microbial nitrogen cycling dynamics in coastal systems a

MICROBIAL NITROGEN CYCLING DYNAMICS IN COASTAL SYSTEMS

A DISSERTATION SUBMITTED TO THE DEPARTMENT OF

ENVIRONMENTAL EARTH SYSTEM SCIENCE AND THE COMMITTEE ON

GRADUATE STUDIES OF STANFORD UNIVERSITY IN PARTIAL

FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF

PHILOSOPHY

Annika Carlene Mosier

May 2011

http://creativecommons.org/licenses/by-nc/3.0/us/

This dissertation is online at: http://purl.stanford.edu/wt558td2044

© 2011 by Annika Carlene Mosier. All Rights Reserved.

Re-distributed by Stanford University under license with the author.

This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 United States License.

ii



http://purl.stanford.edu/wt558td2044

I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.

Christopher Francis, Primary Adviser


Kevin Arrigo


Alexandria Boehm


Craig Criddle


Scott Fendorf

Approved for the Stanford University Committee on Graduate Studies.

Patricia J. Gumport, Vice Provost Graduate Education

This signature page was generated electronically upon submission of this dissertation in electronic format. An original signed hard copy of the signature page is on file inUniversity Archives.

iii

iv

Abstract

Human influence on the global nitrogen cycle (e.g., through fertilizer and

wastewater runoff) has caused a suite of environmental problems including

acidification, loss of biodiversity, increased concentrations of greenhouse gases, and

eutrophication. These environmental risks can be lessened by microbial

transformations of nitrogen; nitrification converts ammonia to nitrite and nitrate,

which can then be lost to the atmosphere as N2 gas via denitrification or anammox.

Microbial processes thus determine the fate of excess nitrogen and yet recent

discoveries suggest that our understanding of these organisms is deficient. This

dissertation focuses on microbial transformations of nitrogen in marine and estuarine

systems through laboratory and field studies, using cutting-edge techniques from

molecular biology, genomics, microbial ecology, and microbiology.

Recent studies revealed that many archaea can oxidize ammonia (AOA;

ammonia-oxidizing archaea), in addition to the well-described ammonia-oxidizing

bacteria (AOB). Considering that these archaea are among the most abundant

organisms on Earth, these findings have necessitated a reevaluation of nitrification to

determine the relative contribution of AOA and AOB to overall rates and to determine

if previous models of global nitrogen cycling require adjustment to include the AOA.

I examined the distribution, diversity, and abundance of AOA and AOB in the San

Francisco Bay estuary and found that the region of the estuary with low-salinity and

high C:N ratios contained a group of AOA that were both abundant and

phylogenetically distinct. In most of the estuary where salinity was high and C:N

ratios were low, AOB were more abundant than AOA—despite the fact that AOA

v

outnumber AOB in soils and the ocean, the two end members of an estuary. This

study suggested that a combination of environmental factors including carbon,

nitrogen, and salinity determine the niche distribution of the two groups of ammonia-

oxidizers.

In order to gain insight into the genetic basis for ammonia oxidation by

estuarine AOA, we sought to sequence the genome of AOA grown in an enrichment

culture (84% pure) from San Francisco Bay; however, standard genome sequencing

methods require large amounts of DNA from pure cultures. These genomic methods

are severely limited because they can only be applied to the ~1% of microbes in nature

that can be isolated in the lab. We overcame this deficiency using novel technology

developed by Dr. Stephen Quake (Stanford University) that relies on microfluidics and

laser tweezing to sequence the genome of a single microbial cell, thereby creating the

potential for genomic sequencing of the ~99% of microbes found in nature that cannot

be isolated. We demonstrated the feasibility of the method by sequencing the genome

of single AOA cells, which proved to represent a new genus within the Archaea. The

genome data revealed that the AOA have genes for both autotrophic and heterotrophic

carbon metabolism, unlike the autotrophic AOB. These AOA may be chemotactic and

motile based on numerous chemotaxis and motility-associated genes in the genome

and electron microscopy evidence of flagella. Physiological studies showed that the

AOA grow aerobically but they also oxidize ammonia at low oxygen concentrations

and may produce the potent greenhouse gas N2O. Continued cultivation and genomic

sequencing of AOA will allow for in-depth studies on the physiological and metabolic

vi

potential of this novel group of organisms that will ultimately advance our

understanding of the global carbon and nitrogen cycles.

Denitrifying bacteria are widespread in coastal and estuarine environments and

account for a significant reduction of external nitrogen inputs, thereby diminishing the

amount of bioavailable nitrogen and curtailing the harmful effects of nitrogen

pollution. I determined the abundance, community structure, biogeochemical activity,

and ecology of denitrifiers over space and time in the San Francisco Bay estuary.

Salinity, carbon, nitrogen and some metals were important factors for denitrification

rates, abundance, and community structure. Overall, this study provided valuable new

insights into the microbial ecology of estuarine denitrifying communities and

suggested that denitrifiers likely play an important role in nitrogen removal in San

Francisco Bay, particularly at high salinity sites.

vii

Acknowledgments

My advisor, Chris Francis, has been the single most influential figure in the

development and execution of my dissertation. He has provided continual guidance

and encouragement, as well as the freedom to pursue my own research interests and

financial support to participate in national and international conferences—an

invaluable learning opportunity.

I am extremely grateful to my dissertation committee members: Scott Fendorf,

Alexandria Boehm, Kevin Arrigo, and Craig Criddle. I thank each of them for their

time, constructive feedback on my research, and interest in my intellectual

development and future success in academia. I also extend my appreciation to my

mentors Mary Beth Mudgett and Karla Kirkegaard who have both provided valuable

conversations about research and academic life.

This research would not have been possible without Paul Salop, Sarah Lowe,

Applied Marine Sciences, the San Francisco Estuary Institute, and the crew of the

R.V. Endeavor. I thank them for allowing us to participate in the San Francisco Bay

RMP sediment cruises, providing access to the RMP data, and assisting in sample

collection.

I thank all of my past and present labmates for their support, friendship, and

assistance in the lab and field. In particular, I look to Alyson, George, Jason, and

Kevan as colleagues and I am enriched from the countless scientific discussions we

have shared over the years.

Lastly, I thank all of my family. I thank my mom who taught me the value of

education at a young age and has provided unwavering support of my academic

viii

pursuits ever since. Above all, to Jeremy, I could never thank you enough for all the

ways that you made this process better, easier, happier. From your refreshing outside

perspective to your exceptional editorial skills, you have been there every step of the

way and I am so grateful to have you in my life.

The work presented in this dissertation was funded by the Diversifying

Academia, Recruiting Excellence Doctoral Fellowship (Stanford University), the

Environmental Protection Agency STAR Graduate Fellowship, and National Science

Foundation grants to Chris Francis.

ix

Table of Contents Abstract ........................................................................................................................ iv Acknowledgments.......................................................................................................vii List of Tables................................................................................................................xi List of Figures .............................................................................................................xii Chapter 1. Introduction ..............................................................................................1

Importance of Nitrogen Cycling in Estuaries .............................................................2 Microbial Biogeochemistry of Ammonia Oxidation..................................................4 Microbial Biogeochemistry of Denitrification ...........................................................7 Dissertation Organization and Chapter Descriptions .................................................8 Figures ......................................................................................................................10 Literature Cited.........................................................................................................11

Chapter 2. Determining the distribution of marine and coastal ammonia-oxidizing archaea and bacteria using a quantitative approach..............................23

Abstract.....................................................................................................................24 Introduction ..............................................................................................................25 Methods for Measuring the Abundance of AOA and AOB within Marine and Coastal Systems........................................................................................................29 Methodological Considerations................................................................................36 Targeting Specific Ecotypes: Quantifying Shallow and Deep Clades of Marine Water Column AOA.................................................................................................38 Acknowledgments ....................................................................................................41 Tables .......................................................................................................................42 Figures ......................................................................................................................43 Literature Cited.........................................................................................................44

Chapter 3. Relative abundance and diversity of ammonia-oxidizing archaea and bacteria in the San Francisco Bay estuary ...............................................................56

Abstract.....................................................................................................................57 Introduction ..............................................................................................................58 Results and Discussion .............................................................................................60 Experimental Procedures..........................................................................................73 Acknowledgements ..................................................................................................80 Tables .......................................................................................................................81 Figures ......................................................................................................................83 Supporting Information ............................................................................................88 Literature Cited.........................................................................................................89

Chapter 4. Genome of a Low-Salinity Ammonia-Oxidizing Archaeon Determined by Single-Cell and Metagenomic Analysis ........................................104

Abstract...................................................................................................................105 Introduction ............................................................................................................107

x

Results and Discussion ...........................................................................................108 Materials and Methods ...........................................................................................125 Acknowledgments ..................................................................................................131 Tables .....................................................................................................................132 Figures ....................................................................................................................135 Supporting Information ..........................................................................................140 Literature Cited.......................................................................................................142

Chapter 5. Physiology of a novel low-salinity type AOA reveals niche adaptation....................................................................................................................................154

Abstract...................................................................................................................155 Introduction ............................................................................................................156 Results and Discussion ...........................................................................................157 Materials and Methods ...........................................................................................167 Acknowledgments ..................................................................................................169 Figures ....................................................................................................................170 Supporting Information ..........................................................................................175 Literature Cited.......................................................................................................176

Chapter 6. Abundance and Potential Rates of Denitrifiers Across the San Francisco Bay Estuary .............................................................................................184

Abstract...................................................................................................................185 Introduction ............................................................................................................186 Results and Discussion ...........................................................................................188 Experimental Procedures........................................................................................198 Acknowledgments ..................................................................................................201 Tables .....................................................................................................................203 Figures ....................................................................................................................204 Supporting Information ..........................................................................................208 Literature Cited.......................................................................................................210

xi

List of Tables Chapter 2 Table 1. Primers and probes for archaeal amoA asssays specific to water column group A and group B marine sequences……………………………………………...42 Chapter 3 Table 1. Physical and chemical properties of San Francisco Bay bottom water and sediments……………………………………………………………………………...81 Table 2. Abundance-based Sørensen-type (Labd) similarities among archaeal and bacterial amoA clone libraries………………………………………………………...82 Chapter 4 Table 1. N. limnia assembly statistics for the consensus genome (metagenome and five single cells), single cell coassembly (five single cells), and three individual cell assemblies…………………………………………………………………………...132 Table 2. Genome features for N. limnia, N. maritimus, and C. symbiosum A…...…132 Table 3. Whole genome nucleotide identity comparisons between N. limnia, N. maritimus, and C. symbiosum A……………………………………………………..133 Table 4. N. limnia genes putatively associated with motility………………………133 Table 5. Nucleotide (NT) and amino acid (AA) identities for key individual genes in N. limnia, N. maritimus, and C. symbiosum A………………………………………134 Chapter 6 Table 1. Abundance-based Sørensen-type (Labd) similarities among nirK (below diagonal) and nirS (above diagonal) clone libraries………………………………...203

xii

List of Figures Chapter 1 Figure 1. Microbial nitrogen cycling in the marine environment (figure from Francis et al., 2007)……………………………………………………………………………10 Chapter 2 Figure 1. Abundance of the shallow and deep archaeal amoA clades within seawater collected at a depth of 891 meters in Monterey Bay, CA……...……………………..43 Chapter 3 Figure 1. Location of sampling sites within San Francisco Bay…...…………….…..83 Figure 2. Log ratio of AOA:β-AOB amoA copy numbers in San Francisco Bay sediments as determined by quantitative PCR………………………………...……...84 Figure 3. Neighbor-joining phylogenetic tree showing the affiliation of archaeal amoA sequences (579 base-pair fragment) from San Francisco Bay and GenBank sequences from other environments. A more detailed phylogeny of the low-salinity group is shown to the left………………………………………………………………………85 Figure 4. Neighbor-joining phylogenetic tree showing the affiliation of bacterial amoA sequences (444 base-pair fragment) from San Francisco Bay and GenBank sequences from other environments………………………...……………………………………86 Figure 5. Canonical correspondence analysis (CCA) of archaeal (left panel) and bacterial (right panel) amoA clone libraries and environmental variables from bottom water and sediments………..…………………………………………………………87 Chapter 4 Figure 1. Phylogeny of ammonia-oxidizing archaea gene sequences. Neighbor-joining phylogenetic tree of (A) archaeal amoA gene sequences and (B) archaeal 16S rRNA gene sequences……………………………………………………………….135 Figure 2. Assembly of the N. limnia datasets. Rarefaction analysis of the N. limnia sequence data showing assembly of the enrichment, single cell, and combined datasets……...……………………………………………………………………….136 Figure 3. Circular representation of the N. limnia genome. From the outer ring to the inner ring: Scaffold breakpoints are indicated by the inside tick marks, predicted genes on the forward strand, predicted genes the reverse strand, G+C content, GC skew, and GGGT skew………………...……………………………………………………….137 Figure 4. Comparison of three AOA genomes. (A) The top panel shows nucleotide identity of blast hits between the reference genomes and the N. limnia genome. Hits

xiii

for N. maritimus and C. symbiosum A are colored according to the position in the reference (query) genome. The arrow direction indicates the hit direction. On the right, box plots summarize the distribution of hits from each reference genome to the N. limnia genome. (B) The positions of the two largest consensus scaffolds are shown. (C) The sequence novelty index, defined as 2 minus the sum of nucleotide identity of blast hits to N. maritimus and C. symbiosum A at each position is plotted. A histogram of the values is present at right. (D) The alignment depth in the final assembly is shown. A histogram of the values is present at right…………………………………….……………………………………….138-139 Chapter 5 Figure 1. BlastP analysis of N. limnia proteins against all metagenomic ORF peptides (in CAMERA database). Sequence hits are color coded by habitats from which the samples were collected……………………………………………………………....170 Figure 2. Ammonia oxidation by N. limnia under different growth conditions….…171

Figure 3. Gas production from N. limnia enrichment cultures grown under different conditions…...……………………...……………………………………………..…172 Figure 4. Motility gene cassette on the N. limnia genome…………..……………..173 Figure 5. Transmission electron micrograph of a typical flagellated cell from the N. limnia enrichment……………………………………………………………….…..174 Chapter 6 Figure 1. Location of sampling sites within San Francisco Bay……………………204 Figure 2. Dissolved inorganic nitrogen concentrations in San Francisco Bay bottom water from 2004 to 2008……………...……………………………………………..205 Figure 3. Quantitative PCR measurements for nirK (left panel) and nirS (right panel) gene copy numbers expressed per gram of sediment (wet weight)………...……….206 Figure 4. Denitrification potential rate measurements from San Francisco Bay sediments in 2007 (left panel) and 2008 (right panel)..……………………………..207

1

Chapter 1. Introduction Annika C. Mosier1

1Department of Environmental Earth System Science, Stanford University, Stanford, CA 94305-4216, USA

2

Importance of Nitrogen Cycling in Estuaries

Estuaries and the vital ecological processes at work within them serve as a

filter of the world’s watersheds, attenuating the impact of pollutants on their local

environments. Center stage as one of the preeminent issues threatening estuarine

systems is nutrient pollution. One of the offending nutrients, dissolved inorganic

nitrogen (DIN; ammonium, nitrate, and nitrite), has been associated with numerous

toxicological and environmental problems. Excess DIN can accelerate eutrophication,

which often leads to hypoxia, the reduction of biodiversity, and new species invasions

(e.g., Ryther and Dunstan, 1971; Fisher et al., 1992; Rabalais and Nixon, 2002). Over

half of the estuaries in the United States experience the effects of eutrophication at

some time each year. Despite the enormous environmental impact of DIN pollution,

relatively little is known about the microbial communities that play critical roles in

transferring nitrogen within estuarine ecosystems.

My research examines microbial nitrogen cycling in San Francisco Bay, one of

the most anthropogenically-altered estuaries in the United States. Since 1970 the

population encircling San Francisco Bay has grown nearly 35%. This expansion and

the expectation of unabated growth in the near-term are exacerbating an already

present dichotomy: dependence on the estuary is increasing as agricultural and urban

stress on it increases. Now more than ever, the delicate health of the estuary hinges on

the effectiveness of regulation to mitigate the impacts of nutrients and other pollutants.

As regulation and education assume increasing importance in sustaining the health of

the estuary, so too does the scientific data that serves as its foundation.

3

San Francisco Bay has historically been considered a high-nutrient, low-

chlorophyll system (Cloern, 2001) where standing stocks of phytoplankton are low

despite high concentrations of inorganic nutrients, including nitrogen. Comparable

nutrient concentrations in Chesapeake Bay cause large phytoplankton blooms and

recurrent hypoxia (e.g., Boesch et al., 2001). Resistance to phytoplankton blooms in

San Francisco Bay has been a result of strong light limitation of photosynthesis by

phytoplankton and rapid consumption of phytoplankton by bivalve mollusks (Cloern,

2001). However, recent studies suggest that this resistance might be changing;

phytoplankton blooms increased in frequency and magnitude beginning after the late

1990s (new seasonal blooms coupled with an increased baseline chlorophyll a; Cloern

et al., 2007). The increase in phytoplankton blooms is a consequence of a decline in

bivalve populations and reduced phytoplankton grazing that coincided with physical

changes in the California Current System (Cloern et al., 2007). Increasing concerns

about the rising frequency and magnitude of phytoplankton blooms, as well as the

decline of pelagic organisms (e.g., delta and longfin smelts, striped bass and threadfin

shad; Sommer et al., 2007) and major shifts in community composition of algae

(Lehman, 2000; Lehman et al., 2005), highlight the importance of microbial nitrogen

removal from this heavily impacted, urbanized estuarine system.

My research in San Francisco Bay is therefore timely and focuses on three

aspects aimed at gaining a better understanding of nitrogen cycling in the estuary: a)

quantifying and describing the diversity of nitrogen cycling microorganisms in the

estuary, b) examining the effect of environmental factors such as salinity, nitrate,

oxygen and temperature on their abundance and rates, and c) cultivating ecologically-

4

relevant organisms to elucidate their physiological, biochemical, and genomic

potential.

Microbial Biogeochemistry of Ammonia Oxidation

Harmful levels of DIN can be diminished through coupled nitrification and

denitrification. Nitrification (the conversion of NH3 to NO3-) links production of NH4

+

to nitrogen removal via denitrification (the conversion of NO3- to N2 gas, which is

rapidly lost to the atmosphere) (Figure 1). It is estimated that over 50% of external

DIN inputs to estuaries are removed by coupled nitrification-denitrification

(Seitzinger, 1990). By removing a large percentage of anthropogenic N pollution

from estuaries, coupled nitrification-denitrification effectively creates a sink for DIN

and thereby plays an important role in lessening the risk of eutrophication.

Nitrification and denitrification thus are particularly significant processes where the

complex interplay of microbial populations determines the fate of excess nutrients in

estuaries.

The oxidation of ammonia (the first step in nitrification) is often the rate-

limiting process in the removal of nitrogen from marine systems (Herbert, 1999). For

over a century (Winogradsky, 1890), scientists have assumed that the most important

contributors to ammonia oxidation were two groups of ammonia-oxidizing bacteria

(AOB): the beta-proteobacteria (βAOB) from the genus Nitrosomonas and

Nitrosospira (Head et al., 1993; Purkhold et al., 2000; Purkhold et al., 2003) and

gamma-proteobacteria (γAOB) from the genus Nitrosococcus (Ward and O'Mullan,

2002). However, recently published research linking nitrification to the Archaeal

5

domain of life has turned this assumption upside-down. A combination of

metagenomic analyses (Venter et al., 2004; Treusch et al., 2005) and the cultivation of

a novel, ammonia-oxidizing, marine crenarchaeote (Könneke et al., 2005) revealed the

first evidence for nitrification within the Archaeal domain and definitively linked the

novel ammonia monooxygenase (amoA) genes to this chemoautotrophic metabolism.

The widespread presence of archaeal amoA genes in marine water columns and

sediments (Francis et al., 2005; Wuchter et al., 2006; Mincer et al., 2007) and the

demonstration that ~83% of the archaeal community in deep ocean waters is

autotrophic (Ingalls et al., 2006) indicates that the ability to oxidize ammonia may be

broadly distributed within the Crenarchaeota and may be biogeochemically important

in marine systems. Studies have shown that ammonia-oxidizing archaea (AOA) may

be up to several 1000-fold more abundant than their well-known bacterial counterparts

in soils, the open ocean, and the Black Sea suboxic zone (e.g., Leininger et al., 2006,

Wuchter et al., 2006, Lam et al., 2007).

Ammonia-oxidizing populations exhibit distinct spatial structure in estuaries.

Studies in the Chesapeake Bay, Plum Island Sound, and Ythan estuaries showed shifts

in AOB diversity along salinity gradients (Francis et al., 2003; Bernhard et al., 2005;

Freitag et al., 2006). All three studies reported Nitrosospira-like sequences at high

salinities and Nitrosomonas-like sequences at low salinities. AOA also showed strong

spatial variability in the Bahía del Tóbari estuary, with distinct AOA communities

within the interior of the bay and at the mouths of the estuary (Beman and Francis,

2006). Nitrification potentials also vary across estuarine salinity gradients with

greatest values observed at low and intermediate salinities (Rysgaard et al., 1999;

6

Bernhard et al., 2007). Although salinity appears to play a role in shaping the

ammonia-oxidizing communities in these environments, many physical and chemical

parameters fluctuate (and covary) across a typical estuarine transect making it difficult

to separate the controlling variables.

The vast majority of amoA gene sequences reported in most estuarine studies

appear to be novel and distinct from sequences of cultivated ammonia oxidizers,

indicating that we know very little about the organisms actually responsible for this

process in nature. Despite the fact that the first βAOB were isolated in the late 1800’s

(Winogradsky, 1890), to date there are large groups of βAOB with no cultivated

representatives, including Nitrosospira cluster 1 and Nitrosomonas cluster 5 (Freitag

and Prosser, 2003; Freitag et al., 2006). These 16S rRNA-based groups dominate

AOB communities in many marine and estuarine systems (McCaig et al., 1999; Bano

and Hollibaugh, 2000; Freitag and Prosser, 2003, 2004) and yet very little is known

about them. The presumption that these groups oxidize ammonia is based on

clustering of their 16S rRNA genes within a monophyletic group containing cultivated

AOB, and on reports of non-persisting enrichment cultures of Nitrosomonas cluster 5

(Stephen et al., 1996). Likewise, only one archaeal ammonia oxidizer has been

isolated—Nitrosopumilus maritimus derived from a tropical marine aquarium

(Könneke et al., 2005). There have been a few other reports of cultivation of AOA in

enrichment cultures (Wuchter et al., 2006; de la Torre et al., 2008; Hatzenpichler et

al., 2008; Mosier and Francis, 2008; Santoro et al., 2008). Cultivation of

environmentally relevant AOB and AOA will be important for understanding the

physiology of these organisms.

7

Microbial Biogeochemistry of Denitrification

Denitrification mediates nitrogen load reduction enabling estuaries to act as

natural nitrogen sinks, curtailing nitrate pollution. Denitrification is the main process

responsible for the conversion of nitrate to dinitrogen gases, which are rapidly lost to

the atmosphere (NO3- → NO2

- → NO → N2O → N2) (Zumft, 1997). The conversion

of nitrate to a less-bioavailable form (dinitrogen gases) can limit excessive growth of

algae that ultimately stimulates hypoxic conditions. The conversion to gaseous

nitrogen forms within the estuary also mitigates the flux of nitrate into the ocean,

preventing secondary pollution of that system.

Recent reports of anammox activity (the anaerobic oxidation of ammonia to N2

via nitrite) have drawn into question the relative significance of denitrification versus

anammox in estuarine sediments. The contribution of N2 formation from anammox

relative to denitrification in estuarine sediments has been observed to range from 1 to

8% in the Thames Estuary (Trimmer et al., 2003), 5 to 24% in Randers Fjord and 0%

in Norsminde Fjord, Denmark (Risgaard-Petersen et al., 2004). Based on these

studies, denitrification appears to represent the dominant nitrogen removal process in

estuarine sediments.

Because denitrifying bacteria are found within a variety of phylogenetically

unrelated groups from over 50 genera (Zumft, 1997), efforts have been directed

toward amplification of functional genes involved in denitrification (e.g., narG, nirK,

nirS, norB, nosZ), rather than the common 16S rRNA-based approach. These genes

have been used to examine denitrifying communities in a variety of environments

8

including: soils (Prieme et al., 2002; Kramer et al., 2006); groundwater (Yan et al.,

2003); coastal aquifer (Santoro et al., 2006); Baltic Sea (Brettar et al., 2006; Hannig et

al., 2006); Arabian Sea (Jayakumar et al., 2004); Black Sea suboxic zone (Oakley et

al., 2007); the deep sea (Tamegai et al., 2007); marine sediments (Braker et al., 2000;

Liu et al., 2003; Hunter et al., 2006; Tiquia et al., 2006); and estuarine sediments

(Nogales et al., 2002).

Nitrite reductase (Nir) is considered the key reductase in the denitrification

pathway because it catalyzes the first committed step to a gaseous product (Zumft,

1997). Two structurally different but functionally equivalent enzymes catalyze nitrite

reduction: NirS, containing iron (cytochrome-cd1); and NirK, containing copper.

Diversity of nirS and nirK genes were recently examined in two estuaries. Relatively

high diversity of expressed nirS genes was observed in the River Colne estuary and

most of the sequences exhibited site specificity (Nogales et al., 2002). Transcripts

were not detected for three of the five denitrification genes found in the sediments

(narG, napA, and nirK), despite the fact that the sediments had high denitrification

rates. The Chesapeake Bay had extensive and unprecedented nirS diversity and

distinct spatial structure (Francis et al., unpublished). Interestingly, nirS richness,

nitrate concentrations, and sediment denitrification rates (N2-flux) all ranged from

extremely high at the low salinity stations to quite low at the most marine station near

the mouth of the Bay.

Dissertation Organization and Chapter Descriptions

This dissertation focuses on microbial transformations of nitrogen in marine

9

systems through laboratory and field studies. Chapters 2, 3, 4, and 6 have been

published in peer-reviewed journals as indicated on the chapter title pages. Chapter 2

describes quantitative approaches used to determine the distribution of marine and

coastal ammonia-oxidizing archaea and bacteria. This chapter also details some of the

current debates about ammonia oxidation in marine systems. Chapter 3 uses

molecular biology techniques to determine the relative abundance and diversity of

ammonia-oxidizing archaea and bacteria in the San Francisco Bay estuary. The

ecology of these organisms is also explored using statistical analyses of environmental

data. In Chapter 4, we analyze the genome of a low-salinity AOA isolated from the

north part of San Francisco Bay. This study represents only the second genome from

a free-living AOA and the third overall AOA genome (the third being a metagenome

from a sponge symbiont). I share the first author position for this article with Paul

Blainey; we both contributed equally to the conception, design, and execution of the

experiments and to the data analysis and interpretation. The physiology of this low-

salinity AOA culture is explored in Chapter 5. Finally, Chapter 6 describes the

abundance and activity of denitrifiers in the San Francisco Bay estuary.

10

Figures

Figure 1. Microbial nitrogen cycling in the marine environment (figure from Francis

et al., 2007).

11

Literature Cited

Bano, N., and Hollibaugh, J.T. (2000) Diversity and distribution of DNA sequences

with affinity to ammonia-oxidizing bacteria of the beta subdivision of the class

Proteobacteria in the Arctic Ocean. Applied and Environmental Microbiology 66:

1960-1969.

Beman, J.M., and Francis, C.A. (2006) Diversity of ammonia-oxidizing archaea and

bacteria in the sediments of a hypernutrified subtropical estuary: Bahia del Tobari,

Mexico. Applied and Environmental Microbiology 72: 7767-7777.

Bernhard, A.E., Donn, T., Giblin, A.E., and Stahl, D.A. (2005) Loss of diversity of

ammonia-oxidizing bacteria correlates with increasing salinity in an estuary system.

Environmental Microbiology 7: 1289-1297.

Bernhard, A.E., Tucker, J., Giblin, A.E., and Stahl, D.A. (2007) Functionally distinct

communities of ammonia-oxidizing bacteria along an estuarine salinity gradient.


Boesch, D.F., Brinsfield, R.B., and Magnien, R.E. (2001) Chesapeake Bay

eutrophication: Scientific understanding, ecosystem restoration, and challenges for

agriculture. Journal of Environmental Quality 30: 303-320.

12

Braker, G., Zhou, J.Z., Wu, L.Y., Devol, A.H., and Tiedje, J.M. (2000) Nitrite

reductase genes (nirK and nirS) as functional markers to investigate diversity of

denitrifying bacteria in Pacific northwest marine sediment communities. Applied And


Brettar, I., Labrenz, M., Flavier, S., Botel, J., Kuosa, H., Christen, R., and Hofle, M.G.

(2006) Identification of a Thiomicrospira denitrificans-like epsilonproteobacterium as

a catalyst for autotrophic denitrification in the central Baltic Sea. Applied And


Cloern, J.E. (2001) Our evolving conceptual model of the coastal eutrophication

problem. Marine Ecology-Progress Series 210: 223-253.

Cloern, J.E., Jassby, A.D., Thompson, J.K., and Hieb, K.A. (2007) A cold phase of the

East Pacific triggers new phytoplankton blooms in San Francisco Bay. Proceedings of

the National Academy of Sciences 104: 18561-18565.

de la Torre, J.R., Walker, C.B., Ingalls, A.E., Konneke, M., and Stahl, D.A. (2008)

Cultivation of a thermophilic ammonia oxidizing archaeon synthesizing crenarchaeol.


Fisher, T., Peele, E., Ammerman, J., and Harding, L. (1992) Nutrient limitation of

phytoplankton in Chesapeake Bay. Marine Ecology Progress Series 82: 51-63.

13

Francis, C.A., Beman, J.M., and Kuypers, M.M.M. (2007) New processes and players

in the nitrogen cycle: the microbial ecology of anaerobic and archaeal ammonia

oxidation. ISME J 1: 19-27.

Francis, C.A., O'Mullan, G.D., and Ward, B.B. (2003) Diversity of ammonia

monooxygenase (amoA) genes across environmental gradients in Chesapeake Bay

sediments. Geobiology 1: 129-140.

Francis, C.A., Roberts, K.J., Beman, J.M., Santoro, A.E., and Oakley, B.B. (2005)

Ubiquity and diversity of ammonia-oxidizing archaea in water columns and sediments

of the ocean. Proceedings of the National Academy of Sciences 102: 14683-14688.




Freitag, T.E., and Prosser, J.I. (2003) Community structure of ammonia-oxidizing

bacteria within anoxic marine sediments. Applied and Environmental Microbiology

69: 1359-1371.

14

Freitag, T.E., and Prosser, J.I. (2004) Differences between betaproteobacterial

ammonia-oxidizing communities in marine sediments and those in overlying water.

Applied and Environmental Microbiology 70: 3789-3793.

Freitag, T.E., Chang, L., and Prosser, J.I. (2006) Changes in the community structure

and activity of betaproteobacterial ammonia-oxidizing sediment bacteria along a

freshwater-marine gradient. Environmental Microbiology 8: 684-696.

Hannig, M., Braker, G., Dippner, J., and Jurgens, K. (2006) Linking denitrifier

community structure and prevalent biogeochemical parameters in the pelagial of the

central Baltic Proper (Baltic Sea). FEMS Microbiology Ecology 57: 260-271.

Hatzenpichler, R., Lebedeva, E.V., Spieck, E., Stoecker, K., Richter, A., Daims, H.,

and Wagner, M. (2008) A moderately thermophilic ammonia-oxidizing crenarchaeote

from a hot spring. Proceedings of the National Academy of Sciences 105: 2134-2139.

Head, I.M., Hiorns, W.D., Embley, T.M., McCarthy, A.J., and Saunders, J.R. (1993)

The phylogeny of autotrophic ammonia-oxidizing bacteria as determined by analysis

of 16S ribosomal RNA gene sequences. Journal of General Microbiology 139: 1147-

1153.

Herbert, R.A. (1999) Nitrogen cycling in coastal marine ecosystems. FEMS

Microbiology Reviews 23: 563-590.

15

Hunter, E.M., Mills, H.J., and Kostka, J.E. (2006) Microbial community diversity

associated with carbon and nitrogen cycling in permeable shelf sediments. Applied

and Environmental Microbiology 72: 5689-5701.

Ingalls, A.E., Shah, S.R., Hansman, R.L., Aluwihare, L.I., Santos, G.M., Druffel,

E.R.M., and Pearson, A. (2006) Quantifying archaeal community autotrophy in the

mesopelagic ocean using natural radiocarbon. Proceedings of the National Academy of

Sciences 103: 6442-6447.

Jayakumar, D.A., Francis, C.A., Naqvi, S.W.A., and Ward, B.B. (2004) Diversity of

nitrite reductase genes (nirS) in the denitrifying water column of the coastal Arabian

Sea. Aquatic Microbial Ecology 34: 69-78.

Könneke, M., Bernhard, A.E., de la Torre, J.R., Walker, C.B., Waterbury, J.B., and

Stahl, D.A. (2005) Isolation of an autotrophic ammonia-oxidizing marine archaeon.

Nature 437: 543-546.

Kramer, S.B., Reganold, J.P., Glover, J.D., Bohannan, B.J.M., and Mooney, H.A.

(2006) Reduced nitrate leaching and enhanced denitrifier activity and efficiency in

organically fertilized soils. Proceedings of the National Academy of Sciences 103:

4522-4527.

16

Lam, P., Jensen, M., Lavik, G., McGinnis, D., Muller, B., Schubert, C. et al. (2007)

Linking crenarchaeal and bacterial nitrification to anammox in the Black Sea.

Proceedings of the National Academy of Sciences 104: 7104-7109.

Lehman, P.W. (2000) Phytoplankon biomass, cell diameter, and species composition

in the low salinity zone of northern San Francisco Bay estuary. Estuaries 23: 216-230.

Lehman, P.W., Boyer, G., Hall, C., Waller, S., and Gehrts, K. (2005) Distribution and

toxicity of a new colonial Microcystis aeruginosa bloom in the San Francisco Bay

Estuary, California. Hydrobiologia 541: 87-99.

Leininger, S., Urich, T., Schloter, M., Schwark, L., Qi, J., Nicol, G. et al. (2006)

Archaea predominate among ammonia-oxidizing prokaryotes in soils. Nature 442:

806-809.

Liu, X.D., Tiquia, S.M., Holguin, G., Wu, L.Y., Nold, S.C., Devol, A.H. et al. (2003)

Molecular diversity of denitrifying genes in continental margin sediments within the

oxygen-deficient zone off the Pacific coast of Mexico. Applied And Environmental

Microbiology 69: 3549-3560.

McCaig, A.E., Phillips, C.J., Stephen, J.R., Kowalchuk, G.A., Harvey, S.M., Herbert,

R.A. et al. (1999) Nitrogen cycling and community structure of proteobacterial beta-

17

subgroup ammonia-oxidizing bacteria within polluted marine fish farm sediments.


Mincer, T.J., Church, M.J., Taylor, L.T., Preston, C., Karl, D.M., and DeLong, E.F.

(2007) Quantitative distribution of presumptive archaeal and bacterial nitrifiers in

Monterey Bay and the North Pacific Subtropical Gyre. Environmental Microbiology 9:

1162-1175.

Mosier, A.C., and Francis, C.A. (2008) Relative abundance and diversity of ammonia-

oxidizing archaea and bacteria in the San Francisco Bay estuary. Environmental


Nogales, B., Timmis, K.N., Nedwell, D.B., and Osborn, A.M. (2002) Detection and

diversity of expressed denitrification genes in estuarine sediments after reverse

transcription-PCR amplification from mRNA. Applied And Environmental


O'Mullan, G.D., and Ward, B.B. (2005) Relationship of temporal and spatial

variabilities of ammonia-oxidizing bacteria to nitrification rates in Monterey Bay,

California. Applied and Environmental Microbiology 71: 697-705.

Oakley, B.B., Francis, C.A., Roberts, K.J., Fuchsman, C.A., Srinivasan, S., and Staley,

J.T. (2007) Analysis of nitrite reductase (nirK and nirS) genes and cultivation reveal

18

depauperate community of denitrifying bacteria in the Black Sea suboxic zone.


Prieme, A., Braker, G., and Tiedje, J.M. (2002) Diversity of nitrite reductase (nirK and

nirS) gene fragments in forested upland and wetland soils. Applied And Environmental


Purkhold, U., Wagner, M., Timmermann, G., Pommerening-Roser, A., and Koops,

H.P. (2003) 16S rRNA and amoA-based phylogeny of 12 novel betaproteobacterial

ammonia-oxidizing isolates: extension of the dataset and proposal of a new lineage

within the nitrosomonads. International Journal of Systematic and Evolutionary


Purkhold, U., Pommerening-Roser, A., Juretschko, S., Schmid, M.C., Koops, H.P.,

and Wagner, M. (2000) Phylogeny of all recognized species of ammonia oxidizers

based on comparative 16S rRNA and amoA sequence analysis: implications for

molecular diversity surveys. Applied and Environmental Microbiology 66: 5368-5382.

Rabalais, N., and Nixon, S. (2002) Preface: nutrient overenrichment of the coastal

zone. Estuaries 25: 639.

