A “.mmu-u
. '0
31516.?

$

.1

an}.
.Pfln

< .115-
.4

1.5:!

 

in,
I

, .
5....“ r.
$5.;
use.»

fihnfiua

.hbilfisfls
.! Z. In}.
3):.-

is.

. nu.
amusu
find gnu-411d
.hérrm
R...
3
a .

Eu

K.

1...;

 

 

W LIBRARY
‘2 Michigan State
W University

 

 

This is to certify that the
dissertation entitled

ANALYSIS OF MICROBIAL COMMUNITIES IN A
CONTAMINATED AQUIFER UNDERGOING URANIUM
BIOREMEDIATION

presented by

ERICK CARDENAS POIRE

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Crop and Soil Sciences -
Environmental Toxicology

 

 

 

 

 

Majy Professor’s §Ignature

 

Date

MSU is an Affirmative 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 K:IProj/Aoc&Pres/ClRC/Dale0ue indd

ANALYSIS OF MICROBIAL COMMUNITIES IN A CONTAMINATED AQUIFER
UNDERGOING URANIUM BIOREMEDIATION

By

Erick Cardenas Poire

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Crop and Soil Sciences — Environmental Toxicology

2009

ABSTRACT

ANALYSIS OF MICROBIAL COMMUNITIES IN A CONTAMINATED AQUIFER
UNDERGOING URANIUM BIOREMEDIATION

By

Erick Cardenas Poire

Contamination by metal and radionuclides is a problem for the Y-12 National
Security Complex in Oak Ridge, Tennessee, and at other Department of Energy (DOE)
facilities and waste sites. In contrast to organic compounds, metals cannot be degraded so
their bioremediation may be possible by controlling their bioavailability. Iron-reducing
bacteria such as Geobacter and sulfate-reducing bacteria such as Desulfovibrio, among
others, are known to utilize metals and radionuclides as terminal electron acceptors.
Uranium (VI) reduction decreases its solubility therefore its movement in water can be

controlled.

In this work, microbial communities of a uranium and nitrate contaminated
aquifer undergoing bioreduction, and located at the Field Research Center (FRC) Area 3
at Oak Ridge, Tennessee, were analyzed through comparative 16S rRNA gene analysis
using both traditional and pyrosequencing of clone libraries. Clone libraries of this gene
revealed sequences belonging to genera known to contain uranium reducers and whose
closest relatives are known to reduce U(VI) and utilize either ethanol, the carbon source
injected as electron donor; acetate, a by-product of ethanol oxidation; or methanol, an
impurity of the injected ethanol solution. The genera detected in clone libraries known to

reduce uranium included Desulfovibrio, Geobacter, Acidovorax, Anaeromyxobacter, and

Desulfosporosinus, among others. Hydrologic characteristics helped to explain the higher
levels of activity of sulfate reducers (SRB), iron reducers (FRB), and nitrate reducers in

certain zones of the site.

Pyrosequencing, a novel tool for microbial ecology studies, was used in the
analysis of the microbial communities of the treatment location in the PRC Area 3.
Differences in community structure along the bioremediation zone were related mostly to
chemical oxygen demand, an indication on how much of the stimulatory carbon source
reached a sampled well. Sequences belonging to genera with U(VI)-reducing ability were
detected throughout the Area 3. The genera detected were the same as in the clone
libraries but the increased sampling depth revealed previously undetected Shewanella and

Deinococcus sequences.

Diversity-based clustering showed four clusters of sequences that were related to
their bioactivity (SRB, FRB, and denitrifiers levels). Indicator species analysis in
addition to the clustering showed several groups that were significantly higher in
abundance in the most active uranium reducing zones. The most significant indicator
species were related to Desulfovibrio, Desulfosporosinus, and Anaeromyxobacter. Based
on indicator species analysis, patterns of abundance, and metabolic ability of their closest
relatives, sulfate-reducing bacteria, especially Desulfovibrio, appear to play the most

important role in uranium reduction at this location at the time of the sampling.

This thesis is dedicated to my family for their support,
To my grandparents, I am sure you are proud
To Eliana, my wife, who started me with the idea of the Ph.D.

and to Luciana, my daughter, for the inspiration to finish it

iv

ACKNOWLEDGEMENTS

I would like to thank Mary Beth Leigh for her mentoring during my early years at
the lab.

Carlos Rodriguez-Minguela, Claribel Cruz-Garcia and Veronica Gruntzig helped me
adapting to the lab, to the sometimes-cold Michigan, and to the research world, I will
never forget that.

I would like to thank the Ribosomal Database Project team with whom I worked
for over a year. They really amazed me with the hard work they do and taught me much
of the comparative 16S rRNA analysis I know. My thanks to Jim Cole, its director, one of
the brightest people I have ever met.

Nowadays, research in environmental microbiology is a multidisciplinary team
effort, and this work was not an exception. This thesis would have not been possible
without the contribution of scientists at the Department of Energy Field Research Center,
Wei-min Wu and Craig Criddle from Stanford University, and Jizhong Zhou from
Oklahoma University.

I would also like to thank my labmates for providing a stimulating and fun work
environment.

This research was supported by the US Department of Energy under grants DE-
FG02-97ER62469 and DE—FG02-99ER62848, and by the US National Institute of

Environmental Health Sciences grant 5P42 ESOO4911-18/19.

PREFACE

Parts of the work for this thesis (chapters two and three) are part of a team effort
between Michigan State University, Stanford University, and the Field Research Center
(FRC) at Oak Ridge National Laboratory.

Stanford University designed and managed the groundwater treatment system
used at the Field Research Center Area 3 and mentioned in this work. This group also
characterized the hydrology of the site. Wei-min Wu, Craig Criddle, J ian Luo, Matthew
Ginder-Vogel, and Peter Kitanidis formed the Stanford team.

Oak Ridge National Laboratory provided us the site, performed biological and
chemical analysis, and retrieved the sediment samples used for microbial community
analysis. Jizhong Zhou, Terry Gentry, Jack Carley, Sue Carroll, David Watson, Baohua
Gu, and Phil Jardine were part of the Oak Ridge team. Christopher Hemme, also from
Oak Ridge, provided the DNA I used to construct one of our gene libraries, and helped us
with useful discussion with his preliminary data from a metagenome constructed for that
sample.

Michigan State University was in charge of the microbial community analysis.
Mary Beth Leigh, Terence Marsh, James Tiedje and I were part of the team. Tony Gaca
helped with the construction of the clone libraries as well as extracting DNA from
sediment samples.

Finally the staff from the Ribosomal Database Project (RDP) provided the

website used for the pyrosequencing analysis, technical support, and custom java scripts

vi

used for this work. Jim Cole, Benli Chai, Qiong Wang, Siddique Kulam-Syed-Mohideen,

and Ryan Farris were valuable members of the RDP for this project.

A portion of this work was previously published in here (1) . This part is not

included in this thesis.

vii

REFERENCES

Wu, W. M., J. Carley, J. Luo, M. Ginder-Volgel, E. Cardenas, M. B. Leigh,
C. Hwang, S. D. Kelly, C. Ruan, L. Wu, J. van Nostand, T. Gentry, K. Lowe,
T. Melhoume, S. Carroll, M. W. Fields, B. Gu, D. Watson, K. M. Kemner, T.
L. Marsh, J. M. Tiedje, J. Zhou, P. K. Fendorf, P. K. Kitanidis, P. M.
Jardine, and C. S. Criddle. 2007. In situ bioremediation of uranium (V I) to
submicromolecular levels and reoxidation by dissolved oxygen. Environ. Sci.

Technol. 41:5716-5723.

viii

TABLE OF CONTENTS

 

 

LIST OF TABLES xii
LIST OF FIGURES xiv
CHAPTER ONE
Introduction ....................................................................................................................... 1
Physicochemical characteristic of uranium and its applications ....................................... l
Uranium contamination .................................................................................................... 2
Uranium reduction as bioremediation strategy ................................................................. 3
Microbial communities involved in uranium bioremediation .......................................... 7
References ......................................................................................................................... 9
CHAPTER TWO
Microbial communities in contaminated sediments, associated with bioremediation of
uranium to submicromolar levels ...................................................................................... 14
Abstract ........................................................................................................................... 15
Introduction .................................................................................................................... 1 6
Materials and methods ................................................................................................... 19
Site description and bioremediation test ...................................................................... 19
Sampling for community analysis ............................................................................... 21
Microbial enumeration ................................................................................................. 24
Tracer test ..................................................................................................................... 25
DNA extraction and community analysis .................................................................... 25
Chemical and analytical methods ................................................................................ 28
Results ............................................................................................................................. 28
In situ biostimulation test ............................................................................................. 28
Groundwater flow pathway .......................................................................................... 31
Microbial enumeration ................................................................................................. 34
Sequence analysis ........................................................................................................ 35
Microbial community structure and major groups detected ........................................ 36
Diversity analysis ......................................................................................................... 42
Discussion ....................................................................................................................... 44
1. Remediation of uranium and metals in the subsurface ............................................ 44
2. Microbial communities ............................................................................................ 44
(2a) Major trophic groups detected ........................................................................... 44
(2b) Carbon source as selective agent and the community of ethanol degraders ..... 45
(2c) Other C sources and contaminants .................................................................... 46
(2d) Putative genera involved in U(VI) reduction .................................................... 48

ix

3. Dynamics of the community .................................................................................... 50

(3a) Patterns of diversity detected ............................................................................. 51
(3b) Emergence of previously undetected populations ............................................. 52
Acknowledgements ......................................................................................................... 54
References ....................................................................................................................... 55
CHAPTER THREE
Massive parallel sequencing of microbial communities in a uranium remediation site ....62
Introduction .................................................................................................................... 62
Materials and methods ................................................................................................... 66
Site description and sampling ...................................................................................... 66
Chemical and analytical methods ................................................................................ 67
DNA extraction and direct sequencing ........................................................................ 68
Pyrosequencing analysis .............................................................................................. 69
Community analysis ..................................................................................................... 70
Multivariate analysis .................................................................................................... 71
Results ............................................................................................................................. 72
Chemical and hydrological characterization ................................................................ 72
Correlation of environmental variables ....................................................................... 78
Pyrosequencing results ................................................................................................. 82
Hierarchical clustering of samples ............................................................................... 86
Microbial community structure .................................................................................... 88
Presence of genera with members known to reduce uranium ...................................... 94
Indicator species ........................................................................................................... 98
Ordination of samples by principal coordinates analysis .......................................... 106
Discussion ..................................................................................................................... 109
Effect of hydraulic control on microbial communities of the treatment area ............ 109
Fate of electron donor injected and populations that responded to the treatment ..... 111
Uranium reduction and its correlation with microbial groups ................................... 113
Patterns of diversity ................................................................................................... 115
Conclusions ................................................................................................................... 1 16
Acknowledgements ....................................................................................................... 1 17
References ..................................................................................................................... l 18
CHAPTER FOUR
New tools for discovering and characterizing microbial diversity .................................. 125
Summary ....................................................................................................................... 126
Introduction ................................................................................................................... 126
Impact of new sequencing technologies in microbial ecology ..................................... 128
A new look at metagenomics ........................................................................................ 131
Array-based approaches for studying microbial diversity ............................................ 132
Microfluidics, manipulations at micro(be) scale .......................................................... 133
The formerly unculturable — New approaches for isolation ......................................... 133
Metatranscriptomics and metaproteomics, fields in their infancy ................................ 135

Future challenges and directions ................................................................................... 138

Reference annotations ................................................................................................... 138
Acknowledgements ....................................................................................................... 142
References ..................................................................................................................... 143
CHAPTER FIVE
Conclusions ...................................................................................................................... 149
APPENDIX ..................................................................................................................... 152
Appendix references ..................................................................................................... 153

xi

LIST OF TABLES

Table 2.1. Chemical properties of groundwater in wells of treatment area before and after
remediation and in control well F W1 06“ ........................................................................... 23

Table 2.2. MPN for three major trophic groups in sediments of treatment zone (day 774)
in comparison with groundwater from control well FW106 ............................................. 35

Table 2.3 Relative contribution of known uranium reducer genera in clone libraries from

sediment samples in Area 3wells" ...................................................................................... 41
Table 2.4. Diversity indices in treated and untreated (FW 106) areas .............................. 42
Table 3.1. Hydrological characterization of the DOE FRC Area 3 wells ......................... 74

Table 3.2. Chemical composition of groundwater from wells from DOE FRC Area 3....75

Table 3.3. Chemical composition of sediments from DOE F RC Area 3 .......................... 77

Table 3.4. Correlation of chemical characteristics of groundwater, sediments and
hydrology of the DOE FRC Area 3 wells .......................................................................... 80

Table 3.5. Sequences per sample and number of groups detected at different levels of
taxonomy ............................................................................................................................ 83

Table 3.6. Detection of phyla poorly represented in 16S rRNA databases in libraries from
FRC area 3 samples ........................................................................................................... 84

Table 3.7. Top 15 indicator species for sample cluster A, their classification and closest
isolated relative ................................................................................................................ 102

xii

Table 3.8. Top 15 indicator species for sample cluster B, their classification and closest
isolated relative ................................................................................................................ 103

Table 3.9. Top 15 indicator species for sample cluster C, their classification and closest
isolated relative ................................................................................................................ 104

Table 3.10. Top 15 indicator species for sample cluster D, their classification and closest
isolated relative ................................................................................................................ 105

Table 3.11. Environmental factor that explain distribution of groups along the samples

.......................................................................................................................................... 109
Table 4.1. Current available sequencing technologies and their features ....................... 129
Table 4.2. Selected resources available for metagenomics, meta transcriptomics and

metaproteomics ................................................................................................................ 1 3 7

xiii

LIST OF FIGURES

Figure 1.1. Mechanism for biological mediated Uranium (VI) dissimilatory

reduction .............................................................................................................................. 5

Figure 2.1. Map of the Area 3 treatment zone depicting the location of the samples wells
and the control well. .......................................................................................................... 21

Figure 2.2. Typical biostimulation of U(VI) reduction by injection ethanol to the
subsurface (days 704 to 714) ............................................................................................. 29

Figure 2.3. A tracer test performed with Br' on days 801 to 801 shows the hydraulic
connection between injection well F W104, extraction well FW026, and MLS wells and
biodegradation of ethanol .................................................................................................. 33

Figure 2.4. Rarefaction curves of the 16S rRNA gene libraries constructred .................. 36

Figure 2.5. Microbial composition of the clone libraries based on the RDP Classifier....38

Figure 2.6. Neighbor-joining tree showing the relationship of selected representatives

(shown in bold) from groups similar to known U-reducing bacteria ................................. 40
Figure 2.7. Clustering of samples according to diversity patterns .................................... 43
Figure 3.1. Scheme of the DOE Field Research Center Area 3 treatment system ............ 67

Figure 3.2. Correlation plot of environmental variables measured for 15 locations at
DOE FRC Area 3 ............................................................................................................... 79

Figure 3.3. Rarefaction curves for F RC Area 3 samples .................................................. 85

xiv

Figure 3.4. Richness estimations for samples from DOE FRC Area 3 as influenced by
evenness and sampling effort ............................................................................................. 86

Figure 3.5. Clustering of samples by abundance-based Sorensen index .......................... 87

Figure 3.6. Relative contribution of most abundant phyla to the microbial community
from 15 samples from DOE FRC Area 3 ........................................................................... 89

Figure 3.7. Relative contribution of the five most abundant phyla present at samples
from the FRC Area 3 .......................................................................................................... 91

Figure 3.8. Class composition of the Proteobacteria from FRC Area 3 samples ............ 92
Figure 3.9. Relative abundance of Deltaproteobacteria orders in FRC Area 3 samples .94

Figure 3.10. Relative contribution of most common uranium reducing genera found in
the samples from DOE FRC Area 3 .................................................................................. 95

Figure 3.1 1. Location of maximum abundance for uranium reducing genera along the
connectivity gradient at FRC Area 3 ................................................................................. 97

Figure 3.12. Principal coordinates ordination on abundance-based Sorensen distances for
samples from the FRC Area 3 .......................................................................................... 107

Figure 4.1. Novel methods to discover and characterize microbial diversity ................. 128

XV

INTRODUCTION

Uranium Contamination and Strategies for its Bioremediation

1. Physicochemical characteristic of uranium and its applications

Uranium (U) is the natural element with the highest atomic number, 92 (protons

in its nucleus), and one of the densest (19.1 g-mL-l). Uranium occurs naturally as a
mixture of several isotopes, mostly 238U (99.28%) followed by 235U (0.71%) and 234U
(0.0054%). The half-life for these isotopes are 4.47 x 109 years, 7.0 x106 years and 2.55
x105 years, respectively. Because of its size, uranium’s nucleus is unstable and decays

naturally 1n a known serles of elements emitting alpha particles. Uranium ranks 42n in

abundance in the Earth crust, and it is distributed worldwide and even present in the
oceans. Uranium is exploited mostly in mining operations in Canada, Australia and

Russia.

The main applications of uranium exploit two of its features: 1) its high density
and 2) the ability of its nucleus to be broken. In the military world, uranium is used for
armor plating and high penetration bullets. By being naturally fissible, it can be used for
energy production in nuclear reactors, and for military purposes in bombs. Other uses

include painting, radioisotope dating and transmission electron microscopy.

In order to make uranium usable for energy production, uranium has to be

transformed to increase the proportion of 235U isotope relative to 23 8U. This process
generates depleted uranium [typically less than 0.2 % of 235U, low enriched uranium (0.2
235 . . . 235 . .
to 3% of U) and ennched uran1um(typ1cally 3 to 5% U)]. Uraruum enrichment
. . . . . 235
creates a high proport1on of depleted uranium dependlng on the natural U

. . . . 235
concentratlon and the targeted final concentration. e.g., enrichlng for 3.6% U for fuel

in a nuclear plant will need approximately 8 kg of natural uranium per kg of enriched

uranium.

2. Uranium contamination

Uranium it is a contaminant in facilities associated with uranium mining and
enrichment. Uranium mining is the most frequent industry that has uranium
contamination problems. In these sites, human exposure to uranium is due to inhalation

and skin contact and the risks relate primarily to the chemical toxicity of U (as a heavy

metal) and secondly to its radioactivity. However, 222Ra, a byproduct of U decay, is

associated with lung cancer (7).

Depleted uranium contamination is associated with uranium enrichment facilities
as well as to the military industry where uranium is used in ammunition or heavy armor
plating. The most frequent exposure route for depleted U is inhalation followed by

ingestion through water (4). Acute exposure to uranium leads to renal toxicity, while

chronic exposure can problems in lungs, the central nervous system, intestines, and
kidneys (3, 7). Depleted uranium is 60% less radioactive than naturally occurring

uranium, and its toxicity is more related to its heavy metal characteristics.

Metals and radionuclides are frequent contaminants in the US Department of
Energy facilities and waste sites (31). One of these sites, the Y-12 National Security
Complex in Oak Ridge, Tennessee, USA, has groundwater contaminated with depleted
uranium as well as nitrate, sulfate, and heavy metals. The main sources of contamination
are four ponds (S-3 ponds) where the Y-l2 uranium enrichment plant stored its wastes for
over 32 years. These unlined ponds had a 9.5 million liter capacity each and receive
mainly acidic uranium nitrate as well as solvents, and wastes from other DOE facilities.
In 1983, the ponds were neutralized with limestone, and their nitrate was removed by
denitrification. Treated water was pumped out of the site and the ponds with their
contaminated sediments were capped with a Resource and Conservation Recovery Act
(RCRA) cover, and paved with asphalt to construct a parking lot (5). Nevertheless, the
areas surrounding the ponds were already contaminated, and intrusion of groundwater to
the ponds promoted the movement of soluble uranium from them (30, 40). Groundwater
at this site is highly acidic (pH 3.5) and contains high levels of inorganic contaminants
such as nitrate and sulfate; radionuclides such as uranium and technetium; and solvents
such as acetone, trichloroethene and tetrachloroethene (41). This site is now used to

develop and study uranium remediation strategies.

3. Uranium reduction as bioremediation strategy

One of the earliest definitions of bioremediation was “ the utilization of
microorganisms for the destruction of chemical pollutants” (1). Though this definition
works fine for the degradation of organic compounds, it does not apply for metals as they
cannot be degraded. However, bioremediation of metals is possible by altering their
bioavailability. This can be done by sequestering metals into organic matrices, by
assimilating the metals into the cells, by changing the solubility of the metals via
oxidation/reduction or complextion with organic compounds, and by modifying the
chemicals associated with uranium in situ (e. g., dissolving insoluble phosphate-uranium

precipitates) (17, 23, 36).

From these, one of the most promising strategies is dissimilatory metal reduction,
a process by which microorganism use metals (e.g., Fe(III), Mn (IV), and Cr (VI)) and
radionuclides (e.g., U(VI) and Tc(VII)) as terminal electron acceptors (18). The electrons
can come from the breakup of organic compounds and, in some special cases, directly
from electrodes (10). Electrons are then transferred from the organisms to the metals
through direct contact such in Geobacter spp. (26); through electron shuttles such in
Shewanella spp. (22); or through reduced products such as hydrogen sulfide (13), reduced
humic acids (16, 27), nitrogen oxides (32), and reduced Fe(II) such as green rust(29)
(Figure 1.1). While oxygen is usually the favored electron acceptor for microbial

communities, these metals can be reduced in anaerobic conditions.

Changes in the oxidation state can trigger changes in solubility in both directions.
Reduction can yield more soluble products (e. g., F e(III) to Fe(II)) or more insoluble ones

(Tc(VI) to Tc(V) and Tc(IV)). In the case of uranium the reduction is desirable because

reduced uranium (U(IV)) is less water soluble; thus, uranium movement through water

can be controlled if populations capable of reducing uranium are stimulated.

Stimulation of uranium reducers can be achieved mainly by injecting an electron
donor and/or carbon source into the aquifer and providing the right conditions for
microbial activity in terms of pH, nutrients and minimizing competing electron acceptors
such as oxygen and nitrate. More recently uranium reduction has been achieved through
direct electron transfer by inserting electrodes, and electrodes coated with iron reducing

bacteria (F RB) into aquifers (10).

  
   
  

@ Organic
electron

donor
Electrons in

 

EICCIIOHS out Ii I 4 Electron”

Figure 1.1. Mechanism for biological mediated Uranium (VI) dissimilatory
reduction. Electron can be donated from electrodes directly (1) or from the
decomposition of organic compounds (2). Form bacterial cells, electron can then be
transferred to uranium by direct contact (3), or through electron shuttles of reduced
compound such as humic acids or hydrogen sulfide (4).

Dissimilatory uranium reduction was first demonstrated for the FRB Geobacter
metallireducens, and the sulfate-reducing bacteria (SRB) Shewanella putrefaciens, both
of them belonging to the Delta class of the Proteobacteria (20). Since then, the ability to
reduce uranium has been detected in at least 18 other genera (3 8) whose affiliations range
from F irmicutes, Deinococcus, Thermus, Actinobacteria, Beta-, Delta- and Gamma-
Proteobacteria as well as Archaea (38). In terms of microbial diversity the ability to
reduce uranium seems widespread through the tree of life; however, for management
purposes the widespread distribution can create challenges in the stimulation of the
desired populations. In addition, is not clear what differences in reduction efficiency exist
among the groups, and how this efficiency relates to environmental conditions such as
pH, oxygen concentrations, and electron donor availability among others. Uranium
reduction as a way of controlling its movement has been tested in pure cultures of F RB
and SRB (19, 21), batch cultures in serum bottles (8, 12, 43), microcosms (28), columns

(11, 39) and in pilot scale projects (2, 15, 37, 41).

Field studies have yielded diverse results depending on the carbon source used,
the delivery method, and the characteristics of the site. In a former uranium mine,
microcosm studies using contaminated sediments found the SRB Desulfosporosinus, and
the fermenter Clostridium associated with uranium reduction (35). The medium used
contained ethanol, lactate, acetate, benzoic acids, glucose, yeast extract and peptones
(35). At a former uranitun ore processing facility at Rifle, Colorado, in situ stimulation
via acetate injection promoted U reduction by FRB Geobacter, but in a second phase,
SRB not capable of U reduction outgrew Geobacter (2, 37). Recently it was reported that

uranium removal at Rifle continued even after acetate amendments stopped (24), and that

this removal was associated with adsorption of uranium to the biomass of a F irmicutes-
related bacteria that likely grew on dead cells from previously stimulated populations
(24). At the Field Research Center at Oak Ridge, Tennessee, a pilot scale project showed
associations between uranium reduction with SRB Desulfovibrio (14, 42, 43), F RB
Geobacter (14, 42), FRB Anaeromyxobacter (6), SRB Desulfosporosinus (6, 14), and

denitrifying Acidovorax (28).

4. Microbial communities involved in uranium bioremediation

In order to study the microbial communities involved in the uranium reduction in
situ, culture independent methods have been predominantly used. The preferred methods
to track specific populations involved the use of the 16S rRNA gene in combination with
terminal restriction fragment length polymorphism (TRFLP), restriction fragment length
polymorphisms (RFLP), denaturing gradient gel electrophoresis (DGGE), and clone
libraries. Restriction enzyme-based methods are simpler and cheaper and can provide a
profile for the whole community that can be later consistently compared. However, it is
hard to associate peaks or bands with specific populations unless a clone library for the
same site is constructed. DGGE provides bands that can be tracked over samples and
later sequenced for their phylogenetic information. However, the variability between and
within gel gradients as well as a limitation in how many bands can be resolved (usually
20 to 40), can make this method challenging to analyze (25). Finally, clone libraries
provide the most sequence information (usually 500 to 800 bases per clone). This

information can be use for phylogeny inference and community reconstruction. However,

library construction is time consuming, expensive, limited in scope, and in some
occasions bias can be introduced in the cloning step (34). All PCR-based methods such
as, TRF LP, DGGE and clone libraries can be biased due to primer coverage, differences

in copy number of a gene in a genome, and extraction of DNA (9).

One recent alternative that is both high throughput and provide sequence
information is massive parallel sequencing. By sequencing hypervariable regions of the
16S rRNA gene, thousand of reads can be generated per sample, and the structure of the
microbial community reconstructed by comparing the sequences to known databases

(33).

In this thesis I use comparative 16S rRNA gene studies of both traditional and
pyrosequencing clone libraries, to study the diversity and structure of the communities
associated with uranium bioremediation at the Oak Ridge National Laboratory Field
Research Center, Area 3. The focal research questions are 1) what are the main
populations associated with U reduction, and 2) how does their abundance, and
presumably the features favoring their selection, relate to the geochemical conditions

present at this site?

REFERENCES

Alexander, M. 1999. Biodegradation and Bioremediation. Academic Press,
Incorporated, San Diego, California.

Anderson, R. T., H. A. Vrionis, I. Ortiz-Bernad, C. T. Resch, P. E. Long, R.
Dayvault, K. Karp, S. Marutzky, D. R. Metzler, A. Peacock, D. C. White, M.
Lowe, and D. R. Lovley. 2003. Stimulating the in situ activity of Geobacter

species to remove uranium from the groundwater of a uranium-contaminated
aquifer. Appl Environ Microbiol 69:5884—91.

Berradi, H., J. M. Bertho, N. Dudoignon, A. Mazur, L. Grandcolas, C.
Baudelin, S. Grison, P. Voisin, P. Gourmelon, and I. Dublineau. 2008. Renal

anemia induced by chronic ingestion of depleted uranium in rats. Toxicol Sci
103:397-408.

Bleise, A., P. R. Danesi, and W. Burkart. 2003. Properties, use and health
effects of depleted uranium (DU): a general overview. J Environ Radioact 64:93-
112.

Brooks, S. C. 2001. Waste Characteristics of the Former S-3 ponds and outline of
uranium chemistry relevant to NABIR Field Research Center studies ORNL/TM-2001/27.
NABIR Field Research Center.

Cardenas, E., W. M. Wu, M. B. Leigh, J. Carley, S. Carroll, T. Gentry, J.
Luo, D. Watson, B. Gu, M. Ginder-Vogel, P. K. Kitanidis, P. M. Jardine, J.
Zhou, C. S. Criddle, T. L. Marsh, and J. M. Tiedje. 2008. Microbial
communities in contaminated sediments, associated with bioremediation of
uranium to submicromolar levels. Appl Environ Microbiol 74:3718-29.

Department of Protection of the Human Environment, W. H. O. 2001.
Depleted Uranium: Sources, Exposure and Health Effects WHO/SDE/PHE/Ol .1.

Finneran, K. T., M. E. Housewright, and D. R. Lovley. 2002. Multiple
influences of nitrate on uranium solubility during bioremediation of uranium-
contaminated subsurface sediments. Environ. Microbiol. 4:510-6.

10.

ll.

12.

13.

14.

15.

16.

17.

18.

Friedrich v. Wintzingerode, U. B. G., Erko Stackebrandt,. 1997.
Determination of microbial diversity in environmental samples: pitfalls of PCR-
based rRNA analysis. FEMS Microbiology Reviews 21:213-229.

Gregory, K. B., and D. R. Lovley. 2005. Remediation and recovery of uranium
from contaminated subsurface environments with electrodes. Environ Sci Technol
39:8943-7.

Cu, B., W. M. Wu, M. A. Ginder-Vogel, H. Yan, M. W. Fields, J. Zhou, S.
Fendorf, C. S. Criddle, and P. M. Jardine. 2005. Bioreduction of uranium in a
contaminated soil column. Environ Sci Technol 39:4841-7.

Holmes, D. E., K. T. Finneran, R. A. O'Neil, and D. R. Lovley. 2002.
Enrichment of members of the family Geobacteraceae associated with stimulation

of dissimilatory metal reduction in uranium-contaminated aquifer sediments.
Appl. Environ. Microbiol. 68:2300-6.

Hua, B., H. Xu, J. Terry, and B. Deng. 2006. Kinetics of uranium(VI) reduction
by hydrogen sulfide in anoxic aqueous systems. Environ Sci Technol 40:4666-71.

Hwang, C., W. Wu, T. J. Gentry, J. Carley, G. A. Corbin, S. L. Carroll, D. B.
Watson, P. M. Jardine, J. Zhou, C. S. Criddle, and M. W. Fields. 2008.
Bacterial community succession during in situ uranium bioremediation: spatial
similarities along controlled flow paths. ISME J.

Istok, J. D., J. M. Senko, L. R. Krumholz, D. Watson, M. A. Bogle, A.
Peacock, Y. J. Chang, and D. C. White. 2004. In situ bioreduction of

technetium and uranium in a nitrate-contaminated aquifer. Environ Sci Technol
38:468-75.

Jang, J. H., B. A. Dempsey, and W. D. Burgos. 2008. Reduction of U(VI) by
Fe(II) in the presence of hydrous ferric oxide and hematite: effects of solid
transformation, surface coverage, and hurrric acid. Water Res 42:2269-77.

Knox, A. S., R. L. Brigmon, D. 1. Kaplan, and M. H. Paller. 2008. Interactions
among phosphate amendments, microbes and uranium mobility in contaminated
sediments. Sci Total Environ 395:63-71.

Lovley, D. R. 1993. Dissimilatory metal reduction. Annu Rev Microbiol 47:263-
90.

10

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Lovley, D. R., and E. J. Phillips. 1992. Reduction of uranium by Desulfovibrio
desulfuricans. Appl. Environ. Microbiol. 58:850-856.

Lovley, D. R., E. J. P. Phillips, Y. A. Corby, and E. R. Landa. 1991. Microbial
reduction of uranium. Nature 350:413-416.

Lovley, D. R., E. E. Roden, E. J. P. Phillips, and J. C. Woodward. 1993.
Enzymatic iron and uranium reduction by sulfate—reducing bacteria. Marine
Geology 113.

