 (30:30:40)] in 120 ml serum bottles (Lofﬂer 1996, Miller

1974, Chapter 4). Wolin vitamins (1 x) and cysteine were the only other potential sources
for carbon in this medium (Wolin 1963). Cultures were transferred following a 14 d
incubation. The cultures were shaken daily at the time visible grth was noted. Four
serial transfers into new media were performed prior to measuring growth by optical
density at 600 nm and total biomass production (TOC) (Chapter 4). Samples for TOC
were collected as a pellet by centrifugation. The supernatant was then combined with
ethanol (2:3) to precipitate soluble organic carbon, centrifuged, and the new pellet was
combined with the ﬁrst and sent for TOC analysis at UC Davis Stable Isotope Facility
(Davis, CA). Eight biological replicates were performed for all samples and submitted
for TOC analysis. Controls consisted of all media conditions without cells, cultures

without electron acceptor addition, and cultures with heat-killed cells.

Culture purity. Culture purity was tested throughout this study in duplicate using
ﬂuorescent in situ hybridization (FISH) on all samples with DAPI, a Cy3-labeled
universal eubacterial rDNA probe, and a Cy5-labeled D. hafniense-speciﬁc rDNA probe
(Bloem 1995, Lanthier 2001). Escherichia coli was used as a positive control for DAPI

and the Cy3-labeled probe and as a negative control for the Cy5-labeled probe. Slides

54

were observed on a Nikon epifluorescent microscope equipped with a 50 W mercury

lamp.

RESULTS and DISCUSSION

In this study, total genome transcriptomic proﬁles were obtained from D. hafniense DCB-

2 cells growing with four different electron acceptors: Fe(III), N03: U(VI), and Se(VI).

Microarray studies on these culture conditions have not only provided valuable
information about key genes/metabolic pathways that respond to each redox condition,
but also offered an opportunity to View the previously inaccessible global metabolism of
pyruvate fermentation by D. hafniense DCB-2. This was possible because transcripts
obtained from each culture were compared against pyruvate-fermenting cultures. The
resulting down-regulated genes from each microarray were then used to deduce and

unravel some of underlying metabolic events in pyruvate-fermenting cells.

Pyruvate-fermenting cell metabolism. From the analysis of the down-regulated gene

lists [623 genes from Fe(III), 643 genes from N03: 857 genes from U(VI), 371 genes

from Se(VI)] and the D. hafniense DCB-2 genome sequence archived in JGI, some key
metabolic features in pyruvate-fermenting D. hafniense DCB-2 were deduced (Figure

3.1).

Carbon cycling. Carbon ﬂow in the cell was determined relative to biogenesis of cell

material and up-regulation of genes under dissimilatory Fe(III) reduction. During

55

pyruvate-fermenting conditions, the tricarboxilic acid cycle (TCA) seemed discontinuous.
Like many strict anaerobes, it lacked genes of the 2-oxoglutarate dehydrogenase
complex. Pyruvate-fermenting cells may instead use bidirectional citrate to 2-
oxoglutarate and oxaloacetate or succinyl-CoA routes. Moreover, the glyoxylate
pathway to replenish oxaloacetate is absent. Instead, citrate lyase, which converts citrate

to oxaloacetate and acetate, appeared active in the cell. Even in pyruvate-fermenting

conditions cells were ﬁxing C02 via the reductive acetyl-CoA pathway (Wood-

Ljungdahl pathway) since the downstream NAD-dependent forrnate dehydrogenase

operon (Dhaf_4268-4271) was induced (Figure 3.1).

Electron ﬂow. Cells appeared to use anaerobic respiration to gain additional energy. It is
evident that some fumarate reductase and DMSO reductase genes are constitutively

expressed. Also, a cytochrome-dependent formate dehydrogenase operon (Dhaf_4616-

4621) seems to provide electrons for energy. There are three putative Nz-ﬁxing

nitrogenase operons in the genome that contain a complete niﬂ-IDK gene unit (Figure

3.6). Apparently, cells are actively using these operons to ﬁx N2 (Dhaf_1350-1360,

Dhaf_1047-1059, and Dhaf_1810-1818). However, it is not clear whether both nitrate

reductase genes and nitrite reductase genes are expressed in pyruvate-fermenting cells.

Ammonia formation through Nz-ﬁxation rather than nitrate ammoniﬁcation seems to be

preferred by D. hafniense under all tested conditions (Figure 3.6). Nitrate could be
reduced to nitrite by periplasmic Nap nitrate reductase (Dhaf_1289), but genes encoding

membrane-bound Nar nitrate reductases are absent in the D. hafniense genome.

56

 

Although not differentially expressed during pyruvate-fermenting conditions, genes for
direct ammoniﬁcation of nitrite are present consisting of two nitrite reductase (Nri)
operons (Dhaf_3630-3631, Dhaf_4234-4235). Also, putative genes for nitric oxide

reductase (Dhaf_2253-2254) and nitrous oxide reductase (Dhaf_1205) of the

denitriﬁcation pathway were found, which may potentially convert nitric oxide to N2.

57

Pyruvate

\WHHHHWWITIM

 

  
 

Gluconeogenesis
ADP

9
H" g :
ATP synthm ATP fru-1 ,6-bisphosphate

 

Nitrogenase, Nif1, Niﬂ, Nif3

, _._....,
l

.-
formate t / ‘1
dehydrogenase _‘_ W d
_ _°°.°{
I oxaloacetate I - 3 t
\ i—

] hydrogenase
\ "2

DMSO
fumarate H succlnate g
,1_ DMSOroductase pus

+
2.‘ 2H+ NADV NADHZ

 
    

 

 

 

 

 

\iIIHHIIIHIHHH

 

 
  

 

fumarate
reductase
1

\\

 

Figure 3.1. Schematic diagram of the key metabolism pathways during fermentation with pyruvate.

Menaquinone is abbreviated as “MK”, cytochrome as “cyt”, and phosphoenolpyruvate as “PEP".

58

Dissimilatory Fe(III)-reducing cell metabolism. Microarray experiments with F e(III)-
respiring cells resulted in 655 up-regulated genes and 523 down-regulated genes (>2 and
<O.5 fold cutoff) against pyruvate-fermenting cell transcriptomes. Fe-reducing
conditions proved to have the largest differential expression in cells of the major

pathways compared to pyruvate fermentation (Figure 3.2).

Carbon cycling. Among the highly induced gene groups (at least >5 fold) were CO
dehydrogenase catalytic genes (induced 50-60 fold). These gene products are part of the

reductive acetyl-CoA pathway (Wood-Ljungdahl pathway) which is the major

mechanism of C02 ﬁxation under anaerobic conditions. Other genes that are involved in

the same pathway were also highly induced (2045 fold). When we consider that

pyruvate-fermenting cells are already actively utilizing this pathway, an additional 50-60

fold induction appears to be extraordinarily high. Although C02 ﬁxation is an energy

demanding process, it could provide a redox balance by consuming excess electrons by

using them to assimilate more reduced cellular material. In fact, under nitrate-reducing

conditions, which also used lactate as the substrate, the C02 ﬁxation pathway was
slightly repressed. This is likely because nitrate can take up eight electrons to form NH3

while Fe(III) reduction to Fe(II) can only accept one electron; giving NO3--reducing cells

less of a burden to deal with excess intracellular electrons.

Many of the other up-regulated genes are associated with central metabolism. These

59

were the genes coding for pyruvate carboxyltransferase, acetaldehyde dehydrogenase,
fumarate reductases, malic enzyme, pyruvate formate-lyase, and ﬁ'uctose-1,6-
bisphosphatase. Both pyruvate carboxyltransferase and ﬁ'uctose-1,6-bisphosphatase are
involved in gluconeogenesis, mediating the carboxylation of pyruvate to oxaloacetate and
the dephosphorylation of ﬁuctose-1,6-bisphosphate to ﬁ'uctose-6-phosphate. Five
fumarate reductase genes located at different loci were moderately induced (2-6 fold) and

a gene putatively encoding NADHzﬂavin oxidoreductase/fumarate reductase was highly

induced by ~20 fold. D. hafniense may use the reverse TCA cycle to ﬁx C02 and

generate energy via fumarate reduction to succinate. Since the genome lacks 2-
oxoglutarate dehydrogenase complex genes, which convert 2-oxoglutarate to succinyl-
CoA in the TCA cycle, succinate formed by ﬁimarate reduction can be either converted
to succinyl-CoA (succinyl-CoA synthetase gene, 9-fold induction) or excreted through

the TRAP dicarboxylate transporter (lO-fold induction) for ﬁiture recycling.

Electron ﬂow. Under F e(III)-respiring conditions, D. hafniense exhibited increased
transcriptional activities on many oxidoreductase, transporter, ﬂavoprotein, and
coenzyme biosynthesis genes. Formate dehydrogenase (Fdh) appears to be a key enzyme
in both carbon and energy ﬂows. Two classes of this enzyme system exist in D.
hafniense (cytochrome-dependent and NAD-dependent). Cytochrome-dependent formate

dehydrogenase (Fdhl) (Dhaf_4616-4621 induced ~7 fold) acts on formate to produce

+ . .
C02, 2H , and two electrons. These electrons are used in the electron transport cham to

reduce a terminal electron acceptor, and the released protons are used to generate energy.

60

Interestingly, under NO3'-, U(VI)-, and Se(VI)-reducing conditions, Fdhl transcription

activity was down-regulated. NAB-dependent formate dehydrogenase (Fdh2a) directs

carbon ﬁxation by catalyzing C02 to produce formate, which is subsequently used in the

Wood-Ljungdahl pathway to generate acetyl-CoA. Expression of Fdh2a-encoding
operon (Dhaf_4268-4271) under Fe(III)-reducing conditions was, to our surprise,
signiﬁcantly decreased (~8 fold down-regulation). BLAST search for another potential
NAD-dependent formate dehydrogenase operon has identiﬁed a second operon (Fdh2b,
Dhaf_1508-1513), which was highly induced (~50 fold) under Fe(III)-reduction.
Induction and repression of this operon seems to be highly speciﬁc since under no other

conditions examined were signiﬁcant changes in its expression detected.

Consistent with our predictions, genes encoding a ferric uptake regulator (Dhaf_1359, ~3
fold) and a ferrous iron transporter (Dhaf_2887, ~8 fold) were up-regulated, as were
lactate dehydrogenase gene and lactate transporter genes. Many such examples have
strengthened the integrity of our microarray results. At this stage, it is too early to
identify the terminal reductase genes involved in dissimilatory Fe(III) reduction.
Potential candidates may be obtained ﬁom more extensive analysis of microarray data

and targeted experimentation.

61

Fe(III) + Lactate
N03'w N02"

\\HWHHWHHHIHH
H+ ADP Nitrate reductase

g A Nitrite reductase
+ I

fru-1,6-bisphosphate

_- -._....,

‘f Nitrogenase. Nif2. Nif3

' +
formate
dehydrogenase ¢
acetyl-CoA ‘—

l -/ hydrogenaee
k H2

29'
2-oxoglutarate
DMSO
—> 5.

ft ,
\

2e° 2H+
transporter

/ fumarate MAD"
rim-@2694 IHH
2H+

Figure 3.2. Schematic diagram of the key metabolism pathways during dissimilatory Fe(III) reduction

 

    

H
ATP synthase ATP

 

 

 

 

 

 

 

\\ HIHHIHHHHH

 

    

NADHZ

     
  

 

with lactate. Menaquinone is abbreviated as “MK”, cytochrome as “cyt”, and phosphoenolpyruvate as

“PEP”. Green arrows indicate genes similarly expressed as pyruvate fermentation. Red arrows are genes

62

up-regulated (in comparison to pyruvate fermentation) in response to dissimilatory Fe( III) reduction.

U(VI)-reducing cell metabolism. First examination of the U(VI) microarray data
immediately revealed that many up-regulated genes were shared with those of the
dissimilatory Fe(III) reduction microarray (Figure 3.3). This is especially true for the
genes involved in coenzyme metabolism and energy metabolism. Examples are the
genes for cobalamin biosynthesis, DMSO reductases, FAD dependent oxidoreductase,
coenzyme F390 synthetase, NiF e hydrogenase, and NADH dehydrogenase. However,
the wide range of induction seen with Fe-respiration (oxidoreductase genes and redox

protein genes) was not observed in U(VI)-reducing conditions.

In other studies of U(VI)-reduction (but not respiring), Desulfovibrio desulfuricans G20
and D. vulgaris, a periplasmic cytochrome c3 was implicated in direct reduction of U(VI)
to U(IV) (Payne 2004, Lovley 1993b). However, insoluble U(IV) was also found in the
cytoplasm (Sani 2006), suggesting that a secondary mode of U(VI) reduction was present
in the cytoplasm. Recently, transposon-mediated mutagenesis of Desulfovibrio
desulfuricans G20 identiﬁed an operon which was implicated in cytoplasmic reduction of
U(VI) (Li 2009). The operon-encoded oxidoreductase, thioredoxin, and thioredoxin
reductase added NADPH and was shown to reduce both U(VI) and Cr(VI) in vitro. One
such operon similar to the cytoplasmic U(VI) reductase of D. desulfuricans G20 is found
in D. hafniense with moderate levels of induction with U(VI) reduction (Dhaf_476-484,

3-7 fold).

63

Cells in the presence of U(VI) had induced transcriptional activities on other gene groups

whose functions include; chemotaxis, ﬂagella biosynthesis, cell wall biosynthesis, DNA

replication, DNA repair, and cell division. While COz-ﬁxing gene activity had increased,

Nz-ﬁxing gene activity had decreased. The acetyl-CoA pool appeared to be provided

from pyruvate (by pyruvate-formate lyase), acetate (by acetyl—CoA synthase), and the
Wood-Ljungdahl pathway. Acetyl-CoA formation from lipid breakdown (B-oxidation by
acetyl-CoA acetyltransferase and others) was severely inhibited, suggesting that cells
were preserving the lipid pool. Furthermore, heavy-metal efﬂux pump (P-type ATPase)

genes were signiﬁcantly induced.

All of these activities suggest that cells were highly stressed, accelerated cell division
while ﬁxing damaged DNA trying to prevent toxicity. There was no indication of
sporulation initiation, demonstrating that cells could tolerate U(VI) and were trying to
adapt to these less favorable conditions. High induction of DNA repair genes (radC and
recF) and different classes of DNA helicase genes suggest that DNA damage had
occurred either by chemicals, radiation, or possibly by intracellular insoluble U(IV)
species. This accelerated cell division attempt, regardless of whether division was
successﬁil, is likely to provide D. hafniense with a better chance of adaptation to U(VI)-
rich environments, through the creation of mutations and recombination events occurring

with DNA replication and repair.

64

U(VI) + Pyruvate

signal molecules. U(VI)

  
 
 
  
 
   

 
  

   

chemotaxis T
Chromosome
Replication

MCP
IIIATP synthaseg and Repair

U(VI)| reductase? U(VI) e?
mag DNA polymerase
g V Helicase
U(W) Repair protein

lllll

insoluble
Cell wall biosynthesrs
2e" 1
2H+

 

co2

formate

dehydrogenase acetate

 

-/ 1
— —u<—_

_........... _Y \

  

IIHHIHIH

 
 

 

  

 

: TCA cycle WOOd pathway P

 

 

 

I fumarate l—yl oxaloacetate I N AD NADH2

s'—@/'HIH\

  

 

 

fumarate

 

vUz

Figure 3.3. Schematic diagram of the key metabolism pathways during dissimilatory U(VI) reduction

65

with pyruvate. Menaquinone is abbreviated as “MK”, and cytochrome as “cyt”. Green arrows indicate
genes similarly expressed as pyruvate fermentation. Red arrows are genes up-regulated (in comparison to

pyruvate fermentation) in response to dissimilatory U(VI) reduction.

Nitrate-reducing cell metabolism. The microarray study on nitrate-reducing cells can

be characterized by two highly induced gene groups that are each involved in
assimilatory sulfate reduction and N2 ﬁxation-like ﬁinction by the Nifl operon (Figure
3.4). The 13 gene operon involved in the formation of H25 from SO42- was highly
induced (8-33 fold). Among these genes, four genes encoded ABC-type sulfate
transporter components: two genes for activation of SO42- by adenylation and

phosphorylation, a thioredoxin gene (Dhaf_4113, ~8 fold) for SO42— reduction to sulﬁte

(8032'), three genes for molybdopterin/thiamine biosynthesis, and two genes for sulﬁte

reductase (Dhaf_4107, 33 fold), which catalyzes H28 formation from 8032-. The Nifl

operon was also highly induced (5-26 fold) including a molybdopterin oxidoreductase

gene (Dhaf_1350, 16 fold) which is located immediately upstream of the Nz-ﬁxing

NifHDK genes and the nitrate (N 03-) transporter genes.

No signiﬁcant induction was observed in the Nap nitrate reductase operon (Dhaf_1289)
or the two Nrf nitrite reductase operons (Dhaf_3630-3631, Dhaf_4234-4235). Nitrate
reduction is preferred when the ratio of electron donor to nitrate is high. Therefore, under

a nitrate-rich environment, the two operons are not likely to be signiﬁcantly up-regulated.