19

Risgaard-Petersen, N., Meyer, R.L., Schmid, M., Jetten, M.S.M., Enrich-Prast, A.,

Rysgaard, S., and Revsbech, N.P. (2004) Anaerobic ammonium oxidation in an

estuarine sediment. Aquatic Microbial Ecology 36: 293-304.

Rysgaard, S., Thastum, P., Dalsgaard, T., Christensen, P.B., and Sloth, N.P. (1999)

Effects of salinity on NH4+ adsorption capacity, nitrification, and denitrification in

Danish estuarine sediments. Estuaries 22: 21-30.

Ryther, J.H., and Dunstan, W.M. (1971) Nitrogen, phosphorus, and eutrophication in

the coastal marine environment. Science 171: 1008-1013.

Santoro, A.E., Boehm, A.B., and Francis, C.A. (2006) Denitrifier community

composition along a nitrate and salinity gradient in a coastal aquifer. Applied And


Santoro, A.E., Francis, C.A., de Sieyes, N.R., and Boehm, A.B. (2008) Shifts in the

relative abundance of ammonia-oxidizing bacteria and archaea across

physicochemical gradients in a subterranean estuary. Environmental Microbiology 10:

1068-1079.

Seitzinger, S.P. (1990) Denitrification in aquatic sediments. In Denitrification in soil

and sediment. Revsbech, N.P., and Sorensen, J. (eds). New York, N.Y.: Plenum Press,

pp. 301-322.

20

Sommer, T., Armor, C., Baxter, R., Breuer, R., Brown, L., Chotkowski, M. et al.

(2007) The collapse of pelagic fishes in the upper San Francisco Estuary. Fisheries 32:

270-277.

Stephen, J.R., McCaig, A.E., Smith, Z., Prosser, J.I., and Embley, T.M. (1996)

Molecular diversity of soil and marine 16S rRNA gene sequences related to beta-

subgroup ammonia-oxidizing bacteria. Applied and Environmental Microbiology 62:

4147-4154.

Tamegai, H., Aoki, R., Arakawa, S., and Kato, C. (2007) Molecular analysis of the

nitrogen cycle in deep-sea microorganisms from the Nankai Trough: genes for

nitrification and denitrification from deep-sea environmental DNA. Extremophiles 11:

269-275.

Tiquia, S.M., Masson, S.A., and Devol, A. (2006) Vertical distribution of nitrite

reductase genes (nirS) in continental margin sediments of the Gulf of Mexico. FEMS

Microbiology Ecology 58: 464-475.

Treusch, A.H., Leininger, S., Kletzin, A., Schuster, S.C., Klenk, H.P., and Schleper, C.

(2005) Novel genes for nitrite reductase and Amo-related proteins indicate a role of

uncultivated mesophilic crenarchaeota in nitrogen cycling. Environmental


21

Trimmer, M., Nicholls, J.C., and Deflandre, B. (2003) Anaerobic ammonium

oxidation measured in sediments along the Thames estuary, United Kingdom. Applied


Venter, J.C., Remington, K., Heidelberg, J.F., Halpern, A.L., Rusch, D., Eisen, J.A. et

al. (2004) Environmental genome shotgun sequencing of the Sargasso Sea. Science

304: 66-74.

Ward, B.B., and O'Mullan, G.D. (2002) Worldwide distribution of Nitrosococcus

oceani, a marine ammonia-oxidizing gamma-proteobacterium, detected by PCR and

sequencing of 16S rRNA and amoA genes. Applied and Environmental Microbiology

68: 4153-4157.

Winogradsky, S. (1890) Sur les organismes de la nitrofication. Comptes Rendus

Biologies 110: 1013-1016.

Wuchter, C., Abbas, B., Coolen, M.J., Herfort, L., van Bleijswijk, J., Timmers, P. et

al. (2006) Archaeal nitrification in the ocean. Proceedings of the National Academy of

Sciences 103: 12317-12322.

Yan, T.F., Fields, M.W., Wu, L.Y., Zu, Y.G., Tiedje, J.M., and Zhou, J.Z. (2003)

Molecular diversity and characterization of nitrite reductase gene fragments (nirK and

22

nirS) from nitrate- and uranium-contaminated groundwater. Environmental


Zumft, W.G. (1997) Cell biology and molecular basis of denitrification. Microbiology

And Molecular Biology Reviews 61: 533-542.

23

Chapter 2. Determining the distribution of marine and coastal ammonia-oxidizing archaea and bacteria using a quantitative approach Annika C. Mosier1 and Christopher A. Francis1

1Department of Environmental Earth System Science, Stanford University, Stanford, CA 94305-4216, USA Published as: Mosier, A.C., and C.A. Francis. 2011. Determining the distribution of marine and coastal ammonia-oxidizing archaea and bacteria using a quantitative approach. Methods in Enzymology 486: 205-221.

24

Abstract

The oxidation of ammonia to nitrite is the first and often rate-limiting step in

nitrification and plays an important role in both nitrogen and carbon cycling. This

process is carried out by two distinct groups of microorganisms: ammonia-oxidizing

archaea and ammonia-oxidizing bacteria. This chapter describes methods for

measuring the abundance of ammonia-oxidizing archaea and bacteria using ammonia

monooxygenase subunit A (amoA) genes, with a particular emphasis on marine and

coastal systems. We also describe quantitative measures designed to target two

specific clades of marine ammonia-oxidizing archaea: the ‘shallow’ and ‘deep’ water

column AOA.

25

Introduction

Nitrification—the microbial conversion of ammonia to nitrate—links the

fixation of atmospheric nitrogen (N2 gas to ammonium) to nitrogen removal processes

(denitrification and anaerobic ammonium oxidation [anammox]), which ultimately

convert nitrate or ammonium to N2 gas. Through tight coupling with nitrogen removal

processes, nitrification plays a critical role in reducing the associated risks of elevated

dissolved inorganic nitrogen (DIN: ammonia, nitrate, and nitrite) often found in

coastal systems, including loss of biodiversity, nuisance/toxic algal blooms, and

depleted dissolved oxygen (e.g., Bricker et al., 1999). In fact, it is estimated that over

50% of external DIN inputs to estuaries are removed by these microbial processes

(e.g., Seitzinger, 1988). Nitrifiers are thought to be primarily autotrophic and thus

also play a role in supplying organic carbon to coastal and marine systems.

Nitrification is carried out by two functional groups (or guilds) of

chemoautotrophic organisms: ammonia-oxidizers convert ammonia to nitrite and then

nitrite-oxidizers convert nitrite to nitrate. Ammonia oxidation is thought to be the

rate-limiting step of the overall reaction, in part because the free energy of the reaction

is significantly higher than that of nitrite oxidation and because nitrite rarely

accumulates in the environment. Two distinct groups of microbes are capable of

ammonia oxidation: (1) the recently discovered ammonia-oxidizing archaea (AOA);

and (2) the ammonia-oxidizing bacteria (AOB), including betaproteobacteria (β-AOB)

from the genera Nitrosomonas and Nitrosospira, as well as gammaproteobacteria (γ-

AOB) from the genus Nitrosococcus.

26

Despite the biogeochemical importance of ammonia oxidation, only recently

have quantitative molecular approaches been employed to determine the relative

abundance of the key microbes responsible for this process—the AOA and AOB.

AOB amoA genes appear to be more abundant than AOA amoA genes in many coastal

and estuarine sediments based on quantitative PCR (qPCR) estimates (Caffrey et al.,

2007; Mosier and Francis, 2008; Santoro et al., 2008; Magalhaes et al., 2009; Wankel

et al., in revision). However, AOA amoA genes are more abundant than AOB amoA

genes in other estuaries and salt marshes (Caffrey et al., 2007; Moin et al., 2009; Abell

et al., 2010; Bernhard et al., 2010). The relative distribution of these two ammonia-

oxidizing groups in estuarine environments has been shown to depend on salinity

(Mosier and Francis, 2008; Santoro et al., 2008), sediment C:N ratio (Mosier and

Francis, 2008), oxygen concentration (Santoro et al., 2008), and chlorophyll-a (Abell

et al., 2010). Other physical and geochemical factors shown to specifically affect

AOA, AOB, or overall nitrification rates will likely impact the ratio of AOA:AOB,

such as sulfide concentrations (Joye and Hollibaugh, 1995), Fe(III) content (Dollhopf

et al., 2005), light (Horrigan and Springer, 1990), temperature (Berounsky and Nixon,

1993), and exogenous inputs from rivers and wastewater treatment plants (Dang et al.,

2010). Based on a review of the current literature, Erguder et al. (2009) proposed

specific AOA and AOB niches corresponding to varying dissolved oxygen, ammonia,

pH, phosphate, and sulfide levels. Substrate availability likely plays a major role in

the relative distribution of AOA versus AOB, particularly since it was recently

demonstrated that the cultivated ammonia-oxidizing crenarchaeote, Nitrosopumilus

maritimus SCM1, has a far greater affinity (lower Km) for ammonium than many AOB

27

(Martens-Habbena et al., 2009). The community composition of AOA and AOB

should be also considered when evaluating the relative abundance of these two groups

across different sites because ammonia-oxidizers often exhibit distinct spatial structure

in coastal/estuarine systems with different phylotypes dominanting at different sites

(e.g., Francis et al., 2003; Bernhard et al., 2005; Beman and Francis, 2006; Freitag et

al., 2006; Mosier and Francis, 2008).

In the open ocean, fixed nitrogen typically enters the deep ocean as ammonia

(NH3/NH4+) but it accumulates as nitrate (Karl, 2007). Nitrifiers rapidly convert the

ammonia into nitrate, and thus are responsible for producing the large nitrate reservoir

(20-40 µM) in the deep waters. Nitrification also plays a critical role in the upper

ocean, both as a source of NO3- to fuel primary production (Wankel et al., 2007; Yool

et al., 2007) and as a source of the greenhouse gas N2O to the atmosphere (Dore et al.,

1998). Interestingly, a number of studies have now shown that AOA amoA genes are

often several 100- to 1000-fold more abundant than AOB amoA genes in the open

ocean (Wuchter et al., 2006; Mincer et al., 2007; Beman et al., 2008; Beman et al.,

2010; Santoro et al., 2010).

It appears that the vast majority of crenarchaea in the open ocean (particularly

in the mesopelagic zone) possess ammonia monooxygenase genes and thus may be

capable of oxidizing ammonia to nitrite. In fact, several oceanic water column studies

have found that archaeal amoA gene copy numbers are equal to or even higher than

archaeal 16S rRNA gene copies, based on qPCR estimates. For instance, archaeal

amoA and marine Group 1 (MGI) crenarchaeal 16S rRNA genes have a ratio of ≥1 at

the San Pedro Ocean Time-series (SPOT) site in the Southern California Bight

28

(Beman et al., 2010), the Guaymas and Carmen Basins in the Gulf of California

(Beman et al., 2008), the central California Current (Santoro et al., 2010), Monterey

Bay (Mincer et al., 2007), the central Pacific Ocean (Church et al., 2010), and station

ALOHA near Hawaii (Mincer et al., 2007). Wuchter et al. (2006) also found similar

results when comparing amoA gene copy numbers to crenarchaeal cell counts

generated by catalyzed reporter deposition-fluorescence in situ hybridization (CARD-

FISH). In a whole-genome shotgun library from 4,000-m depth at Station ALOHA,

Konstantinidis et al. (2009) showed that the ratio of individual crenarchaeal ammonia

monooxygenase sequence reads to crenarchaeal 16S rRNA sequence reads was

consistent with a 1:1 amoA:16S rRNA ratio. Taken together, all of these results imply

that most MGI crenarchaea contain amoA genes and thus are likely capable of

oxidizing ammonia. Indeed, MGI crenarchaeal 16S rRNA gene copies and group A

(see below) archaeal amoA gene copies were significantly correlated with measured

15NH4+oxidation rates in the Gulf of California (Beman et al., 2008), providing further

evidence for marine crenarchaeal nitrification. Considering that previous studies have

shown that MGI crenarchaea constitute ~10-40% of the total microbial community in

the ocean below the euphotic zone (Karner et al., 2001; Herndl et al., 2005; Teira et

al., 2006; Kirchman et al., 2007), the crenarchaea are likely the dominant ammonia-

oxidizers in the open ocean and ultimately responsible for the deep nitrate reservoir.

Studies in the North Atlantic (Agogué et al., 2008), Eastern Mediterranean Sea

(de Corte et al., 2009) and Antarctic (Kalanetra et al., 2009) have suggested that some

crenarchaea lack the amoA gene because of low 16S rRNA:amoA gene ratios, and may

thus be incapable of ammonia oxidation. However, all three of these studies used

29

amoA qPCR primers developed by Wuchter et al. (2006), which have several

mismatches to marine AOA and therefore may lead to spurious quantitative results

(primer mismatches analyzed and discussed by Konstantinidis et al., 2009; Church et

al., 2010; and Santoro et al., 2010). Thus it seems that the majority of crenarchaea in

the ocean likely possess the capacity to oxidize ammonia. Even still, some marine

crenarchaea may utilize alternate metabolic pathways. For instance, isotopic

(Ouverney and Fuhrman, 2000; Herndl et al., 2005; Teira et al., 2006; Varela et al.,

2008) and genomic (Hallam et al., 2006; Walker et al., 2010; Blainey et al., 2011)

studies have suggested that AOA may be capable of mixotrophic growth.

Additionally, urease genes have been reported in the Cenarchaeum symbiosum A

genome (Hallam et al., 2006) and in a crenarchaeal genomic scaffold from the Pacific

Ocean (Konstantinidis et al., 2009), suggesting that at least some marine AOA may

have the potential to use urea as an energy source. It is also possible that the archaeal

AMO proteins act on substrates other than, or in addition to, ammonia. Further

studies, including cultivation-based approaches, will be necessary to confirm whether

or not AOA derive energy from sources other than ammonia.

Methods for Measuring the Abundance of AOA and AOB within Marine and

Coastal Systems

Sample Collection

Sediment cores for molecular analyses are generally collected from a Young-

modified, Van Veen grab sampler (deployed off a research vessel) using sterile, cut-

off 5-cc syringes. Syringe cores are collected towards the interior of the sediment grab

30

in order to avoid disturbed sediment near the outer edges of the sampler. Larger cores

can be collected using larger syringes (e.g., 60-cc), PVC piping, or acrylic tubing.

Water samples are collected at discrete depths with a Niskin or multi-bottle rosette

sampler. Water samples (1-4 liters) are vacuum-filtered onto a 25 mm, 0.2 µm filter

and filters are placed in bead-beating tubes. The water samples can also be size

fractionated using filters with different pore sizes (e.g., 10 µm pore size filter to

remove larger plankton). For filters specified for RNA extraction, 600 µl RLT Buffer

(Qiagen) with 1% ß-mercaptoethanol is added to the tube containing the filter (Santoro

et al., 2010). Water filters and sediment cores for DNA and RNA analyses are

immediately frozen on liquid nitrogen or dry ice until permanent storage at -80°C.

DNA Extraction

We routinely use the FastDNA SPIN Kit for Soil (MP Biomedicals) to recover

total community DNA from sediment samples (Francis et al., 2005; Beman and

Francis, 2006; Santoro et al., 2008; Mosier and Francis, 2008). The kit produces

‘PCR-ready’ genomic DNA after bead-beating lysis, sample homogenization and

protein solubilization, and DNA purification. DNA is extracted from approximately

0.5g of sediment from the upper 5 mm of the syringe cores. All steps in the

manufacturer’s protocol are followed, except that the purified DNA is eluted in 100 µl

DNase/RNase-free water.

For water samples, we use the DNA extraction method described by Santoro et

al. (2010). Briefly, sucrose-EDTA lysis buffer and 10% sodium dodecyl sulfate are

added to the bead-beating tubes containing the frozen filters. Following bead beating,

31

proteinase K is added and filters are incubated at 55°C for approximately 4 h. The

supernatant is then purified using DNeasy columns (Qiagen) following the

manufacturer’s protocol with the incorporation of an additional wash step with wash

buffer AW2 (Qiagen). The purified DNA is eluted in 200 µl of DNase/RNase-free

water.

RNA Extraction and cDNA Synthesis

Total RNA extracts are analyzed to determine which amoA genes are actively

expressed in the environment. RNA is extracted from sediment samples using the

RNA PowerSoil Total RNA Isolation Kit (MO BIO). RNA from filtered water

samples is extracted according the methods described by Santoro et al. (2010), which

are based on modifications of the RNeasy Kit (Qiagen). RNA extracts are

immediately treated with DNase I to remove any residual DNA. Complimentary DNA

(cDNA) is generated from mRNA using the SuperScript III first strand cDNA

synthesis kit (Invitrogen).

PCR Screening and Gene Sequencing

Mixed template DNA and cDNA is PCR amplified with primers targeting

betaproteobacterial and archaeal amoA: amoA-1F* (Stephen et al., 1999) or amoA-1F

and amoA-2R (Rotthauwe et al., 1997) for ß-AOB; Arch-amoAF and Arch-amoAR

(Francis et al., 2005) for AOA. Clone libraries of amoA gene fragments are

constructed using a TOPO TA cloning kit (Invitrogen) and individual clones are

sequenced. Both nucleic acid and amino acid sequences are subjected to phylogenetic

32

analysis. Sequences are manually aligned with GenBank sequences using MacClade

(http://macclade.org) and phylogenetic trees are constructed in ARB (Ludwig et al.,

2004).

Quantifying Archaeal and Bacterial amoA Genes and Transcripts

Quantitative PCR (qPCR) is used to estimate the number of archaeal and

bacterial amoA gene or transcript copies in sediment or water samples. It is advisable

to analyze replicate DNA or RNA extractions for each sediment or water sample to

minimize biases associated with sample heterogeneity and extraction efficiency.

Reactions are carried out in a qPCR thermal cycler such as the StepOnePlus™ Real-

Time PCR System (Applied Biosystems).

A standard curve is prepared from a dilution series of control template of

known concentration, such as a plasmid containing a cloned gene of interest, genomic

DNA, or cDNA. We use plasmids containing cloned amoA PCR amplicons that have

been extracted with the Qiagen Miniprep Spin Kit. Plasmids are linearized with the

NotI restriction enzyme, purified, and quantified using the Quant-iT™ Broad Range

DNA assay with the Qubit Fluorometer (Invitrogen). The standard curve should

include at least four data points spanning the range of template concentrations in a

given set of samples (often several orders of magnitude). Following amplification, the

threshold cycle (Ct) value for each standard dilution is plotted against the logarithm of

the copy number of each dilution, generating the standard curve. The efficiency of the

qPCR reaction is calculated using the equation below:

33

Efficiency = [10(–1/slope)]–1]100

When the efficiency is perfect (100%), the slope of the standard curve is –3.32,

indicating that there is a perfect doubling of target amplicon every cycle. Correlation

coefficients (R2) for the standard curve should be very close to 1 (ideally > 0.985).

Melt curves (for SYBR assays) and gels should be analyzed to check the specificity of

amplification.

Sample nucleic acids are simultaneously amplified alongside the standards.

Sample Ct values are compared to the standard curve to estimate the copy numbers

within each reaction. All sample reactions are performed in triplicate and an average

value is calculated. An outlier may be removed from some triplicate measurements to

maintain a standard deviation of less than 10-15% for each sample. The number of

genes or transcripts in the original sample can then be calculated using the initial

volume of water filtered or weight of sediment extracted, the total amount of nucleic

acids recovered from the extraction, and the amount of nucleic acids added to each

qPCR reaction.

SYBR Green qPCR assays are commonly used for amplification of archaeal

and bacterial amoA genes or transcripts. Although TaqMan assays have advantages

over SYBR assays, divergence of the archaeal and bacterial amoA genes precludes

development of universal TaqMan probes that capture the full phylogenetic diversity

of each gene. Primers Arch-amoAF and Arch-amoAR (Francis et al., 2005) are

frequently used for AOA amoA quantification and amoA-1F and amoA-2R

(Rotthauwe et al., 1997) for ß-AOB amoA quantification. Other archaeal amoA

primers used for qPCR in marine studies (e.g., Bernhard et al., 2010; Church et al.,

34

2010; Mincer et al., 2007; Moin et al., 2009) include CrenAmoAQ-F (Mincer et al.,

2007) and CrenAmoAQModF (Moin et al., 2009). Variations on the amoA-1F and

amoA-2R primers have been used in some marine studies (e.g., Dang et al., 2010 and

Magalhaes et al., 2009) for ß-AOB amoA quantification: amoA-1F* (Stephen et al.,

1999) and amoA-2R’ (Okano et al., 2004). Primer amoA-3F and amoB-4R (Purkhold

et al., 2000) have been used to quantify γ-AOB (Lam et al., 2009).

Reaction conditions for SYBR assays using primers Arch-amoAF and Arch-

amoAR for AOA amoA quantification and amoA-1F and amoA-2R for ß-AOB amoA

quantification follow. qPCR reactions are prepared on a cold block (~4°C) to avoid

non-specific priming. AOA amoA qPCR is performed in a 25 µl reaction mixture with

a known amount of template DNA (typically 0.2-10 ng), 0.4 µM of each primer, 2

mM MgCl2, 1.25 Units AmpliTaq DNA polymerase (Applied Biosystems), 1 µl

passive reference dye, 40 ng µl-1 BSA and 12.5 µl Failsafe Green Premix E

(Epicentre). The qPCR protocol is as follows: 3 min at 95°C, then 32 cycles

consisting of 30 s at 95°C, 45 s at 56°C and 1 min at 72°C with a detection step at the

end of each cycle. ß-AOB amoA qPCR is performed in the same manner but with 0.3

µM of each primer, Failsafe Green Premix F (Epicentre), and no addition of BSA or

MgCl2. The PCR protocol is as follows: 5 min at 95°C, then 34 cycles consisting of

45 s at 94°C, 30 s at 56°C, 1 min at 72°C and a detection step for 10 s at 80.5°C.

Other Potential Target Genes for Ammonia-Oxidizers

The sheer number of amoA sequences in GenBank (e.g., currently over 8,000

archaeal amoA sequences) makes these genes excellent candidates for phylogenetic

35

and quantitative studies. However, there are also a number of other genes that can

potentially be used to target ammonia-oxidizers. Primers have been designed to

amplify the amoB and amoC subunits of the ammonia monooxygenase enzyme for

both AOA and AOB (Purkhold et al., 2000; Norton et al., 2002; Könneke et al., 2005;

Junier et al., 2008b; Junier et al., 2008a). Genes encoding other enzymes involved in

key nitrogen transformations have been amplified from AOB, including the copper-

containing dissimilatory nitrite reductase gene (nirK; Casciotti and Ward, 2001) and

the nitric oxide reductase gene (norB; Casciotti and Ward, 2005). However, primers

for these genes must to be optimized for qPCR assays to ensure that they are only

targeting AOB, while still capturing the full range of sequence diversity within AOB.

Notably, putative nirK homologues have recently been identified in marine (Walker et

al., 2010), low-salinity (Blainey et al., 2011), and soil AOA (Treusch et al., 2005;

Bartossek et al., 2010), which may prove to be powerful molecular markers for

characterizing and quantifying AOA in marine and estuarine systems (Lund and

Francis, unpublished).

Beyond nitrogen cycling genes, carbon fixation genes can also be targeted for

phylogenetic and quantitative studies of ammonia-oxidizers. While RuBisCO genes

can be used to examine AOB (Sinigalliano et al., 2001; Utåker et al., 2002), AOA

apparently lack RuBisCO genes (Hallam et al., 2006; Walker et al., 2010; Blainey et

al., 2011). Instead, genes encoding the acetyl-CoA carboxylase enzyme (e.g., accC)

involved in the 3-hydroxypropionate/4-hydroxybutyrate autotrophic carbon fixation

cycle have been used as molecular markers for CO2 assimilation linked to archaeal

autotrophy (Auguet et al., 2008). While this enzyme is found in all sequenced AOA

36

genomes (Hallam et al., 2006; Walker et al., 2010; Blainey et al., 2011), it is also

found in other (non-ammonia-oxidizing) crenarchaea, so the primer sets should be

further refined for a qPCR assay specific to the AOA. In general, we prefer using C-

cycling functional genes as a complement to (rather than a replacement for) amoA-

based quantitiave analyses of AOA and AOB in marine and coastal systems, as amoA

genes are directly linked to the process of ammonia oxidation.

Methodological Considerations

It is important to note that different DNA extraction protocols may yield

different gene copy numbers. For instance, Leininger et al. (2006) demonstrated that

the use of different bead-beating times yielded different abundances of AOA and AOB

amoA gene copies. In four different soil samples, bead-beating times of 30 s for

quantification of AOA amoA and between 120-150 s for quantification of bacterial

amoA yielded the highest abundances. Thus, care should be taken when comparing

absolute copy numbers across different samples and/or studies, however, general

patterns and trends can still yield valuable insights into the dynamics of microbial

populations.

PCR inhibition should also be considered in quantitative assays. The effects of

inhibition can be seen by comparing the amplification of a positive control with the

amplification of the positive control plus environmental template: inhibitors in the

environmental template would decrease the amplification efficiency of the positive

control. Additionally, PCR inhibition can be detected by screening amplification of

varying concentrations of template DNA (e.g., 0.5, 1, 5, 10 ng): the presence of

37

inhibitors at higher DNA concentrations can result in lower quantification values.

When setting up a new qPCR experiment, careful consideration should be

given to primer and probe choice and design. The specificity of the primers/probes is

critical to ensure that the assay only quantifies the target gene of interest. If the

primers are too specific, the qPCR results will underestimate the abundance of the

target gene; however, if the primers are too broad, various non-target sequences will

be amplified (e.g., amplify both amoA and pmoA sequences) and abundance will be

overestimated. We design new primers/probes (generally between 15-30 bp in length)

by manually visualizing a multi-sequence nucleotide alignment and determining

candidate oligonucleotide regions. After taking G+C content, potential secondary

structure, and annealing temperature into consideration, we run a BLAST search on

the primer/probe sequences to determine which targets they might anneal to.

Candidate primers/probes are then scanned against the entire database of sequences

using ARB to determine specificity in silico.

The qPCR assay and thermal cycle program are then optimized. Particular

attention is given to the choice of qPCR premix, primer/probe concentrations, and

MgCl2 concentrations. A variety of qPCR premixes are commercially available (e.g.,

Epicentre’s Failsafe Green Premix A-L, Invitrogen’s AccuPrime SuperMix I-II,

Applied Biosystem’s Power SYBR Green PCR Master Mix, etc.) and each is made

with a proprietary recipe with varying concentrations of MgCl2, PCR enhancers (e.g.,

betaine), BSA, dNTPs, etc. Therefore, it is necessary to test several different premixes

when optimizing a new qPCR assay because some will undoubtedly work better than

others. Primer concentrations influence the amplification efficiency and specificity of

38

the reaction. Low concentrations can decrease PCR efficiency, whereas high

concentrations can cause nonspecific priming and primer-dimer formation. MgCl2

concentrations influence DNA polymerase activity and fidelity, DNA denaturation

temperature, primer annealing, PCR specificity, and primer-dimer formation.

Generally speaking, excess magnesium can cause non-specific amplification and

insufficient magnesium reduces the overall yield. Finally, it is imperative to sequence

the amplification products from the qPCR assay to ensure that only the product of

interest is amplified under the optimized conditions.

Fluorescent in situ hybridization (FISH) is an alternative approach for

quantifying AOA and AOB in marine and coastal systems. FISH requires 16S rRNA

gene probes that have high specificity for AOA or AOB and yet capture the full

phylogenetic diversity with each group. Functional gene- and mRNA-based FISH

approaches are in development but are still not widely accessible. We prefer to use

qPCR to quantify AOA and AOB because of the high throughput nature of the

method, the ability to rapidly examine multiple genes/organisms in the same samples,

and the ease of use with multiple sample types (e.g., sediment, water column, soil).

Targeting Specific Ecotypes: Quantifying Shallow and Deep Clades of Marine

Water Column AOA

The vast majority of archaeal amoA genes from marine water columns fall into

two phylogenetically distinct clades: group A and group B (first identified by Francis

et al., 2005). These clades are distinct from N. maritimus, which falls in the

marine/coastal sediment clade (approximately 82% nucleotide identity between group

39

A and N. maritimus amoA sequences and 75% between group B and N. maritimus).

Group A amoA sequences appear to represent a primarily surface water ecotype

(predominately found at depths <200 m) and group B amoA sequences represent a

deep-water ecotype (predominately found at depths >200 m) (Hallam et al., 2006;

Mincer et al., 2007; Beman et al., 2008; Santoro et al., 2010). Crenarchaeal depth-

specific groups have also been identified from phylogenetic analyses of 16S rRNA

sequences (García-Martínez and Rodríguez-Valera, 2000; Massana et al., 2000). The

deep group B AOA are predominantly found in waters below ~200 m but they may

periodically be transported to the surface waters during upwelling events. In the

central California Current (off the coast of Monterey Bay), the deep group B amoA

sequences were abundant in clone libraries from the upper water column (25–100 m)

at a site experiencing upwelling, whereas group B sequences comprised only a small

percentage of the sequences (down to 150 m) at an offshore site without upwelling

(Santoro et al., 2010). Interestingly, deep group B amoA gene transcripts were not

detected at 25 m depth at the upwelling site, suggesting that these AOA may indeed be

specifically adapted to deep waters (Santoro et al., 2010).

In a study of the water column of the Gulf of California—a biologically

productive subtropical sea bordered by the Baja California peninsula and mainland

Mexico—Beman et al. (2008) designed amoA PCR primers/assays to specifically

quantify shallow (group A) versus deep (group B) clades of planktonic AOA, using

SYBR Green reaction chemistry (Table 1). Clade-specific qPCR analysis using these

assays revealed that AOA communities in the upper water column (0–100 m) were

composed exclusively of group A. In contrast, group B amoA genes comprised over

40

50% of the archaeal amoA copies within most of the deeper sampling depths within

the oxygen minimum zone (300-650m). Notably, the sum of the group A and B

assays correlated with the general AOA amoA qPCR assay (r2=0.89, slope=1.15).

Thus, although the extensively used Arch-amoAF/R primer set was designed based on

a limited number of sequences (Francis et al., 2005) and undoubtedly miss some

archaeal amoA sequences (de la Torre et al., 2008), these primers still work

remarkably well as a general proxy for AOA.

For certain applications (e.g., high-throughput and/or automated qPCR

screening), it is particularly useful, if not essential, to have TaqMan assays available

for quantifying specific genes of interest. Toward this end, we have designed

TaqMan-based qPCR assays to target the shallow group A and deep group B archaeal

amoA clades. The assays are specific for amoA sequence types from each group with

no cross-amplification. For both assays, the qPCR protocol is as follows: 94°C for 75

s, then 42 cycles consisting of 94°C for 15 s and 55°C for 30 s. Each 25 µL reaction

contains: 12.5 µL AccuPrime™ SuperMix I (Invitrogen), 1 µL template DNA, 2.5

mM MgCl2, forward primer, reverse primer, and probe. Primers and probes for the

individual assays are shown in Table 1.

In addition to standard qPCR analyses, these AOA qPCR assays can be

deployed on the Environmental Sample Processor (ESP; Scholin et al., 2001; Scholin

et al., 2009; http://www.mbari.org/ESP/). The ESP is a field-deployable system that

can quantify specific genes of interest (Scholin et al., 2009;

http://www.mbari.org/esp/mfb/mfb.htm; http://www.mbari.org/ESP/PCR/pcr.htm),

detect 16S rRNAs without amplification (recently demonstrated in Monterey Bay, CA;

41

Preston et al., 2009), and measure relevant environmental parameters using bundled

chemical and physical sensors. In collaboration with Christopher Scholin's group

(MBARI), our qPCR assays are being used on the ESP to assess fine scale variations

in the abundance of the shallow group A and deep group B archaeal amoA clades in

surface and deep waters in Monterey Bay, CA. Preliminary data shows that at 891

meters, deep group B archaeal amoA gene copies are several orders of magnitude

more abundant than the shallow group A clade (Figure 1). The abundance of both the

shallow and deep clades fluctuates significantly over time (Figure 1). Persistent

measurements of these amoA genes in Monterey Bay and elsewhere will illuminate

marine AOA dynamics on a time continuum that cannot be captured by traditional

discrete shipboard sampling. Quantitative data generated by the ESP has the potential

to refine, if not dramatically change, our understanding of the molecular microbial

ecology of the oceanic nitrogen cycle.

Acknowledgments

We thank two anonymous reviewers for constructive feedback on this

manuscript and Christopher Scholin for useful comments on the ESP text. We also

thank both Christopher Scholin and Christina Preston for providing access to samples

collected in Monterey Bay. This work was supported in part by National Science

Foundation grants MCB-0604270, OCE-0825363, and OCE-0847266 (to C.A.F.), as

well as an EPA STAR Graduate Fellowship (to A.C.M.).

42

Tables

Primer/Probe Sequence (5'-3') Rxn Conc. Reference

WCA-amoA-F ACACCAGTTTGGCTWCCDTCAGC 0.5 !M modified from Beman et al., 2008

WCA-amoA-R TCAGCCACHGTGATCAAATTG 0.333 !M this study

WCA-amoA-P FAM-ACTCCGCCGAACAGTATCA-BHQ1 0.4 !M this study

Arch-amoAFB CATCCRATGTGGATTCCATCDTG 1.5 !M Beman et al., 2008

WCB-amoA-R AAYGCAGTTTCTAGYGGATC 1.0 !M this study

WCB-amoA-P FAM-CCAAAGAATATYAGCGARTG-BHQ1 0.4 !M this study

Arch-amoAFA ACACCAGTTTGGYTACCWTCDGC 0.4 !M Beman et al., 2008

Arch-amoAR GCGGCCATCCATCTGTATGT 0.4 !M Francis et al., 2005

Arch-amoAFB CATCCRATGTGGATTCCATCDTG 0.4 !M Beman et al., 2008

Arch-amoAR GCGGCCATCCATCTGTATGT 0.4 !M Francis et al., 2005

Taqman assay: deep water column group B archaeal amoA

SYBR assay: shallow water column group A archaeal amoA

SYBR assay: deep water column group B archaeal amoA

Taqman assay: shallow water column group A archaeal amoA

Table 1. Primers and probes for archaeal amoA assays specific to water column group A and group B

marine sequences.

43

Figures

Figure 1. Abundance of the shallow and deep archaeal amoA clades within seawater

collected at a depth of 891 meters in Monterey Bay, CA. Note difference in scales (y-

axis) for group A versus B archaeal amoA qPCR data.

!"

#"

$"

%"

&"

'!"

'#"

'$"

'%"

'&"

!"

(!!"

'!!!"

'(!!"

#!!!"

#(!!"

)!!!"

)(!!"

$!!!"

$(!!"

(!!!"

'!*'*!+" '!*)*!+" '!*(*!+" '!*%*!+" '!*)!*!+"

!"#$$%&'&#()*'+%$,-.'/*%,0'1'#*+"#)#$'!"#$%+%02)!3-4'!)#&#()*'

5))0'&#()*'+%$,-.'/*%,0'6'#*+"#)#$'!"#$%+%02)!3-4'!)#&#()*'

5#()'+%$$)+()5'

,--."/012."3" 4567718"/012."9" 4567718"/012."9":";-6420-,"<2="<-718",-=->?1@"7A;A="

44

Literature Cited

Abell, G.C.J., Revill, A.T., Smith, C., Bissett, A.P., Volkman, J.K., and Robert, S.S.

(2010) Archaeal ammonia oxidizers and nirS-type denitrifiers dominate sediment

nitrifying and denitrifying populations in a subtropical macrotidal estuary. The ISME

Journal 4: 286-300.

Agogué, H., Brink, M., Dinasquet, J., and Herndl, G.J. (2008) Major gradients in

putatively nitrifying and non-nitrifying Archaea in the deep North Atlantic. Nature

456: 788-791.

Auguet, J.-C., Borrego, C.M., Bañeras, L., and Casamayor, E.O. (2008) Fingerprinting

the genetic diversity of the biotin carboxylase gene (accC) in aquatic ecosystems as a

potential marker for studies of carbon dioxide assimilation in the dark. Environmental


Bartossek, R., Nicol, G.W., Lanzen, A., Klenk, H.-P., and Schleper, C. (2010)

Homologues of nitrite reductases in ammonia-oxidizing archaea: diversity and

genomic context. Environmental Microbiology 12: 1075-1088.



Mexico. Applied and Environmental Microbiology 72: 7767-7777.

45

Beman, J.M., Popp, B.N., and Francis, C.A. (2008) Molecular and biogeochemical

evidence for ammonia oxidation by marine Crenarchaeota in the Gulf of California.

The ISME Journal 2: 429-441.

Beman, J.M., Sachdeva, R., and Fuhrman, J.A. (2010) Population ecology of

nitrifying Archaea and Bacteria in the Southern California Bight. Environmental





Bernhard, A.E., Landry, Z.C., Blevins, A., De La Torre, J.R., Giblin, A.E., and Stahl,

D.A. (2010) Abundance of ammonia-oxidizing archaea and bacteria along an estuarine

salinity gradient in relation to potential nitrification rates. Applied and Environmental


Bricker, S.B., Clement, C., Pirhalla, D., Orlando, S., and Farrow, D. (1999) National

Estuarine Eutrophication Assessment: Effects of Nutrient Enrichment in the Nation's

Estuaries. NOAA, National Ocean Service, Special Projects Office and the National

Centers for Coastal Ocean Science Silver Spring, MD: 71.