Marsili, E., D. B. Baron, I. D. Shikhare, D. Coursolle, J. A. Gralnick, and D.
R. Bond. 2008. Shewanella secretes flavins that mediate extracellular electron
transfer. Proc Natl Acad Sci U S A 105:3968-73.

Merroun, M. L., and S. Selenska-Pobell. 2008. Bacterial interactions with
uranium: An environmental perspective. J Contarn Hydrol 102:285-95.

N'Guessan, A. L., H. A. Vrionis, C. T. Resch, P. E. Long, and D. R. Lovley.
2008. Sustained removal of uranium from contaminated groundwater following
stimulation of dissimilatory metal reduction. Environ Sci Technol 42:2999-3004.

Nakatsu, C. H. 2007. Soil Microbial Community Analysis Using Denaturing
Gradient Gel Electrophoresis. Soil Sci Soc Am J 71:562-571.

Nevin, K. P., and D. R. Lovley. 2000. Lack of production of electron-shuttling
compounds or solubilization of Fe(III) during reduction of insoluble Fe(III) oxide
by Geobacter metallireducens. Appl Environ Microbiol 66:2248-51.

Nevin, K. P., and D. R. Lovley. 2000. Potential for Nonenzymatic Reduction of

Fe(III) via Electron Shuttling in Subsurface Sediments. Environ. Sci. Technol.
34:2472-2478.

Nyman, J. L., T. L. Marsh, M. A. Ginder-Vogel, M. Gentile, S. Fendorf, and
C. Criddle. 2006. Heterogeneous response to biostimulation for U(VI) reduction
in replicated sediment microcosms. Biodegradation V17:303-316.

O'Loughlin, E. J., S. D. Kelly, R. E. Cook, R. Csencsits, and K. M. Kemner.
2003. Reduction of uranium(VI) by mixed iron(II)/iron(III) hydroxide (green
rust): formation of U02 nanoparticles. Environ Sci Technol 37:721-7.

11

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

Phillips, D. H., D. B. Watson, Y. Rob, T. L. Mehlhorn, J. W. Moon, and P. M.
Jardine. 2006. Distribution of uranium contamination in weathered fractured
saprolite/shale and ground water. J Environ Qual 35: 1715-30.

Riley, R. G., and J. M. Zachara. 1992. Chemical contaminants on DOE lands
and selection of contaminant mixtures for subsurface science research. US.

Department of Energy, Washington, DC.

Senko, J. M., J. D. Istok, J. M. Suflita, and L. R. Krumholz. 2002. In-situ
evidence for uranium immobilization and remobilization. Environ Sci Technol
36:1491-6.

Sogin, M. L., H. G. Morrison, J. A. Huber, D. Mark Welcb, S. M. Huse, P. R.
Neal, J. M. Arrieta, and G. J. Herndl. 2006. Microbial diversity in the deep sea
and the underexplored. Proc Natl Acad Sci U S A 103:12115-20.

Sorek, R., Y. Zhu, C. J. Creevey, M. P. Francine, P. Bork, and E. M. Rubin.
2007. Genome-wide experimental determination of barriers to horizontal gene
transfer. Science 318: 1449-52.

Suzuki, Y., S. D. Kelly, K. M. Kemner, and J. F. Banfield. 2003. Microbial
populations stimulated for hexavalent uranium reduction in uranium mine
sediment. Appl Environ Microbiol 69: 1337-46.

Turick, C. E., A. S. Knox, C. L. Leverette, and Y. G. Kritzas. 2008. In situ
uranium stabilization by microbial metabolites. J Environ Radioact 99:890-9.

Vrionis, H. A., R. T. Anderson, I. Ortiz-Bernad, K. R. O'Neill, C. T. Resch,
A. D. Peacock, R. Dayvault, D. C. White, P. E. Long, and D. R. Lovley. 2005.
Microbiological and geochemical heterogeneity in an in situ uranium

bioremediation field site. Appl Environ Microbiol 71:6308-18.

Wall, J. D., and L. R. Krumholz. 2006. Uranium reduction. Annu Rev
Microbiol 60: 149-66.

Wan, J., T. K. Tokunaga, E. Brodie, Z. Wang, Z. Zheng, D. Herman, T. C.
Hazen, M. K. Firestone, and S. R. Sutton. 2005. Reoxidation of bioreduced
uranium under reducing conditions. Environ. Sci. Technol. 39:6162-9.

l2

40.

41.

42.

43.

Watson, D. B., W. E. Doll, T. J. Gamey, J. R. Sheeban, and P. M. Jardine.
2005. Plume and lithologic profiling with surface resistivity and seismic

tomography. Ground Water 43: 169-77.

Wu, W. M., J. Carley, M. Fienen, T. Mehlhorn, K. Lowe, J. Nyman, J. Luo,
M. E. Gentile, R. Rajan, D. Wagner, R. F. Hickey, B. Gu, D. Watson, 0. A.
Cirpka, P. K. Kitanidis, P. M. Jardine, and C. S. Criddle. 2006. Pilot-scale in
situ bioremediation of uranium in a highly contaminated aquifer. 1. Conditioning
of a treatment zone. Environ Sci Technol 40:3978-85.

Wu, W. M., J. Carley, T. Gentry, M. A. Ginder-Vogel, M. Fienen, T.
Mehlhorn, H. Yan, S. Carol], M. N. Pace, J. Nyman, J. Luo, M. E. Gentile,
M. W. Fields, R. F. Hickey, B. Gu, D. Watson, 0. A. Cirpka, J. Zhou, S.

F endorf, P. K. Kitanidis, P. M. Jardine, and C. S. Criddle. 2006. Pilot-scale in
situ bioremediation of uranium in a highly contaminated aquifer. 2. Reduction of
u(VI) and geochemical control of u(VI) bioavailability. Environ Sci Technol
40:3986-95.

Wu, W. M., B. Gu, M. W. Fields, M. Gentile, Y.-K. Ku, H. Yan, S. Tiquias, T.
Yan, J. Nyman, J. Zhou, P. M. Jardine, and C. S. Criddle. 2005 . Uranium (VI)
Reduction by Denitrifying Biomass. Bioremediation 9:46-61.

13

CHAPTER TWO

Cardenas, E., W. M. Wu, M. B. Leigh, J. Carley, S. Carroll, T. Gentry, J. Luo, D.
Watson, B. Gu, M. Ginder-Vogel, P. K. Kitanidis, P. M. Jardine, J. Zhou, C. S.
Criddle, T. L. Marsh, and J. M. Tiedje. 2008. Microbial communities in contaminated
sediments, associated with bioremediation of uranium to submicromolar levels. Applied

and Environmental Microbiology 74:3718-29.

l4

Microbial Communities in Contaminated Sediments Associated with

Bioremediation of Uranium to Submicromolar Levels

ABSTRACT

Microbial enumeration, 16S rRNA gene clone libraries and chemical analysis were

used to evaluate the in situ biological reduction and immobilization of uranium (VI) in a
long term experiment (more than two years) conducted at a highly uranium contaminated
site (up to 60 mg/l and 800 mg/kg solids) of the US Department of Energy in Oak Ridge,
TN. Bioreduction was achieved by conditioning groundwater above ground and then
stimulating growth of denitrifying, Fe(III)-reducing, and sulfate-reducing bacteria in situ
through weekly injection of ethanol to the subsurface. After nearly two years of
intermittent injection of ethanol, aqueous U levels fell below the US EPA maximum
contaminant level (MCL) for drinking water and groundwater (<30 ug/l or 0.126 pM).
Sediment microbial communities from the treatment zone were compared with those from
a control well without biostimulation. Most Probable Number estimations indicated that
microorganisms implicated in bioremediation accumulated in the sediments of the
treatment zone but were either absent or in very low numbers in an untreated control area.
Organisms belonging to genera known to include U(Vl) reducers were detected, including
Desulfovibrio, Geobacter, Anaeromyxobacter, Desulfosporosinus and Acidovorax spp.
The predominant sulfate-reducing bacterial species were Desulfovibrio spp. while the iron
reducers were represented by F erribacterium spp and Geothrix spp. Diversity-based

clustering revealed differences between treated and untreated zones and also within

15

samples of the treated area. Spatial differences in community structure within the
treatment zone were likely related to hydraulic pathway and electron donor metabolism

during biostimulation.

INTRODUCTION

Metal and radionuclide are common groundwater contaminants at facilities and
waste sites of the US Department of Energy (DOE) occurring in more than 50% of these
locations (32). One of these sites, the former Y-12 National Security Complex at Oak
Ridge, Tennessee, contains uranium in concentrations as high as 60 mg/l (252 pM) in
groundwater and 800 mg/kg in sediments (48). To control the migration of the uranium,
microbial reduction of U(VI) to sparingly soluble and immobile U(IV) has been proposed
as a promising approach (1, 2, 12, 17, 22, 42, 48). Bioreduction of U(Vl) to U(IV) has
been reported by pure and mixed cultures of iron(III)-reducers (FeRB) such as Geobacter
spp., sulfate—reducers (SRB) such as Desulfovibrio, Desulfoporosz’nus, and
Desulfotomaculum spp. and denitrifying bacteria such as Acidovorax spp. among others

(45).

In recent years, researchers have evaluated this approach through laboratory-scale
experiments using batch serum bottles (10, 14, 51), microcosms (29), and sediment
columns (11, 46). Reduction and immobilization of uranium in the mentioned laboratory
experiments has been observed and confirmed by X-ray near-edge absorption
spectroscopy (XANES) analysis (11, 29, 40, 41, 46). Geobacter spp. and Geothrix spp.

were found to be associated with Fe(III) and U(VI) reduction in field conditions (5, 14,

16

46, 50). Control of the microbial community structure may be one of key issues for the
long-term reduction and stabilization of uranium in situ (1). While bench-scale tests
provide valuable information about the feasibility of bioreduction, they cannot replicate
all of the heterogeneity and complexity of the subsurface. Hydrogeology is a factor that
can contribute to the complexity of field treatments by creating gradients of resources in
situ, affecting the microbial diversity and thus, potentially the remediation process. Little
is known about the effect of groundwater flow on microbial community diversity during
and after bioremediation, even when the kinetics of U(VI) reduction should be dependent
on the microbial community and geochemical conditions. The effect that hydrogeology

has on microbial diversity should be addressed.

Pilot studies are the next step in demonstrating the feasibility of the in situ
uranium remediation approach (12, 14, 17, 37, 48, 49). DOB does not provide specific
uranium target levels, but US EPA regulates the maximum contaminant level (MCL) for
drinking and groundwater at 30 ug/l (0.126 uM). At the Field Research Center (F RC) site
of the Environmental Remediation Sciences Program (ERSP) of the DOE, Oak Ridge,
TN, test facilities have been constructed for the remediation of uranium contaminated
groundwater and sediments using various approaches (1, 17, 49). Prior to the tests, the
microbial community structure was characterized by 16S rRNA gene clone libraries
made from groundwater samples in Areas 1 and 3 and the majority of the sequences
(>73% in Area 1 and >65% in Area 3) were found to be Proteobacteria belonging to the
genera Azoarcus and Pseudomonas (9). Microbial community analyses of Areas 1 and 2
identified Geobacter spp. and Anaeromyxobacter spp. among others as the potential

uranium reducers (28). In these areas of moderate uranium contamination (0 to 5.8 uM),

l7

removal of nitrate and aqueous U(VI) was stimulated by the injection of ethanol or
glucose to the subsurface using a push-pull approach (17). In the extremely contaminated
Area 3, a long-term (> 2 years) bioremediation test has been performed. This area is
located near former S-3 Ponds and contains high levels of nitrate (up to 200 mM) and U
(up to 250 uM), aluminum(12-13mM) and calcium (22-25mM) (48, 49). Prior to
remediation, the microbial levels in groundwater of this area were extremely low
probably due to low pH (3.6) and high levels of nitric acid. A combination of remediation.
approaches was used to remove U(VI) reduction inhibitors (nitrate and Ca) and condition
the area for metal reduction by raising the pH and providing a carbon source. Using these
approaches, low U(VI) concentrations below MCL were achieved and a new microbial

community capable of uranium reduction was established (48-50).

The objective of our study was to characterize the microbial community arising in a
earlier study where successful U(VI) bioremediation was demonstrated (49). We used
16S rRNA genes sequence analysis to determine the structure of the bacterial consortia
present in the treatment area where low uranium levels were achieved during
biostimulation and in a control samples from outside the treated area. The community
data were integrated with geochemical and hydraulic data to provide insight about
environmental variables that profoundly influence the remediation process. We were able
to identify key bacterial groups associated with successful reduction of U(VI) in the
subsurface and correlate their spatial relationship with hydrogeology and geochemistry in

the treatment zone.

18

MATERIAL AND METHODS

Site description and bioremediation test. The bioremediation test was
performed in Area 3 of the DOE F RC at the Y-12 National Security Complex, Oak
Ridge, TN as previously reported (24, 48). The field system consisted of an outer
groundwater recirculation loop (injection well FW024 and extraction well FW103) that
isolates an inner groundwater loop (injection well FW104 and extraction well FW026)
preventing penetration by highly contaminated groundwater from the source zone (Figure
2.1). The hydraulic control afforded by this system created a controlled in situ
bioreactor. Reduction of U(VI) to U(IV) was accomplished through ethanol injection to
the inner loop. Injection and extraction wells (FW104 and FW026) had a 10.2 cm
diameter and a depth of 14.6 m below ground surface (bgs) with 2.5 m screened intervals
between 11.28 and 13.77 m. Multilevel sampling (MLS) wells F W100, F W101 and
FW102 were used to monitor hydrogeology and remediation performance. The MLS
wells contained seven separate sampling tubes (diameter =1.9 cm) at different depths bgs.
In this study, MLS wells FW101-2 (13.7 m bgs), FW101-3 (12.2 m bgs), FW102-2 (13.7
m bgs) and F W102-3 (12.2 m bgs) were selected for routine monitoring of remediation
performance because of their hydraulic connection to FW104 (24, 48). The recirculation
flow rates in the inner loop were 0.45 l/min (injection F W1 04, extraction F W026). The
rates for the outer loop were 1.35 l/min (injection at FW024) and 0.45 l/rnin( extraction at
FW103). Additional clean water was injected to F W024 at 0.7-0.9 l/min in order to
minimize entry of ambient groundwater (48). This clean water was a mixture of tap water
and groundwater treated by an aboveground system to remove nitrate via a bioreactor

(48). The remediation test was started on August 23, 2003 by preconditioning the site

19

(day 0). During the initial 137 days, water was pumped from the subsurface, pH
adjusted, treated to remove aluminum, calcium and nitrate (this by a denitrification
bioreactor) and then re-injected. Calcium and nitrate was removed to avoid formation of
stable Ca-U-CO3 products, U(IV) re-oxidation by nitrites, and nitrate competition as
terminal electron acceptor. Ex situ treatment was used to avoid clogging by nitrogen gas,
biomass (due to denitrification), and calcium and aluminum precipitates (due to pH
adjustment) (48).

As result, the pH increased from 3.6 to around 6.0, and nitrate and U(VI)
concentrations decreased to around 0.5 mM and 5 uM, respectively (48). After the
conditioning phase, ethanol was added as an electron donor to stimulate U(VI)
bioreduction starting on January 7, 2004 (day 137) (49). An ethanol solution (industrial
grade, containing ethanol 88.12%, methanol 4.65% and water 7.23%, w/w) with a
Chemical Oxygen Demand (COD) to weight ratio of 2.1 was prepared in a storage tank
with 6.9-9.8 g COD/l. This solution was normally injected at F W1 04 over a 48-hour

period each week to give a COD 120-150 mg/l at FW104.

20

Outer loop \ FW102
lnnerloop\1 FW101
. FW103... / o FW024
FW105 / \
FWAHJZES III I:VV1(M4

 

 

 

 

 

FW100
Water flow 44'
hfleters
0 1.5 3 6
FW106
0 Wells
I Multisampling Wells

 

 

 

Figure 2.1. Map of the Area 3 treatment zone depicting the location of the sampled wells
and the control well. FW024: outer loop injection; F W1 04: inner loop injection; FW026:
inner loop extraction; FW103: outer loop extraction well. Wells F W100, FW101 and
FW102 are multilevel sampling wells. FW106: control well. Contamination source is
approximately 20 m to the right.

Sampling for community analysis. Sediment samples for microbial community
analysis were collected on day 774 (October 5, 2005) from the inner loop injection
FW104, extraction F W026, and two MLS wells at two depths F W101-2, F W101-3,
FW102-2 and F W102-3. To collect the samples, a smooth PVC surge block (10 cm or
1.8 cm in diameter) was inserted into the well and then lifted up and down in a rapid
plunging motion. This motion drew sediment from the soil matrix surrounding the well

screens into the well. The surge blocks were rinsed with groundwater in between samples

21

to remove attached solids. The sediment slurry that settled to the bottom of the wells was
pumped to a 2 L glass bottle under Ar atmosphere and the bottle was then sealed with a
rubber stopper. The slurry was transferred to the laboratory and centrifuged to separate
the sediments from the water. The pellets were frozen at -80 °C prior to being shipped on
dry ice to Michigan State University. Fresh sediment slurry was also collected
anaerobically in a 27-ml glass pressure tube for the most probable number (MPN)
enumeration. Samples taken by this method are a mixture of the sediment along 1 m
depth of the well (screened interval) and could contain sediment in the deep matrix where

electron donor did not reach.

Before remediation, bioactivity in the treatment zone was extremely poor due to
low pH (3.2-3.6) and high levels of nitric acid and uranium. DNA extraction from
untreated sediments repeatedly failed. In order to compare microbial communities before
and after biotreatrnent, DNA was collected by filtering 1700 liters of groundwater from
the F W106 well which is located 12 m away from the treatment zone in parallel with
treated sediments. Groundwater from FW106 has similar composition to that in the
treatment zone before remediation, i.e. high levels of nitric acid and uranium and low pH

(Table 2.1).

22

80:838. bomb 2: no 833 CA: 35 :03 .838? on“ 8 5:00:88 0:28:83 mamas—8828 :05
9 888888 888885 28 8:83 .83 mmz<x a a .3388 82 a 9:: 88 888 8838...... 88288 .2. 988 a: .88 .8 .808

m: >86 no 2052,385805 88.8 ”m Axum >83 2: 3w 8.8 8588 Roam ._tg<-b88nom E :83 p.53 moifimmv 8288:8882 20.83 ”8....

 

 

 

a. 8882 88 8 8.8 88 2 888 8.2 8.8 8.8 88.8 3888 8;:
.xév 88 88 88 88 888 88 888 88 82 88 88 m 8838
8 8882 :8 R8 888 88.8 8.2 88.8 888 88 8882 8.8 <
88 28 88 88 88 $8 88 :8 8.8 82 88.8 888 m 38;:
8 88.5 888 2 8.8 8.8 88 88.8 88.8 88. _ m 88.: _ 8.8 <
8 888 888 888 :8 888 88 £8 88 _8._ 88.8 88 m 8.88;:
8 88.: a. a: 8.: 8.8 a. 88.8 8.8 88.8 88.5 8.8 <
t 88 88 888 :8 N88 88 8.8 :8 82 88.8 88 m 8-8.3.8
8 8882 a: a. 88 8.8 a. 88.8 88 8.8 88.5 8.8 <
a 28 888 88 88 $8 888 88 8.8 88 88.8 88 m 8-2: 3..
8 888.2 :8 2 8.2 8.8 8.2 88.8 88 8.2 888: 8.8 <
6 R8 .88 88 88.8 888 88 88 88 82 88.8 8.8 m 8;:
8 8882 :8 a. 88.: 8_ .8 88.2 88.8 88;: 88. _ m 8888 88.8 <
r. 28 2.: 25 2:. 2:. 26 28 25 28 28
98:8 2 88 =2 82 8 2 am -6 28:88 2882 :8 .8238 =83
823.8

:83 75:8 2: can 8:860:88 coca new 883 828 8585828 .80 8:83 2: E 82828::on 80 828.59er 303.25 .~.N 25,—.

23

Microbial Enumeration. Denitrifying bacteria, FeRB and SRB were estimated
using the MPN technique with five tubes for each dilution. Anaerobic pressure tubes (27
ml) containing 10 ml basal medium were sealed with butyl rubber stoppers with

aluminum caps. The basal medium contained the following components (per liter):

NH4C1, 0.5g; NaCl, 0.4g; NaHCO3, 0.55g; and mineral solution, 100 ml. The mineral
solution contained (per liter): MnClz, 0.4 g; MgSO4, 1.5 g; CaClz, 0.5 g; and yeast

extract, 0.02g. The medium was prepared under a Nz-COZ (99:1, vol/vol) atmosphere

and distributed to each pressure tube (10 ml per tube). After autoclaving, a sterile trace
element solution (0.4 ml) and a sodium trimetaphosphate solution (50 mM, 0.025 ml) was

added into each tube to obtain pH 7.0. The trace element solution contained (per liter):

HCl (12N), 6.4 ml; FeC12-4H20, 0.3g; ZnSO4-H20, 0.1g; MnSO4, 0.085g; boric acid,

0.06g; CoC12-6H20, 0.02g; CuSO4, 0.004g; NiSO4-6H20, 0.028 g; and NaMoO3-2H20,

0.04g. The electron acceptor solution was added to the tubes to obtain a final

concentration of 5 mM Fe(III)-citrate for FeRB, 8.76 mM of sodium thioglycollate and

33 mM of F eSO4 for SRB, and 9.9 mM of KNO3 for denitrifiers, respectively. Ethanol

solution (1 M) was added to each tube to give a final concentration of 10 mM.
Groundwater from FW106 was pumped from the wells into anaerobic pressure tubes pre-
filled with nitrogen gas. Sediment slurries were collected in pressure tubes under
anaerobic conditions. The sample was then serially diluted in MPN tubes. The tubes were
incubated at room temperature for two months. Tubes were compared to controls for

scoring as positive or negative for production of gas in denitrifying tubes, color change in

24

F eRB tubes, and production of black Fe(II) sulfide precipitates in SRB tubes.

Tracer test. A tracer study was performed to characterize the groundwater flow
in the treatment zone. The hydraulic flow path is expected to affect the delivery of
nutrients and thus, affect the metabolism of ethanol in the groundwater and the microbial
community diversity. Sodium bromide, a conservative tracer, was injected through the

FW104 well together with ethanol from days 801 to 803. An ethanol/NaBr tracer solution

with COD/Br- of 2.46 g/g was prepared and injected into the recirculation line of the

inner loop, resulting in injected concentrations in well FW104 of 50 mg/l bromide and 1.0

mM of ethanol. The tracer test continued for 50.75 h. Samples were periodically taken
from the injection, extraction and MLS wells for analysis of Br-, COD, ethanol and

ac etate .

DNA extraction and community analysis. DNA was extracted from 0.5 g of
sediments with the Fast soil prep kit (MoBio Inc., San Diego, CA, USA) following the
manufacturer’s instructions. DNA was used to amplify 16S rRN A genes using the
universal primers 27F (5’- AGAGTTTGATCMTGGCTCAF-3’) and 1392R (5’-
ACGGGCGGTGTGTRC-3’) in a Stratagene thermal cycler (Stratagene, La Jolla, CA,

USA). The PCR reaction (50 pl) contained 20 mM Tris-HCL (pH 8.4), 50 mM KCl, 3

mM MgC12, 0.1 pg/ul BSA, 10 pmol of each primer, 0.2mM each dNTP , 1.25 U taq

polymerase (Invitrogen Life Technologies, Carlsbad, CA, USA) and 20 ng of purified

25

DNA. The PCR cycling conditions were: 95°C for 5 min; 95 °C for 1 min, 59 °C for 1
min, 72°C for 1min 405, for 28 cycles; 72 °C for 10 min. PCR products were analyzed in
a 1.5% (wt/vol) agarose TAE gel to confirm the size of the product. Four replicate PCR
reactions were generated for each DNA extract and then were compiled to address
variability that may be introduced by PCR bias. The PCR products (200pl) were
concentrated to 30 ill with a PCR purification kit (Qiagen Inc., Valencia, CA, USA). The
concentrated products were then run in a 1% agarose gel, excised, extracted with the
QIAquick Gel Extraction Kit (Qiagen Inc., Valencia, CA, USA) and eluted to 30 pl with
EB buffer (10 mM Tris-Cl, pH 8.5). An additional step was taken to add poly-adenine
overhangs to the PCR products to facilitate cloning (10 min at 72 °C, 18.8 mM Tris HCl
pH8.4, 47mM KCl, 0.93 mM dATP and 0.5 U of taq polymerase; Qiagen Inc., Valencia,
CA, USA). Products were cloned using the Topo TA Cloning kit for sequencing
following manufacturer’s instructions (Invitrogen Life Technologies, Carlsbad, CA,
USA). Single extension sequencing was conducted by Macrogen Inc (Seoul, Korea) from
the internal primer 27F. The sequences were submitted to to the GenBank under

accession numbers EF692646-EF693732.

Raw sequencing files (Abl) were submitted to the myRDP application of the
Ribosomal Database project where the nucleotide sequence was determined, the quality
(Q score) of each base stored and vector sequence removed (7). The resulting sequences
were aligned to the RDP model and rRN A distances matrices were generated with the
Jukes-Cantor distance correction. Operational taxonomic units (OTUs) were defined at
97% sequence identity. The distance matrices were used to calculate alpha diversity

indices (Chaol, Simpson and Shannnon) and rarefaction curves using Distance-Based

26

Operational Taxonomic Unit and Richness (DOTUR) determination program (35). The

diversity indices were rarified to account for differences in number of sequences per

library.

The distribution of the OTUs in the different libraries was used in EstimateS (8)
to make comparisons based on diversity patterns using Bray-Curtis and Sorensen beta
diversity indices. These indices were used to cluster the samples according to their
distances using MEGA 3.1 (20). The composition was normalized to account for
differences in number of clones in the different libraries. I-LIBSHUFF (36, 38) was used
to compare the clone libraries according to Good’s homologous and heterologous
coverage. This approach provides a quantitative comparison of 16S rRNA gene clone

libraries from environmental samples (3 8).

A mask using the quality value (Q>20) was used and the resultant sequences were
used for classification with RDP classifier using a 80% confidence value. MALLARD
(4) was used to detect sequences with anomalies such as chimeras. The putative
chimeras were later re-evaluated using the RDP Sequence Match with a suspicious-free
and near-full-length data set and with the Pintail program (3). Sequences confirmed as

anomalous with Pintail were excluded from the analyses.

Phylogenetic trees were constructed using distance and maximum likelihood
methods. Aligned sequences were downloaded from the RDP, aligned manually and the
non-model positions were masked from the alignment. The Jukes-Cantor distance
correction, full gap deletion and a bootstrap test using 10000 replicates and random seeds

were used to construct the trees in MEGA v3.1 (20).

27

Chemicals and analytical methods. The source and quality of chemicals used in
the field test were described previously (48, 49). COD was used as an overall indicator to
monitor the consumption of electron donors (ethanol, its metabolite acetate and others).

COD, sulfide and Fe(II) were determined using a Hach DR 2000 spectrophotometer
(Hach Chemical, Loveland, CO). Anions (including N03: Br-, Cl-, S042. and P043.)

were analyzed with an ion chromatograph equipped with an IonPac AS-14 analytical
column and an AG-14 guard column (Dionex DX-120, Sunnyvale, CA), metals (Al, Ca,
Fe, Mn, Mg, U and K etc.) were determined using an inductively coupled plasma mass
spectrometer (ICPMS) (Perkin Elmer ELAN 6100), and U reduction state was
determined with XANES as described elsewhere (49, 50). Ethanol and acetate were
determined by a HP5 890A gas chromatograph equipped with a flame ionization detector
and an 80/ 120% Carbopack BDA column (Supelco Division, Sigma-Aldrich Corp., St.

Louis, MO) using Helium as carrier gas.

RESULTS

In situ biostimulation test. After one year of biostimulation, nitrate concentration
in the treatment zone dropped to a level of nearly zero. Increase in sulfide concentration
and appearance of F e(II) in groundwater suggested that sulfate and iron reduction were
occurring (50). Microbial activity during biostimulation was determined by monitoring
changes in aqueous concentration of electron donor added (as COD), sulfide and

uranium. A representative time course of biostimulation is shown in Figure 2.2. On day

28

704, ethanol was injected to FW104, causing increase COD concentrations in all major

MLS wells (Figure 2.2A).

I80 +E -E ‘+E ‘_
8150* E A

 

Oi 000C)

E12010 23° 020::

d90'I 0A .o O

060.3,.“2 A

030 --" 2"";
"A O i o
053°010833138 p 15300 0935

 

aJ

704 706 708 710 71'2 714 71
1.00

B

SUlfide, mM
.9
S

 

0.01
704 706 708 710 7l2 714 716
0.37
2 Q
1 9. . .
8.02.130 9 o .
g 800083. 0 0° 9; 0..
E A" 4344.13 0 A 0 A AA AA dooo
201%.. M . . .“
I. I I I . I ' I
I I I - .
0.0

 

 

704 706 708 710 71‘2 71'4 716
Time, days

9FW101-2 AFWlOl-B I FW102-2

O FW102-3 El FW104 O FW026

Figure 2.2. Typical biostimulation of U(VI) reduction by injecting ethanol to subsurface
(day 704 to 714). (A) COD concentrations. (B) Sulfide concentrations. (C) Uranium
concentration changes in MLS. U concentration in the injection well FW104 was 0.5 pM
during this test period and is not shown due to scale. +E indicates start and —E indicates
stop of ethanol biostimulation.

29

The COD in FW104 was mainly attributed to ethanol (>80%) and a small amount of
acetate (<30 uM) while in MLS and F W026 wells, the COD came mainly from acetate
(data not shown). No ethanol was detected in the MLS wells and FW026 well and
methanol concentrations were below detection limit (<50 uM) for all the cases. On day
706, ethanol injection was stopped and the COD concentrations in MLS wells decreased
rapidly. The same pattern was observed again on day 711 to 713 when ethanol was
reinjected. Sulfide concentrations in all MLS wells increased after ethanol was injected
and decreased after injection stopped (Figure 2.2 B) but remained at detectable levels
(>20 uM). Sulfate concentrations in MLS wells decreased when ethanol was injected and
rebounded when the injection stopped (data not shown) indicating the presence of SRB
activity in the subsurface. Uranium concentration in all MLS wells decreased after
ethanol was delivered and slightly rebounded when ethanol was not injected (Figure
2.2C). However, the uranium levels after the rebound were lower than those before
ethanol injection. This rebound was likely due to uranium being carried in the
recirculated water (~0.5 um) and to the lack of electron donor for U(VI) reduction when
ethanol injection stopped. During the ten-day test in figure 2.2, uranium concentration in
the injection well F W104 was around 0.5 pM (data not shown in figure 2.2C) which was
much higher than those in the MLS wells. Uranium levels in F W1 01-3 and F W102-2
were near or below EPA MCL (0.126 uM) throughout the 10 day-test period while U
levels in FW101-2 and FW102-3 dropped even below 0.126 uM after ethanol injection.
The relatively higher U levels in the latter two wells are likely due to the lack of enough

electron donor after ethanol injection stopped. Lower U concentrations in the MLS wells

30

were achieved later as shown in Table 2.1. Aqueous F e(II) concentrations were 10-20
pM in the MLS wells before ethanol injection, dropped below 5 pM during ethanol
injection as sulfide concentration increased and then slowly rebounded after ethanol
injection stopped and sulfide concentration decreased (data not shown). The decrease in
Fe(II) concentration during ethanol injection is likely due to formation of more F eS
precipitates by reaction with the produced sulfide. The FeS precipitates likely

accumulated in sediments.