66

Because the Nap nitrate reductase is located in the periplasm, this system may be used for
the reduction of external nitrate. During this process, ATP-generating proton

accumulation across the membrane may not be possible since protons released are likely
to be used immediately for H20 production. Together with the periplasmic Nrf nitrite
reductase, the Nap nitrate reductase system in D. hafniense may ﬁinction as an electron

sink as well as produce NH3. These two operons were signiﬁcantly induced (4-7 fold)

under Fe(III)-reducing conditions, where excess amounts of electrons were predicted to
form intracellularly because of the use of lactate substrate and the low electron-accepting

capacity of Fe(III).

67

Nitrate + Lactate

4 Sulfate transporter

llllllllilll

 
  
   
 

   
   
 
   
  
   

-$> -'l$> Nitrate transporter
PPi ADP _
b... =
" i
homoserine NADp’ H20 ,

Nitrogenase, Nif1

3;. "<—
formate :02 t t
dehydrogenase 4—

malate
TC A cycle \ NH3
I 2-oxoglutarate 9" glutamate I
DMSO

fumarate <—> succinate g
1_DMSOreductase DMS

+
2" 2H+ "ADM NADHZ

 

 

 

 

 

\HIIIIHHHHHH

 

 

 
   

      
 

fumarate
reductase

 

  

Figure 3.4. Schematic diagram of the key metabolism pathways during nitrate reduction with lactate.
Menaquinone is abbreviated as “MK”, cytochrome as “cyt”, adenylyl sulfate as “APS”, and phospho-

adenylyl sulfate as “PAPS”. Green arrows indicate genes similarly expressed as pyruvate fermentation.

68

Red arrows are genes up-regulated (in comparison to pyruvate fermentation) in response to nitrate

reduction.

Se(VI)-reducing cell metabolism. Out of 305 genes that were up-regulated during
Se(VI)-reducing conditions, 122 genes were of unknown functions. These unknown
genes were, in many cases, among the most highly induced genes. Since Se(VI) cultures
were provided pyruvate as the carbon and energy source, the numbers of genes that were
up-regulated and the overall induction rates were relatively low compared to pyruvate

fermentation (Figure 3.5).

Genes showing signiﬁcant induction include the genes for a specialized RNA polymerase

sigma-E factor (9 fold), transcriptional regulators, ABC transporters, N2 -ﬁxing

nitrogenase operon Nif3 (3 fold), ﬁimarate reductases (3 fold), glutamine synthase (6

fold), and PBSX prophage genes (2-8 fold). Whereas, genes showing transcription

reduction encoded succinyl—CoA synthetase (0.1 fold), COz-ﬁxation (0.2 fold), ﬂagella

biosynthesis (0.2 fold), NADH dehydrogenase (0.3 fold), menaquinone biosynthesis (0.3

fold), sulﬁte reductase (0.4 fold) and sulfate transporter (0.4 fold). These results may

indicate that cells no longer had to ﬁx as much C02 in order to balance their internal

redox state since Se(VI) has a large electron-accepting capacity. It is evident that the

dependency on NADH-mediated energy generation decreased. Cells also appeared to ﬁx

cellular nitrogen through Nz-ﬁxation and glutamine synthesis.

69

Escherichia coli K-12 carries a putative oxidoreductase gene, Yng, required for selenate
reduction (Bébien 2002), and the Ygﬂ< enzyme has three distinct domains: a N-terminal
NAD-binding domain, a central pyridine nucleotide-disulﬁde redox domain, and a C-
terminal [4Fe-4S]-binding domain. Two other genes in the same operon, YgﬂVIN were
predicted to encode proteins with a FAD-binding domain and a [2Fe-ZS]-binding domain,
respectively. The anaerobic bacterium Thauera selenatis contains the serABC operon
encoding a periplasmic heterotrimeric selenate reductase which shows high substrate
speciﬁcity unable to reduce nitrate, chlorate or sulfate (Krafft 2000). The SerABC genes
encode subunits with Mo active site, Fe-S clusters, and b-type cytochrome, respectively.
The Se(VI) reductase in a Gram-negative facultative anaerobe, Enterobacter cloacae
SLDla-l, is a membrane-bound heterotrimeric complex that faces the periplasm, and
contains Mo, heme, and nonheme iron as prosthetic groups and a b-type cytochrome in
the active complex (Ridley 2006, Watts 2003). The membrane-bound Se(VI) reductase
showed broader substrate speciﬁcity. In D. hafniense cultures grown with Se(VI), a
potential candidate gene encoding the b-type cytochrome protein was not identiﬁed.

However, three oxidoreductase genes that were up-regulated by 2-3 fold were found

(Dhaf_313, 480, 2375).

70

Se(VI) + Pyruvate

Se(VI) —> 9.,

 

  
   
  
    
  

 

'n
/ Chromosome
“(M-8"”: -\.

Unknown protein
Unknown protein

 
 
      

HHIIIIHHW“

 

 

Nitrogenase. NIQ
16 ATP

    
 

.

m... —» =
Y

t TCA cycle (:02
-_—P NAD+ NADHZ

.rtamf5s;@°'v Will

W

 

16 ADP

 

-/

 

 

 

H l||||| llH

 

 

 

2H”

Figure 3.5. Schematic diagram of the key metabolism pathways during Se(VI) reduction with pyruvate.
Menaquinone is abbreviated as “MK”, and cytochrome as “cyt”. Solid green arrows indicate g-es

similarly expressed as pyruvate fermentation. Red arrows are genes up-regulated (in comparison to

71

pyruvate fermentation) in response to Se(VI) reduction. The vesicle is external in D. hafniense, though for

simplicity depicted here as internal.

 

Fe(III)

Z
0
w I

U(VI) Se(VI)

 

N03' Nitrate
l _ reductase
"I’z Nitrite

 

reductase l
Nitrite
NH3 Reductase ll

 

 

N2 Nitrogenase l

 

 

 

1 "use"... it

000000:

 

00.000

 

NH3 Nitrogenase Ill

 

 

 

 

 

 

 

 

Figure 3.6. Induced expression of the genes involved in nitrate reduction and nitrogen ﬁxation
(pathways shown at left). Relative induction is shown in grey scales (light to dark; low to high induction).
No evidence was found which supports gene expression of nitrate and nitrite reductase genes in pyruvate-

fermenting cells. Therefore, relevant cells are shown in white.

Nitrogenase activity. As mentioned brieﬂy, D. hafniense can ﬁx N2, and three

nitrogenase operons are present in the genome, Niﬂ (Dhaf_1350-1360), Nii2
(Dhaf_1047-1059), and Nif3 (Dhaf_1810-18l8). These operons all contain a NifHDK
catalytic gene unit, but they differ in other genes residing in each operon. All of these
operons appeared to be expressed in pyruvate-fermenting cells. The eight-gene operon,
Nifl, which contains three ABC-type nitrate transporter genes, was strongly repressed

under Fe(III)-respiring conditions (Figure 3.6). The remaining two operons stayed at

72

similar levels of expression (slightly lower with Nif3 and slightly higher with Nif2). The

presence of nitrate transporter genes in the Nifl operon suggests that Nifl nitrogenase

system may be positively regulated by N03: This is supported by the fact that under
NOf-reducing conditions, the Nifl operon was highly up-regulated (5-26 fold). Also,

under NO3--reducing conditions, neither the Nap nitrate reductase system nor Nrf nitrite

reductase system were induced above the cutoff level (>2). In this respect, it is possible

that the Niﬂ operon has evolved for nitrate-reduction utilization, and not for the true N2-

ﬁxing pathway.

Determination of an active acetyI-CoA reduction pathway. Turbid growth was
observed with M. thermoacetica, the positive acetogen control, consistently within 7 days
in all 30:30:40 headspace combinations. Growth of D. hafniense under these same
headspace conditions was undetectable until cultures were shaken, revealing light growth.
The biomass produced was 25 to 50% less than that by M. thermoacetica (Table 3.1).
The vast majority of the D. hafniense biomass appeared to be extracellular soluble
organic carbon because very large amounts of precipitate formed following ethanol
addition during TOC sample processing. Growth of either species in carbonate-buffered
ﬁ'eshwater medium (CBF) with lactate was undetectable by all methods other than TOC,
which was used to determine the amount of growth attributable to inoculum alone. This
was selected as both D. hafniense and M. thermoacetica are reportedly unable to ferment
lactate alone, thus any biomass production above this indicated the organism was making

use of the provided terminal electron acceptor (Christiansen 1996, Brum 1988). Ten-fold

73

 

Table 3.1. Effect of acetogenic conditions
of cell yields of M. thermoacetica and D. hafniense

 

Carbon source Biomass produced

 

Organism Gas mixture Gas ratio Medium (20 mM) (mg/L)
D . hafniense C02:N2 5:95 DOB-1 pyruvate 15.1
D. hafniense COZ:N2 20:80 CBF lactate 0.295
D. hafniense CO:N2 40:60 DOB-1 - 1 .44
D . hafniense CO:COZ:N2 30:30:40 DOB-1 - 13.9
D. hafniense H2:CO:N2 30:30:40 DCB-1 - 33.1
D . hafniense H2:COZ:N2 30:30:40 DCB-1 - 1 3.3
D. hafniense COZ:N2 40:60 DCB-1 - 1 .00
M. thermoacetica COZ:N2 5:95 DCB-1 pyruvate 34.3
M. thennoacetica COZ:N2 20:80 CBF lactate 3.19
M. thermoacetica CO:N2 40:60 DOB-1 - 4.10
M. thermoacetica CO:C02:N2 30:30:40 DOB-1 - 46.5
M. thermoacetica H2:CO:N2 30:30:40 DOB-1 - 54.5
M. thermoacetica H2:COZ:N2 30:30:40 DOB-1 - 24.7
M . thermoacetica COZ:N2 40:60 DCB-1 - 3.16

 

more biomass was produced in CBF with lactate alone for M. thermoacetica than with D.
hafniense (Table 3.1), revealing that either M. thermoacetica possessed more residual

nutrient carry-over from inoculation, or more likely that it was better able to use the

components present in the CBF medium (e.g. vitamins, 20% C02, etc.) to yield low

amounts of growth. Moreover, M. thermoacetica produced twice as much biomass with

pyruvate fermentation alone as did D. hafniense (Table 3.1).

All gas combinations for both species produced more biomass than CBF with lactate (the

amount of biomass attributable to inoculation) (Table 3.1). It is thought that even in the

absence of a typical electron donor, e. g. hydrogen gas, D. hafniense may utilize C02 and

CO as electron donors and acceptors for growth. The amount of gaseous carbon was

74

twice the amount in CBF medium with C02:N2 (20:80). A 5-fold increase in biomass

was observed in CO:N2, and a 3-fold increase in C02:N2 (Table 3.1). M. thermoacetica

remained unchanged with C02:N2, and increased by 33% with CO:N2. In the presence

of H2:CO:N2, both species produced the most cell material, followed by CO:C02:N2 and

H2:C02:N2. The CO:C02:N2 yields are considerably less than those found by Daniels

(1993) (146 mg/L); however, those studies were performed with an undeﬁned medium

containing yeast extract. The low amount of grth ﬁom CO:N2 for either D. hafniense

or M. thermoacetica is not surprising. When M. thermoacetica was previously cultured

in CO:N2, glucose and yeast extract were provided, and when glucose was removed,

cultures did not grow reproducibly (Kerby 1983). It was an unexpected result that
CO:C02:N2 would facilitate more signiﬁcant biomass production than H2:C022N2 [when
compared to results found by Daniel (1990)], which indicates a preference for CO as both
the electron donor and acceptor for acetogenesis. Controls of media alone did not

produce detectable biomass, thus all biomass was produced by the bacteria. Some

biomass production by D. hafniense and M. thermoacetica was observed in cultures with

only CO:N2 or C02:N2. These ﬁndings indicate that CO and C02 might be used both as

electron acceptors and donors but with better biomass yields when H2 and C0 are

provided for the electron donor and acceptor (30:30:40 combination). This evidence
strongly suggests D. hafniense is capable of growth as an acetogen, though this may be

weak acetogenic grth due to the bulk of its ﬁxed carbon forming from small soluble

75

 

Figure 3.7. SEM of D. hafniense DCB-2 exposed to diﬁ‘erent electron acceptors. (A) Curved rod
morphology during pyruvate fermentation. (B) Curved rod morphology during nitrate respiration. (C)
Elongated rod exhibited during U(VI) exposure. (D) Elongated rod with vesicles during growth with
Se(VI). (E) Streptobacilli made of shorter rods during Fe(III) respiration. (F) SEM of Fe(III) biomass.
White arrow points to cell in G. (G) Backscatter image of Fe(III) biomass from E. White arrow points to

cell in F.

extracellular polymeric substances rather than cells.

Gene expression corresponding to cell morphology differences. Cell morphology
observed during metal reduction appears to mirror changes in gene expression. In the
absence of metal-reducing conditions, similar curved rods were found (2.5 pm with both
pyruvate fermentation and nitrate respiration) (Figure 3.3A and B). The presence of both

Se(VI) and U(VI) in medium with pyruvate resulted in cells elongating to nearly twice

76

their normal cell length (averaging 4.7 and 4.8 pm per cell respectively) (Figure 3.3C and
D). This observation seems to correlate with the gene expression of cells trying to limit
exposure to toxic U(VI). While further investigation would be required to determine if
this was the case for Se(VI)-reducing conditions, repressed grth has been found with
Se(VI) (Chapter 2). Se(VI) exposed cells also produced extracellular vesicles of
selenium (Chapter 4). Perhaps the genes responsible for this adaptation are in the 122
genes of unknown function. The highly expressed RelB antitoxin-like gene (Dhaf_0577)
might ﬁrnction in part of these processes. Shorter cells in a streptobacilli formation were
present with Fe(III) (Fig 3.3E). Surrounding these chains of cells was large amounts of
biomass containing elements of high atomic weight (Figure 3.3F and G). Upon
investigation with EDS, bound iron was determined to be the cause of the heavy elements
detected in the biomass (data not shown). This high amount of extracellular material
production reﬂects the large induction of the Wood-Ljungdahl pathway and
gluconeogenesis (Figure 3.2). A large number of oxidoreductases and a ferrous
transporter gene were highly expressed, this along with the induction of the carbon-
ﬁxation pathway provide evidence for the Fe(II)-bound in biomass as seen in Figure

3.3G.

Conclusions. In summary, the largest contrast in gene expression existed between
pyruvate fermentation and Fe(III) reduction. Genes expressed with U(VI)-reducing

conditions were similar to those of Fe(III)-reducing conditions, while Se(VI)-reducing
conditions had many highly expressed genes in common with NO3'-reducing conditions,

though most were of unknown ﬂinction. This difference may be in part due to the

77

number of electrons reduced per molecule of metal; however, ﬁrrther research would be

required to verify this interpretation.

The physiology of D. hafniense DCB-2 has been expanded to such diverse grth as
fermentation, halorespiration, metalorespiration, and now acetogenesis. The growth of
D. hafniense in conditions providing only a gaseous electron donor, electron acceptor,
and carbon source have shown that it is an acetogen; however, this grth was weak
compared to the profuse cell material produced by M. thermoacetica. Moreover, CO was
found to be the preferred carbon source for cell material production during acetogenesis.
Further studies of metal-reduction supported by gaseous carbon may be of interest, as
bioremediation with D. hafniense may be less expensive using gaseous carbon to support

growth.

ACKNOWLEDGEMENTS
Research was supported by the U.S. Department of Energy, Ofﬁce of Science,
Environmental Remediation Science and Genomes to Life Programs. The collaboration
of Dr. Sang-Hoon Kim of Michigan State University (East Lansing, MI) was greatly

appreciated in operon expression analyses and creation of ﬁgures within this chapter.

78

 

REFERENCES

Bébien, M., J. Kirsch, V. Me'jean, and A. Verméglio. 2002. Involvement of a putative
molybdenum enzyme in the reduction of selenate by Escherichia coli. Microbiology
148:3865-3872.

Bloem, J. 1995. Fluorescent staining of microbes for total counts. In: Akkermans A.D.L.,
Elsas J .D., Bruijn F.J. (eds), Molecular Microbial Ecology Manual. Kluwer Academic
Press, Dordrecht, The Netherlands, pp. 1-12.

Brumm, P. J. 1988. Fermentation of single and mixed substrates by the parent and an
acid-tolerant mutant strain of Clostridium thermoaceticum. Biotech. Bioeng. 32:444—450.

Christiansen, N. and F. K. Ahring. 1996. Desulﬁtobacterium hafniense sp. nov.,
anaerobic, reductively dechlorinating bacterium. Int. J. Syst. Bacteriol. 46:442-448.

Daniel, S. L. and H. L. Drake. 1990. Characterization of the H2- and CO-dependent
chemolithotrophic potentials of the acetogens Clostridium thermoaceticum and
Acetogenium kivui. J. Bacteriol 17224464-4471.