46

Caffrey, J.M., Bano, N., Kalanetra, K., and Hollibaugh, J.T. (2007) Ammonia

oxidation and ammonia-oxidizing bacteria and archaea from estuaries with differing

histories of hypoxia. The ISME Journal 1: 660-662.

Casciotti, K.L., and Ward, B.B. (2001) Dissimilatory nitrite reductase genes from

autotrophic ammonia-oxidizing bacteria. Applied and Environmental Microbiology

67: 2213-2221.

Casciotti, K.L., and Ward, B.B. (2005) Phylogenetic analysis of nitric oxide reductase

gene homologues from aerobic ammonia-oxidizing bacteria. FEMS Microbiology

Ecology 52: 197-205.

Church, M., Wai, B., Karl, D., and Delong, E. (2010) Abundances of crenarchaeal

amoA genes and transcripts in the Pacific Ocean. Environmental Microbiology 12:

679-688.

de la Torre, J.R., Walker, C.B., Ingalls, A.E., Konneke, M., and Stahl, D.A. (2008)

Cultivation of a thermophilic ammonia oxidizing archaeon synthesizing crenarchaeol.


De Corte, D., Yokokawa, T., Varela, M.M., Agogué, H., Herndl, G.J. (2009) Spatial

distribution of Bacteria and Archaea and amoA gene copy numbers throughout the

water column of the Eastern Mediterranean Sea. The ISME Journal 3: 147-158.

47

Dang, H., Li, J., Chen, R., Wang, L., Guo, L., Zhang, Z., and Klotz, M.G. (2010)

Diversity, abundance, and spatial distribution of sediment ammonia-oxidizing

Betaproteobacteria in response to environmental gradients and coastal eutrophication

in Jiaozhou Bay, China. Applied And Environmental Microbiology 76: 4691-4702.

Dore, J.E., Popp, B.N., Karl, D.M., and Sansone, F.J. (1998) A large source of

atmospheric nitrous oxide from subtropical North Pacific surface waters. Nature 396:

63-66.

Erguder, T., Boon, N., Wittebolle, L., Marzorati, M., and Verstraete, W. (2009)

Environmental factors shaping the ecological niches of ammonia-oxidizing archaea.

FEMS Microbiology Reviews 33: 855-869.

Francis, C.A., O'Mullan, G.D., and Ward, B.B. (2003) Diversity of ammonia





of the ocean. Proceedings of the National Academy of Sciences 102: 14683-14688.

48



freshwater-marine gradient. Environmental Microbiology 8: 684-696.

García-Martínez, J., and Rodríguez-Valera, F. (2000) Microdiversity of uncultured

marine prokaryotes: the SAR11 cluster and the marine Archaea of Group I. Molecular

Ecology 9: 935-948.

Hallam, S.J., Mincer, T.J., Schleper, C., Preston, C.M., Roberts, K., Richardson, P.M.,

and DeLong, E.F. (2006) Pathways of carbon assimilation and ammonia oxidation

suggested by environmental genomic analyses of marine Crenarchaeota. PLoS Biology

4: e95.

Herndl, G.J., Reinthaler, T., Teira, E., van Aken, H., Veth, C., Pernthaler, A., and

Pernthaler, J. (2005) Contribution of Archaea to total prokaryotic production in the

deep Atlantic Ocean. Applied and Environmental Microbiology 71: 2303-2309.

Junier, P., Kim, O.-S., Hadas, O., Imhoff, J.F., and Witzel, K.-P. (2008a) Evaluation

of PCR primer selectivity and phylogenetic specificity by using amplification of 16S

rRNA genes from betaproteobacterial ammonia-oxidizing bacteria in environmental

samples. Applied and Environmental Microbiology 74: 5231-5236.

49

Junier, P., Kim, O.S., Molina, V., Limburg, P., Junier, T., Imhoff, J.F., and Witzel,

K.P. (2008b) Comparative in silico analysis of PCR primers suited for diagnostics and

cloning of ammonia monooxygenase genes from ammonia-oxidizing bacteria. FEMS


Karl, D.M. (2007) Microbial oceanography: paradigms, processes and promise.

Nature Reviews Microbiology 5: 759-769.

Karner, M.B., DeLong, E.F., and Karl, D.M. (2001) Archaeal dominance in the

mesopelagic zone of the Pacific Ocean. Nature 409: 507-510.

Kirchman, D.L., Elifantz, H., Dittel, A.I., Malmstrom, R.R., and Cottrell, M.T. (2007)

Standing stocks and activity of Archaea and Bacteria in the western Arctic Ocean.

Limnology and Oceanography 52: 495-507.



Nature 437: 543-546.

Konstantinidis, K.T., Braff, J., Karl, D.M., and Delong, E.F. (2009) Comparative

metagenomic analysis of a microbial community residing at a depth of 4,000 meters at

station ALOHA in the North Pacific subtropical gyre. Applied and Environmental


50

Lam, P., Lavik, G., Jensen, M.M., van de Vossenberg, J., Schmid, M., Woebken, D.,

et al. (2009) Revising the nitrogen cycle in the Peruvian oxygen minimum zone.


Leininger, S., Urich, T., Schloter, M., Schwark, L., Qi, J., Nicol, G.W. et al. (2006)


806-809.

Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Yadhukumar et al.

(2004) ARB: a software environment for sequence data. Nucleic Acids Research 32:

1363-1371.

Magalhaes, C.M., Machado, A., and Bordalo, A.A. (2009) Temporal variability in the

abundance of ammonia-oxidizing bacteria vs. archaea in sandy sediments of the Douro

River estuary, Portugal. Aquatic Microbial Ecology 56: 13-23.

Martens-Habbena, W., Berube, P.M., Urakawa, H., de La Torre, J.R., and Stahl, D.A.

(2009) Ammonia oxidation kinetics determine niche separation of nitrifying Archaea

and Bacteria. Nature 461: 976-979.

51

Massana, R., DeLong, E.F., and Pedrós-Alió, C. (2000) A few cosmopolitan

phylotypes dominate planktonic archaeal assemblages in widely different oceanic

provinces. Applied and Environmental Microbiology 66: 1777-1787.

Mincer, T.J., Church, M.J., Taylor, L.T., Preston, C.M., Karl, D.M., and DeLong, E.F.


Monterey Bay and the North Pacific Subtropical Gyre. Environmental Microbiology 9:

1162-1175.

Moin, N.S., Nelson, K.A., Bush, A., and Bernhard, A.E. (2009) Distribution and

diversity of archaeal and bacterial ammonia oxidizers in salt marsh sediments. Applied





Norton, J.M., Alzerreca, J.J., Suwa, Y., and Klotz, M.G. (2002) Diversity of ammonia

monooxygenase operon in autotrophic ammonia-oxidizing bacteria. Archives of


Okano, Y., Hristova, K.R., Leutenegger, C.M., Jackson, L.E., Denison, R.F.,

Gebreyesus, B. et al. (2004) Application of real-time PCR to study effects of

52

ammonium on population size of ammonia-oxidizing bacteria in soil. Applied And


Ouverney, C.C., and Fuhrman, J.A. (2000) Marine planktonic archaea take up amino

acids. Applied and Environmental Microbiology 66: 4829-4833.

Preston, C.M., Marin, R., Jensen, S.D., Feldman, J., Birch, J.M., Massion, E.I. et al.

(2009) Near real-time, autonomous detection of marine bacterioplankton on a coastal

mooring in Monterey Bay, California, using rRNA-targeted DNA probes.


Purkhold, U., Pommerening-Roser, A., Juretschko, S., Schmid, M.C., Koops, H-.P.,



molecular diversity surveys. Applied and Environmental Microbiology 66: 5368-5382.

Rotthauwe, J.H., Witzel, K.P., and Liesack, W. (1997) The ammonia monooxygenase

structural gene amoA as a functional marker: Molecular fine-scale analysis of natural

ammonia-oxidizing populations. Applied and Environmental Microbiology 63: 4704-

4712.



53

physicochemical gradients in a subterranean estuary. Environmental Microbiology 10:

1068-1079.

Santoro, A.E., Casciotti, K.L., and Francis, C.A. (2010) Activity, abundance and

diversity of nitrifying archaea and bacteria in the central California Current.


Scholin, C.A., Massion, E.I., Wright, D.K., Cline, D.E., Mellinger, E., and Brown, M.

(2001) Aquatic Autosampler Device. US patent 6187530.

Scholin, C.A., Doucette, G.J., Jensen, S., Roman, B., Pargett, D., Marin, R. et al.

(2009) Remote detection of marine microbes, small invertebrates, harmful algae, and

biotoxins using the Environmental Sample Processor (ESP). Oceanography 22: 158-

167.

Seitzinger, S.P. (1988) Denitrification in freshwater and coastal marine ecosystems:

Ecological and geochemical significance. Limnology and Oceanography 33: 702-724.

Sinigalliano, C.D., Kuhn, D.N., Jones, R.D., and Guerrero, M.A. (2001) In situ reverse

transcription to detect the cbbL gene and visualize RuBisCO in chemoautotrophic

nitrifying bacteria. Lett Appl Microbiol 32: 388-393.

54

Stephen, J.R, Chang, Y.J., Macnaughton, S.J., Kowalchuk, G.A., Leung, K.T.,

Flemming, C.A., and White, D.C. (1999) Effect of toxic metals on indigenous soil

beta-subgroup proteobacterium ammonia oxidizer community structure and protection

against toxicity by inoculated metal-resistant bacteria. Applied and Environmental


Teira, E., Lebaron, P., van Aken, H., and Herndl, G.J. (2006) Distribution and activity

of Bacteria and Archaea in the deep water masses of the North Atlantic. Limnology

and Oceanography 51: 2131-2144.

Treusch, A.H, Leininger, S., Kletzin, A., Schuster, S.C., Klenk, H-.P., and Schleper,

C. (2005) Novel genes for nitrite reductase and Amo-related proteins indicate a role of



Utåker, J.B., Andersen, K., Aakra, A., Moen, B., and Nes, I.F. (2002) Phylogeny and

functional expression of ribulose 1,5-bisphosphate carboxylase/oxygenase from the

autotrophic ammonia-oxidizing bacterium Nitrosospira sp. isolate 40KI. Journal Of

Bacteriology 184: 468-478.

Varela, M.M., van Aken, H.M., Sintes, E., and Herndl, G.J. (2008) Latitudinal trends

of Crenarchaeota and Bacteria in the meso- and bathypelagic water masses of the

Eastern North Atlantic. Environmental Microbiology 10: 110-124.

55

Walker, C.B., La Torre, J.R.D., Klotz, M.G., Urakawa, H., Pinel, N., Arp, D.J. et al.

(2010) Nitrosopumilus maritimus genome reveals unique mechanisms for nitrification

and autotrophy in globally distributed marine crenarchaea. Proceedings of the

National Academy of Sciences 107: 8818-8823.

Wankel, S.D., Kendall, C., Pennington, J.T., Chavez, F.P., and Paytan, A. (2007)

Nitrification in the euphotic zone as evidenced by nitrate dual isotopic composition:

Observations from Monterey Bay, California. Global Biogeochemical Cycles 21:

GB2009.

Wankel, S.D., Mosier, A.C., Hansel, C.M., Paytan, A., and Francis, C.A. Spatial

variability in nitrification rates and ammonia-oxidizing microbial communities in the

agriculturally-impacted Elkhorn Slough estuary. In revision.

Wuchter, C., Abbas, B., Coolen, M.J.L., Herfort, L., van Bleijswijk, J., Timmers, P. et

al. (2006) Archaeal nitrification in the ocean. Proceedings of the National Academy of

Sciences 103: 12317-12322.

Yool, A., Martin, A.P., Fernandez, C., and Clark, D.R. (2007) The significance of

nitrification for oceanic new production. Nature 447: 999-1002.

56

Chapter 3. Relative abundance and diversity of ammonia-oxidizing archaea and bacteria in the San Francisco Bay estuary Annika C. Mosier1 and Christopher A. Francis1


Published as: Mosier, A.C., and C.A. Francis. 2008. Relative abundance and diversity of ammonia-oxidizing archaea and bacteria in the San Francisco Bay estuary. Environmental Microbiology 10(11): 3002-3016.

57

Abstract

Ammonia oxidation in marine and estuarine sediments plays a pivotal role in

the cycling and removal of nitrogen. Recent reports have shown that the newly

discovered ammonia-oxidizing archaea can be both abundant and diverse in aquatic

and terrestrial ecosystems. In this study, we examined the abundance and diversity of

ammonia-oxidizing archaea (AOA) and betaproteobacteria (β-AOB) across

physicochemical gradients in San Francisco Bay—the largest estuary on the west coast

of the United States of America. In contrast to reports that AOA are far more

abundant than β-AOB in both terrestrial and marine systems, our quantitative PCR

estimates indicated that β-AOB amoA (encoding ammonia monooxygenase subunit A)

copy numbers were greater than AOA amoA in most of the estuary. AOA were only

more pervasive than β-AOB in the low-salinity northern region of the estuary. Both

AOA and β-AOB communities exhibited distinct spatial structure within the estuary.

AOA amoA sequences from the north part of the estuary formed a large and distinct

low-salinity phylogenetic group. The majority of the β-AOB sequences were closely

related to other marine/estuarine Nitrosomonas-like and Nitrosospira-like sequences.

Both ammonia-oxidizer community composition and abundance were strongly

correlated with salinity. Ammonia-oxidizing enrichment cultures contained AOA and

β-AOB amoA sequences with high similarity to environmental sequences. Overall,

this study significantly enhances our understanding of estuarine ammonia-oxidizing

microbial communities and highlights the environmental conditions and niches under

which different AOA and β-AOB phylotypes may thrive.

58

Introduction

Anthropogenic influence on the global nitrogen cycle has resulted in a suite of

environmental problems including global acidification, loss of biodiversity, increased

concentrations of the potent greenhouse gas N2O, and eutrophication of terrestrial and

aquatic systems (Vitousek et al., 1997; Gruber and Galloway, 2008). Eutrophication

is of particular concern in estuaries—over half of the estuaries in the United States of

America experience its effects (82 of 138 studied; Bricker et al., 1999). Harmful

levels of nitrogen (N) in estuaries can be diminished through tightly coupled processes

in the microbial nitrogen cycle; nitrification converts ammonia to nitrite and nitrate,

which can then be lost to the atmosphere as N2 gas via denitrification or anammox.

Coupled nitrification-denitrification removes up to 50% of external dissolved

inorganic nitrogen (DIN; ammonium, nitrate, and nitrite) inputs to estuaries

(Seitzinger et al., 2006), thereby reducing the risk of eutrophication. The fate of

excess nitrogen in estuaries is thus determined by the microbially-driven nitrogen

cycle and yet recent discoveries suggest that our understanding of these processes is

deficient (Francis et al., 2007).

Removal of DIN from aquatic systems is often directly dependent on the

supply of products from nitrification (Jensen et al., 1993; Jensen et al., 1994). It has

long been assumed that the most important contributors to ammonia oxidation—the

first and rate-limiting step in nitrification—are two groups of ammonia-oxidizing

bacteria (AOB): the betaproteobacteria (β-AOB) from the genus Nitrosomonas and

Nitrosospira (e.g., Head et al., 1993; Purkhold et al., 2000; Purkhold et al., 2003) and

gammaproteobacteria (γ-AOB) from the genus Nitrosococcus (e.g., Ward and

59

O'Mullan, 2002). Distinct groups of Nitrosospira-like and Nitrosomonas-like AOB

(uncultivated as of yet) are commonly detected in marine and estuarine systems (e.g.,

Francis et al., 2003; Bernhard et al., 2005; O'Mullan and Ward, 2005). Recently, a

combination of metagenomic and genomic analyses (Venter et al., 2004; Treusch et

al., 2005; Hallam et al., 2006b; Hallam et al., 2006a) and the cultivation of a novel

ammonia-oxidizing marine crenarchaeote (Könneke et al., 2005) revealed strong

evidence for nitrification within the domain Archaea, including identification of genes

putatively encoding archaeal ammonia monooxygenase subunits (amoA, amoB, and

amoC). Active gene expression of archaeal amoA has since been detected in

enrichment cultures (Hatzenpichler et al., 2008; Santoro et al., 2008), microcosms

(Treusch et al., 2005), and the natural environment (Treusch et al., 2005; Leininger et

al., 2006; Lam et al., 2007).

The abundance of open ocean Crenarchaeota (~1.0 x 1028 cells, equivalent to

20% of picoplankton in the global ocean; Karner et al., 2001) and the demonstration

that >80% of the archaeal metabolism in the deep ocean is autotrophic (Ingalls et al.,

2006) implies that AOA may be significant contributors to ammonia oxidation in

marine settings. Indeed, archaeal amoA genes are found in marine water columns and

sediments (Francis et al., 2005; Wuchter et al., 2006; Herfort et al., 2007; Lam et al.,

2007; Mincer et al., 2007; Nakagawa et al., 2007; Beman et al., 2008), estuaries

(Francis et al., 2005; Beman and Francis, 2006; Caffrey et al., 2007), and a coastal

aquifer (Francis et al., 2005; Santoro et al., 2008). Archaeal amoA gene copy numbers

are greater than bacterial amoA in the North Atlantic and North Sea (Wuchter et al.,

2006), Monterey Bay and Hawaii (Mincer et al., 2007), the Japan Sea (Nakagawa et

60

al., 2007), the Gulf of California (Beman et al., 2008) and several estuaries (Caffrey et

al., 2007). The dynamics of these two distinct ammonia-oxidizing groups (AOA and

AOB) are likely to be complex—as they compete for a common substrate and

ecological niche—and additional studies are needed to discern the specific physical

and geochemical conditions under which AOA or AOB are more abundant and/or

diverse.

Natural gradients of salinity, nutrients, dissolved oxygen, temperature,

turbidity and substrate, amongst others, make estuaries ideal systems in which to

address such questions. In the present study, we examined the abundance and

diversity of AOA and β-AOB (based on amoA genes) in sediments of San Francisco

Bay, the largest estuary on the west coast of the United States of America

(encompassing nearly 178,000 km2). San Francisco Bay’s historical resistance to the

harmful consequences of nutrient enrichment has recently weakened due to changes in

interdecadal oceanic regimes (Cloern et al., 2007), making microbial cycling of the

large standing stock of DIN even more important. We analyzed sediments from sites

spanning a range of physical and chemical conditions in the Bay allowing us to

examine spatial, geochemical and temporal (between two summers) effects on

ammonia-oxidizer community structure and abundance.

Results and Discussion

Environmental setting

San Francisco Bay effectively connects two individual estuaries to the Pacific

Ocean (Fig. 1). The northern portion of the bay—encompassing classic estuarine

61

gradients of salinity and nutrients—originates at the San Joaquin and Sacramento

River Delta, which drains approximately 40% of the area of the entire state of

California. The southern half of the bay is considered a weakly mixed lagoonal

system with wastewater treatment plant effluent as the primary source of freshwater

(Hager and Schemel, 1996). In this study, sampling sites covered four distinct

regions: North Bay (BG20, BG30, SU02, SU16, BF21), San Pablo Bay (BD31, SP19),

Central Bay (BC11, CB20, CB18), and South Bay (SB02, BA41, SB18, LSB2, BA10)

(Fig. 1). Physical and chemical conditions varied throughout the estuary but did not

change considerably between 2004 and 2005 (Table 1). Salinity in the North Bay did

not exceed 10 practical salinity units (psu) and was less than 0.6 psu at the riverine

sites (BG20 and BG30). Salinity in the rest of the estuary ranged from 22 to 31 psu.

Total C:N ratios were highest in the North Bay, likely due to allochthonous riverine

inputs (Jassby et al., 1993; Jassby and Cloern, 2000; Kennish, 2000), compared to

lower ratios in the Central and South Bay where autochthonous organic matter inputs

are typically more dominant (Jassby et al., 1993; Canuel et al., 1995; Kennish, 2000).

Bottom water nitrite and nitrate concentrations were highest at the southernmost sites

(2.1-3.7 µM nitrite and 34-88 µM nitrate at BA10 and LSB2) where there are long

residence times and high loading from wastewater treatment plants (Kimmerer, 2004;

Wankel et al., 2006).

AOA and β-AOB abundance

Abundance of archaeal and betaproteobacterial amoA genes was determined at

13 sites distributed throughout the Bay from August 2004 and 8 sites from August

62

2005 (Table S1). AOA amoA gene copy numbers ranged from 5.6 x 102 to 4.3 x 106

copies per µg DNA and from 1.4 x 104 to 3.9 x 107 copies per g of sediment (wet

weight). β-AOB amoA ranged from 9.1 x 103 to 4.2 x 106 copies per µg DNA and from

3.1 x 104 to 5.3 x 107 copies per g of sediment (wet weight). AOA amoA gene copy

numbers (normalized to sediment wet weight) were most abundant in the North Bay

(BG20, BG30, SU02, SU16, BF21) and then decreased nearly 70-fold in the rest of the

estuary (abundances averaged across regions compared). β-AOB amoA copy numbers

were lowest at BG20, BG30, and SU02 in the north and then increased 50-fold along

the rest of the estuarine transect. In San Francisco Bay, AOA amoA copy numbers

spanned a broader range than observed in coastal aquifer sediments (~105 copies per g

of sediment wet weight; Santoro et al., 2008), whereas β-AOB amoA copy numbers

were similar to those in salt marsh and coastal aquifer sediments (104 to 106 copies per

g sediment wet weight; Dollhopf et al., 2005; Santoro et al., 2008).

In contrast to numerous recent studies suggesting that AOA are more abundant

than AOB in marine and terrestrial ecosystems (Leininger et al., 2006; Wuchter et al.,

2006; Lam et al., 2007; Mincer et al., 2007; Nakagawa et al., 2007; Beman et al.,

2008), our quantitative PCR estimates indicated that β-AOB amoA copy numbers

were greater than AOA amoA in most of the San Francisco Bay estuary (Fig. 2). β-

AOB amoA were on average 215 times more abundant than AOA amoA at all sites

from San Pablo Bay to South Bay in both 2004 and 2005 (average log10 ratio of

AOA:β-AOB = -2.1). However, at BG20, BG30, and SU02 in the North Bay, AOA

amoA were 47 times more abundant than β-AOB amoA on average (average log10 ratio

of AOA:β-AOB = 1.5). The westernmost North Bay sites (SU16 and BF21) marked a

63

transition point where AOA and β-AOB amoA gene abundances were nearly identical

(average log10 ratio of AOA:β-AOB = 0.1; salinity = 6-9 psu). Following a similar

pattern to that described here, coastal aquifer sediment β-AOB were more abundant

than AOA (up to 30-fold) at high salinities but less abundant than AOA at low

salinities (Santoro et al., 2008). This is particularly striking given the fact that the

coastal aquifer shift in AOA/AOB abundance occurred over just tens of meters,

whereas the San Francisco Bay transition occurred over tens of kilometers. Caffrey et

al. (2007) also found that β-AOB were more abundant than AOA at many of the

estuarine sites examined, particularly in the Weeks Bay estuary.

AOA and β-AOB richness and phylogenetic diversity

Archaeal and betaproteobacterial amoA sequences were recovered from 13

sites from August 2004 and 5 sites from August 2005 chosen as representatives of the

main regions in the estuary. From all environmental clone libraries combined, we

recovered 67 unique archaeal amoA operational taxonomic units (OTUs) and 41

unique bacterial amoA OTUs defined at a 5% distance cutoff. Within each individual

site, between 4-11 AOA OTUs and 3-11 β-AOB OTUs were observed (Table S2). In

some cases, these numbers may have been higher if more clones were sequenced,

based on non-asymptotic rarefaction curves (Fig. S1). β-AOB richness estimates for

libraries from a given site (4-15 predicted OTUs by Chao1; Table S2) were within the

same range reported in other estuaries (Francis et al., 2003; Beman and Francis, 2006)

and, similar to Chesapeake Bay, were highest at sites in the low-salinity North Bay

region. Overall, archaeal amoA Chao1 richness estimates (4-46 predicted OTUs;

64

Table S2) were consistently higher than bacterial amoA, with three exceptions

(4BG20, 4SU02, 5BG30; 4BA10 had equal AOA and β-AOB estimates). On average,

the number of observed AOA OTUs at a site represented 71% of the total number of

AOA OTUs predicted (Chao1 estimates) to occur at that site. For β-AOB, 86% of the

predicted OTUs were observed. The highest β-AOB richness and lowest AOA

richness occurred at North Bay sites in both 2004 (site SU02) and 2005 (site BG30).

Archaeal amoA Shannon diversity was higher than bacterial amoA at 12 of the 18

sites. AOA and β-AOB Shannon diversity did not change greatly between 2004 and

2005 (average difference ≤0.5).

Phylogenetic analysis of archaeal amoA sequences showed that a large

proportion (174 out of 415) fell into a cluster comprised almost entirely of

coastal/estuarine sediment sequences from Northern California, along with the amoA

sequence of Nitrosopumilus maritimus (Könneke et al., 2005) (Fig. 3, Fig. S2).

Approximately 15% of the 415 archaeal amoA sequences from San Francisco Bay

sediments fell into the soil/sediment group (Francis et al., 2005), which includes the

soil metagenomic (fosmid) clone 54d9 (Treusch et al., 2005) derived from

uncultivated Group 1.1b Crenarchaeota (the most ubiquitous lineage of Crenarchaeota

in terrestrial environments; Ochsenreiter et al., 2003). Interestingly, none of the

5BG30, 4BG20, 4BG30, 4SU02, or 4CB18 sequences fell into the soil/sediment

group.

Archaeal amoA sequences from the North Bay (BG20, BG30, SU02, and

SU16) formed a large and distinct low-salinity cluster (Fig. 3). Some sequences from

San Pablo Bay also fell into this group (1 from 4SP19, 3 from 4BD31, and 13 from

65

5BD31); despite having relatively high salinities at the time of sampling, salinity in

this region is often oligo- or mesohaline in the winter and spring due to high flow from

the San Joaquin and Sacramento rivers (http://sfbay.wr.usgs.gov/access/wqdata). This

may imply that these putative low-salinity AOA phylotypes are tolerant of higher

salinities. GenBank sequences within this clade were from the rhizosphere of

freshwater macrophytes (Herrmann et al., 2008), coastal aquifer sediments at

freshwater and freshwater-saltwater interface sites (Santoro et al., 2008), and a

wastewater treatment plant (WWTP) in McMinnville, OR (Park et al., 2006).

Sequences from the sandy flat of the Douro River estuary (Magalhaes et al.,

unpublished; direct GenBank submission) grouped closely with San Francisco Bay

sequences. The entire Douro River estuary has been shown to become river-like

during high flow (Vieira and Bordalo, 2000), and mesophilic Crenarchaeota 16S

rRNA sequences have been previously reported from brackish (1 to 5 psu) intertidal

sediments from this estuary (Abreu et al., 2001). Two sequences from South San

Francisco Bay (SF04-LSB002S-H02 and one sequence from low-salinity sediments

near the Palo Alto WWTP; Park et al., 2006) were also found in this cluster.

Bacterial amoA sequences from San Francisco Bay showed substantial overlap

with Chesapeake Bay, Plum Island Sound, and Bahía del Tóbari estuaries (Francis et

al., 2003; Bernhard et al., 2005; Beman and Francis, 2006) (Fig. 4, Fig. S3, and Fig.

S4). A majority of the bacterial amoA sequences fell into two distinct groups with

other estuarine sequences in the ‘estuarine/marine Nitrosospira-like’ and

‘estuarine/marine Nitrosomonas-like’ clusters. Sequences from BG20 and BG30 in

the North Bay fell almost exclusively into a ‘low salinity Nitrosomonas ureae- and N.

66

oligotropha-like’ group along with sequences from low salinity sites in Plum Island

Sound and Chesapeake Bay (Francis et al., 2003; Bernhard et al., 2005). Studies in

the Chesapeake Bay, Plum Island Sound, and Ythan (Scotland) estuaries reported

Nitrosospira-like amoA sequences at high salinities and Nitrosomonas-like sequences

at low salinities (Francis et al., 2003; Bernhard et al., 2005; Freitag et al., 2006). In

contrast to the predominantly Nitrosomonas-dominated freshwater sites in San

Francisco Bay (BG20 and BG30), sequences from the other North Bay sites (SU02

and SU16 with salinities of 2.9 and 6.3 psu, respectively) fell almost exclusively (38

of 42 sequences) into the estuarine/marine Nitrosospira-like group, along with

sequences from sites with salinities ≥22.0 psu. In the rest of San Francisco Bay (San

Pablo Bay to South Bay with salinities ≥22.0 psu), sequences were distributed

between both the estuarine/marine Nitrosospira-like (219 sequences) and

estuarine/marine Nitrosomonas-like (117 sequences) groups.

AOA and β-AOB community structure

There was significant compositional overlap between the 2004 and 2005 clone

libraries for both archaeal and bacterial amoA (Table 2), based on abundance-based

Sørensen’s indices of similarity (Labd) greater than 0.5 (i.e., for OTUs found in at least

one of the communities, there is a 50% probability that a randomly selected OTU is

found in both communities; Schloss and Handelsman, 2006). Likewise, at four out of

five sites in the Bahía del Tóbari estuary, AOA communities sampled 9 months apart

were virtually indistinguishable (Beman and Francis, 2006). β-AOB at a high-salinity

site in the Plum Island Sound estuary showed little variability over 3 years, while mid-

67

and low-salinity sites were less consistent among sampling dates (Bernhard et al.,

2005). Although there was little variation in AOA and β-AOB community

composition in San Francisco Bay between two summers, seasonal influence

(particularly higher precipitation and lower temperature in the winter months) on

diversity and community structure remain uncharacterized.

Ammonia oxidizers often exhibit strong spatial structure in estuaries. In the

Bahía del Tóbari estuary, AOA communities differ between the interior and mouths of

the bay (Beman and Francis, 2006). Studies in the Chesapeake Bay, Plum Island

Sound, and Ythan (Scotland) estuaries have shown shifts in AOB communities along

salinity gradients (Francis et al., 2003; Bernhard et al., 2005; Freitag et al., 2006).

Distinct spatial patterns were also evident in the San Francisco Bay estuary (Table 2).

For archaeal amoA, there was little compositional overlap between the North Bay and

the Central and South Bays (Labd < 0.5). All sites within the North Bay (SU16, SU02,

BG30, and BG20) had high similarity with each other but not with any other sites in

the estuary and, conversely, Central and South Bay sites (CB20, CB18, SB02, SB18,

LSB2, and BA10) overlapped with each other but not with North Bay sites. In 2004,

the central-most site with a direct connection to the ocean (4BC11) only exhibited

significant overlap with one other site (4BD31). In 2004, the San Pablo Bay sites

(BD31 and SP19) showed high similarity to Central and South Bay sites and low

similarity to the North Bay sites. However, in 2005, the San Pablo Bay sites were

much more compositionally similar to the North Bay sites. For bacterial amoA, the

two northernmost river sites (BG20 and BG30) had no significant overlap with any

other sites in the estuary. 4BG30 had low similarity (Labd < 0.18) to all other sites

68

except the same site sampled in 2005 (Labd = 0.72). There was high similarity between

the other North Bay sites (SU02, SU16) and the San Pablo Bay sites (BD31, and

SP19). High compositional overlap occurred amongst sites from San Pablo, Central

and South Bays.

Relationship between environmental variables and ammonia-oxidizer community

structure and abundance

Many physical and geochemical factors influence nitrification, including

oxygen concentrations (e.g., Christensen et al., 1990), sulfide concentrations (Joye and

Hollibaugh, 1995), Fe(III) content (Dollhopf et al., 2005), salinity (Rysgaard et al.,

1999; Caffrey et al., 2003; Bernhard et al., 2007), C:N ratios (Ballinger et al., 2002),

availability of ammonium (e.g., Jones and Hood, 1980), light (Horrigan and Springer,

1990), and temperature (Berounsky and Nixon, 1993). In estuarine sediments, salinity

plays a particularly significant role in shaping β-AOB diversity and population

structure (Francis et al., 2003; Bernhard et al., 2005). In our study, Canonical

Correspondence Analysis (CCA) showed that salinity and temperature had the greatest

correlation with archaeal amoA community composition of the 27 factors tested, while

the bacterial amoA community composition was most correlated with C:N ratio,

salinity, and nickel concentrations (Fig. 5). In both cases, the North Bay sites grouped

independently from all other regions.

Because microbial nitrogen cycling processes involve metalloenzymes, trace

metal availability may be a key factor regulating nitrogen transformations (Morel and

Price, 2003). For example, copper limitation decreases the activity of the copper-

69

containing enzyme nitrous oxide reductase (NosZ) in denitrifiers (Granger and Ward,

2003). Likewise, nickel (Ni) incorporation is required for microbial synthesis of

active urease enzymes (Mobley et al., 1995). Many AOB, including various

Nitrosomonas, Nitrosospira, and Nitrosococcus species, have Ni-containing urease

enzymes and are capable of using urea for growth (Koper et al., 2004). The

correlation between β-AOB community composition and sediment nickel

concentrations in San Francisco Bay is thus intriguing and warrants further

exploration.

Environmental factors also influence the abundance of ammonia oxidizers in

estuaries. In San Francisco Bay, the log ratio of AOA:β-AOB amoA copy numbers

was strongly correlated to bottom water salinity (Spearman correlation: r = -0.85, P =

0.00, n = 20). When analyzed individually, AOA amoA abundance (gene copies per g

sediment) was negatively correlated with salinity (Spearman correlation: r = -0.79, P =

0.00, n = 21) and β-AOB amoA abundance was positively correlated with salinity

(Spearman correlation: r = 0.77, P = 0.00, n = 20). At sites where AOA amoA were

more abundant than β-AOB amoA, the bottom water salinity averaged 2 psu (range =

0.2-9 psu), whereas when β-AOB amoA were more abundant than AOA amoA the

average salinity was 27 (range = 22-31 psu). The transition sites where AOA and β-

AOB amoA abundances were similar had intermediate salinity values.

Salinity is a particularly important parameter for ammonia oxidation in

estuaries, in part because of its influence on NH4+ adsorption in sediments (Boatman

and Murray, 1982; Boynton and Kemp, 1985). As was found in this study, salinity

was strongly correlated to the ratio of AOA and β-AOB in coastal aquifer sediments

70

(Santoro et al., 2008), and AOB abundance in Plum Island Sound estuary sediments

was lowest at low salinity (Bernhard et al., 2007). However, in contrast to these

studies, abundance of AOA was positively correlated with salinity across six different

estuaries, while AOB abundance showed no correlation with salinity (Caffrey et al.,

2007). It is unclear whether this discrepancy is due to variation in physical and

chemical conditions in the estuaries studied or to methodological differences.

The abundance of AOA and β-AOB amoA and the log ratio of AOA:β-AOB

were all strongly correlated with C:N ratios (Spearman correlation > ±0.7; Table S3).

However, because C:N ratios can covary with salinity (e.g., riverine inputs in the

North Bay likely cause both high C:N ratios and low salinity), partial correlation

analyses were used to separate the individual effect of these variables (Table S4).

When controlling for C:N ratios, AOA amoA abundance and the log ratio of AOA:β-

AOB were still significantly and strongly correlated to salinity. Correlation between

salinity and β-AOB amoA abundance decreased when controlling for C:N ratios and

yet there was no correlation between C:N ratios and abundance when controlling for

salinity.

The abundance of AOA amoA was also strongly correlated with lead (Pb)

concentrations and percent clay (Table S3). Partial correlation between salinity, lead,

and clay (Table S4) makes it difficult to establish a direct link between these variables

and AOA amoA abundance. It is possible that other parameters not measured here

also influence AOA and β-AOB, including sediment macrofauna (e.g., bioturbation),

porewater sulfide, porewater NH4+ and porewater dissolved oxygen concentrations.

Even so, as has been shown previously for several other estuaries (e.g., Francis et al.,

71

2003; Bernhard et al., 2005; Bernhard et al., 2007; Santoro et al., 2008), salinity

appears to play a central role in shaping the structure of the ammonia-oxidizer

community (both in terms of composition and abundance) in San Francisco Bay.

AOA and β-AOB enrichments

The vast majority of amoA gene sequences reported in most cultivation-

independent estuarine studies appear to be novel and distinct from sequences of

cultivated ammonia oxidizers, indicating that we know little about the specific

organisms actually responsible for this process in nature. Ammonia-oxidizing

enrichment cultures were initiated from 5 San Francisco Bay sites that spanned the

estuarine gradient and overlapped with our environmental samples from 2004 and

2005. After 6 months, archaeal amoA was successfully amplified from all 5

enrichments, while bacterial amoA only amplified from 4 enrichments (excluding the

freshwater site BG30). Archaeal amoA sequences from all 5 enrichments showed

significant compositional overlap with environmental sequences from sediments at the

same sites (>90% of sequences belong to shared OTUs with the 2004 and/or 2005

environmental sequences from the same site; also highlighted in the phylogenetic trees

in Fig. 3 and Fig. S2), suggesting that these enriched AOA may be representative of

estuarine Crenarchaeota.