The reduced sediment samples showed black color (F W104) or dark green color
(MLS). Reduced U(IV) was detected by XANES in the sediments samples from
injection well and the MLS (Table 2.1). In the F W026 well, the U(IV) content was
below the detection limit of XANES (<10% of total U). The highest content of reduced

uranium was found in the injection well FW104 (50).

Groundwater flow pathway. The injection of the NaBr'/ethanol solution to

F W104 lasted for 50.75 h. Bromide concentration increased in the MLS and extraction
(FW026) wells with different recovery ratio and mean travel time (Figure 2.3). These
results indicate that all MLS wells were hydraulically connected to FW104 (Figure 2.3A).
The Br' recovery percentages were: FW101-2, 93%; FW101-3, 60%; FW102-2, 94%;
FW102-3, 93%; and FW026, 50% (Figure 2.3A). FW101-2, FW102-2 and FW102-3
received more than 93% of water injected to FW104, while F W101-3 and FW026
received 50% and 60% of water from FW104, respectively. The rest water was from

surrounding areas. Mean travel times from F W104 to the different wells were: FW101-2,

31

2.84 h; FW101-3, 18.4 h; FW102-2, 11.6 h; FW102-3, 3.7 h; and FW026, 46 h (23).
Therefore, groundwater injected to FW104 reached FW101-2 and FW102-3 more rapidly
than other two wells. The fraction of groundwater flow from injection well to each MLS
well and the mean travel time may significantly influence microbial community in the
MLS wells as discussed later. During the tracer test, COD concentrations in MLS and
FW026 wells increased after ethanol injection (Figure 2.38). The measured COD
concentrations in the MLS wells were significantly lower than those calculated based on
the Br' concentration and the COD/Br' ratio in the NaBr-ethanol tracer solution used. The
difference was likely due to biodegradation of ethanol and acetate in the subsurface.
Ethanol was observed only in the FW104 well during the tracer test. The COD in MLS
and F W026 was mainly from acetate (>80%) based on acetate vs. COD measured.
During the tracer test, sulfide concentrations increased in all MLS wells and U (V 1)
concentrations decreased in the same trend as shown in figure 2.2 (data not shown).
Another separate tracer test by injecting bromide to outerloop FW024 indicated that
innerloop extraction well FW026 received 17% of water injected to FW024 (23). The

infiltration of water from outerloop may also influence the microbial community.

32

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

80
704—A 'FWIO4 __
5 AFw026
J
60 Emma
_ O OFw101-3 ____.
E 50 w to
an I. O D 0FW102-2
é 40"?8‘9 o AFw102-3 #
Cl
0 0 Q
:2 30m: ..00 00
a o
g 20 o ‘ A
m JD; AAA
I 9 Q
150 200
I FW104
A FW026 —
cr FW101-2
Q . FW101-3
an O FW102-2—
E, A FW102-3
a FW101-2—
8 ——r=wror-3
FW102-2
- - - FW102-3

 

 

0
010 20 30 4 50 60 70 80 90100
Time (hr)
Figure 2.3. Tracer test performed with Bromide on day 801-803 shows the hydraulic

connection between injection well F W104, extraction well FW026 and MLS and
biodegradation of ethanol. A. Changes in bromide concentrations (23) B. Comparison of

measured COD (marks) and calculated COD (lines) concentrations based on COD vs Br-
ratio (2.46 g/g).

33

Microbial enumeration. MPN results indicated that after biostimulation for
nearly two years, increased levels of denitrifiers, SRB and F eRB were present in the
sediments in the bioreduced area in comparison with control well FW106 (Table 2.2) that
did not undergo stimulation and shares similar geochemical composition with the active

area wells (Table 2.1). After biostimulation, the microbial concentrations in the treatment
. . . . 7 8 6 8 5 7
area were (1n cells/g sedrment): denrtrrfiers, 10 -10 ; SRB, 10 -10 ;and FeRB, 10 ~10 .

he highest levels for all three trophic groups were found in inner loop injection well

F W104 where ethanol was injected. Relatively low levels were found in extraction well
FW026. Consistent with the tracer study, FW101-2 (more connected to the injection well
F W104 based on the tracer study) showed higher bacterial counts than FW102-2 (less
connected) (Table 2.2). MPN counts from FW106 (groundwater) showed levels of
denitrifier as low as 3.3 cells/ml groundwater (49). Neither SRB nor F eRB were found in
F W106. Comparison with sediment MPN at day 453 showed a decrease in SRB and

denitrifiers and no change for FeRB in FW104. MPN count at day 774 for all three
groups remained around 108 cell/g sediment in F W1 04 (49). Integration of MPN at day
774 and three earlier timepoints (days 278, 354, 453) for FW101-2 showed an increase in
the counts overtime with microbial levels at 774 around 107 cell/g sediment, one order

lower than F W104. Conclusions from the MPN analysis must be qualified because we
contrast sediments with groundwater samples. However, prior MPN studies using
samples from the same well show a one log difference in counts in sediment versus

groundwater samples (data not shown).

34

Table 2.2. MPN for three major trophic groups in the sediments of the treatment zone
(day 774) in comparison with groundwater from control well F W106.

 

Cells / g sediment

 

Well Denitrifiers FeRB SRB
FW104 7.23 x 108 9.40 x 107 1.53 x 108
FW101-2 1.54 x 108 2.05 x 107 2.06 x 108
FW102-2 2.39 x 107 5.48 x 105 1.87 x 107
FW026 1.10x107 1.93 x106 1.07x106
FW106“ 3.3 cell/ml ND ND

 

F eRB: Iron reducing bacteria; SRB: Sulfate reducing bacteria. ND: None detected. Wells

are arranged according to their descending hydraulic connection to the injection well
(FW104) based on the tracer studies. *Data on day 278 (49).

Sequence analyses. After discarding putative chimeric sequences, an average of
155 sequences per sample were used for each library. Rarefaction analysis at 97%
similarity levels (Figure 2.4) showed that the estimated coverage (rarified number of
OTUs divided by rarified Chaol estimator) ranged from 36 to 58% in the stimulated area
and was 93% in the background area (FW106). Even with low coverage in the
biostimulated zone, clear differences in diversity were observed. Highest diversity was
found in all the wells of the biostimulated area; working at 97% similarity for OTUs the
number of OTUs ranged from 41.2 to 91.4 (rarified values). Only 6.6 OTUs were found

in the background area FW106 (rarified value).

35

 

120

 

 

  
  
  

 

 

 

o —FW026
00 O -- 'FWlOl -2
1 - ,
I -—-—er01 -3 I’ll:
CI -- --FW102 -2 x",
.. I
5 80 -. ------ FW102 -3 ,r’
['5 <>— ~er04
3 60 ‘ alt — -FW106
3'
E
:z

 

b
O

 

 

 

 

120 150 180
Number of Sequences

Figure 2.4. Rarefaction curves of the 16S rRNA gene libraries constructed. OTUs were
defined at 97% sequence identity. The library from the untreated control (FW106) was
close to complete sampling (93% coverage).

Microbial community structure and major groups detected. All libraries were
dominated by members of the phylum proteobacteria, with Betaproteobacteria and
Deltaproteobacteria being the most dominant proteobacterial classes in the biostimulated
zone (Figure 2.5). In contrast, the background area was dominated by

Gammaproteobacteria. Bacteria belonging to groups known to include U(VI)-, nitrate-,
sulfate- and Fe(III)-reducers were detected in the active area but not in the background

area library.

Sequences from the genera Desulfovibrio, Geobacter, Anaeromyxobacter,

36

Acidovorax and Desulfosporosinus were detected in the libraries. These genera are
known to include U(VI) reducers and additionally can contribute with one or more of the
following activities: iron (III)-, sulfate- and nitrate reduction (Figure 2.6). The
contribution of the putative species responsible for the uranium reduction and their
richness (measured as number of OTUs in a genus) was highly variable (Table 2.3).
Geobacter and Desulfovibrio were detected in all the libraries of the active area. In five
of the six samples from the active area, the relative contribution of Desulfovibrio was
higher than that from Geobacter (Table 2.3). Anaeromyxobacter, Acidovorax and

Desulfosporosinus sequences were not present in all the libraries.

Nitrate reducers from different taxonomic lineages were present in the libraries of
the active zone but not in the untreated area (FW106). Most of the nitrate reducers were
members of the Proteobacteria phylum e. g. F erribacterium (2-27%), T hiobacillus (0-
29%), Sphingomonas (0-6%), Desulfovibrio (4-l6%), Azoarcus (0-5%), Acidovorax (0-
4%), and Pseudomonas (0-1%) among others. The denitrifying Acidobacteria Geothrix
was also detected in all the libraries of the active area (4-17%). Nitrate respiration is not
exclusive to these groups and the mentioned bacteria are also know for using other

electron acceptors,

Iron(III) reducers were represented by bacteria from three groups:
Deltaproteobacteria (Anaeromyxobacter and Geobacter), Betaproteobacteria
(Ferribacterium and Thiobacillus) and the Acidobacteria Geothrix. F erribacterium was

the most abundant FeRB contributing an average of 17% to the libraries.

Other commonly found soil bacteria like Acidobacteria, Actinobacteria,

37

Planctomycetes and Verrucomicrobia were also present in the samples. These bacteria

were present in the active area and not in the untreated area.

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l00°/o

8- ) r > >
890%" wig"; -
:— 80% ‘ —
E .
g 709 -- . '-

0 I ..4} '3: ‘j
E :31:
'5 60%, . ..0_’i —
8 331: ho}; .
LE 50% - -’ " '- coo -
o 9..
= 0.. .
.9 40% ‘ '-
E
E 30% " '-
§ I
o 20% ‘ '—
.2
§ 10% 4

0%

 

 

‘ FW026 FW101-2 FW101-3 FW102-2 FW102-3 FW104 FW106
Wells

U. Bacteria - Chloroflexi

  

I Alphaproteobacteria

:-I
Other D Bactemides D Gammaproteobacteria
5“; . _ . . .
R§ Actmobacterza Aczdobacterza @ Dellaproteobacterza
a F irmicutes [IIIII] U. Proteobacteria I Betaproteobacleria

Figure 2.5. Microbial composition of the clone libraries based on the RDP Classifier.
The “other” group category includes the phyla Spirochaetes, Gemmatimonadetes,
Verrucomicrobia, Chlamydia, Planctomycetes, Nitrospira, Cyanobacteria and the
proposed phyla, OP] 1, OPlO, BRCl and TM7. Bacteria that could not be assigned with
the 80% confidence bootstrap value were included in an artificial “Unclassified Bacteria”
(U. Bacteria) taxon. Wells are arranged according to their descending hydraulic
connection to the injection well (F W104) based on the tracer studies.

 

38

Figure 2.6. Neighbor joining tree showing the relationship of selected representatives
(shown in bold) from the groups similar to known U-reducing bacteria. Metabolic
abilities of the clones’ closest cultivated relatives are indicated. Non-model positions
from the 16S rRN A were masked and J ukes-Cantor distance correction used. Type strains
have a “(T)” label. Bootstrap values (10000 repetitions) are displayed if larger than 50%.
(+) Present in closest relatives, (-) absent in closest relatives, * activity found some
species of the genus, ? = unknown. The range of relative contribution to the different
samples is also shown.

39

 

8... _ 82 9.8 2882.8 $888.8
HOQOU -0 HOHQOOOQ -0 E negmitau maemmoaomaoKNaaoQ _
3?: .33 .flnaewnoaomnobaufillroc
mu: «~33—
+\. + + + + *1 + §=~ueaemue§auu§ _UaNN 82.883 3.3—=25 3.
3158:888— 8
€e§3odee¢Ro~oI

758288§JR
.\ 24 n: E 2838:8886 82
- . - . $188.88.? 8
.. + A. + + 5.333 a- n? «:95 d8 omeauoNotvoSH—socD ma
81588388
r €153.38— ? 8
. . .x. I. 1 82E 83.388 823.35%. S
A. A. 1T + I + + 50 9‘ Qagevg Aw ~"fivg— _INNAvfiBh—L
. 8 8..
(TON AbacuzoMSefifi afloacoRAEoLoucV
Gus—2. 383.,— 1.2:

a 8815 538;?— .
33.: once. 28 Eaauflamotfiaobaqu 3
- - . - 2 .15 :333... 8

+ 1. + + 8.8.8.835 88 28 83888.8 8,2838% 8
3158288 82
r «mac .8... 8.8.8.8833 8. 8

N m. - m0 AC aeooaheh-3o§o~oecoob I.
FeSESEJH—r
{cu—1c ~10 Soofizma 882586 2:

- - 315.8385...
- +\ + + + + 03336 AC nozowfifi 8.238% 9: cc—
§158243E 88

Nm .2932 $3236
..\o «1: _mm in mocgonaonk no no
315 8333.8: 8
822833. :38. E 8.838 822833... 88. _III
ccvm buzgoaoeex 3.238333%

mama—M a§~n£o3o§£~ LSoucozehoam 3 mad

 

 

 

fl

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.
+
+
+
.
+
.

fir

 

 

 

 

I

 

 

'
‘

Methanol
Ethanol
Acetate
U M)
sulfate
NItrate
Fe (III)

 

8:288. meg—8:5

5923 8 88:88 338w Bob A23 5 :30an 8038858952 @2028 no 8:588:28 05 wages—m ooh wEEoT—onnwmoz .e.~ 0.53...—

40

 

_ o o o _ N Rm €§s§§£ m §§§uw

 

 

.wzzmto .Q
o v N o o _ .Emfimfitau .Q n mxzwmoaomwobzme
m _ o. v o_ w 2 n:5§£am\ .b a 538mg
0 N m m w _ . azuzmmfiafih .V N amuuunoxkficxmuxw
MEEQEQAMQ. D
Emuzhmhfizx .D
mmzmmfifi .b
333 6
_ _ S _ _ m $852? 6 n 333.3
mzoaofifau .Q
o n v S 2 0 £83 .Q n etfiafiafi
3.95m m-_o_>>n_ N-No_>>n_ m-NS>?.m N‘s—33m veg/m
833—2
ax; :25 8a 8:33.80 gnu—om toga—0% “memo—U mDEO 3:00

 

68.8qu 03 and ® £03839 83o£ “memo? can Abra—Ea $3 v mDHO 26E: no .8583: 2F i=9:
n no.2 2: 5 8.958 Beam—.3. SP: 8295: 28.9 E .223» Seacoaéaiaaua =32:— ue Aekvueuafihaeo «>323— .m.~ ~35.

41

Diversity analyses. All libraries from the stimulated zone showed greater
diversity than the untreated zone based on both evenness and richness values as indicated
by the diversity indices used (Table 2.4). One single OTU dominated the untreated area
sample (FW106). This member of the Xanthomonadaceae family was also detected in a
metagenome experiment for well FW106, where it was shown to carry a variety of metal

resistance genes (13).

Table 2.4. Diversity indices for the different wells in the treated and untreated area

 

 

 

(F W106).
Index
Chaol Shannon (H) Simpson (l-D)

Well (LCI, HCI) (LCI, HCI)

FW104 125.5 (91.7, 200.6) 3.7 (3.5, 3.9) 0.97
FW101-2 76.5 (54.4, 137.9) 2.9 (2.7, 3.2) 0.90
FW102-3 87.8 (64.6, 147.7) 3.4 (3.2, 3.6) 0.95
FW102-2 250 (168.8, 415.1) 4.3 (4.2, 4.4) 0.99
FW101-3 114.3 (81.1, 192.9) 3.5 (3.3, 3.7) 0.95
FW026 124.9 (88.3, 210.4) 3.7 (3.5, 3.8) 0.97
FW106 6.0 (6.0, 6.0) 0.7 (0.5, 0.9) 0.31

 

LCI and HCI are rarefied 95% lower and higher confidence intervals (provided by the
DOTUR application). Wells are arranged according to their descending hydraulic
connection to the injection well (FW104) based on the tracer studies.

Diversity-based clustering revealed that the F W106 community formed a cluster
separated from the treatment zone wells (Figure 2.7). The two different clustering
methods used resulted in congruent topologies. High similarity was observed between
communities from FW102-3 and FW104, which were 53% similar according to a
calculated Bray—Curtis index. l—LIBSHUFF comparison indicated that these two libraries

were not significantly different (p=0.05). All the other pairwise comparison were non-

42

significant (libraries were significantly different, p=0.05).. Additional comparison of
F W104 and FW102-3 communities with the LIBCOMPARE function of the RDP
showed non—significant differences at (p=0.01) at all the different levels of taxonomy
from phylum to genus. No clear relation was observed between the alpha diversity

indices and the hydrology.

 

 

FW026
FW101-3
FW102-3
FW104
FW101-2
FW102-2
FW106

 

 

 

 

 

 

 

 

 

 

 

*-—01——* Sorensen Distance

 

 

FW026

FW101-3
{ FW102-3

FW 104

FW101-2

FW102-2

FW106

T Bray-Curtis Distance

,1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2.7. Clustering of the samples according to diversity patterns. The topology was
similar when using the Sorensen index (Presence/Absence) and the Bray-Curtis index
(Presence/Absence and abundance). The indices were normalized to account for
differences in total number of sequences per library.

43

DISCUSSION

1. Remediation of uranium and metals in the subsurface

Microbial reduction of metals including uranium has been the proposed
alternative to control the mobility of the contaminants in the groundwater (45). In this
field experiment, ethanol injection successfully created a reducing environment capable
of achieving uranium remediation and immobilization to levels below US EPA drinking
water MCL (49). The biostimulation process changed the structure of the microbial
communities from a small and low-diversity community to a more abundant and diverse
community with bacteria capable of reducing the contaminants nitrate, sulfate and U(VI)

in situ.

2. Microbial communities

2.a Major trophic groups detected

Microbial enumeration together with clone libraries revealed that after
biostimulation viable FeRB, SRB and denitrifiers had grown in the treated area where
uranium, nitrate, sulfate, ferric compounds and perhaps other compounds served as
electron acceptors. At least three major microbial trophic groups appear to be involved in
the bioremediation of the area. The first group, denitrifiers, can remove nitrate and
provide a favorable low redox environment for U(VI) reducers, FeRB and SRB. These
latter two groups may contribute to the remediation process by having members capable

of reducing not only Fe(III) and sulfate but also U(VI). Microbial enumeration indicated

44

that these three trophic groups were present in the sediments at high levels after
biostimulation. In contrast, the untreated area sample (F W106) showed the presence of
low numbers of denitrifiers (3.3 cells/ml) in groundwater but neither SRB nor FeRB.
Previous reports on groundwater MPN for this stimulated area (same wells, earlier
timepoints) also show these three groups but the microbial counts were lower that our
timepoint (49). The cell count in the untreated sediments is unknown but it could be
significantly low since DNA extraction was unsuccesful. Clone libraries permitted

detailed speciation of these trophic groups.

2.b. Carbon source as selective agent and the community of ethanol degraders

The use of ethanol as a carbon source likely increased the diversity of taxa and
physiologies. Injected ethanol provided both the carbon source and the electron donor in
the forms of ethanol, acetate (metabolic intermediate) and methanol (an impurity in
industrial ethanol). Electron donor consumption and metabolism were likely performed
by denitrifiers, FeRB and SRB (Figure 2.6). Based on groundwater analysis during tracer
study, we hypothesize that the SRB were mainly responsible for ethanol consumption
and acetate release between the injection well and MLS wells (where no nitrate was
detected and sediments showed black or dark green color), while FeRB and denitrifiers
utilized mainly acetate between MLS wells and the extraction well. At this well there was
evidence of groundwater infiltration from the outer loop and the sediments were green-
yellow in color. Close relatives of the SRB detected (Desulfovibrio spp. and

Desulfosporosinus spp.) utilize ethanol and not acetate when doing sulfate reduction

45

though few strains of both groups can also use acetate (19, 33). The release of metabolic
acetate can be used later by other species, including Geobacter. This hypothesis is
supported by sulfide accumulation during ethanol injection (SRB activity) (Figure 2.2)
and the detection of mostly acetate (and not ethanol) as carbon source in less connected
wells. The sequential utilization of ethanol and release of acetate likely create a niche for
acetate-consuming F eRB and acetate-consuming denitrifiers. The first group was
represented in our libraries by Geobacter spp., Geothrix spp. and Anaeromyxobacter spp.
These bacteria are capable of acetate and ethanol degradation mainly by Fe(III)-reduction
as well as by denitrification (2, 6, 34). The second group is represented by denitn'fiers
such as Acidovorax spp., and T hauera spp. These Betaproteobacteria members utilize
acetate and ethanol as electron donors and have been previously found in the

aboveground denitrification reactor used to pre-treat groundwater at this site (16).

Methanol utilization may play a minor role because its small contribution (5% in
the ethanol solution used), nevertheless it can be used for denitrification by almost all
denitrifiers and also for sulfate reduction by some Desulfosporosinus spp. (33) and

Desulfovibrio carbinolicus (25).

2.c. Other C sources and contaminants

In addition to the C sources used to stimulate the microbial activities, other C
sources such as aromatic and chlorinated compounds, humic acids and cellular

components are present in the contaminated area.

Aromatic (phenols) and chlorinated compounds (TCE) were detected in the

46

groundwater prior to biostimulation but mainly removed during site preconditioning (48).
These compounds are still present in the groundwater outside the treatment area and
could have infiltrated to the inner loop (based on tracer test results) and provided
additional carbon sources, electron donor and acceptors to the underground communities.
Several detected groups of different phylogenetic affiliation show some potential for
dechlorination or aromatic compounds degradation. In the case of dechlorination, at least
50% of the Geobacteraceae sequences were related to G. lovleyi, an isolate known to use
chlorinated compounds such as TCE and PCB as electron acceptors (39). Close relatives
of the chloro-phenol respirer Anaeromyxobacter dehalogenans 2CP-1 (34) and PCB
respirer Desulfosporosinus meridiei were also detected (33). Aromatic compounds can
also be used as carbon sources by some bacteria of the Acidovorax genus which was
detected in the sediments. The Betaproteobacteria Pseudomonas sp. strain P51 can
degrade chlorinated benzenes (43). Aromatic compounds and solvents present at the
beginning of the bioremediation operation could have influenced the initial communities
before the biostimulation began though their concentration is generally low and it is not

clear if they can support growth at levels present at the site.

Humic substances can be used as electron acceptors (21). Sequences related to the
humic-acid reducers Geothrix sp. (6) and G. humireducens (21) were detected in the
libraries. Reduced humic substances can potentially abiotically reduce UWI) far from the
bacteria and they have been shown effective in increasing F e(III)-reduction in subsurface

environments (27).

In addition, other bacterial groups not directly related to bioremediation, as far as

we know, were found in the sediments e. g. Planctomycetes, C hloroflexi, Actinobacteria,

47

etc. They are likely involved in the degradation or digestion of dead cells, soil humics,
and extracellular substances produced during biostimulation in the subsurface.
Chloroflexi members are known to grow in filaments, produce hydrogen and are
proposed to be involved in dead cell recycling (52). This group can potentially contribute
to the stability to the bacterial community by promoting the formation of biofilms and
can facilitate bioremediation by transferring electrons in the form of hydrogen to other
groups more likely involved in bioremediation such as Desulfovibrio spp. The
Chloroflexi group was the only group not belonging to the F eRB, SRB or denitrifiers

present in all the libraries from the active zone.

2.d. Putative genera involved in U(VI) reduction

Analyses of uranium oxidation state in sediments by XANES confirmed that the
decrease in groundwater U concentration was due to reduction of U(VI) to U(IV).
Elevated total U and U(IV) percentages of total U (18, 49, 50) were found in the inner
loop injection well FW104 consistently with effective bioremediation (Table 2.1). In this
study, we found a variety of previously reported UWI)-reducing bacteria present in the
reduced sediments including FeRB Geobacter spp. and Anaeromyxobacter spp., SRB
Desulfovibrio spp. and Desulfosporosinus spp. as well as the denitrifier Acidovorax spp.
Our results suggest that uranium reduction cannot be attributed to a single group and it is

very likely that this role is taken by several different bacteria.

The contribution to the uranium reduction based solely on the relative frequency

in our libraries and the reported literature ranks the contributor groups as following:

48

Desulfovibrio > Geobacter > Anaeromyxobacter > Desulfosporosinus > Acidovorax.

Desulfovibrio spp. were detected in all the samples of the active area and their
contribution peaked in wells with high percentage of reduced uranium. This group was
the most abundant group with reported uranium reducing abilities and was found in the
sediment samples. Even though there are no reports of U(VI)-reduction by the closest
relatives of the cloned sequence, the high frequencies of these bacteria in the sediment

clones (up to 16%) suggest a substantial role in the groundwater ecosystem.

Geobacter spp. were found in all sediment sample of the active area. Our
sequences were closely related to the uranium reducer G. lovleyi (39) and to the humic-
acid reducer G. humireducens (21). Additionally, Geobacter spp. was found to be
associated with U(VI) reduction at sites contaminated with uranium and U ore where

acetate was added (2, 28, 44).

Anaeromyxobacter sequences were detected in all sediment samples from the
active area except for FW026. These sequences were related to the known U(VI) reducer
A. dehalogenans (47) and to a clone from a uranium mine sediment were uranium

reduction was demonstrated (40).

Desulfosporosinus sequences were present in half the libraries of the active area
and the closest isolate to most of the sequences was D. orientis, a SRB also known for
reducing Fe(III), nitrate and U(VI) (26, 40). Despite their relative low contribution to the

fatal community, Desulfosporosinus spp. may play a minor role in U(VI) reduction and a
‘0) gger role in the long term stability of the reduced uranium as they can form spores and

survive under starvation conditions.

49

Acidovorax sequences were found in half the libraries of the active area. An
Acidovorax sp. was shown capable of U(VI) reduction in microcosms tests with
sediments from the FRC (29). Nitrate and nitrite has shown being able to re—oxidize and
remobilize Fe(III) (3 7) and the presence of this denitrifier could contribute to the removal

of these competing electron acceptors and ensure the stability of the reduced uranium.

Geothrix spp. sequences were found in all the libraries of the active zone. This
iron(III) reducer can use humic acids as electron acceptor, and has been found during
uranium reduction events in the F RC Area 2 using 16S rRNA gene microarrays and
enrichment studies (5). The Geothrix genus has not been characterized for U(VI)
reduction, but based on the number of clones retrieved, it was an important member of
the community. It is possible that its contributed to U(VI) reduction indirectly via

reduced humic acids or reduced iron(II) compounds.

Abiotic U(VI) reduction may also play a role under our operational conditions
(i.e. pH 5 .8-6.6 and HCO3- < 5 mM). Sulfide, the end product of SRB; and green rusts,
products of F eRB can both reduce U(VI) to U(IV) (15, 30). Indirect U(VI) reduction by
reduced humic substances can also be contributing an important piece given the constant
presence of F eRB humic-reducing Geothrix spp. in all the tested sediments from the

active area. Therefore, the activity of SRB and FeRB could also indirectly contribute to

the reduction of U(VI) and maintenance of a stable and low level of uranium.

3. Dynamics of the community

50

3.a. Patterns of diversity detected

Microbial diversity varied in the sediments of the treated zone (Table 2.4). The
utilization of standardized techniques in all of our community analyses allows us to
conclude that the observed differences were due to real differences in the community, and
not the consequence of PCR bias. The irregular pattern observed is likely due to the
heterogeneity of groundwater flow and the distribution of contaminants as shown by the
tracer studies. Microbial enumeration analyses for the three trophic groups studied
(denitrifiers, FeRB and SRB) showed a decline in cell counts going from the injection
well and its more connected wells, to the less connected wells (Table 2.2). Even though

replication did not allow for statistical conclusions, some apparent trends were observed.

The different approaches used to study the diversity and to compare the
communities yielded consistent results and the clustering analyses showed a topology
that was in agreement with the tracer studies depicting the groundwater flow. Because the
flow of the water is not homogenous, gradients of electron donors are expected as the
microbial communities consume and convert the injected ethanol solution. The hydrology
clearly affected the microbial counts (MPN) but no clear relationship between the water
flow and the alpha diversity indices was found. Though this may be due to incomplete
sampling, it is more likely that the ethanol injection created a selective pressure for
specific functions (like iron reduction, ethanol utilization, etc.) but not for specific
bacteria. Bigger gradients and more divergent communities would be expected in natural
systems where the water flow is not controlled. Having more divergent communities in
an area of remediation adds additional layers of complexity that can make the monitoring

of the performance more difficult. Thus, control of the hydrology is key to have a more

51

homogenous response to the bioestimulation.

When analyzing the carbon and electron acceptors, more ethanol-consuming
organisms were detected in the more connected wells. Consistent with the expected
sequential electron acceptor utilization (49), more denitrifiers where present in highly
connected wells and FeRB were more abundant in the wells with lower connectivity. The
exception to this pattern was Desulfovibrio, a SRB that showed high relative abundance
in more connected wells, possibly due to its ability to utilize the injected ethanol.
Desulfovibrio spp. were more abundant than Desulfosporosinus spp., also a SRB capable
of using ethanol. The presence of Desulfovibrio in the aboveground bioreactor (16) could
have given Desulfovibrio spp. an initial competitive advantage over Desulfosporosinus
spp. by a continuous inoculation of Desulfovibrio spp. carried in the treated water from
the bioreactor. On the other hand the ability of Desulfosporosinus to sporulate and
degrade methanol may account for its survival. Overall the microbial methods were in
agreement with the hydrological studies. However, more study is needed to understand

the relation quantitatively.

3.b. Emergence of previously undetected populations

After preconditioning by pH adjustment and removal of inhibitors, the ethanol
injection increased the microbial diversity of the subsurface as shown by diversity indices
that consider richness and abundance (Table 2.4). The activity of the new complex
bacterial community created favorable conditions where F eRB, SRB and denitrifiers

thrive in contrast with the untreated area where they were either absent or in very low

52

levels.

It is very unlikely that the microbial community observed was fully derived from
indigenous species that survived the extreme conditions of the area. The most likely
source of new colonists was the upper soil where nitrate is low and pH neutral. These
bacteria could have been transported by either natural groundwater infiltration or forced
recirculation during the treatment. Geobacter and Anaeromyxobacter spp. have been
found in FRC Areas 1 and 2 using iron-reducer enrichments with acetate and lactate (31).
Geobacter spp. and Geothrix spp. have been also detected in the F RC Area 2, using high
density 16S rRNA gene arrays (5). Pseudomonas spp. and Azoarcus spp. have been
detected in the FRC Area 3 with 16S rRNA clone libraries prior to biostimulation (9).

Therefore, these microorganisms are present in pH neutral soils at the site.

A second likely microbial source could be the aboveground bioreactor system.
This reactor worked for 400 days to remove nitrate from the groundwater and the treated
nitrate-free water was reinjected into the subsurface (48, 49). The reinjected water likely
carried some bacteria to the treated area although it was filtered. Sequences related to
Desulfovibrio, Thauera, Azoarcus, F erribacterium and Acidovorax have been previously
reported for the bioreactor (16) and were detected in the collected sediments of the active

area in this study.

Our results suggest that biostimulation efforts successfully fostered communities
comprised of a variety of bacterial groups (Geobacter, Desulfovibrio, Geothrix,
Anaromyxobacter, Desulfosporosinus and Acidovorax) involved in the groundwater

remediation process. This study provides a detailed view of the differences and

53

similarities among the microbial communities throughout the active area that correlated
with the path of groundwater flow depicted by tracer studies. The results demonstrated
that microbial communities can be established by in situ biostimulation with an electron
donor to achieve successful reduction of U(VI) concentration below US EPA MCL.
These findings contribute to an improved understanding of the composition, variability
and controls on microbial communities in the subsurface associated with a successful
bioremediation process and provide a foundation for future implementation and

monitoring efforts applied to this and other contaminated sites.