Gorby, Y. A., T. J. Beveridge, and R. P. Blakemore. 1988. Characterization of the
bacterial magnetosome membrane. J. Bacteriol. 170:834-841.

He, 2., L. Wu, X. Li, M. W. Fields, and J. Zhou. 2005. Empirical establishment of
oligonucleotide probe design criteria. Appl. Environ. Microbiol. 71:3753-3760.

Kerby, R. and J. G. Zeikus. 1983. Growth of Clostridium thermoaceticum on H2/C02
or CO as energy source. Curr. Microbio]. 8:27-30.

Kostka, J. E., D. D. Dalton, H. Skelton, S. Dollhopf, and J. W. Stucki. 2002. Grth
of iron(III)—reducing bacteria .on clay minerals as the sole electron acceptor and

comparison of growth yields on a variety of oxidized iron forms. Appl. Environ.
Microbiol. 68:6256-6262.

Krafft, T., A. Bowen, F. Theis, and J. M. Macy. 2000. Cloning and sequencing of the
genes encoding the periplasmic-cytochrome B-containing selenate reductase for Thauera
selenatis. DNA Sequence 10:365-377.

Lanthier, M.,. R. Villemur, F. Lepine, J-G. Bisaillon, and R. Beaudet. 2001.
Geographic distribution of Desulﬁtobacterium frappieri PCP—l and Desulﬁtobacterium
spp. in soils from the province of Quebec, Canada. FEMS Micro. Ecol. 36:185-191.

Li, X., Z. He, and J. Zhou. 2005. Selection of optimal oligonucleotide probes for

microarrays using multiple criteria, global alignment and parameter estimation. Nucleic
Acids Res. 33:6114-6123.

79

 

Li, X and L. Krumholz. 2009. Thioredoxin is Involved in U(VI) and Cr(VI) Reduction
in Desulfovibrio desulfuricans G20. J. Bacteriol. published online ahead of print.

Lbffler, F. E., R. A. Sanford, and J. M. Tiedje. 1996. Initial characterization of a
reductive dehalogenase from Desulﬁtobacterium chlororespirans C023. Appl. Environ.
Microbiol. 62:4982-4985.

Lovley, D. R. 1993a. Dissimilatory metal reduction. Annu. Rev. Microbiol. 47:263-290.

Lovley, D. R., P. K. Widman, J. C. Woodward and E. J. Phillips. 1993b. Reduction of
uranium by cytochrome c3 of Desulfovibrio vulgaris. Appl. Environ. Microbiol. 59:3572-
3576.

Lovley, D. R. and E. J. P. Phillips. 1988. Novel mode of microbial energy metabolism:

organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl.
Environ. Microbiol. 54:1472-1480.

Madsen, T. and D. Licht. 1992. Isolation and characterization of an anaerobic
chlorophenol-transforming bacterium. Appl. Environ. Microbiol. 58:2874-2878.

Miller, T. L. and W. J. Wolin. 1974. A serum bottle modiﬁcation of the Hungate
technique for cultivating obligate anaerobes. Appl. Microbiol. 27:985-987.

Niggemyer, A., S. Spring, E. Stackebrandt, and R. F. Rosenzweig. 2001. Isolation and
characterization of a novel As(V)-reducing bacterium: implications for arsenic
mobilization and the genus Desulfitobacterium. Appl. Environ. Microbiol. 67:5568-5580.

Payne, R. B., L. Casalot, T. Rivere, J. H. Terry, L. Larsen, B. J. Giles and J. D.
Wall. 2004. Interaction between uranium and the cytochrome c3 of Desulfovibrio
desulfuricans strain G20. Arch. Microbiol. 181 :398-406.

Ridley, H., C. A. Watts, D. J. Richardson, and C. S. Butler. 2006. Resolution of
distinct membrane-bound enzymes ﬁom Enterobacter cloacae SLDl a-l that are

responsible for selective reduction of nitrate and selenate oxyanions. Appl. Environ.
Microbiol. 72:5173-5180.

Riley, R.G., J. M. Zachara, and F. J. Wobber. 1992. Chemical contaminants on DOE
lands and selection of contaminant mixtures for subsurface science research. DOE/ER-
0547T. U.S. Dept. of Energy. Subsurface Science Program, Washington, DC.

Rooney-Varga, J. N., R. T. Anderson, J. L. Fraga, D. Ringelberg, and D. R. Lovley.
1999. Microbial communities associated with anaerobic benzene degradation in a
petroleum-contaminated aquifer. Appl. Environ. Microbiol. 65:3056-3064.

Sani, R. K., B. M. Peyton and A. Dohnalkova. 2006. Toxic effects of uranium on
Desulfovibrio desulfurz'cans G20. Environ. Toxicol. Chem. 25: 123 l— 1238.

80

Watts, C. A., H. Ridley, K. L. Condie, J. T. Leaver, D. J. Richardson, and C. S.
Butler. 2003. Selenate reduction by Enterobacter cloacae SLDla-l is catalyzed by a

molybdenum-dependent membrane-bound nitrate reductase. FEMS Microbiol. Lett.
228:273-279.

Wolin, E. A., R. S. Wolfe, and M. J. Wolin. 1963. Viologen dye inhibition of methane
formation by Methanobacillus omelianskii. J. Bacteriol. 872993-998. .

81

Chapter IV

Exterior Membrane-Bound Selenium Vesicles Produced by Desulﬁtobacterium

hafniense DCB-2 in Response to Selenate Exposure.

ABSTRACT
Desulﬁtobacterium hafniense DCB-2, a Gram-positive bacterium with abilities to reduce
N, S, Fe, Mn and chlorinated compounds, grew in minimal medium with selenate
concentrations up to 25 mM, and reduced both selenate and selenite to insoluble Se(O)
forms. When grown with selenate, the cells produced extracellular vesicles containing
the selenium, while the cells contained no detectable selenium. Gram, lipophilic,
live/dead® and DAPI stains conﬁrmed that the Se—containing vesicles possessed
membranes with cell wall but lacked detectable DNA. Red and black selenium
precipitates were produced when grown with selenate, while only a red precipitate was
produced from growth with 1 mM selenite. Se XANES established that both red and
black precipitates were elemental selenium, and XRD showed that the red precipitates are
XRD amorphous and the black precipitates are slightly more ordered with several small
broad peaks consistent with elemental selenium. Extrusion of membrane vesicles that
bind the reduced selenium appears to be a mechanism to protect cells ﬁ'om selenium

toxicity as well as results in insoluble selenium.

82

INTRODUCTION
Selenium is found in four oxidation states in nature: Se(-II) (sodium selenide and
hydrogen selenide), Se(O) (colloidal, elemental selenium), Se(IV) (selenite, selenium
dioxide and selenious acid), and Se(VI) (selenate and selenic acid). Although an
essential trace element, selenium is toxic at only three to ﬁve times its required
concentration for nutrition, producing clinical symptoms such as “blind staggers”, hair
loss, and nail disﬁgurement (Diplock 1981, NAS 1976, Young 1981). Elemental
selenium is neither absorbed by the intestine nor taken up by plants; however, selenite
and selenate are soluble and taken up by plants. Soils elevated in these ions can cause
elevated Se concentrations in plants and toxicity in organisms ingesting them.
Distribution of selenium in soil varies regionally, ranging from 0.1 to 4.3 ppm in US soils
with a mean of 0.3 ppm (Shacklette 1974). Selenate is the most bioavailable form and
predominates in most soils above pH 5.5 (Elrashidi and Adriano 1987, Neal 1987 and

1989, Shamberger 1981).

Bioreduction to insoluble Se(O) using bacteria has been proposed as a means of selenate
and selenite bioremediation ﬁom sites with toxic levels; thereby decreasing the risk to
humans, plants, and animals (Losi 1997). Bacterial species capable of reducing selenate
have been identiﬁed in many phylogenetic groups. Some can use selenium reduction to
support grth as a terminal electron acceptor, while others reduce selenium without
supporting growth (Stolz 1999). Microorganisms capable of reducing selenate for
respiratory grth include some Crenarcheota, B-, y-, 8-, s— Proteobacteria, and both low

and high G+C Gram-positive bacteria (Stolz 2002, Narasingarao 2007, Huber 2000).

83

However, the processes by which bacteria reduce selenium are poorly understood.

Several bacterial genera have been studied to gain insight into bioreduction of selenate.
In Sulfospirillum barnesii, selenate reduction is localized in the membrane ﬁaction, while
a soluble selenate reductase has been isolated from Thauera selenatis (Stolz 2002).
There is no consistent pattern as to where bacteria localize reduced (insoluble) selenium
since it is found in the cytoplasm, cell wall, or the surrounding medium, and in various
combinations of these three in different species (Milne 1998). Whole cells and cell-free
extracts of Clostridium pasteurianum reduce selenate to elemental selenium, but cell-
associated selenium was not found (Harrison 1980). Oremland (2004) reported Bacillus
selenitireducens, S. barnesii, and Sulfospirillum shriftii produce uniform nanospheres of
elemental selenium following selenium reduction, which were located both internal and
external to the cell. These elemental selenium spheres were either red or black, indicative
of reduction state, for each bacterial species, but were never both colors. Sulfate and
nitrate reduction pathways have been shown to also reduce selenate in some bacteria
(Milne 1998). In S. barnesii, however, selenate and nitrate induced different reduction
pathways, though both pathways could reduce selenite and nitrite (Oremland 1994). The
dissimilatory sulfur reduction pathway was also proposed to reduce selenite, but not

selenate.

We studied selenium reduction in Desulﬁtobacterium hafniense DCB-2, a Gram-positive,
spore-forming anaerobe with a broad capacity for using inorganic electron acceptors for

growth, including iron, manganese, arsenic, nitrate, sulﬁte, but not sulfate (unpublished

84

data and Christiansen 1996). Here we focus on its selenium reducing abilities, its
reduction products, and the selenium-containing vesicles it produces when grown with
selenate. We then propose a morphologically-based model of selenate reduction in which

external vesicles play a key role.

METHODS
Culture conditions. Desulﬁtobacterium hafniense DCB-2 was cultivated using strict

anaerobic technique at 37°C in a modiﬁed DCB-l medium: (per liter) 1 g NaCl, 0.5 g

MgClz, 0.2 g KH2P04, 0.3 g KCl, 0.3 g NH4C1, 15 mg CaClz, with 1 m1 trace elements

solution SL-lO, 1 ml selenite tungstate solution, 0.25 mg resazurin, and 10 mM HEPES,

pH 7.0 (Lﬁfﬂer 1996, Tschech 1984, Widdel 1983). Salts and trace elements were

mixed, heated and cooled under a gas mixture of N2:COz (95:5) (Miller 1974). Cysteine

(0.08 g) was added, and the medium was buffered to pH 7.0 with NaHCO3. The medium

was then dispensed into 25 ml Balch tubes, 120 ml or 1 L serum bottles; all sealed with
butyl-rubber stoppers and aluminum crimp tops, and autoclaved for 40 min. Prior to
inoculation, Wolin vitamins, (Wolin 1963), sodium pyruvate (pyruvate), and sodium
selenate were added from separate sterilized stocks to yield ﬁnal concentrations of 1
ml/L, 20 mM and 1 mM, respectively. Sodium selenate concentrations up to 25 mM and
sodium selenite concentrations up to 1 mM were tested, but 1 mM sodium selenate was
used in this study unless otherwise stated. The controls were medium without selenium
inoculated with a live culture, medium with selenium but without inoculation, medium

with selenium and heat-killed cells inoculated as a cell pellet, and a heat-killed culture to

85

which selenium was then added. Neither selenium precipitate nor any evidence of

selenate or selenite reduction was observed in any of the controls.

Growth. Cultures were grown at 37°C without shaking. Samples were removed

aseptically from cultures and controls using syringes in an anaerobic glove box

containing N21H2 (95:5) at time 0 and every 6 h, for 126 h. Portions of the samples were

used immediately for optical density, organic acid analysis, colony counts, and
microscopy. The remainder of the sample was frozen at -20°C for further analyses. All

culture manipulations were performed in an anaerobic glove box.

Bacterial grth was monitored by optical density at 600 nm and by consumption and
production of organic acids as monitored by HPLC. Samples for HPLC were collected
and ﬁltered through a 0.45 pm membrane under anaerobic conditions. Organic acids
were quantiﬁed at 210 nm on a Hewlett-Packard 1050 HPLC system equipped with a
heated C18 column using 0.013 N sulfuric acid as the solvent pumped at a rate of 0.6
ml/min. Standard curves of pyruvate, acetate, lactate and combinations of these were
made from 40 mM down to 0.1 mM without loss of detection sensitivity. Colony counts

were determined on R2A medium with 10 mM pyruvate alter anaerobic incubation for 10

days in a N22H2 (95:5) atmosphere at 30°C.

Microscopy. Gram stains of vesicles sampled over the time course were observed on

two replicate slides per time point with each slide consisting of two replicate samples and

86

one control mixture of Escherichia coli and Staphylococcus aureus. Cells of wet mounts
and Gram stains were observed under bright ﬁeld on a Nikon epiﬂuorescent microscope.
The E. coli and S. aureus control consistently showed Gram-negative and Gram-positive
staining, respectively. Portions of the same samples were used for DAPI staining,
scanning electron microscopy (SEM), transmission electron microscopy (TEM), and
energy dispersive spectroscopy (EDS). Duplicate samples were stained for DNA using

the DAPI protocol of Bloem (1995) and viewed on a Nikon epiﬂuorescent microscope.

Culture purity was tested in duplicate by comparing ﬂuorescent in situ hybridization
(FISH) staining on all samples using a Cy3-labeled universal eubacterial rDNA probe,
Cy5-labeled D. hafniense speciﬁc rDNA probe (Lanthier 2001), and DAPI staining
(Bloem 1995). E. coli was used as a quality control for all dyes and as a negative control
for the D. hafniense-speciﬁc rDNA probe. Slides were observed on a Nikon

epiﬂuorescent microscope equipped with a 50 W mercury lamp.

For the live/dead® stain, propidium iodide and SYTO 9 (Molecular Probes, Eugene,
Oregon) were used, while N-(3-triethy1ammoniumpropyl)-4- (p-diethylaminophenyl-
hexatrienyl) pyridinium dibromide (FM 4—64) was used as the lipophilic dye (Molecular
Probes). Live/dead® staining, lipophilic staining, SEM, TEM, and EDS were done on
samples taken at 24, 48 and 72 h. Stained cultures were viewed using an Olympus
Flroiew 1000 laser scanning confocal microscope (Olympus America Inc., Center
Valley, PA). Bright ﬁeld microscopy provided target conﬁrmation. For viewing

membrane disruption in the presence of dyes, 33% formamide (Sigma Aldrich, St. Louis,

87

MO) was applied to avoid compromising dye ﬂuorescence.

Three samples were prepared with gold coating for SEM and three with carbon coating
for EDS. Samples were ﬁxed in 2.5% paraformaldehyde, 2.5% glutaraldehyde and 0.1 M
cacodylate buffer, pH 7.4 (Electron Microscopy Sciences, Hatﬁeld, PA). A drop of

sample was added to 1% poly-L-lysine (Sigma Aldrich) coated, 12 mm, circular cover

slips and left to adhere for 5 min prior to being gently rinsed with sterile H20. Slow

dehydration was achieved through an ethanol series of 1 min in each of 10%, 25%, 50%,
75%, 85%, 90%, 95%, and twice in 100%. A Balzers critical point dryer (Balzers,
Lichtenstein) was used for drying before mounting the cover slips on aluminum stubs and
coating for viewing. Gold was sputter coated using an EMSCOPE SCSOO sputter coater
(Ashford, Kent, Great Britain), and the carbon was coated using a carbon string
evaporator (EFFA MkII Carbon Coater, Ernest F. Fullam Inc., Latham, New York).
Samples were viewed on a JEOL 6400V (Japan Electron Optics Laboratories, Tokyo,
Japan) microscope with a LaB6 emitter (Noran EDS). The average number of cells per
stub viewed at 24 and 72 h of selenium exposure was 1,960 and 3,570 respectively, for a
total of over 11,000 and 21,000 cells. At least 20 photos per stub were acquired, for

documentation.

Three sample replicates per time point were prepared for TEM. Each ﬁxed sample was
embedded in 2% agarose and secondarily stained with 1% osmium tetroxide (Sigma
Aldrich) in 0.1 M cacodylate buffer. An acetone series of l min in each of 10%, 25%,

50% 75%, 85%, 90%, 95%, and twice in 100% was used for sample dehydration. The

88

sample was then embedded in the epoxy resin Poly/Bed 812, cut using a MTX
ultramicrotome (RMC, Tucson, AZ), and sections were placed on formvar-coated copper
grids. These sections were then stained with 1% uranyl acetate followed by lead citrate
with sodium hydroxide pellets present (Venable 1965). The grids were imaged on a
JEOL 2200FS 200 keV emission TEM EDS (Japan Electron Optics Laboratories, Tokyo,
Japan). The number of cells viewed averaged 1,790 per copper grid at 72 h of selenium
exposure for a total of over 5,370 cells viewed. A minimum of 20 photos per condition

were acquired for documentation.