To date there are large groups of β-AOB with no cultivated representatives,

including the 16S rRNA-based Nitrosospira cluster 1 and Nitrosomonas cluster 5

(Freitag and Prosser, 2003; Freitag et al., 2006) found to dominate AOB communities

in many marine and estuarine systems (e.g., McCaig et al., 1999; Bano and

72

Hollibaugh, 2000; Freitag and Prosser, 2003, 2004). The presumption that these

groups oxidize ammonia is based on clustering of their 16S rRNA genes within a

monophyletic group containing cultivated AOB, and on reports of non-persisting

enrichment cultures of Nitrosomonas cluster 5 (Stephen et al., 1996). We found

bacterial amoA sequences in our enrichment cultures from both the estuarine/marine

Nitrosospira-like and Nitrosomonas-like groups (Fig. 4, Fig. S3 and Fig. S4)

commonly detected in estuarine systems (Francis et al., 2003; Bernhard et al., 2005).

Additional efforts, including continued cultivation over a longer time period, will be

necessary to definitively confirm ammonia oxidation by these AOB groups.

Conclusions

In summary, we have provided evidence for niche specialization of AOA and

β-AOB in estuary sediments from San Francisco Bay. Both groups exhibit distinct

spatial structure and little temporal variability between two summers. AOA and β-

AOB community composition and abundance are highly correlated with salinity. β-

AOB community composition is also correlated with sediment nickel concentrations,

which is intriguing given the nickel requirement for active urease enzymes. β-AOB

amoA genes are far more abundant than AOA amoA genes in much of the estuary

where salinity is high and C:N ratios are low. In contrast, the region of the estuary

with low-salinity and high C:N ratios contains a group of AOA that are both abundant

and phylogenetically distinct amongst San Francisco Bay ammonia oxidizers, forming

a low-salinity amoA gene cluster. β-AOB amoA sequences at the riverine sites (BG20

and BG30) are primarily Nitrosomonas-like, whereas estuarine/marine Nitrosospira-

73

like and Nitrosomonas-like amoA sequences predominate the rest of San Francisco

Bay.

Experimental Procedures

Sample collection

Sediment and water samples were collected from San Francisco Bay in

northern California during the summers of 2004-2006 in conjunction with the

Regional Monitoring Program (RMP) managed by the San Francisco Estuary Institute

(http://www.sfei.org/). Annual cruises took place aboard the R/V Endeavor (operated

by the Bureau of Land Management). Sediment cores were collected from a Young-

modified, Van Veen grab using sterile, cut-off 5-cc syringes and stored on dry ice until

permanent storage at –80ºC. Bottom water samples for nutrient analyses were

immediately 0.2 µm-filtered (Nalgene PES or Millipore MillexGV) and frozen at –

20ºC. In 2006, sediment samples were collected for ammonia-oxidizing enrichment

cultures and stored at 4ºC until inoculation.

Environmental parameters

Water column profiles of temperature, salinity and dissolved oxygen (DO)

were measured at each site using a SeaBird™ SBE-19 CTD. Bottom water samples

from 2004 were previously analyzed for nitrate (NO3-), nitrite (NO2

-), and ammonium

(NH4+) (Wankel et al., 2006). In 2005, nutrients were measured on a QuikChem 8000

Flow Injection Analyzer (Lachat Instruments). The RMP analyzed sediment samples

for pH, trace elements (Ag, Al, As, Cd, Cu, Fe, Hg, MeHg, Mn, Ni, Pb, Se and Zn),

grain size, total organic carbon and total nitrogen. C:N ratios were calculated using

74

total organic carbon and total nitrogen percentages.

Quantification of amoA gene abundance

Quantitative PCR (Q-PCR) was used to estimate the number of archaeal and

bacterial amoA copies in sediments at 13 sites in 2004 (4BG20, 4BG30, 4SU02,

4SU16, 4BD31, 4SP19, 4BC11, 4CB20, 4CB18, 4SB02, 4SB18, 4LSB2, 4BA10) and

8 sites in 2005 (5BG20, 5BG30, 5SU02, 5BF21, 5BD31, 5BC11, 5BA41, and

5BA10). The number preceding the sampling site name indicates the year sampled

(e.g., 4BA10 and 5BA10 correspond to site BA10 sampled in 2004 and 2005,

respectively). DNA was extracted from approximately 250-300 mg of sediment from

the upper 5 mm of the cores using the FastDNA SPIN kit for soil (MP Biomedicals,

Inc.) with a 30 second bead beating time. Duplicate DNA extractions were performed

for each sediment core. Reactions were carried out using a StepOnePlus™ Real-Time

PCR System (Applied Biosystems). Primers Arch-amoAF and Arch-amoAR (Francis

et al., 2005) were used for archaeal amoA quantification and amoA-1F and amoA-2R

(Rotthauwe et al., 1997) for bacterial amoA quantification. Plasmids containing

cloned amoA PCR amplicons were extracted with the Qiagen Miniprep Spin Kit for

use in standard curves: archaeal amoA clone SF06E-BC11-A02 and bacterial amoA

clone SF06E-BA10-C02. Sediment and standard DNA was quantified using the

Quant-iT™ High-Sensitivity and Broad Range DNA assays, respectively, with the

Qubit Fluorometer (Invitrogen). Standard curves spanned a range from 11 to 1.07 x

105 amoA copies per µl for the AOA assay and from 16 to 1.62 x 105 amoA copies per

µl for the β-AOB assay. All sample and standard reactions were performed in

triplicate and an average value was calculated. An outlier was removed from some

75

triplicate measurements to maintain a standard deviation of less than 15% for each

sample. Melt curves were generated after each assay to check the specificity of

amplification. PCR efficiencies were 90-99% (average 94%) for archaeal amoA and

76-94% (average 86%) for bacterial amoA. Correlation coefficients (R2) for both

assays averaged 0.99 (standard deviation of 0.003). The ratio of AOA amoA to β-AOB

amoA was calculated using copy numbers normalized to µg DNA and then log10

transformed.

Archaeal amoA Q-PCR was performed in a 25 µl reaction mixture with 0.2-63

ng template DNA, 0.4 µM of each primer, 2 mM MgCl2, 1.25 Units AmpliTaq® DNA

polymerase (Applied Biosystems), 1 µl passive reference dye, 40 ng µl-1 BSA, and

12.5 µl Failsafe Green Premix E (Epicentre). The PCR protocol was as follows: 3 min

at 95°C, then 32 cycles consisting of 30 s at 95°C, 45 s at 56°C, and 1 min at 72°C

with a detection step at the end of each cycle.

Bacterial amoA Q-PCR was performed in a 20 µl reaction mixture with 0.2-23

ng template DNA, 0.3 µM of each primer, 1 Unit AmpliTaq® DNA polymerase

(Applied Biosystems), 0.8 µl passive reference dye, and 10 µl Failsafe Green Premix

F (Epicentre). The PCR protocol was as follows: 5 min at 95°C, then 34 cycles

consisting of 45 s at 94°C, 30 s at 56°C, 1 min at 72°C, and a detection step for 10 s at

80.5°C.

Ammonia-oxidizing community composition

AOA and β-AOB amoA diversity was analyzed at 18 of the sites used for

qPCR: 4BG20, 4BG30, 4SU02, 4SU16, 4BD31, 4SP19, 4BC11, 4CB20, 4CB18,

76

4SB02, 4SB18, 4LSB2, 4BA10, 5BG30, 5BD31, 5BC11, 5BA41, and 5BA10. Total

community DNA was extracted as described above. Sediment DNA extracts were

PCR amplified with primers targeting archaeal and bacterial amoA: Arch-amoAF and

Arch-amoAR (Francis et al., 2005) for AOA; and amoA-1F* (Stephen et al., 1999)

and amoA-2R (Rotthauwe et al., 1997) for β-AOB. Clone libraries of amoA gene

fragments were constructed using the TOPO TA cloning® kit (Invitrogen, Inc.) and 17

to 31 clones from each library were sequenced (ABI 3100 Capillary Sequencer).

Nucleotide sequences were aligned with GenBank sequences using MacClade version

4.08 (Maddison and Maddison, 2000). A 582-bp region of the sequence alignment

was used for archaeal amoA richness and diversity analyses and a 579-bp region was

used for phylogenetic analyses. For bacterial amoA, a 444-bp region was used for

richness, diversity and phylogenetic analyses.

Jukes-Cantor distance matrices were generated in PAUP 4.0b10 (Sinauer

Associates). Operational Taxanomic Units (OTUs), diversity indices (Shannon and

Simpson), and nonparametric richness estimates (Chao1 and ACE; Chao, 1984; Chao

and Lee, 1992; Chao et al., 1993) were determined using the furthest neighbor

algorithm in DOTUR (Schloss and Handelsman, 2005). Abundance-based Sørensen-

type similarity indices (Labd) (Chao et al., 2005, 2006) were calculated using SONS

with 10,000 iterations (Schloss and Handelsman, 2006). Labd values were used to

estimate community composition similarity between libraries (Labd standard error for

all AOA and β-AOB comparisons was < 0.32). For this analysis, site SB18 from 2004

and site BA41 from 2005 were compared because of their close physical proximity to

each other (2.7 kilometers apart). OTUs were defined at a ≤5% nucleic acid sequence

77

difference. Neighbor joining phylogenetic trees (based on Jukes-Cantor correction)

were constructed in ARB (Ludwig et al., 2004). Distance and parsimony bootstrap

analyses (100 replicates) were performed using PAUP version 4.0b10 (Swofford,

2003).

Statistical Analyses

Correlation between amoA abundance and environmental variables was

calculated using SPSS, version 11.0.4. Non-parametric Spearman correlation and

partial correlation analyses were used for AOA amoA copies per g sediment, β-AOB

amoA copies per g sediment, and the log10 ratio of AOA:β-AOB. Bottom water

parameters included in the statistical analyses were: salinity (psu), temperature (ºC),

dissolved oxygen (mg/L), nitrite+nitrate (µM), nitrite (µM), nitrate (µM), ammonium

(µM), DIN (µM), water depth (m), and pH. Sediment parameters included in the

statistical analyses were: TOC (%), TN (%), C:N ratio, Ag (mg/Kg), Al (mg/Kg), As

(mg/Kg), Cd (mg/Kg), Cu (mg/Kg), Fe (mg/Kg), Hg (mg/Kg), MeHg (ug/Kg), Mn

(mg/Kg), Ni (mg/Kg), Pb (mg/Kg), Se (mg/Kg), Zn (mg/Kg), % Clay, % Fine, %

Granule and Pebble, % Sand, and % Silt.

Correlations between amoA community composition and environmental

parameters were analyzed by Canonical Correspondence Analysis (CCA) using the

program Canoco (ter Braak and Smilauer, 2002). CCA was chosen to determine if the

community structure is more strongly related to specific environmental variables than

expected by chance. Detrended CCA showed that a unimodal response model better

approximated both AOA and β-AOB species relationship to the explanatory variables

(maximum gradient length exceeds 4 S.D.). Relative abundance of sequences in each

78

OTU (defined at 0.05 distance) was used as species input (i.e., the number of

sequences in an OTU divided by the total number of sequences in the library).

Significant environmental variables were determined by forward selection (P < 0.05

for AOA; P ≤ 0.06 for β-AOB). Bottom water parameters included in the forward

selection were: salinity (psu), temperature (ºC), dissolved oxygen (mg/L),

nitrite+nitrate (µM), nitrite (µM), nitrate (µM), ammonium (µM), DIN (µM), water

depth (m), and pH. Sediment parameters included in the forward selection were: TOC

(%), TN (%), C:N ratio, As (mg/Kg), Cd (mg/Kg), Cu (mg/Kg), Hg (mg/Kg), Mn

(mg/Kg), Ni (mg/Kg), Pb (mg/Kg), Se (mg/Kg), Zn (mg/Kg), % Clay, % Fine, %

Granule and Pebble, % Sand, and % Silt. Analyses used non-transformed data and

biplot scaling.

LC scores (linear combinations of the environmental variables) were used as

points in the CCA ordination plots. In the plots, the length of the environmental line

corresponds to the strength of the relationship of that variable with the community

composition, whereas the perpendicular position of the sampling site points relative to

the environmental line is an approximate ranking of species response curves to that

environmental variable (McCune and Grace, 2002). The first canonical axis and all of

the canonical axes taken together were significant (P ≤ 0.002) based on Monte Carlo

tests (499 permutations under reduced model). The variance of the species-

environment relation represented in each ordination was: axis 1 = 70.2% and axis 2 =

29.8% for AOA; axis 1 = 49.8% and axis 2 = 32.5% for β-AOB. The variance of the

species data represented in each ordination was: axis 1 = 19.0% and axis 2 = 8.1% for

AOA; axis 1 = 20.4% and axis 2 = 13.3% for β-AOB.

79

Cultivation of estuarine ammonia-oxidizers

Ammonia-oxidizer enrichment cultures were initiated in 2006 from 5 sites

overlapping with the samples collected in 2004 and 2005 (BG30, BD31, BC11, BA41,

and BA10). Surface sediments were inoculated into a modified version of Synthetic

Crenarchaeota Medium (Könneke et al., 2005) made with 500 µM ammonium

chloride, 1 mL selenite/tungstate solution, 1 mL vitamin solution (Balch et al., 1979),

1 mL sodium bicarbonate (1M), 10 mL KH2PO4 (4g g L-1), 1 mL trace metals (Biebl

and Pfennig, 1978), and 988 mL artificial seawater (35.1 g L-1 sodium chloride, 24.7 g

L-1 magnesium sulfate, 2.9 g L-1 calcium chloride, and 1.5 g L-1 potassium chloride)

diluted to 10 or 32 psu salinity. Cultures were grown in static culture flasks or serum

vials in the dark at room temperature (approximately 22°C). Approximately 10% of

the culture volume was transferred into fresh media 2-3 times over 6 months (BA41

was only transferred once following original transfer from a sediment slurry to media),

at which point DNA was extracted from enrichment cultures as previously described

(Santoro and Boehm, 2007; Santoro et al., 2008). Archaeal and bacterial amoA genes

were amplified, cloned, sequenced and phylogenetically analyzed as described above.

Compositional overlap with sequences from the environmental amoA libraries from

the same site was determined using the Vobs statistic in SONS at 0.05 distance, which

represents the fraction of sequences found in shared OTUs (Schloss and Handelsman,

2006).

Nucleotide sequence accession numbers

Sequences reported in this study have been deposited in GenBank under

accession numbers EU650799 to EU651305 (archaeal amoA) and EU651306 to

80

EU651796 (bacterial amoA). Sequence names in GenBank are referred to as SF04 and

SF05 for sampling years 2004 and 2005 and SF06E for enrichment cultures. For sites

with longer names, the original site name was used in GenBank (SU002S abbreviated

here as SU02, SU016S = SU16, SPB019S = SP19, CB020S = CB20, CB018S =

CB18, SB002S = SB02, SB018S = SB18, LSB002S = LSB2).

Acknowledgements

We thank Paul Salop, Sarah Lowe, Applied Marine Sciences, the San

Francisco Estuary Institute, and the crew of the R.V. Endeavor for allowing us to

participate in the San Francisco Bay RMP sediment cruises, providing access to the

RMP data, and assisting in sample collection. We thank Scott Wankel for sampling

and performing nutrient analyses from the 2004 cruise. We greatly appreciate those

who provided advice and comments on the manuscript (A. Santoro, G. Wells, M.

Beman, J. Smith, S. Ying, and J. Lee). This work was supported in part by National

Science Foundation Grants MCB-0433804 and MCB-0604270 (to C.A.F.) and an

EPA STAR Graduate Fellowship (to A.C.M.).

81

Tables

SiteDepth

(m)

Salinity

(psu)

Temp.

(ºC)

DO

(mg/L)

Nitrite

(!M)

Nitrate

(!M)

Ammonium

(!M)C:N* pH*

4BG20 9.0 0.4 21.7 6.7 0.71 16.33 4.25 12.0 6.9

4BG30 10.2 0.4 22.5 7.0 0.48 17.84 2.50 13.0 6.6

4SU02 11.7 2.9 21.7 8.0 0.52 22.22 3.65 11.9 7.3

4SU16 2.4 6.3 21.6 8.0 0.68 29.43 3.93 11.0 7.0

4BD31 6.0 25.2 19.5 7.0 0.79 26.46 4.44 9.0 6.6

4SP19 2.6 25.4 19.8 7.0 0.77 26.66 4.68 8.2 6.7

4BC11 5.8 30.4 19.0 6.7 0.87 19.92 12.27 7.2 6.8

4CB20 6.5 30.5 19.9 6.6 1.04 21.82 7.13 7.9 6.9

4CB18 11.0 30.3 21.0 6.4 1.08 23.64 4.94 8.1 7.0

4SB02 2.5 30.8 20.2 6.4 0.13 1.19 2.98 7.9 6.9

4SB18 3.1 29.7 21.3 6.0 0.33 4.85 1.70 7.8 6.9

4LSB2 6.7 26.3 20.4 5.2 2.16 49.44 9.17 8.2 7.2

4BA10 1.9 22.0 20.9 6.1 3.70 87.59 12.05 7.9 7.1

5BG20 8.9 0.2 21.3 8.8 1.80 21.12 6.50 24.3 7.0

5BG30 1.1 0.6 21.5 8.8 0.53 16.40 2.38 24.7 6.7

5SU02 11.5 9.0 20.2 8.6 1.16 24.73 2.73 -- 7.1

5BF21 2.6 9.1 20.5 8.5 0.88 20.01 1.84 14.1 7.0

5BD31 6.7 25.0 19.0 8.0 0.52 16.86 2.86 9.3 7.0

5BC11 6.9 30.3 17.3 8.0 0.49 12.06 14.94 8.6 7.0

5BA41 2.5 27.0 20.6 7.6 0.60 4.56 9.36 9.4 6.8

5BA10 3.9 24.2 21.0 7.7 2.11 33.97 12.93 9.5 7.3*measured in sediments (all others from bottom water).

Table 1. Physical and chemical properties of San Francisco Bay bottom water and sediments. Sampling

sites are presented by year (2004 followed by 2005) and ordered by location within the estuary (from north to

south).

82

4B

G2

04

BG

30

4S

U0

24

SU

16

4B

D3

14

SP

19

4B

C1

14

CB

20

4C

B1

84

SB

02

4S

B1

84

LS

B2

4B

A1

05

BG

30

5B

D3

15

BC

11

5B

A4

15

BA

10

4B

G2

00

.17

0.2

00

.09

0.0

90

.09

0.0

00

.07

0.0

00

.00

0.0

00

.00

0.0

00

.49

0.0

90

.07

0.0

00

.00

4B

G3

00

.93

0.1

10

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.72

0.0

00

.00

0.0

00

.00

4S

U0

20

.95

0.8

60

.91

0.8

00

.88

0.0

90

.16

0.0

00

.07

0.0

70

.00

0.0

90

.18

0.9

30

.38

0.0

70

.08

4S

U1

60

.86

0.6

30

.66

0.6

80

.78

0.0

00

.17

0.0

00

.00

0.0

60

.00

0.0

60

.06

0.8

40

.18

0.0

60

.00

4B

D3

10

.13

0.1

70

.18

0.4

50

.95

0.3

80

.61

0.3

00

.49

0.4

20

.52

0.5

00

.06

0.9

00

.81

0.3

50

.46

4S

P1

90

.07

0.0

80

.08

0.4

50

.65

0.3

10

.79

0.1

70

.55

0.4

80

.48

0.5

00

.06

0.9

71

.00

0.2

50

.49

4B

C1

10

.00

0.0

00

.00

0.1

60

.54

0.2

50

.61

0.7

90

.98

0.7

80

.76

0.7

00

.00

0.2

90

.85

0.4

60

.71

4C

B2

00

.00

0.0

00

.00

0.0

70

.85

0.6

00

.37

0.7

00

.79

0.6

00

.87

0.5

60

.00

0.4

70

.76

0.5

70

.70

4C

B1

80

.00

0.0

00

.00

0.0

00

.53

0.4

80

.27

0.8

10

.95

0.8

00

.83

0.6

10

.00

0.0

60

.83

0.6

40

.67

4S

B0

20

.00

0.0

00

.00

0.0

00

.48

0.5

70

.26

0.8

10

.75

1.0

00

.97

0.8

90

.00

0.4

41

.00

0.8

00

.98

4S

B1

80

.00

0.0

00

.00

0.0

00

.53

0.3

70

.43

0.8

50

.71

0.6

60

.87

0.8

60

.00

0.4

00

.95

0.6

30

.76

4L

SB

20

.09

0.1

00

.10

0.2

40

.54

0.6

10

.29

0.6

60

.70

0.6

80

.75

0.9

90

.00

0.1

71

.00

0.6

20

.99

4B

A1

00

.00

0.0

00

.00

0.0

60

.70

0.5

70

.42

0.9

00

.80

0.8

40

.72

0.6

60

.00

0.4

30

.85

0.6

40

.92

5B

G3

00

.60

0.7

20

.63

0.6

70

.18

0.0

80

.00

0.0

00

.00

0.0

00

.00

0.1

00

.00

0.0

60

.00

0.0

00

.00

5B

D3

10

.65

0.5

40

.74

0.8

70

.69

0.5

90

.42

0.2

90

.44

0.4

00

.24

0.4

60

.42

0.5

31

.00

0.1

30

.36

5B

C1

10

.00

0.0

00

.00

0.0

70

.79

0.4

70

.60

0.5

50

.24

0.4

20

.27

0.5

10

.36

0.0

00

.34

0.7

30

.82

5B

A4

10

.00

0.0

00

.00

0.3

00

.49

0.3

50

.72

0.5

30

.50

0.4

90

.52

0.6

00

.64

0.0

00

.43

0.3

80

.57

5B

A1

00

.00

0.0

00

.00

0.3

00

.50

0.7

10

.60

0.7

00

.30

0.3

60

.30

1.0

00

.64

0.0

00

.50

0.9

60

.68

lib

rary

Ta

ble

2.

Ab

un

dan

ce-b

ased

Sø

ren

sen

-ty

pe

(La

bd)

sim

ilar

itie

s am

on

g a

rch

aeal

an

d b

acte

rial

am

oA

clo

ne

lib

rari

es.

La

bd

val

ues

gre

ater

th

an o

r eq

ual

to

0.5

are

in

bo

ld a

nd

sh

aded

. L

ibra

ries

are

pre

sen

ted

by

yea

r (2

00

4 a

nd

20

05

) an

d o

rder

ed b

y l

oca

tio

n w

ith

in t

he

estu

ary

(fr

om

no

rth

to

so

uth

).

83

Figures

Figure 1. Location of sampling sites within San Francisco Bay. Sites within

geographic regions are color coded: North Bay – red; San Pablo Bay – yellow; Central

Bay – green; South Bay – blue. Image courtesy of and adapted from the U.S.

Geological Survey.

NORTH BAY

SAN PABLO BAY

CENTRAL BAY

SOUTH BAY

BG30

SU16

SP19

BD31

BC11

CB20

CB18

LSB2

BA10

SB18

SB02

BA41

BF21

SU02 BG20

PACIFIC

OCEAN

NORTH

15 km

84

Figure 2. Log ratio of AOA:β-AOB amoA copy numbers in San Francisco Bay

sediments as determined by quantitative PCR. qPCR log ratios from replicate DNA

extractions were averaged and error bars show one standard deviation of the replicates.

Ratios from 2004 are shown in black and 2005 in white. Shaded boxes show the range

of bottom water salinities and sediment C:N ratios.

85

Figure 3. Neighbor-joining phylogenetic tree showing the affiliation of archaeal

amoA sequences (579 base-pair fragment) from San Francisco Bay and GenBank

sequences from other environments. A more detailed phylogeny of the low-salinity

group is shown to the left. Clusters containing sequences from San Francisco Bay are

shaded light grey. Colored circles show the number of environmental clones from

each region and colored stars show enrichment culture sequences (see legend). The

number of San Francisco Bay sequences includes those described by Francis et al.

(2005). Trees are midpoint rooted. Significant bootstrap values (> 50) are shown in

italics at branch nodes (Neighbor joining bootstrap values are denoted above the line

and parsimony values are below the line for the low-salinity tree. Only parsimony

values are shown in the cluster tree).

86

Figure 4. Neighbor-joining phylogenetic tree showing the affiliation of bacterial

amoA sequences (444 base-pair fragment) from San Francisco Bay and GenBank

sequences from other environments. Clusters containing sequences from San

Francisco Bay are shaded light grey. Colored circles represent the number of

environmental clones from each region and colored stars show enrichment culture

sequences (see legend). The tree is midpoint rooted. Significant bootstrap values (>

50; parsimony analysis) are shown in italics at branch nodes.

87

Figure 5. Canonical correspondence analysis (CCA) of archaeal (left panel) and

bacterial (right panel) amoA clone libraries and environmental variables from bottom

water and sediments.

88

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Figure S1. Rarefaction curves for AOA (top panel) and β-AOB (bottom panel) amoA

clone libraries. OTUs were defined at a 5% sequence cutoff.

Figure S2. Neighbor-joining phylogenetic tree of the archaeal amoA sediment group.

Sequences from San Francisco Bay are color coded by region (see legend) and

GenBank sequences are in black. Significant bootstrap values (> 50; parsimony

analysis) are shown in italics at branch nodes.

Figure S3. Neighbor-joining phylogenetic tree of the bacterial amoA estuarine/marine

Nitrosospira-like group. Sequences from San Francisco Bay are color coded by

region (see legend) and GenBank sequences are in black. Significant bootstrap values

(> 50; parsimony analysis) are shown in italics at branch nodes.

Figure S4. Neighbor-joining phylogenetic tree of the bacterial amoA estuarine/marine

Nitrosomonas-like group. Sequences from San Francisco Bay are color coded by

region (see legend) and GenBank sequences are in black. Significant bootstrap values

(> 50; parsimony analysis) are shown in italics at branch nodes.

89

Literature Cited

Abreu, C., Jurgens, G., De Marco, P., Saano, A., and Bordalo, A.A. (2001)

Crenarchaeota and Euryarchaeota in temperate estuarine sediments. J Appl Microbiol

90: 713-718.

Balch, W.E., Fox, G.E., Magrum, L.J., Woese, C.R., and Wolfe, R.S. (1979)

Methanogens: Reevaluation of a unique biological group. Microbiol Rev 43: 260-296.

Ballinger, S.J., Head, I.M., Curtis, T.P., and Godley, A.R. (2002) The effect of C/N

ratio on ammonia oxidising bacteria community structure in a laboratory nitrification-

denitrification reactor. Water Sci Technol 46: 543-550.

Bano, N., and Hollibaugh, J.T. (2000) Diversity and distribution of DNA sequences

with affinity to ammonia-oxidizing bacteria of the beta subdivision of the class

Proteobacteria in the Arctic Ocean. Appl Environ Microbiol 66: 1960-1969.



Mexico. Appl Environ Microbiol 72: 7767-7777.

90

Beman, J.M., Popp, B.N., and Francis, C.A. (2008) Molecular and biogeochemical

evidence for ammonia oxidation by marine Crenarchaeota in the Gulf of California.

ISME J 2: 429-441.



Environ Microbiol 7: 1289-1297.

Bernhard, A.E., Tucker, J., Giblin, A.E., and Stahl, D.A. (2007) Functionally distinct

communities of ammonia-oxidizing bacteria along an estuarine salinity gradient.


Berounsky, V.M., and Nixon, S.W. (1993) Rates of nitrification along an estuarine

gradient in Narragansett Bay. Estuaries 16: 718-730.

Biebl, H., and Pfennig, N. (1978) Growth yields of green sulfur bacteria in mixed

cultures with sulfur and sulfate reducing bacteria. Arch Microbiol 117: 9-16.

Boatman, C.D., and Murray, J.W. (1982) Modeling exchangeable NH4+ adsorption in

marine sediments: Process and controls of adsorption. Limnol Oceanogr 27: 99-110.

91

Boynton, W.R., and Kemp, W.M. (1985) Nutrient regeneration and oxygen

consumption by sediments along an estuarine salinity gradient. Mar Ecol Prog Ser 23:

45-55.

Bricker, S.B., Clement, C.G., Pirhalla, D.E., Orlando, S.P., and Farrow, D.R.G. (1999)

National estuarine eutrophication assessment: Effects of nutrient enrichment in the

nation's estuaries. In. Silver Spring, MD: NOAA, National Ocean Service, Special

Projects Office and the National Centers for Coastal Ocean Science, p. 71.

Caffrey, J.M., Harrington, N., Solem, I., and Ward, B.B. (2003) Biogeochemical

processes in a small California estuary. 2. Nitrification activity, community structure

and role in nitrogen budgets. Mar Ecol Prog Ser 248: 27-40.

Caffrey, J.M., Bano, N., Kalanetra, K., and Hollibaugh, J.T. (2007) Ammonia

oxidation and ammonia-oxidizing bacteria and archaea from estuaries with differing

histories of hypoxia. ISME J 1: 660-662.

Canuel, E.A., Cloern, J.E., Ringelberg, D.B., Guckert, J.B., and Rau, G.H. (1995)

Molecular and isotopic tracers used to examine sources of organic matter and its

incorporation into the food webs of San Francisco Bay. Limnol Oceanogr 40: 67-81.

Chao, A. (1984) Nonparametric-estimation of the number of classes in a population.

Scandinavian Journal Of Statistics 11: 265-270.

92

Chao, A., and Lee, S.M. (1992) Estimating the number of classes via sample coverage.

J Am Stat Assoc 87: 210-217.

Chao, A., Ma, M.C., and Yang, M.C.K. (1993) Stopping rules and estimation for

recapture debugging with unequal failure rates. Biometrika 80: 193-201.

Chao, A., Chazdon, R.L., Colwell, R.K., and Shen, T.J. (2005) A new statistical

approach for assessing similarity of species composition with incidence and

abundance data. Ecol Lett 8: 148-159.

Chao, A., Chazdon, R.L., Colwell, R.K., and Shen, T.J. (2006) Abundance-based

similarity indices and their estimation when there are unseen species in samples.

Biometrics 62: 361-371.

Christensen, P.B., Nielsen, L.P., Sorensen, J., and Revsbech, N.P. (1990)

Denitrification in nitrate-rich streams: diurnal and seasonal variation related to benthic

oxygen metabolism. Limnol Oceanogr 35: 640-652.

Cloern, J.E., Jassby, A.D., Thompson, J.K., and Hieb, K.A. (2007) A cold phase of the

East Pacific triggers new phytoplankton blooms in San Francisco Bay. Proc Natl Acad

Sci USA 104: 18561-18565.

93

Dollhopf, S.L., Hyun, J.H., Smith, A.C., Adams, H.J., O'Brien, S., and Kostka, J.E.

(2005) Quantification of ammonia-oxidizing bacteria and factors controlling

nitrification in salt marsh sediments. Appl Environ Microbiol 71: 240-246.

Francis, C.A., O’mullan, G.D., and Ward, B.B. (2003) Diversity of ammonia








of the ocean. Proc Natl Acad Sci USA 102: 14683-14688.

Freitag, T.E., and Prosser, J.I. (2003) Community structure of ammonia-oxidizing

bacteria within anoxic marine sediments. Appl Environ Microbiol 69: 1359-1371.

Freitag, T.E., and Prosser, J.I. (2004) Differences between betaproteobacterial

ammonia-oxidizing communities in marine sediments and those in overlying water.

Appl Environ Microbiol 70: 3789-3793.

94



freshwater-marine gradient. Environ Microbiol 8: 684-696.

Granger, J., and Ward, B.B. (2003) Accumulation of nitrogen oxides in copper-limited

cultures of denitrifying bacteria. Limnol Oceanogr 48: 313-318.

Gruber, N., and Galloway, J.N. (2008) An Earth-system perspective of the global

nitrogen cycle. Nature 451: 293-296.

Hager, S., and Schemel, L. (1996) Dissolved inorganic nitrogen, phosphorus and

silicon in South San Francisco Bay; I. Major factors affecting distributions. In San

Francisco Bay: The ecosystem; Further investigations into the natural history of San

Francisco Bay and Delta with reference to the influence of man. Hollibaugh, J.T. (ed):

American Association for the Advancement of Science, p. 542.

Hallam, S.J., Mincer, T.J., Schleper, C., Preston, C.M., Roberts, K., Richardson, P.M.,

and DeLong, E.F. (2006a) Pathways of carbon assimilation and ammonia oxidation

suggested by environmental genomic analyses of marine Crenarchaeota. PLoS Biology

4: e95.

95

Hallam, S.J., Konstantinidis, K.T., Putnam, N., Schleper, C., Watanabe, Y., Sugahara,

J. et al. (2006b) Genomic analysis of the uncultivated marine crenarchaeote

Cenarchaeum symbiosum. Proc Natl Acad Sci USA 103: 18296-18301.

Hatzenpichler, R., Lebedeva, E.V., Spieck, E., Stoecker, K., Richter, A., Daims, H.,

and Wagner, M. (2008) A moderately thermophilic ammonia-oxidizing crenarchaeote

from a hot spring. Proc Natl Acad Sci USA 105: 2134-2139.

Head, I.M., Hiorns, W.D., Embley, T.M., McCarthy, A.J., and Saunders, J.R. (1993)

The phylogeny of autotrophic ammonia-oxidizing bacteria as determined by analysis

of 16S ribosomal RNA gene sequences. J Gen Microbiol 139: 1147-1153.

Herfort, L., Schouten, S., Abbas, B., Veldhuis, M.J., Coolen, M.J., Wuchter, C. et al.

(2007) Variations in spatial and temporal distribution of Archaea in the North Sea in

relation to environmental variables. FEMS Microbiol Ecol 62: 242-257.

Herrmann, M., Saunders, A.M., and Schramm, A. (2008) Archaea dominate the

ammonia-oxidizing community in the rhizosphere of the freshwater macrophyte

Littorella uniflora. Appl Environ Microbiol 74: 3279-3283.

Horrigan, S.G., and Springer, A.L. (1990) Oceanic and estuarine ammonium oxidation

- effects of light. Limnol Oceanogr 35: 479-482.

96

Ingalls, A.E., Shah, S.R., Hansman, R.L., Aluwihare, L.I., Santos, G.M., Druffel,

E.R.M., and Pearson, A. (2006) Quantifying archaeal community autotrophy in the

mesopelagic ocean using natural radiocarbon. Proc Natl Acad Sci USA 103: 6442-

6447.

Jassby, A.D., and Cloern, J.E. (2000) Organic matter sources and rehabilitation of the

Sacramento-San Joaquin Delta (California, USA). Aquatic Conservation: Marine and

Freshwater Ecosystems 10: 323-352.

Jassby, A.D., Cloern, J.E., and Powell, T.M. (1993) Organic carbon sources and sinks

in San Francisco Bay: variability induced by river flow. Mar Ecol Prog Ser 95: 39-54.

Jensen, K., Revsbech, N.P., and Nielsen, L.P. (1993) Microscale Distribution of

Nitrification Activity in Sediment Determined with a Shielded Microsensor for

Nitrate. Appl Environ Microbiol 59: 3287-3296.

Jensen, K., Sloth, N.P., Risgaard-Petersen, N., Rysgaard, S., and Revsbech, N.P.

(1994) Estimation of Nitrification and Denitrification from Microprofiles of Oxygen

and Nitrate in Model Sediment Systems. Appl Environ Microbiol 60: 2094-2100.

Jones, R.D., and Hood, M.A. (1980) Effects of temperature, pH, salinity, and

inorganic nitrogen on the rate of ammonium oxidation by nitrifiers isolated from

wetland environments. Microb Ecol 6: 339-347.

97

Joye, S.B., and Hollibaugh, J.T. (1995) Influence of sulfide inhibition of nitrification

on nitrogen regeneration in sediments. Science 270: 623-625.

Karner, M.B., DeLong, E.F., and Karl, D.M. (2001) Archaeal dominance in the

mesopelagic zone of the Pacific Ocean. Nature 409: 507-510.

Kennish, M.J. (ed) (2000) Estuary Restoration and Maintenance: the National

Estuary Program. Boca Raton, FL: CRC Press LLC.

Kimmerer, W.J. (2004) Open Water Processes of the San Francisco Estuary: From

Physical Forcing to Biological Responses. San Francisco Estuary and Watershed

Science 2: 1-147.



Nature 437: 543-546.

Koper, T.E., El-Sheikh, A.F., Norton, J.M., and Klotz, M.G. (2004) Urease-encoding

genes in ammonia-oxidizing bacteria. Appl Environ Microbiol 70: 2342-2348.

98

Lam, P., Jensen, M.M., Lavik, G., McGinnis, D.F., Muller, B., Schubert, C.J. et al.

(2007) Linking crenarchaeal and bacterial nitrification to anammox in the Black Sea.

Proc Natl Acad Sci USA 104: 7104-7109.

Leininger, S., Urich, T., Schloter, M., Schwark, L., Qi, J., Nicol, G.W. et al. (2006)


806-809.

Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Yadhukumar et al.

(2004) ARB: a software environment for sequence data. Nucleic Acids Res 32: 1363-

1371.