Acknowledgements

The microbial study was funded by the US DOE under grant DE-FG02-97ER62469 and
the fieldwork was funded by the US. DOE Environmental Remediation Sciences
Program (ERSP) under grant DOE C05-OOOR22725. Dr. Mary Beth Leigh was supported
by US NSF Postdoctoral Fellowship during this study. The authors thank Hui Yan,
Kenneth Lowe and Tonia Mehlhorn for help in field and analytic work and Benli Chai
from the Ribosomal Database Project for his help in formatting data. We appreciate Dr.
Chris Hemme for providing a groundwater DNA sample and Anthony Gaca for the

assistance in the construction of the clone libraries.

54

REFERENCES

Anderson, R. T. 2006. DOE genomics: Applications to in situ subsurface
bioremediation. Remediation Journal 17:23-38.

Anderson, R. T., H. A. Vrionis, I. Ortiz-Bernad, C. T. Resch, P. E. Long, R.
Dayvault, K. Karp, S. Marutzky, D. R. Metzler, A. Peacock, D. C. White, M.
Lowe, and D. R. Lovley. 2003. Stimulating the in situ activity of Geobacter

species to remove uranium from the groundwater of a uranium-contaminated
aquifer. Appl. Environ. Microbiol. 69:5884-91.

Ashelford, K. E., N. A. Chuzhanova, J. C. Fry, A. J. Jones, and A. J.
Weightman. 2005. At least 1 in 20 168 rRN A sequence records currently held in

public repositories is estimated to contain substantial anomalies. Appl. Environ.
Microbiol. 71:7724-36.

Ashelford, K. E., N. A. Chuzhanova, J. C. Fry, A. J. Jones, and A. J.
Weightman. 2006. New Screening Software Shows that Most Recent Large 16S

rRNA Gene Clone Libraries Contain Chimeras. Appl. Environ. Microbiol.
72:5734-5741.

Brodie, E. L., T. Z. Desantis, D. C. Joyner, S. M. Back, J. T. Larsen, G. L.
Andersen, T. C. Hazen, P. M. Richardson, D. J. Herman, T. K. Tokunaga, J.
M. Wan, and M. K. Firestone. 2006. Application of a high-density
oligonucleotide microarray approach to study bacterial population dynamics
during uranium reduction and reoxidation. Appl. Environ. Microbiol. 72:6288-98.

Coates, J. D., D. J. Ellis, C. V. Gaw, and D. R. Lovley. 1999. Geothrix
fermentans gen. nov., sp. nov., a novel Fe(III)-reducing bacterium from a
hydrocarbon-contaminated aquifer. Int. J. Syst. Bacteriol. 49 Pt 4:1615-22.

Cole, J. R., B. Chai, R. J. Farris, Q. Wang, A. S. Kulam-Syed-Mohideen, D.
M. McGarrell, A. M. Bandela, E. Cardenas, G. M. Garrity, and J. M. Tiedje.
2007. The ribosomal database project (RDP-II): introducing myRDP space and
quality controlled public data. Nucleic Acids Res. 35:D169-72.

55

10.

11.

12.

13.

14.

15.

16.

Colwell, R. K. 2005. EstimateS: Statistical estimation of species richness and
shared species from samples. Version 7.5. Persistent url
<http://purl.0c1c.org/estimates>.

 

Fields, M. W., T. Yan, S. K. Rhee, S. L. Carroll, P. M. Jardine, D. B. Watson,
C. S. Criddle, and J. Zhou. 2005. Impacts on microbial communities and
cultivable isolates from groundwater contaminated with high levels of nitric acid-
uranium waste. FEMS Microbiol. Ecol. 53:417-28.

Finneran, K. T., M. E. Housewright, and D. R. Lovley. 2002. Multiple
influences of nitrate on uranium solubility during bioremediation of uranium-
contarninated subsurface sediments. Environ. Microbiol. 4:510-6.

Cu, B., W. M. Wu, M. A. Ginder-Vogel, H. Yan, M. W. Fields, J. Zhou, S.
F endorf, C. S. Criddle, and P. M. Jardine. 2005. Bioreduction of uranium in a
contaminated soil column. Environ. Sci. Technol. 39:4841-7.

Hazen, T. C., and H. H. Tabak. 2005. Developments in Bioremediation of Soils
and Sediments Polluted with Metals and Radionuclides: 2. Field Research on
Bioremediation of Metals and Radionuclides. Reviews in Environmental Science
and Biotechnology V4: 157-183.

Hemme, C. L., T. J. Gentry, L. Wu, M. W. Fields, C. Detter, K. Barry, D.
Watson, C. W. Schadt, P. Richardson, T. Hazen, J. Tiedje, E. Rubin, and J.
Zhou. 2006. Analysis of a Microbial Community Metagenome from Uranium-
Contaminated Groundwater, 106th Gen. Meet. Am. Soc. Microbiol. May 21-25,
2006, Orlando, FL.

Holmes, D. E., K. T. Finneran, R. A. O'Neil, and D. R. Lovley. 2002.
Enrichment of members of the family Geobacteraceae associated with stimulation

of dissimilatory metal reduction in uranium-contaminated aquifer sediments.
Appl. Environ. Microbiol. 68:2300-6.

Hua, B., H. Xu, J. Terry, and B. Deng. 2006. Kinetics of uranium(V I) reduction
by hydrogen sulfide in anoxic aqueous systems. Environ. Sci. Technol. 40:4666-
71.

Hwang, C., W. M. Wu, T. J. Gentry, J. Carley, S. L. Carroll, C. Schadt, D.
Watson, P. M. Jardine, J. Zhou, R. F. Hickey, C. S. Criddle, and M. W.
Fields. 2006. Changes in bacterial community structure correlate with initial

56

17.

18.

19.

20.

21.

22.

23.

24.

25.

operating conditions of a field-scale denitrifying fluidized bed reactor. Appl.
Microbiol. Biotechnol. 71:748-60.

Istok, J. D., J. M. Senko, L. R. Krumholz, D. Watson, M. A. Bogle, A.
Peacock, Y. J. Chang, and D. C. White. 2004. In situ bioreduction of

technetium and uranium in a nitrate-contaminated aquifer. Environ. Sci. Technol.
38:468-75.

Kelly, S. D., K. M. Kemner, J. Carley, C. Criddle, P. M. Jardine, T. L.
Marsh, D. Phillips, D. Watson, and W.-M. Wu. 2008. Speciation of Uranium in

Sediments before and after In situ Biostimulation. Environ. Sci. Technol.
42:1558-1564.

Krieg, N. R., and J. G. Holt. 1984. Bergey's manual of systematic bacteriology.
Williams & Wilkins, Baltimore.

Kumar, S., K. Tamura, and M. Nei. 2004. MEGA3: Integrated software for
Molecular Evolutionary Genetics Analysis and sequence alignment. Brief.
Bioinform. 5: 150-63.

Lovley, D. R., J. D. Coates, E. L. Blunt-Harris, E. J. P. Phillips, and J. C.
Woodward. 1996. Humic substances as electron acceptors for microbial
respiration. Nature:445-448.

Lovley, D. R., E. J. P. Phillips, Y. A. Gorby, and E. R. Landa. 1991. Microbial
reduction of uranium. Nature 350:413-416.

Luo, J., W.-M. Wu, J. Carley, C. Ruan, B. Gu, P. M. Jardine, C. S. Criddle,
and P. K. Kitanidis. 2007. Hydraulic performance analysis of a multiple
injection-extraction well system. Journal of Hydrology 336:294.

Luo, J., W. M. Wu, M. N. Fienen, P. M. Jardine, T. L. Mehlhorn, D. B.
Watson, 0. A. Cirpka, C. S. Criddle, and P. K. Kitanidis. 2006. A Nested-Cell
Approach for In Situ Remediation. Ground Water 44:266-274.

Nanninga, H. J., and J. C. Gottschal. 1987. Properties of Desulfovibrio
carbinolicus sp. nov. and Other Sulfate-Reducing Bacteria Isolated from an
Anaerobic-Purification Plant. Appl. Environ. Microbiol. 53:802-809.

57

26.

27.

28.

29.

30.

31.

32.

33.

34.

Nevin, K. P., K. T. Finneran, and D. R. Lovley. 2003. Microorganisms
associated with uranium bioremediation in a high-salinity subsurface sediment.
Appl. Environ. Microbiol. 69:3672-5.

Nevin, K. P., and D. R. Lovley. 2000. Potential for Nonenzymatic Reduction of
Fe(III) via Electron Shuttling in Subsurface Sediments. Environ. Sci. Technol.
34:2472-2478.

North, N. N., S. L. Dollhopf, L. Petrie, J. D. Istok, D. L. Balkwill, and J. E.
Kostka. 2004. Change in bacterial community structure during in situ

biostimulation of subsurface sediment cocontaminated with uranium and nitrate.
Appl. Environ. Microbiol. 70:4911-20.

Nyman, J. L., T. L. Marsh, M. A. Ginder-Vogel, M. Gentile, S. Fendorf, and
C. Criddle. 2006. Heterogeneous response to biostimulation for U(VI) reduction
in replicated sediment microcosms. Biodegradation V17 :303-316.

O'Loughlin, E. J., S. D. Kelly, R. E. Cook, R. Csencsits, and K. M. Kemner.
2003. Reduction of uranium(VI) by mixed iron(II)/iron(III) hydroxide (green
rust): formation of U02 nanoparticles. Environ. Sci. Technol. 37 :721-7.

Petrie, L., N. N. North, S. L. Dollhopf, D. L. Balkwill, and J. E. Kostka. 2003.
Enumeration and characterization of iron(III)-reducing microbial communities
from acidic subsurface sediments contaminated with uranium(VI). Appl. Environ.
Microbiol. 69:7467-79.

Riley, R. G., and J. M. Zachara. 1992. Chemical contaminants on DOE lands
and selection of contaminant mixtures for subsurface science research. US.
Department of Energy, Washington, DC.

Robertson, W. J., J. P. Bowman, P. D. F ranzmann, and B. J. Mee. 2001.
Desulfosporosinus meridiei sp. nov., a spore-forming sulfate-reducing bacterium
isolated from gasolene-contaminated groundwater. Int. J. Syst. Evol. Microbiol.
51:133-40.

Sanford, R. A., J. R. Cole, and J. M. Tiedje. 2002. Characterization and
description of Anaeromyxobacter dehalogenans gen. nov., sp. nov., an aryl-
halorespiring facultative anaerobic myxobacterium. Appl. Environ. Microbiol.
68:893-900.

58

35.

36.

37.

38.

39.

40.

41.

42.

43.

Schloss, P. D., and J. Handelsman. 2005. Introducing DOTUR, a computer
program for defining operational taxonomic units and estimating species richness.
Appl. Environ. Microbiol. 71: 1501-6.

Schloss, P. D., B. R. Larget, and J. Handelsman. 2004. Integration of microbial

ecology and statistics: a test to compare gene libraries. Appl. Environ. Microbiol.
70:5485—92.

Senko, J. M., J. D. Istok, J. M. Suflita, and L. R. Krumholz. 2002. In-situ
evidence for uranium immobilization and remobilization. Environ. Sci. Technol.
36: 1491-6.

Singleton, D. R., M. A. Furlong, S. L. Rathbun, and W. B. Whitman. 2001.
Quantitative Comparisons of 168 rRNA Gene Sequence Libraries from
Environmental Samples. Appl. Environ. Microbiol. 67:4374-4376.

Sung, Y., K. E. Fletcher, K. M. Ritalahti, R. P. Apkarian, N. Ramos-
Hernandez, R. A. Sanford, N. M. Mesbah, and F. E. Loffler. 2006. Geobacter
lovleyi sp. nov. strain 82, a novel metal-reducing and tetrachloroethene-
dechlorinating bacterium. Appl. Environ. Microbiol. 72:2775-82.

Suzuki, Y., S. D. Kelly, K. M. Kemner, and J. F. Banfield. 2004. Enzymatic
U(VI) reduction by Desulfosporosinus species. Radiochimica acta 92:11-16.

Suzuki, Y., S. D. Kelly, K. M. Kemner, and J. F. Banfield. 2003. Microbial
populations stimulated for hexavalent uranium reduction in uranium mine
sediment. Appl. Environ. Microbiol. 69:1337-46.

Tabak, H. H., P. Lens, E. D. van Hullebusch, and W. Dejonghe. 2005 .
Developments in Bioremediation of Soils and Sediments Polluted with Metals and
Radionuclides - 1. Microbial Processes and Mechanisms Affecting
Bioremediation of Metal Contamination and Influencing Metal Toxicity and
Transport. Reviews in Environmental Science and Biotechnology V4:115-156.

van der Meer, J. R., R. I. Eggen, A. J. Zehnder, and W. M. de Vos. 1991.
Sequence analysis of the Pseudomonas sp. strain P51 tcb gene cluster, which
encodes metabolism of chlorinated catechols: evidence for specialization of
catechol 1,2-dioxygenases for chlorinated substrates. J. Bacteriol. 173:2425-34.

59

44.

45.

46.

47.

48.

49.

50.

51.

Vrionis, H. A., R. T. Anderson, I. Ortiz-Bernad, K. R. O'Neill, C. T. Resch,
A. D. Peacock, R. Dayvault, D. C. White, P. E. Long, and D. R. Lovley. 2005.
Microbiological and geochemical heterogeneity in an in situ uranium
bioremediation field site. Appl. Environ. Microbiol. 71:6308-18.

Wall, J. D., and L. R. Krumholz. 2006. Uranium reduction. Annu. Rev.
Microbiol. 60: 149-66.

Wan, J., T. K. Tokunaga, E. Brodie, Z. Wang, Z. Zheng, D. Herman, T. C.
Hazen, M. K. Firestone, and S. R. Sutton. 2005. Reoxidation of bioreduced
uranium under reducing conditions. Environ. Sci. Technol. 39:6162-9.

Wu, Q., R. A. Sanford, and F. E. Loffler. 2006. Uranium(VI) reduction by
Anaeromyxobacter dehalogenans strain 2CP-C. Appl. Environ. Microbiol.
72:3608-14.

Wu, W. M., J. Carley, M. Fienen, T. Mehlhorn, K. Lowe, J. Nyman, J. Luo,
M. E. Gentile, R. Rajan, D. Wagner, R. F. Hickey, B. Gu, D. Watson, 0. A.
Cirpka, P. K. Kitanidis, P. M. Jardine, and C. S. Criddle. 2006. Pilot-scale in
situ bioremediation of uranium in a highly contaminated aquifer. 1. Conditioning
of a treatment zone. Environ. Sci. Technol. 40:3978-85.

Wu, W. M., J. Carley, T. Gentry, M. A. Ginder-Vogel, M. Fienen, T.
Mehlhorn, H. Yan, S. Carol], M. N. Pace, J. Nyman, J. Luo, M. E. Gentile,
M. W. Fields, R. F. Hickey, B. Gu, D. Watson, 0. A. Cirpka, J. Zhou, S.
Fendorf, P. K. Kitanidis, P. M. Jardine, and C. S. Criddle. 2006. Pilot-scale in
situ bioremediation of uranium in a highly contaminated aquifer. 2. Reduction of
u(VI) and geochemical control of u(VI) bioavailability. Environ. Sci. Technol.
40:3986-95.

Wu, W. M., J. Carley, J. Luo, M. Ginder-Volgel, E. Cardenas, M. B. Leigh,
C. Hwang, S. D. Kelly, C. Ruan, L. Wu, J. van Nostand, T. Gentry, K. Lowe,
T. Melbourne, S. Carroll, M. W. Fields, B. Gu, D. Watson, K. M. Kemner, T.
L. Marsh, J. M. Tiedje, J. Zhou, P. K. F endorf, P. K. Kitanidis, P. M.
Jardine, and C. S. Criddle. 2007. In situ bioremediation of uranium (VI) to

submicromolecular levels and reoxidation by dissolved oxygen. Environ. Sci.
Technol. 41:5716-5723.

Wu, W. M., B. Gu, M. W. Fields, M. Gentile, Y. K. Ku, H. Yan, S. Tiquias, T.
Yan, J. Nyman, J. Zhou, P. M. Jardine, and C. S. Criddle. 2005. Uranium (VI)
Reduction by Denitrifying Biomass. Bioremediation 9:46-61.

60

52.

Yamada, T., Y. Sekiguchi, S. Hanada, H. Imachi, A. Ohashi, H. Harada, and
Y. Kamagata. 2006. Anaerolinea thermolimosa sp. nov., Levilinea
saccharolytica gen. nov., sp. nov. and Leptolinea tardivitalis gen. nov., sp. nov.,
novel filamentous anaerobes, and description of the new classes Anaerolineae
classis nov. and Caldilineae classis nov. in the bacterial phylum Chloroflexi. Int.
J. Syst. Evol. Microbiol. 56:1331-40.

61

CHAPTER THREE

Massive parallel sequencing of microbial communities in a uranium remediation site

INTRODUCTION

Metals and radionuclides are two major contaminants at Department of Energy
(DOE) facilities and waste sites (45). One of these sites, the former Y-12 National
Security Facility at Oak Ridge, Tennessee, stored its waste from uranium enrichment
operation in four unlined ponds of 9.5 million liter capacity each for over 31years. The
ponds received acidic uranium nitrate (~30% uranium nitrate) from its uranium
enrichment operations as well as waste from other DOE sites. After 31 years of
operations the ponds were neutralized, subject to nitrate removal, and its water treated for
nitrate removal before the ponds were capped (the contaminated sludges remained). The
contamination, however, was not contained within the ponds but expanded to nearby
areas through groundwater movement. The nearby Area 3 contains uranium in

concentrations as high as 250 uM in groundwater and 800 mg/kg in sediments (8, 58).

Since metals cannot be degraded, other strategies are needed to reduce their
bioavailability. One of these alternatives is microbial dissimilatory reduction, an
anaerobic process in which microbes use metals such as Fe(III), Mn(IV), Cr(VI) , Se(VI)
and radionuclides such as U(VI), Tc(VII) as final electron acceptors to support their
growth. Metal reduction involves changes in oxidation state than can trigger changes in
water solubility making some metals more soluble and easier to remove from the system

(e.g. Mn), and others less soluble, hence less mobile and less bioavailable, like in the

62

case of uranium. Since uranium reduction was first discovered it has been proposed as an

alternative to control U bioavailability (2, 3, 23, 29, 34, 58).

The ability to reduce U(VI) to U(IV) was first shown in the Fe(II) reducing
bacteria (FRB) Geobacter metallireducens (34) and later in the sulfate- reducing bacteria
(SRB) Desulfovibrio desulfuricans (33). Since then, U(VI) reducers have been found in a
variety of phylogenetic groups such as Delta-, Beta— and Gamma-Proteobacteria,
Firmicutes, Deinococci, Actinobacteria among others groups (55). Beside the ability to
reduce uranium, there is no common theme among uranium reducing bacteria as they

represent metabolically different groups such as FRB, SRB, ferrnenters, and denitrifiers.

Uranium reduction for bioremediation purposes has been tested successfully with
pure cultures of Fe(III)-reducing bacteria such as Geobacter spp.; sulfate-reducing
bacteria such as Desulfovibrio, Desulfoporosinus, Shewanella, and Desulfotomaculum
spp.; ferrnenters such as Clostridium; heterotrophs such as Deinococcus spp. and
denitrifying bacteria such as Acidovorax spp., among others (55). More recently, batch
serum bottles (18, 25, 61), microcosms (43), and sediments columns (20, 56) have been
used to study this approach in controlled conditions. Field scale studies have recently
started at a former uranium ore processing facility (Old Rifle site, now part of the
Uranium Mill Tailings Remedial Action program of the DOE) at Rifle, Colorado (54),
and at the Field Research Center (FRC) at Oak Ridge, Tennessee, the site containing

residue from the former Y-12 uranium enrichment facility (3, 29, 58).

Studies of uranium reduction have shown response of F RB and/or SRB, though

the stimulatory conditions were different. Geobacter was shown to be abundant at the

63

Old Rifle site, where uranium reduction was promoted through acetate injection into the
subsurface (3, 25, 42, 54). A different study also at the Old Rifle Site found enrichments
of FRB (Geobacter spp.) and also SRB (Desulfobacterales) (54). A recent publication
from this site has shown that uranium removal from groundwater continued even after
acetate injection stopped due to uranium absorption into biomass of F irmicutes-related
bacteria (40). Though not expected to be uranium reducers but dead-cell recyclers, their
effect in controlling uranium bioavailability is real and promising. In field studies from
the FRC Area 3 where ethanol was used as stimulatory electron donor, Desulfovibrio spp.
and Geobacter spp. have been associated to the uranium (10, 27); and Geobacter and
Geothrix have been found in enrichments using Area 2 samples with lactate as electron
donor (6). In anaerobic enrichments of contaminated sediments of a uranium mine,
sequences related to the SRB Desulfosporosinus, and fermentative Clostridium (53)

appeared to be major contributors to U(VI) reduction.

This present work studies the approach taken at the F RC Area 3 for uranium and
nitrate bioremediation. The bioremediation strategy was based on a hydraulic control
system that creates an area, isolated from groundwater intrusion and which allows
electron donors to be injected into the system to promote growth of local communities of
uranium reducers. The system was created using a series of wells that pump and
recirculate the groundwater. Over a two year period, the treatment was successful in
removing nitrate and conditioning the subsurface for the establishment of uranium
reducing communities (58). After conditioning, ethanol was injected and uranium levels
in groundwater dropped. X-ray Absorption Near Edge Structure (XANES) confirmed that

uranium was being reduced, discarding the possibility that uranium removal from water

64

was due to sorption or complexation with inorganic compounds (59). Microbial surveys
of sediments from six wells from Area 3 detected genera known to harbor U(VI)-
reducing members such as Geobacter, Desulfovibrio, Aneromyxobacter,
Desulfosporosinus, and Acidovorax after uranium reduction was established in the zone
(10) . Microbial counts, measured as Most Probable Number, correlated with the
hydraulic path; however, the proportion of U present as U(IV) did not follow a linear
relationship with the hydraulic connection (10). This suggests that factors other than
electron donor availability are influencing the presence and/or activity of the uranium
reducing populations. A more recent study tracked the microbial communities of F RC
Area 3 groundwater over a 1.5 year period and found that environmental factors such as
nitrate, uranium, sulfide, and ethanol were correlated with particular bacterial populations
(27). More interestingly, the engineering adjustments to control dissolved oxygen and
deliver nutrients to the subsurface were also found to be significant in explaining the

biological variability (27).

In this study we explored the role of environmental conditions, including
hydrology, on the structure of the sediment microbial communities of FRC Area 3 in a
period of high uranium reduction. The use of sediments allows for a more precise spatial
characterization than using groundwater samples where filtering large volumes of water
is required because of low biomass. Additionally, groundwater samples may not reflect
the attached communities present in sediments, especially in oligotrophic aquifers such as

the FRC (22).

For this purpose we thoroughly surveyed the microbial communities present in

the area to include a wider range of environmental conditions. We used massive parallel

65

16S rRNA gene sequencing to obtain a deeper survey of the microbial communities, and
higher confidence in the detection indicator species. We were able to detect groups
associated with uranium reduction, to explain differences in community structure with
environmental factors, and to manage with high confidence the massive number of
sequences to provide candidate groups likely important for the bioremediation of the

Area 3.

MATERIAL AND METHODS

Site description and sampling. Samples were retrieved from Area 3 of the DOE
F RC at the Y-12 National Security Complex, Oak Ridge, Tennessee, USA. In this area, a
hydraulic control system was used to promote in situ bioremediation in a controlled
fashion. The system consisted of an outer groundwater recirculation loop (injection well
F W024 and extraction well FW103) that isolated an inner groundwater loop (injection
well F W104 and extraction well FW026) preventing penetration by highly contaminated
groundwater from the source zone (Figure 3.1). Well F W105 was used to feed water to
the above ground treatment system and provided additional hydraulic control to the outer
loop (58). The outer loop was expected to respond to changes in groundwater movement,
while the inner loop’s shape and flow was expected to remain stable (5 8). The
remediation started on August 23th, 2003 (day 0). During the first 137 days, groundwater
was pumped and treated ex situ to adjust pH, remove Al, Ca and nitrate (via a

denitrification bioreactor). This process removed nitrate competition as a terminal

electron acceptor, avoided formation of stable Ca-U-CO3 products and clogging by gas

and biomass due to denitrification. Reduction of U(VI) to U(IV) was accomplished

66

through ethanol injection to the inner loop, starting on day 137. Multilevel sampling
(MLS) wells, F W100, F W101 and F W102 were used to monitor hydrogeology and
remediation performance over seven levels (-1 to -7). Except for the multisarnpling wells,
the wells were 14.6 m depth with screens between 11.28 m and 13.77 m. Multisampling

wells retrieved sediments from 15.2 , 13.7, 12.2, and, 10. 7 m for levels -1, -2, -3, and -4,

respectively. Water and sediments samples were retrieved on October 5th, 2005(day 775)

as previously described (10).

 

FW101

 

 

Groundwater I
flow f FW026

I
O
FW105 fl
\
FW103
3

. Multilevel sampling wellsL Milfilfi
, O

0 Well 1L 2 ;-

Figure 3.1. Scheme of the DOE Field Research Center Area 3 treatment system.
Wells FW104 and FW024 are injection wells, FW026,F W105 and F W103 are extraction
wells. Wells F W100, F W101, and F W102 are multilevel sampling wells. The controlled
hydraulic scheme creates an outer cell (2) that protects an inner cell (1) from intrusion of
contaminated groundwater from the background (3). The stimulatory electron donor,
ethanol, is injected through FW104 into the inner cell. The terms inner/outer loop and
inner/outer cells are used interchangeably throughout this article.

 

 

 

 

 

Chemicals and analytical methods. Chemical oxygen demand (COD), sulfide

67

and F e(II) were determined using a Hach DR 2000 spectrophotometer (Hach Chemical,
Loveland, CO). Anions (including NO3-, Br-, Cl-, SO42- and P043) were analyzed with

an ion chromatograph as previously described (10). Metals (Al, Ca, Fe, Mn, Mg, U and K
etc.) were determined using an inductively coupled plasma mass spectrometer (ICPMS)
(Perkin Elmer ELAN 6100), and U reduction state was determined with XANES as
described elsewhere (59, 60). The proportion of Fe(II) to the total Fe was measured as
HCl (10%) extractable. Two hydrology parameters that described the connectivity of the
different locations to the injection well were measured with a tracer study that introduced
a conservative tracer (bromide) at the injection well (FW104) were previously
determined (3 5). Tracer recovery (%) was considered a proxy for the amount of the
stimulatory electron donor received at each location. Similarly, mean travel time (from
FW104) was used as proxy for the composition of the electron donor as ethanol is

converted to acetate by the microbial communities.

DNA extraction and direct sequencing. DNA was extracted from 0.5 g of
sediments with the Fast soil prep kit (MoBio Inc., San Diego, CA, USA) following the
manufacturer’s instructions. Ribosomal RNA genes were amplified using a primer set
that flanked the v4 hypervariable region of the 16S rRNA gene at corresponding
Escherichia coli positions 563 and 802. Primers 563F (5’-
GCCTCCCTCGCGCCATCAG(barc0de)AYTGGGYDTAAAGVG-3’) and 802R (5’-
GCCTTGCCAGCCCGCTCAGTACNVGGGTATCTAATCC-3’) were a combination

of universal eubacterial primers and sequence adaptors (in bold) used in the amplification

68

step of the 454 Roche pyrosequencing process. The forward primer additionally
contained a short run of nucleotides used as bar codes. The bar codes were designed to be

distinct from each other by at least two nucleotide changes.

Primers were dual HPLC-purified (Integrated DNA Technologies, Coralville, IA).

Each PCR mixture contained 1 uM of each primer (dual HPLC-purified, Integrated DNA

Technologies, Coralville, IA), 1.8 mM MgC12, 0.2 mM for each dNTP, 3 ug BSA (New

England Biolabs, Beverly, MA), 1 unit of F astStart High Fidelity PCR System enzyme
blend (Roche Applied Science, Indianapolis, IN), and 10 ng of DNA template. Each
sample was run in triplicate. The PCR cycling conditions were: 95°C for 3 min; 95 °C for
453, 57 °C for 455, 72°C for 1min, for 30 cycles; 72 °C for 4 min. PCR products were run
in a 1% (wt/vol) agarose Tris-acetate-EDTA gel and bands between 270-300 bp were
excised. Bands for each triplicate were pooled together for gel purification with the
QIAquick Gel Extraction Kit (Qiagen Inc., Valencia, CA, USA). After gel extraction,
products were cleaned for a second time with the QIAquick PCR purification kit (Qiagen 3
Inc., Valencia, CA, USA) and eluted with 20 ul of EB buffer (10 mM Tris-Cl, pH 8.5).
Clean products were quantified using the Nanodrop ND-IOOO spectrophotometer

(N anoDrop Technologies, Wilmington, DE) and mixed in equal amounts, 20 mg each for

direct sequencing by Genome Sequencer FLX System (454 Life Sciences).

Pyrosequencing analysis. Sequences were processed with the Ribosomal
Database Project (RDP) Pyrosequencing Pipeline (14). Sequences were first trimmed to

remove the pyrosequencing adaptor sequences, and then sorted according to their bar

69

codes (used in the PCR amplification). Two quality filters were applied to remove
sequences with aberrant lengths and to discard sequences with more than two changes in
the primer portion. The RDP also provided a complete linkage clustering tool and
secondary-stucture—based alignment tool based on INFERNAL version 8.1

(http://inl‘emal.ianeliaorgf) and a 16S rRNA secondary structure model (9).

Community analysis. Sequences from different samples were classified using
RDP classifier with a 80% bootstrap confidence (57). To rapidly separate sequences into
taxonomic groups, RDP’s Classifier was used at a 50% confidence. e.g. to obtain all
F irmicutes from sample A. Sequences from that particular taxon from different samples
were later aligned and clustered (complete linkage) in order to track the distribution of a

specific group among samples.

Operational taxonomic units (OTUs) were defined from 0% to 20% sequence
dissimilarity (distance) to calculate richness of individual samples at different levels of
taxonomy. The results were rarified to the lowest number of sequences per sample from
the group of samples being compared. When an individual phylum or class was studied,
OTUs were defined at 5% distance (~genus level). When an individual genus was

studied, OTUs were defined at 1% distance (species level).

To compare individual samples according to their diversity patterns (beta
diversity analysis), a modified Sorensen index was used. This index has been corrected
by using abundance information to account for unseen species in the survey (11).

Clustering of samples was done using the average neighbor clustering algorithm on the

70

modified Sorensen index matrix. To find groups that represent a specific branch of the
clustering tree, indicator species analysis was used on the resulting topology (16). This
method was preferred because it provides a consistent approach to analyze the high
number of sequences per sample. The indicator value algorithm considers both frequency
of occurrence, and relative abundance, and its higher when a species is present in all the
samples and highly abundant in one specific group (16), this avoids considering indicator
species groups that appear only on one location. Since we test each OTU for significance
in the indicator species analysis, and multiple tests could increase the chance of finding
significant results by chance, we used a false discovery rate (FRD) approach (5). The
FDR estimates the chance of reporting a false positive in all the significant results (q-
value) instead of using excessively conservative corrected p-values. Q values were
generated from p-values using the freely available software package QVALUE Ver. 1.0

(51, 52) using the bootstrap method.