XANES and XRD analyses. For x-ray absorption near edge structure (XANES)

analysis, biomass samples containing selenium were prepared in an anaerobic glove box

of N2:H2 (95:5). The biomass was captured on a 0.2 pm ﬁlter (Millipore, Bedford, MA),

wrapped with Kapton ﬁlm, and sealed air tight with Kapton tape. The x-ray
measurements were made on the moist biomass paste within 24 h of preparation to
prevent sample dehydration and oxygen exposure. The samples were kept in the anoxic
chamber until a few hours before measurement and then transferred to the beamline in a
Mason jar sealed under anoxic conditions. The Kapton sealed samples were exposed to
the atmosphere during the x-ray measurement for 1 to 2 h. The Se K-edge XANES
measurements were made at beamline locations MRCAT 10-ID (Segre 2000) and PNC-
CAT 20-BM at the Advanced Photon Source at Argonne National Laboratory. At
MRCAT, the insertion device was tapered, and the Si(III) double crystal monochromator

was slew scanned. Several quick EXAF S measurements in transmission mode were

89

averaged for each sample. The incident ionization chamber was ﬁlled with N2, and the

transmission ionization chamber was ﬁlled with Aer2 (50:50). A Rh mirror was used to

remove harmonic x-rays. The x-ray beam size of 1 mm X 1 mm was incident on the
sample. The beamline parameters were similar at PNC-CAT 20-BM, where the Si(III)
monochromator was detuned 15% to enhance harmonic rejection, and scans were
collected in step scanning mode in ﬂuorescence with a thirteen element Ge solid state
detector. The same selenium metal standard was measured with each scan and used to

align the sample scans to each other.

The XANES spectra were processed using Athena (Ravel 2005) and IFEFFIT software
(Newville 2001). The spectra were aligned to each other and the reference spectra.
Multiple scans from the same sample were merged to increase the data quality. Linear
combinations of the control sample and the sample showing the most reduction (D.
hafniense black precipitate) were used to estimate the amount of Se in the initial
oxidation state that was not reduced. The Se(O) standard was not used because of the
differences measured in the white line shape and intensity, the strong peak in the spectra
at the absorption edge, as compared to the samples. This measured difference may
indicate small particle size within the samples compared to the standard and/or more

disorder in the measured samples as compared to the standard.

X-ray difﬁaction (XRD) was performed on the ﬁltered samples ﬁom synchrotron

analysis of the red and black precipitates ﬁom D. hafniense in DCB-l medium using a

90

Rigaku Miniﬂex x-ray diffractometer with a Cu anode with step size of 002°C and a
2theta range ﬁom 5 to 90°C. Spectra lines were compared with the lines from selenium

standard ICSD-40016.

RESULTS
Growth in the presence of Se. Desulﬁtobacterium hafniense grew when incubated with
sodium selenate at concentrations up to 25 mM, the highest concentration tested, but the
doubling time increased with concentrations greater than 1 mM (Figure 4.1A). A red
precipitate was produced when grown with all concentrations of sodium selenate and
with 1 mM sodium selenite. More red precipitate was produced over time at all selenium
concentrations. This particle formation accounted for the increased standard error in the
optical density measurement (Figure 4.1B). Cell counts more accurately reﬂect cell
grth and were used to calculate doubling times (Figure 4.1A). Pyruvate was
consumed by D. hafniense to produce acetate (Figure 4.1B) similar to cells were grown

without selenium. No precipitate was produced in any of the controls.

91

 

 

 

351
soi f
i
E 25 —
5
g 20 ~ i
E 15 ‘ .
‘3’ l i
o 10 i
1 . ii
51
o . . .
0.01 0.1 1 10 100
[Selenate] (M)
B
30 _ ix—PyWe‘ —— 0.2
l o Acetate 1 ~ 0.18
25 f l o Turbidity t 0.16
I i L—— E
C
A O
E E
2' .é‘
a a
.5. 8
O.
3 o

 

 

0 50 100 150 200
Time (hrs)

Figure 4.1. Growth of D. hafniense in sodium selenate medium. (A) Effect of sodium selenate
concentration on growth rate: (A) the zero concentration is displayed as 0.01 to accommodate the log

scale. (B) Growth with 1 mM sodium selenate measured by optieal density, pyruvate consumption and

92

acetate production. Standard error is shown for all measurements as vertical lines.

 

Figure 4.2. Morphological changes of D. hafniense cells when grown under pyruvate fermentation
conditions. (A) SEM of a cell grown on pyruvate with no selenium. (B-F) Cells grown with 1 mM
selenate. (B) SEM of elongated cells with buds or spheres at 24 h. (C) SEM of cells showing an increase
in sphere number and development at 72 h. White arrows in C and E indicate spheres veriﬁed to be
attached to cells via visible ﬁlament connections. (D) SEM of spheres both detached and cell-associated at
72 h. (E) SEM at 72 h of ‘beads on a string’ (ﬁlament with Se spheres). (F) TEM at 72 h of cells
surrounded by spheres (very dark circles). Black arrows point to Sc spheres where we found Se and lipid-

binding osmium predominant arormd the outer edge. This was conﬁrmed by EDS.

Sphere formation. Bright ﬁeld microscopy of D. hafniense cells revealed an abundance
of refractive spheres on and around the cells (Figure 4.2) but only when grown with
sodium selenate. Images of cells taken over time during growth with 1 mM sodium

selenate were viewed under SEM and TEM to determine patterns of morphological

93

change, including sphere production. Under SEM, selenium-grown cells were elongated
(Figure 4.2B) compared to cells not exposed to selenate (Figure 4.2A). At 24 h, small
raised buds, or potential spheres, were visible on the cell surface (Figure 4.2B). After 48
h, spherical structures appeared on the surface of cells or were protruding on elongated
ﬁlament structures away from the cell where further buds were observed. Detached
spheres were approximately 200 nm in diameter as measured by SEM. At 72 h of
exposure, cells had multiple 200 nm spheres both attached and detached ﬁ'om the cells
(Figure 4.2C and D). Electron microscopy showed ﬁlament connections between cells
and spheres. New spheres appeared to repeatedly form in localized spots producing
‘beads on a string’ (Figure 4.2E). Selenium as determined by TEM EDS was found only
associated with spheres and was not detectable in cells (detection limit 0.01%) (Figure
4.2F). Moreover, the spheres were found only external to the elongated cell. The
selenium signal intensity of spheres was highest at sphere edges with different degrees of
intensity internally. The majority of spheres; however, possessed strong selenium signal
intensity throughout the sphere as determined by TEM EDS and shown by their dark

pigmentation (Figure 4.2F).

Production of two elemental Se precipitates. In D. hafniense cultures grown with
selenate, two different colored precipitates were noted, red and black, with red being the
most predominant form. When the red and black precipitates were found together, the
red precipitate formed prior to the black. In rare cases, cultures appeared to exclusively
produce black precipitates, but upon closer observation, red precipitate was present. All

precipitates were produced with the same medium under identical conditions. Parallel

94

experiments using a separate basal medium in which only the trace elements and the
gaseous head space varied produced only red precipitate (data not shown), suggesting
that subtle differences inﬂuence the precipitate form. In all cultures, no organism other

than D. hafniense was found.

The Se reduction state was investigated from Se XANES spectra of the red and black
precipitates along with a control sample of medium with selenium but without
inoculation (negative control) and a Se metal standard (reference) (Figure 4.3A). The
negative control spectrum has a strong white line, which is a strong peak in the spectra at
the absorption edge. The white line was not visible in the D. hafniense black precipitate;
this sample was used as the fully reduced standard for the linear combination ﬁt (LCF).
The white line ﬁom the negative control spectrum is barely apparent in the D. hafniense
red precipitate spectrum. From LCF calculations, the oxidation state of Se in the red

precipitate is 2 :l: 10%.

95

A

 

5 ..
—o. M. Black
‘ 0 0.11.1. Red
4 _ —Lci=
88 Metal Reference
" —Negatlve Control

 

Normalized X-ray Absorption

0g

 

 

I W I I I

l l I
12640 12650 12660 12670 12680
X-ray Energy (eV)

 

Se Red Precipitate
Se Black Precipitate
2000 _ - Selenium ICSD-40016

 

b ..
'6 1500—
C

9 .1
E 1000—

l

 

or
o
o
1

o
l

 

 

 

 

2theta

Figure 4.3. (A) Se K-edge XANES spectra from sterile selenium medium control (negative control), D.

96

hafniense cultures that formed black and red precipitates and the Sc metal reference made at the APS at
Argonne National Laboratory. The measured spectra for red precipitate are shown as symbols, and the
LCF is shown as a line. (B) XRD spectra of red and black selenium precipitates from D. hafniense grown

in 1 mM sodium selenate compared to the XRD pattern of metallic elemental selenium.

The structure of both precipitates was determined by XRD (Figure 4.3B). Large particles
of metallic selenium did not form in the red or black precipitates, as evidenced by the
absence of strong diffraction peaks in the measured XRD spectra based on the calculated
pattern of metallic selenium. This occurred even though the Sc XANES spectra indicate
predominately elemental Se (Figure 4.3A). Instead, the red precipitates are almost XRD
amorphous as no XRD signal was detected (deﬁned by the absence of any strong
diffraction peaks) (Fig 4.3B). For the black precipitates, some of the predicted lines are
visible as small broad peaks indicating that some limited ordering of selenium metal
particles may have occurred in this sample (Figure 4.3B). Moreover, as different XRD
patterns were observed (i.e. no signal from the red precipitates, and small broad signals
from the black precipitates), we suggest that the increased ordering in the precipitates is
responsible for the color shift from red (XRD amorphous elemental selenium) to black

(slightly crystalline elemental selenium).

Chronology of spherical, membrane-bound Se vesicles. Spherical bodies were visible
on the outer surface of elongated cells exposed to selenate (Figure 4.4A). Higher atomic
weights were present where spherical precipitates were located when imaged using SEM
backseatter (Figure 4.4B). SEM EDS showed the selenium was concentrated in the

spheres (Figure 4.4C -arrow 2), and was not detected in the cell itself (Figure 4.4C -arrow

97

1). The detached spheres, more prominent at 72 h, surrounded hexagonal crystals (Figure
4.4D). The crystals appeared to be composed of 200 nm spherical components that were
of similar intensity in TEM backscatter (Figure 4.4E). TEM EDS (data not shown)
conﬁrmed that the hexagonal crystals were selenium. A TEM EDS line scan (Figure
4.4F and G) shows that selenium was below the detection limit in cells, but produced the
highest Se signals in locations corresponding to the spheres. Osmium, which binds to
lipids, was elevated not only at the edges of the cells, but also along edges of the spheres.
This provides evidence that a membrane was surrounding the spheres, indicating they can
be termed vesicles (Fig 4.2F and 4.4F). The carbon signal remained high throughout the
cells and at regions around the vesicles, evidence consistent with lipids surrounding the
vesicles. Similar results were found on eight separate TEM EDS scans that sampled 265
vesicles, 9 cells, and 14 hexagonal crystals. All evidence pointed to selenium sequestered

inside the membrane bound vesicles and not in the cells.

To further discern if the selenium spheres were membrane-bound selenium vesicles, 1
mM selenate grown cultures were stained with a ﬂuorescent lipophilic dye (FM—464).
Under the confocal microscope, the dye was concentrated at the perimeter of selenium
spheres and cells indicating the presence of a membrane structure also surrounding the
selenium spheres (vesicles). A loss over time of the ﬂuorescence surrounding cells, cell-
associated vesicles, and detached vesicles following formamide addition demonstrated
the vesicles were membrane-bound as formamide disrupts membranes. Gram stains of
Se vesicle-producing cultures over time showed that both cells and vesicles stained

Gram-positive, suggesting that the Gram-positive envelope also surrounded the

98

 

Al NaAl

     
  

O
0.01.000 2.000 1.000 2.000
Energy(lteV)

   

o ”10.56““ ‘
3 Distance on Line Scan (pm) 4

Figure 4.4. Representative SEM EDS and TEM EDS mapping of D. hafniense cells grown in 1 mM
sodium selenate. (A) SEM of Se exposed cell displaying buds and vesicles. (B) SEM backscatter image of
A. Arrows indicate regions of the image scanned by EDS. Arrow 1 points to the cell, while arrow 2 points

to an associated vesicle. (C) EDS of cell in B at portions indicated by arrows 1 and 2. (D) TEM of

99

hexagonal crystals (cr) and surrounding vesicles (v). (E) TEM backscatter image of hexagonal crystals.
(F) TEM of vesicles and cells. Line depicts the TEM EDS line scan through four vesicles and two cells
from points 3 to 4. (G) TEM EDS line scan of G illustrating changes in carbon, osmium, and selenium
ﬁ'om points 3 to 4 on the line. Locations of the vesicles (v) and cells (c) seen in (F) are represented on the

line.
vesicles. Gram-negative structures were not found.

The red ﬂuorescence of propidium iodide of the live/dead® stain provided evidence of a
once living structure. Active transport across a membrane was indicated by green
ﬂuorescence of SYTO 9 dye. Green ﬂuorescence after staining with both dyes was
observed in vesicles attached to cells or cell debris, however, unattached vesicles
possessed no ﬂuorescence indicating active transport was ﬁrnctional in the former. When
formamide was added, cells lost their green ﬂuorescence and only red ﬂuorescence was
visible indicative of nucleic acids. All vesicles lost green ﬂuorescence upon formamide
addition, but only cell-associated vesicles showed some red ﬂuorescence. This suggests
that unattached vesicles may still possess nucleic acids, which may readily leak out upon
membrane disruption with formamide. Attached vesicles may still exchange nucleic
acids with the cell, and thus their nucleotide content is more stable. But it can not be
ruled out that this is due to the reﬂecting red ﬂuorescence of cells and not due to vesicles.
D. hafniense cells not exposed to selenate, which do not produce vesicles, exhibited
similar cellular responses to the dyes and formamide, indicating that in general nucleic

acid and membrane function of the cell are not changing their reactivity with the dyes.

100

Since the vesicles stained with the live/dead® stain, nucleic acids must be present.
Whether the vesicles contained DNA and were perhaps mini-cells was investigated using
DAPI on duplicate samples taken at 20 time points. For all preparations, DAPI was only
detected in the cell and not in detached vesicles. Whether DAPI stained cell-associated
vesicles was unclear due to the ﬂuorescence intensity of the cells. As the DAPI
ﬂuorescence faded; however, ﬂuorescence was seen solely in cells and not in associated

vesicles regardless of proximity. Presence of RNA was later conﬁrmed (data not shown).

DISCUSSION
Oremland (2004) reported that some Gram-negative and Gram-positive bacteria are
capable of reducing both selenate and selenite, producing spheres of elemental selenium
as well as precipitates of reduced selenium inside or on the surface of bacterial cells.
Here we show that D. hafniense DCB-2, a Gram-positive bacterium, produced
extracellular vesicles in the presence of selenate and selenite, and EDS and XANES data
show that these vesicles contained the reduced selenium produced by the culture.
Bacterial growth is not inhibited by selenium until concentrations exceed 1 mM selenate
(Figure 4.1). No selenium was detected in the cells. Hence, vesicle production and

selenium fate are closely linked.

D. hafniense could not have reduced selenate and selenite by the sulfate reduction
pathway since it is incapable of sulfate reduction (Christiansen 1996). Although, D.
hafniense can reduce nitrate, nitrate-reduction, which is inducible by nitrate in this strain,

was not induced in this study. In previous studies, T. selenatis was found unable to

101

 

reduce selenite without the presence of nitrate (Milne 1994). Thus, the absence of nitrate
in the medium with D. hafniense likely eliminated the potential for facilitated or passive
reduction by the nitrate reduction pathway. Moreover, vesicles were not observed under
any of the control grth conditions or during grth with nitrate where nitrate reduction
was induced (data not shown). Thus, selenium reduction is likely by a non-sulfate and
non-nitrate pathway and in concordance with vesicle production, perhaps as a

detoxiﬁcation mechanism.

Previous work demonstrated that only one type of elemental selenium precipitate was
generated for each bacterial species capable of Se reduction (Oremland 2004). Here we
show that D. hafniense is capable of reducing selenate to two different colored elemental
selenium precipitates, red and black, with red more abundant than the black. Most pure
cultures of selenium-reducing bacteria, as well as a sediment slurry, produce only the red
precipitate (Oremland 1989). Selenium reduction rates were previously reported to
change due to differences in temperature, media, carbon source, and substrate (Lortie
1992); however, these factors were held constant in our experiments. The ability of D.
hafniense to also produce black selenium precipitate may be due to altered metabolism or

minor differences in trace elements or gaseous head space of the medium.

Spheres of elemental selenium were produced by D. hafniense during the reduction of
sodium selenate. Moreover, selenium-grown cells were elongated with an increasing
number of spherical selenium vesicles exterior to the cell over time. Interior selenium

vesicles were never found. As we determined these exterior selenium spheres were

102

membrane bound, the term vesicles was applied. While it had been speculated previously
by Oremland (2004) that selenium nanospheres were membrane bound, conclusive

evidence was missing until now.