Maddison, D.R., and Maddison, W.P. (2000) MacClade 4: Analysis of phylogeny and

character evolution. Sunderland, Massachusetts: Sinauer Associates.

McCaig, A.E., Phillips, C.J., Stephen, J.R., Kowalchuk, G.A., Harvey, S.M., Herbert,

R.A. et al. (1999) Nitrogen cycling and community structure of proteobacterial beta-

subgroup ammonia-oxidizing bacteria within polluted marine fish farm sediments.


McCune, B., and Grace, J.B. (2002) Analysis of Ecological Communities. Gleneden

Beach, Oregon: MjM Software Design.

99

Mincer, T.J., Church, M.J., Taylor, L.T., Preston, C., Karl, D.M., and DeLong, E.F.


Monterey Bay and the North Pacific Subtropical Gyre. Environ Microbiol 9: 1162-

1175.

Mobley, H.L., Island, M.D., and Hausinger, R.P. (1995) Molecular biology of

microbial ureases. Microbiol Rev 59: 451-480.

Morel, F.M.M., and Price, N.M. (2003) The biogeochemical cycles of trace metals in

the oceans. Science 300: 944-947.

Nakagawa, T., Mori, K., Kato, C., Takahashi, R., and Tokuyama, T. (2007)

Distribution of cold-adapted ammonia-oxidizing microorganisms in the deep-ocean of

the northeastern Japan Sea. Microbes Environ 22: 365-372.

O'Mullan, G.D., and Ward, B.B. (2005) Relationship of temporal and spatial

variabilities of ammonia-oxidizing bacteria to nitrification rates in Monterey Bay,

California. Appl Environ Microbiol 71: 697-705.

Ochsenreiter, T., Selezi, D., Quaiser, A., Bonch-Osmolovskaya, L., and Schleper, C.

(2003) Diversity and abundance of Crenarchaeota in terrestrial habitats studied by 16S

RNA surveys and real time PCR. Environ Microbiol 5: 787-797.

100

Park, H.D., Wells, G.F., Bae, H., Criddle, C.S., and Francis, C.A. (2006) Occurrence

of ammonia-oxidizing archaea in wastewater treatment plant bioreactors. Appl


Purkhold, U., Wagner, M., Timmermann, G., Pommerening-Roser, A., and Koops,

H.P. (2003) 16S rRNA and amoA-based phylogeny of 12 novel betaproteobacterial

ammonia-oxidizing isolates: extension of the dataset and proposal of a new lineage

within the nitrosomonads. Int J Syst Evol Microbiol 53: 1485-1494.

Purkhold, U., Pommerening-Roser, A., Juretschko, S., Schmid, M.C., Koops, H.P.,



molecular diversity surveys. Appl Environ Microbiol 66: 5368-5382.

Rotthauwe, J.H., Witzel, K.P., and Liesack, W. (1997) The ammonia monooxygenase

structural gene amoA as a functional marker: Molecular fine-scale analysis of natural

ammonia-oxidizing populations. Appl Environ Microbiol 63: 4704-4712.




101

Santoro, A.E., and Boehm, A.B. (2007) Frequent occurrence of the human-specific

Bacteroides fecal marker at an open coast marine beach: relationship to waves, tides

and traditional indicators. Environ Microbiol 9: 2038-2049.



physicochemical gradients in a subterranean estuary. Environ Microbiol 10: 1068-

1079.

Schloss, P.D., and Handelsman, J. (2005) Introducing DOTUR, a computer program

for defining operational taxonomic units and estimating species richness. Appl Environ

Microbiol 71: 1501-1506.

Schloss, P.D., and Handelsman, J. (2006) Introducing SONS, a Tool for Operational

Taxonomic Unit-Based Comparisons of Microbial Community Memberships and

Structures. Appl Environ Microbiol 72: 6773-6779.

Seitzinger, S., Harrison, J.A., Bohlke, J.K., Bouwman, A.F., Lowrance, R., Peterson,

B. et al. (2006) Denitrification across landscapes and waterscapes: a synthesis. Ecol

Appl 16: 2064-2090.

102

Stephen, J.R., McCaig, A.E., Smith, Z., Prosser, J.I., and Embley, T.M. (1996)

Molecular diversity of soil and marine 16S rRNA gene sequences related to beta-

subgroup ammonia-oxidizing bacteria. Appl Environ Microbiol 62: 4147-4154.

Stephen, J.R., Chang, Y.J., Macnaughton, S.J., Kowalchuk, G.A., Leung, K.T.,

Flemming, C.A., and White, D.C. (1999) Effect of toxic metals on indigenous soil

beta-subgroup proteobacterium ammonia oxidizer community structure and protection

against toxicity by inoculated metal-resistant bacteria. Appl Environ Microbiol 65: 95-

101.

Swofford, D.L. (2003) PAUP*. Phylogenetic Analysis Using Parsimony (*and Other

Methods). Sunderland, Massachusetts: Sinauer Associates.

ter Braak, C.J.F., and Smilauer, P. (2002) CANOCO Reference Manual and

CanoDraw for Windows User's Guide: Software for Canonical Community Ordination

(version 4.5). Ithaca, NY: Microcomputer Power.

Treusch, A.H., Leininger, S., Kletzin, A., Schuster, S.C., Klenk, H.P., and Schleper, C.


uncultivated mesophilic crenarchaeota in nitrogen cycling. Environ Microbiol 7: 1985-

1995.

103

Venter, J.C., Remington, K., Heidelberg, J.F., Halpern, A.L., Rusch, D., Eisen, J.A. et

al. (2004) Environmental genome shotgun sequencing of the Sargasso Sea. Science

304: 66-74.

Vieira, M., and Bordalo, A. (2000) The Douro estuary (Portugal): a mesotidal salt

wedge. Oceanologica Acta 23: 585-594.

Vitousek, P.M., Aber, J.D., Howarth, R.W., Likens, G.E., Matson, P.A., Schindler,

D.W. et al. (1997) Technical report: Human alteration of the global nitrogen cycle:

Sources and consequences. Ecol Appl 7: 737-750.

Wankel, S.D., Kendall, C., Francis, C.A., and Paytan, A. (2006) Nitrogen sources and

cycling in the San Francisco Bay Estuary: A nitrate dual isotopic composition

approach. Limnol Oceanogr 51: 1654-1664.

Ward, B.B., and O'Mullan, G.D. (2002) Worldwide distribution of Nitrosococcus

oceani, a marine ammonia-oxidizing gamma-proteobacterium, detected by PCR and

sequencing of 16S rRNA and amoA genes. Appl Environ Microbiol 68: 4153-4157.

Wuchter, C., Abbas, B., Coolen, M.J., Herfort, L., van Bleijswijk, J., Timmers, P. et

al. (2006) Archaeal nitrification in the ocean. Proc Natl Acad Sci USA 103: 12317-

12322.

104

Chapter 4. Genome of a Low-Salinity Ammonia-Oxidizing Archaeon Determined by Single-Cell and Metagenomic Analysis Paul C. Blainey*1, Annika C. Mosier*2, Anastasia Potanina1, Christopher A. Francis2 and Stephen R. Quake1 *contributed equally 1Deptartment of Bioengineering, Stanford University, Stanford, CA 94305, USA and Howard Hughes Medical Institute 2Department of Environmental Earth System Science, Stanford University, Stanford, CA 94305-4216, USA Published as: Blainey, P.C.*, Mosier, A.C.*, Potanina, A., Francis, C.A., and S.R. Quake. 2011. Genome of a Low-Salinity Ammonia-Oxidizing Archaeon Determined by Single-Cell and Metagenomic Analysis. PLoS ONE 6: e16626. *contributed equally.

105

Abstract

Ammonia-oxidizing archaea (AOA) are thought to be among the most

abundant microorganisms on Earth and may significantly impact the global nitrogen

and carbon cycles. We sequenced the genome of AOA in an enrichment culture from

low-salinity sediments in San Francisco Bay, using a combined single-cell and

metagenomic approach. Five single cells were isolated inside an integrated

microfluidic device using laser tweezers. Their genomic DNA was amplified by

multiple displacement amplification (MDA) in 50 nL volumes inside the microfluidic

device and then sequenced. This microscopy-based approach to single-cell genomics

minimizes contamination and allows correlation of high-resolution cell images with

genomic sequences. The genome of this AOA, which we designate Candidatus

Nitrosoarchaeum limnia SFB1, is ~1.77 Mb with >2100 genes and a G+C content of

32%. Across the entire genome, the average nucleotide identity to Nitrosopumilus

maritimus, the only AOA in pure culture, is ~70%, suggesting this AOA represents a

new genus of mesophilic Crenarchaeota. Phylogenetically, the 16S rRNA and

ammonia monooxygenase subunit A (amoA) genes of this AOA are most closely

related to sequences reported from a wide variety of freshwater ecosystems. Like N.

maritimus, the low-salinity AOA appear to have an ammonia oxidation pathway

distinct from ammonia oxidizing bacteria (AOB). The genome lacks homologues for

hydroxylamine oxidoreductase or cytochromes c552 and c554, but does encode

numerous copper-binding proteins, including multicopper oxidases and small blue

copper-containing proteins, which may constitute important components of the AOA

electron transfer system. In contrast to other described AOA, these low-salinity AOA

106

appear to be motile, based on the presence of numerous motility- and chemotaxis-

associated genes in the genome. In-depth studies on the physiological and metabolic

potential of this novel group of organisms may ultimately advance our understanding

of the global carbon and nitrogen cycles.

107

Introduction

Mesophilic Crenarchaea (first discovered by Fuhrman et al. [2] and DeLong

[1]) are among the most abundant microorganisms on the planet. [1-2]Crenarchaea

constitute ~10-40% of the total microbial community in the ocean below the euphotic

zone [3-6]. In fact, it is estimated that there are 1.0x1028 crenarchaeal cells in the

world’s oceans compared to 3.1x1028 bacterial cells [3]. In soil systems, Crenarchaea

generally comprise 1-5% of the total prokaryotic community [7-9], but can reach

upwards to 38% in some systems (e.g., temperate acidic forest soils) [10]. It appears

that the vast majority of Crenarchaea in soils and the open ocean possess ammonia

monooxygenase (amoA) genes and thus may be capable of oxidizing ammonia to

nitrite, based on quantitative PCR estimates of 16S rRNA and amoA genes,

fluorescent in situ hybridization, and metagenomic analyses [11-18]. Recent

phylogenetic and genomic analyses suggest that these mesophilic Crenarchaea should

be recognized as a third archaeal phylum, the Thaumarchaeota [19-20].

AOA have proven difficult to isolate in culture, as evidenced by the fact that

only one pure culture [21] has been documented despite reports of several enrichment

cultures [12,22-25]. Many questions remain about the physiology, metabolic and

biogeochemical functions, evolutionary history, and ecological niche partitioning of

AOA. Genome sequencing is a powerful tool that can be used to address these

outstanding questions, yet traditional sequencing approaches require pure cultures or

extreme enrichment of a homogeneous population. Genome sequencing of individual

cells presents a new approach for exploring the genetic makeup of microorganisms

present in heterogeneous mixtures without the need for cell viability or growth.

108

Single-cell genomics can be used independently on uncultured microbes [26-27] or as

a complement to metagenomic studies on enrichment cultures and environmental

samples by aiding in sequence binning and genome reconstruction [28].

Here, we used a combination of single-cell genomic and metagenomic

approaches to sequence the genome of a novel, low salinity-type AOA. The single-

cell approach is useful for obtaining organism-specific sequence datasets from

heterogeneous samples, and furthermore provides an insight into micro-heterogeneity

at the population level. The genomic data reveal sequence features not observed in

other AOA, as well as similarities to published AOA genomes, which speak to the

metabolic potential and environmental relevance of these organisms.


Enrichment of ammonia-oxidizing archaea from low-salinity sediments

An ammonia-oxidizing enrichment culture initiated from sediments in the low-

salinity region of San Francisco Bay was dominated by AOA; CARD-FISH showed

that Archaea (Arch915 probe) accounted for approximately 84% of cells in the

enrichment and the remaining 16% were Bacteria (Eub338 I-III probes). The AOA

grew chemoautotrophically by aerobic ammonia oxidation to nitrite. Archaeal amoA

and 16S rRNA gene sequences from the AOA enrichment culture were

phylogenetically distinct (bootstrap supported) from N. martimus [Figure 1], and were

instead most closely related to environmental clones derived from freshwater

ecosystems. In total, over 500 GenBank amoA sequences fell within this phylogenetic

cluster, 79% of which were from low salinity habitats. Another 7% of the sequences

109

were from habitats that experience low salinity at some point in the year (e.g., San

Pablo Bay sites in San Francisco Bay), and 4% of the sequences were from soil. Most

of the remaining sequences were from salt marsh sediments [29], suggesting that this

AOA phylotype may be tolerant of elevated salinity. Nevertheless, the vast majority

of amoA sequences within this clade were from freshwater habitats.

On the basis of these results and the genome analysis described herein, we

propose the provisional classification of this novel archaeon as Candidatus

Nitrosoarchaeum limnia SFB1. The name refers to low-salinity habitat of this

archaeon and to its ability to oxidize ammonia to nitrite. Nitrosoarchaeum comes

from the Latin masculine adjective 'nitrosus' for nitrous and limnia is derived from the

Greek word 'limne' for freshwater (as in limnology) and the strain name SFB1 refers to

its isolation from San Francisco Bay. The short description of Candidatus

Nitrosoarchaeum limnia is a mesophilic, chemolithoautotrophic ammonia-oxidizing

archaeon belonging to the Thaumarchaeota phylum.

On-chip cell sorting

Based on this information suggesting that N. limnia is novel and distinct from

N. maritimus, we employed a combined single cell sorting and metagenomic approach

to obtain the genome sequence of these AOA. A variety of single-cell approaches to

microbial genomics are currently being applied [26-27,30-31]. While all rely on the

multiple displacement amplification (MDA) reaction [32] for the amplification of

unknown genomic DNA sequences, the means by which individual cells are isolated

distinguish the various approaches. Microfluidic flow [30], flow cytometry [31,33],

110

and micromanipulation [27] have all been successfully applied to isolate single cells

for genomic amplification, although each approach has strengths and weaknesses. A

challenge in this case was the tiny, submicron size of the AOA in the enrichment

culture.

We implemented laser tweezer [34-35] based cell sorting inside a microfluidic

device to combine the benefits of a sealed, integrated system for sorting and

amplification, small reaction volumes, and minimum sorting volume. The imaging

and trapping system was adapted to enable visualization and trapping of tiny cells.

This required the application of a high-numerical aperture phase-contrast objective

with high infrared transmission. All reagents were carefully filtered to remove

extraneous particles that could be confused with cells. These features of the optical

sorting method specifically and effectively address the three classes of contamination

with the potential to compromise single cell amplification reactions, namely: the

laboratory environment, amplification reagents, and cells/DNA from the sample itself.

Using this microfludic and laser trapping device, single cells smaller than one micron

in length were sorted from the AOA enrichment culture. The cells were lysed 'on-

chip' using proteinase K and alkaline lysis, and the genomes of each individual cell

were amplified by MDA. Amplified cells were screened for archaeal amoA and 16S

rRNA genes to confirm affiliation with AOA. Based on PCR screening, we selected

the MDA products of five individual AOA cells for genomic sequencing. In addition,

due to the highly enriched nature of the ammonia-oxidizing culture (with respect to the

AOA), we also carried out MDA on a benchtop scale using cells from the enrichment

culture as the template. We directly sequenced the enrichment MDA products to

111

obtain sequences from the 'bulk' enrichment culture metagenome. This combined

single cell/metagenomic approach enabled us to test the single cell assemblies for

completeness and the metagenome for contamination.

Sequencing and assembly

MDA-amplified genomic DNA from the five individual AOA cells was

sequenced using a combination of shotgun (SG) and paired-end (PE) DNA

pyrosequencing approaches on the Roche/454 FLX Titanium platform. MDA

products from both individual cells and the bulk enrichment were used to create

sequencing libraries, which were quantitated using digital PCR [36] and sequenced.

We analyzed 483,350 SG and 56,122 PE reads (119 Mb) obtained from five individual

cells and 160,632 SG and 18,118 PE reads (37 Mb) from the enrichment sample

[table 1]. The G+C contents of the single-cell reads formed a major peak centered at

32.4%, an even lower G+C content than N. maritimus (34.2%) [supplemental figure

1]. To compare the dispersion of the G+C content, we simulated shotgun reads from

N. maritimus with the same average read length as the experimentally derived

datasets. The dispersion of the read G+C content from the single-cell reads was

essentially identical to that found of N. maritimus, confirming that the individual AOA

cells were amplified without contamination from bacterial cells or DNA. The G+C

content of the enrichment reads demonstrated a principal mode similar to the single

cell datasets and N. maritimus (suggesting a large proportion of AOA in the

enrichment), as well as a higher G+C component attributed to sequences from the

bacteria present in the culture.

112

Individually, three of the single-cell datasets gave assemblies in excess of one

megabase from sequencing depths of 10x to 30x, based on an expected genome size of

about 1.7 Mb. The implied 60% genomic coverage is typical in our experience with

single-cell genomic data at such depths of sequencing. The limited genomic coverage

of the single-cell assemblies owes to the phenomenon of MDA amplification bias,

wherein certain regions of the genome are amplified to a greater extent than other

regions. The less-amplified regions are not efficiently sampled by the shotgun

sequencing procedure, and extremely high sequence coverage is required to recover

the under-represented sequences. Interestingly, each of these single-cell assemblies

represents a different 60% of the target genome, indicating that the amplification bias

profile differs from cell to cell. Thus, while further sequencing and efforts to improve

the assembly suffer diminishing returns with respect to an individual cell, great strides

in assembly quality are quickly won by combining the reads obtained from several

individual cells. By pooling the data from the five individual cells sequenced, we

obtain an assembly of 136 elements (26 scaffolds and 110 contigs) for a total of 1.69

Mb of sequence, 1.37 Mb of which was captured in the scaffolds (scaffold N50 = 156

kb) [table 1]. A combination of rarefaction analysis [Figure 1] and comparison to the

metagenomic data indicates (Table 1) that this single-cell assembly represents >95%

of the N. limnia genome.

The metagenome from the bulk enrichment was used as an independent

validation of the single cell data and to supplement the single cell analysis. To

combine the strengths of each dataset—the pure single cell AOA reads with no

bacterial sequences and the bulk enrichment reads with less amplification bias—we

113

generated a high-quality consensus assembly. First, we used five single-cell datasets

to establish a stringent G+C content filter and eliminate high G+C content (>46%)

reads from the enrichment dataset, purging much of the bacterial sequence

contribution prior to assembly of the data. Next, we mapped the single-cell reads to a

de novo assembly of the filtered enrichment reads to identify and reject chimeric

sequences, the occurrence of which is expected to be independent in different datasets.

The small pool of high-G+C reads was mapped against N. maritimus and C.

symbiosum A to recover high G+C content conserved sequences. The overall

consensus assembly contained 2 scaffolds and 29 contigs for a total of 1.77 Mb

(contig N50 = 94 kb) [Table 1]. More than 99% of the assembled bases were

represented by the two scaffolds (1.36 Mb and 0.39 Mb in length). The overall

consensus assembly represents greater than 97% of the genome, based on rarefaction

analysis. The elements from both assemblies were oriented and ordered by homology

to N. maritimus. Orientation was determined by maximizing the sum of the alignment

score to this reference, and the elements ordered according to the hit length-weighted

average subject position of the homologous regions that were co-orientated with the

reference.

Microheterogeneity and sequencing (MDA) error

Comparing homologous regions of the single cell assemblies with each other

and with the enrichment dataset, we consistently find 99.9% average nucleotide

identity. The identity result is a lower limit because of the possibility of errors

introduced in the amplification and sequencing procedures. Newbler reports 5904

114

bases in the assembly as having quality scores lower than 40. If we conservatively

estimate the average quality score of these lower-quality aligned bases at 10 and the

average quality score of the high-quality aligned bases at 40, 234 errors arising from

ambiguous alignments and low-quality raw bases are predicted. Based on independent

measurements (data not shown), we expect a minor contribution of additional errors

based on the MDA procedure, on the order of 1/100,000 bp, or fewer than 20 bases

across the N. limnia genome. Thus, it seems that microheterogeneity among the

enrichment AOA must account for a substantial portion of the SNPs predicted among

the individual cells.

Genome size estimate

Subsets of the enrichment dataset were assembled in a rarefaction analysis to

determine the approximate genome size of N. limnia [figure 2]. The total assembly

size smoothly asymptotes below 1.80 Mb, indicating that the N. limnia assembly

represents an essentially complete genome at 1.77 Mb. We also rarefied the single-

cell data, which we expected to assemble less efficiently than the enrichment dataset,

due to the effect of greater amplification bias in the single cell MDA reactions

compared with the bulk MDA of the enrichment sample. This expectation was borne

out, but the single-cell data does asymptote at approximately the same level. This

suggests that the single-cell dataset contains all the genomic sequences, albeit with

less genomic information per Mb as a result of representation bias. Assembly using

only the single cell data yields a genome that is estimated to be 95% complete [Table

1].

115

Gene prediction and annotation

Gene prediction was carried out on the N. limnia assembly using the FgenesB

platform (Softberry, Inc.), which called 2180 genes, including 2130 protein-coding

ORFs and 50 RNA genes [Table 2]. In total, 86.34 % of the assembled bases were

predicted to code for proteins. The N. limnia genome encodes 45 transfer RNA

(tRNA) genes (42 identified by FgenesB and 3 additional tRNA sequences identified

by comparison with N. maritimus). tRNAs for 18 amino acids were predicted, which

further included two tRNAs with unidentified codon/amino acid specificity. Single

copies of the full-length 16S, 23S, 5S, and 5.8S ribosomal RNA genes exist in the

assembly. The average nucleotide identities of the 16S and 23S genes to N.

maritimus are of 97% and 94%, respectively. In addition, a conserved noncoding

RNA with 93% identity to N. maritimus was identified.

Codon usage and nucleotide skew

The N. limnia codon usage was compared with that of N. maritimus and C.

symbiosum A, with a nucleotide-matched scrambled control as a reference (data not

shown). The codon usage of N. limnia is very similar to that of N. maritimus, which

has a similar G+C content, and distinct from both the high G+C content C. symbiosum

A and the randomized control. This is also the case for N. limnia CDS lacking

homology to database sequences, supporting the identification of these regions as

protein-coding genes.

116

The G+C skew in the N. limnia genome exhibits an irregular pattern

comparable to that observed in other archaeal genomic sequences [figure 3] [37].

Analysis of cumulative G+C and GGG+T skew failed to reveal an obvious origin of

replication, although several strong transition points exist in the genome.

Gene prediction and percent identity comparisons

The N. limnia genome has a coding density of 1.20 protein coding genes per kb

of sequence data, which is slightly higher than the coding density for N. maritimus

(1.09) or C. symbiosum A (0.99) [Table 2]. The N. limnia genome shares 1378 genes

with N. maritimus and 1112 genes with C. symbiosum A (based on Reciprocal Best

Hit [RBH] analysis). The 1378 genes that were RBHs with N. maritimus genes

represent 64.7% of the coding genes predicted in the N. limnia assembly. Conversely,

76.8% of the N. maritimus coding genes were identified as RBHs in the N. limnia

assembly. Only 54.3 % of coding genes in the N. limnia assembly were RBHs with C.

symbiosum A, despite the larger genome size of C. symbiosum A, indicating the

greater evolutionary distance that exists between N. limnia and C. symbiosum A than

exists between N. limnia and N. maritimus.

When compared across the entire length of the genome, N. limnia was distinct

from both N. maritimus and C. symbiosum A [Table 3 and Figure 4]. The average

nucleotide identity across overlapping regions of the N. limnia and N. maritimus

genomes was 76.3% [Table 3]. When scoring non-overlapping genomic regions at an

arbitrary 25% nucleotide identity, the overall nucleotide identity declines to 68.1%.

The percent identity between the N. limnia and C. symbiosum A genomes was even

117

lower, as expected due to the divergent total nucleotide content: 67.4% identity across

overlapping regions and 33.3% when including non-overlapping regions [Table 3].

The cutoff for defining a new species was recently proposed to be 95-96% average

nucleotide identity between two genomes [38]-an updated method of species

circumspcriptions compared to DNA-DNA hybridization. Thus, the much lower

percent identity between N. limnia, N. maritimus and C. symbiosum A across the entire

genome supports the classification of N. limnia as a new genus of AOA. The N.

limnia and N. maritimus genomes shared extensive regions of synteny (gene order),

whereas N. limnia and C. symbiosum A showed very little synteny across the genomes

[Figure 4 and Supplemental Figure 2].

Motility

One of the most striking features of the N. limnia genome, relative to N.

maritimus and C. symbiosum, is the presence of numerous motility-associated genes,

including those required for Flp pilus assembly, flagellar assembly, and chemotaxis

[Table 4]. In total, N. limnia has 38 motility-associated genes, many of which are

clustered in a locus on the largest scaffold [Figure 4]. This is a larger number than

observed in many other archaea including Sulfolobus solfataricus, Aeropyrum pernix,

and Archaeoglobus fulgidus, which all have fewer predicted motility genes (15-20

each). The N. limnia flagellar genes share homology with thermophilic

Crenarchaeota. In addition, the majority of the cells in the AOA enrichment culture

were motile. This appears to be a distinctive trait from N. maritimus and C.

symbiosum A, both of which lack gene suites for assembly of flagella. N. limnia was

118

isolated from estuarine sediments, raising the possibility that this AOA requires

motility to position itself in response to varying substrate and/or oxygen

concentrations in the surface sediments. In contrast, N. maritimus was isolated from

gravel in a tropical fish tank and C. symbiosum A is a symbiont of a marine sponge,

where perhaps motility is less essential. Determining how widespread motility and

chemotaxis are within phylogenetically diverse AOA genera (from water columns,

soils, sediments, etc.) represents an exciting new avenue of research.

Protection from osmotic shock

The N. limnia genome contains one large-conductance mechanosensitive

channel protein and two small-conductance mechanosensitive channel proteins.

Mechanosensitive (MS) channels are found in bacteria, archaea, and eukaryotes. MS

channel proteins protect cells from hypoosmotic shock when a cell transitions from a

high osmolarity environment to a low osmolarity environment [39]. This transition

causes a rapid influx of water into the cells, which increases the internal turgor

pressure and places considerable strain on the cytoplasmic membrane and cell wall.

To release this pressure, the MS channels open, allowing the release of low molecular

mass solutes (K+, glutamate and compatible solutes) out of the cell. This rapidly

decreases the concentration of osmotically-active solutes in the cytoplasm, minimizing

the influx of water and thereby preventing cell lysis. The C. symbiosum A and N.

maritimus genomes appear to lack large-conductance MS channels, although they each

encode at least one small-conductance MS channel protein. N. limnia was isolated

from sediments in the San Francisco Bay estuary that likely experience regular

119

fluctuations in salinity and may depend on these MS channel proteins for protection

from osmotic shock.

The N. maritimus genome contains genes coding for ectoine biosynthesis,

representing the first description of this process in Archaea [40]. Bacteria and

Archaea accumulate many different organic osmolytes, including ectoine, in response

to high osmotic pressure [41]. Notably, the N. limnia genome is missing all four genes

for ectoine biosynthesis (ectA, ectB, ectC, and ectD). Although the N. limnia genome

is not closed, the fact that the genome is missing all four genes suggests that N. limnia

does not use ectoine biosynthesis as a mechanism for dealing with osmotic stress.

These genomic differences may in turn reflect niche differentiation between N.

maritimus (marine) and N. limnia (low-salinity).

Genes conferring resistance

Several phage integrase proteins and one phage SPO1 DNA polymerase-

related protein were identified in the N. limnia genome. However, no obvious

CRISPRs (clustered regularly interspaced short palindromic repeats) were detected in

the genome, which confer resistance to phage infection in many archaeal species,

despite the fact that CRISPRs are found within ~90% of archaeal genomes [[42] and

references within]. Thus, N. limnia may be particularly susceptible to phage infection.

Three transposase genes were also present in the genome and may be an additional

source of genetic exchange and rearrangement. Several genes putatively conferring

antibiotic resistance were found in the N. limnia genome (e.g. beta-lactamase domain-

containing proteins and an antibiotic biosynthesis monooxygenase). The presence of

120

several metal transporters and proteins implicated in metal resistance may reflect

metal exposure in the San Francisco Bay sediments; indeed, there are many potentially

toxic metals in the sediments of North San Francisco Bay [23].

Novel cell division system

In Euryarchaeota and Korarchaeota (as well as most bacteria), cell division is

mediated by FtsZ proteins, which are homologous to tubulin, the building block of the

microtubule cytoskeleton in eukaryotes [43-44]. The sequenced genomes of

hyperthermophilic Crenarchaeota lack FtsZ genes. Instead, it appears that all

Crenarchaeota (with the exception of the Thermoproteales, where the cell division

machinery remains uncharacterized) use a novel Cdv cell division system that is

related to the related to the eukaryotic ESCRT-III sorting complex [44-45]. However,

both the FtsZ and Cdv systems of cell division are evident in the AOA genomes,

including N. limnia, N. maritimus, and C. symbiosum A, as well as the draft genome of

Nitrososphaera gargensis [20]. The striking difference in cell division between these

AOA and all other Crenarchaea supports the proposal that they belong to a third

archaeal phylum, the Thaumarchaeota [19-20].

Chromosome structure and organization

Prokaryotic chromosomal DNA is highly compacted to form a nucleus-like

structure—the nucleoid. Bacteria and Archaea have to compact their genomic DNA

by over 1000-fold [46] by utilizing a wide variety of DNA binding chromatin proteins.

These proteins not only serve to compact the genomic DNA, but also in maintaining

121

the genome in a state that allows access for gene expression, DNA replication and

repair. The two major archaeal chromatin proteins are the archaeal histone and Alba

proteins. Archaeal histones act as spools that DNA wraps around (similar to their

eukaryotic homologs), whereas Alba proteins form extended filamentous fibers

containing DNA [47]. All archaea have either one of these proteins, and many encode

both proteins [48]. The N. limnia genome codes for one histone protein (Nlim1809)

and two Alba proteins (Nlim0610 and Nlim1213).

Carbon metabolism

The N. limnia genome sequence suggests that this organism uses a modified

version of the 3-hydroxypropionate/4-hydroxybutyrate pathway for carbon fixation, as

in N. maritimus and C. symbiosum A [40, 55]. In addition to pathways for autotrophic

carbon fixation, genes putatively inferring the function of organic carbon consumption

have been identified in the genomes of N. limnia, N. maritimus and C. symbiosum A.

Several isotopic studies have suggested that natural populations of AOA may be

capable of mixotrophic growth [4-5,49-51]. It remains to be determined how

important heterotrophy and/or mixotrophy is to AOA in natural environments.

Ammonia oxidation and respiratory chain

In ammonia-oxidizing bacteria (AOB), ammonia is oxidized to hydroxylamine

by ammonia monooxygenase (AMO) composed of α, β, and γ subunits encoded by the

amoA, amoB, and amoC genes, respectively. Hydroxylamine oxidoreductase (HAO)

then oxidizes hydroxylamine to nitrite, generating four electrons that are transferred to

122

the terminal oxidase and AMO via cytochromes c554 and cM552 and the ubiquinone

pool ([52], [53] and references therein).

The N. limnia genome contains genes putatively encoding the α, β, and γ

subunits of the archaeal AMO protein. The gene order in the AMO operon is

conserved amongst N. limnia, N. maritimus [40], and the Sargasso Sea metagenomic

fragments [54]: amoA, hypothetical protein, amoC, followed by amoB. The N. limnia

AMO gene sequences have relatively low identities to N. maritimus, ranging from 80-

91% nucleotide identity and 77-96% amino acid identity [Table 5].

While the details of the ammonia oxidation and electron transfer pathways

remain to be elucidated, AOA, including N. maritimus, C. symbiosum A and N. limnia,

appear to have an ammonia oxidation pathway distinct from AOB [40,55-56]. The N.

maritimus, C. symbiosum A and N. limnia genomes lack homologues for HAO or

cytochromes c554 and cM552, and it remains to be demonstrated whether

hydroxylamine is an intermediate in the pathway. AOA in pure culture produce nitrite

from the oxidation of ammonia [21,57] but the genome sequence annotations do not

include the key HAO enzyme necessary to produce nitrite. The archaeal AMO protein

may not produce hydroxylamine (explaining the missing HAO protein) or the AOA

may contain novel enzymes responsible for the oxidation of hydroxylamine. In either

case, redox reactions in AOA are likely more reliant on copper than iron as seen in the

AOB with the iron-rich HAO enzyme and cytochromes [40,55]. The three AOA

genomes all contain numerous copper-containing proteins, including multicopper

oxidases (MCOs) and small blue copper-containing proteins. These predicted copper-

containing proteins might form the basis of electron transfer in AOA with

123

functionality similar to cytochromes in AOB [40,55]. The N. limnia genome contains

11 blue (type 1) copper domain-containing proteins, 1 cupredoxin blue copper protein,

and 4 multicopper oxidases [Supplemental Table 1]. Alternative pathways for

ammonia oxidation have been proposed by Walker et al. [40]. The N. limnia genome

does not further illuminate the exact mechanism for ammonia oxidation.

Two ammonium transporter genes were identified in the N. limnia genome

(Nlim1421 and Nlim1564). If the AMO protein accesses ammonia outside of the

cytoplasmic membrane (as in AOB), the AOA would not necessarily need ammonium

transporters. However, the RBHs exist between the N. limnia ammonium permeases

and both N. maritimus and C. symbiosum A. The existence of conserved ammonium

transporters in all three genomes suggests a necessary function for this activity. Arp et

al. (2007) [52] proposed that ammonium transporters in AOB might facilitate the

regulation of AMO and HAO that is induced by ammonium or to assist in the uptake

of ammonia for nitrogen assimilation.

Many AOB are able to use urea as an alternative source of ammonia and

carbon dioxide for growth [58]. Ureolysis may therefore be an ecological advantage in

environments with low pH and/or fluctuating ammonia and carbon dioxide

availability. Like N. maritimus [40], there was no evidence that N. limnia has the

ability to use urea as an alternative source of ammonia, because no genes encoding

urea transporters or urease enzymes were identified in the genome. Interestingly,

however, both urea transporter and urease genes were present in the C. symbiosum A

genome [55]. The symbiotic C. symbiosum A may use ureolysis as a mechanism to

remove nitrogenous waste from its host, the marine sponge Axinella mexicana [56].

124

Konstantinidis et al. [16] also reported urease genes in a crenarchaeal genomic

scaffold from 4,000-m-depth at station ALOHA in the Pacific Ocean (along with

amoA, amoB, amoC, and ammonia permease genes), suggesting that urease activity is

a feature of at least some marine water column AOA.

Conclusion

We used an approach combining metagenomic and single cell data to produce

a high quality and complete (but not fully closed) genome of a mesophilic AOA from

a low-salinity estuarine environment – Nitrosoarchaeum limnia. Single-cell genome

sequencing enabled discovery and assembly of 95% of the complete N. limnia

genome; when this data was combined with metagenomic reads derived from the bulk

enrichment culture, greater than 97% of the genome was determined. The genomic

sequence reflects several systems present in the two other AOA genome sequences

available, including the machinery for cell division and ammonia oxidation,

establishing common elements that define this functionally related group of archaea

belonging to the proposed archaeal phylum Thaumarchaeota. The N. limnia genome

also reveals features unique to this environmental AOA, including a substantial suite

of genes for flagellar biosynthesis and chemotaxis, and the presence of a large

mechanosensitive channel that may be crucial for fitness in the estuarine environment.

Several large regions of the genome lacking homology to N. maritimus and C.

symbiosum may encode other functional genes important to the survival of free-living,

estuarine AOA. Cultivation and genomic sequencing of environmentally-relevant

AOA portends further physiological and metabolic studies of this novel and pervasive

125

group of organisms that will ultimately advance our understanding of the global

carbon and nitrogen cycles.

Materials and Methods

Cultivation of archaeal ammonia-oxidizers

Ammonia-oxidizer enrichment cultures were initiated from sediments collected

in northern part of the San Francisco Bay estuary in 2006 (site SU001S; latitude

38°5'55.75"N, longitude 122° 2'47.40"W). Surface sediments were inoculated into a

modified low-salinity version of Synthetic Crenarchaeota Medium [21], as described

by Mosier and Francis (2008) [23]. Cultures were grown in static culture flasks or

serum vials in the dark at room temperature (approximately 22°C). Approximately

10% of the culture volume was routinely transferred into fresh media.

The presence of ammonia-oxidizing archaea was confirmed by gene

sequencing and fluorescent in situ hybridization (FISH). Briefly, DNA was extracted

from the enrichment culture and archaeal 16S rRNA (Arch21F and Arch958R primers;

[1]) and archaeal amoA (Arch-amoAF and Arch-amoAR primers; [59]) genes were

PCR amplified, cloned and sequenced. Catalyzed reporter deposition fluorescence in

situ hybridization (CARD-FISH; e.g., [60]) was carried out with horseradish

peroxidase (HRP)-labeled probes Arch915 [61] and EUB338 I-III [62-63]. Filters

were treated with lysozyme or proteinase K for permeabilization of bacteria and

archaea, respectively.

