Poster Abstracts

Florent Angly and Forest Rohwer
San Diego State University
fangly@projects.sdsu.edu

Uncultured viral diversity
There are an estimated 10³¹ viral particles on Earth, most of which are phages. As microbial predators, phages structure their host communities and influence global biogeochemical cycles like the carbon cycle. Determining viral diversity has traditionally been impeded by the inability to cultivate more than 1% of the microbial hosts. Unlike microbes, which can be studied using molecular markers like the 16S rDNAs, viruses do not have a ubiquitous marker. The advent of metagenomics, however, provides new ways to estimate the previously veiled viral diversity based on sequence assemblies.

Overlaps between sequences represent the resampling of a given genome and provide information about community diversity. The contig spectrum is a census of the contigs of different sizes obtained from assembly of viral metagenomic DNA sequences. It can readily be modeled (http://biome.sdsu.edu/phaccs) using the Generalized Lander-Waterman algorithm to make estimates of local community diversity (alpha diversity): i) richness, or number of different species, ii) evenness, or equitability of the species, and iii) Shannon-Wiener index, a combination of richness and evenness. As an extension to the contig spectrum approach, it is possible to assemble sequences from different metagenomes to generate the so-called mixed contig spectrum, which can be modeled as previously to produce geographic-scale diversity estimates (gamma-diversity). By extending the approach further, a Monte-Carlo analysis can be performed on a cross-contig spectrum. In that case, only contigs formed by sequences from different metagenomes are counted and used to assess the difference in diversity between communities (beta-diversity) by considering two parameters: i) how many species are common to the communities, or percentage shared, and ii) how many of the most abundant species have a different abundance rank, or percentage permuted.

We have developed methods to estimate diversity using mathematical modeling of contig spectra. Using these approaches, it is possible determine local diversity (alpha diversity), difference between communities (beta diversity), and geographical-scale diversity (gamma diversity). A study of four viral metagenomes from different oceanic provinces showed that marine viral communities have a large majority of species in common and that reshuffling of the 30% most abundant species accounts for differences between communities. In the Northern Line Islands, the viral diversity on four reefs was correlated with human activities and seems to support the Intermediate Disturbance Hypothesis. Overall, oceanic viral richness is believed to be as high as a few hundred thousand of species. Further analyses of diversity from metagenomes will enable ecologists to better understand how biogeochemical factors shape the structure of microbial communities and what rules define global diversity patterns.

Beltran Rodriguez Brito, Linda Wegley, Matt Haynes, Mike Furlan, Kristen Marhaver, Jennifer Mueller, Florent Angly, Tuong Tran, Michelle Hartman, Veronica Casas, Shailaja Sriganesh, Steve Rayhawk, Ben Felts, Jim Nulton, Robert Edwards, Peter Salamon, Joseph Mahaffy, Mya Breitbart, and Forest Rohwer
San Diego State University
beltran.rodriguezbrito@gmail.com

A metagenomic analysis of a solar saltern in Southern California
Metagenomic analysis of samples in a solar saltern in southern California were collected using TFF filters and its DNA extracted. The analysis of sequence data show population and diversity changes as the salinity increases. The microbial fraction is comprised mostly of Bacteria, giving way to Archaea as the salinity increases. The medium salinity sample seems to be the less diverse, with low salinity being the most diverse environment.

Giuseppe D'Auria, Ravindra Pushker, Jose Carlos Alba-Casado, and Francisco Rodríguez-Valera
University Miguel Hernandez, Alicante, Spain
gdauria@umh.es

Micro-Mar: A database for dynamic representation of marine microbial biodiversity
Prokaryotes represent a major component of all marine ecosystems. Their reluctance to grow in pure culture brought during the last decades a massive amount of PCR-based data about these microbial communities. To analyze this large dataset from a global point of view, new tools are required that would allow the scientific community to approach problems from bioprospecting to development of evolutionary and speciation models.

Micro-Mar database was born to gather molecular data from several oceanic locations, previously published in generic databases. The sequences included are stored with their geographic origin and, where provided by the authors, with physicochemical parameters. Scrupulous attention was also given to the taxonomic affiliation of the stored sequences. This way searches can be focused into specific prokaryotic taxonomic groups.

Micro-Mar provides a friendly interface that allows users to search sequences within the database by free text search, geographic location, or BLAST analysis of user sequences. The results coming from the various search systems are displayed in a classic tabular format or a Dynamic Map providing a clear representation of geographic origin for the best BLAST hits. The Micro-Mar database stores 16S, ITS, 23S and several CDS sequences for a total of 1,241 sequences related to Archaea and 8,875 to Bacteria superkingdom. The database is online at http://egg.umh.es/micromar/.