Moreover, the vesicles produced outside the cell by D. hafniense were found to be bound
by a membrane that was attached to the cell by a ﬁlament. Unless selenium metal is
capable of interacting with crystal violet, which has not been shown previously, results
indicate the vesicles likely contain cell wall and are formed by a budding event. We
propose that selenium—reducing membrane proteins are present in regions of vesicle
formation since the ‘bead on a string’ vesicle formations were preferentially localized,
‘hot spots’, on the cell surface. The periplasmic associated selenate reductase of Thauera
selenatis and the presence of selenate reductase activity pointedly in the membrane of

Sulfospirillum barnesii give precedence to this hypothesis (Schréder 1997, Stolz 1997).

We tested whether the vesicles were created for the purposes of selenium sequestration
and reduction. The high abundance of only selenium spectroscopic patterns in the
vesicles and the lack of vesicles in selenium-ﬂee cultures suggest that vesicles resulted in
protecting the cell from selenium toxicity. The lack of DAPI and propidium iodide in
detached vesicles and in protruding attached vesicles indicates no ﬁrnctional DNA was
present in these vesicles. We speculate that DNA is also not present in vesicles closely
associated with the cell membrane as a fading DAPI ﬂuorescence signal was only
detected to emanate from the cell and not ﬁom any attached structure. The red

ﬂuorescence of detached vesicles stained with the lipophilic dye indicates the presence of

103

 

a vesicle membrane. SYTO-9 and propidium iodide of the live/dead® stain binds DNA
and RNA to produce green and red ﬂuorescence. Green ﬂuorescence was observed in
vesicles attached to cells or cell debris, but upon formamide addition the green
ﬂuorescence quickly became red ﬁom propidium iodide and then was lost entirely. Thus,
these vesicles likely contain some RNA, which upon disruption of the vesicle membrane
by formamide is lost to the surrounding matrix. Presence of active transport across the
vesicle membrane may also be supported by the presence of green ﬂuorescence inside
vesicles. The lack of ﬂuorescence in completely detached vesicles still possessing
membrane supports the breakdown of the RNA without replenishment. Thus, we
conclude the main purpose of the vesicles is to sequester and subsequently reduce

selenium.

Previous studies have found bacteria, such as Bacillus selenitireducens and
Selenihalanaerobacter shriftii, capable of producing both external and internal selenium
spheres. However, only extracellular selenium vesicles were produced by D. hafniense
cultures. Since selenium levels were below EDS detection limits in cells, it was not
sequestered in cells, but selenium molecules may still induce the vesicle response. Based
on our observations of D. hafniense growth in the presence of selenate, we propose a
model describing external vesicle formation and their fate. First, the cell detects
selenium in the environment and in response begins manufacturing a pool of membrane
associated selenium inﬂux pumps, reductases, and other specialized proteins. These
selenium-reducing proteins ﬂock to locales on the cell envelope which initiates bud

formation. In an effort to separate the potentially toxic selenium area, the cell completely

104

 

surrounds the selenium with membrane, forming a vesicle. To make room for ﬁrrther
reduction structures, a fully developed vesicle is then pushed away from the cell leaving a
thin ﬁlament connection between the vesicle and cell. A new bud forms at the site of the
ﬁlament and develops into a vesicle that has high concentrations of selenium-reducing
proteins. These ‘hot spots’ for vesicle formation result in the ‘beads on a string’
morphology of multiple attached vesicles. At some point the ﬁlament and associated
vesicles break away ﬁom the cell. The detached vesicles may continue to reduce
selenium as long as there is reductant and ﬁinctional reductase. The vesicles eventually
slough off the inactive membrane, and with proper conditions the selenium coalesces into

larger hexagonal selenium crystals.

Vesicles have been produced by other bacteria in response to a variety of stimuli, such as
iron, e. g. the production of internal magnetosomes for magnetotaxis (Gorby 1988). Here
the metal response appears to be a detoxiﬁcation reducing the selenium to an insoluble
form. Determining the amount of reduction 6 vesicle performs remains to be answered.
Do the vesicles remain active until their potential is exhausted, until a vesicle has reached
capacity, or until a vesicle is released from a ﬁlament connection to the cell? With
respect to the ﬁnal question, our confocal microscopy data of the live/dead® stain and
TEM EDS supports the continued presence of selenium-reducing activity in vesicles even
after separation from the cell. Moreover, if these vesicles are capable of reducing
selenium when detached from the cell, the potential remains for sustained and dispersed

remediation of selenium in the environment.

105

This research may prove of interest to those in selenium bioremediation for bioreduction
of mobile selenium in the environment. The nano-scale selenium vesicle product also
lends itself to nanotechnology. The electronics and glass industries in particular, which
require high purities of selenium and comprise 10% and 30% of the US selenium market,
respectively (Butterman 2004), may be interested in the potential of vesicles to reduce
selenium when detached from the microbial cell. As the global selenium demand is
projected to continually increase (Backman 2008), perhaps bioreduction will prove useﬁil

in selenium recovery and puriﬁcation.

ACKNOWLEDGEMENTS
Research was supported by the U.S. Department of Energy, Office of Science,
Environmental Remediation Science and Genomes to Life Programs. I am grateful to the
beamline staff at Argonne National Laboratory (Argonne, IL), in particular Dr. Mali
Balasubramanian at Paciﬁc Northwest Consortium Collaborative Access Team (PNC-
CAT) Sector 20-BM and Dr. Soma Chattopadhyay at Material Research Collaborative
Access Team (MR-CAT) Sector 10-ID for there assistance with the x-ray measurements.
MRCAT operations are supported by the Department of Energy and the MR—CAT
member institutions. Use of the Advanced Photon Source was supported by the U.S.
Department of Energy, Ofﬁce of Science, Ofﬁce of Basic Energy Sciences, under
Contract No DE-AC02-06CH11357. The author would especially like to thank Dr.
Shelly Kelly for assistance at Argonne National Laboratory performing beamline and

XRD studies, evaluating XANES data, and providing Figure 4.3.

106

REFERENCES

Backman, C-M. 2008. Global supply and demand of metals in the future. J. Tox. Env.
Health. 71:1244-1253.

Bloem, J. 1995. Fluorescent staining of microbes for total counts. In: Akkermans A.D.L.,
Elsas J.D., Bruijn F.J. (eds), Molecular Microbial Ecology Manual. Kluwer Academic
Press, Dordrecht, The Netherlands, pp. 1-12.

Butterman, W. C. and R. D. Brown Jr. 2004. Mineral commodity proﬁle: selenium.
US Geological Survey. Open ﬁle report 03-018.

Christiansen, N. and F. K. Ahring. 1996. Desulﬁtobacteriurn hafniense sp. nov.,
anaerobic, reductively dechlorinating bacterium. Int. J. Syst. Bacteriol. 46:442-448.

Diplock, A. T. 1981. Metabolic and Functional Defects in Selenium Deﬁciency. Philos.
Trans. R Soc. London Ser. B 294:105-117.

Elrashidi, M. A., and D. C. Adriano. 1987. Effect of redox potential and pH on
chemical speciation of inorganic selenium in soils. Sixth International Conference on
Heavy Metals in the Environment, New Orleans. Proceedings Vol. 1, pp. 107-109. CEP
Consultants Ltd., Edinburgh.

Gorby, Y. A., T. J. Bcvcridge, and R. P. Blakemore. 1988. Characterization of the
bacterial magnetosome membrane. J. Bacteriol. 170:834-841.

Harrison, G. 1., E. J. Laishley, and H. R. Krouse. 1980. Stable isotope ﬁactionation by
Clostridium pasteurianum. 3. effect of SeO32- on the physiology and associated sulfur
isotope fractionation during SO32— and S042 reductions. Can. J. Microbiol. 26:952-958.

Huber, R., M. Sacher, A. Vollmann, H. Huber, and D. Rose. 2000. Respiration of
arsenate and selenate by hyperthermophilic archaea. Syst. Appl. Microbiol. 23:305-14.
Lanthier, M., R. Villemur, F. Lepine, J-G. Bisaillon, and R. Beaudet. 2001.
Geographic distribution of Desulﬁtobacterium frappieri PCP-l and Desulﬁtobacterium
spp. in soils from the province of Quebec, Canada. FEMS Micro. Ecol. 36:185-191.

Loffler, F. E., R. A. Sanford, and J. M. Tiedje. 1996. Initial characterization of a
reductive dehalogenase ﬁ'om Desulﬁtobacterium chlororespirans C023. Appl. Environ.
Microbiol. 62:4982-4985.

Lortie, L., W. D. Gould, S. Rajan, R. G. L. McCready, and K.-J. Cheng. 1992.
Reduction of selenate and selenite to elemental selenium by a Pseudomonas stutzeri
isolate. Appl. Environ. Microbiol. 58:4042-4044.

Losi, M. E., and W. T. Frankenberger Jr. 1997. Bioremediation of selenium in soil and
water. Soil. Sci. 162:692-702.

107

 

Miller, T. L. and W. J. Wolin. 1974. A serum bottle modiﬁcation of the Hungate
technique for cultivating obligate anaerobes. Appl. Microbiol. 27:985-987.

Milne, J. B. 1998. The uptake and metabolism of inorganic selenium species. In: W. T.
Frankenberger and R. A. Engberg. (eds), Environmental Chemistry of Selenium. Marcel
Decker, New York, pp. 459-478.

Narasingarao, P., and M. M. Haggblom. 2007. Pelobacter seleniigenes sp. nov., a
selenate-respiring bacterium. Int. J. Syst. Evol. Microbiol 57:1937-1942.

NAS. 1976. Medical and biological effects of environmental pollutants - selenium.
National Academy of Sciences, Washington, DC.

Neal R. H., and G. Sposito. 1989. Selenate adsorption on alluvial soils. Soil Sci. Soc.
Am. J. 53:70-74.

Neal, R. H., G. Sposito, K. M. Holtzclaw, and S. J. Traina. 1987. Selenite adsorption
on alluvial soils: 1. Soil composition and pH effects. Soil Sci. Soc. Am. J. 51:1161-1165.

Newville, M. 2001. IFEFFIT: interactive XAF S analysis and FEFF ﬁtting. J. Synch. Rad.
8:322-324.

Oremland, R. S., J. T. Hollibaugh, A. S. Maest, T. S. Presser, L. G. Miller, C. W.
Culbertson. 1989. Selenate reduction to elemental selenium by anaerobic bacteria in
sediments and culture: biogeochemical signiﬁcance of a novel, sulfate independent
respiration. Appl. Environ. Microbiol. 55:2333-2343.

Oremland, R. S. J. Switzer Blum, C. W. Culbertson, P. T. Visscher, L. G. Miller, P.
Dowdle, and F. E. Strohmaier. 1994 Isolation, growth, and metabolism of an obligately
anaerobic, selenate-respiring bacterium, strain SES-3. Appl. Environ. Microbiol.
60:3011-3019.

Oremland, R. S., M. J. Herbel, J. Switzer Blum, S. Langley, T. J. Beveridge, P. M.
Ajayan, T. Sutto, A. V. Ellis, and S. Curran. 2004. Structural and spectral features of

selenium nanospheres produced by Se-respiring bacteria. Appl. Environ. Microbiol.
70:52-60.

Ravel, B., and M. Newville. 2005. ATHENA, ARTEMIS, HEPHAESTUS: data analysis
for X-ray absorption spectroscopy using IFEFFIT. J. Synch. Rad. 12:537-541.

Schriider, 1., S. Rech, T. Kralft, and J. M. Macy. 1997. Puriﬁcation and

characterization of the selenate reductase ﬁom Thauera selenatis. J. Biol. Chem.
272:23765-23768.

108

 

Segre, C. U., N. E. Leyarovska, L. D. Chapman, W. M. Lavender, P. W. Plag, A. S.
King, A. J. Kropf, B. A. Bunker, K. M. Kemner, P. Dutta, R. S. Druan, J. Kaduk.
2000. The MRCAT insertion device beamline at the advanced photon source.
Synchrotron Rad. Inst. CP521:419-422.

Shacklette, H. T., J. G. Boerngen, and J. R. Keith. 1974. Survey Circular 692. US
Department of Interior, Washington, DC.

Shamberger, R. J. 1981. Selenium in the environment. Sci Total Environ. 17:59-74.
Stolz, J. F., T. Gugliuzza, J. Switzer Blum, R. Oremland, and F. M. Murillo. 1997.
Differential cytochrome content and reductase activity in Geospirillum barnesii strain

SES-3. Arch. Microbiol. 167:1-5.

Stolz, J. F., P. Basu, and R. S. Oremland. 2002. Microbial transformation of elements:
the case of arsenic and selenium. Int. MicrobioL 5:201-207.

Tschech, A., and N. Pfennig. 1984. Grth yield increase linked to caffeate reduction in
Acetobacterium woodii. Arch. Microbiol. 137: 163-167.

Venable, J. H., and R. Coggeshall. 1965. A simpliﬁed lead citrate stain for use in
electron microscopy. J. Cell. Bio. 25:407-408.

Widdel, F., G.-W. Kohrin, and F. Mayer F. 1983. Studies on dissimilatory sulfate-
reducing bacteria that decompose fatty acids. Arch. Microbiol. 134:286-294.

Wolin, E. A., R. S. Wolfe, and M. J. Wolin. 1963. Viologen dye inhibition of methane
formation by Methanobacillus omelianskii. J. Bact. 87:993-998.

Young, V. R. 1981. Selenium: a case for its essentiality. N. Eng. J. Med. 304:1228-1230.

109

 

Chapter V

A Comparison of As(V), Fe(III) and Se(VI) Reduction by the Acetogens

Desulﬁtobacterium hafniense DCB-2 and Moorella thermoacetica

ABSTRACT
Two bacterial species, Desulﬁtobacterium hafniense DCB-2 and Moorella thermoacetica,
were compared for growth in minimal media with metal-reducing conditions. M.
thermoacetica was capable of growth in the presence of Cd(II), As(V) and Se(VI) but,
unlike D. hafniense, it did not grow with lactate and soluble Fe(III) as the sole electron
acceptor. It also reduced As(V to H1) and Se(VI to 0) similar to D. hafniense, although

less selenium was mineralized.

INTRODUCTION
Bacteria are capable of utilizing a large variety of electron acceptors for growth.
Acetogens possess the unique ability to ﬁx carbon to create biomass during which the
carbon serves as the electron acceptor for respiration. This process has been heavily

studied in thermophilic, Gram-positive Moorella thermoacetica (previously known as
Clostridium thermoaceticum) with both C02 and CO as carbon sources and electron
acceptors (Daniel 1990, Kerby 1983). Metals have never been tested as terminal electron

acceptors for M. thermoacetica; however, functional cytochromes and menaquinone have

been found as well as its ability to use nitrate, nitrite and thiosulfate as terminal electron

110

acceptors (Gottwald 1975, Drake 2004). Studies with metals in trace amounts for grth
determined that F e(II) was required for growth. Though Se(IV) was not required, it did

enhance production of formate dehydrogenase (Andreesen 1973).

Recently, microarray data of Desulﬁtobacterium hafniense DCB-2 has detected the
Wood-Ljungdahl carbon ﬁxation pathway to be highly expressed and functional with
genes similar to M. thermoacetica (Chapter 3). In fact, D. hafniense resides in the same
phylogenetic family as M. thermoacetica. D. hafniense and M. thermoacetica are both
spore-formers that were isolated ﬁ'om excrement and have been found in soils (Fontaine
1942, Madsen 1992, Lanthier 2004). D. hafniense is a Gram-positive, dehalogenator that
has been found to reduce many different electron acceptors, including metals. Both
species grow well fermenting pyruvate, but unlike D. hafniense, M. thermoacetica is
unable to ferment fumarate (Andreesen 1973, Christiansen 1996). The phylogenetic
similarities between these two organisms suggested the possibility that M. thermoacetica
might be capable of metal reductions as well We tested the ability of the long studied
acetogen (M. thermoacetica) to grow with and respire Cd(II), As(V), Se(VI), and Fe(III)

and compared these results to D. hafniense.

METHODS
Bacterial strains. M. thermoacetica and D. hafniense DCB-2 were originally grown in

modiﬁed DCB-l medium fermenting sodium pyruvate (pyruvate) with a head space of
C02:N2 (5:95) (Brumm 1988, Ldfﬂer 1996, Chapter 4). Cultures were incubated at

37°C for D. hafniense and 55°C for M. thermoacetica. These fermenting cultures served

111

 

as inocula for growth with metals. Culture purity was checked using microscopy and

grth on R2A plates supplemented with 10 mM pyruvate in an anoxic chamber of

H22N2 (4:96) (Chapter 4).

Experimental determination of growth. Growth was determined by cell counts, optical
density at 600 nm, HPLC of organic acids, and total biomass production (TOC) (Chapters
2, 3 and 4). Two replicate samples were sent for TOC analysis at UC Davis Stable
Isotope Facility (Davis, CA) following freeze-drying of combined sample pellets from
initial centriﬁigation and centrifugation of ethanol (2:3) precipitated sample supematants.
Controls consisted of media conditions without cells, cultures without electron acceptor
addition, and culture conditions with heat-killed cells. When testing M. thermoacetica
for grth with metals and metal reduction abilities, D. hafniense served as the positive

control.