Multivariate analysis. Environmental variables were standardized (z-score) and
tested for correlation using a principal component analysis (PCA) on their correlation
matrix. This result was used to find variables correlated to uranium reduction, sulfate
reduction and iron reduction, to create a model on how the treatment changes the
geochemical characteristics, and to select environmental for direct gradient analysis. For
direct gradient testing, environmental variates were normalized using a log(x+1) or
arcsin(\/x) transformations, pH was left unchanged. Hellinger transformation was used in
some cases on OTUs abundance data to make it suitable even if when many zeros were
present (32). Principal coordinates analysis (PCoA) and non-dimensional scaling, was

performed on untransformed species data.

71

Statistical analyses were conducted using the R environment v2.80
(http://www.R-project.0rg.) with packages Vegan v1.8-8 (50), BiodiversityR (13),
Labdsv v1.3-1 (46) (indicator species analysis), and QVALUE v1.1 (available at

http://genomics.princeton.edu/storeylab/qvalue/) (51, 52).

RESULTS
Chemical and hydrological characterization. Analysis of DOE FRC Area 3

hydrology (Table 3.1), groundwater (Table 3.2) and sediments (Table 3.3) showed
differences in geochemical and hydrological composition due to the creation of the three
zones (inner loop, outer loop and background) by the treatment system. One of the
hydrology parameters, percentage of tracer recovery, an indication of how much water
from the injection well FW104 reached to a particular well, was significantly higher in
the inner loop (p<0.01) than in the outer loop. Mean travel time, the second hydrology
parameter and an indication on how long does it takes groundwater to move from the
injection well FW104 was not significantly different in the two loops. When grouping the
two multilevel sampling (MLS) wells from the inner loop (FW101 and FW102)
according to the depth they sampled, wells at level -1 and -4 received significantly less
water from FW104 (p<0.05), and were less connected (p<0.001) than wells from levels - -
2 and -3. This is likely because the injection at F W1 04 occurred at depths between levels

-2 and -3.

When compared to the outer loop, inner loop groundwater was richer in sulfate
(1K0 - 01), Fe (p<0.01), K (p<0.01), and chloride (p<0.05); and more depleted in sulfide

(1K0 - 05). This trend remained when compared to the background area (FW105) though

72

no statistical comparison was possible (only one sample from background well existed).
Samples from depths 2 and 3 of the inner loop were significantly higher in K (p<0.01),

and Fe (<0.05) and lower in chloride (p<0.05) than samples from levels 1 and 4.

Sediment analysis showed reduced U(IV) only in wells from the inner loop, and a
higher proportion of reduced Fe(II) in the inner loop (p=0.08). Sulfide levels of sediments
were also higher in the inner loop, although not significantly. These three trends
remained when adding the background well FW105, though no statistical test was

possible.

Overall, the bioremediation system seems to have influenced the composition of
the groundwater by creating gradients in electron donor availability (due to differences in
connectivity) and maintaining the isolation of the inner loop from contaminated
background groundwater. The differences found can be related to the activity of SRB and

F RB as observed in changes in sulfate/sulfide and total Fe/reduced Fe(II).

73

Table 3.1. Hydrological characterization of the DOE FRC Area 3 wells.

 

Tracer studies results

 

 

 

Well* Mean travel time Tracer recovery
(hr) (%)
FW104 0 100
FW101-1 387 18
FW101-2 2.9 93
FW101-3 17.9 60
FW101—4 298 18
FW102-l 222 3
FW102-2 1 1.6 94
FW102-3 3.7 94
F W102-4 790 6
FW026 8 1 50
"#07024 """"""""""""""""" 7345 """""""
FW100—1 21 1 9
F W100-2 223 12
F W100-3 43 18
FW100-4 211 8
"Ell/.193. __________________________________ 7-8 ......................... 1.9 ..............
FW105 710 6
Inner loop (meand: SE.) 174.1 :h 82.3 53.6 i 12.6
Outer loop (meant SE) 1 17.3 d: 44.1 10.3 :1: 1.9
Background 710 6
P-value* * 0.622 0.008

 

Results from hydrology studies using bromide (conservative tracer) on days 801 to 802

(November 1St to 2nd, 2005). Tracer was injected at well F W104, tracer recovery and

travel time at this well are 100% and 0 h by definition. The treatment system created
three different areas: inner loop (FW104, FW026 and multilevel wells F W101 and
FW102), outer loop (FW024, FW103 and multilevel well FW100), and a local

background (represented by FW1 05).*F or multilevel sampling wells the well location is

followed by the depth where the sample was taken. * *Comparison of inner and outer
loops was done with a two-tailed t-Test, alpha=0.05.

74

 

 

 

 

42 ~26 83 a; mac who a? 33 -.~ 8; so A: 8.4 8;:
who -o.o 33 ~55 :3 a3 mn~ 83 :.~ Sod So _2 ~qm SSE
of Sod 82. 83 :3 ~26 -.~ 82 e~.~ 32 8d a: 9% «.825
Ed Rod 83 :3 as 22 $~ 83 -.~ -o.o cod 9: an 22.5
who Sad 83 ~23 moo wed 8.~ 88d wm~ Sod cod 42 com woes/a
3m ~23 Sod ~23 86 3o 32 m~qo m~.~ 83 cod m~._ Se 33:
a; Ego Sod 23 $3 26 $~ ~85 ~m.~ _ 5d 23 w? :3 48.5
85 ES 89o 23 ~26 3o 3~ m~no 82 Sod 36 o~._ Ed 89?.
go 985 ~85 83 _~.o who :1 5% 3..~ ~56 cod m~._ 3.... 4-823...
~90 $3 Sod ~23 .3 m2 ~o.~ 53 :.~ 53 one o: as ~-~o§m
23 ES 98¢ :36 ad a: 8; $3 ~m.~ _ :3 e _ .c :2 as 32.5
2.2 $3 :3 23 -.~ 2.0 a; an; 3~ 23% cod 3o £8 78;:
$5 5.3 83 83 was 93 :.~ 83 a..~ 83 85 :1 3m 4-8.3.2
$5 Sod mood ~35 m2 :3 o~.~ _ _ 3 on~ 83 :.o 8.. Se 32.5
$5 E3 486 ~55 one on: ~o.~ 9.3 -.~ Sod as 2: m3 323...
om.~ Sod wood ~55 was 93 ~02 $3 ~v.~ o~no cod n3 who 78;:
85 ~86 ES :3 ~26 3o S.~ $3 v~.~ coed one w: 26 3;:
8 :2 on 2 m2 v. «z a -6 32.2 8.2.5 Beam 2; .133

 

N225 533.6555 mo :oEmomEoU 325.20

 

.m 8:. um... won :3... £2. 8...: 3:55....» .3 858388 .3255 .~.m 03..
s

75

.modumna—m 48,—L 3:8 -95 a was: 25¢ 33 302 8:5 25 55: mo camtmmfioouw * dew—8 $3 29:8 2:

20:3 fiaoc 05 an 330:8 2 5:80. :03 2: 233 wEEEmm 35:22: 8m * A3. 3m 3 338835 vcsohwv—omn was .89 3m

:95 ESE—:8 can 8:5”— .vmoamv 82 530 Aoomam Ea 9:23 m=oB 65:22: ES amok/n— .vogzmv Q02 55: ”£8“ “:80ng
085 @888 88?? Sesame; .Eq 38x3 2:. 5 Ba mcosmbcoocoo .Amoom .Em 89809 SE mac no 3552 803 838nm

 

 

 

 

wnmd cued w 56 _Nwd Sod mood god 026 mmod Cvd god mood omod **o=_w> -m
S: A as .o coed mum ~ .0 mmd mmd 3N; nwmmd voomd So. .N c Q..— 86 cczewxomm
To 36 o de vod mod Ed ommd Nod womd cod mod Ed Amm HSBEV
H H H H H H H H H H H H H Q8— 330
mm; vod mood Rod mvd amd Sum $6 and Nomd cod 9: vafi
mm; god mood Sod mud Ed wmd .36 mod mmva 36 _d _md Amm HamoEV
H H H H H H H H H H H H H 82 BE:
S.N owed fled _mod 36 wwd Wm cwwd own mama 36 no; 36
«0 £2 on _< 32 M «Z D -5 @8me 0325 Sufism in *5?
92:1? $339580 mo $583.00 32825
)/
Q.

.53 .~.~ 2...;

76

Table 3.3. Chemical composition of sediments from DOE FRC Area 3.

 

Chemical Characteristics of Sediments

 

 

 

U 1v / Total Fe 11 /
We11* Bioactivity (mg/kg) To(tal)U Fe T6191 )Fe (cg/(111;) fig:
(%) (g/kg) (%)
FW104 High 10.3 61 199.1 0.53000 337.9 23.000
FW101-1 Low 0.051 0 48.3 0.02640 8.6 0.000
FW101-2 High 1.249 54 43.3 0.27400 23.9 0.667
FW101-3 High 1.825 51 29.4 0.29400 24.2 0.265
FW101-4 Low 0.619 0 42.6 0.04660 27.1 0.000
FW102-1 Low 0.041 0 39.4 0.00560 22.6 0.000
FW102-2 High 0.517 17 33.7 0.30900 20 0.557
FW102-3 High 0.875 30 36.4 0.28700 22.5 0.431
FW102-4 Low 0.481 0 37.1 0.00660 12.7 0.000
"If-W939 ............. 9.4.3.9193? ......... L??- .......... 9 ......... 9121.-----9219999 ...... 9.2.27. ...... 99.3.9---
FW024 Low 0.371 0 50.8 0.21900 33.2 0.070
FWIOO-l Low 0.215 0 55.6 0.00142 10.5 0.000
FW100-2 Medium 0.98 0 32.4 0.12640 7.5 0.000
FW1 003 Medium 1.098 0 36.2 0.06360 13.4 0.000
FW100-4 Low 1.501 0 34.4 0.05450 23.8 0.000
“Ii-“(ESP ............... 9.9-“Y .......... 9.26.9.3 .......... 9 ......... 9i-2-----9.-.99§1.9 ...... 3.2.-.1 ....... 9.9-99---

FW105 Low 0.978 0 47.8 0.00480 30.7 0.000
Inner loop 1.718 21.3 55.6 0.19272 52.2 2.495
(mean :1: SE) i i i i i i

- 0.97 8.1 16 0.05515 31.8 2.28
Omr 100p 0.8 42.4 0.07834 20.1 0.012
(mean t SE) i i i i i

- 0.197 0 3.9 0.03374 4.6 0.012
Background - 0.978 0 47.8 0.0048 30.7 0
P-value* * - 0.377 ND 0.441 0.157 0.343 0.304

 

Samples retrieved on day 774 (October 5th, 2005). Treatment system created three

different areas: inner loop (FW104, F W026 and MLS FWIOO and FW200), outer loop
(FW024, F W103 and MLS FW100), and a local background (represented by FW105).

*F or multilevel sampling wells the well location is followed by the depth where the

samme was taken. * *Comparison of inner and outer loops with a two tailed t-Test,
a1Pha=0.05.ND, mathematically not defined. Bioactivity is a qualitative descriptor based
on Most Probable Number (MPN) results for iron reducing bacteria, sulfate reducing
bacteria and denitrifiers. This variable is used more as a general guide rather than a
Strictly quantitative indicator. The MPN methodology is described elsewhere (10).

77

Correlation of environmental variables. Correlation analysis showed that the
environmental variables measured were highly correlated. PCA on the correlation matrix
(Figure 3.2) showed that most of the variables were collinear with at least one other

variable as shown in the correlation matrix (Table 3.4).

Hydrology factors (mean travel time and tracer recovery) seemed to influence key
variables such as pH. Mean travel time was correlated with pH, A1 and U in groundwater.
It is not clear if U levels in groundwater were influenced by pH directly or controlled by
microbial activity. The percentage of tracer recovery was correlated with sulfide levels in

groundwater, potassium and iron in groundwater.

Reduced uranium in sediments (U(VI)/Total U) was strongly correlated with
bioactivity and reduced iron and moderately correlated with sulfide levels in sediments,
uranium levels in sediments, iron in sediments and chemical oxygen demand (COD).
Nitrate was highly correlated (r>0.9) with Na, chloride, Mg, Mn, and Ca. Nitrate was also
correlated with chloride and negatively with sulfate but with smaller strength (F 0.72 and
-O.72 respectively). The correlation of reduced U(IV) with variables related to FRB and
SRB and not with nitrate reduction suggests that FRB and SRB are more important for

U(VI) reduction at this location than nitrate reducing populations.

Based on the correlation analysis, the following variables were kept: pH, uranium

in sediment, tracer travel time, nitrate, and COD in sediments.

78

 

 

  
   
 

01-7 Sulfate

bu

P U(Sed)

"‘ Fe(sed) Sulfide(Sed)
COD(Se )

w

9 Red.U(Sed)
.c-J Red.Fe(Sed)
° Bioactivity -
Sulfide (Gw) .

 

 

 

-° Tracer recovery Chloride
U" i
Nitrate
Ca Mg
I l i 1 1

-1.0 -O.5 0.0 0.5 1.0 1.5
PCI

Figure 3.2. Correlation plot of environmental variables measured for 15 locations at
DOE FRC Area 3. Variables were z—score transformed prior to correlation analysis.
Bioactivity was converted as following, High=3, Medium=2 and Low=1.

79

ts and

imen

f groundwater, sed

lcharactenstics 0

Table 3.4 Correlation of chemica

hydrology of the DOE FRC Area 3 wells.

 

.0269: Bow:- .mm
053 $295 :60: . _ N
:83 6.653. .8

$63 moo .2

:83 tom museum .2
sea em .2

663 .5 Bose-é .e.
:63 3 .2

$3385 .3

6 .2

:2 .2

is em .:

_< .2

w: .6
v. L
NZ 5

9,3 2 .e
625.6 .m
35:2 .v
633 .m
63:3 .~
:d ._

 

03min?

dmd
vod-
vod-
odd-
2d
Nod-
vmd
mod-
:d
mmd
2d
hmd
end-
dmd
_ _d
_md
5.?
mod-
dmd
mmd
mmd-
dd.—

1

mmd-
N. d
mdd-
_dd-
odd-
edd-
mNd-
:3
Ed-
mud-
Ed-
3d-
mvd
dud-
de-
vod-
nmd
mod-
2.d-
mmd-
dd.—

dad
mvd-
dvd
_vd
_vd
mmd
ovd
mvd
_md
mmd-
mod
and
Ed-
3d-
and
m _ d-
de-
w _ d-
w _ d-
dd.—

emd-
add
mod-
N. d-
de-
med-
2 d-
m _ d-
mmd-
wad
Nod
ddd
fidd-
odd
Ed-
3d
5d-
mud
dd. _

cmd-
vmd
Ndd-
fidd-
bod
mod-
N_ d
wdd-
N _ d
mod
med
mod-
_dd-
wed
o _ d-
mud
dd
8. _

dvd-
Ndd

add-
add-
mmd-
udd-
mmd-
hod-
mmd-
odd-
3d
2d-
mwd
d. d-
_ _d-
de-
dd.—

dmd-
_Nd-
mdd-
add-
udd-
Nod-
_ _d-
Ed-
2d-
wwd
bud
Nod
Rd-
mod
mmd-
do.—

and
mmd-
Ed
2d
2d
3d
oNd
m2...
2d
3d-
mod-
3d
3 d-
und-
8..

dmd- and. and cod dmd-
3d and cvd- odd bdd
add- :d- mmd ddd 2d-
N_d- dod- «.md Ndd- Ed-
de- 3.? «_d 2d- mmd-
Ndd- :d- and Vdd Nod-
Ed- 2d- dmd Vdd- N_d-
2d- hod- _md vod- :d-
mmd- de- odd Ed- mmd-
odd Cd- mdd- dwd dd.—
wwd vod mmd dd.—
Ndd- 3.? dc.—
2d- do.—
do.—

9 10 l l 1 2 l3

 

80

Table 3.4 (cont’d)

 

5‘68. sum; .mm
9:: _o>m.= cam—2 .5
3me 853m .8
€me 80 .2

8% ..£ 803% .3
:3 ma .2

€me .2 “.838”. .9
sad 2 .9
€885 .3

8 .9

:2 .NF

is 8 .2

_< .2

92 .m

x .Q

«2 s

is a Q

$8.5 .m

222 L

853 .m

03.5 .N

In d

03a€m>

 

cmd mvd vvd wmd 3d ovd mvd de-

dmd- 2d-
wmd wad
mmd mod
mad dud
mmd 3d
mad 3d
mvd dd. _

dd.—

Zd- 3d-
and mod
and mod
mad and
dmd dd.—

dd.—

Rd. 3.? 2d- dd.—
wod odd dd.—

ccd dd.—

dd.—

dd. _

14 15 l6 17 100 19 20 21 22

 

*Percentage of U(IV) to total U. * *Percentage of Fe(II) to total Fe. Variables 1 to 13

were measured in groundwater (gw) samples, variables 14 to 20 in sediments (sed), and
variables 21 and 22 in tracer studies. Pearson correlation is significant if greater than

0.482 (two tailed, alpha=0.05)

81

Pvrosecmencing results

The 16S rRNA gene surveys produced a total of 97, 610 sequences for 17 samples
that covered a wide range of conditions in the F RC Area 3 including wells from the inner
and outer loops, three MLS wells at four different depths and one well (F W105) outside
the treatment area. The high throughput provided a wide view of the microbial diversity
present at the site (Table 3.5). The surveys were powerful enough to detect phyla poorly
represented in the databases (Table 3.6). In some of these cases there were more
sequences in the libraries for a given phyla than in the complete Ribosomal Database

Project dataset, though RDP Classifier tools is better trained for well studied groups.

Two samples yielded a low sequence count and were discarded from the analysis
(F W100-l and F W101-1). Without considering those samples there was no significant
difference in microbial richness recovered between inner and outer loops. Rarefaction
curves (Figure 3.3) suggests that even with thousands of sequences per sample the
sampling would be incomplete. This was not surprising given the high microbial diversity
expected for non-extreme environments (17, 26). Estimates of coverage (data not shown)
were similar to previous experiments that used ~100 clones per library (10). Given that
we obtained 20 to 80 times more sequences per sample, it is clear that rarefaction curves
are only informative when a plateau has been reached, otherwise the coverage estimation

is deemed to fail, especially if using non-parametric models such as Chaol.

82

Table 3.5. Sequences per sample and number of groups detected at different levels
of taxonomy.

 

 

 

Well Sequences Phylum Class Order Family Genus
FW104 6671 1 8 20 34 66 99
FW101-1* 52 7 10 11 16 15
FW101-2 6769 15 19 34 67 94
FW101-3 6458 18 20 34 34 107
FW101-4 5450 17 21 40 91 146
FW102-1 8566 18 21 41 36 165
FW102-2 8881 18 20 39 92 170
FW102-3 5954 16 20 37 75 118
FW102-4 7118 20 25 43 91 143
EYE/92.6 _______ §L3§ _____ 1.5. _ _ - _ .22 ..... 3 £5 ..... 7.4 _____ 1.09 _ _
FW024 8121 15 18 36 68 92
FW100-1* 83 13 16 13 16 17
FW100-2 2050 17 21 40 79 95
FW100—3 5564 18 24 41 82 121
FW100-4 5451 20 34 40 89 169
-EVYLQ} _______ 23.4.5 _____ L7_____29 _____ 3 .7 ..... 8.0 ..... 1.2.3--
FW105 6889 17 21 39 77 122
Whole set 97610 26 29 57 151 251

 

Classification based on RDP Classifier at 80% bootstrap confidence. *Discarded from
later analyses.

Table 3.6. Detection of phyla poorly represented in 168 rRNA databases in libraries
from FRC area 3 samples

 

 

Found in Present in Average identity to closest
Phyla library RDP* isolated relative
WS3 7 203 84.5%
TM 7 185 883 77.1%
Tenericutes 7 2570 91 .8%
0p] 0 2 369 82.0%
OD] 545 195 74.6%
Nitrospira 6 2045 95.6%
Lentisphaerae 1 6 271 92.7%
Gemmatimonas 2099 1 572 89.9%
F usobacteria 23 1929 86.6%
F ibrobacter 23 3 64 95 .6%
Deinococcus— T hermus 13 1305 95.0%
C hlamydiae 4076 654 91.3%
BRC] 121 86 82.2%

 

*RDP Release 10, Update 6 of 12/3/2008 containing 715,637 aligned 16S rRNA gene
sequences. Comparison is against the complete RDP dataset (Both partial and near-full
length sequences, from clones or isolates, of good or suspicious quality. Sequences were
assigned to a particular phylum using RDP Classifier with 80% bootstrap confidence.

Richness (number of different OTUs at 95% similarity) correlated with the

. 2 .
number of sequences per Site (r =0.6) as expected when coverage 1S not complete.

However, richness in some samples was high albeit low sequence counts. This is
probably due to a more even community structure that makes it easier to capture a wider
range of richness (Figure 3.4), as well as incomplete sampling. Nevertheless, the
communities were generally dominated by few OTUs, in 10 most abundant OTUs in any
sample accounted for 40% of the sequences for that particular sample. Singletons,
sequences with abundance equal to one in one library, represented on average 50.1% of

the richness for any given site.

84

 

 

   

 

 

 

 

 

1400
.\
1200 «“31
a "FW103
E1000 FW100-3
8 FW102-4
g 800 - FWIOO-Z
g FW105
3.600 _ FW100-2
'3 FW104
D FW102-3
400 - _FW026
0 l I r l I I I I I 1
O O O O O O O O O O O
O O O O O O O O O O
O O O O O O O O O O
.— N m V W \O l\ 00 05 2
Sequences

Figure 3.3. Rarefaction curves for FRC Area 3 samples. The shape of the curves
suggests that a plateau has not been reach and incomplete sampling may occurred.

85

'D'S/Smax +

 

gowsvwnscvwwsmw
;§§§§§§§§§§§§§§
u—BBBBBBBLHLHBBBWU-
mmmmmmm [amt-L.

Figure 3.4. Richness estimations for samples from DOE FRC Area 3 as influenced
by evenness and sampling effort. S = Richness (number of different groups), n =

number of sequences sampled, J-eveness = Eveness indicator (range 0 to l), 11max and

Smax refer to the maximum number of sequences and richness the complete set of
samples, respectively. Changes in sampling correlated with changes in richness. In some
wells, e.g.FW100—2, richness was still high even when sampling was low due to a more
even community.

Hierarchical clustering of samples

Clustering of the samples based on their microbial community yielded four
clusters (Figure 3.5). The clustering was consistent with differences in bioactivity and
connectivity. Cluster A contained only samples with low bioactivity and poor
connectivity to the injection well. Some samples from this group came from the inner
loop but at depths (levels -1 and -4) that make them less likely to have received the
electron donor (injected between levels -2 and -3). Cluster B contained all the samples
with the highest bioactivity and U(VI) reduction ability. Three subgroups were present in
cluster B, with subgrouping consistent with pH levels suggesting that even when the

Communities respond to the treatment in a similar fashion, there is variation that can be

86

attributed to differences in pH, non homogenous starting populations or other variables.
Clusters C and D contained wells with medium to low bioactivity and intermediate
connectivity. Samples present in clusters B, C and D showed similar pH levels within

clusters but cluster A did not, having a wide range of pH.

— FW105 (4.66)

— FW103 (5.92)

— FW102-1(6.4) A

— FW101-4(5.6)

— FW100-4(5.8)

~-— FW102-4(4.4)

_, FW104(5.8)
—_{: FW026 (5.7) B1

— FW101-3 (6.1)

_L— FW101-2(6.2) B2 B

FW102-3 (6.2)
,_ FW102-2 (6.5) B3

— FW100-3 (5.9)
— FW024(5.9)| J C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

— FW100-2 (5.5) D

l l l l l
0.2 0.3 0.4 0.5 0.6
Sorensen abundance-based distance

 

Figure 3.5. Clustering of samples by abundance-based Sorensen index. Well name
and (pH) shown. Four clusters were defined (A to D). Cluster A was comprised of wells
with low activity and connectivity to the carbon injection well. Cluster B has the most
active communities and the highest connectivity. Clusters C and D contained wells with
medium activity. In clusters B and CD, the subgroups seemed to be influenced by pH.
Hierarchical clustering used an average neighbor algorithm.

87

Microbial communig structure

The microbial communities were dominated by Proteobacteria and
Acidobacteria, this agrees with previous surveys of the site (10, 27). On average
Proteobacteria and Acidobacteria together contributed 52% of the sequences, with
combined contributions ranging from 24% at FW100-2 to 74% at F W024. Proteobacteria
contributed 39% on average, with a maximum of 71% at F W024 and minimum of 18% at
FW100-2. At F W024, 46% of the sequences were related to Sulfuricurvum, an anaerobic

sulfide oxidizer that can use nitrate and hydrogen as electron donor (30).

Acidobacteria contributed 13% of the sequences with a maximum of 33% at
FW026 and minimum of 3% at FW026.The Actinobacteria contribution was around 6%
with the exception of one well (FW100—2) where its contribution was 36%. The
Proteobacteria contribution varied the most, followed by Actinobacteria, Acidobacteria

and F irmicutes (Figure 3.6).

88

 

O\ \l

O O
l

0

(%g,
0

 

A
O
l
|--l I I I
O
0

Relative abundance
N U)
o d
l l
0
-- I 0

#
O
l

 

 

 

 

 

 

 

 

 

O
l

 

|--
'[li'

.115 " Inj-"I
(DI
[i' o 0
III
HI
.0
II

01d
13V
”V

Figure 3.6. Relative contribution of most abundant phyla to the microbial
community from 15 samples from DOE FRC Area 3. Boxplots show medians, lower
and upper quartile (box limits), range of observations (whiskers) and outliers if any
(circles). Pro = Proteobacteria, Aci = Acidobacteria, Act = Actinobacteria, Fir =

F irmicutes, Chla = Chlamydiae, Ver = Verrucomicrobia, Bac= Bacteroidetes, Gem =
Gemmatimonadetes, Chlo = Chloroflexi, Pla = Planctomycetes, Other = other phyla.
Unclassified bacteria are not shown.

89

When comparing the structure of samples in the inner and outer loops,
Acidobacteria (p<0.01), Chloroflexi (p<0.05) and Chlamydiae(p<0.05) were significantly
higher in the inner loop, while F ibrobacter was significantly lower (p<0.05) in this zone.
F ibrobacter is a rare phylum originally found in the gut of herbivores where it plays a
role in the anaerobic degradation of cellulose. It has been detected recently outside the
gut in landfills (3 7). Our putative F ibrobacter sequences were on average 95.6% identical

with its closest isolated relative in the RDP.

When comparing the four sample clusters produced in the clustering analysis,
cluster D was dominated by Actinobacteria and cluster C by Proteobacteria (Figure 3.7).
Clusters A and B shared a similar profile, however cluster B had a significantly higher
proportion of Chloroflexi (p<0.01), Chlamydiae (p<0.01), and BRCI (p<0.01), and a
significantly lower proportion of Verrucomicrobia (p<0.05), Actinobacteria (p<0.05),
TM7(p<0.05), and Bacteroidetes (p<0.05) than cluster A. Chloroflexi are a group of
facultative aerobes known to subsist on carbohydrates and yeast extract and thought to
have a role in cellular matter degradation (62), therefore, its abundance in wells with
higher bacterial biomass could be due not to its ability to utilize the stimulatory carbon
source but to recycle dead cells. Verrucomicrobia are ubiquitous in soil, with tolerance to
acidic conditions and oligotrophy and thus a good competitor under these limiting

conditions at the FRC.

90

>
CD
0
U

100%

90%

--....... ‘
U.;-__—_—_——

80%

$5:

I
I

O
.\
V — ‘
E’ ' 4 " —
Ad

E 70% I - — - 2,:
.9 'N
a o W I I I
3 60A) _, I

II.

M m .. '.' -

a 3:1 «"3 N” “M \A. - vs.
3500/44-W M.. MW av-w —
o .. U. A, H. M. H, a. M:- .,. I

= :': I ‘A "
.'. . U' V '. V V‘ V _ “V V‘
-:-..v~we-v~ Mw-MN m
40%.;.;...;.vx. .9“ MWWM .~ ;. -——-
= In. 0.: In 1'. m '
g -:- :-: -: w ..v~ “WW“ -2-
0 1:3 :3: 1' P53 :1: -' :3' W N '11 "I
0 30% ~3:3~:1: :5 3:3 :3: .3 ': W W '--
5 555 33: 5= *1‘~
M

20%

n.’
a)

10%

 

 

 

 

 

 

 

 

 

0% :-: H-:- :u ':.:'u:-':-' '-:.':- ':-:'-:-I ' I I
New“ , W?’ W” B”

sslsiiaa tis‘aaag‘“ a ‘1“ at“
I Other Bacteria El Verrucomicrobia

I F irmicutes D Actinobacteria
Acidobacteria El Proteobacteria

Figure 3.7. Relative contribution of the five most abundant phyla present at samples
from the FRC Area 3. Samples are ordered into four clusters (A, B, C and D) according
to the clustering analysis described earlier. Other Bacteria include Chlamydiae,
Chloroflexi, Bacteroidetes, Gemmatimonadetes, Planctomycetes, Tenericutes, Nitrospira,
F usobacteria, Deinococcus-Thermus, Deferribacteres, TM 7, Spirochaetes, WS3, 0D],
0P10, F ibrobacteres, BRCI, Cyanobacteria, Lentisphaerae, and unclassified Bacteria.
Not all the mentioned phyla were not present in all sites.

91

 

 

1 w '70
A 90%
80%

70%

40%

l

N u
2 o.
.\ 0\

Relative contribution to

 

10%
0%
5»
D U. Proteobacteria I Epsilon
I Delta 13 Alpha
E] Gamma a Beta

Figure 3.8. Class composition of the Proteobacteria from FRC Area 3 samples. U.
Proteobacteria accounts for sequences classified as Proteobacteria but that could not be
assigned with a 80% confidence to any of the classes of the Proteobacteria using RDP’s
Classifier. Samples are ordered into four clusters (A, B, C and D) according to the
clustering analysis described earlier.

Taxonomic classes of the Proteobacteria phyla varied among the sites (Figure
3.8). When comparing sample clusters A and B fiom the clustering analysis (inactive vs
active), cluster B was enriched in Deltaproteobacteria (p<0.01), while cluster A had a
significantly higher proportion of Gammaproteobacterz'a and Alphaproteobacteria.

Cluster D was similar to cluster A in terms of abundance of Gammaproteobacteria.

92

When comparing the inner and outer loops, we found Deltaproteobacteria in higher
proportions (p<0.05) at the inner loop, while Epsilonproteobacteria were significantly
lower (p<0.05) in this zone. The great variability observed in Proteobacteria is probably
due to the range of conditions surveyed in terms of pH and carbon availability (measured
as chemical oxygen demand due to connectivity differences. Also the proteobacteri are

the most metabolically and ecologically diverse group of bacteria (19).

Given that the Deltaproteobacteria class has many of the known uranium
reducers such as Geobacter (Desulfomoronales), Desulfovibrio (Desulfovibrionales), and
Anaeromyxobacter (Myxococcales), we evaluated if orders of this class were present in
the samples and if they were distributed in the same abundance in samples from inactive
wells versus samples from active wells. The results showed that the cluster B’s
Deltaproteobacteria were enriched in members of the orders Myxococcales (p=0.01),
Desulfobacterales (p<0.05) and Desulfovibrionales (p<0.01) in contrast to low activity

wells.