Phylogenetic trees

126

Archaeal amoA nucleotide sequences were aligned with GenBank sequences

using MacClade version 4.08 [64]. Archaeal 16S rRNA gene sequences were aligned

using the NAST alignment server [65]. Sequences were then manually adjusted and

aligned to the Greengenes 16S rRNA gene database [66] in ARB [67]. For

phylogenetic analyses, a 582-bp region of the sequence alignment was used for amoA

and a 1272-bp region for 16S rRNA. Neighbor joining (based on Jukes-Cantor

correction) and parsimony phylogenetic trees were constructed in ARB [67] with 1000

bootstrap replicates. The set root option in ARB was used to manually root the trees

between the sediment and water column cluster sequences.

Microfluidic device and cell sorting

Microfludidic devices with 32 units for nanoliter MDA reactions were

produced by the Stanford microfluidics foundry. These devices are similar to those

described in Marcy et al. [26]. The chip was pre-treated for 10 minutes with pluronic

F127 at 0.2% in 1x PBS before filling with 1x PBS containing 0.01% Tween-20 and

0.01% pluronic F127 to reduce cell adhesion. The sample was pre-treated with 0.5 - 1

µg/mL proteinase K at room temperature for 10 minutes and then washed in 1x PBS

and resuspended in 50:50 1X PBS:Ethanol. BSA was added to the treated cells (0.1

mg/mL final concentration). Individual cells were sorted one at a time from the bulk

sample using the laser trap, through a two-valve gate, opening one valve at a time to

allow the trapped cell, but not fluid flow, to pass through (see supplemental movie).

Each trapped cell was moved about 1 mm from the bulk sample to the reaction

chamber using the laser trap.

Single-cell amplification

127

The Repli-G midi MDA reagents (Qiagen) were used to amplify individual

cells in 50-60 nL volumes on the device, yielding 107 - 108 genomic equivelants of

dsDNA product. First, cells contained in 0.75 nL PBS with 0.02% Tween-20 were

flushed into the first lysis chamber with 3 nL buffer DLB (supplemented with 0.1 M

dithiothreitol) to complete lysis and denature the genomic DNA. Then, ~50 nL of

reaction mix (45 µL were prepared from 29 µL Repli-G reaction buffer, 10 µL 20 mM

H2O with 0.6% tween, 2 µL Repli-G enzyme, and 2 µL Repli-G stop solution) was

added to each of the 32 reactions. The chip was then transferred to a hot plate set to

32°C and incubated overnight. This was carried out by fitting the outboard product

ports on the chip with plastic pipet tips (P10 size), and by flushing the products into

the pipet tips with the TRIS solution plumbed into the reagent port at 8 psi.

Six µl of each first-round reaction product were re-amplified using the Repli-G

midi kit (Qiagen). Additionally, 1 µl of the proteinase K-treated cell solution used for

sorting was amplified. The template solution was denatured by the addition of 3.5 µl

buffer DLB for 5 minutes at room temperature. The solution was neutralized by

adddition of 3.5 µl stop solution. A reaction mix consisting of 29 µl reaction buffer,

10 µl water, and 1 µl enzyme was prepared on ice, and then added to the denatured

template. Reactions were incubated at 30°C for 12 hours and then diluted 10-fold in

10 mM TRIS with 0.02% Tween-20 and stored at -60°C.

Shotgun and paired-end library prep

Shotgun libraries were prepared from approximately 5 µg of the second round

MDA product according to the Roche/454 protocol for "Titanium" shotgun libraries

with the following modifications. Custom barcoded adaptor oligos (IDT) were used to

128

enable pooling multiple libaries in a single emulsion PCR reaction and picotiter plate

region during sequencing. To obtain dsDNA sequencing libraries and shorten the

library preparation process, the library immobilization, Fill-in, and single-stranded

library isolation steps (steps 3.7 - 3.9) were omitted.

Paired end libraries were prepared from approximately 5 µg of the second

round MDA product according to the Roche/454 protocol for "Titanium 3kb span"

libraries with the following modifications. Custom barcoded adaptor oligos (IDT)

were used in step 3.9.2 to enable pooling multiple libaries in a single emulsion PCR

reaction and picotiter plate region during sequencing. To obtain dsDNA sequencing

libraries and shorten the library preparation process, the library immobilization and

single-stranded library isolation steps (steps 3.12.1 and 3.12.2) were omitted.

Sequencing library quantification

Sequencing libraries were quantified using digital PCR as previously reported

[36], with the exceptions that 48.770 digital arrays (Fluidigm) were used for the

microfluidic dPCR step, and that amplification primers complimentary to the Titanium

adaptor sequences were used. Briefly, serial dilutions of the sequencing libraries were

made in 10 mM TRIS buffer with 0.02% Tween-20. 48 sample preparations were

then combined per the Fluidigm dPCR protocol with a reaction buffer containing

thermostable DNA polymerase, dNTPs, GE sample loading reagent (Fluidigm), and

the primers and probe necessary to carry out the universal Taqman

amplification/detection scheme. The samples were loaded in the chip and run on the

Biomark thermocycler for 45 cycles. Sample analysis was carried out using the

default parameters for dPCR analysis using the Fluidigm analysis software. The

129

quantitated libaries were then diluted to 2 x 106 molecules per microliter in 10 mM

TRIS with 0.02% Tween-20 and aliquotted for storage at -60°C.

Genome sequencing

DNA pyrosequencing of the shotgun and paired-end libraries was carried out

on the Roche 454 Genome Sequencer FLX instrument using "Titanium" chemistry.

Emulsion PCR was carried out with DNA:Bead ratios of 0.3:1 for the shotgun libraries

and 2.5:1 for the paired-end libraries.

Sequence assembly

Reads from the shotgun and paired end pyrosequencing runs were separated by

sample and trimmed using the sfffile tool (Roche) permitting one error in each 10 bp

barcode. The G+C content of the single-cell reads formed a major peak centered at

32.4%, about 2% lower than the average G+C content of N. maritimus (34.2%). The

dispersion of the read G+C content was essentially identical to that of N. maritimus,

with the latter sampled at the same average read length. The few reads with G+C

content greater than 46% were mapped to the N. maritimus genome, and mapping

reads (90% identity over 40 bases) were combined with the low G+C content reads for

subsequent analyses. The reads from the enrichment sample were then assembled de

novo using Newbler (Roche). All assembly operations were carried out with Newbler

v2.3 using default parameters. Paired end reads from the enrichment sample tagged

by the assembler as 'chimeric' were excluded from subsequent operations. Reads from

the single cells were then mapped to the initial enrichment assembly, and 'chimeric'

reads excluded. Contigs made up of single-cell reads were then sampled as 550-mers

with 50 base overlaps and included as reads in the final de novo assembly of the

130

single-cell data. A second, ‘consensus’, de novo assembly was also perfomed using

the single cell pyrosequencing reads (~120 Mb) and the reads from the enrichment

(~37Mb). A small number of short outlier contigs were excluded from the final

assembly due to the identification of sequences from known contaminants from the

MDA kit, such as Delftia acidovorans.

Gene annotation

Genes were predicted on the assembled genome sequence using the FgenesB

program (Softberry, Inc.). Automated FgenesB annotations were manually refined by

reviewing Reciprocal Best Hits (RBH) to Nitrosopumilus maritimus and

Cenarchaeum symbiosum A and BLAST searches against the GenBank database.

Individual proteins of interested were evaluated against “InterPro: the integrative

protein signature database” [68] using InterProScan. RBH were identified using the

blastp algorithm of Blast+ in conjunction with custom Matlab code. RBH are

classified as "high confidence” when the forward hit covered at least 60% of the query

sequence for the gene pair. In total, 1200 genes were functionally annotated and 607

have strong homology to another organism but no assigned function. 323 unannotated

genes (156 of which are 30 aa in length or less) were labeled as "hypothetical

proteins."

Genome sequence analysis

CRISPRFinder [69] and the CRISPR recognition tool (CRT) [70] was used to

search for Clustered Regularly Interspaced Short Palindromic Repeats (CRISPRs)

using the default settings. Synteny plots were created using the Nano+Bio-Center

computational biology software (http://nbc3.biologie.uni-kl.de/). Comparisons were

131

made at the DNA level using Nucmer and then the protein level using Promer by

translating the DNA sequences in all six reading frames. Analyses of nucleotide

content and sequence read statistics were performed using Blast+ on a 64 bit Linux

server and using custom Matlab and Perl code.

Acknowledgments

We thank Richard White for shotgun library preparation and sequencing and

Lolita Penland for sequencing the paired-end libraries. We greatly appreciate Eoghan

Harrington for allowing the use of and providing support for SmashCell in beta release

form (Harrington, Bioinformatics 2010, submitted), and would like to thank Jianbin

Wang and Joshua Weinstein for sharing their analysis of errors arising from MDA of

single template molecules. This work was supported in part by National Science

Foundation Grants MCB-0604270 and OCE-0847266 (to C.A.F.), NIH Director’s

Pioneer Award and NIH 5R01HG004863-02 (to S.R.Q.) and an EPA STAR Graduate

Fellowship (to A.C.M.). The funders had no role in study design, data collection and

analysis, decision to publish, or preparation of the manuscript.

132

Tables

assembly statistic Consensus Single cells only Cell 90507.23 Cell 90507.21 Cell 90507.3

raw reads 693,589 530,748 70,156 200,938 138,886

paired-end reads 33,725 26,953 3,018 517 2,388

raw bases 150,994,537 118,796,782 17,107,411 52,341,561 29,999,202

assembly bases 1,769,573 1,690,404 1,094,113 1,039,820 1,041,604

number of elements 31 136 355 253 348

number of scaffolds 2 26 68 76 83

scaffold bases 1,757,035 1,366,809 852,837 892,230 834,370

scaffold N50 1,363,261 155,965 20,543 16,443 14,147

largest scaffold 1,363,261 263,850 70,423 49,773 41,297

contig bases 1,740,160 1,664,128 1,090,610 1,039,764 1,031,668

contig N50 94,192 28,182 14,827 14,834 9,613

unscaffolded contigs 29 110 287 177 265

largest contig 230,803 135,296 70,423 49,773 27,947

Table 1. N. limnia assembly statistics for the consensus genome (metagenome and five single cells), single

cell coassembly (five single cells), and three individual cell assemblies.

Specifications N. limnia N. maritimus C. symbiosum A

Size, bp 1,769,573 1,645,259 2,045,086

Average G + C content 32 34 58

Predicted genes 2,179 1,840 2,066

Gene density, /kb 1 1 1

Average ORF length (bp) 717 825 924

Coding percentage, % 86 90 91

ORF content

Total number of ORFs 2,130 1,795 2,017

Predicted functional 1,200 1,082 994

Hypotheticals 930 713 1,023

RNA genes

16S-23S rRNA operon 1 1 1

5S rRNA 1 1 1

tRNAs 45 38 45

other RNA 2 2 1

Table 2. Genome features for N. limnia, N. maritimus, and C. symbiosum A.

133

Whole genome comparison N. maritimus C. symbiosum A

number of bases in corresponding regions 1,180,862 398,085

% identity of corresponding regions 76.30% 67.40%

whole genome average % identity* 68.10% 33.30%

Table 3. Whole genome nucleotide identity comparisons between N. limnia, N. maritimus ,

and C. symbiosum A.

* given 25% identity in non-corresponding regions

Motility Associated Genes Number of Genes

Flagella:

Archaeal flagella assembly protein J 1

Archaeal flagellins 5

Predicted ATPases involved in biogenesis of archaeal flagella 1

Type IV secretory pathway, VirB11 components, and related ATPases

involved in archaeal flagella biosynthesis 1

Chemotaxis:

Chemotaxis protein histidine kinase and related kinases 2

Chemotaxis protein; stimulates methylation of MCP proteins 1

Chemotaxis response regulator containing a CheY-like receiver domain

and a methylesterase domain 2

Chemotaxis signal transduction protein 1

CheY-like receiver 19

Methyl-accepting chemotaxis protein 2

Methylase of chemotaxis methyl-accepting proteins 1

Pilus:

Flp pilus assembly protein TadC 2

Table 4. N. limnia genes putatively associated with motility.

134

Individual gene comparison AA Identity NT Identity AA Identity NT Identity

16S rRNA n/a 97% n/a 94%

23S rRNA n/a 94% n/a 89%

ammonia monooxygenase subunit A 95% 89% 95% 78%

ammonia monooxygenase subunit B 92% 88% 76% 75%

ammonia monooxygenase subunit C 96% 91% 93% 81%

AMO-associated hypothetical protein** 77% 80% 73% 72%***

nitrite reductase (nirK) 85% 85% -- --

** ammonia monooxygenase operon-associated hypothetical protein

*** <90% sequence coverage

Table 5. Nucleotide (NT) and amino acid (AA) identities for key individual genes in N. limnia ,

N. maritimus, and C. symbiosum A.

N. limnia vs. N. maritimus N. limnia vs. C. symbiosum A

135

Figures

Figure 1. Phylogeny of ammonia-oxidizing archaea gene sequences. Neighbor-

joining phylogenetic tree of (A) archaeal amoA gene sequences and (B) archaeal 16S

rRNA gene sequences. Grey boxes highlight the putative low-salinity group.

Significant bootstrap values (≥ 50) from 1000 replicates are shown in italics at branch

nodes. Neighbor-joining bootstrap values are above the line and parsimony values are

below the line.

136

Figure 2. Assembly of the N. limnia datasets. Rarefaction analysis of the N. limnia

sequence data showing assembly of the enrichment, single cell, and combined

datasets.

137

Figure 3. Circular representation of the N. limnia genome. From the outer ring to the

inner ring: Scaffold breakpoints are indicated by the inside tick marks, predicted genes

on the forward strand, predicted genes the reverse strand, G+C content, GC skew, and

GGGT skew. On the outer two rings, protein coding sequences are colored gray,

tRNA genes in blue, rRNA genes in red, and noncoding RNA genes in green.

138

Figure 4. Comparison of three AOA genomes. (A) The top panel shows nucleotide

identity of blast hits between the reference genomes and the N. limnia genome. Hits

for N. maritimus and C. symbiosum A are colored according to the position in the

reference (query) genome. The arrow direction indicates the hit direction. On the

right, box plots summarize the distribution of hits from each reference genome to the

N. limnia genome. (B) The positions of the two largest consensus scaffolds are

shown. (C) The sequence novelty index, defined as 2 minus the sum of nucleotide

identity of blast hits to N. maritimus and C. symbiosum A at each position is plotted.

139

A histogram of the values is present at right. (D) The alignment depth in the final

assembly is shown. A histogram of the values is present at right.

140



Supplemental Figure 1. GC content of raw 454 reads from the 5 single cells and the

AOA enrichment. The GC content of the filtered set of reads used for the final

assembly of N. limnia and simulated reads from the N. maritimus genome are also

shown.

Supplemental Figure 2. Synteny plots of the N. limnia, N. maritimus, and C.

symbiosum A genomes.

Supplemental Table 1. Genes in N. limnia that are putatively involved in ammonia

oxidation and respiration (Table modified from Walker et al., 2010).

Supplemental Movie 1. Movie shows last stage of sorting cell 90507.21 (occurring

in the area on the device beyond the second gate valve). The laser trap is fixed near

the end of the fiducial arrow. In the movie, the cell is moved some 10s of micron

before the stage is stopped. The trap beam is interrupted when the stage is stopped,

freeing the cell to diffuse away. The pixels visible in the fiducial arrow are 0.106

micron wide.

141

Supplemental Dataset. Genbank file Nlimnia_consensus_CAT.gb contains the

concatenated N. limnia genome sequence and annotation. The accession number will

be provided when the database submission is complete.

142

Literature Cited

1. DeLong EF (1992) Archaea in coastal marine environments. Proc Natl Acad Sci U

S A 89: 5685-5689.

2. Fuhrman JA, McCallum K, Davis AA (1992) Novel major archaebacterial group

from marine plankton. Nature 356: 148-149.

3. Karner MB, DeLong EF, Karl DM (2001) Archaeal dominance in the mesopelagic

zone of the Pacific Ocean. Nature 409: 507-510.

4. Herndl GJ, Reinthaler T, Teira E, van Aken H, Veth C, et al. (2005) Contribution of

Archaea to total prokaryotic production in the deep Atlantic Ocean. Appl Environ

Microbiol 71: 2303-2309.

5. Teira E, Lebaron P, van Aken H, Herndl GJ (2006) Distribution and activity of

Bacteria and Archaea in the deep water masses of the North Atlantic. Limnology and

Oceanography 51: 2131-2144.

6. Kirchman DL, Elifantz H, Dittel AI, Malmstrom RR, Cottrell MT (2007) Standing

stocks and activity of Archaea and Bacteria in the western Arctic Ocean. Limnology

and Oceanography 52: 495-507.

143

7. Buckley DH, Graber JR, Schmidt TM (1998) Phylogenetic analysis of

nonthermophilic members of the kingdom Crenarchaeota and their diversity and

abundance in soils. Applied and Environmental Microbiology 64: 4333-4339.

8. Sandaa RA, Enger O, Torsvik V (1999) Abundance and diversity of Archaea in

heavy-metal-contaminated soils. Applied and Environmental Microbiology 65: 3293-

3297.

9. Ochsenreiter T, Selezi D, Quaiser A, Bonch-Osmolovskaya L, Schleper C (2003)

Diversity and abundance of Crenarchaeota in terrestrial habitats studied by 16S RNA

surveys and real time PCR. Environmental Microbiology 5: 787-797.

10. Kemnitz D, Kolb S, Conrad R (2007) High abundance of Crenarchaeota in a

temperate acidic forest soil. Fems Microbiology Ecology 60: 442-448.

11. Leininger S, Urich T, Schloter M, Schwark L, Qi J, et al. (2006) Archaea

predominate among ammonia-oxidizing prokaryotes in soils. Nature 442: 806-809.

12. Wuchter C, Abbas B, Coolen MJ, Herfort L, van Bleijswijk J, et al. (2006)

Archaeal nitrification in the ocean. Proc Natl Acad Sci U S A 103: 12317-12322.

144

13. Mincer TJ, Church MJ, Taylor LT, Preston C, Kar DM, et al. (2007) Quantitative

distribution of presumptive archaeal and bacterial nitrifiers in Monterey Bay and the

North Pacific Subtropical Gyre. Environmental Microbiology 9: 1162-1175.

14. Herfort L, Schouten S, Abbas B, Veldhuis MJW, Coolen MJL, et al. (2007)

Variations in spatial and temporal distribution of Archaea in the North Sea in relation

to environmental variables. Fems Microbiology Ecology 62: 242-257.

15. Beman JM, Popp BN, Francis CA (2008) Molecular and biogeochemical evidence

for ammonia oxidation by marine Crenarchaeota in the Gulf of California. Isme

Journal 2: 429-441.

16. Konstantinidis KT, Braff J, Karl DM, DeLong EF (2009) Comparative

Metagenomic Analysis of a Microbial Community Residing at a Depth of 4,000

Meters at Station ALOHA in the North Pacific Subtropical Gyre. Applied and


17. Church MJ, Wai B, Karl DM, DeLong EF (2010) Abundances of crenarchaeal

amoA genes and transcripts in the Pacific Ocean. Environmental Microbiology 12:

679-688.

145

18. Beman JM, Sachdeva R, Fuhrman JA (2010) Population ecology of nitrifying

Archaea and Bacteria in the Southern California Bight. Environmental Microbiology

12: 1282-1292.

19. Brochier-Armanet C, Boussau B, Gribaldo S, Forterre P (2008) Mesophilic

Crenarchaeota: proposal for a third archaeal phylum, the Thaumarchaeota. Nature

Reviews Microbiology 6: 245-252.

20. Spang A, Hatzenpichler R, Brochier-Armanet C, Rattei T, Tischler P, et al. (2010)

Distinct gene set in two different lineages of ammonia-oxidizing archaea supports the

phylum Thaumarchaeota. Trends Microbiol 18: 331-340.

21. Könneke M, Bernhard AE, de la Torre JR, Walker CB, Waterbury JB, et al. (2005)

Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature 437: 543-546.

22. Santoro AE, Francis CA, de Sieyes NR, Boehm AB (2008) Shifts in the relative

abundance of ammonia-oxidizing bacteria and archaea across physicochemical

gradients in a subterranean estuary. Environ Microbiol 10: 1068-1079.

23. Mosier A, Francis C (2008) Relative abundance and diversity of ammonia-


microbiology 10: 3002-3016.

146

24. de la Torre JR, Walker CB, Ingalls AE, Konneke M, Stahl DA (2008) Cultivation

of a thermophilic ammonia oxidizing archaeon synthesizing crenarchaeol. Environ

Microbiol 10: 810-818.

25. Hatzenpichler R, Lebedeva EV, Spieck E, Stoecker K, Richter A, et al. (2008) A

moderately thermophilic ammonia-oxidizing crenarchaeote from a hot spring. Proc

Natl Acad Sci U S A 105: 2134-2139.

26. Marcy Y, Ouverney C, Bik EM, Losekann T, Ivanova N, et al. (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: 11889-11894.

27. Woyke T, Tighe D, Mavromatis K, Clum A, Copeland A, et al. (2010) One

bacterial cell, one complete genome. Plos One 5: e10314.

28. Dick GJ, Andersson AF, Baker BJ, Simmons SL, Thomas BC, et al. (2009)

Community-wide analysis of microbial genome sequence signatures. Genome Biol 10:

R85.

29. Moin NS, Nelson KA, Bush A, Bernhard AE (2009) Distribution and diversity of

archaeal and bacterial ammonia oxidizers in salt marsh sediments. Appl Environ

Microbiol 75: 7461-7468.

147

30. Marcy Y, Ishoey T, Lasken RS, Stockwell TB, Walenz BP, et al. (2007) Nanoliter

reactors improve multiple displacement amplification of genomes from single cells.

PLoS Genet 3: 1702-1708.

31. Woyke T, Xie G, Copeland A, Gonzalez JM, Han C, et al. (2009) Assembling the

marine metagenome, one cell at a time. Plos One 4: e5299.

32. Dean FB, Nelson JR, Giesler TL, Lasken RS (2001) Rapid amplification of

plasmid and phage DNA using Phi 29 DNA polymerase and multiply-primed rolling

circle amplification. Genome Res 11: 1095-1099.

33. Rodrigue S, Malmstrom RR, Berlin AM, Birren BW, Henn MR, et al. (2009)

Whole genome amplification and de novo assembly of single bacterial cells. Plos One

4: e6864.

34. Ashkin A, Dziedzic JM, Bjorkholm JE, Chu S (1986) Observation of a single-

beam gradient force optical trap for dielectric particles. Opt Lett 11: 288.

35. Neuman KC, Block SM (2004) Optical trapping. Rev Sci Instrum 75: 2787-2809.

36. White RA, 3rd, Blainey PC, Fan HC, Quake SR (2009) Digital PCR provides

sensitive and absolute calibration for high throughput sequencing. BMC Genomics 10:

116.

148

37. Myllykallio H, Lopez P, Lopez-Garcia P, Heilig R, Saurin W, et al. (2000)

Bacterial mode of replication with eukaryotic-like machinery in a hyperthermophilic

archaeon. Science 288: 2212-2215.

38. Richter M, Rossello-Mora R (2009) Shifting the genomic gold standard for the

prokaryotic species definition. Proc Natl Acad Sci U S A 106: 19126-19131.

39. Booth IR, Edwards MD, Black S, Schumann U, Miller S (2007) Mechanosensitive

channels in bacteria: signs of closure? Nature Reviews Microbiology 5: 431-440.

40. Walker CB, de la Torre JR, Klotz MG, Urakawa H, Pinel N, et al. (2010)

Nitrosopumilus maritimus genome reveals unique mechanisms for nitrification and

autotrophy in globally distributed marine crenarchaea. Proc Natl Acad Sci U S A 107:

8818-8823.

41. Burg MB, Ferraris JD (2008) Intracellular organic osmolytes: function and

regulation. J Biol Chem 283: 7309-7313.

42. Horvath P, Barrangou R (2010) CRISPR/Cas, the immune system of bacteria and

archaea. Science 327: 167-170.

149

43. Margolin W (2005) FtsZ and the division of prokaryotic cells and organelles. Nat

Rev Mol Cell Biol 6: 862-871.

44. Lindas AC, Karlsson EA, Lindgren MT, Ettema TJ, Bernander R (2008) A unique

cell division machinery in the Archaea. Proc Natl Acad Sci U S A 105: 18942-18946.

45. Samson RY, Obita T, Freund SM, Williams RL, Bell SD (2008) A role for the

ESCRT system in cell division in archaea. Science 322: 1710-1713.

46. Holmes VF, Cozzarelli NR (2000) Closing the ring: links between SMC proteins

and chromosome partitioning, condensation, and supercoiling. Proc Natl Acad Sci U S

A 97: 1322-1324.

47. Lurz R, Grote M, Dijk J, Reinhardt R, Dobrinski B (1986) Electron microscopic

study of DNA complexes with proteins from the Archaebacterium Sulfolobus

acidocaldarius. EMBO J 5: 3715-3721.

48. Luo X, Schwarz-Linek U, Botting CH, Hensel R, Siebers B, et al. (2007) CC1, a

novel crenarchaeal DNA binding protein. J Bacteriol 189: 403-409.

49. Ouverney CC, Fuhrman JA (2000) Marine planktonic archaea take up amino acids.


150

50. Ingalls AE, Shah SR, Hansman RL, Aluwihare LI, Santos GM, et al. (2006)

Quantifying archaeal community autotrophy in the mesopelagic ocean using natural

radiocarbon. Proc Natl Acad Sci U S A 103: 6442-6447.

51. Varela MM, van Aken HM, Sintes E, Herndl GJ (2008) Latitudinal trends of

Crenarchaeota and Bacteria in the meso- and bathypelagic water masses of the Eastern

North Atlantic. Environ Microbiol 10: 110-124.

52. Arp, D., Chain, P., and Klotz, M. (2007) The impact of genome analyses on our

understanding of ammonia-oxidizing bacteria. Annual Review of Microbiology 61:

503-528.

53. Hooper A, Mai T, Balny C (2005) Kinetics of reduction by substrate or dithionite

and heme-heme electron transfer in the multiheme hydroxylamine oxidoreductase.

European Journal of Biochemistry 141: 565-571.

54. Venter JC, Remington K, Heidelberg JF, Halpern AL, Rusch D, et al. (2004)

Environmental genome shotgun sequencing of the Sargasso Sea. Science 304: 66-74.

55. Hallam SJ, Mincer TJ, Schleper C, Preston CM, Roberts K, et al. (2006) Pathways

of carbon assimilation and ammonia oxidation suggested by environmental genomic

analyses of marine Crenarchaeota. PLoS Biol 4: e95.

151

56. Hallam SJ, Konstantinidis KT, Putnam N, Schleper C, Watanabe Y, et al. (2006)

Genomic analysis of the uncultivated marine crenarchaeote Cenarchaeum symbiosum.

Proc Natl Acad Sci U S A 103: 18296-18301.

57. Martens-Habbena W, Berube P, Urakawa H, de La Torre J, Stahl D (2009)

Ammonia oxidation kinetics determine niche separation of nitrifying Archaea and

Bacteria. Nature 461: 976-979.

58. Burton SA, Prosser JI (2001) Autotrophic ammonia oxidation at low pH through

urea hydrolysis. Appl Environ Microbiol 67: 2952-2957.

59. Francis CA, Roberts KJ, Beman JM, Santoro AE, Oakley BB (2005) Ubiquity and

diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean.

Proc Natl Acad Sci U S A 102: 14683-14688.

60. Pernthaler A, Pernthaler J (2007) Fluorescence in situ hybridization for the

identification of environmental microbes. Methods Mol Biol 353: 153-164.

61. Stahl DA, Amann RI (1991) Development and application of nucleic acid probes

in bacterial systematics. In: Stackebrandt E, Goodfellow M, editors. Sequencing and

hybridization techniques in bacterial systematics. Chichester, England: John Wiley

and Sons. pp. 205-248.

152

62. Amann RI, Zarda B, Stahl DA, Schleifer KH (1992) Identification of individual

prokaryotic cells by using enzyme-labeled, rRNA-targeted oligonucleotide probes.


63. Daims H, Bruhl A, Amann R, Schleifer KH, Wagner M (1999) The domain-

specific probe EUB338 is insufficient for the detection of all Bacteria: development

and evaluation of a more comprehensive probe set. Systematic and Applied


64. Maddison, D.R., and Maddison, W.P. (2000) Macclade 4: Analysis of Phylogeny

and Character Evolution. Sunderland, MA, USA: Sinauer Associates.

65. DeSantis TZ, Jr., Hugenholtz P, Keller K, Brodie EL, Larsen N, et al. (2006)

NAST: a multiple sequence alignment server for comparative analysis of 16S rRNA

genes. Nucleic Acids Res 34: W394-399.

66. DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, et al. (2006)

Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible

with ARB. Appl Environ Microbiol 72: 5069-5072.

67. Ludwig W, Strunk O, Westram R, Richter L, Meier H, et al. (2004) ARB: a

software environment for sequence data. Nucleic Acids Res 32: 1363-1371.

153

68. Hunter S, Apweiler R, Attwood TK, Bairoch A, Bateman A, et al. (2009) InterPro:

the integrative protein signature database. Nucleic Acids Res 37: D211-215.

69. Grissa I, Vergnaud G, Pourcel C (2007) CRISPRFinder: a web tool to identify

clustered regularly interspaced short palindromic repeats. Nucleic Acids Res 35: W52-

57.

70. Bland C, Ramsey TL, Sabree F, Lowe M, Brown K, et al. (2007) CRISPR

recognition tool (CRT): a tool for automatic detection of clustered regularly

interspaced palindromic repeats. BMC Bioinformatics 8: 209.

154

Chapter 5. Physiology of a novel low-salinity type AOA reveals niche adaptation Annika C. Mosier1 and Christopher A. Francis1


155

Abstract

Ammonia oxidation in marine and terrestrial ecosystems plays a pivotal role in

the cycling of nitrogen and carbon. Recent discoveries have shown that ammonia-

oxidizing archaea (AOA) are both abundant and diverse in these systems, and yet very

little is known about their physiology. Here we report the physiology of a novel low-

salinity type AOA enriched from the San Francisco Bay estuary, ‘Candidatus

Nitrosoarchaeum limnia’ strain SFB1. N. limnia has a much slower growth rate than

Nitrosopumilus maritimus, the only AOA isolate described to date. There was no

change in growth rate or ammonia oxidation when N. limnia was grown under lower

oxygen conditions (5.5% oxygen). Although N. limnia was capable of growing at

75% seawater salinity, growth and ammonia oxidation were inhibited. Allylthiourea

(ATU), a known inhibitor of bacterial ammonia oxidation, inhibited ammonia

oxidation by N. limnia but did not affect the growth rate. We confirmed the presence

of flagella, as suggested by various flagellar genes in the N. limnia genome, using

electron microscopy. Together these findings support the hypothesis that N. limnia is

adapted to specific low-salinity niches where motility may provide an ecological

advantage.

156

Introduction

The recent discovery of ammonia-oxidizing archaea (AOA) capable of

converting ammonia to nitrite has raised many questions about their physiology and

ecology relative to ammonia-oxidizing bacteria (AOB). AOA are found in almost

every environment on Earth, including ocean waters, estuaries, salt marshes,

sediments, soils, hot springs, hydrothermal vents, caves, corals and sponges,

wastewater treatment plants, groundwater, lakes and rivers. AOA outnumber AOB in

many of these environments, often by orders of magnitude (based on quantitative PCR

estimates).

Phylogenetic analysis suggests that different AOA sequence types are found in

different habitats; for example, most soil sequences are phylogenetically distinct from

marine water column sequences. AOA phylogeny and abundance are often correlated

with specific environmental variables (e.g., Erguder et al., 2009; Schleper, 2010 and

references within). A broad-scale biogeography survey of over 8,000 archaeal

ammonia monooxygenase (amoA) sequences from GenBank confirmed that different

habitats contain distinct AOA populations, providing strong evidence for niche

partitioning (Biller et al., in prep).

Physiological studies on AOA cultures are required to further test functional

differences between distinct populations of AOA. Such studies have the potential to

highlight specific traits selected for different habitats. For instance, Nitrosopumilus

maritimus (the only AOA isolate described to date) has a remarkably high affinity for

ammonia and appears to be adapted to life under extreme nutrient limitation (Martens-

157

Habbena et al., 2009). N. maritimus may therefore be able to outcompete AOB in

environments with low ammonia concentrations.

Here, we describe the physiology of a novel, low-salinity type AOA enriched

from the San Francisco Bay estuary, ‘Candidatus Nitrosoarchaeum limnia’ strain

SFB1. The N. limnia genome revealed features that may be crucial for fitness in the

estuarine environment, including a substantial suite of genes for flagellar biosynthesis

and chemotaxis, and the presence of a large mechanosensitive channel protein

involved in protection against osmotic stress (Blainey et al., 2011). Our aim here was

to use the physiological profile of this organism to better understand its potential niche

distribution in the environment.


Description of the isolation source

The N. limnia enrichment culture was initiated from sediments in San

Francisco Bay at site SU001S in the northern part of the estuary (Blainey et al., 2011).

At the time of sampling, nutrient concentrations in the bottom water just above the

sediments were 2 µM ammonia, 14 µM nitrate, and 0.9 µM nitrite. Salinity was 7.9,

temperature was 21.6ºC, and sediment C:N ratio was 15.8 (data courtesy of the San

Francisco Bay Regional Monitoring Program). Interestingly, quantitative PCR

suggested that AOA were, on average, 34 times more abundant than AOB in this low-

salinity region of the estuary (Mosier and Francis, 2008). Archaeal amoA sequences

from this region fell into a phylogenetically distinct cluster (Mosier and Francis, 2008)

with sequences from other low-salinity sites, including groundwater (Reed et al.,

158

2010), lakes, rivers (direct GenBank submission; Liu,Z., Huang,S., Sun,G. and

Xu,M.), freshwater macrophytes (Herrmann et al., 2008), and the Douro River estuary

(Magalhaes et al., 2009). The N. limnia archaeal amoA sequence clustered in this

phylogenetic group and its 16S rRNA gene sequence also clustered with sequences

from freshwater ecosystems (Blainey et al., 2011).

Low salinity, estuarine ecotype

Biller et al (in prep) identified six distinct ecotypes amongst AOA amoA

sequences from coastal sediments, lakes, and rivers with strong signals from salinity

and environmental setting. The six ecotypes predominantly corresponded to

populations at 1) high salinity estuary sites; 2) low salinity estuary sites; 3) high

salinity surf zone sites; 4) low salinity surf zone sites; 5) high salinity sites within

estuaries, salt marshes, and heathland pools; and 6) low salinity lake sites. Over 94%

of the sequences from the low salinity region of the San Francisco Bay estuary where

N. limnia was isolated fell within the low-salinity, estuarine ecotype defined by Biller

et al. (in prep). N. limnia is more than 99% identical to sequences from the estuary

sediments (based on amoA genes), and therefore is representative of this low-salinity,

estuarine ecotype.

Habitat distribution of metagenomic blast hits to N. limnia proteins

In order to further explore the biogeography of the N. limnia ecotype, proteins

indentified in the N. limnia genome (Blainey et al., 2011) were queried against “All

Metagenomic ORF peptides” using blastp in CAMERA (Seshadri et al., 2007). Blastp

159

hits to each N. limnia protein were then plotted by habitat type (based on sample

metadata provided by CAMERA). Over 70% of the protien hits were to coastal

sequences (mangrove, estuary, lagoon, coastal ocean) and another 15% hit freshwater

sequences (Figure 1). Among the proteins involved in ammonia oxidation, carbon

cycling, and phophorus aquistion, 82% hit coastal and freshwater sequences. Only

11% of the N. limnia proteins hit open ocean sequences, including 84 hypothetical

proteins, transcription and translation proteins, radical SAM proteins, and others. This

is further evidence that marine AOA are genetically distinct from coastal AOA, which

likely translates into physiological differences as well. Less than 1% hit soil

sequences, which may in part be due to high sequence divergence between

marine/coastal AOA and soil AOA, as well as low representation of soil sequences in

the CAMERA database.

Physiology of N. limnia

N. limnia enrichment cultures were grown under varying conditions, including

low oxygen, high salinity, and in the presence of a nitrification inhibitor. The control

cultures of N. limnia grew chemoautotrophically by aerobic ammonia oxidation to

nitrite with near stoichiometric conversion of ammonia to nitrite (i.e., one mole of

ammonia was converted to one mole of nitrite) (Figure 2). The growth rate was

0.0063 h-1 with a doubling time of 110 h. N. limnia grew much slower than N.

maritimus, which has a growth rate of 0.027 h-1 and a doubling time of 26 h (Martens-

Habbena et al., 2009).

160

When N. limnia was grown under lower oxygen conditions (5.5% oxygen),

there was no significant change in either the oxidation of ammonia to nitrite, growth

rate (0.0065 h-1), or doubling time (107 h) (Figure 2). N. maritimus was unable to

grow under oxygen limitation (Martens-Habbena et al., 2009), although the oxygen

concentration that limited growth was not reported. The lower limit of oxygen that

would support growth of N. limnia is unclear at this time but the data suggest some

capacity to survive under low oxygen conditions.