Measurements of growth with metals. Both strains were grown with Cd(II), As(V),

Se(VI), and Fe(III) in two different media: modiﬁed DCB-l with 20 mM pyruvate and

Wolin vitamins (1 x) in C02:N2 (5:95) and carbonate-buffered freshwater medium (CBF)

with 20 mM sodium lactate (lactate) in C02:N2 (20:80) (Ibfﬂer 1996, Lovley 1996,

Wolin 1963, and Chapters 2 and 4). The electron acceptors were presented as 50 mM
ferric citrate, 1 mM sodium arsenate, and 1 mM sodium selenate. Mineralization of
selenium was measured by inductively coupled plasma mass spectrometry (ICP-MS)

(Chapter 2). Upon precipitate production in cultures, reduction of the electron acceptors

112

was determined using x-ray absorption spectroscopy (XAS) (Chapter 2).

RESULTS
The ability of M. thermoacetica to grow with Cd(II), As(V), Se(VI), and Fe(III) were
compared to cultures unexposed to metals (Table 5.1). Cultures grown in DCB-l with
As(V) produced 38% more biomass than unexposed cultures. Growth with As(V) as the
sole electron acceptor (CBF medium with lactate) was found with twice the amount of
cell material being produced compared to the amount attributable to inoculation (CBF
with lactate alone). However, this grth with As(V) in CBF had only 17% (5-fold less)
the amount of biomass as growth in DCB-l with As(V). Similar biomass production
levels were found in DCB-l medium with pyruvate for both Cd(II) and Se(VI), which
were about 33% less than pyruvate-fermenting cultures alone. Although previously
found to grow and affect precipitation of Cd(II) in complex media (Cunningham 1993),
M. thermoacetica demonstrated repressed growth in our study with Cd(II). Grth did

not occur using Fe(III) as a sole terminal electron acceptor in CBF medium with lactate.

The generation times for M. thermoacetica DCB-l medium with and without metal were
all within 30 nrin of each other (Table 5.1). Fermentative growth without metal had the
fastest generation time of 3.5 h, followed by As(V), Cd(II), and Se(VI). No grth was
detected when grown in CBF medium with lactate alone or with Fe(III) and lactate. The

change in time between DCB-l and CBF media with As(V) was nearly 3-fold. This

113

 

Table 5.1. Generation times and cell yields of M. thermoacetica and D. hafniense with metals

 

 

Metal concentration Carbon source Biomass produced Generation time

Metal (mM) Organism Medium (20 mM) (mg/L) (h)

- - M. thermoacetica DOB-1 pyruvate 34.3 3.47
Se (VI) 1 M. thermoacetica DOB-1 pyruvate 22.4 3.88
Cd (II) 1 M. themmacetica DOB-1 pyruvate 24.1 3.85
As (V) 1 M. thermoacetica DCB-1 pyruvate 47.2 3.63

- - D . hafniense DCB-1 pyruvate 15.1 8.95
Se (VI) 1 D. hafniense DCB-1 pyruvate 18.8 9.01
Cd (ll) 1 D. hafniense DOB-1 pymvate 15.1 8.9
As (V) 1 D. hafniense DOB-1 pyruvate 28.30 7.78

- - M. thermoacetica CBF lactate 3.19 unable to determine
Fe (III) 50 M. thermoacetica CBF lactate -** unable to determine
As (V) 1 M. thermoacetica CBF lactate 8.30 10.2

- - D. hafniense CBF lactate 0.295 unable to determine
Fe (III) 50 D. hafniense CBF lactate 1330 4.49
As (V) 1 D. hafniense CBF lactate 32.1 7.08

 

" Generation time could not be determined for these conditions due to lack of measureable cell growth
** Biomass was undetectable for M. thennoacetica in CBF with Fe (III) above background

along with the biomass production data suggest that growth is considerably more strained

with lactate and As(V) as the sole electron acceptor (Table 5.1).

This contrasted grth in DCB-l by D. hafniense which had an equal amount of biomass
production with Cd(II) (15.1 mg/L) and greater amount with Se(VI) (18.1 mg/L) than
pyruvate fermentation alone, while grth with As(V) produced 46% more biomass
(Table 5 .1). While M. thermoacetica, which had longer generation times in all cultures
with metals, D. hafniense had similar generation times with Cd(II) and Se(VI) to cultures
unexposed to metal. Moreover, the generation time for D. hafniense with As(V) was
faster in DCB-l medium than unexposed cultures. In CBF medium, D. hafniense had a
considerably faster generation time with As(V) (21% faster) and Fe(III) (2-fold increase)
and As(V) (21% faster), while M. thermoacetica grew 2.9-fold slower with As(V) and

had no growth with F e(III).

114

 

Table 5.2. Se (VI) mineralization ability of M. thermoacetica and D. hafniense

 

 

Metal Metal in Percent metal in solution
concentration solution relative to "no cells"
_(_)_rganism Medium (mM) (ppm) (ppm/ppm)
No cells 008-1 1 mM Se (Vl) 78.1 100
M. thermoacetica DCB-1 1 mM Se (VI) 16.3 20.9
D. hafniense DOB-1 1 mM Se (VI) 0.18 0.230

 

Reduction analyses using XAS for precipitates in cultures with Cd(II), As(V), or Se(IV)
showed Cd(II) was not reduced, while As(V to IE) and Se(VI to 0) were reduced (data
not shown). These results are similar to those found with D. hafniense. Cd(II) was
previously shown to be precipitated as CdS by M. thermoacetica; however, these
experiments were performed in undeﬁned medium (Cunningham 1993). Unlike D.
hafniense, which produced both red and black precipitates with Se(VI), M. thermoacetica
produced only red precipitates (Chapters 2 and 4). Extracellular vesicles on M.
thermoacetica were observed and were similar to those produced by D. hafniense, but
only a lO-fold increase in biomass above inoculation was measured for M. thermoacetica,
while there was a 50-fold increase for D. hafniense. Tests for removal of Se from
solution using ICP-MS found M. thermoacetica was able to mineralize 80% (Table 5.2).
This illustrates it was less effective at selenium mineralization than D. hafniense, which

removed 99.9% ﬁom solution.

Studies with M. thermoacetica in ferric citrate yielded no precipitate or biomass
production above levels attributable to inoculation. These results along with lack of
Fe(III) reduction indicate that M. thermoacetica was unable to grow with soluble Fe(III)

as the sole electron acceptor with lactate. This greatly contrasts the profuse biomass

115

production coupled to iron reduction observed with D. hafniense, and is surprising as the
formate dehydrogenase of M. thermoacetica possesses 30% amino acid homology to
corresponding subunits of the soluble F e(III) reductase of Geobacter sulfurreducens
(Kaufrnann 2001). The proﬁrse amount of biomass production by D. hafniense in

cultures with Fe(III) is likely attributable to a combination of carbon substrates

supporting growth, such as lactate with citrate, C02 ﬁxation, and acetate (Chapter 2).

DISCUSSION
M. thermoacetica was determined to have depressed growth with Cd(II) and Se(VI).
Extracellular vesicles similar to those of D. hafniense were seen when grown in the
presence of Se(VI), but grth was depressed. Vesicle production by M. thermoacetica
(Table 5.2) and in other studies with D. hafniense (Chapter 4), Sulfospirillum barnesii,
Sulfospirillum shriftii, and Bacillus selenitireducens (Oremland 2004) do suggest that this
method may be a general approach for cells to mineralize and detoxify selenium. Despite
having a higher optimum growth temperature, M. thermoacetica was unable to produce
the increase in biomass observed with D. hafniense in medium with As(V), Se(VI),
Cd(II), and Fe(III). It appeared that with the considerably slower generation time and
signiﬁcantly lower biomass yields of M. thermoacetica in CBF with As(V) and lactate,
fermentation with pyruvate is a preferred mode of growth. M. thermoacetica is typically
cultivated with yeast extract to produce greater than 30% more in biomass yields, despite
being able to ferment molecules such as oxalate and pyruvate (Daniel 1993). From our
study, using yeast extract, peptone, tomato juice, and other complex nutrients in medium

are unnecessary, which is important as they can affect metal reduction results.

116

 

This study discovered new metabolic capabilities in M. thermoacetica by comparing it to
known abilities in a species from the same family (D. hafniense) that possessed a
homologous major carbon pathway. Large similarities exist [e.g. acetogenic ability,
reduction of As(V) and Se(VI), and altered ligand binding with Cd(II)] as do stark
contrasts [e.g. grth yields with metals, and reduction of soluble Fe(III)]. These
comparisons can help to determine the role of these bacteria in the environment. The
ﬁndings may also have impacts for bioremediation of thermophilic environments,

particularly with Fe(III) and As(V) since As(V) adsorbs to F e(III).

ACKNOWLEDGEMENTS
Use of the Advanced Photon Source was supported by the U.S. Department of Energy,
Ofﬁce of Science, Ofﬁce of Basic Energy Sciences, under Contract No DE-AC02-
O6CH11357. The assistance of Dr. Shelly Kelly was greatly appreciated during beamline
studies and performing XAS analyses at Argonne National Laboratory (Argonne, IL).
Dr. Ed O’Loughlin of the Molecular Environmental Science Group in the Biosciences
Division at Argonne National Laboratory was gracious to collect ICP-MS data. The
technical assistance of Dr. Mali Balasubramanian at Paciﬁc Northwest Consortium
Collaborative Access Team (PNC-CAT) Sector 20-BM at Argonne National Laboratory

was also appreciated during beamline studies.

117

REFERENCES

Andreesen, J. R., A. Schaupp, C. Neurauter, A. Brown, and L. G. Ljungdahl. 1973.
Fermentation of glucose, fructose, and xylose by Clostridium thermoaceticum; effect of

metals on grth yield, enzymes, and the synthesis of acetate ﬁom C02. J. Bacteriol.
114:743-751.

Brumm, P. J. 1988. Fermentation of single and mixed substrates by the parent and an
acid-tolerant mutant strain of Clostridium thermoaceticum. Biotech. Bioeng. 32:444-450.

Christiansen, N. and F. K. Ahring. 1996. Desulﬁtobacterium hafniense sp. nov.,
anaerobic, reductively dechlorinating bacterium. Int. J. Syst. Bacteriol. 46:442-448.

Cunningham, D. P. and L. L. Lundie Jr. 1993. Precipitation of cadmium by
Clostridium thermoaceticum. Appl. Environ. Microbiol. 59:7-14.

Daniel, S. L. and H. L. Drake. 1990. Characterization of the H2- and CO-dependent

chemolithotrophic potentials of the acetogens Clostridium thermoaceticum and
Acetogenium kivui. J. Bacteriol 172:4464-4471.

Daniel, S. L. and H. L. Drake. 1993. Oxalate- and glyoxylate-dependent grth and
acetogenesis by Clostridium thermoaceticum. Appl. Environ Microbiol. 59:3062-3069.

Fontaine, F. E., W. H. Peterson, E. McCoy, M. J. Johnson, and G. J. Ritter. 1942. A
new type of glucose fermentation by Clostridium thermoaceticum. J. Bacteriol 43:701-
715.

Gottwald, M., J. R. Andreesen, J. Legal], and L. G. Ljungdahl. 1975. Presence of
cytochrome and menaquinone in Clostridium formicoaceticum and Clostridium
thermoaceticum. J. Bacteriol. 122:325-328.

Kaufmann, F. and D. R. Lovley. 2001. Isolation and characterization of a soluble
NADPH-dependent Fe(III) reductase from Geobacter sulfurreducens. J. Bacteriol.
183:4468-4476.

Kerby, R. and J. G. Zeikus. 1983. Grth of Clostridium thermoaceticum on H2/C02
or CO as energy source. Curr. Microbiol. 8:27-30.

Lanthier, M., R. Villemur, F. Lepine, J-G. Bisaillon, and R. Beaudet. 2001.
Geographic distribution of Desulﬁtobacterium frappieri PCP-1 and Desulﬁtobacterium
spp. in soils from the province of Quebec, Canada. FEMS Micro. Ecol. 36:185-191.

Lbfﬂer, F. E., R. A. Sanford, and J. M. Tiedje. 1996. Initial characterization of a

reductive dehalogenase ﬁ'om Desulﬁtobacterium chlororespirans C023. Appl Environ.
Microbiol. 62:4982-4985.

118

Lovley, D. R. and E. J. P. Phillips. 1988. Novel mode of microbial energy metabolism:
organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl.
Environ. Microbiol. 54: 1472-1480.

Madsen, T. and D. Licht. 1992. Isolation and characterization of an anaerobic
chlorophenol-transforming bacterium. Appl. Environ. Microbiol. 58:2874-2878.

Miller, T. L. and W. J. Wolin. 1974. A serum bottle modiﬁcation of the Hungate
technique for cultivating obligate anaerobes. Appl. Environ. Microbiol. 27:985-987.

Oremland, R. S., M. J. Herbel, J. Switzer Blum, S. Langley, T. J. Beveridge, P. M.
Ajayan, T. Sutto, A. V. Ellis, and S. Curran. 2004. Structural and spectral features of

selenium nanospheres produced by Se-respiring bacteria. Appl. Environ. Microbiol.
70:52-60.

Wolin, E. A., R. S. Wolfe, and M. J. Wolin. 1963. Viologen dye inhibition of methane
formation by Methanobacillus omelianskii. J. Bacteriol. 87:993-998.

119

 

Chapter VI

SUMMARY AND CONCLUSIONS

Microbial interactions with metals can alter mobility and toxicity of a metal in the
environment. Studies of these interactions can aid bioremediation of metal contaminated
sites. The studies conducted in this dissertation have focused on exploring this
bioremediation potential in the dehalogenator Desulﬁtobacterium hafniense DCB-2. The
main objective of this research was to investigate the potential of using D. hafniense as
the ﬁrst model Gram-positive organism for dissimilatory metal reduction (Chapter 1).
Having been found in metal-contarninated soils, it was selected to further understand
Gram-positive microbial interactions with metals as previous metal reduction research
has primarily focused on Gram-negative bacteria. These are not the ﬁrst studies to report
metal-reducing abilities of this bacterium; however, they are the most extensive and
exhaustive proving direct interactions with the metals. This is in large part due to the
detail in methods and controls, which we developed to take into account both passive and
active interactions, true perpetuation of cultures in experimental conditions, and limiting

outside compounding factors in measurements.

Three speciﬁc objectives were addressed during this study (Chapter 1). The ﬁrst
objective, “to determine the spectrum of metals reduced and which metals support grth

by respiratory coupling of metal reduction to energy production”, was addressed in

120

 

Chapter 2. Growth of D. hafniense in the presence of metals was examined both under
pyruvate fermentation conditions and respiration conditions with lactate. Nine metals
were provided as terminal electron acceptors, including: As(V), Cd(II), Co(III), Cr(VI),
Cu(II), Fe(III), Ni(II), Se(VI), and U(VI). During pyruvate fermentation, depressed
growth from metal toxicity effects were observed at the following metal concentrations:
500 11M Cr(VI), 1 mM Cd(II), 1 mM Ni(II), 1 mM Cu(II), 2.5 mM and above Se(VI), and
all concentrations tested of U(VI). Altered binding of ligands to metals was observed
with cells in As(V), Cd(II), Co(III), Cu(II), Fe(III), Ni(II), Se(VI), and U(VI) (Chapter 2).
Respiration of metals with lactate occurred with As(V), Fe(III), and weakly with Cu(II)

using nitrate as a positive respiration control.

Four of the nine metals tested were found to be conclusively reduced by D. hafniense
during valence state analysis at the Advanced Photon Source synchrotron at Argonne
National Laboratory (Argonne, IL) (Chapter 2). Fe(III) was reduced when presented both
in a soluble form (ferric citrate) and as a solid (ferric oxide). Cr(VI) was reduced as well,
but the reaction was so fast upon introduction to cultures that it was unable to be
determined if cells were directly involved. When XANES analysis was examined for
Cd(II), Co(III), Cu(H), and Ni(II), the metals were found not to be reduced, but rather
only to have alternate ligand binding partners. The potential for incomplete reduction, or
reduction of some portion of the metal introduced, below detection limits of the
synchrotron exists, as growth was present using Cu(II) as a sole electron acceptor with
lactate and signiﬁcant loss of solubility was present for Co(III). Moreover, the color

changes observed for Co(III) cultures are typical of previous reports of its reduction by

121

 

bacteria. Alternate media, carbon sources, or metal complexes introduced (e.g. using
Co(HI) EDTA rather than Co(III) chloride) may facilitate conclusive results to determine

the reduction of Co(HI), Cr(VI), and Cu(II) by D. hafniense.