93

A B C D

 

 

 

 

 

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(0.88%) (8.46%) (0.80%)(0.54%)
A100% ”1!!! , I
\- _ '5? 35:-2:} ., -Z:--I
2% 90% 23:}: “5:31;” ‘ ‘25:: :Ij-I:j::'i—:j:§:-I:‘ —:::II:'
3.... :2: 1.::: «2:2: 2:: :.:r::- :. 1: f
§ 70% — ~ — ZfiEj- - ~ —+ '7'-
s 60% :1: . .::2: 333:: -—— -
:3 500/ . _2-31: ~_: _.¥::-_ . _:-;:3::-_ __ _ __ _
3 ° -::-- __ - -::--: -:=:--:
= ;_-_', .555; 1. .1} {3. :2”: 2:3} f}.
.3 40% ‘ - — .-_. -
5 '. ...-._ .5}:-
‘E 30% - - _' —: u - 3: ~ I;:-;:-.; —— ~ —- -
=1 -.'-'. 4:? -I 3'3
5 *5 i-‘-' t: . _ ‘5
°‘ 0% 2E- .. at b -I'- at t» a 2% it" at ::e .22 e?

I Unc. Deltaproteobacteria El Syntrophobacterales

’00?

El Desulfovibrionales Cl Desulfomoronales
El Desulfobacterales I Bdellovibrionales
Myxococcales

Figure 3.9. Relative abundance of Deltaproteobacteria orders in FRC Area 3
samples. Samples are ordered in the four clusters (A to D) that were defined previously
in clustering analysis. The average contribution of the Deltaproteobacteria to the total
community in that specific cluster of samples is given in parenthesis. U.
Deltaproteobacteria are sequences that cannot be assigned with >80% confidence to one
of the orders of the Deltaproteobacteria class by RDP classifier.

Presence of genera with members known to reduce uranium

94

Genera known to contain U(VI) reducers were detected in the samples.
Desulfovibrio was on average the most abundant one followed by Desulfosporosinus,
Geobacter, Anaeromyxobacter and CIostridium. The first four genera were present in all
wells with high bioactivity (Figure 3.10), Clostridium was present in these wells but its

abundance was higher at a low activity well from cluster A.

 

12

 

I Geobacter
Desulfovibrio
B A naeromyxobacter

Desulfosporosinus
I Clostridium

 

g—o
o
a
I

 

 

 

 

 

co

\\\\\\\\\\\\

 

 

 

\\\\\\\\\\\\\\\\\\\
I

 

 

Relative contribution to to library (%)
h 0‘

 

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

IIIII——_—
IIIIIIIIIIII'IIIIIII’IIIIIII.

\\\\\\\\\\\\\\\

   

, : i E:
’ l '5 . ‘S:: l
O 1 W :9 1.: ’5.‘ . ‘
9" @‘M'v x," o) my @‘W xwwfi'v Q6”
\Q \Q \Q \Q \Q .\Q \Q \Q
N‘éeas si‘ksaa $8 at“
C D

Figure 3.10. Relative contribution of most common uranium reducing genera found
in the samples from DOE FRC Area 3. Samples are ordered in the four clusters (A to
D) defined previously in clustering analysis. Cluster B contains samples with highest
bioactivity.

95

Abundance of Geobacter, Anaeromyxobacter ,Desulfosporosinus, and
Anaeromyxobacter sequences were higher at wells with higher connectivity, as measured
by tracer recovery (Figure 3.11). The location where the population peaked in abundance
differed for each group (order: Desulfovibri0> Geobacter > Anaeromyxobacter >
Desulfosporosinus), suggesting that some niche specialization may be have occurred
along the electron donor gradient. The location of the peak for Geobacter is consistent
with a study that found elevated Geobacter abundance in clone libraries when organic

carbon (measured as COD) was elevated (27).

Four other genera with known uranium reducers were also detected:
Desulfitobacterium, Shewanella, Acidovorax, and Deinococcus. They, however, were not
present in all the wells where uranium reduction was established and their relative
abundance was generally low (0.1 to 1%). Hence, their role in uranium reduction is

difficult to establish and for that they were not considered in the analysis.

96

 

     
 

 

 

  

 

 

 

 

 

   

 

 

 

 

 

 

 

 

-O- Geobacter
100 d ‘o. 4 -D-Desu(fi)vibrio
:3 ' . -&' Anaemmyxobacter
E 80 t .0-Desulfospomsinus ‘
g S V ' -I- C lostndzum
,1 g “m- Tracer recovery
as t
s2 w -
ii
5% 4° ‘ ‘
é‘
a. 20
o .

 

$5®W3Wb® 153' $5.1?
Wg‘“ Q9§4§§y$4®

We“
$3®§®<¢Q<€M M3 Miss

Figure 3.11. Location of maximum abundance for uranium reducing genera along
the connectivity gradient at FRC Area 3. Plotted are the ratio of abundance to

maximum abundance for selected groups and the proportion of tracer recovery, all in a 0
to 100% scale.

97

Indicator species

Indicator species analysis is an approach that can greatly reduce the complexity of
the dataset by finding sequences that represent samples, or cluster of samples.
Additionally, the indicator value is provided with an indication of statistical confidence.
We used abundance information for 4719 OTUs over 15 samples to obtain indicator

species.

The first level of analysis focused on the four sample clusters found by clustering
analysis (Clusters A, B, C, and D). Cluster A (mostly inactive wells) was represented
mostly by Betaproteobacteria and Gammaproteobacteria OTUs (Table 3.7), two classes
known to have nitrate reducing organisms. The top 15 representatives (ranked according
to descending indicator value) contained three Rhodanobacter and two Castellinella
OTUs. Rhodanobacter, a Gammaproteobacteria, was previously detected in high
abundance in F W106, a location in Area 3 meters away from the treatment zone (10).
FW106 has not received any stimulatory electron donor and it is still acidic @H 3.6) and
highly contaminated. A metagenome analysis of a FW106 groundwater sample
additionally revealed a Rhodanobacter sp. with a wide variety of metal resistance genes
and indication that these genes were horizontally transmitted (24). Castellinella is a
denitrifying Betaproteobacteria that has been found in biostimulated wells from the FRC
Area 1 where it was the dominant nirK-containing denitrifier (49). Area 1 is also close to
the main contamination source at the PRC, and has acidic conditions as well as high
nitrate levels. The main differences between these areas 1 and 3 reside in the U levels,

with Area 3 being higher, and the treatment system used. Area I experienced a pull-push

98

approach with ethanol while Area 3 used the hydraulic control system previously
described. Castellinella isolates from FRC Area 1 were recovered under acidic (pH 4.5)
and neutral (pH 7.5) conditions. It was suggested that this group may outcompete other
denitrifiers under acidic conditions characteristic of the FRC contaminated areas (49).
Castellinella has also been detected in microcosm enrichment from the FRC Area 2 (1).
The rest of the top 15 representatives of cluster A did not include genera known to be
involved in bioremediation of U or nitrate. A Comamonas sp. was also found in top 15
ranking at #2, this genus is known for their ability to grow on organic compounds such as
phenols. These compounds were present in the contamination source (58) and may be
still present in locations poorly connected to the injection well. The main physiological
theme in cluster A representatives was their resistance to harsh conditions like those from

the FRC.

The top 15 representative OTUs of cluster B contained seven OTUs related to
U(VI)-reducing genera (Table 3.8). The wells in this cluster are the most active in
microbial activity, and are also the only locations where U(VI) reduction has been
detected. Out of the seven OTUs, four belonged to Desulfosporosinus (SRB F irmicutes),
two to Desulfovibrio (SRB Deltaproteobacteria) and one to Anaeromyxobacter (FRB

Deltaproteobacteria).

Two denitrifiers, Vogesella (#3) and T hauera (#12) were also present as top
indicator species. Thauera spp. have been previously found in the denitrifying bioreactor
used ex situ to precondition the site by removing nitrate (28). Since aboveground
treatment system was connected to the subsurface, it is likely that T hauera spp, were

injected with the treated water.

99

Bacteria related to iron reduction/oxidation had a high indicator value for this
cluster of sequences: Gallionella (#15, Ind.value 0.93, p=0.001), Thiobacillus (#16,
Ind.value=0.928, p=0.001), Geothrix (#22, Ind.value=0.883, p=0.001), and
F erribacterium (#28, Ind.value=0.883 p=0.115). Geobacter, a F RB also known to reduce
U(VI), ranked slightly lower at position 53 (indicator value 0.73, p=0.065). A special
case is that of T hiobacillus, a bacteria capable of reducing iron and nitrate but also able to
oxidize iron and uranium (the later while reducing nitrate) (4). Its presence in the top of
the indicator list suggest that it plays a role in the most active uranium reducing

communities and may influence the mobility of uranium in situ.

Two amoeba endosymbiont relatives were found in high in the indicator species
ranking. The putative endosymbiont come from two different phyla ,Chlamydiae and
Proteobacteria, and share the same likely host, Acanthamoeba, a common free-living
protozoa that feeds on bacteria, algae and yeast in the environment (3 6). Considering that
the sediments with high activity in occasions clog the wells with biomass (10), it is quite
possible that predatory populations of amoeba exist at these locations, though this

remains to be established directly by microscopy, cultivation, or PCR based methods.

Clusters C and D did not contain known U(VI) reducers in their top 15 ranking
taxa (Table 3.9 and Table 3.10, respectively). In cluster D, Geobacter was positioned #21
(p=0.065) and Desulfovibrio #99 (p=0.109). These results suggest that even when
populations capable of reducing U(VI) exist in these locations, they are not good
representatives of the main selection trends in these sites. In both C and D, the most

frequent groups in the top of the list were related to Alpha- and Gammaproteobacteria.

100

Indicator species analysis was also performed on the cluster B. Subgroup B2 was
formed by three samples, FW101-2, FW101-2 and FW102-3, and was consistently high
in uranium reduction. In these three samples Desulfovibrio was a top indicator species
where it ranked #1 (p=0.05 5) and #12 Q3=O.O3), respectively. Desulfosporosinus ranked
#8 at B2 but the statistical confidence was low (p=0.317). In these samples Desulfovibrio
spp. were always in the 10 most abundant species. Interestingly, Geothrix, an iron
reducer Acidobacteria was also highly present in these samples, and always more

abundant than Desulfovibrio.

It is also likely that there are taxa not previously described as U(VI) reducers
present and actively reducing uranium at this site. Candidates for these roles are iron
reducers such as Geothrix (Acidobacteria) and F erribacterium (Betaproteobacteria). The
indicator list for this group of samples could potentially be used to target specific groups
in isolation efforts since the uranium reduction ability seems to be more widespread than

originally described.

101

Table 3.7. Top 15 indicator species for sample cluster A, their classification and
closest isolated relative.

 

 

3
Ind. - 1 2 .
Rank Value vailue Classification Closest isolate in RDP ”:22?
Treponema Parvibaculum sp. psclO;
0‘ ”’00 0-002 (spirochaetes) EU930870 90.7
C omamonas Bacterium r19; ABOZl327
02 0'997 0'002 (Betaproteobacteria) 99_ 5
U. Sphingobacteriales Niastella sp. RHYL-67;
03 09” 0'008 (Bacteroidetes) EU917053 94.]
C astellaniella C astellam'ella defragrans;
04 0'985 0‘00] (Betaproteobacteria) 62Car; AJ005449 99. 5
Rhodanobacter Cimanggu media isolate 88;
05 0.973 0.005 (Gammaproteobacteria) Cimanggu media 88; B;
AF229452 99.0
Rhodanobacter Rhodanobacter ginsengisoli;
0‘ “'9‘“ “'0‘” (Gammaproteobacteria) GRl7-7; EF166075
99.0
U. Rhizobiaceae Ensifer xinjiangensis; Rx22;
07 0'938 0'004 (Alphaproteobacteria) AF250353
99.0
Opitutus Opitutus sp. VeSml3; X99392
08 0'904 0'064 (Verrucomicrobia) 97.1
Rhodanobacter Cimanggu media isolate 88;
09 0.886 0.002 (Gammaproteobacteria) Cimanggu media 88; B;
AF229452 99.0
Gp6 (Acidobacteria) Unidentified eubacterium clone
‘0 0.872 “-004 sto7; A1229181 339
C astellaniella C astellam'ella sp. 7.5A2;
1 1 0'866 0'002 (Betaproteobacteria) EF175378 99. 5
Gp6 (Acidobacteria) Marine gamma proteobacterium
‘2 0-847 0-006 HTCC2246; AY386337 g5_3
13 0.843 0.065 U. Bacteria Sigifigsogrggnas spJQl-3; 32 3
U. Verrucomicrobia/es Bacterium Ellin516; AY960779
l4 0'833 0'07] (Verrucomicrobia) 33_7
U. Hypomicrobiaceae Devosia ginsengisoli; Gsoil 326;
15 0‘833 0'00? (Alphaproleobacteria) ABZ71045 9&0

 

Ind.value = Indicator value, p-value = probability of obtaining as high an indicator values
as observed over 1000 iterations. 1 Classification based on RDP Classifier with a 80%
bootstrap confidence. 2 Closest relative was found with RDP SeqMatch function using an

isolates-only dataset. 3 Sequence identity with the closest isolate from the RDP.
Unclassified sequences (U) were not able to be assigned to a lower taxon with the 80%
confidence. Wells in this cluster: F W100-4, F W101-4, FW102-1, FW102-4, FW103, and
F W1 05. Groups in bold were significant (q-value <0.05) under the false discovery rate
analysis.

102

Table 3.8. Top 15 indicator species for sample cluster B, their classification and
closest isolated relative.

 

 

3
Ind. - l 2 '
Rank Value vzilue Classification Closest isolate in RDP Ides/Sty
0
Desulfosparosinus Desulfosporosinus sp. Sapy;
0‘ "00° 0‘00] (Firmicutes) AF159120 99.0
Desulfosporosinus Desulfosporosinus sp. Sapy;
02 M97 “'0‘” (Firmicutes) AF159120 99,0
03 0 979 0 001 U. Burkholderialas Vogesella indigofera; ATCC
' ' (Betaproteobacteria) 19706T; A8021385 95,6
U. Peptococcaceae Anaerobic bacterium Prop2;
0“ 0377 0'00] (Firmicutes) AY756143 92 7
U. Chlamydia/es Endosymbiont of Acanthamoeba sp.
05 0974 “002 (Chlamydiae) uwczz; AF083616 91,7
Desulfosporosinus Desulfosporosinus sp. Sapy;
06 0'97] 0'00] (Firmicutes) AF159120 98.0
Desulfovibrio Desulfovibrio putealis (T); DSM
07 0'97" 0'00] (Deltaproteobacteria) 16056; AY574979 99,5
08 0.968 0.00] U. Firmicutes Clostridium sp. 9B4; AY554416 92.7
Desulfovibrio Sulfate-reducing bacterium F 1-
09 0367 0'00] (Deltaproteobacteria) 7b; AJ012594 99,5
. Desulfomonile limimaris (T);
10 0.962 0.002 U. Bacteria AF230531 86.3
11 0.959 0.001 U. BRCI Bacterium Ellin37l; AF498753 82.8
12 0.957 0.006 U. Betaproteobacteria Thauera sp. R-28312;AM084110 99.0
Anaerom xobacter Anaeromyxobacter dehalogenans
13 0.948 0.001 ( Del 1‘ roJI’eobacteria) (1‘); 2CP-l; ATCC BAA-258;
“p AF382396 96.1
. Endosymbiont of Acamhamoeba
14 0.943 0.001 U. Alphaproteobactena polyp haga; AF132138 90.2
. Gallionellaferruginea subsp.
15 0.929 0.001 U. Betaproteobactena caps iferr iformans; DQ386262 96.6

 

Ind.value = Indicator value, p-value = probability of obtaining as high an indicator values
as observed over 1000 iterations. 1 Classification based on RDP Classifier with a 80%
bootstrap confidence. 2 Closest relative was found with RDP SeqMatch function using an

isolates-only dataset. 3 Sequence identity with the closest isolate from the RDP.
Unclassified sequences (U) were not able to be assigned to a lower taxon with the 80%
confidence. Wells in this cluster: FW026, FW101-2, FW101-3, FW102-2, FW102-3, and
FW104. Groups in bold were significant (q-value <0.05) under the false discovery rate
analysis.

103

Table 3.9. Top 15 indicator species for sample cluster C, their classification and
closest isolated relative.

 

 

3
1nd. - 1 2 .
Rank Value vzilue Classification Closest isolate in RDP [dig/n?
0
U. Chlamydiales Neochlamydia hartmannellae (T);
01 1000 0009 (Chlamydia) All-15p; AF 1 77275 84'9
02 1.000 0.013 g . Legionella sp. 1129; ABOS8910 90.7
ammaproteobacterta
. _ Tuber borchii symbiont b-Z43; b-
03 1.000 0.010 U. Bacterozdetcs Z431; AF233292 89.3
Rhodocista zospirillum sp. DA10-2;
04 1 'OOO 0'0 1 3 (A Iphaproteobacleria) AY 1 l 8225 93 '2
. Pseudophormidium sp.
05 1.000 0.008 GpV (C yanobacterza) ANT.PROGRESS2.2; AY 493 583 95.6
. C handromyces [anuginosus (T)'
7 9
06 1.000 0.01.. U. Bacteria Sy t2; A1233939 82.9
. C lostridium paradoxum; DSM
07 0.981 0.043 U. Bacteria 7308T; para2;Z69941 79.0
08 0.968 0.001 S“bd'v's'°'i' 3 . Bacterium Ellin514; AY960777 94.6
( Verrucomicrobia)
. Neochlamydia hartmannellae (T);
09 0.964 0.003 U. Bacteria Alep; AF177275 85.2
Geobacter sulfurreducens PC A;
10 0.957 0.00] 0D] AE017180 75.6
Rhabdochlamydia Rhabdochlamydia crassificans;
' 1 ”“947 0'008 (Chlamydiae) CRIBOI; AY928092 93'2
. Sulfate-reducing bacterium R-
12 0.946 0.010 U. Bacteria PropAl;A.101259l 82.0
13 0.943 0.010 Gp3 (Acidobacteria) Bacterium Ellin37l; AF498753 89.3
. Desulfohalobium utahense (T);
14 0.931 0.086 U. Proteobacteria EtOH3; DQ067421 82.9
15 0.931 0.006 U. Alphaproteobacteria 305p "”1”" amazoneme; 21R; 89.3

AY741 146
Ind.value = Indicator value, p-value = probability of obtaining as high an indicator values

as observed over 1000 iterations. 1 Classification based on RDP Classifier with a 80%
bootstrap confidence. 2 Closest relative was found with RDP SeqMatch function using an

isolates-only dataset. 3 Sequence identity with the closest isolate from the RDP.
Unclassified sequences (U) were not able to be assigned to a lower taxon with the 80%
confidence. Wells in this cluster: F W024 and FW100-3. Groups in bold were significant
(q-value <0.05) under the false discovery rate analysis.

104

Table 3.10. Top 15 indicator species for sample cluster D, their classification and

closest isolated relative.

 

 

3
1 d. - 2 ‘
Rank Vglue vzilue Classification Closest isolate in RDP Ide(r01/t01)ty
. Candidatus Brocadia sp.
1 1.000 0.070 U. Bacteria 40; AM285341 81.5
U. Lachnospiraceae C lostridium herbivorans
2 1'000 00% (Firmicutes) (T); 54408; L34418 95'6
Verrucomicrobia
3 1.000 0.064 U. Bacteria bacterium YM31-066; 87.8
AB372850
. Unidentified eubacterium
4 1.000 0.069 U. Bacteria clone BSV28; A1229190 82.0
Flavimonas Pseudomonas sp. SV4;
5 Low 0072 (Gammaproteobacteria) AF251334 92'2
. Candidatus Brocadia sp.
6 1.000 0.070 U. Bacteria 40; AM285341 80.5
. Bacillus sp. BR (T);
7 1.000 0.053 U. Bacteria AM050346 82.0
. Butyrate-producing
8 1.000 0.060 fggffutr’fnicmbia) bacterium A2-194; 95.6
0 AJ270473
. Treponema porcinum (T);
9 1.000 0.064 U. Bacteria 14V28; AY518274 88.3
U. Actinomycetales Actinobaculum sp.
10 1'0“) 0073 (Actinomycetes) P2P_19 P1; AY207066 94'6
lsosphaera Isosphaera -like str. C2-
1 1 1000 0062 (Planctomycetes) 3; AF239692 99.5
Desulforomonas Delta proteobacterium
12 1'000 0'066 (Deltaproteobacteria) 112 (T); AY835388 95'6
Sphaerobacter
13 1.000 0.079 U. Bacteria thermophilus; DSM 78.4
20745T; AJ420142
U. C lostridiales C lostridium termitidis;
‘4 1'0“) mm (Firmicutes) DSM 5396; X71854 9L7
Treponema T reponema bryantii (T);
15 ”mo 0'064 (Spirochaetes) RUS-l; M57737 88'7

 

Ind.value = Indicator value, p-value = probability of obtaining as high an indicator values

as observed over 1000 iterations. ] Classification based on RDP Classifier with a 80%
bootstrap confidence. 2 Closest relative was found with RDP SeqMatch function using an

isolates-only dataset. 3 Sequence identity with the closest isolate from the RDP.

Unclassified sequences (U) were not able to be assigned to a lower taxon with the 80%
confidence. Cluster formed by well FW100-2. Groups in bold were significant (q-value
<0.05) under the false discovery rate analysis.

105

Ordination of samples by principal coordinates analysis

Principal coordinates analysis showed an ordination that was consistent with
previous clustering analysis as shown with the minimum spanning tree overlaying the
ordination (Figure 3.12 left panel). When testing the ordination against gradients of
environmental factors, chemical oxygen demand (COD) in sediments, sulfide in
sediments, total iron in sediments, and uranium levels in sediments were significant in
explaining the ordination (p<0.05). However, we previously established that colinearity
was present in these variables. We hypothesize that COD is the independent variable of
this group. COD represents the organic carbon content in sediments, and also probably an
indication of electron donor availability in this electron-donor—limited environment. COD
likely influenced sulfide, iron, and uranium levels through the activity of sulfate-, iron-
and uranium-reducing microorganism that used the organic compounds. This results
agrees with our hypothesis that the injection of the stimulatory carbon, ethanol, helped

shape the community structure.

106

 

 

A FW024 ‘

FW1 00-3

  
  

 

 

 

(%L0 91) zvooa
to,1'0.0'011'0.Z'0.£'0‘v'019'0 1

 

 

 

 

 

 

FW026
FW102-4
FW100-4
FW101-2
FW101 FW101-3
0.4 0.2 010 0'2 0'.4
PCOAl (22.07%)
B Fw’Tz4 ‘ COD
3 Sulfide (Sed)
1 FW100-3 U (Sed)
i e(Sed)
w ‘1
8 8
B « mime-2
f; 5 ‘ii’ . FW103
S o“ FW026
3‘: z...
‘ FW102-4
.0 FW104
°_ FW100-4 ' I
p o 0 FW105 FW102-2 FW101-2
“ o w102-1
- FW101-4.1; FW102-3. FW101-3
9 l 1 1
N 0.4 0'2 0.0 0.2 0.4
PCOAl (22.07%)

 

 

 

Figure 3.12. Principal coordinates ordination on abundance-based Sorensen
distances for samples from the FRC Area 3. First two dimensions together explain
38% of the original variability. The grouping of samples is consistent with previous
clustering results as shown with a minimum spanning tree on top of the ordination (panel
A). Vectors that explained significantly the ordination (p<0.05) are shown (panel B),
though vectors were collinear (correlated). Cluster A (0), cluster B (I), cluster C (A),
and D (*).

107

We also used non metric dimensional scaling (NMDS) to study the distribution of
specific taxonomic groups likely related to the U(VI) reduction process. This model was
preferred because it does not assume a specific distribution of species along a gradient. It
is not clear if the measured environmental gradients were pronounced enough for our
groups of interest to react to; thus, we cannot assume that our populations have a
unimodal distribution (single peak along a gradient). In this case we tested only five main
groups of variables, they were called pH, nitrate, chemical oxygen demand (COD), mean
travel time and U in sediments. Mean travel time is a hydrology terms that refers to the
time water from the injection wells takes to travel to a given location, the more connected
the location the shorter the travel time. Because the variables were correlated, these
names actually represent several collinear variables. The names used are to our best
estimation of what is the independent variable for the group, though it is possible that a
collinear variable is responsible for the response. Four of the five groups of variables

were significant for at least one group of organisms (Table 3.11).

COD and Uranium levels in sediment were the most common factor to be found
significant in explaining the variability of the groups of interest. COD explained
significantly (p<0.05) the distribution of Bacteria, Anaeromyxobacter, and Desulfovibrio
along the samples. Uranium levels in sediment were highly significant in explaining the
distribution of Deltaproteobacteria, Desulfovibrio and Desulfosporosinus (Table 3.11).
However, this variable is correlated with sulfide and iron levels in sediments, so part of
the response could be due to changing levels of sulfate and/or iron. Nitrate was weakly

significant (p<0.1) in explaining the distribution of Geobacter OTUs. Nitrate was highly

108

correlated with sulfate, na, mg, Mn, chloride and Ca. The results suggest that SRB
Desulfovibrio and Desulfosporosinus, as well as FRB Anaeromyxobacter are responsible
for the uranium levels, and that they responded to the injection of the stimulatory electron

donon

Table 3.11 Environmental factor that explain distribution of groups along the
samples.

 

COD Travel

 

Group Nitrate U(sed) Jsed) time
Acidobacteria - . _
Gp8 - * . -
Betaproteobacteria - . . -
Deltaproteobacteria - - - .
Anaeromyxobacter - ** * *
Desulfovibrio - ** * *
Geobacter . - - -
F irmicutes - * * .
Desulfosporasinus - ** . -
Verrucomicrobia - — - -
Bacteria - . * -

 

Environmental factor that significantly explain the distribution of clusters in NMDS
ordination space. °p<0.1,*P<0.05, **p<0.01. Similarity threshold: Genus level analysis=99%,
Phyla and class level = 95%. pH was not significant in explaining the distribution of any of the
groups studied.

DISCUSSION

Effect of hydraulic control on microbial communities of the treatment area

The establishment of a controlled hydraulic system created an opportunity to study

the bioremediation of a contaminated aquifer that went through different stages from

109

denitn'fication to iron reduction to sulfate reduction (27, 58, 59). We observed differences
in the spatial scale that are a consequence of the hydraulic flow regime and the injection
of ethanol as the stimulatory electron donor. The geochemical gradients shaped the
microbial communities to resist toxicity, denitrify, reduce iron and reduce uranium in
different zones of the study area, creating differences in the spatial patterns. Other work
that tracked the succession of communities in FRC Area 3 since the stimulation started
determined the establishment of these different types of communities (toxicity
resistant/denitrifiers/metal reducers) but in a temporal scale (27). In biological terms, the
establishment of spatial gradients increased the diversity of metabolism for the site
creating a biological transition zone. This zone potentially works as buffer zone,
protecting the inner loop from intrusion of contaminated water if the hydraulic system is

perturbed and could be important in the long term bioremediation of the site.

Besides the hydraulic control and ethanol injection, an additional factor in shaping the
microbial communities of the site is the conditioning phase that used a fluidized bed
bioreactor to remove nitrate (competing electron acceptor). This bioreactor worked for
over 4 months treating water ex situ and reinjected treated water as well as microbial cells
into the study area (28). This bioreactor was initially inoculated with biomass from the Y-
12 wastewater treatment plant (300 m away from Area 3, also at the FRC) (28), and
biomass from a pilot study FBR that showed denitrification as well as U(VI) reduction
(61) . The pilot scale reactor was also inoculated with site water (TBP-16 well the FRC).
In both reactors, pilot and full scale, ethanol was used as electron donor and populations
of SRB such as Desulfovibrio spp. were detected (28, 61) . In the pilot scale bioreactor,

uranium reduction was also present and additions of molybdate (a SRB inhibitor)

110

decreased sulfate reduction as well as U(VI) reduction, suggesting that SRB were partly
responsible for the uranium reduction (61). Other groups detected in the bioreactors that
were also detected in the sediment samples include Acidovorax (denitrifier, also U(VI)
reducer), Thauera (denitrifier Betaproteobacteria), Sporomusa (acetogen F irmicutes),
Stenotrophomonas, Pseudomonas, Decholormonas (denitrifier Betaproteobacteria),
Hydrogenophaga (hydrogen oxidizing Betaproteobacteria), and Clostridium among
others (28, 61). No FRB known to reduce uranium where detected in either bioreactor.
However, Geobacter and Anaeromyxobacter, both U(VI)-reducing F RB, were detected in
enrichments of sediments of Area 1 of the FRC (42). The bioreactor is likely a source
inoculum in the subsurface ecosystem, helping U(VI)-reducing SRB to establish in the
aquifer and compete with local FRB populations that are more likely in lower numbers.
The F BR inoculums is probably a reason why SRB are predominant in this location, and
U(VI) reduction works for this site. Since not all SRB reduce U(VI) and many SRB can
utilize ethanol as electron donor, the success formula of this site may not occur in other
sites. In a different site also contaminated with uranium, injection of acetate stimulated an
early response of F RB (Geobacter spp.) that reduced uranium but later succession with

SRB Desulfobacterales decreased uranium reduction (25).

Fate of electron donor injected and populations that responded to the treatment

Several studies have shown that ethanol is the best electron donor for uranium reduction .
at the FRC (1, 59). Ethanol, however, is rapidly converted into acetate at the injection

point, acetate accumulates and it is later consumed (1 , 10, 39). The oxidation of ethanol

111

to acetate and its release to the environment is consistent with the activity of SRB such as
Desulfovibrio, and Desulfosporosinus (31, 47). Additionally when ethanol has been used
as electron donor in studies at the F RC, sulfate and uranium reduction occur
simultaneously (1, 38, 59). This suggests that at the FRC, SRB are more important in
uranium reduction than FRB. In our analysis, SRB Desulfovibrio and Desulfosporosinus
correlated with uranium reduction in ordination analysis and are characteristic of sites

with high U(VI) reducing abilities in the indicator species.

This however does not discount that FRB partially contributed to the reduction
even though acetate provides less energy than ethanol. Akob and colleagues found that
F RB Geobacter spp. reacted to ethanol amendments in microcosm enrichment from Area
2 that showed U(VI) reduction (59). In our case F RB Geothrix, F erribacterium, and
Geobacter were also abundant in sites with high UW I) reducing activity. The three of
them are known to utilize acetate to reduce iron but U(VI) reduction has been
demonstrated for Geobacter only (34). In all the active wells, the F RB Geothrix was the
more abundant than any SRB. At the injection well, F W1 04, where ethanol is the main
electron donor, Geothrix and F erribacterium where the most dominant FRB, being more
abundant than Geobacter. In this well, reduced iron was at its maximum, so competition
for oxidized iron, or unknown abilities of the specific F erribacterium and Geothrix spp.

could have excluded Geobacter.

In later sample wells when acetate is the main electron donor, Geobacter peaked
in abundance. After Geobacter, Anaeromyxobacter (also FRB) and Desulfovibrio (SRB)
peaked in abundance. The fact that the SRB were less abundant in these wells could be

due to sulfate reduction being less favorable than iron reduction when acetate is the

112

electron donor. Another possible explanation is that since acetate oxidation is not a
frequent characteristic of Desulfovibrio, this SRB could be growing as the hydrogen-
consuming partner in a syntrophic association with an acetate-consuming partner. This
type of association has been demonstrated for pure cultures of Geobacter and

Desulfovibrio (15).

In general, it seems that SRB dominate these samples because of their ability to
oxidize ethanol, the time of sampling in the succession event, and the initial advantage
given by the F BR. FRB likely use mainly acetate to reduce iron and benefit from the
presence of iron oxidizers (aerobic and anaerobic) at the site. This dynamic seems to be
present in our site since many iron oxidizing genera such as Gallionella and T hiobacillus
were detected in the most active wells. This dynamic could also be applied to sulfur since
Sulfuricurvum, an anaerobic sulfide oxidizer was present in high abundance but only in

one site (FW024).