Christelle Desnues, Beltran Rodriguez-Brito, Florent Angly, Scott Kelley, Selina Liu, Steven Rayhawk, Matthew Haynes, Robert Edwards, Mya Breitbart, Janet Siefert, Valeria Souza, Pamela Reid, and Forest Rohwer
San Diego State University
cdesnues@yahoo.fr

Phage communities in stromatolites: What metagenomics could reveal about diversity, adaptations, and evolution?
Stromatolites are usually defined as “organosedimentary structures predominantly accreted by sediment trapping, binding, and/or in situ precipitation as a result of the growth and metabolic activity of microorganisms” (Walter, 1976). Fossil stromatolites, dating back from ~3.5 GY, have been found in the geological record, probably revealing the most ancient form of life on Earth. Actually, modern stromatolites can be found in a few places, such as hypersaline waters (Shark Bay, Australia), fresh water where calcium carbonate precipitation occurs (Cuatro Ciénegas, Mexico), or an open marine environment (Exuma Cay, Bahamas). From a geological view, it has been suggested that modern stromatolites could reflect what happened in their ancient analogues and, therefore, shed light on Earth’s history.

Phages (viruses that infect prokaryotes) are known to play an important role in determining the diversity, adaptation, and evolution of prokaryotes. Even if they are highly abundant in aquatic environments, phage communities have never been studied in stromatolites. One of the limits in the study of phage diversity is the fact that only 0.1-1 % of their host could be cultured using conventional techniques. The genomic analysis of DNA extracted directly from communities in environmental samples (i.e., metagenomic) help to circumvent this limitation.

Here we present a comparative metagenomic analysis of phage communities from two stromatolites collected in fresh water (a pool, Poza Azul, and a river, Rio Mesquites, in Mexico) and one stromatolite collected in a marine environment (Highborne Cay, Bahamas). The 81,687,957 bp of DNA generated from the three libraries were BLASTed against different databases giving insights on structural and functional diversity of the phage community. As expected, most of the sequences don’t match anything in the databases. Based on the “known” sequences, the diversity of the phage community appears relatively low for Rio mesquites, medium for Highborne Cay, and very high for Poza Azul. Surprisingly, most of the sequences in the Poza Azul stromatolite are related to phages infecting marine bacteria even if this environment has not been in contact with the ocean for millions years.

Phages often provide genes important to the life cycle and adaptation of their bacterial hosts. For example, cyanophages have been found to carry photosynthetic genes that are expressed during infection to enhance bacterial photosynthesis. Functional analyses of the Highborne Cay phage metagenome revealed sequences related to a hydrophilic protein belonging to the Late Embryogenesis Abundant (LEA) protein family. These proteins are known to accumulate under conditions of extreme desiccation in plants, nematodes, fungi, and bacteria. As stromatolites in Highborne Cay form in the near-shore beach zone where they are alternately buried and uncovered at low tide, one suspects that, during infection, phage could transmit this gene to their host or at least avoid their death or even enhance their growth.

Results provided here offer us for the first time insights into genetic diversity, population structure, ecology, and the evolution of phage communities in stromatolites. Lateral gene transfer among phages and their hosts undoubtedly occurs in these complex microbial natural communities, potentially promoting the capabilities of the organisms that are involved.

Elizabeth Dinsdale, Olga Pantos, Steve Smriga, Robert Edwards, Florent Angly, Elysa Brown, Matthew Haynes, Lutz Krause, Rebecca Vega Thurber, Farooq Azam, and Forest Rohwer
San Diego State University
Elizabeth_dinsdale@hotmail.com

Dramatic change in coral reef microbial communities across a gradient of human disturbance
Microbes are key energy and nutrient generators in the marine environment, but their interactions with the health of ecosystems, like coral reefs, are virtually unknown. Here we describe the abundance, taxonomic composition, and potential metabolic function of the microbial communities from relatively pristine to moderately disturbed coral reefs. Increasing human population was associated with a 10-fold increase in microbial community and a functional transition from a community composed of autotrophic and heterotrophic microbes to a community dominated by heterotrophs, including many potential pathogens. The transition was supported by increases in nutrient levels that enabled faster growth of xenogenic microbes.

Chris Dupont (SIO/UCSD), Song Yang (UCSD/SDSC), Brian Palenik (SIO/UCSD), and Phillip E. Bourne (UCSD/SDSC)
Scripps Institution of Oceanography and San Diego Supercomputer Center, UCSD
cdupont@ucsd.edu
Modern proteomes contain putative imprints of ancient shifts in trace metal geochemistry
Due to the rise in atmospheric oxygen 2.3 billion years ago (Gya) and subsequent changes in oceanic redox state over the last 2.3-1 Gya, trace metal bioavailability in marine environments has changed dramatically. While theorized to have influenced the biological usage of metals leaving discernable genomic signals, a thorough and quantitative test of this hypothesis has been lacking.  Using structural bioinformatics and whole genome sequences, the Fe, Zn, Mn, and Co-binding metallomes of 23 Archaea, 233 Bacteria, and 57 Eukarya were constructed.  These metallomes reveal that the overall abundances of these metal-binding structures scale to proteome size as power laws with a unique set of slopes for each Superkingdom of Life.  The differences in the power slopes describing the abundances of Fe, Mn, Zn, and Co-binding proteins in the proteomes of Prokayotes and Eukaryotes are similar to the theorized changes in the abundances of these metals following the oxygenation of oceanic deep waters.  This suggests that Prokarya and Eukarya evolved in anoxic and oxic environments respectively, a hypothesis further supported by structures and functions of Fe-binding proteins in each Superkingdom.  Also observed is a proliferation in the diversity of Zn-binding protein structures involved in protein-DNA and protein-protein interactions within Eukarya, an event unlikely to occur in either an anoxic or euxinic environment where Zn concentrations would be vanishingly low.  We hypothesize that these conserved trends are proteomic imprints of changes in trace metal bioavailability in the ancient ocean that highlight a major evolutionary shift in biological trace metal usage.