The second speciﬁc objective was “to determine whether there are differences in the cell
morphology caused by reduction of different metals, and if speciﬁc adaptations occurred
to aid metal reduction” (Chapter 1). This research was conducted using Fe(III), Se(VI),
and U(VI), which were all conclusively reduced (Chapters 3 and 4). Using both bright-
ﬁeld and electron microscopy, unusual cell morphologies (that reﬂected gene expression)
were observed in the presence of respiration with Fe(III) and fermentation with U(VI)
and Se(VI) (Chapters 3 and 4). With 50 mM Fe(III), D. hafniense presented itself as
chains of cells, while with 0.5 mM U(VI) and 1 mM Se(VI) the cell was elongated
signiﬁcantly as compared to growth with pyruvate fermentation alone. These chains of
cells observed with Fe(III) and lactate were shorter than pyruvate fermenting cells, and
were surrounded by large amounts of exopolysaccharides, which bound the reduced
F e(II) (Chapter 3). The shorter cell, and thus larger surface area of the cell indicates that
D. hafniense may be rapidly dividing and likely thriving while reducing Fe(III), which is
supported by Fe(III)-reducing cultures having roughly half the generation time as
pyruvate-fermenting cultures (Chapter 2). In contrast, cultures grown with U(VI) had a
signiﬁcantly slower generation time than with pyruvate fermentation alone (Chapter 2).
An elongation of the cell was observed and is believed to occur due to cells resisting

potential metal toxicity (Chapter 3).

122

 

Extracellular vesicles were observed both free- ﬂoating and cell-associated in the presence
of Se(VI) (Chapter 4). Se(VI)-reducing cultures produced two different colors (red and
black) of reduced selenium, which differed by the ordering of their crystal structures.
Spectroscopy techniques coupled to microscopy revealed that Se was being sequestered
within these vesicles and not within the cell. Osmium, lipophilic, and 1ive/dead® staining
provided strong evidence that the vesicles were membrane-bound (Chapter 4). The
vesicles were positive for Gram-staining, signifying vesicles possessed cell wall material.
Live/dead® stain, which has a strong binding speciﬁcity for DNA and RNA, ﬁrrther
implicated the presence of nucleic acids within vesicles. DAPI did not stain vesicles
demonstrating DNA was not present in the vesicles. Puriﬁed vesicles were shown to
contain RNA by gel electrophoresis (with and without nucleases) and by UV absorption
analysis. Time-course studies revealed the formation of vesicles over time and their
connection to the cell via a ﬁlament. We have proposed a model for vesicle development
as it seems to occur in stages (Chapter 4). This morphological shift to produce vesicles
and the reduction ability of attached and detached vesicles should be a primary focus of

further research on Se(VI) reduction in D. hafniense.

The third research objective was “to identify genes and major pathways involved with
Fe(III), Se(VI), and U(VI) reduction” when compared to pyruvate fermentation using
microarray data (Chapter 1). During Fe(III) reduction, large amounts of carbon were
diverted to gluconeogenesis, which corresponds to the observed cell morphology with
abundant biomass production (Chapter 3). Two different nitrogenase operons were found

to be up-regulated during respiration with one functioning primarily during Fe(III)

123

T

reduction and the other during nitrate reduction (Chapter 3). Several ﬁrmarate and
DMSO reductases as well as cytochromes were also highly up-regulated, which indicates
their involvement in Fe(III) reduction. U(VI)-reducing cultures had many similar genes
expressed as during Fe(III) reduction (Chapter 3). However, U(VI) reduction appeared to
highly stress cells, up-regulating DNA repair mechanisms and cell wall biosynthesis
compared to pyruvate-fermenting cultures. The possibility of cells attempting to evade
high concentrations of uranium exists as chemotaxis and ﬂagellar genes were highly
expressed as well. This chemotactic response to uranium could be studied further in real-

tirne using controlled cell chambers with microscopy. Nitrate-reducing cultures

unexpectedly expressed gene groups annotated as assimilatory sulfate reduction and N2

ﬁxation-like function by the Nifl operon. Further investigation, possibly by knock-out
mutation, must be explored to conclude the role of these expression study ﬁndings with
Fe(III), U(VI), and nitrate. Many questions still surround Se(VI) reduction, as most of
the genes identiﬁed as highly expressed compared to pyruvate fermentation possessed
unknown function (Chapter 3). None of the known Se(VI)-reducing genes found in other
bacteria were up-regulated in D. hafniense. While some genes potentially involved in
Se(VI) reduction were found, we think focus should be directed to identifying the
proteins of the selenium-containing vesicles for which we have developed a protocol

(Appendix C).

Genetic studies of cultures grown with these reduced metals by rrricroarrays revealed a
highly complex use of carbon. One highly expressed gene (compared to pyruvate-

fermenting conditions) was carbon monoxide dehydrogenase, which is used by the

124

I .

acetyl-CoA reduction pathway, or Wood-Ljungdahl pathway (Chapter 3). Acetogenic

grth by D. hafniense was conﬁrmed with H2:CO:N2 (30:30:40), H2:COZ:N2

(30:30:40), and CO:C02:N2 (30:30:40) (Chapter 3). This pathway was previously shown

to be active in the phylogenetically related acetogen, Moorella thermoacetica, and it was

therefore used as a positive control for the experiment. The use of this C02 ﬁxation

pathway is likely to account for the need to shuttle excess electrons produced by the
reduction of the metals. This newly discovered metabolic ability of D. hafniense to grow
acetogenically solely using the Wood-Ljungdahl pathway also has profound implications

for future research on this organism.

In related studies to this work, it was discovered that D. hafniense can form bioﬁlms,
which can help establish and maintain it within the environment (a characteristic
desirable for bioremediation) (Appendix A). These bioﬁlms were observed to form on
two distinct surfaces (silica and activated carbon beads) under both fermentation
conditions with pyruvate and respiration conditions with Fe(III) and lactate. While this
observation requires future quantiﬁcation, it introduces the possibility of bioﬁlm research

with D. hafniense.

To exemplify the role of D. hafniense as a model Gram-positive metal-reducer, metal
reduction experiments were also performed with M. thermoacetica. These experiments

showed that Se(VI) and As(V) were reduced to similar products made by D. hafniense.

Both bacteria also formed Cd(II) sulﬁde when presented with CdCl, but remained

125

 

 

 

unchanged in controls without live cells. This had previously been found with M.
thermoacetica, and further illustrates the similar reactions both bacteria are performing in
the presence of metals. In strong contrast, however, soluble F e(III) was not reduced by

M. thermoacetica when presented with lactate (Chapter 5).

The guiding hypothesis for this research was that D. hafniense is capable of reducing
metals for energy gain through dissimilatory, including respiratory, metal reduction. I
have found that this hypothesis is true for the metals As(V), Fe(III), Se(VI), and U(VI),
and possibly true for Co(III), Cr(VI), and Cu(II). Additionally, I hypothesized that D.
hafniense is capable of reducing the same metals as those reduced by model, Gram-
negative, metal-reducing organisms Geobacter sulfurreducens and Shewanella oneidensis
MR-l. Of the nine metals tested, D. hafniense was found to reduce the same metals as
both Gram-negative model organisms. Moreover, Cr(VI) was reduced by both Gram-
negative organisms, and the likelihood that D. hafniense reduces it is very high. Co(IH)
reduction was found with S. oneidensis MR-l and G. sulfurreducens when presented as
Co(III) EDTA, and judging by our color reaction and growth studies, D. hafniense may
also reduce Co(III) EDTA effectively. As was found with S. oneidensis MR—l, D.
hafniense is able to convert the ligand binding to cadmium and copper. However, further

research must be done to conclude how and why this is occurring.

In this work, examination of the metal reducing abilities of D. hafniense was the main
objective, which was facilitated by the recently released genome sequence. This

organism is of particular interest to bioremediation as a spore-former capable of

126

 

halorespiration and respiration of metals, which we have coined metalorespiration.

Beyond the main objective we have discovered a previously unknown acetogenic ability

of D. hafniense with both CO and C02 (Chapter 3). The capability of this bacterium to

form single species bioﬁlm was also found (Appendix A). These two signiﬁcant ﬁndings

aid in the potential importance it can play in bioremediation.

However, the most valuable outcome from this research is that D. hafniense is now
poised as the only model, metal-reducing, Gram-positive, organism to which we have
shown other Gram-positive bacteria can effectively be compared (Chapter 5). Studies
with M. thermoacetica show how research on D. hafniense can inform us about the
potential biology in related bacterial species. This investigation also provides a necessary
reference for interpretation of systems microbiology. Future research should continue to
delve deeper into the metal-reducing abilities of D. hafniense, such as by discovering the
proteins directly involved when it performs metal-reduction, and test its effectiveness in
situ during metal reduction bioremediation. By studying D. hafniense and its interactions
with metals, D. hafniense is poised to become the ﬁrst Gram-positive dissimilatory metal-

reducing model organism.

127

 

Appendix A

Bioﬁlm-Forming Capacities of Desulﬁtobacteriurn hafniense DCB-Z

Desulﬁtobacterium haﬁriense DCB-2 was tested for bioﬁhn formation under conditions
of fermentation (modiﬁed DCB-l medium) and respiration (ferric citrate medium) using
two different surfaces (Dupont and Siran beads) (Ldfﬂer 1996, Lovley 1988, and
Chapters 2 and 4). Dupont beads are plastic coated activated carbon spheres ranging
from 3 to 5 mm in diameter (Figure A.1A). Siran beads are extremely rough and pitted
silica glass beads ranging ﬁ'om 2 to 3 mm in diameter (Figure A.1B and C). The beads
were laid 2.5 cm deep with a 1 cm cover of medium in ﬂee-standing serum tubes. The
medium was replaced every 2 to 3 days using a 22 G spinal needle to avoid disturbing

bacteria attaching to the beads and forming bio ﬁlms.

D. hafniense growth and bioﬁlm formation were observed under all conditions
throughout 25 medium transfers (Figure A2). The bioﬁhn settled in amongst the beads
unless disturbed, although the bioﬁlm consistency and abundance varied for each
condition (Figure A.2A and B). Ferric citrate medium demonstrated characteristic
reduction coloration (burnt orange to brown to olive to yellow) with cultures on both
Dupont and Siran beads (Figure A.2C and D). The largest accumulated bioﬁlrn occurred
in ferric citrate medium on Siran beads (Fig A.2E). Cell morphology varied similarly

depending on media and substrate. Bacteria were easily retrieved in quantity ﬁ'om

128

 

 

Figure A.1. Bright ﬁeld microscopic images (lOOX magniﬁcation) of beads. (A) Cross-section of a
Dupont activated carbon bead showing the thin white plastic surface coating. (B). Side view of a Siran
glass bead illustrating its rough surface. (C) Top view of a Siran glass bead showing its deeply pitted

surface.

Dupont beads. These pelagic bacteria were clustered in streptobacilli formations not
typical of D. hafniense during growth fermenting pyruvate. Pelagic bacteria were
not readily recovered from the Siran beads. However, those recovered were in
streptobacilli clusters or isolated. Removal of D. hafniense bioﬁlm from the highly pitted
Siran beads was signiﬁcantly more difﬁcult than retrieving bioﬁlm samples from the
smoother Dupont beads. This difﬁcult retrieval may have caused some bacteria to break
out of clusters to appear as isolates. The Dupont beads may not have provided as
intricate a physical support as the Siran beads. Furthermore, ferric citrate grown D.
hafniense were signiﬁcantly smaller than normal pyruvate fermenting cells despite
pelagic or sissile growth. Scanning electron microscopy has conﬁrmed the streptobacilli

shape of D. hafniense cells grown in ferric citrate and noted their smaller size

129

 

Figure A.2. Bioﬁlm cultures with Dupont and Siran beads aﬂer 25 medium transfers. (A) Undisturbed
(left) and shaken (right) cultures on Siran beads in DCB-l with pyruvate. (B). Undisturbed (left) and
shaken (right) cultures on Dupont beads in DCB-l with pyruvate. (C) Culture grown on Dupont beads
prior to medium transfer (left) and following transfer (right) of ﬂesh ferric citrate medium. (D) Culture
grown on Siran beads prior to medium transfer (left) and following transfer of ﬂesh ferric citrate medium.

(E) Above view of the bioﬁlm produced by a culture grown on Siran beads in ferric citrate medium.

(Chapter 3). Energy-dispersive spectroscopy (EDS) has shown that iron appears to be
sequestered in the extracellular polymeric substance and not the D. hafniense cell during

its reduction (Chapter 3).

130

 

These results conclusively show D. hafniense is able to produce bioﬁhn, which is more
proﬁise under conditions of F e(III) reduction as the sole terminal electron acceptor.
Implications for enhanced bioremediation by bacteria due to bio ﬁlm formation have been
proposed (Peacock 2004 and Singh 2006). As D. hafniense is a dehalogenator, metal
reducer, and acetogen, the effect on augmenting the bioremediation of toxic compounds
could be substantial. Further research on D. hafniense should investigate changes in its
gene expression and the efﬁciency of metal reduction or dehalogenation during growth as

a bio ﬁlm.

131

 

REFERENCES

Loffler, F. E., R. A. Sanford, and J. M. Tiedje. 1996. Initial characterization of a
reductive dehalogenase from Desulﬁtobacterium chlororespirans C023. Appl. Environ.
Microbiol. 62:4982-4985.

Lovley, D. R. and E. J. P. Phillips. 1988. Novel mode of microbial energy metabolism:
organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl.
Environ. Microbiol. 54:1472-1480.

Peacock, A. D., Y-J. Chang, J. D. Istok, L. Krumholz, R. Geyer, B. Kinsall, D.
Watson, K. L. Sublette, and D. C. White. 2004. Utilization of microbial bioﬁhns as
monitors of bioremediation. Micro. Ecol. 47:284-292.

Singh, R., D. Paul, and R. K. Jain. 2006. Bioﬁlms: implications in bioremediation.
Trends Microbio]. 14:389-397. '

132

"T'

Appendix B

Minimal Medium and Culturing Methods for Desulﬁtobacterium hafniense DCB-2

Here, the two minimal media used for culturing Desulﬁtobacterium hafniense DCB-2 are
provided along with descriptions of materials used and how to grow cultures. All media
are designed for making 1 l, but can be adapted for more. The ﬁrst medium, a modiﬁed
DCB-l medium was used with pyruvate in my studies to provide fermentative growth
conditions (Lbffler 1996, Chapter 4). The second medium, carbonate-buffered
freshwater medium, does not mimic freshwater but is modiﬁed from a medium described
by Lovley (1988). This medium was used with lactate and an additional electron
acceptor (i.e. metals or nitrate) for studies during anaerobic respiration. Culturing
methods, described here, can also be adapted for use with other anaerobic organisms.
While these methods describe the use of butyl rubber stoppers, alternate culturing
conditions (e. g. dechlorination studies) may require the use of different stoppers (e. g.
Teﬂon-coated rubber stoppers) to avoid possible absorption of chemicals with the
stoppers. Balch (1979) provides further information for Hungate system construction and
assistance with modifying these methods for other anaerobic organisms (i.e.

methano gens).

133

 

Modiﬁed DCB-l medium (DCB-l)

(modiﬁed from Léffler 1996)

MilliQ H20 964.6 ml
10% NaCl 10 ml
10% MgClz 5 ml
10% KH2P04 2 ml
10% KC] 3 ml
10% NH4C1 3 ml
10% CaClz 150 pl
SeWo (Selenite tungstate solution) 1 ml
trace elements (SL-10) 1 ml
0.1 % resazurin 250 pl
1 .M HEPES, pH 7.0 10 ml

Directions for DCB-I medium:
1. Boil all ingredients under reﬂux for 10 min with 95:5 of szCOz (hence forth referred

to as under gas). Be certain to have boiling chips in your media. Two sterilized mini stir
bars work ﬁne. A water bath may not be necessary during boiling under reﬂux; heat may
be directly applied to most round bottom ﬂasks.

2. Stand at room temperature (10-30 min) to cool glass prior to putting in ice bath to
bring to room temperature (~15 min) all while under gas.

3. Stir slowly under gas while adding 0.16 g cysteine or (0.06 g cysteine and 0.12 g
NazS)
4. Continue stirring slowly under gas while adding 2.52 g NaHCO3

5. Measure pH (~7.15). This is easiest with a portable pH meter. Adjust pH with ~3
pellets of NaOH if necessary.
6. Under gas dispense media and cap bottles (e. g. with butyl rubber stoppers and

134

aluminum tear-off seals for serum bottles).
7. Autoclave DCB-l media for 40 min on dry/ fast exhaust for serum bottles.
*A liquid/ slow exhaust may be preferred for bottle integrity of containers over 100 ml.

8. Store at room temperature.

SeWo (selenite tungstate solution)

(from Tschech 1984)

NaOH 0.5 g
Na28603 X 5 H20 3 mg
Na2W04 x 2 H20 4 mg
MilliQ H20 1 liter

Final pH of the solution should be 6.9 - 7.1.

Store at room temperature.

135

Trace element solution SL-lO:

(ﬁom Widdel 1983)

HCl (25%; 7.7 M) 10.00 ml
FeClz x 4 H20 1.50 g
ZnClz 70.00 mg
MnC12 x 4 H20 100.00 mg
H3BO3 6.00 mg
CoC12 x 6 H20 190.00 mg
CuC12 x 2 H20 2.00 mg
NiClz x 6 H20 24.00 mg
NazMoO4 x 2 H20 36.00 mg
MilliQ H20 990.00 m1

First dissolve FeC12 in the HCl, and dilute in water. Add and dissolve the other salts in

order. Finally, bring volume up to 1 1. Store at 4°C away from light.