Uranium reduction and its correlation with microbial groups

Comparison of the levels of uranium reduction in different wells with the
locations where the putative iron reducers peaked, suggests that multiple populations are
responsible for uranium reduction. The highest uranium reduction (61%) was at the
injection site FW104, where ethanol was rapidly converted to acetate (likely by SRB) and
FRB Geothrix and F erribacterium dominated. In slightly less connected wells, FW102-
2, and FW102-3, Geobacter peaked in abundance. In these wells the average uranium

reduction was 24%. Finally at wells FW101-2 and FW101-3, uranium-reducing

113

 

candidates with sulfate-reducing activities peaked in abundance. Those wells had 54%
and 51% of reduced uranium, respectively. We hypothesize that F RB are responsible in
the more connected wells for uranium reduction, and SRB are responsible for uranium
reduction at the injection site and less connected points. This agrees with previous
findings of Geobacter and Geothrix being associated with high levels of COD and no
dissolved oxygen control (27). In our case more connected wells are the equivalent of
that because they have more carbon available and a higher chance of having dissolved
oxygen (due to injection process). Desulfovibrio spp. were associated with dissolved
oxygen control in previous studies (27), this agrees with our findings of Desulfovibrio
peaking in less connected wells where oxygen is less likely to be present due to its

consumption in earlier wells.

It remains unclear if the FRB F erribacterium and Geothrix are involved directly
in uranium reduction or if they are only benefitting from the added electron donor while
reducing iron or humic acids. Since reduced iron compounds (from FRB )(44) as well as
reduced humic acids (from a humic reducer such as Geothrix) (12, 41) can potentially
reduce uranium abiotically, F erribacterium and Geothrix activities could be linked to
uranium reduction. Geothrix has been associated with uranium reduction in other studies
of the FRC, though no direct evidence of uranium reduction exists to date (7, 10, 27). The

role of humic acids as electron shuttles for uranium reduction also remains unknown (39).

Clustering, direct gradient ordination, and indicator species analysis suggests that in
more active communities, Deltaproteobacteria, and more specifically its SRB members
seemed to be playing a bigger role in the immobilization of uranium at the time of the

sampling. In less active communities, the lack of electron donors has selected for species

114

 

capable of life under acidic and toxic conditions. Our indicator species analysis suggests
that the groups more likely to reflect the uranium-reducers are related to Desulfovibrio,
Desulfosporosinus, Anaeromyxobacter, and Geobacter. Differences in uranium reducing
populations at the active locations can be attributed mainly to differences in connectivity

to the injection well that likely influenced that carbon available for the microbes.

There are populations that were not related to uranium reduction but were a
consequence of the establishment of high number of bacteria in the active zone, e. g.
populations likely involved in recycling dead cell material such as Chloroflexi were

significantly higher in active sediments.

Patterns of diversity

The communities coming from samples with reduced uranium were not the richest
communities. This is likely due to a combination of incomplete sampling and differences
in evenness. Uneven communities reflect on one extreme a thriving community with
highly enriched groups (e. g. SRB in FW104) and on the other extreme a community
struggling to survive toxic conditions (e.g. F W105). Clustering by diversity patterns was
more informative, and showed that communities responsible for uranium reduction were
similar in composition suggesting that the same main groups in these communities were
reacting to the treatment. The treatment changed the subsurface by creating better
conditions for microbes in terms of pH, heavy metals, and electron donor availability.
Electron donor availability was probably the most important change since energy source

is commonly the limiting factor in uranium reduction in groundwater (59). The impact of

115

the treatment can be tracked to groups that increased in abundance such as C hloroflexi,
Acidobacteria, Deltaproteobacteria, Desulfovibronales, Myxococcales and
Desulfobacterales. In general the treatment shaped the community to provide specific

functional groups for nitrate, iron, sulfate, and uranium reduction.

When those databases are compared to our previous clone libraries results, some
differences in abundance are noted. Since we used two different systems in terms of
primers, product size, and amplification conditions; some differences in abundance can
be due to primer coverage (primer bias), sampling, and regions flanking the template
region (21). Additionally, the lack of a cloning step could potentially avoid cloning bias.
Cloning bias has been reported for F irmicutes rRNA genes cloned into a
Gammaproteobacteria host (48). In our case, Desulfosporosinus contribution comprised
1% of clone libraries and 8% of the pyrosequencing produced, using the same DNA

extract. This difference is more likely attributable to difference in bias and sampling.

Beside avoiding cloning bias, our methodology also allowed us to explore more
deeply the microbial diversity of the site, and to track specific populations that occur at
low frequency (e. g. 0.5% for Anaeromyxobacter) and whose trends were previously

difficult to follow.

CONCLUSIONS

The results from a much larger number of 16S rRNA gene sequences allowed us
to more comprehensively study the microbial communities in response to treatment than

was previously feasible, and to compare their patterns of diversity with a statistical

116

confidence.

Clustering analysis showed differences in community structures that can be
related to connectivity and pH among wells as well as organic carbon. In contrast with
other sites where Geobacter populations are more abundant, our results suggest that SRB
such as Desulfovibrio were more involved in uranium reduction at this particular location
at the time of sampling, this is likely due to the choice of electron donor, its metabolism

by SRB and F RB, and the use of a bioreactor that likely injected SRB into the aquifer.

ACKNOWLEDGMENTS

The microbial study was funded by the US. DOE under grant DE-FG02-97ER62469,
and the field work was funded by the US. DOE Environmental Remediation Sciences
Program under grant DOE C05-OOOR22725. Mary Beth Leigh who was a collaborator in
this work was supported by a US. NSF postdoctoral fellowship. We thank Hui Yan,
Kenneth Lowe, and Tonia Mehlhom for help with field and analytic work and Benli Chai
and Qiong Wang from the Ribosomal Database Project for help in analyzing such large

sequence dataset.

117

References

Akob, D. M., H. J. Mills, T. M. Gihring, L. Kerkhof, J. W. Stucki, A. S.
Anastacio, K. J. Chin, K. Kusel, A. V. Palumbo, D. B. Watson, and J. E.
Kostka. 2008. Functional diversity and electron donor dependence of microbial

populations capable of U(VI) reduction in radionuclide-contaminated subsurface
sediments. Appl Environ Microbiol 74:3159-70.

Anderson, R. T. 2006. DOE genomics: Applications to in situ subsurface
bioremediation. Remediation Journal 17:23-38.

Anderson, R. T., H. A. Vrionis, I. Ortiz-Bernad, C. T. Resch, P. E. Long, R.
Dayvault, K. Karp, S. Marutzky, D. R. Metzler, A. Peacock, D. C. White, M.
Lowe, and D. R. Lovley. 2003. Stimulating the in situ activity of Geobacter
species to remove uranium from the groundwater of a uranium-contaminated
aquifer. Appl Environ Microbiol 69:5884—91.

Beller, H. R. 2005 . Anaerobic, nitrate-dependent oxidation of U(IV) oxide
minerals by the chemolithoautotrophic bacterium Thiobacillus denitrificans. Appl
Environ Microbiol 71:2170-4.

Benjamini, Y., and Y. Hochberg. 1995. Controlling the False Discovery Rate: A
Practical and Powerful Approach to Multiple Testing. J. R. Stat. Soc. B 57:289-
300.

Brodie, E. L., T. Z. Desantis, D. C. Joyner, S. M. Back, J. T. Larsen, G. L.
Andersen, T. C. Hazen, P. M. Richardson, D. J. Herman, T. K. Tokunaga, J.
M. Wan, and M. K. Firestone. 2006. Application of a high-density
oligonucleotide microarray approach to study bacterial population dynamics
during uranium reduction and reoxidation. Appl Environ Microbiol 72:6288-98.

Brodie, E. L., T. Z. Desantis, D. C. Joyner, S. M. Back, J. T. Larsen, G. L.
Andersen, T. C. Hazen, P. M. Richardson, D. J. Herman, T. K. Tokunaga, J.
M. Wan, and M. K. Firestone. 2006. Application of a high-density
oligonucleotide microarray approach to study bacterial population dynamics
during uranium reduction and reoxidation. Appl. Environ. Microbiol. 72:6288-98.

118

10.

ll.

12.

13.

14.

15.

16.

Brooks, S. C. 2001. Waste Characteristics of the Former S-3 ponds and outline of
uranium chemistry relevant to NABIR Field Research Center studies ORNL/TM-
2001/27. NABIR Field Research Center.

Cannone, J. J., S. Subramanian, M. N. Schnare, J. R. Collett, L. M. D'Souza,
Y. Du, B. Feng, N. Lin, L. V. Madabusi, K. M. Muller, N. Pande, Z. Shang,
N. Yu, and R. R. Gutell. 2002. The comparative RNA web (CRW) site: an
online database of comparative sequence and structure information for ribosomal,
intron, and other RNAs. BMC Bioinformatics 3:2.

Cardenas, E., W. M. Wu, M. B. Leigh, J. Carley, S. Carroll, T. Gentry, J.
Luo, D. Watson, B. Gu, M. Ginder-Vogel, P. K. Kitanidis, P. M. Jardine, J.
Zhou, C. S. Criddle, T. L. Marsh, and J. M. Tiedje. 2008. Microbial
communities in contaminated sediments, associated with bioremediation of
uranium to submicromolar levels. Appl Environ Microbiol 74:3718-29.

Chao, A., R. L. Chazdon, R. K. Colwell, and T. J. Shen. 2006. Abundance-
based similarity indices and their estimation when there are unseen species in
samples. Biometrics 62:361-71.

Coates, J. D., D. J. Ellis, C. V. Gaw, and D. R. Lovley. 1999. Geothrix
fermentans gen. nov., sp. nov., a novel Fe(III)-reducing bacterium from a
hydrocarbon-contaminated aquifer. Int. J. Syst. Bacteriol. 49 Pt 4: 1615-22.

Coe, R. K. a. R. 2005. Tree diversity analysis. A manual and software for
common statistical methods for ecological and biodiversity studies. World
Agroforestry Centre (ICRAF), Nairobi (Kenya).

Cole, J. R., Q. Wang, E. Cardenas, J. Fish, B. Chai, R. J. Farris, A. S.
Kulam-Syed-Mohideen, D. M. McGarrell, T. Marsh, G. M. Garrity, and J.
M. Tiedje. 2009. The Ribosomal Database Project: improved alignments and new
tools for rRNA analysis. Nucleic Acids Res 37:D141-5.

Cord-Ruwisch, R., D. R. Lovley, and B. Schink. 1998. Growth of geobacter
sulfurreducens with acetate in syntrophic cooperation with hydrogen-oxidizing
anaerobic partners. Appl Environ Microbiol 64:2232-6.

Dufrene, M., and P. Legendre. 1997. Species Assemblages and Indicator
Species: The Need for a Flexible Asymmetrical Approach. Ecological
Monographs 67 :345-366.

119

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Dykhuizen, D. E. 1998. Santa Rosalia revisited: why are there so many species of
bacteria? Antonie Van Leeuwenhoek 73:25-33.

Finneran, K. T., M. E. Housewright, and D. R. Lovley. 2002. Multiple
influences of nitrate on uranium solubility during bioremediation of uranium-
contaminated subsurface sediments. Environ. Microbiol. 4:510-6.

Garrity, G., J. Staley, D. Boone, R. Castenholz, D. Brenner, and N. Krieg
(ed.). 2005. Bergey's Manual of Systematic Bacteriology, Second Edition ed, vol.
Volume Two: The Proteobacteria. Springer — Verlag.

Gu, B., W. M. Wu, M. A. Ginder-Vogel, H. Yan, M. W. Fields, J. Zhou, S.
Fendorf, C. S. Criddle, and P. M. Jardine. 2005. Bioreduction of uranium in a
contaminated soil column. Environ. Sci. Technol. 39:4841-7.

Hansen, M., T. Tolker-Nielsen, M. Givskov, and S. Molin. 1998. Biased 16S
rDNA PCR amplification caused by interference from DNA flanking the template
region. FEMS Microbiology Ecology 26:141-149.

Hazen, T., L. Jiménez, G. Lopez de Victoria, and C. Fliermans. 1991.
Comparison of bacteria from deep subsurface sediment and adjacent groundwater.
Microbial Ecology 22:293-304.

Hazen, T. C., and H. H. Tabak. 2005. Developments in Bioremediation of Soils
and Sediments Polluted with Metals and Radionuclides: 2. Field Research on

Bioremediation of Metals and Radionuclides. Reviews in Environmental Science
and Biotechnology V4:157-183.

Hemme, C. L., T. J. Gentry, L. Wu, M. W. Fields, C. Detter, K. Barry, D.
Watson, C. W. Schadt, P. Richardson, T. Hazen, J. Tiedje, R. 8, E., and J.
Zhou. 2006. Analysis of a Microbial Community Metagenome from Uranium-
Contaminated Groundwater, Florida.

Holmes, D. E., K. T. Finneran, R. A. O'Neil, and D. R. Lovley. 2002.
Enrichment of members of the family Geobacteraceae associated with stimulation
of dissimilatory metal reduction in uranium-contaminated aquifer sediments.
Appl. Environ. Microbiol. 68:2300-6.

Hong, S. H., J. Bunge, S. O. Jeon, and S. S. Epstein. 2006. Predicting microbial
species richness. Proc Natl Acad Sci U S A 103:117-22.

120

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

Hwang, C., W. Wu, T. J. Gentry, J. Carley, G. A. Corbin, S. L. Carroll, D. B.
Watson, P. M. Jardine, J. Zhou, C. S. Criddle, and M. W. Fields. 2009.
Bacterial community succession during in situ uranium bioremediation: spatial
similarities along controlled flow paths. ISME J 3:47-64.

Hwang, C., W. M. Wu, T. J. Gentry, J. Carley, S. L. Carroll, C. Schadt, D.
Watson, P. M. Jardine, J. Zhou, R. F. Hickey, C. S. Criddle, and M. W.
Fields. 2006. Changes in bacterial community structure correlate with initial

operating conditions of a field-scale denitrifying fluidized bed reactor. Appl.
Microbiol. Biotechnol. 71 :748-60.

Istok, J. D., J. M. Senko, L. R. Krumholz, D. Watson, M. A. Bogle, A.
Peacock, Y. J. Chang, and D. C. White. 2004. In situ bioreduction of

technetium and uranium in a nitrate-contaminated aquifer. Environ Sci Technol
38:468-75.

Kodama, Y., and K. Watanabe. 2004. Sulfuricurvum kujiense gen. nov., sp.
nov., a facultatively anaerobic, chemolithoautotrophic, sulfur-oxidizing bacterium

isolated from an underground crude-oil storage cavity. Int J Syst Evol Microbiol
54:2297-2300.

Krieg, N. R., and J. G. Holt. 1984. Bergey's manual of systematic bacteriology.
Williams & Wilkins, Baltimore.

Legendre, P., and E. Gallagher. 2001. Ecologically meaningful transformations
for ordination of species data. Oecologia 129:271-280.

Lovley, D. R., and E. J. Phillips. 1992. Reduction of uranium by Desulfovibrio
desulfuricans. Appl. Environ. Microbiol. 58:850-856.

Lovley, D. R., E. J. P. Phillips, Y. A. Gorby, and E. R. Landa. 1991. Microbial
reduction of uranium. Nature 350:413-416.

Luo, J., W.-M. Wu, J. Carley, C. Ruan, B. Gu, P. M. Jardine, C. S. Criddle,
and P. K. Kitanidis. 2007. Hydraulic performance analysis of a multiple
injection-extraction well system. Journal of Hydrology 336:294-302.

Marciano-Cabral, F., and G. Cabral. 2003. Acanthamoeba spp. as Agents of
Disease in Humans. Clin. Microbiol. Rev. 16:273-307.

121

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

McDonald, J. E., R. J. Lockhart, M. J. Cox, H. E. Allison, and A. J.
McCarthy. 2008. Detection of novel Fibrobacter populations in landfill sites and

determination of their relative abundance via quantitative PCR. Environ
Microbiol 10:1310-9.

Michalsen, M. M., B. A. Goodman, S. D. Kelly, K. M. Kemner, J. P.
McKinley, J. W. Stucki, and J. D. Istok. 2006. Uranium and technetium bio-
immobilization in intermediate-scale physical models of an in situ bio-barrier.
Environ Sci Technol 40:7048-53.

Mohanty, S. R., B. Kollah, D. B. Hedrick, A. D. Peacock, R. K. Kukkadapu,
and E. E. Roden. 2008. Biogeochemical processes in ethanol stimulated
uranium-contaminated subsurface sediments. Environ Sci Technol 42:43 84-90.

N'Guessan, A. L., H. A. Vrionis, C. T. Resch, P. E. Long, and D. R. Lovley.
2008. Sustained removal of uranium from contaminated groundwater following
stimulation of dissimilatory metal reduction. Environ Sci Technol 42:2999-3004.

Nevin, K. P., and D. R. Lovley. 2000. Potential for Nonenzymatic Reduction of
Fe(III) via Electron Shuttling in Subsurface Sediments. Environ. Sci. Technol.
34:2472-2478.

North, N. N., S. L. Dollhopf, L. Petrie, J. D. Istok, D. L. Balkwill, and J. E.
Kostka. 2004. Change in bacterial community structure during in situ
biostimulation of subsurface sediment cocontaminated with uranium and nitrate.
Appl. Environ. Microbiol. 70:4911-20.

Nyman, J. L., T. L. Marsh, M. A. Ginder—Vogel, M. Gentile, S. Fendorf, and
C. Criddle. 2006. Heterogeneous response to biostimulation for U(VI) reduction
in replicated sediment microcosms. Biodegradation V17:303-316.

O'Loughlin, E. J., S. D. Kelly, R. E. Cook, R. Csencsits, and K. M. Kemner.
2003. Reduction of uranium(VI) by mixed iron(II)/iron(III) hydroxide (green
rust): formation of UO2 nanoparticles. Environ Sci Technol 37:721-7.

Riley, R. G., and J. M. Zachara. 1992. Chemical contaminants on DOE lands

and selection of contaminant mixtures for subsurface science research. US.
Department of Energy, Washington, DC.

Roberts, D. W. 2007. Labdsv: Ordination and Multivariate Analysis for Ecology.

122

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Robertson, W. J., J. P. Bowman, P. D. Franzmann, and B. J. Mee. 2001.
Desulfosporosinus meridiei sp. nov., a spore-forming sulfate-reducing bacterium

isolated from gasolene-contaminated groundwater. Int. J. Syst. Evol. Microbiol.
51:133-40.

Sorek, R., Y. Zhu, C. J. Creevey, M. P. Francino, P. Bork, and E. M. Rubin.
2007. Genome-wide experimental determination of barriers to horizontal gene
transfer. Science 318:1449-52.

Spain, A. M., A. D. Peacock, J. D. Istok, M. S. Elshahed, F. Z. Najar, B. A.
Roe, D. C. White, and L. R. Krumholz. 2007. Identification and isolation of a
Castellaniella species important during biostimulation of an acidic nitrate- and
uranium-contaminated aquifer. Appl Environ Microbiol 73:4892-904.

Stevens, J. O. a. R. K. a. P. L. a. B. O. H. a. M. H. H. 2007. Vegan: Community
Ecology Package.

Storey, J., J. Taylor, and D. Siegmund. 2004. Strong control, conservative point
estimation and simultaneous conservative consistency of false discovery rates: a
unified approach. Journal of the Royal Statistical Society Series B 66: 187-205.

Storey, J., and R. Tibshirani. 2003. Statistical significance for genomewide
studies. Proceedings of the National Academy of Sciences of the United States of
America 100:9440-9445.

Suzuki, Y., S. D. Kelly, K. M. Kemner, and J. F. Banfield. 2003. Microbial
populations stimulated for hexavalent uranium reduction in uranium mine
sediment. Appl Environ Microbiol 69:1337-46.

Vrionis, H. A., R. T. Anderson, I. Ortiz-Bernad, K. R. O'Neill, C. T. Resch,
A. D. Peacock, R. Dayvault, D. C. White, P. E. Long, and D. R. Lovley. 2005.
Microbiological and geochemical heterogeneity in an in situ uranium
bioremediation field site. Appl Environ Microbiol 71:6308-18.

Wall, J. D., and L. R. Krumholz. 2006. Uranium reduction. Annu Rev
Microbiol 60: 149-66.

Wan, J., T. K. Tokunaga, E. Brodie, Z. Wang, Z. Zheng, D. Herman, T. C.
Hazen, M. K. Firestone, and S. R. Sutton. 2005. Reoxidation of bioreduced
uranium under reducing conditions. Environ. Sci. Technol. 39:6162-9.

123

57.

58.

59.

60.

61.

62.

Wang, Q., G. M. Garrity, J. M. Tiedje, and J. R. Cole. 2007. Naive Bayesian
classifier for rapid assignment of rRNA sequences into the new bacterial
taxonomy. Appl Environ Microbiol 73:5261-7.

Wu, W. M., J. Carley, M. Fienen, T. Mehlhorn, K. Lowe, J. Nyman, J. Luo,
M. E. Gentile, R. Rajan, D. Wagner, R. F. Hickey, B. Gu, D. Watson, 0. A.
Cirpka, P. K. Kitanidis, P. M. Jardine, and C. S. Criddle. 2006. Pilot-scale in
situ bioremediation of uranium in a highly contaminated aquifer. 1. Conditioning
of a treatment zone. Environ Sci Technol 40:3978-85.

Wu, W. M., J. Carley, T. Gentry, M. A. Ginder-Vogel, M. Fienen, T.
Mehlhorn, H. Yan, S. Caroll, M. N. Pace, J. Nyman, J. Luo, M. E. Gentile,
M. W. Fields, R. F. Hickey, B. Gu, D. Watson, 0. A. Cirpka, J. Zhou, S.

F endorf, P. K. Kitanidis, P. M. Jardine, and C. S. Criddle. 2006. Pilot-scale in
situ bioremediation of uranium in a highly contaminated aquifer. 2. Reduction of
u(VI) and geochemical control of u(VI) bioavailability. Environ Sci Technol
40:3986-95.

Wu, W. M., J. Carley, J. Luo, M. Ginder-Volgel, E. Cardenas, M. B. Leigh,
C. Hwang, S. D. Kelly, C. Ruan, L. Wu, J. van Nostand, T. Gentry, K. Lowe,
T. Melhourne, S. Carroll, M. W. Fields, B. Gu, D. Watson, K. M. Kemner, T.
L. Marsh, J. M. Tiedje, J. Zhou, P. K. Fendorf, P. K. Kitanidis, P. M.
Jardine, and C. S. Criddle. 2007. In situ bioremediation of uranium (VI) to

submicromolecular levels and reoxidation by dissolved oxygen. Environ. Sci.
Technol. 41:5716-5723.

Wu, W. M., B. Gu, M. W. Fields, M. Gentile, Y. K. Ku, H. Yan, S. Tiquias, T.
Yan, J. Nyman, J. Zhou, P. M. Jardine, and C. S. Criddle. 2005. Uranium (VI)
Reduction by Denitrifying Biomass. Bioremediation 9:46-61.

Yamada, T., Y. Sekiguchi, S. Hanada, H. Imachi, A. Ohashi, H. Harada, and
Y. Kamagata. 2006. Anaerolinea thermolimosa sp. nov., Levilinea saccharolytica
gen. nov., sp. nov. and Leptolinea tardivitalis gen. nov., sp. nov., novel
filamentous anaerobes, and description of the new classes Anaerolineae classis

nov. and Caldilineae classis nov. in the bacterial phylum Chloroflexi. Int J Syst
Evol Microbiol 56: 133 1-40.

124

CHAPTER FOUR

Cardenas, E., and J. M. Tiedje. 2008. New tools for discovering and characterizing

microbial diversity. Curr Opin Biotechnol 19:544-9.

125

New tools for discovering and characterizing microbial diversity

Summary

To discover and characterize microbial diversity, approaches based on new
sequencing technologies, novel isolation techniques, microfluidics and metagenomics
among others are being used. These approaches have contributed to discovery of novel
genes from environmental samples, to massive characterization of functional and
phylogenetic genes and to isolation of members of formerly uncultured yet ubiquitous
groups like Verrucomicrobia, Acidobacteria, OP10 and methanogenic Archaea. Cheaper
sequencing is key in this process by making available applications that were previously
restricted to big research centers, complementing previously available methodologies and
potentially replacing some of them. The new tools are reshaping the way we study the
environment, increasing the resolution at which microbial communities, their
complexities and dynamics, can be studied to reveal their genetic potential and their

functional diversity.

Introduction

Discovering new microbes and characterizing their functions are major goals in
the study of microbial diversity. Historically, this was achieved through cultivation and
subsequent characterization of strains. Decades later, culture-independent methods have
provided new tools to study the microbial world. Recently, metagenomics is providing a

gene-based exploration of the community as a whole while isolation and cultivation

126

continue as the main-stay for testing metabolic abilities and more detailed genomic
studies of individual members of this community. As more and more environments were
sampled, it became evident that the majority of the microbes have yet to be cultured (30),
driving the development of new methods to unveil the unknown (Figure 4.1). These
methods benefit from advances in sequencing, isolation and microfluidics and allow
researchers to study microbial communities in their natural environments and with higher

resolution.

127

I ‘_ . _ I 1
,3 “*"-m“‘*"“‘ 'A ‘ * ' «g. .g..mu‘:'m W Mao” await-mu.

,.. w. . o
W ’\.~ P ‘I w
a, r .

 

 

 

I '~
- . r" , 1
Community Structure 1 Genetic Potential
' (Who is there?) FE (What can they do?)
' i
. - Phylogenetic arrays i - Metagenomics
" ' 16S rRNAtag sequencing l E - Functional gene arrays
- Community Image analysis 1 ‘ - Neural Gene Networks
l. (CMEIAS) - E
3*- ~ . a}
.gA‘Ifi'éh.:A-utouur, M. «New; .23; 5.1-...)"13'Jui‘; .. -33... g; ’2’;.“Qmsru .-.-~.u.-...x-...-e..

fir
K;
.‘
6'

Gene expression Functional characterization

(What are they doing?) '3

4‘1.

— Novel isolation methods

- Metatranscriptomics ~ - Stable isotope probing
- Metaproteom . ' NaHOSIMS
; , . CARD-FISH

 

 

Fig: 4.1. Novel methods to discover 21:1 ficharfiacterize microbial diversity. In boldjr:
the methods reviewed in this article. The image showing the morphological diversity
from an anaerobic bioreactor community is courtesy of Frank Dazzo.
Impact of the new sequencing technologies in microbial ecology

As we write this report, new sequencing technologies are changing the way we
study microbial communities. Traditionally, the study of genes from natural
environments included cloning DNA into a vector, inserting that vector into a host,
screening, and Sanger sequencing. Sequence-by-synthesis methods provide faster,
cheaper and simpler methods for (meta)genome sequence that bypass the cloning bias
and labor-intensive Sanger method. New sequencing technologies differ in type of
output, read length, speed, and error rate distribution (Table 4.1); for a more detailed

review see ref. (34). Single molecule sequencing, expected soon by Helicos and Pacific

Biosciences, will provide another leap in sequence capacity with the latter company

128

projecting reads as long as 10 kilobases. The different sequencers are being used for
different applications: longer reads being preferred for genome sequencing and gene
surveys while shorter reads being the choice for transcriptomes, small RNAs analysis and
comparison of closely related strains or strain variants. These methods compensate their
shortcomings in read length and error bias with massive output (and thus coverage),
speed and simplicity. The new methodologies create a situation where the limitation is
not the ability to produce sequence data but the ability to store and analyze it in new

revealing ways.

Table 4.1. Currently available sequencing technologies and their features. Not included
in the table are technologies under development by Visigen Biotechnologies, Polonator,
Pacific Biosciences, Intelligent Biosystems, Complete Genomics. N/A, not available.

 

 

Method Output per run Read length System described
(per day) (bases)
52-86 kB ABI 3730xl DNA
sang“ (0.69-2.1MB) 550'900 Analyzer. 96 Well
Roche 454 Life Sciences
Pyrosequencing 400-600 MB 400 GS FLX Titanium
Series Reagents
. N/A The HeliScope Single
Helicos (600-2160 MB) 30 Molecule Sequencer
SOLiD 6 GB (600MB) 35 SOLiD system
Illumina Illumina Genome
(Solexa) 3GB (600MB) 36 Analyzer

 

Currently pyrosequencing is favored for microbial genome sequencing and
metagenomic studies. It is valuable for sequencing regions that are technically difficult
(e. g. due to strong secondary structure or high GC content) and for covering regions
recalcitrant to cloning in Escherichia coli (36). In one example, a Sanger/pyrosequencing

hybrid approach completely eliminated gaps in two genomes and reduced gaps by 86% in

129

 

four other genomes (12). In the large sequencing centers, the Sanger method is being
replaced as the main sequencing method for bacterial genomes though it remains an
important tool for closing them. See also (32) for a more historical perspective of 454

sequencing.

A major advantage of the new sequencing technologies is the substantial
reduction in cost, democratizing genomics and making massive gene surveys feasible as a
way to discover new microbes and genes. This is especially useful for the study of
microbial communities by 16S rRNA gene analysis, the most commonly used species
proxy. By combining primers that target hypervariable regions of the 168 rRN A gene
with nucleotide barcodes, it is now more feasible to survey environments with thousands
of sequences at the time. The hypervariable regions targeted are short enough (100-3 50
bases) to be covered by some of the new sequencing technologies yet long enough to be
informative for classification by the current rRNA databases (e.g. Ribosomal database
project (RDP), Greengenes, Silva), though they are poor for phylogeny in comparison to
the full length 168 rRN A gene (~1500 bases). The RDP has reported accuracy in
classification of partial sequences to the genus level for 400-base reads and to the family
level for ZOO-base reads (44). The first in the new wave of massive 16S rRNA gene
studies was of deep sea sediments using the V6 hypervariable region (35). Since then,
other regions have been used to study temperate and tropical soils (31). Highlighting the
magnitude of the species diversity question, in one case 750,000 sequences were not
enough to recover all the microbial diversity present in deep sea communities near

hydrothermal vents (15). Longer reads will allow targeting of regions that provide a

130

better classification (44) and to use the sequences for more accurate phylogenetic studies.
A complementary approach to massive sequencing is the adoption of nucleotide barcodes
in the amplification primers. In this method, samples from different origins can be mixed
in one run and after sequencing their data can be separated according to their barcode (22,
25). This approach decreases the cost per sample since more samples can be pooled in the
sequencing run rather than sequencing fewer samples to greater depth. For example,
using 454 F LX pyrosequencing 400,000 reads are usually generated; by using 80
barcodes, 5000 reads per sample can be obtained. This number of sequences is at least

one order of magnitude higher than those from traditional clone libraries.

Transcriptomics is also being transformed by the new sequencing technologies.
Gene prediction algorithms, probe design, cloning, and hybridization steps involved in
traditional transcriptomic analyses can all be bypassed by massively parallel direct
sequencing of cDNA and subsequently mapping the reads to a reference genome to
quantify known transcripts and even detect putative new genes. For example, Illumina
(Solexa) sequencing has been used to detect genes, intron usage and alternative initiation
codons in yeast (23). In a more recent article, 512 new genes were predicted for the
nitrogen-fixing plant symbiont Sinorhizobium meliloti after a transcriptome analysis that

used pyrosequencing (19).