The N. limnia genotype is most often found in low salinity environments,

although it can be found in environments with high salinity (e.g., salt marshes) or

environments that experience high salinity at some points during the year (e.g., San

Pablo Bay in the San Francisco Bay estuary where the salinity fluctuates seasonally

depending on the volume of freshwater, riverine flow relative to the marine surge). To

test whether N. limnia is tolerant of higher salinity, we grew the enrichment culture at

75% seawater salinity (salinities commonly seen in estuaries). N. limnia grew at high

salinity, yet there was a longer lag time, incomplete oxidation of ammonia to nitrite,

slower growth rate (0.0035 h-1), and longer doubling time (198 h) (Figure 2). Thus, it

appears that N. limnia is capable of surviving at more marine salinities but with

hindered growth that may make it susceptible to being outcompeted by other ammonia

oxidizers.

We also tested the sensitivity of N. limnia to allylthiourea (ATU), which is a

known inhibitor of bacterial ammonia oxidation. At concentration of 172 µM ATU,

N. limnia had no change in growth rate (0.0063 h-1) or doubling time (110 h) but did

show incomplete oxidation of ammonia (Figure 2). Ammonia oxidation by the

161

moderately thermophilic AOA, Nitrososphaera gargensis, was also partially inhibited

at 100 µM ATU (Hatzenpichler et al., 2008). Varied levels of inhibition of ammonia

oxidation were observed in marine water column nitrification rate assays with 86 µM

ATU (Santoro et al., 2010). Each of these ATU concentrations is known to

completely inhibit AOB (Hooper and Terry, 1973; Ginestet et al., 1998) and yet AOA

appear to be much less susceptible to its inhibitory effects. This may be due to

structural differences in the ammonia monoxygenase enzyme between AOA and AOB

since ATU acts directly on this protein. The use of ATU to inhibit nitrification in

marine and terrestrial studies should be reconsidered.

Nitrogen gas production

Nitrous oxide (N2O) is a highly potent greenhouse gas and is currently the

most important contributor to stratospheric ozone depletion (Ravishankara et al.,

2009). N2O concentrations in the atmosphere from natural and anthropogenic sources

may be rising at unprecedented rates (Codispoti, 2010). Ammonia oxidizers are one

of the primary contributors to natural N2O emissions. Under well-oxygenated

conditions, AOB produce N2O at low rates as a by-product of ammonia oxidation. At

low oxygen concentrations, AOB reduce nitrite to N2O via nitrite reductase (Nir) and

nitric oxide reductase (Nor) proteins in a process called nitrifier-denitrification (Arp

and Stein, 2003). Most AOB studied to date have genes coding for the copper-

containing form of nitrite reductase (nirK; Cantera and Stein, 2007).

It appears that many but not all AOA also have nirK genes, providing a

potential pathway for N2O production in AOA. While missing in C. symbiosum,

162

archaeal nirK genes have been identified in N. maritimus, as well as in soils, oceans,

stream biofilms, and lakes (Treusch et al., 2005; Yooseph et al., 2007; Bartossek et al.,

2010; Walker et al., 2010). One nirK gene (Nlim_1007) was identified in the N.

limnia genome (based on analysis of the conserved metal-binding residues within

multicopper oxidases, as described by Bartossek et al. 2010). The N. limnia nirK

sequence shares only 83% and 85% identity to the N. maritimus nirK sequences at the

amino acid level. Using degenerate PCR primers targeting archaeal nirK genes, we

detected a number of N. limnia-like nirK genes in sediments from North San Francisco

Bay (Lund & Francis, unpublished).

Because of the presence of a nirK gene in the N. limnia genome, we measured

N2O gas production from the enrichment cultures (Figure 3). N2O concentrations

increased more than 3.5-fold over the 53 day experiment. Interestingly, N2O

concentrations increased by almost 13-fold in the low oxygen treatment. Carbon

dioxide (CO2) concentrations also increased over the experiment but to a lesser extent

than N2O (average of 1.9-fold across all treatments versus an average of 6.1-fold for

N2O). Oxygen concentrations did not change significantly over time so the increases

in N2O and CO2 are likely not an artifact of the sampling setup (e.g., contamination by

air, vacuum created in the serum vials, etc.). Because the enrichment culture contains

approximately 15-20% bacteria, the production of N2O by N. limnia cannot be

confirmed. However, the N. limnia enrichment conditions are unlikely to support the

growth of other microorganisms known to produce N2O, including AOB, denitrifiers,

and methane oxidizers. A substantial fraction of upper-ocean N2O production may be

163

attributable to archaeal nitrification (Santoro et al., 2010), further suggesting a possible

role of AOA in N2O production.

Motility

The N. limnia genome contains over 38 genes associated with motility,

including those required for Flp pilus assembly, flagellar assembly, and chemotaxis

(Blainey et al., 2011). More than 50% of the motility genes identified in the genome

were found on one gene cassette (Figure 4) containing 13 genes associated with

chemotaxis followed by 8 genes associated with archaeal flagella. The other seven

non-motility genes within the cassette have no annotated function; however, five

genes contained signal peptide or transmembrane domains [identified by InterProScan

(Hunter et al., 2009)]. It is possible that these proteins are also involved in motility or

chemotaxis.

The presence of flagella was confirmed by transmission electron microscopy

(Figure 5). Because >85% of the cells in the enrichment are AOA and the majority of

cells viewed by TEM had flagella (and uniform morphology), we are confident that N.

limnia has a flagellum. The cells appeared as straight rods with an average diameter

of 0.24 µm and length of 0.77 µm (range 0.19 - 0.27 µm x 0.55 - 1.00 µm).

Motility appears to be a distinctive trait of N. limnia; N. maritimus and C.

symbiosum A both of which lack chemotaxis and flagella gene suites. N. limnia was

isolated from estuarine sediments, raising the possibility that it uses motility to

position itself in response to varying substrate and/or oxygen concentrations in the

surface sediments. In contrast, N. maritimus was isolated from gravel in a tropical fish

164

tank and C. symbiosum A is a symbiont of a marine sponge, where motility is perhaps

less essential. Interestingly, in the blastp comparison against all metagenomic proteins

(Figure 1), all of the N. limnia flagellar proteins, both pilus proteins, and 54% of the

chemotaxis proteins had highest hits with sequences from mangroves.

Phosphorus uptake

The N. limnia genome contains genes for phosphorus acquisition using the

high-affinity inorganic phosphate-specific transporter (Pst) system. The Pst system

(Wanner, 1993; Oganesyan et al., 2005) includes proteins that transfer inorganic

phosphorus through the inner membrane (PstA and PstC; Nlim_0923 and Nlim_0924),

an ATPase that energizes transport (PstB; Nlim_0922), a periplasmic inorganic

phosphorus-binding protein (PstS; Nlim_0927), and an uptake regulator (PhoU;

Nlim_0921, Nlim_1362, Nlim_1441, and Nlim_0925). N. limnia does not appear to

have the sensor or response regulator proteins in the Pst system (PhoR and PhoB). N.

limnia may also have the ability to use the low-affinity inorganic phosphate-specific

transporter (Pit) system. The Pit system consists of a single membrane protein

(putatively Nlim_0661 or Nlim_0662) and is the preferential transport system when

inorganic phosphorus is in excess (Wanner, 1993; Gebhard et al., 2009). Unlike N.

maritimus (Walker et al., 2010), the N. limnia genome has no genes for transport of

organic phosphorus. The northern region of San Francisco Bay where N. limnia was

isolated has lower phosphate concentrations (2.7 µM on average) than the rest of the

estuary (Wankel et al., 2006); however, phosphorus does not limit phytoplankton

165

productivity (Peterson et al., 1985; Jassby, 2008). Thus, N. limnia may not require the

utilization of organic phosphorus since inorganic phosphorus is in excess.

Carbon metabolism

Six autotrophic CO2 fixation pathways are known: the reductive pentose

phosphate cycle (Calvin–Benson–Bassham cycle), the reductive citric acid cycle

(Arnon–Buchanan cycle), the reductive acetyl-CoA pathway (Wood–Ljungdahl

pathway), the 3-hydroxypropionate/malyl-CoA cycle, the 3-hydroxypropionate/4-

hydroxybutyrate cycle (Berg et al., 2007), and the dicarboxylate/4-hydroxybutyrate

cycle (Huber et al., 2008). All Crenarchaeota and Thaumarchaeota (N. maritimus and

C. symbiosum A) appear to use either the 3-hydroxypropionate/4-hydroxybutyrate

cycle or the dicarboxylate/4-hydroxybutyrate cycle, but not both (based on

experimental evidence or genome annotations) (Hallam et al., 2006; Berg et al., 2007;

Jahn et al., 2007; Huber et al., 2008; Berg et al., 2010; Walker et al., 2010). It was

previously thought that the fundamental difference between these two autotrophic

pathways was their different sensitivity to oxygen: the 3-hydroxypropionate/4-

hydroxybutyrate cycle can tolerate oxygen, whereas the dicarboxylate/4-

hydroxybutyrate cycle is oxygen sensitive. However, recent evidence suggests that

the distribution of these pathways correlates with 16S rRNA-based phylogeny, rather

than with aerobic or anaerobic metabolism (Berg et al., 2010).

The N. limnia genome contains genes that putatively encode enzymes involved

in most steps of the 3-hydroxypropionate/4-hydroxybutyrate pathway (Supplemental

Figure 1). In this cycle, one acetyl-CoA and two bicarbonate molecules are converted

166

to succinyl-CoA via 3-hydroxypropionate. Succinyl-CoA is then reduced to 4-

hydroxybutyrate and converted into two acetyl-CoA molecules by the key enzyme 4-

hydroxybutyryl-CoA dehydratase. The N. limnia genome is missing enzymes

responsible for the conversion of malonyl-CoA to propionyl-CoA. However, as

proposed by Berg et al. (2007), various alcohol dehydrogenases, aldehyde

dehydrogenases, acyl-CoA synthetases, or enoyl-CoA hydratases may fulfill the same

functions.

Many enzymes of the tricarboxylic acid (TCA) cycle are encoded by the N.

limnia genome. The TCA cycle is involved in the oxidative breakdown of

carbohydrates, lipids, and proteins. The enzyme citrate synthase transfers the two-

carbon acetyl group in acetyl-CoA to the four-carbon compound oxaloacetate to form

the six-carbon compound citrate. Citrate is subsequently oxidized in a stepwise

manner to CO2, supplying NADH for oxidative phosphorylation and other metabolic

processes. Alternatively, the TCA cycle can run in the reverse direction producing

acetyl-CoA from two molecules of CO2 as a means of autotrophy. Acetyl-CoA is then

reduced to pyruvate, from which other central metabolites can be formed. In the

reductive TCA cycle, the enzyme citrate lyase is required to split citrate into acetyl-

CoA and oxaloacetate. Genes potentially encoding steps in both the oxidative and

reductive TCA cycle were identified in the N. limnia genome. However, only one

subunit of the citrate lyase protein, required for the reductive TCA cycle, was

identified. Additionally, genes encoding the enzyme in step five of the cycle between

2-oxoglutarate and succinyl-CoA (2-oxoglutarate ferredoxin oxidoreductase) could not

be confirmed. Taken together, N. limnia may utilize a horseshoe version of the TCA

167

cycle—in the oxidative direction from citrate to 2-oxoglutarate and in the reductive

direction from oxaloacetate to succinate—as described in C. symbiosum A (Hallam et

al., 2006).

Materials and Methods

Cultivation of ammonia oxidizing archaea

Enrichment cultures of Candidatus Nitrosoarchaeum limnia SFB1 were

maintained as previously described (Blainey et al., 2011). Briefly, the cultivation

media contained 500 µM ammonium chloride, 1 mL selenite/tungstate solution, 1 mL

vitamin solution (Balch et al., 1979), 1 mL sodium bicarbonate (1M), 10 mL KH2PO4

(4g g L-1), 1 mL trace metals (Biebl and Pfennig, 1978), and 988 mL artificial

seawater diluted to ~25% seawater salinity.

Physiology experiments

AOA enrichment cultures were grown in sealed serum vials in the dark at room

temperature. The AOA were grown under different treatment conditions: high salinity

media, freshwater media, ~5% oxygen, ~3% oxygen, ~2% oxygen, and with the

addition of 172 µM ATU, an ammonia oxidation inhibitor. Each treatment included

three replicates and a negative control containing only media (with the exception of

the allylthiourea treatment, which did not include a negative control). The high

salinity media was made as described above using artificial seawater diluted to ~75%

seawater salinity. The freshwater media was made as described above except that

minimal salts were added instead of artificial seawater (1 g L-1 NaCl, 0.4 g L-1

MgCl2•6H2O, 0.1 g L-1 CaCl2•2H2O and 0.5 g L-1 KCl; as described by de la Torre et

168

al., 2008). The N. limnia enrichment culture was innoculated into the media at a 1%

volume ratio.

Two mL of the culture was collected periodically for nutrient analysis and

frozen at –20ºC. Nitrite and ammonium concentrations were measured on a

QuikChem 8000 Flow Injection Analyzer (Lachat Instruments). Three mL of

headspace was sampled periodically to measure oxygen, nitrous oxide, carbon dioxide,

and methane gases using the Shimadzu 2014 gas chromatograph.

Flourescent in situ hybridization

For in situ hybridization, cultures were fixed in 2% formaldehyde and filtered

onto 0.2 µm polycarbonate membranes (Millipore). Catalyzed reporter deposition

fluorescence in situ hybridization (CARD-FISH; e.g., Pernthaler and Pernthaler, 2007)

was carried out with horseradish peroxidase (HRP)-labeled probes Arch915 (Stahl and

Amann, 1991) and EUB338 I-III (Amann et al., 1992; Daims et al., 1999). Filters

were treated with lysozyme or proteinase K for permeabilization of bacteria and

archaea, respectively.

Electron microscopy

Scanning and Transmission Electron Microscopy were performed at the Cell

Sciences Imaging Facility at Stanford University. For Scanning Electron Microscopy

(SEM), samples were fixed overnight in 2% glutaraldehyde with 4%

paraformaldehyde (PFA) in 0.1M sodium cacodylate buffer (pH 7.3). 100 µl of

preserved samples were allowed to settle onto poly-L-lysine coated 12mm circular

glass coverslips, before post-fixation (1 hr) in 1% aqueous Osmiumtetroxide. Samples

were then dehydrated in a graded ethanol series (50-70-80-90-100% ethanol; 15

169

minutes each), before critical point drying with liquid CO2 in a Tousimis Autosamdri-

814 apparatus (Tousimis, Rockville, MD, USA). Specimens were mounted onto 12-

mm aluminum stubs with colloidal graphite and sputter-coated with a 100 Å layer of

Au/Pd using a Denton Desk II Vacuum unit. Visualization was performed with a

Hitachi S-3400N Variable Pressure (VP) SEM operated at 10-15 kV, working distance

7 to 8 mm, and secondary electron detector under high-vacuum conditions.

For transmission electron microscopy (TEM) with negative staining, 10 µl

samples (unfixed or fixed in 4% PFA with 2% glutaraldehyde in 0.1M NaCacodylate,

pH 7.3) were spotted onto glow-discharged formvar-coated 100 mesh copper TEM

grids and allowed to settle for 5 minutes. Grids were then stained with 1% Uranyl

Acetate for 2 minutes, and dried overnight before visualization with a JEOL 1230

TEM, operated at 80 kV.

Fragment recruitment

Protein sequences indentified in the N. limnia genome (Blainey et al., 2011)

were blasted against the “All Metagenomic ORF peptides” database in CAMERA

(Seshadri et al., 2007). Blastp searches in CAMERA used the NCBI default blastall

parameters with one database alignment per query. Blast hits to each N. limnia protein

were then plotted by habitat type (based on sample metadata provided by CAMERA).

Acknowledgments

We thank L.M. Joubert at the Stanford Cell Sciences Imaging Facility for

performing electron microscopy. This work was funded by National Science

Foundations grants (OCE-0847266) to C.A.F.

170

Figures

Figure 1. Blastp analysis of N. limnia proteins against all metagenomic ORF peptides

(in CAMERA database). Sequence hits are color-coded by habitats from which the

samples were collected. N. limnia proteins are ordered along the length of the genome

(x-axis).

!"

#!"

$!"

%!"

&!"

'!"

(!"

)!"

*!"

+!"

#!!"

!" '!!" #!!!" #'!!" $!!!"

!"#$"%&'()"%*&+',-./0&!1'

!"#$%&'%(#2#3&"4%'(5',3#)"#")'/.3%6'."%6&7'38'6"%39"1'

,-./0"123.-"

4256-78."

9/3:2-;"

<26775"

=72/32>"[email protected]"

?A.5"[email protected]"

102>.",2>>"

B;C-730.-D2>"E.53"

=7-2>"

F.G26."F>:C6."

H@IC"4I5."J-2I526."

F7I>"

171

Figure 2. Ammonia oxidation by N. limnia under different growth conditions.

!"

#!"

$!!"

$#!"

%!!"

%#!"

&!!"

&#!"

'!!"

'#!"

#!!"

!" (" $'" %$" %)" &#" '%" '*" #+"

µ!"

#$%&"

'('"

,-'"

,.%"

!"

#!"

$!!"

$#!"

%!!"

%#!"

&!!"

&#!"

'!!"

'#!"

#!!"

!" (" $'" %$" %)" &#" '%" '*" #+"

µ!"

#$%&"

)*+"(,%-./"

,-'"

,.%"

!"

#!"

$!!"

$#!"

%!!"

%#!"

&!!"

&#!"

'!!"

'#!"

#!!"

!" (" $'" %$" %)" &#" '%" '*" #+"

µ!"

#$%&"

'01"

,-'"

,.%"

!"

#!"

$!!"

$#!"

%!!"

%#!"

&!!"

&#!"

'!!"

'#!"

#!!"

!" (" $'" %$" %)" &#" '%" '*" #+"

µ!"

#$%&"

23-4"5$63/37%"

,-'"

,.%"

172

Figure 3. Gas production from N. limnia enrichment cultures grown under different

conditions.

!"!#

!"$#

%"!#

%"$#

&"!#

&"$#

!# %!# &!# '!# (!# $!# )!#

!"#$%&'(%)"*+(,--./(

012'(

!"#$%&'(%)"*+(

*+*# *,-# ./01#234/5/67# 89:#+;70<5#

!#

%!!!#

&!!!#

'!!!#

(!!!#

$!!!#

)!!!#

=!!!#

>!!!#

!# %!# &!# '!# (!# $!# )!#

31$4%5(*"%)"*+(,--./(

012'(

31$4%5(*"%)"*+(

*+*# *,-# ./01#234/5/67# 89:#+;70<5#

173

Figure 4. Motility gene cassette on the N. limnia genome. Chemotaxis-associated

genes are colored in blue and flagellar-associated genes are colored in orange. Other

genes are color-coded in grey or brown, depending on the presence of signal peptide

or transmembrane regions in the peptide sequence.

Other (contains signal peptide and/or tranmembrane regions)

Chemotaxis Archaeal Flagella Other

Nlim_0478

Nlim_0466 Nlim_0467 Nlim_0468 Nlim_0469


Nlim_0474 Nlim_0475 Nlim_0476 Nlim_0477 Nlim_0478

Nlim_0485 Nlim_0486 Nlim_0487 Nlim_0488 Nlim_0489


Nlim_0479 Nlim_0480 Nlim_0481

Nlim_0483Nlim_0482 Nlim_0484

174

Figure 5. Transmission electron micrograph of a typical flagellated cell from the N.

limnia enrichment.

175


Supplemental Figure 1. a) Proposed autotrophic 3-hydroxypropionate/4-

hydroxybutyrate cycle (adapted from Berg 2007). b) Genes from N. limnia, N.

maritimus, and C. symbiosum A putatively involved in each step of the pathway.

acetyl-CoAmalonyl-CoA

malonate semialdehyde

3-hydroxypropionate

3-hydroxypropionyl-CoA

acryloyl-CoA

propionyl-CoA

methylmalonyl-CoA

succinyl-CoA

succinate semialdehyde

4-hydroxybutyrate

4-hydroxybutyryl-CoA

crotonyl-CoA

3-hydroxybutyryl-CoA

acetoacetyl-CoA

acetyl-CoA

HCO3-

HCO3-

A.

B.

1) acetyl-CoA carboxylase

2) malonyl-CoA reductase

3) malonate semialde-hyde reductase

4) 3-hydroxypropionyl-CoA synthetase

5) 3-hydroxypropionyl-CoA dehydratase

6) acryloyl-CoA reductase

7) propionyl-CoA carboxylase

8) methylmalonyl-CoA epimerase; methylmalonyl-

CoA mutase9) succinyl-CoA reductase

10) succinate semialde-hyde reductase

11) 4-hydroxybutyryl-CoA synthetase

12) 4-hydroxybutyryl-CoA dehydratase

13) crotonyl-CoA hydratase

14) 3-hydroxybutyryl-CoA dehydrogenase

15) acetoacetyl-CoA !-ketothiolase

1 master1195, master1194, master1193 Nmar_0272, Nmar_0273, Nmar_0274 CENSYa_1660, CENSYa_1661, CENSYa_1662

2 unidenti"ed unidenti"ed unidenti"ed





7 master1195, master1194, master1193 Nmar_0272, Nmar_0273, Nmar_0274 CENSYa_1660, CENSYa_1661, CENSYa_1662

8 master0650,master0656, master0652 Nmar_0953, Nmar_0958, Nmar_0954 CENSYa_0475, CENSYa_0482, CENSYa_0476

9 master1367, master0974, master2129 Nmar_1608 unidenti"ed

10 master0812, master2009 Nmar_1110, Nmar_0161 CENSYa_0549

11 master1257 Nmar_0206 CENSYa_0021



14 master0708 Nmar_1028 CENSYa_041315 master0440, master1392 Nmar_0841, Nmar_1631 CENSYa_0587, CENSYa_1780

Step N. limnia N. maritimus putative homolog C. symbiosusm A putative homolog

176

Literature Cited

Amann, R.I., Zarda, B., Stahl, D.A., and Schleifer, K.H. (1992) Identification of

individual prokaryotic cells by using enzyme-labeled, rRNA-targeted oligonucleotide

probes. Applied and Environmental Microbiology 58: 3007-3011.

Arp, D.J., and Stein, L.Y. (2003) Metabolism of inorganic N compounds by ammonia-

oxidizing bacteria. Crit Rev Biochem Mol Biol 38: 471-495.

Balch, W.E., Fox, G.E., Magrum, L.J., Woese, C.R., and Wolfe, R.S. (1979)

Methanogens: Reevaluation of a unique biological group. Microbiological Reviews

43: 260-296.

Bartossek, R., Nicol, G.W., Lanzen, A., Klenk, H.-P., and Schleper, C. (2010)

Homologues of nitrite reductases in ammonia-oxidizing archaea: diversity and

genomic context. Environmental Microbiology 12: 1075-1088.

Berg, I., Kockelkorn, D., Buckel, W., and Fuchs, G. (2007) A 3-hydroxypropionate/4-

hydroxybutyrate autotrophic carbon dioxide assimilation pathway in Archaea. Science

318: 1782-1786.

Berg, I.A., Ramos-Vera, W.H., Petri, A., Huber, H., and Fuchs, G. (2010) Study of the

distribution of autotrophic CO2 fixation cycles in Crenarchaeota. Microbiology

(Reading, Engl) 156: 256-269.

177

Biebl, H., and Pfennig, N. (1978) Growth yields of green sulfur bacteria in mixed

cultures with sulfur and sulfate reducing bacteria. Archives of Microbiology 117: 9-16.

Blainey, P.C.*, Mosier, A.C.*, Potanina, A., Francis, C.A., and Quake, S.R. (2011)

Genome of a Low-Salinity Ammonia-Oxidizing Archaeon Determined by Single-Cell

and Metagenomic Analysis. PLoS ONE 6: e16626. *contributed equally.

Cantera, J., and Stein, L. (2007) Molecular diversity of nitrite reductase genes (nirK)

in nitrifying bacteria. Environmental Microbiology 9: 765-776.

Codispoti, L.A. (2010) Oceans. Interesting times for marine N2O. Science 327: 1339-

1340.

Daims, H., Brühl, A., Amann, R., Schleifer, K.H., and Wagner, M. (1999) The

domain-specific probe EUB338 is insufficient for the detection of all Bacteria:

development and evaluation of a more comprehensive probe set. Systematic and

Applied Microbiology 22: 434-444.

de la Torre, J., Walker, C., Ingalls, A., Konneke, M., and Stahl, D. (2008) Cultivation

of a thermophilic ammonia oxidizing archaeon synthesizing crenarchaeol.


178

Erguder, T.H., Boon, N., Wittebolle, L., Marzorati, M., and Verstraete, W. (2009)

Environmental factors shaping the ecological niches of ammonia-oxidizing archaea.

FEMS Microbiology Reviews 33: 855-869.

Gebhard, S., Ekanayaka, N., and Cook, G. (2009) The low-affinity phosphate

transporter PitA is dispensable for in vitro growth of Mycobacterium smegmatis.

BMC microbiology 9: 254.

Ginestet, P., Audic, J., Urbain, V., and Block, J. (1998) Estimation of nitrifying

bacterial activities by measuring oxygen uptake in the presence of the metabolic

inhibitors allylthiourea and azide. Applied and Environmental Microbiology 64: 2266-

2268.

Giri, A.V., Anishetty, S., and Gautam, P. (2004) Functionally specified protein

signatures distinctive for each of the different blue copper proteins. BMC

Bioinformatics 5: 127.

Hallam, S., Mincer, T., Schleper, C., Preston, C., Roberts, K., Richardson, P., and

DeLong, E. (2006) Pathways of carbon assimilation and ammonia oxidation suggested

by environmental genomic analyses of marine Crenarchaeota. PLoS Biology 4: e95.

179

Hatzenpichler, R., Lebedeva, E., Spieck, E., Stoecker, K., Richter, A., Daims, H., and

Wagner, M. (2008) A moderately thermophilic ammonia-oxidizing crenarchaeote

from a hot spring. Proceedings of the National Academy of Sciences 105: 2134-2139.

Herrmann, M., Saunders, A.M., and Schramm, A. (2008) Archaea dominate the

ammonia-oxidizing community in the rhizosphere of the freshwater macrophyte

Littorella uniflora. Applied and Environmental Microbiology 74: 3279-3283.

Hooper, A.B., and Terry, K.R. (1973) Specific inhibitors of ammonia oxidation in

Nitrosomonas. Journal Of Bacteriology 115: 480-485.

Huber, H., Gallenberger, M., Jahn, U., Eylert, E., Berg, I.A., Kockelkorn, D. et al.

(2008) A dicarboxylate/4-hydroxybutyrate autotrophic carbon assimilation cycle in the

hyperthermophilic Archaeum Ignicoccus hospitalis. Proceedings of the National

Academy of Sciences 105: 7851-7856.

Hunter, S., Apweiler, R., Attwood, T.K., Bairoch, A., Bateman, A., Binns, D. et al.

(2009) InterPro: the integrative protein signature database. Nucleic Acids Research 37:

D211-215.

Jahn, U., Huber, H., Eisenreich, W., Hügler, M., and Fuchs, G. (2007) Insights into the

autotrophic CO2 fixation pathway of the archaeon Ignicoccus hospitalis:

180

comprehensive analysis of the central carbon metabolism. Journal Of Bacteriology

189: 4108-4119.

Jassby, A. (2008) Phytoplankton in the Upper San Francisco Estuary: Recent biomass

trends, their causes and their trophic significance. San Francisco Estuary & Watershed

Science 6: 1–24.

Magalhaes, C.M., Machado, A., and Bordalo, A.A. (2009) Temporal variability in the

abundance of ammonia-oxidizing bacteria vs. archaea in sandy sediments of the Douro

River estuary, Portugal. Aquatic Microbial Ecology 56: 13-23.

Martens-Habbena, W., Berube, P.M., Urakawa, H., De La Torre, J.R., and Stahl, D.A.

(2009) Ammonia oxidation kinetics determine niche separation of nitrifying Archaea

and Bacteria. Nature 461: 976-979.

Mosier, A., and Francis, C. (2008) Relative abundance and diversity of ammonia-



Oganesyan, V., Oganesyan, N., Adams, P., Jancarik, J., Yokota, H., Kim, R., and Kim,

S. (2005) Crystal structure of the "PhoU-like" phosphate uptake regulator from

Aquifex aeolicus. Journal Of Bacteriology 187: 4238.

181

Pernthaler, A., and Pernthaler, J. (2007) Fluorescence in situ hybridization for the

identification of environmental microbes. Methods Mol Biol 353: 153-164.

Peterson, D., Smith, R., Hager, S., Harmon, D., Herndon, R., and Schemel, L. (1985)

Interannual variability in dissolved inorganic nutrients in Northern San Francisco Bay

Estuary. Hydrobiologia 129: 37-58.

Ravishankara, A.R., Daniel, J.S., and Portmann, R.W. (2009) Nitrous Oxide (N2O):

The Dominant Ozone-Depleting Substance Emitted in the 21st Century. Science 326:

123-125.

Reed, D.W., Smith, J.M., Francis, C.A., and Fujita, Y. (2010) Responses of ammonia-

oxidizing bacterial and archaeal populations to organic nitrogen amendments in low-

nutrient groundwater. Applied and Environmental Microbiology 76: 2517-2523.

Santoro, A.E., Casciotti, K.L., and Francis, C.A. (2010) Activity, abundance and

diversity of nitrifying archaea and bacteria in the central California Current.


Schleper, C. (2010) Ammonia oxidation: different niches for bacteria and archaea?

The ISME Journal 4: 1092-1094.

182

Seshadri, R., Kravitz, S.A., Smarr, L., Gilna, P., and Frazier, M. (2007) CAMERA: a

community resource for metagenomics. PLoS Biology 5: e75.

Stahl, D., and Amann, R. (1991) Development and application of nucleic acid probes

in bacterial systematics. In Nucleic acid techniques in bacterial systematics.

Stackebrandt, E., and Goodfellow, M. (eds). Chichester, England: John Wiley & Sons

Ltd., pp. 205–248.

Treusch, A., Leininger, S., Kletzin, A., Schuster, S., Klenk, H., and Schleper, C.




Walker, C.B., de la Torre, J.R., Klotz, M.G., Urakawa, H., Pinel, N., Arp, D.J. et al.

(2010) Nitrosopumilus maritimus genome reveals unique mechanisms for nitrification

and autotrophy in globally distributed marine crenarchaea. Proc Natl Acad Sci USA

107: 8818-8823.

Wankel, S., Kendall, C., Francis, C., and Paytan, A. (2006) Nitrogen sources and


approach. Limnology and Oceanography 51: 1654-1664.

183

Wanner, B.L. (1993) Gene regulation by phosphate in enteric bacteria. J Cell Biochem

51: 47-54.

Yooseph, S., Sutton, G., Rusch, D.B., Halpern, A.L., Williamson, S.J., Remington, K.

et al. (2007) The Sorcerer II Global Ocean Sampling expedition: expanding the

universe of protein families. PLoS Biology 5: e16.

184

Chapter 6. Abundance and Potential Rates of Denitrifiers Across the San Francisco Bay Estuary Annika C. Mosier1 and Christopher A. Francis1

1Department of Environmental Earth System Science, Stanford University, Stanford, CA 94305-4216, USA Published as: Mosier, A.C., and C.A. Francis. 2010. Denitrifier abundance and activity across the San Francisco Bay estuary. Environmental Microbiology Reports 2(5): 667-676. Note: An additional paragraph that did not appear in this publication was added to the Environmental context section (see Footnote).

185

Abstract

Over 50% of external dissolved inorganic nitrogen inputs to estuaries are

removed by denitrification—the microbial conversion of nitrate to nitrogen gas under

anaerobic conditions. In this study, denitrifier abundance, potential rates, and

community structure were examined in sediments from the San Francisco Bay estuary.

Abundance of nirK genes (encoding Cu-containing nitrite reductase) ranged from 9.7

x 103 to 4.4 x 106 copies per gram sediment, while the abundance of nirS genes

(encoding cytrochrome-cd1 nitrite reductase) ranged from 5.4 x 105 to 5.4 x 107 copies

per gram sediment. nirK gene abundance was highest in the riverine North Bay,

whereas nirS gene abundance was highest in the more marine Central and South Bays.

Denitrification potential (DNP) rate measurements were highest in the San Pablo and

Central Bays and lowest in the North Bay. nirS-type denitrifiers may be more

biogeochemically important than nirK-type denitrifiers in this estuary because DNP

rates were positively correlated with nirS abundance, nirS abundance was higher than

nirK abundance at every site and time point, and nirS richness was higher than nirK

richness at every site. Statistical analyses demonstrated that salinity, organic carbon,

nitrogen and several metals were key factors influencing denitrification rates, nir

abundance, and community structure. Overall, this study provides valuable new

insights into the abundance, diversity and biogeochemical activity of estuarine

denitrifying communities and suggests that nirS-type denitrifiers likely play an

important role in nitrogen removal in San Francisco Bay, particularly at high salinity

sites.

186

Introduction

Denitrifying bacteria are widespread in coastal and estuarine environments and

account for a significant reduction of external inorganic nitrogen inputs (Seitzinger,

1988). Under anaerobic conditions, denitrifying bacteria reduce nitrate to nitrogen

gases, which are rapidly released to the atmosphere. Denitrification thereby

diminishes the amount of bioavailable nitrogen and curtails the harmful effects of

nitrogen pollution in estuaries, such as stimulated phytoplankton production, depleted

dissolved oxygen, declines in fish populations and loss of biodiversity (e.g., Bricker et

al., 1999; Rabalais, 2002; Diaz and Rosenberg, 2008). Globally, estuarine

denitrification removes approximately 8 Tg of nitrogen per year (Seitzinger et al.,

2006). Key factors known to influence the occurrence and rates of denitrification

include the supply of nitrate and organic carbon, oxygen concentration, temperature,

and pH, among others (e.g., Christensen et al., 1990; Kemp et al., 1990; Nielsen et al.,

1995; Cornwell et al., 1999; Saleh-Lakha et al., 2009). However, despite the well-

established biogeochemical significance of denitrification in estuarine systems,

questions remain about how the underlying microbial communities relate to key

environmental variables and denitrification activity, particularly in nutrient-rich

systems like San Francisco Bay.

Efforts to characterize denitrifying communities are most often directed toward

analysis of denitrification functional genes rather than 16S rRNA genes, because

denitrifiers are found within a variety of phylogenetically unrelated groups from over

50 genera (Zumft, 1997). Nitrite reductase is a particularly important enzyme in the

denitrification pathway because it catalyzes the first committed step to a gaseous

187

product (the reduction of nitrite to nitric oxide; Zumft, 1997). The nitrite reduction

step distinguishes denitrification from other forms of nitrate metabolism, which can be

carried out by a wide range of organisms (Shapleigh, 2006). Nitrite reductase occurs

in two structurally different but functionally equivalent forms: NirK, containing

copper; and NirS, containing iron (cytochrome-cd1). nirK and nirS genes have been

used to examine denitrifying communities in a variety of environments (e.g., Braker et

al., 2000; Braker et al., 2001; Nogales et al., 2002; Prieme et al., 2002; Liu et al.,

2003; Yan et al., 2003; Jayakumar et al., 2004; Sharma et al., 2005; Hannig et al.,

2006; Santoro et al., 2006; Tiquia et al., 2006; Oakley et al., 2007; Tamegai et al.,

2007). Although these studies have resulted in an extensive database of nitrite

reductase sequences, relatively few nirK sequences have been recovered from

estuarine ecosystems and very few studies have attempted to quantify both nirK and

nirS genes.

The relative contribution of denitrification to nitrogen (N) removal from

aquatic systems has been drawn into question with the discovery of anammox as an

alternate N-loss pathway (the anaerobic oxidation of ammonium and nitrite to N2 gas).

Anammox has been shown to be a major source of N-loss in oxygen-deficient marine

water columns (e.g., Kuypers et al., 2005; Francis et al., 2007; Hamersley et al., 2007;

Jensen et al., 2008; Lam et al., 2009). However, in estuaries the measured

contribution of N2 formation from anammox relative to denitrification has been low.

N-loss from anammox ranged from 0 to 25% in several estuaries including the Thames

Estuary (Trimmer et al., 2003), Randers and Norsminde Fjords (Risgaard-Petersen et

al., 2004), Chesapeake Bay (Rich et al., 2008), and Plum Island Sound (Koop-

188

Jakobsen and Giblin, 2009). Based on these studies, denitrification appears to

represent the dominant nitrogen removal process in estuarine sediments.

In this study, we determined the abundance, community structure, and

biogeochemical activity of denitrifiers in San Francisco Bay—the largest estuary on

the west coast of the United States (encompassing an area of ~1200 km2 and draining

a watershed that covers 40% of the area of California). Specifically, we combined

potential denitrification rate measurements, quantitative PCR estimates, and functional

gene sequencing to characterize the variation of nirK- and nirS-type denitrifiers across

the estuarine gradient and assess the relationship with environmental factors in the

estuary.