136

Carbonate-Buffered Freshwater medium (CBF)
(modiﬁed ﬁ'om Lovley 1988)

MilliQ H20 981.5 ml

10% NH4C1 (or 0.25 g) 2.5 ml

10% KC] (or 0.1 g) 1 ml

Mineral Mix 10 ml
NaH2P04 x H20 0.6 g
NaHC03 2.5 g

Resazurin (depending on use)

Directions for CBF medium:

1. Bring water to boil using reﬂux under 80:20 of N2:C02 (hence forth referred to as

gas).

2. Allow water to boil 10 min per liter of medium made. Cool water under gas.

3. Once water is at room temperature, add NH4C1, KCl, and the Mineral Mix.
4. Dissolve the NaH2P04 X H20 in about 3 ml of MilliQ water. Filter this and add the

dissolved NaHzPO4 8 H20 to the medium.
5. To achieve the optimum pH of 6.8 - 7.0, add about 5 ml of MilliQ water to the
NaHCO3 and begin to dissolve it. While monitoring the pH of the medium, add the

NaHCO3 (ﬁltering it ﬁrst) until the optimum pH is maintained.

6. Under gas dispense media and cap bottles (e. g. with butyl rubber stoppers and
aluminum tear-off seals for serum bottles).

7. Autoclave the medium for at least 20 min on dry/ fast exhaust for serum bottles.

*A liquid/ slow exhaust may be preferred for bottle integrity of containers over 100 ml.

8. Store at room temperature.

*For 50 mM ferric citrate medium, ferric citrate is ﬁrst dissolved in the water. This is

137

 

then boiled under reﬂux with all components except NaHzPO4 X H20 and NaHCO3,

which are slowly added to cooled medium to adjust the pH. (One may also need to use
10 N NaOH to achieve optimum pH.) Care must be taken to not exceed the optimum pH

as Fe can fall out of solution.

138

Mineral Mix for CBF medium

(from Lovley 1988)

Nitrilotriacetic acid 1.5 g
FeSO4 X 7 H20 100 mg
ZnC12 130 mg
MnSO4 X H20 500 mg
H3BO3 10 mg
CoClz X 6 H20 100 m8
CuSO4 X 5 H20 10 mg
NiC12 X 6 H20 24 mg
NazMoO4 25 mg
MgSO4 X 7 H20 3 g
NaCl 1 g
CaClz X 2 H20 100 mg
AlK(SO4)2 X 12 H20 10 mg
Na2WO4 X 2 H20 25 mg
MilliQ H20 990 ml

Nitrilotriacetic acid (NTA) must be added and dissolved ﬁrst to 900 ml MilliQ water. As
NTA can be used as a carbon source by some microbes, it can be substituted with HCl.
When using HCl as a substitute for NTA, use only enough to keep the metals dissolved.
Add the ingredients in the order they are listed above. Do not add the next ingredient

until the previous ingredient has fully dissolved. Bring the ﬁnal volume to 1 l with

139

MilliQ water after all ingredients have been added. Store at 4°C away from light.

140

 

Wolin vitamin solution: (1000x)
(ﬁ’om Wolin 1963)

Biotin 2 mg
Folic acid 2 mg
Pyridoxine-HCI 10 mg
Thiamine-HCl x 2 H20 5 mg
Riboﬂavin 5 mg
Nicotinic acid 5 mg
D-Ca-pantothenate 5 mg
Vitamin B12 0.1 mg
p-Aminobenzoic acid 5 mg
Lipoic acid 5 mg
MilliQ H20 1 liter

Add all ingredients in order to 900 ml of MilliQ water. Wait for the previous ingredient
to dissolve prior to adding the next ingredient. Once all ingredients have dissolved, bring
the ﬁnal volume to 1 l. Dispense the 1 lof 1000X vitamin solution into tubes and store at
-20°C away from light. This maintains vitamin integrity and allows you to unthaw only

as necessary.

Prior to use, dilute the 1000X vitamin solution 1/ 10, ﬁlter it through a 0.22 pm ﬁlter, and

then gas it (100X ﬁnal concentration). Vitamins should not be autoclaved. Store at 4°C

away ﬁ'om light.

141

Carbon sources for Desulﬁtobacterium haﬁriense

The stock carbon source for DCB-l medium it is 2 M sodium pyruvate (pyruvate), and
for CBF medium it is 2 M sodium lactate (lactate). These carbon source stocks should be
ﬁltered through a 0.22 pm ﬁlter, and then gassed out to become anaerobic (described
below). The ﬁnal concentration desired for carbon stocks is 100X (2 M) for easy
dispensing to 20 mM in culture bottles. The carbon stocks are not autoclaved. Store
carbon stocks at 4°C. Note: grth in DCB-l medium with pyruvate is fermentative,

while respiration conditions are provided in CBF medium with lactate.

Gassing out items that can not be autoclaved.
(i.e. vitamins and carbon sources).
1. Filter sterilize the liquid (i.e. dissolved nutrient) through a 0.22 pm ﬁlter into a clean

sterile serum bottle.

2. Gas the solution at least 10 min for 100 ml of liquid at 95:5 of N22C02 (or the desired

gas ratio) being sure that the gas needle is fully submerged and bubbling through the
liquid at all times. The length of time will vary on the ﬂow rate of your gas; however, to
alleviate this problem I typically gas 15-20 min per 100 ml. The gas is considered
sterilized as it passes through the hot copper catalyst of the Hungate gas station.

3. Cap the bottles (butyl rubber stoppers and crimp closed with aluminum seals), and '

these newly anaerobic liquids are ready for use.

142

Making anaerobic gas bottles
Bottles containing only sterile anaerobic gas are made for transfers of any anaerobic
material not performed in an anaerobic chamber. (For use see below “Culturing D.

hafniense in DCB-l Medium”).
1. Put clean gas canulas/ needles running gas at 95:5 of N2:C02 (or desired gas ratio) into

a clean, sterile serum bottle to be capped.

2. Allow gas to circulate about 3 min.

3. Holding the butyl rubber stopper over the top and canula try to create a back pressure
of gas in the bottle for about 30 sec with the canula about half an inch remaining in the
bottle.

4. In one swift move, plunge in the stopper while removing the canula (maintain gas
pressure).

5. Crimp on aluminum seal and autoclave the bottles 40 min on dry/ fast exhaust and use.
6. Store at room temperature.

7. Dispose of bottle after about ﬁfteen 1 m1 syringe withdrawals of gas from the bottle or

until it feels difﬁcult to draw gas ﬁ‘om the bottle (whichever comes ﬁrst).

143

 

Feeding Desulﬁtobacterium hafniense

As Desulﬁtobacterium hafniense is a spore-former, old cultures may be revitalized by
adding vitamins and a source of carbon. Cultures at 37°C may also be maintained in a
state of constant growth by feeding once every 5-7 days with a carbon source and once a
month with vitamins. To do this, simply follow steps 2 through 16 (skipping step 15) in
“Culturing Desulﬁtobacterium hafniense in DCB-l medium” (below); however, instead
of using fresh DCB-l medium, you will add pyruvate and vitamins to your culture you

wish to feed.

144

 

Culturing Desulﬁtobacterium hafniense in DCB-l medium

You will need the following: DCB-l medium, 2 M pyruvate, 100x vitamins, anaerobic
gas, and stock culture (your inocuhrm). These should all be anaerobic prior to starting.
Other materials include (per 100 ml culturing bottle): three 1 ml syringes, three 22 G
needles, ethanol, ﬂame source, and weighing spatula (to open aluminum crimp tops)

(Figure B.1A).

 

Figure B.1. Anaerobic inoculation process. (A) The materials necessary for culturing into sterile

medium. Only DCB-l medium is pictured with an anaerobic gas bottle and a bottle dropper of ethanol.
The vitamin and carbon source stock bottles are not shown. (B) The position of the hand just prior to
removal of the needle from the anaerobic gas bottle is shown. The index ﬁnger is on the syringe plunger
ready to expel gas to prevent oxygen from coming in the needle. (C) The expelling of the needle contents
into a serum bottle. (D) The mixing and measuring of the liquid from a serum bottle for subsequent

transfer. The serum bottle is held upside down to obtain the liquid.

Assume inoculating a 100 ml bottle of DCB-l medium.
1. To inoculate a sealed serum bottle of DCB-l medium, tear back a portion of each seal

using the perforated circular top ﬁrrthest from the tear off edge to expose the butyl rubber

145

stopper. This is where you will enter the bottle with your needle. (I use a weighing
spatula to open them)

2. Put three to four drops of ethanol on the tops of all bottles to be involved in the
inoculation process, and ﬂame the ethanol off to sterilize the tops. For our purposes we
will start with anaerobic gas, 2 M pyruvate and DCB-l medium.

3. Repeat step 2 two times for a total of 3 ethanol burnings on each bottle top.

4. During your last time of ﬂaming the tops (tops should be hot by now), unwrap the
syringe barrel and quickly attach the needle (touching no part of the barrel and ﬁtting nor
the needle). (If in doubt about having touched the needle at all or syringe barrel junction
then discard the unit into a sharps container. Do not reuse it.)

5. Just as ethanol ﬂames are extinguished, plunge needle into anaerobic gas bottle and
take up gas (Fig B.lB). (Anaerobic gas is always the ﬁrst bottle in order to exchange the
oxygen item the syringe barrel.)

6. Remove the needle and slowly void the gas into the air while plunging into the next
bottle’s stopper (in this case, 2 M pyruvate) (Figure B.1B and C).

7. Expel all gas, ﬂip the bottle over and pull out pyruvate while shaking bottle to mix
contents (Fig B. 1D).

8. Flick the syringe barrel to get all bubbles to go to the needle and expel bubbles back
into the pyruvate bottle.

9. Measure out lml of 2 M pyruvate (1/ 100 dilution).

10. Remove the needle ﬁom the pyruvate bottle (do not expel the liquid as you did the
gas) and plunge the needle into the DCB-l medium bottle’s stopper. (Be sure the stopper

is still hot or it should be reﬂamed before removing the needle ﬁ'om the pyruvate bottle.)

146

ll. Expel the needle contents into the DCB-l medium bottle and remove the needle
(Figure B.1C).

12. Discard the used needle and syringe barrel into a sharps container.

13. Reﬂarne the tops of all bottles used with ethanol.

14. Repeat step 2 through 13 using 100X vitamins instead of 2 M pyruvate (use same
volumes) for a 1/ 100 dilution of the anaerobic vitamin stock.

15. Repeat step 2 through 13 using the stock culture (your inoculum) instead of 2 M
pyruvate (use same volumes) for a 1/ 100 dilution of the stock culture.

16. Mix and incubate the newly inoculated culture (optimum grth at 37°C; slower

grth at lower temperatures).

*Do not expect cell densities as high as to Escherichia coli.

MFor CBF medium follow the same procedure substituting 2 M lactate for 2 M pyruvate.
Additionally, a step 14a should be added, in which the electron acceptor (i.e. metals or
nitrate) is added by repeating steps 2 through 13. Thus, a fourth needle and syringe barrel
are needed.

***For CBF medium using ferric citrate as the electron acceptor no additional step 143 is

required as the Fe is added during the process of making the CBF medium.

147

 

 

Figure B.2. Black butyl rubber stoppers with crimped aluminum seals on a specially designed 3-port 1 l

ﬂask. These 3-port ﬂasks were used to hold maximally 500 ml anaerobically.

MATERIAL SOURCES

Source for black butyl rubber stoppers (13 X 20 mm): Geo-Microbial Technologies Inc.,
PO. Box 132, East Main Street, Ochelata, OK 74051. Catalog# 1313.

Source for tear-off aluminum crimp top seals (20 mm): distributed by VWR. Catalog#
16171-851. Manufactured by Wheaton Science Products, 1501 N. 10th Street Millville,
NJ 08332-2093. Catalog# 224193-01

148

REFERENCES

Balch, W. E., G. E. Fox, L. J. Magrum, C. R. Woese, and R. S. Wolfe. 1979.
Methanogens: reevaluation of a unique biological group. Microbiol. Rev. 43:260-296.

thﬂer, F. E., R. A. Sanford, and J. M. Tiedje. 1996. Initial characterization of a
reductive dehalogenase ﬁ'om Desulﬁtobacterium chlororespirans C023. Appl. Environ.
Microbiol. 62:4982-4985.

Lovley, D. R. and E. J. P. Phillips. 1988. Novel mode of microbial energy metabolism:
organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl.
Environ. Microbiol. 54:1472-1480.

Madsen, T. and D. Licht. 1992. Isolation and characterization of an anaerobic
chlorophenol-transfornnng bacterium. Appl. Environ. Microbiol. 58:2874-2878.

Miller, T. L. and W. J. Wolin. 1974. A serum bottle modiﬁcation of the Hungate
technique for cultivating obligate anaerobes. Appl. Microbiol. 272985-987.

Tschech, A., and N. Pfennig. 1984. Growth yield increase linked to caffeate reduction in
Acetobacterium woodii. Arch. Microbiol. 137:163-167.

Widdel, F., G.-W. Kohrin, and F. Mayer F. 1983. Studies on dissimilatory sulfate-
reducing bacteria that decompose fatty acids. Arch. MicrobioL 134:286-294.

Wolin, E. A., R. S. Wolfe, and M. J. Wolin. 1963. Viologen dye inhibition of methane
formation by Methanobacillus omelianskii. J. Bacteriol. 87:993-998.

149

 

Appendix C

Protocols for Selenium Vesicles Isolation from Se-Reducing Cultures of

Desulﬁtobacterium hafniense DCB-2 with RNA and Protein Analyses

The development of selenium-ﬁlled vesicles by Desulﬁtobacterium hafniense DCB-2
was found during growth with selenate and selenite (Chapter 4). The following is a
protocol developed to isolate the vesicles from the cells for ﬁnther downstream analyses,
such as RNA analyses or the identiﬁcation of proteins in the vesicular membrane (also
described below). Proteomics of cells (isolated from the Sc vesicles) may also be desired

in future work, and a protocol is described below.

Se vesicles isolation. D. hafniense cultures grown in 1 mM Se(VI) using strict anaerobic

technique (gas mixture of N2:C02 (95:5)) at 37°C in a modiﬁed DCB-l medium with

pyruvate and Wolin vitamins were spun 30 min at 12,000 g (L6fﬂer 1996, Wolin 1963,
Chapter 4). The pellet was resuspended in phosphate buffered saline (PBS) and sonicated
with 1 sec pulses for 20 min with 120 W output (W-385 Ultrasonics Processor, Misonix,
Farmingdale, NY). The sonicated material was layered onto a step gradient column
consisting of 100% and 50% Percoll® layers spun for 6 min at 2,200 g. The Percoll®
gradients were fractionated into 1 ml aliquots and spun for 10 min at 12,000 g. The pellet
was then washed three times with PBS and spun for 10 min at 12,000 g to remove the

Percoll®.

150

 

RNA isolation and protein preparation. Replicate pellets of puriﬁed Se vesicles were
used for protein and RNA studies. RNA was extracted ﬁ'om the Se vesicles using the
RNeasy kit (QIAGEN, Valencia, CA). For protein studies of the Sc vesicles, the pellet
was resuspended in PBS, beta-mercapto ethanol, and two fold loading buffer containing
SDS following the ReadyPrep Protein Extraction Kit (Bio-Rad Laboratories Inc.,
Hercules, CA). Dissolution was achieved by heating to 90°C for 10 min. This was then

spun at 12,000 g for 5 min with the upper lysate being loaded for SDS-PAGE.

Proteomic preparation and analysis. Proteomic analyses should be performed against
positive control cells, which for our study were pyruvate-fermenting cells. Cells
harvested via centrifugation (and following the vesicle isolation protocol above if
desired) are broken using a French press. Soluble and membrane proteins are then run on
separate 2-D SDS-PAGE gels for each condition. Spots are selected based on gel
comparisons between the test (metal-exposed) condition and the positive control for MS-
DS analysis. These spots are then excised, gel puriﬁed, and run on a Thermo-Finnigan

LTQ-FT mass spectrometer (Thermo Scientiﬁc, Waltham, MA) to identify the proteins.

ACKNOWLEDGEMENTS

The assistance of Dr. Sang-Hoon Kim of Michigan State University was greatly

appreciated in the development of these protocols.

151

REFERENCES
Lﬁffler, F. E., R. A. Sanford, and J. M. Tiedje. 1996. Initial characterization of a
reductive dehalogenase ﬁ'om Desulﬁtobacterium chlororespirans C023. Appl. Environ.
Microbiol. 62:4982-4985.

Wolin, E. A., R. S. Wolfe, and M. J. Wolin. 1963. Viologen dye inhibition of methane
formation by Methanobacillus omelianskii. J. Bact. 87:993-998.

152

 

5"
Am3
will“
H|