A new look at metagenomics

Metagenomics, the study of microbial communities based on its genetic material

(6), has paved the way for the discovery of new genes, proteins and biochemical

131

pathways. Genes involved in anaerobic methane oxidation were discovered with a
metagenomics effort that targeted an association of Archaea and sulfate-reducing bacteria
(13). Metagenomics has proved more valuable in recovering complete genomes in
systems with low diversity, such as the acid drainage from an abandoned mine (43) or in
highly enriched cultures such as the first genome of the annamox (anaerobic ammonia
oxidation) group, Kuenenia stuttgartiensis (41). When applied to more diverse
environments, such as soils, the yield of assembled genomes has been poor (3 8).
However, assembly is not necessary to make some important inferences. One particularly
valuable use of a metagenome can be as the basis for metranscriptomics (9) and

metaproteomics (28).

Array-based approaches for studying microbial diversity

The study of the microbial community structure and its functional diversity has
benefited from the use of microarrays that target phylogenetic markers and functional
genes. Arrays provide high throughput, specificity and relative quantification free of
cloning bias. A functional gene array, Geochip (14), has been helpful in the study of
nitrogen and carbon cycles in Antarctic soils (47) and PCB contaminated soils (18).
Arrays targeting the 16S rRN A gene have been used to study microbial communities
involved in uranium reduction and reoxidation (4). Phylogenetic arrays easily detect more
microbial diversity than traditional clone libraries (7). Nevertheless, probe design limits
detection to what is already known, or what can be placed on a chip, and hence
microarrays are tools better suited to generally characterize ecosystems rather than to

discover new genes and functions.

132

 

Microfluidics, manipulation at micro(be) scale

Microfluidics, the manipulation of small volumes, is becoming a powerful tool in
microbial ecology. By working at dimensions relevant to microbes it has been possible to
study specific microbes via isolation, gene amplification or genome reconstruction.
Microfluidics allowed for individual members of termite hindgut flora to be sorted and
screened by multiplex PCR (24). By targeting the 16S rRNA gene and functional genes it
was possible to unambiguously link phylogeny and function (no cultivation required),
and to detect horizontal gene transfer events (24). This approach has been coupled with
whole-genome amplification steps that target more genes and to sequence much of an
organism (3 7). A genome from the candidate phylum TM7 has been reconstructed from
human mouth samples with the help of a microfluidic device that selected for rod-like

morphology, amplified individual genomes and screened for TM7 signatures (20).

The formerly unculturables — New approaches for isolation

Renewed efforts in isolation are being taken to capture more of the microbial
functional diversity. These novel methods demonstrate that culturing is often possible if
niche conditions are mimicked. In contrast with traditional cultivation, the novel isolation
conditions try to better reflect natural environments in terms of nutrients (composition
and concentration), oxygen levels, pH and natural associations with other
microorganisms. These novel isolation methods also benefit from advances in

microfluidics.

133

Novel lineages of Actinobacteria, Acidobacteria, Proteobacteria and
Verrucomicrobia were isolated using diluted nutrient broth solidified with gellan (16).
Gellan is a good alternative solidifying reagent producing gels that are clearer (easier
screening) and more stable than agar under a wider range of temperatures and pHs.
Gellan also worked as a carbon source when isolating the first cultured member of the
OP10 division from soils (40). In one study, colony counts increased up to 21.3 times
when replacing agar with gellan in culture medium (42). In terms of oxygen
concentration, microaerophillic conditions have led to increased recovery of termite gut
bacteria (46) and soil Acidobacteria (8).

In situ and trap cultivation introduce sterile media to natural environments to use
local conditions for enrichment. Using in situ cultivation with diffusion chambers, the
diversity of isolates recovered from freshwater sediments increased and isolates from
some rarely cultivated groups, such as Acidobacteria and Verrucomicrobia, were
recovered (3). Trap cultivation has been used to capture novel soil Actinobacteria, known

secondary metabolite producers, whose filaments penetrate the medium when growing

(10).

Because organisms rarely live in pure cultures, cocultivation is another strategy
for isolation. This is well established for hydrogenotrophic methanogenic Archaea and
their partners. In these associations Archaea consume hydrogen, a product of incomplete
substrate oxidation in anaerobic environments and make the oxidation reaction

thermodynamically feasible (17). Through cocultivation, a globally abundant methanogen

134

(with a propionate-oxidizing, H2 producing partner) (33), and an anaerobic phenol
degrader (and its hydrogen consuming Archaeal syntroph) (27) have been isolated.
Carbon exchange occurring in natural associations can also be used for isolation
purposes. Novel heterotrophic partners of algae have been isolated by adding algae-
excreted organic material to culture media (21) or by creating defined media to simulate

the exudates’ composition (45).

Another novel isolation approach uses of a combination of high throughput
screening, microfluidics and dilution-to-extinction techniques originally used by Casida
(5). In a current version of this approach(39), small volumes of inocula, sometimes
picoliters, are added to sterile medium, typically in a 96-well plate format. By having few
cells per well, the competition with fast growers is greatly reduced. After incubation
different preparations can be screened for growth using fluorescent staining or group-
specific probes. The most dominant group in marine ecosystems (SARll clade) (29) and
novel isolates of Proteobacteria, Actinobacteria, F irmicutes and F lavobacteria-
Cytophaga (11) have been obtained using this approach. There appears to be no reason
why most of the formerly unculturable microbes cannot be made culturable with
sufficient study, cleverness and patience. Study of cultured isolates remains the most
straightforward method to characterize an organism’s functions and evaluate its

biotechnological potential.

Metatranscriptomics and metaproteomics, fields in their infancy

135

The study of gene expression and translation into proteins within natural
environments are two emerging fields in microbial ecology which hold special promise
for studying function. Recovering mRN A and proteins from natural environments is
challenging, yet several groups have pace-setting examples. The first metatranscriptome
study recovered transcripts encoding Archaea and Bacteria genes involved in carbon,
sulfur and nitrogen cycling from marine and freshwater bacterioplankton communities
(26). A marine metatranscriptome study found that hypothetical genes were among the
most highly expressed in Prochlorococcus spp. and that a large fraction of the transcripts
recovered were novel, indicating the immense metabolic diversity hidden in the oceans
(9). In a soil transcriptome, even though most of the transcripts coded for housekeeping
proteins, genes linked to soil processes were also found (1). In addition, 32% of
eukaryote transcripts recovered carried novel genes (1), again highlighting the vast
unknown within hypothetical genes. In the case of metaproteomics, proteins identified as
chlorocatechol dioxygenases were recovered from a chlorobenzene-contaminated aquifer,

a finding that agrees with the predicted degradation pathway of chlorobenzene (2).

Many challenges remain in metagenomic, metatranscriptomic and metaproteomic
approaches, including DNA, RNA and protein extraction, mRN A instability, low
abundance, and low proportion of mRNA in total RNA. As these are still novel
approaches, new solutions to these technical problems are expected to arise, which would
allow them to reach their full potential of characterizing functional diversity. Databases

and software tools (Table 4.2) are essential and their further development is needed to

136

deal with the growing bottleneck in the analysis of metagenomic, metatranscriptomic and

metaproteomic data.

Table 4.2. Selected resources available for metagenomics, metatranscriptomics and

 

 

mass spectrometry
data to identify
proteins from primary
sequence databases

 

 

 

 

 

 

 

 

 

 

 

metaproteomics.
Resource Features Website
Data repository and
CAMERA bioinformatics tools httgflcameracal illnet
resource
Integrated
Microbial .
Genomes with Comparative . . . . . . . . .
. s , r ’ /
Microbiome metagenome analysrs http/hmgjgdoegov, CQl-bm/m.» 1113”].ng
and annotation tools
samples
(IMG/M)
Human Oral Database on bacteria
Microbiome from the human oral h_ttp://www.h0md.org
Database cavity
SEED Comparative genomics http://wwwtheseedoru
envuonment '
Metagenome Analysis http://www-ab.infomiatikuni-
MEGAN . , - . ,. ,
Software tuebmgen .de/sottwai e/ m «3221 n/
Automated
33%c;mfs ifs‘intagggllliciing analysrs http://metagenomics.nmpdr.org/
environment
PFAM Collection of protein http://pfamjaneliaorgj
families using multiple
sequence alignments
and hidden Markov
models
Phylofacts Structural http://phylogenomics.berkelevedu/phvlofacts/indexphp
phylogenomic
encyclopedia for
protein functional and
structural classification
Uniprot Protein sequence and http://wwwuniprot.org1
fimctional information
Mascot Search engine that uses http://www.matrixscience.comKhome.html

137

Future challenges and directions

Massively parallel sequencing, metagenomics, metatranscriptomics,
metaproteomics, and isolation will allow us to study deeper and deeper layers of the
microbial diversity, including genes behind novel or valued functions. These methods
promise to broaden the use of (meta)genomics and offer the chance to study specific
groups that were previously unknown and uncharacterized. We can now ask more
complex questions such as what is the role of low abundant organisms?, can we estimate
the total diversity of the system and assess its value?, how important is it to know in what
organism particular genes reside?, and more importantly, how can we use this
information to better characterize, manage and/or harvest even greater functional

diversity. The ball is in our court.

Reference annotations

° of special interest ” of outstanding interest

Sogin ML, Morrison HG, Huber JA, Mark Welch D, Huse SM, Neal PR, Arrieta JM,
Hemdl GJ: Microbial diversity in the deep sea and the underexplored. Proc Natl

AcadSci USA 2006, 103:12115-12120.

"First paper on massive parallel l6S rRNA gene surveys. Deep sea samples were studied

by targeting the V6 hypervariable region. By comparing against a custom V6 database, it

138

was found that 80% of the sequences were divergent by 10% or more, suggesting an

immense diversity present in rare members of the community.

Nagalakshmi U, Wang Z, Waern K, Shou C, Raha D, Gerstein M, Snyder M: The
transcriptional landscape of the yeast genome defined by RNA sequencing. Science

2008, 320:1344-1349.

"Transcriptomic study using massive parallel sequence of cDNA. By comparing the
transcripts against the yeast genome it was possible to detect alternative splicing,

initiation sites and delimit the boundaries of genes.

Mao C, Evans C, Jensen RV, Sobral BW: Identification of new genes in
Sinorhizobium meliloti using the Genome Sequencer F LX system. BMC Microbiol

2008, 8:72.

- Massive sequencing of cDNA using pyrosequencing. Putative new genes were found,
these genes were likely missed by prediction algorithm because they were short or had

low scores against known databases.

F rias-Lopez J, Shi Y, Tyson GW, Coleman ML, Schuster SC, Chisholm SW, DeLong
EF: Microbial community gene expression in ocean surface waters. Proc Natl Acad

Sci USA 2008, 105:3805-3810.

"A study of DNA and cDNA from the same marine community. Proof that gene
expression in complex systems if possible by mapping the transcripts to the metagenome.

Interestingly, some of the more highly expressed genes were hypothetical and a large

139

fraction of cDNA reads were not present in databases, suggesting that large part of the

microbial diversity remains hidden in the oceans.

He Z, Gentry TJ, Schadt CW, Wu L, Liebich J, Chong SC, Huang Z, Wu W, Gu B,
Jardine P, et al.: GeoChip: a comprehensive microarray for investigating

biogeochemical, ecological and environmental processes. ISME J 2007, 1:67-77.

°Description of the functional gene array Geochip.

Brodie EL, Desantis TZ, Joyner DC, Baek SM, Larsen JT, Andersen GL, Hazen TC,
Richardson PM, Herman DJ, Tokunaga TK, et al.: Application of a high-density
oligonucleotide microarray approach to study bacterial population dynamics during

uranium reduction and reoxidation. Appl Environ Microbiol 2006, 72:6288-6298.

°First example of the use of high-density phylogenetic arrays to profile microbial
communities in an uranium contaminated aquifer. The results suggest a role of member

of the Geobacter spp. Geothrixfermentans in uranium reduction.

Stepanauskas R, Sieracki ME: Matching phylogeny and metabolism in the uncultured

marine bacteria, one cell at a time. Proc Natl Acad Sci U S A 2007, 104:9052-9057.

°An approach that combines flow cytometry, whole genome amplification and screening
to allow single cells to be separated and their genome amplified. The method allows

targeting metabolic and phylogenetic markers in the same uncultured microbe.

140

Marcy Y, Ouvemey C, Bik EM, Losekann T, Ivanova N, Martin HG, Szeto E, Platt D,
Hugenholtz P, Relman DA, et al.: Dissecting biological "dark matter" with single-cell

genetic analysis of rare and uncultivated TM7 microbes from the human mouth.

Proc Natl Acad Sci USA 2007, 104:11889-11894.

"A microfluidic device that allows selection of specific uncultured bacterium and

amplification and recovery of their genome.

Stott MB, Crowe MA, Mountain BW, Smirnova AV, Hou S, Alam M, Dunfield PF:
Isolation of novel bacteria, including a candidate division, from geothermal soils in

New Zealand. Environ Microbiol 2008.

- A fine example of novel isolation methods that combine low nutrient conditions, and

alternative gelling agents to recover novel bacteria from soils.

Bollmann A, Lewis K, Epstein SS: Incubation of environmental samples in a diffusion
chamber increases the diversity of recovered isolates. Appl Environ Microbiol 2007,

73:63 86-6390.

0 By using a diffusion chamber incubated in situ, rarely recovered groups from

Deltaproteobacteria, Verrucomicrobia, Spirochaetes, and Acidobacteria were isolated.

141

Stingl U, Tripp HJ, Giovannoni SJ: Improvements of high-throughput culturing
yielded novel SARll strains and other abundant marine bacteria from the Oregon

coast and the Bermuda Atlantic Time Series study site. ISME J 2007, 1:361-371.

0- Excellent example of high-throughput dilution-to-extinction culturing yielding novel

isolates from the globally abundant SARll group.

Acknowledgments

This work was supported by the US. DOE under grants DE-FG02-97ER62469 and DE-
FG02-99ER62848, and NIEHS grant 5P42 ESOO491 1-18/19. The authors thank Brian

Campbell, Dieter Tourlousse and Patrick Chain for useful comments and discussion.

142

REFERENCES

Bailly, J., L. F raissinet-Tachet, M. C. Vemer, J. C. Debaud, M. Lemaire, M.
Wesolowski-Louvel, and R. Marmeisse. 2007. Soil eukaryotic functional
diversity, a metatranscriptomic approach. ISME J 1:632-42.

Benndorf, D., G. U. Balcke, H. Harms, and M. van Bergen. 2007. Functional
metaproteome analysis of protein extracts from contaminated soil and
groundwater. ISME J 1:224-34.

Bollmann, A., K. Lewis, and S. S. Epstein. 2007. Incubation of environmental
samples in a diffusion chamber increases the diversity of recovered isolates. Appl
Environ Microbiol 73:63 86-90.

Brodie, E. L., T. Z. Desantis, D. C. Joyner, S. M. Baek, J. T. Larsen, G. L.
Andersen, T. C. Hazen, P. M. Richardson, D. J. Herman, T. K. Tokunaga, J.
M. Wan, and M. K. Firestone. 2006. Application of a high-density
oligonucleotide microarray approach to study bacterial population dynamics
during uranium reduction and reoxidation. Appl Environ Microbiol 72:6288-98.

Casida, L. E., Jr. 1965. Abundant Microorganism in Soil. Appl Microbiol
13:327-34.

Committee on Metagenomics: Challenges and Functional Applications, N. R.
C. 2007. The New Science of Metagenomics: Revealing the Secrets of our
Microbial Planet. National Academies Press, Washington, DC.

DeSantis, T. 2., E. L. Brodie, J. P. Moberg, I. X. Zubieta, Y. M. Piceno, and
G. L. Andersen. 2007. High-density universal 16S rRNA microarray analysis
reveals broader diversity than typical clone library when sampling the
environment. Microb Ecol 53:371-83.

Eichorst, S. A., J. A. Breznak, and T. M. Schmidt. 2007. Isolation and
characterization of soil bacteria that define Terriglobus gen. nov., in the phylum
Acidobacteria. Appl Environ Microbiol 73:2708-17.

143

10.

11.

12.

13.

14.

15.

16.

17.

Frias-Lapez, J., Y. Shi, G. W. Tyson, M. L. Coleman, S. C. Schuster, S. W.
Chisholm, and E. F. Delong. 2008. Microbial community gene expression in
ocean surface waters. Proc Natl Acad Sci U S A 105:3805-10.

Gavrish, E., A. Bollmann, S. Epstein, and K. Lewis. 2008. A trap for in situ
cultivation of filamentous actinobacteria. J Microbiol Methods 72:257-62.

Gich, F., K. Schubert, A. Bruns, H. Hoffelner, and J. Overmann. 2005.
Specific detection, isolation, and characterization of selected, previously

uncultured members of the freshwater bacterioplankton community. Appl Environ
Microbiol 71:5908-19.

Goldberg, S. M., J. Johnson, D. Busam, T. Feldblyum, S. Ferriera, R.
Friedman, A. Halpern, H. Khauri, S. A. Kravitz, F. M. Laura, K. Li, Y. H.
Rogers, R. Strausberg, G. Sutton, L. Tallan, T. Thomas, E. Venter, M.
Frazier, and J. C. Venter. 2006. A Sanger/pyrosequencing hybrid approach for
the generation of high-quality draft assemblies of marine microbial genomes. Proc
Natl Acad Sci U S A 103:11240-5.

Hallam, S. J., N. Putnam, C. M. Preston, J. C. Detter, D. Rokhsar, P. M.
Richardson, and E. F. DeLong. 2004. Reverse methanogenesis: testing the
hypothesis with environmental genomics. Science 305: 1457-62.

He, Z., T. J. Gentry, C. W. Schadt, L. Wu, J. Liebich, S. C. Chang, Z. Huang,
W. Wu, B. Gu, P. Jardine, C. Criddle, and J. Zhou. 2007. GeoChip: a
comprehensive microarray for investigating biogeochemical, ecological and
environmental processes. ISME J 1:67-77.

Huber, J. A., D. B. Mark Welch, H. G. Morrison, S. M. Huse, P. R. Neal, D.
A. Butterfield, and M. L. Sogin. 2007. Microbial population structures in the
deep marine biosphere. Science 318:97-100.

Janssen, P. H., P. S. Yates, B. E. Grintan, P. M. Taylor, and M. Sait. 2002.
Improved culturability of soil bacteria and isolation in pure culture of novel
members of the divisions Acidobacteria, Actinobacteria, Proteobacteria, and
Verrucomicrobia. Appl Environ Microbiol 68:2391-6.

Kamagata, Y., and H. Tamaki. 2005. Cultivation of Uncultured Fastidiaus
Microbes. Microbes and Environments 20:85-91.

144

18.

19.

20.

21.

22.

23.

24.

25.

26.

Leigh, M. B., V. H. Pellizari, O. Uhlik, R. Sutka, J. Rodrigues, N. E. Ostrom,
J. Zhou, and J. M. Tiedje. 2007. Biphenyl-utilizing bacteria and their fimctional

genes in a pine root zone contaminated with polychlorinated biphenyls (PCBs).
ISME J 1:134-48.

Mao, C., C. Evans, R. V. Jensen, and B. W. Sobral. 2008. Identification of new
genes in Sinorhizobium meliloti using the Genome Sequencer F LX system. BMC
Microbial 8:72.

Marcy, Y., C. Ouverney, E. M. Bik, T. Losekann, N. Ivanova, H. G. Martin,
E. Szeto, D. Platt, P. Hugenholtz, D. A. Relman, and S. R. Quake. 2007.
Dissecting biological "dark matter" with single-cell genetic analysis of rare and
uncultivated TM7 microbes from the human mouth. Proc Natl Acad Sci U S A
104:11889-94.

Mayali, X., P. J. Franks, and F. Azam. 2008. Cultivation and ecosystem role of
a marine raseabacter clade-affiliated cluster bacterium. Appl Environ Microbial

74:2595-603.

Meyer, M., U. Stenzel, S. Myles, K. Prufer, and M. Hofreiter. 2007. Targeted
high-throughput sequencing of tagged nucleic acid samples. Nucleic Acids Res
35:e97.

Nagalakshmi, U., Z. Wang, K. Waern, C. Shou, D. Raha, M. Gerstein, and
M. Snyder. 2008. The transcriptional landscape of the yeast genome defined by
RNA sequencing. Science 320:1344-9.

Ottesen, E. A., J. W. Hong, S. R. Quake, and J. R. Leadbetter. 2006.
Microfluidic digital PCR enables multigene analysis of individual environmental
bacteria. Science 314: 1464-7.

Parameswaran, P., R. Jalili, L. Tao, S. Shokralla, B. Gharizadeh, M.
Ronaghi, and A. Z. Fire. 2007. A pyrosequencing-tailored nucleotide barcode

design unveils opportunities for large-scale sample multiplexing. Nucleic Acids
Res 35:e130.

Poretsky, R. S., N. Bano, A. Buchan, G. LeCleir, J. Kleikemper, M.
Pickering, W. M. Pate, M. A. Moran, and J. T. Hollibaugh. 2005. Analysis of

microbial gene transcripts in environmental samples. Appl Environ Microbiol
71:4121-6.

145

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

Qiu, Y. L., S. Hanada, A. Ohashi, H. Harada, Y. Kamagata, and Y.
Sekiguchi. 2008. Syntrophorhabdus aromaticivorans gen. nov., sp. nov., the first
cultured anaerobe capable of degrading phenol to acetate in obligate syntrophic

associations with a hydrogenotrophic methanogen. Appl Environ Microbiol
74:2051-8.

Ram, R. J., N. C. Verberkmoes, M. P. Thelen, G. W. Tyson, B. J. Baker, R.
C. Blake, 2nd, M. Shah, R. L. Hettich, and J. F. Banfield. 2005. Community
proteomics of a natural microbial biofilm. Science 308:1915-20.

Rappe, M. S., S. A. Cannon, K. L. Vergin, and S. J. Giovannoni. 2002.
Cultivation of the ubiquitous SARll marine bacterioplankton clade. Nature
418:630-3.

Rappe, M. S., and S. J. Giovannoni. 2003. The uncultured microbial majority.
Annu Rev Microbiol 57:369-94.

Roesch, L. F., R. R. Fulthorpe, A. Riva, G. Casella, A. K. Hadwin, A. D.
Kent, S. H. Daroub, F. A. Camargo, W. G. Farmerie, and E. W. Triplett.

2007. Pyrosequencing enumerates and contrasts soil microbial diversity. ISME J
1:283-90.

Rothherg, J. M., and J. H. Leamon. 2008. The development and impact of 454
sequencing. Nat Biotechnol 26:1 1 17-24.

Sakai, 8., H. Imachi, Y. Sekiguchi, A. Ohashi, H. Harada, and Y. Kamagata.
2007. Isolation of key methanogens for global methane emission from rice paddy

fields: a novel isolate affiliated with the clone cluster rice cluster 1. Appl Environ
Microbiol 73:4326-31 .

Shendure, J., and H. Ji. 2008. Next-generation DNA sequencing. Nat
Biotechnol 26:1 135-45.

Sogin, M. L., H. G. Morrison, J. A. Huber, D. Mark Welch, S. M. Huse, P. R.
Neal, J. M. Arrieta, and G. J. Herndl. 2006. Microbial diversity in the deep sea
and the underexplored. Proc Natl Acad Sci U S A 103:12115-20.

Sorek, R., Y. Zhu, C. J. Creevey, M. P. Francino, P. Bork, and E. M. Rubin.
2007. Genome-wide experimental determination of barriers to horizontal gene
transfer. Science 318:1449-52.

146

37.

38.

39.

40.

41.

42.

43.

44.

Stepanauskas, R., and M. E. Sieracki. 2007. Matching phylogeny and

metabolism in the uncultured marine bacteria, one cell at a time. Proc Natl Acad
Sci U S A 104:9052-7.

Steward, G. F., and M. S. Rappe. 2007. What's the 'meta' with metagenomics?
ISME J 1:100-2.

Stingl, U., H. J. Tripp, and S. J. Giovannoni. 2007. Improvements of high-
throughput culturing yielded novel SAR11 strains and other abundant marine

bacteria from the Oregon coast and the Bermuda Atlantic Time Series study site.
ISME J 1:361-71.

Stott, M. B., M. A. Crowe, B. W. Mountain, A. V. Smirnova, S. Hou, M.
Alam, and P. F. Dunfield. 2008. Isolation of novel bacteria, including a
candidate division, from geothermal soils in New Zealand. Environ Microbiol.

Strous, M., E. Pelletier, S. Mangenot, T. Rattei, A. Lehner, M. W. Taylor, M.
Horn, H. Daims, D. Bartol-Mavel, P. Wincker, V. Barbe, N. Fonknechten, D.
Vallenet, B. Segurens, C. Schenowitz-Truong, C. Medigue, A. Collingro, B.
Snel, B. E. Dutilh, H. J. Op den Camp, C. van der Drift, I. Cirpus, K. T. van
de Pas-Schoonen, H. R. Harhangi, L. van Niftrik, M. Schmid, J. Keltjens, J.
van de Vossenberg, B. Kartal, H. Meier, D. Frishman, M. A. Huynen, H. W.
Mewes, J. Weissenhach, M. S. Jetten, M. Wagner, and D. Le Paslier. 2006.
Deciphering the evolution and metabolism of an anammox bacterium from a
community genome. Nature 440:790-4.

Tamaki, H., Y. Sekiguchi, S. Hanada, K. Nakamura, N. Nomura, M.
Matsumura, and Y. Kamagata. 2005. Comparative analysis of bacterial
diversity in freshwater sediment of a shallow eutrophic lake by molecular and
improved cultivation-based techniques. Appl Environ Microbiol 71:2162-9.

Tyson, G. W., J. Chapman, P. Hugenholtz, E. E. Allen, R. J. Ram, P. M.
Richardson, V. V. Solovyev, E. M. Rubin, D. S. Rokhsar, and J. F. Banfield.
2004. Community structure and metabolism through reconstruction of microbial
genomes from the environment. Nature 428:37-43.

Wang, Q., G. M. Garrity, J. M. Tiedje, and J. R. Cole. 2007. Naive Bayesian
classifier for rapid assignment of rRNA sequences into the new bacterial
taxonomy. Appl Environ Microbiol 73:5261-7.

147

45.

46.

47.

Watanabe, K., M. Imase, H. Aoyagi, N. Ohmura, H. Saiki, and H. Tanaka.
2008. Development of a novel artificial medium based on utilization of algal
photosynthetic metabolites by symbiotic heterotrophs. J Appl Microbiol.

Wertz, J. T., and J. A. Breznak. 2007. Physiological ecology of Stenoxybacter

acetivorans, an obligate microaerophile in termite guts. Appl Environ Microbiol
73:6829-41.

Yergeau, E., S. Kang, Z. He, J. Zhou, and G. A. Kowalchuk. 2007. Functional
microarray analysis of nitrogen and carbon cycling genes across an Antarctic
latitudinal transect. ISME J 1:163-79.

148

CHAPTER FIVE

CONCLUSIONS

Comparative l6S rRNA gene analyses have allowed us to explore the diversity of the
microbial world present in the Field Research Center Area 3 where uranium remediation
is being applied. The results from both studies have suggested that bacteria related to
sulfate-reducing bacteria, Desulfovibrio and Desuljbsporosinus and iron reducing
bacteria are directly involved in uranium reduction in situ. Desulfovibrio seems to be the
most important group given its relative abundance in wells where uranium reduction

occurs, and also because the rapid conversion of ethanol to acetate observed in this site.

In addition to the uranium reducing groups, other bacteria seems to contribute
important functions to the community such as the ability to eliminate nitrate, a competing
electron (e.g. Thauera); create resistance structure such as biolfilms (e.g. Chloroflexi),
recycle nutrients (e. g. Chloroflexi, Acanthoamoeba) and provide uranium reduction
indirectly through reduced iron (F RB, e. g. F erribacterium), sulfur (SRB, e. g.
Desulfosporosinus) and nitrogen compound (Denitrifiers e. g Acidovorax ) as well as

reduced humic acids (Geothrix).

One surprising find was the abundance of iron-oxidizing populations such as
T hiobacillus and Gallionella. Thiobacillus can oxidize uranium while reducing nitrate,
Gallionella can oxidize iron under microaerophillic conditions, and iron oxides can
directly reoxidize uranium. The effect of iron oxidizers on uranium stability needs further

attention.

149

The ability of Acidobacteria iron reducers such as Geothrix, to reduce uranium
also needs further study. In our experiments, Geothrix-related sequences were associated
with the wells of high uranium reducing activity. These bacteria can reduced humic acids
and through them potentially reduced uranium, yet the ability to reduce uranium directly
is unknown. The use of Acidobacteria in uranium reduction would be a useful application
given their ability to grow under a wide range of pH, resist environmental stresses and

grow on oligotrophic conditions.

In terms of diversity, what started as a low diversity community of a heavily
contaminated site has changed to a diverse community that can perform the task we
stimulated it for, uranium reduction. Even when the response was heterogeneous i.e.
different uranium reducing population responded in different sites, the diverse response
was partially explained by hydrology parameters. Using the hydrology information
helped us to understand the gradients present in the site and better understand how the
communities respond to the stimulation. We have demonstrated that hydrology
information is no longer relevant only for engineering purposes but can also be used to

explain the natural variability in the microbial communities.

Method wise, massive parallel sequencing has shown to be powerful in detecting
minor members of the community that seemed to respond to stimulation such as
Anaeromyxobacter and Desulfosporosinus, it also potentially avoided cloning bias.
Additionally, massive sequencing provides the opportunity to perform multivariate
analyses that are powered by high sequence counts. It is expected that in the future this
technology will be widely adopted for the analysis of microbial communities, this works

uses for the first time a combination of massive sequencing and indicator species

150

approach to provide a consistent and statistically sound method to detect candidate
species for a given groups. We expect this approach to be more adopted to deal with the

high complexity of massive parallel sequencing databases.

Overall this work has answered some questions regarding the identity of the
populations involved in uranium reduction but has also raised new question regarding the
role of non-uranium reducers on the stability of reduced uranium, and their role in the
community. The massive aspect has potentially shown areas of the microbial diversity
where isolation efforts should be made in order to obtain microorganism relevant to the
cleanup of this site e. g. in the isolation of new Acidobacteria associated with uranium

reduction.

151

APPENDIX

My research has greatly benefited from a year of work in the Ribosomal Database
Project where I developed and produced video tutorials to increase the adoption of the
Functional Gene Database / Repository as part of the NIEHS Research translation core of
the Michigan State University Superfund group. Additionally I led in the development of
a new feature of the Ribosomal Database Project, Assignment Generator. This
application provides instructors the chance to create a realistic assignment on

comparative 16S rRNA gene analysis.

These two projecta, the tutorials and the Assignment Generator, have been
published as part of the Ribosomal Database Project’s papers in the journal Nucleic acid

research in 2007(1) and 2009 (2), respectively.

152

REFERENCES

Cole, J. R., B. Chai, R. J. Farris, Q. Wang, A. S. Kulam-Syed-Mohideen, D.
M. McGarrell, A. M. Bandela, E. Cardenas, G. M. Garrity, and J. M. Tiedje.
2007. The ribosomal database project (RDP-II): introducing myRDP space and
quality controlled public data. Nucleic Acids Res. 35:D169-72.

Cole, J. R., Q. Wang, E. Cardenas, J. Fish, B. Chai, R. J. Farris, A. S.
Kulam-Syed-Mohideen, D. M. McGarrell, T. Marsh, G. M. Garrity, and J.
M. Tiedje. 2009. The Ribosomal Database Project: improved alignments and new
tools for rRNA analysis. Nucleic Acids Res 37:D141-5.

153

     

     

      

      

IIIIIII STATE NI

‘6 VERSITY LIBRARIES
11111311111er Mlflljllflllflljlmblflll

1293 0