Environmental context

Denitrifiers were analyzed within sediments collected annually during the

summers (July-August) of 2004-2008. Sediments and bottom water were also

analyzed for ancillary variables at each site and time point. Sampling sites covered

four distinct regions: North Bay (BG20, BG30, BF21), San Pablo Bay (BD31, SP19),

Central Bay (BC11, CB18), and South Bay (SB02, BA41, SB18, LSB2, BA10) (Fig.

1). The North Bay originates at the San Joaquin and Sacramento River Delta and

encompasses classic estuarine gradients of salinity and nutrients depending on the

relative amounts of tidal intrusion and freshwater flow. The South Bay is weakly

mixed with wastewater treatment plant effluent as the primary source of freshwater

(Hager and Schemel, 1996), resulting in more marine salinities year-round. Both of

189

these hydrologically distinct portions of the estuary exchange water in the Central

Bay, which has a direct connection to the Pacific Ocean.

During the summertime sampling from 2004 to 2008, bottom water salinity

was less than 1 psu at the riverine sites in the North Bay (BG20 and BG30) and

increased to greater than 25 psu in the Central and South Bays (Table S1). Nitrate,

nitrite, and ammonium concentrations measured across 44 to 48 sites in the estuary

varied little over the five-year period (Fig. 2). Nitrate ranged from 0.02-87.6 µM,

nitrite ranged from 0.03-4.5 µM, and ammonium ranged from 0.2-19 µM. Nitrate and

nitrite concentrations were relatively constant throughout most of the estuary and then

peaked in the South Bay, whereas ammonium concentrations were consistently low in

the North Bay and quite variable throughout the rest of the estuary. Over the five

years, the highest concentrations of nitrate, nitrite, ammonium, and total DIN occurred

in the South Bay, which has long residence times and high loading from wastewater

treatment plants (Kimmerer, 2004; Wankel et al., 2006).

Nutrients and salinity have been shown to display linear relationships during

periods of high river flow to northern San Francisco Bay (Hager and Schemel, 1992;

Wilkerson et al., 2006), because the rate of supply of nutrients by rivers is much larger

than internal estuarine supply and removal rates (Peterson et al. 1985). During the

summers of 2004 to 2008, our data along the entire estuary (not just the northern

portion) showed no linear relationship between salinity and bottom water nitrate,

nitrite, nitrate plus nitrite, or total DIN. This is likely because river flow is low during

the summer months and the northern and southern portions of the bay are effectively

hydrodynamically distinct (e.g., (Wankel et al., 2006). However, ammonia increased

190

linearly with salinity in the estuary (Pearson correlation: r = 0.44, P = 0.01, n = 35).

Higher ammonium concentrations in the high salinity region of the estuary may be the

result of greater nutrient inputs (e.g., from wastewater treatment plants), less

biological uptake of ammonium, or greater rates of ammonium production via organic

matter mineralization. When analyzed from San Pablo Bay to South Bay (excluding

the North Bay), salinity increased linearly with decreasing concentrations of nitrate

(Pearson correlation: r = -0.58, P = 0.01, n = 20), nitrate plus nitrite (Pearson

correlation: r = -0.58, P = 0.01, n = 20), and total DIN (Pearson correlation: r = -0.52,

P = 0.02, n = 20). This pattern was also seen in San Francisco Bay across a larger

number of sites in 2004 (Wankel et al., 2006)†.

Sediment characteristics analyzed from 2004-2007 were found to vary across

the estuary gradient (Table S2). Total C:N sediment ratios were consistently high in

the North Bay (average 15.7) and low in the rest of the estuary (average 8.9). The

percentage of clays (R2= 0.68, P < 0.005, n = 28), fines (R2= 0.51, P < 0.005, n = 28),

and silt (R2= 0.17, P = 0.03, n = 28) increased logarithmically with distance from the

head of the estuary from North to South Bay, whereas the percent sand decreased (R2=

0.64, P < 0.005, n = 28). Distributions of 12 different metals (Ag, As, Cd, Cu, Fe, Hg,

MeHg, Mn, Ni, Pb, Se, and Zn) along the estuary varied considerably (Table S2).

Abundance of nitrite reductase genes

Based on in silico analysis of our extensive 2004 database of estuarine nirK

and nirS sequences from San Francisco Bay (see below) and GenBank sequences, we

† This paragraph did not appear in the original publication.

191

designed and optimized quantitative PCR (qPCR) primers and assays for quantifying

the abundance of these genes. Abundance of nitrite reductase genes (nirK and nirS) in

surface sediments was determined at 7 sites spanning the estuary gradient during the

summers of 2004-2008. Abundance of nirK genes ranged from 9.7 x 103 to 4.4 x 106

copies per gram sediment and from 3.7 x 103 to 5.8 x 105 copies per µg of DNA (Fig.

3 and Table S3). There was no statistically significant change in nirK abundance over

the five-year period at each individual site or across the entire estuary. When

averaged over the five-year period, abundance of nirK genes per gram of sediment

was highest at BG20 and BF21 in the North Bay and consistently low across the San

Pablo, Central and South Bay sites.

Abundance of nirS genes ranged from 5.4 x 105 to 5.4 x 107 copies per gram

sediment and from 3.6 x 105 to 5.7 x 106 copies per µg of DNA (Fig. 3 and Table S3).

These values are within the range reported in Chesapeake Bay (2 x 105 to 7 x 107

copies per gram sediment; Bulow et al., 2008) and the Colne estuary (~104 to 107

copies per gram sediment; Smith et al., 2007). In San Francisco Bay, there was no

statistically significant change in nirS abundance over the five-year period at each

individual site or across the entire estuary. During all five years, abundance of nirS

genes was highest in the Central and South Bays and lowest in the North Bay.

Notably, at every site and time point, nirS copies per gram of sediment were higher

than nirK copies (ranging from 5 to 266-fold higher), a trend that was observed in the

subtropical Fitzroy estuary (Abell et al., 2009).

When averaged over the five-year period, nirK gene abundance was highest in

the North Bay, whereas nirS gene abundance was highest in the Central and South

192

Bays and lowest in the North Bay. nirS gene copy numbers were also highest in the

mesohaline sediments from central Chesapeake Bay (Bulow et al., 2008), whereas

nirS gene copy numbers were highest at the head of the Colne estuary (Smith et al.,

2007). Both nirK and nirS gene abundances were relatively constant throughout the

subtropical Fitzroy estuary (Abell et al., 2009). Although geographic,

physical/chemical, and biological differences between these estuaries likely confound

the identification of common patterns and trends in denitrifier abundance, this

quantitative nirK/S dataset represents the first insight into the spatial and inter-annual

distribution of denitrifying genes in the San Francisco Bay estuary.

Denitrification rate potentials

To assess the metabolic capacity of San Francisco Bay sediment communities

to carry out denitrification, denitrification potential (DNP) rate measurements were

performed at 7 sites (same sites used for qPCR anlayses) during the summers of 2007

and 2008. DNP rates ranged from 0.1 to 54 mg N2O-N kg-1 day-1 and from 0.3 to 44

mg N2O-N kg-1 day-1 upon the addition of chloramphenicol (Fig. 4). Chloramphenicol

was added to inhibit synthesis of new denitrifying enzymes while the activity of

previously existing denitrifying enzymes was being measured (Smith and Tiedje,

1979). DNP rates were higher in 2007 than in 2008 at all sites except BA41. DNP

rates were consistently highest in the San Pablo and Central Bays and lowest in the

North Bay, while the South Bay had intermediate rates.

In San Francisco Bay sediments, DNP rates were positively correlated with

nirS abundance (copies per gram sediment; Spearman correlation: ρ = 0.57, P = 0.03,

193

n = 14), whereas nirK abundance (copies per gram sediment) showed no correlation

with DNP rates. However, when expressed in terms of copies per µg DNA, nirK was

negatively correlated with DNP rates (Spearman correlation: ρ = -0.62, P = 0.02, n =

14). Several lines of evidence suggest that nirS-type denitrifiers may be

biogeochemically important in this estuary, including the fact that DNP rates were

positively correlated with nirS abundance, nirS abundance was higher than nirK

abundance at every site and time point, and nirS richness was higher than nirK

richness at every site examined (Fig. S4, Table S4, and Supplementary Material).

Together, this may suggest that denitrifiers harboring NirS play a greater role in N-

removal from the system compared to denitrifiers harboring NirK.

Nitrite reductase gene richness and phylogenetic diversity

nirK and nirS gene sequences were analyzed at 12 sites from 2004 (7 sites used

for the multi-year qPCR analyses and 5 additional sites distributed throughout the

estuary). From all clone libraries combined, we recovered 112 unique nirK

operational taxonomic units (OTUs) and 169 unique nirS OTUs defined at a 5%

distance cutoff (from a total of from 383 nirK sequences and 360 nirS sequences; 28-

45 sequences per site) (Supplementary Material and Table S4).

Across all sites, nirK sequences shared considerable phylogenetic similarity

with sequences from the water column and sediments of two lakes and the Baltic Sea

(O.-S. Kim, P. Junier, J. Imhoff and K.-P. Witzel, unpublished, direct GenBank

submission) (Fig. S1). Approximately 80% of the nirK sequences from the North Bay

riverine BG20 and BG30 sites grouped closely (73-100% amino acid identity) with

194

sequences from soil (e.g., Throback et al., 2007), activated sludge (e.g., Hallin et al.,

2006), and lakes (e.g., Junier et al., 2008). The majority of the sequences from the rest

of the estuary fell into the previously described nirK I-IV groups (Santoro et al.,

2006), but did not correspond to particular salinity or nutrient conditions as observed

in coastal aquifer sediments (Santoro et al., 2006).

nirS sequences from San Francisco Bay grouped phylogenetically with

sequences from many other estuaries (Fig. S2), including: Chesapeake Bay (Francis,

O’Mullan, Cornwell, and Ward, unpublished, direct GenBank submission; Bulow et

al., 2008); Bahía del Tóbari, Mexico (J.M. Beman and C.A. Francis, unpublished;

direct GenBank submission); Changjiang estuary (H. Dang and J. Wang, unpublished,

direct GenBank submission); the Colne estuary (Nogales et al., 2002); Puget Sound

(Braker et al., 2000); and a coastal aquifer/subterranean estuary (Santoro et al., 2006).

Most of the sequences from the low-salinity North Bay grouped independently from

other sites in San Francisco Bay but closely to sequences from low-salinity sites in

Chesapeake Bay (CB1 and CT1, Choptank River, MD). There was also considerable

overlap between high salinity (>20 psu) sites in both of these large North American

estuaries. However, the most abundant and highly expressed nirS phylotypes in

Chesapeake Bay (Bulow et al., 2008) were not as well represented in the San

Francisco Bay sediments (<5% of the sequences fell within this phylogenetic cluster).

Interestingly, mRNA nirS clones from the Colne estuary (Nogales et al., 2002) were

found in many of the phylogenetic clusters containing San Francisco Bay DNA

sequences.

195

Community structure of nitrite reductase genes

Some spatial structure was evident in the distribution of San Francisco Bay

nirK and nirS OTUs (Table 1) in the summer of 2004, based on abundance-based

Sørensen’s indices of similarity (Labd) (Schloss and Handelsman, 2006). For nirK and

nirS, the riverine North Bay (BG20 and BG30) sites showed almost no overlap with

any of the sites in San Pablo, Central, and South Bay (Labd < 0.07). For nirK and nirS,

BF21 in the North Bay was more similar to San Pablo Bay sites than to the other

North Bay sites or to sites in the Central and South Bay. The San Pablo Bay nirK and

nirS communities (BD31 and SP19) exhibited significant compositional overlap (Labd

≥ 0.5) with each other and with sites in the southernmost part of the estuary (LSB2 and

BA10). There was significant similarity (Labd ≥ 0.5) between most Central and South

Bay sites for both nirK and nirS. Interestingly, a previous study on ammonia-

oxidizing archaea and bacteria (AOA and AOB) communities (based on ammonia

monooxygenase subunit A, amoA, genes) at the same sites found similar patterns of

spatial distribution within the estuary; North Bay sites were similar to each other but

not with any other sites in the estuary and, conversely, Central and South Bay sites

overlapped with each other but were dissimilar to the North Bay sites (Mosier and

Francis, 2008).

Relationship between environmental variables and denitrifier community structure,

abundance, and potential rates

Denitrification potential rates (with chloramphenicol treatment) and the

abundance and community composition of nirK and nirS genes were compared to

196

environmental variables measured in the sediments and water column of San

Francisco Bay. Salinity had a strong and significant positive correlation with nirS

abundance (Spearman correlation for copies per gram sediment: ρ = 0.71, P < 0.001, n

= 35). DNP rates also had a strong, but slightly less significant, positive correlation

with salinity (Spearman correlation: ρ = 0.74, P = 0.002, n = 14). For nirK

abundance, salinity had a strong and significant negative correlation with nirK

abundance when expressed in terms of copies per µg DNA (Spearman correlation: ρ =

-0.65, P < 0.001, n = 33), but no correlation in terms of copies per gram sediment.

Salinity also showed a relationship with denitrification rates in other estuaries (e.g.,

Kemp et al., 1990; Rysgaard et al., 1999; Dong et al., 2000), although not always in

the same direction or magnitude—perhaps due to methodological differences in rate

measurements, divergent communities and responses, and/or unclear statistical results

from covarying variables within the estuaries.

Because most denitrifiers are heterotrophic, carbon and nitrogen are likely to

be important determinants of community composition and structure. In San Francisco

Bay sediments, canonical correspondence analysis (CCA) showed that sediment total

organic carbon (TOC) and C:N ratio had the greatest correlation with nirK community

composition of the 31 factors tested, while the nirS community composition was most

strongly correlated with only C:N ratio (Fig. S3). In the CCA plots, the perpendicular

position of the sampling sites relative to the environmental line is an approximate

ranking of species response curves to that environmental variable (McCune and Grace,

2002). For both nirK and nirS, the North Bay sites grouped independently from all

other regions. The Central and South Bay sites clustered together, reflecting

197

similarities in community composition and sediment carbon and nitrogen content.

Additionally, nirS abundance had a strong and significant negative correlation with

sediment C:N ratios (Spearman correlation: ρ = -0.69, P < 0.001, n = 21) and a

positive correlation with total nitrogen (Spearman correlation: ρ = 0.71, P < 0.001, n =

21). Together these results suggest that sediments with high C:N ratios in San

Francisco Bay have distinct nirS phylotypes that are not particularly abundant relative

to the rest of the estuary.

Several metals were correlated with denitrifier abundance and potential rates in

the estuary. nirS abundance (copies per gram sediment) was correlated with lead

(Spearman correlation: ρ = 0.60, P = 0.001, n = 28). nirK abundance (copies per µg

DNA) had a strong and significant negative correlation with silver (Spearman

correlation: ρ = -0.65, P = 0.001, n = 22) and methyl-mercury (Spearman correlation:

ρ = -0.80, P < 0.001, n = 26). Additionally, sediment cadmium concentrations were

negatively correlated with DNP rates in 2007 but with less statistical significance

(Spearman correlation: ρ = -0.86, P = 0.01, n = 7; cadmium concentrations were not

available for 2008). The bioavailability of metals and differences in denitrifying

microbial community tolerances may impact the toxicity on benthic denitrifier

abundance and activity.

While other variables not measured in this study could also affect

denitrification in San Francisco Bay (and we were unable to deduce the potential

effects of covariation), it appears that salinity, sediment C:N ratios, and several metals

may significantly influence denitrifying (nirK and nirS) communities in San Francisco

Bay. Notably, salinity and C:N ratios were also correlated with the ammonia-

198

oxidizing community in this estuary (Mosier and Francis, 2008). Salinity was

correlated with AOA and AOB amoA abundance and community structure, nirK and

nirS abundance, and denitrification potential rates. C:N ratio was correlated with

AOA and AOB amoA abundance, AOB amoA community composition, and nirK and

nirS community composition. Together these studies demonstrate a first-order insight

into nitrification and denitrification across chemical/physical gradients in the largest

estuary on the west coast of the United States.

Experimental Procedures

Sample collection and environmental variables

Sediment and water samples were collected from San Francisco Bay in

northern California during the summers (July-August) of 2004-2008 in conjunction

with the Regional Monitoring Program (RMP) managed by the San Francisco Estuary

Institute (http://www.sfei.org/). Water column profiles of temperature, salinity and

dissolved oxygen (DO) were measured at each site using a SeaBird™ SBE-19 CTD

(RMP). Bottom water samples from 2004 were previously analyzed for nitrate (NO3-),

nitrite (NO2-), and ammonium (NH4

+) (Wankel et al., 2006). For years 2005-2008,

nutrients were measured on a QuikChem 8000 Flow Injection Analyzer (Lachat

Instruments). The RMP analyzed sediment samples for years 2004-2007 for pH, trace

elements (Ag, Al, As, Cd, Cu, Fe, Hg, MeHg, Mn, Ni, Pb, Se and Zn), grain size, total

organic carbon and total nitrogen. C:N ratios were calculated using total organic

carbon and total nitrogen percentages.

Denitrification potential assays

199

Denitrification potential rates were determined from triplicate sediment

samples collected from 7 sites in the summers of 2007 and 2008 using the acetylene

block technique (Sorensen, 1978). Approximately 15 g of wet sediment

(homogenized from the upper 3 cm of 60 ml syringe cores) and 25 mL of

denitrification potential solution (1 mM glucose and 1mM NaNO3 in 0.2 µm filtered

site water) were added to serum bottles. An additional set of triplicate sediment

samples was treated with 0.3 mM chloramphenicol. Serum bottles were sealed and

bubbled with N2 gas for 5 minutes. 30 mL of acetylene gas was added to each bottle.

Bottles were shaken for 1 hour at room temperature. The headspace was sampled for

gas at four time points over the hour. Gas samples were analyzed for N2O on a

Shimadzu GC-14A gas chromatograph fitted with a 63Ni electron capture detector.

Quantification of nirS and nirK gene abundance

Quantitative PCR (qPCR) was used to estimate the number of nirK and nirS

copies in sediments collected during the summers of 2004-2008 at 7 sites

(Supplementary Material). The number preceding the sampling site name in Table S3

indicates the year sampled (e.g., 4BA10 and 5BA10 correspond to site BA10 sampled

in 2004 and 2005, respectively). DNA was extracted from approximately 250-300 mg

of sediment from the upper 5 mm of the cores using the FastDNA SPIN kit for soil

(MP Biomedicals, Inc.) with a 30 second bead beating time at speed 5.5. Duplicate

DNA extractions were performed for each sediment core. Reactions were carried out

using a StepOnePlus™ Real-Time PCR System (Applied Biosystems). Primers used

for nirK quantification were nirK-q-F (5’-TCATGGTGCTGCCGCGYGA; a shorter

version of the nirK583FdegCF forward primer from (Santoro et al., 2006) and

200

nirK1040 (Henry et al., 2004). Primers used for nirS quantification were nirS1F

(Braker et al., 1998) and nirS-q-R (5’-TCCMAGCCRCCRTCRTGCAG), an upstream

and significantly modified version of nirS3R (Braker et al., 1998).

Denitrifier community composition

nirK and nirS diversity was analyzed at 12 of sites from the summer of 2004

(Supplementary Material): 4BG20, 4BG30, 4BF21, 4BD31, 4SP19, 4BC11, 4CB18,

4SB02, 4BA41, 4SB18, 4LSB2, and 4BA10. Total community DNA was extracted as

described above. Sediment DNA extracts were PCR amplified with primers targeting

nirK and nirS: nirK583FdegCF (5’-TCATGGTGCTGCCGCGYGANGG; (Santoro et

al., 2006) and nirK5R (Braker et al., 1998); nirS1F and nirS6R (Braker et al., 1998).

Clone libraries of nirK and nirS gene fragments were constructed using the TOPO TA

cloning® kit (Invitrogen, Inc.) and 28 to 45 clones from each library were sequenced

(ABI 3100 Capillary Sequencer).

Statistical analyses

SPSS, version 17.0 was used for parametric and non-parametric correlation

analyses. nirK abundance, nirS abundance and denitrification potentials were

correlated with bottom water and sediment variables from each site and time point.

Bottom water variables included: salinity (psu), temperature (ºC), dissolved oxygen

(mg/L), nitrite+nitrate (µM), nitrite (µM), nitrate (µM), ammonium (µM), DIN (µM),

water depth (m), and pH. Sediment variables included: TOC (%), TN (%), C:N ratio,

Ag (mg/Kg), Al (mg/Kg), As (mg/Kg), Cd (mg/Kg), Cu (mg/Kg), Fe (mg/Kg), Hg

(mg/Kg), MeHg (µg/Kg), Mn (mg/Kg), Ni (mg/Kg), Pb (mg/Kg), Se (mg/Kg), Zn

(mg/Kg), % Clay (<0.0039 mm), % Fine (<0.0625 mm), % Granule and Pebble (2.0 to

201

<64 mm), % Sand (0.0625 to <2.0 mm), and % Silt (0.0039 to <0.0625 mm).

Correlations between nirK and nirS community composition and

environmental variables were analyzed by Canonical Correspondence Analysis (CCA)

using the program Canoco (ter Braak and Smilauer, 2002) (Supplementary Material).

Relative abundance of sequences in each OTU (defined at 0.05 distance) was used as

species input (i.e., the number of sequences in an OTU divided by the total number of

sequences in the library). Significant environmental variables were determined by

forward selection (P < 0.05). Bottom water variables included in the forward

selection were those described above as well as silicate (µM) and phosphate (µM).

Sediment variables included in the forward selection were all of those described

above, except Ag (mg/Kg) and Al (mg/Kg).

Nucleotide sequence accession numbers

Sequences reported in this study have been deposited in GenBank under

accession numbers GQ454031 to GQ454413 for nirK and GQ453671 to GQ454030

for nirS.

Acknowledgments

We thank Paul Salop, Sarah Lowe, Applied Marine Sciences, the San

Francisco Estuary Institute, and the crew of the R.V. Endeavor for allowing us to

participate in the San Francisco Bay RMP sediment cruises, providing access to the

RMP data, and assisting in sample collection. We thank S. Wankel for sampling and

performing nutrient analyses from the 2004 cruise and K. Roberts, J. Smith, S. Ying,

and G. Wells for assisting in field sampling. We greatly appreciate those who

202

provided advice and comments on the manuscript (A. Santoro, J. Smith, and G.

Wells). This work was supported in part by National Science Foundation Grants

MCB-0433804, MCB-0604270 and OCE-0847266 (to C.A.F.) and an EPA STAR

Graduate Fellowship (to A.C.M.).

203

Tables

204

Figures

Figure 1. Location of sampling sites within San Francisco Bay. Sites are shaded by

geographic region: North Bay, San Pablo Bay, Central Bay, and South Bay. Image

courtesy of and adapted from the U.S. Geological Survey.

205

Figure 2. Dissolved inorganic nitrogen concentrations in San Francisco Bay bottom

water from 2004 to 2008. Sites are ordered from approximate distance from the head

of the estuary.

206

Figure 3. Quantitative PCR measurements for nirK (left panel) and nirS (right panel)

gene copy numbers expressed per gram of sediment (wet weight). Values represent

averages of two duplicate DNA extractions (except where noted in Table S3).

207

Figure 4. Denitrification potential rate measurements from San Francisco Bay

sediments in 2007 (left panel) and 2008 (right panel). Light gray columns are

measurements with no treatment; dark gray columns are measurements with the

addition of chloramphenicol. Error bars represent duplicate or triplicate

measurements.

208



Supplementary Material. Additional text from the Results and Discussion and

Experimental Procedures sections.

Figure S1. Neighbor-joining phylogenetic tree of nirK sequences. Sequences from

San Francisco Bay are color coded by region (see legend) and GenBank sequences are

in black. Significant bootstrap values (≥ 50) are shown in italics at branch nodes.

Colored circles show the number of environmental clones from each region. Numbers

within each grey cluster represent the total number of San Francisco Bay and

GenBank sequences within that group.

Figure. S2. Neighbor-joining phylogenetic tree of nirS sequences. Sequences from

San Francisco Bay are color coded by region (see legend) and GenBank sequences are

in black. Significant bootstrap values (≥ 50) are shown in italics at branch nodes.

Colored circles show the number of environmental clones from each region. Numbers

within each grey cluster represent the total number of San Francisco Bay and

GenBank sequences within that group.

Figure S3. Canonical correspondence analysis (CCA) of nirK (left panel) and nirS

(right panel) sequence data and environmental variables from bottom water and

sediments.

209

Figure S4. Rarefaction curves for nirK (top panel) and nirS (bottom panel) clone

libraries. OTUs were defined at a 5% sequence cutoff.

Table S1. Physical and chemical properties of San Francisco Bay bottom water

averaged over five years (2004-2008).

Table S2. Physical and chemical properties of San Francisco Bay sediments averaged

over four years (2004-2007).

Table S3. Quantitative PCR measurements for nirK and nirS expressed per µg of

DNA and grams of sediment (wet weight). Averages and standard deviations (stdev)

are of two duplicate DNA extractions (except where noted). The number preceding

the sampling site name indicates the year sampled (e.g., 4BA10 and 5BA10

correspond to site BA10 sampled in 2004 and 2005, respectively).

Table S4. Diversity and richness estimates for nirK and nirS clone libraries. OTUs

were defined at 5% nucleic acid sequence difference. Libraries are ordered by

location within the estuary (from north to south).

210

Literature Cited

Abell, G.C., Revill, A.T., Smith, C., Bissett, A.P., Volkman, J.K., and Robert, S.S.

(2009) Archaeal ammonia oxidizers and nirS-type denitrifiers dominate sediment

nitrifying and denitrifying populations in a subtropical macrotidal estuary. The ISME

Journal: doi:10.1038/ismej.2009.1105.

Braker, G., Fesefeldt, A., and Witzel, K.P. (1998) Development of PCR primer

systems for amplification of nitrite reductase genes (nirK and nirS) to detect

denitrifying bacteria in environmental samples. Applied And Environmental


Braker, G., Zhou, J.Z., Wu, L.Y., Devol, A.H., and Tiedje, J.M. (2000) Nitrite

reductase genes (nirK and nirS) as functional markers to investigate diversity of

denitrifying bacteria in Pacific northwest marine sediment communities. Applied And


Braker, G., Ayala-del-Rio, H.L., Devol, A.H., Fesefeldt, A., and Tiedje, J.M. (2001)

Community structure of denitrifiers, bacteria, and archaea along redox gradients in

Pacific Northwest marine sediments by terminal restriction fragment length

polymorphism analysis of amplified nitrite reductase (nirS) and 16S rRNA genes.


211

Bricker, S.B., Clement, C.G., Pirhalla, D.E., Orlando, S.P., and Farrow, D.R.G. (1999)

National estuarine eutrophication assessment: Effects of nutrient enrichment in the

nation's estuaries. In. Silver Spring, MD: NOAA, National Ocean Service, Special

Projects Office and the National Centers for Coastal Ocean Science, p. 71.

Bulow, S.E., Francis, C.A., Jackson, G.A., and Ward, B.B. (2008) Sediment denitrifier

community composition and nirS gene expression investigated with functional gene

microarrays. Environmental Microbiology 10: 3057-3069.

Christensen, P.B., Nielsen, L.P., Sorensen, J., and Revsbech, N.P. (1990)

Denitrification in nitrate-rich streams: diurnal and seasonal variation related to benthic

oxygen metabolism. Limnology and Oceanography 35: 640-652.

Cornwell, J.C., Kemp, W.M., and Kana, T.M. (1999) Denitrification in coastal

ecosystems: methods, environmental controls, and ecosystem level controls, a review.

Aquatic Ecology 33: 41-54.

Diaz, R.J., and Rosenberg, R. (2008) Spreading dead zones and consequences for

marine ecosystems. Science 321: 926-929.

Dong, L.F., Thornton, D.C.O., Nedwell, D.B., and Underwood, G.J.C. (2000)

Denitrification in sediments of the River Colne estuary, England. Marine Ecology-

Progress Series 203: 109-122.

212



oxidation. The ISME Journal 1: 19-27.

Hager, S., and Schemel, L. (1996) Dissolved inorganic nitrogen, phosphorus and

silicon in South San Francisco Bay; I. Major factors affecting distributions. In San

Francisco Bay: The ecosystem; Further investigations into the natural history of San

Francisco Bay and Delta with reference to the influence of man. Hollibaugh, J.T. (ed).

San Francisco, CA: American Association for the Advancement of Science, p. 542.

Hallin, S., Throback, I.N., Dicksved, J., and Pell, M. (2006) Metabolic profiles and

genetic diversity of denitrifying communities in activated sludge after addition of

methanol or ethanol. Applied and Environmental Microbiology 72: 5445-5452.

Hamersley, M.R., Lavik, G., Woebken, D., Rattray, J.E., Lam, P., Hopmans, E.C. et

al. (2007) Anaerobic ammonium oxidation in the Peruvian oxygen minimum zone.


Hannig, M., Braker, G., Dippner, J., and Jurgens, K. (2006) Linking denitrifier

community structure and prevalent biogeochemical parameters in the pelagial of the

central Baltic Proper (Baltic Sea). FEMS Microbiology Ecology 57: 260-271.

213

Henry, S., Baudoin, E., Lopez-Gutierrez, J.C., Martin-Laurent, F., Baumann, A., and

Philippot, L. (2004) Quantification of denitrifying bacteria in soils by nirK gene

targeted real-time PCR. Journal Of Microbiological Methods 59: 327-335.

Jayakumar, D.A., Francis, C.A., Naqvi, S.W.A., and Ward, B.B. (2004) Diversity of

nitrite reductase genes (nirS) in the denitrifying water column of the coastal Arabian

Sea. Aquatic Microbial Ecology 34: 69-78.

Jensen, M.M., Kuypers, M.M.M., Lavik, G., and Thamdrup, B. (2008) Rates and

regulation of anaerobic ammonium oxidation and denitrification in the Black Sea.


Junier, P., Kim, O.S., Witzel, K.P., Imhoff, J.F., and Hadas, O. (2008) Habitat

partitioning of denitrifying bacterial communities carrying nirS or nirK genes in the

stratified water column of Lake Kinneret, Israel. Aquatic Microbial Ecology 51: 129-

140.

Kemp, W.M., Sampou, P., Caffrey, J., Mayer, M., Henriksen, K., and Boynton, W.R.

(1990) Ammonium Recycling Versus Denitrification in Chesapeake Bay Sediments.


214

Kimmerer, W.J. (2004) Open Water Processes of the San Francisco Estuary: From

Physical Forcing to Biological Responses. San Francisco Estuary and Watershed

Science 2: 1-147.

Koop-Jakobsen, K., and Giblin, A. (2009) Anammox in Tidal Marsh Sediments: The

Role of Salinity, Nitrogen Loading, and Marsh Vegetation. Estuaries and Coasts 32:

238–245.

Kuypers, M.M., Lavik, G., Woebken, D., Schmid, M., Fuchs, B.M., Amann, R. et al.

(2005) Massive nitrogen loss from the Benguela upwelling system through anaerobic

ammonium oxidation. Proceedings of the National Academy of Sciences 102: 6478-

6483.

Lam, P., Lavik, G., Jensen, M.M., van de Vossenberg, J., Schmid, M., Woebken, D. et

al. (2009) Revising the nitrogen cycle in the Peruvian oxygen minimum zone.


Liu, X.D., Tiquia, S.M., Holguin, G., Wu, L.Y., Nold, S.C., Devol, A.H. et al. (2003)

Molecular diversity of denitrifying genes in continental margin sediments within the

oxygen-deficient zone off the Pacific coast of Mexico. Applied And Environmental


215

McCune, B., and Grace, J.B. (2002) Analysis of Ecological Communities. Gleneden

Beach, Oregon: MjM Software Design.




Nielsen, K., Nielsen, L.P., and Rasmussen, P. (1995) Estuarine nitrogen retention

independently estimated by the denitrification rate and mass-balance method: a study

of Norsminde Fjord, Denmark Marine Ecology-Progress Series 119: 275-283.

Nogales, B., Timmis, K.N., Nedwell, D.B., and Osborn, A.M. (2002) Detection and

diversity of expressed denitrification genes in estuarine sediments after reverse

transcription-PCR amplification from mRNA. Applied And Environmental


Oakley, B.B., Francis, C.A., Roberts, K.J., Fuchsman, C.A., Srinivasan, S., and Staley,

J.T. (2007) Analysis of nitrite reductase (nirK and nirS) genes and cultivation reveal

depauperate community of denitrifying bacteria in the Black Sea suboxic zone.


216

Prieme, A., Braker, G., and Tiedje, J.M. (2002) Diversity of nitrite reductase (nirK and

nirS) gene fragments in forested upland and wetland soils. Applied And Environmental


Rabalais, N.N. (2002) Nitrogen in aquatic ecosystems. Ambio 31: 102-112.

Rich, J.J., Dale, O.R., Song, B., and Ward, B.B. (2008) Anaerobic ammonium

oxidation (anammox) in Chesapeake Bay sediments. Microbial Ecology 55: 311-320.

Risgaard-Petersen, N., Meyer, R.L., Schmid, M., Jetten, M.S.M., Enrich-Prast, A.,

Rysgaard, S., and Revsbech, N.P. (2004) Anaerobic ammonium oxidation in an

estuarine sediment. Aquatic Microbial Ecology 36: 293-304.




Saleh-Lakha, S., Shannon, K.E., Henderson, S.L., Goyer, C., Trevors, J.T., Zebarth,

B.J., and Burton, D.L. (2009) Effect of pH and temperature on denitrification gene

expression and activity in Pseudomonas mandelii. Applied and Environmental


217

Santoro, A.E., Boehm, A.B., and Francis, C.A. (2006) Denitrifier community

composition along a nitrate and salinity gradient in a coastal aquifer. Applied And


Schloss, P.D., and Handelsman, J. (2006) Introducing SONS, a Tool for Operational

Taxonomic Unit-Based Comparisons of Microbial Community Memberships and

Structures. Applied and Environmental Microbiology 72: 6773-6779.

Seitzinger, S., Harrison, J.A., Bohlke, J.K., Bouwman, A.F., Lowrance, R., Peterson,

B. et al. (2006) Denitrification across landscapes and waterscapes: a synthesis.

Ecological Applications 16: 2064-2090.

Seitzinger, S.P. (1988) Denitrification in freshwater and coastal marine ecosystems:

Ecological and geochemical significance. Limnology and Oceanography 33: 702-724.

Shapleigh, J.P. (2006) The denitrifying prokaryotes. In Prokaryotes, pp. 769-792.

Sharma, S., Aneja, M.K., Mayer, J., Munch, J.C., and Schloter, M. (2005) Diversity of

transcripts of nitrite reductase genes (nirK and nirS) in rhizospheres of grain legumes.


Smith, C.J., Nedwell, D.B., Dong, L.F., and Osborn, A.M. (2007) Diversity and

abundance of nitrate reductase genes (narG and napA), nitrite reductase genes (nirS

218

and nrfA), and their transcripts in estuarine sediments. Applied and Environmental


Smith, M.S., and Tiedje, J.M. (1979) Phases of denitrification following oxygen

depletion in soil. Soil Biology and Biochemistry 11: 261-267.

Sorensen, J. (1978) Denitrification Rates in a Marine Sediment as Measured by the

Acetylene Inhibition Technique. Applied and Environmental Microbiology 36: 139-

143.

Tamegai, H., Aoki, R., Arakawa, S., and Kato, C. (2007) Molecular analysis of the

nitrogen cycle in deep-sea microorganisms from the Nankai Trough: genes for

nitrification and denitrification from deep-sea environmental DNA. Extremophiles 11:

269-275.

ter Braak, C.J.F., and Smilauer, P. (2002) CANOCO Reference Manual and

CanoDraw for Windows User's Guide: Software for Canonical Community Ordination

(version 4.5). Ithaca, NY: Microcomputer Power.

Throback, I.N., Johansson, M., Rosenquist, M., Pell, M., Hansson, M., and Hallin, S.

(2007) Silver (Ag+) reduces denitrification and induces enrichment of novel nirK

genotypes in soil. FEMS Microbiology Letters 270: 189-194.

219

Tiquia, S.M., Masson, S.A., and Devol, A. (2006) Vertical distribution of nitrite

reductase genes (nirS) in continental margin sediments of the Gulf of Mexico. FEMS


Trimmer, M., Nicholls, J.C., and Deflandre, B. (2003) Anaerobic ammonium

oxidation measured in sediments along the Thames estuary, United Kingdom. Applied


Wankel, S.D., Kendall, C., Francis, C.A., and Paytan, A. (2006) Nitrogen sources and


approach. Limnology And Oceanography 51: 1654-1664.

Yan, T.F., Fields, M.W., Wu, L.Y., Zu, Y.G., Tiedje, J.M., and Zhou, J.Z. (2003)

Molecular diversity and characterization of nitrite reductase gene fragments (nirK and

nirS) from nitrate- and uranium-contaminated groundwater. Environmental


Zumft, W.G. (1997) Cell biology and molecular basis of denitrification. Microbiology

And Molecular Biology Reviews 61: 533-542.

Documents

microbial nitrogen cycling dynamics in coastal systems a