Emiley A. Eloe, Federico M. Lauro, Rudi F. Vogel, and Douglas H. Bartlett
Scripps Institution of Oceanography, UCSD
eeloe@ucsd.edu

Flagellar motility at high hydrostatic pressure: Functional genomics of Photobacterium profundum strain SS9 and comparative studies of piezophilic flagellin sequences
Pressure is a fundamental and unique physical parameter that has influenced the evolution and distribution of life in the deep-sea environment. The psychrotolerant, moderately piezophilic deep-sea bacterium Photobacterium profundum strain SS9 has been identified as a model organism for studies involving high pressure adaptations. The recent genome sequencing and comparative genomics of two other strains of P. profundum (DSJ4 and 3TCK) have yielded significant insight into the molecular basis of piezophily (Vezzi et al., 2005; Campanaro et al., 2005). In this study, motility has been selected for further investigation since it is considered one of the most pressure-sensitive cellular processes.

The mechanism of flagellar motility in P. profundum strain SS9 was investigated through generation of in-frame flagellin and motor-protein deletion mutants. SS9 has two main gene clusters for motility and flagellar assembly, one of which is lacking in the pressure-sensitive P. profundum strain 3TCK. Mutant phenotype characterization involved an indirect growth-based motility assay and direct swimming speed measurements as a function of pressure using the newly developed HPDS high-pressure microscopic cell (Hartmann et al., 2004; Frey et al., 2006). Flagellin sequence comparisons from other piezophile isolates and phylogenetically related species were analyzed and structurally modeled against the Salmonella typhimurium phase 1 flagellin (Samatey et al., 2001; Yonekura et al., 2003). The results indicate unique adaptations required to maintain motility at high hydrostatic pressure for deep-sea autochthonous species.

Vincent C. Henrich1, Parke Rublee2, and Michael M. Marshall2
1Center for Biotechnology, Genomics, and Health Research
2University of North Carolina, Greensboro
vincent_henrich@uncg.edu

A method for rapid characterization of microbial communities in freshwater samples
Characterization of known and unidentified microorganisms that occur in freshwater sources relies on a variety of methods.  A prototype microarray using 50-mer oligonucleotides derived from a variable portion of SSU rDNA was developed for characterizing DNA extracted en masse from freshwater microbial field collections.  Probe sequences representing variable portions of known eukaryotic and prokaryotic rDNA sequences were obtained from GenBank. Also, clones containing rDNA from undescribed microbes were obtained from rDNA amplicon libraries derived from freshwater microbial samples (see poster by Rublee et al., this conference).  The microarray includes probes for infectious agents, known microbial species such as cyanobacteria, indicators associated with specific environmental pollutants, and several “unknown” sequences.  The method begins with extraction of DNA from microbes recovered by filtration followed by a multiplex PCR amplification, fluorescent labeling, and hybridization with the microarray.  The microarray generates specific signals and allows semi-quantification of abundant rDNAs.  By using differential fluorescent labeling, the possibility for comparing environmental and/or reference samples is also described.

Brian Hopkinson
Scripps Institution of Oceanography, UCSD
bhopkins@ucsd.edu

Genomic and experimental evidence for heme uptake by marine bacteria
Although iron is an essential micronutrient regulating growth of microorganisms in large areas of the ocean, its chemistry and cycling are difficult to study due to iron’s low oceanic concentration, its particle reactive nature, and a range of other technical challenges. Analysis of marine microbial genomes presents an alternative approach to understanding iron in the ocean, allowing organisms’ metabolic capabilities to describe routes of iron transformation occurring in the marine environment.  As a well defined and abundant intracellular iron pool, heme iron is likely an important component in the iron cycle and could be a significant source of iron for organisms, especially those involved in nutrient recycling pathways such as heterotrophic bacteria. Toward a better understanding of heme’s potential importance in the iron cycle, we used Hidden Markov Models to search the genomes of more than 50 marine bacteria for heme transport components and conducted heme uptake experiments with genome-sequenced bacteria. A number of bacteria are capable of a taking up 55Fe-heme, and putative heme transporters have been identified, most convincingly in the gamma- and alpha-proteobacteria. While results are preliminary, it appears that the ability to take up heme iron is more common among bacteria believed to be associated with particles, consistent with the low-seawater solubility but relatively high-intracellular concentration of heme.

Federico Lauro and Douglas H. Bartlett
Scripps Institution of Oceanography, UCSD
flauro@ucsd.edu

Genomic perspectives in piezomicrobiology
The structure of complex microbial assemblages in the ocean is heavily influenced by variations in physicochemical factors such as light, organic matter availability and composition, temperature, and hydrostatic pressure. The latter, increasing by approximately 1 atmosphere every 10 meters depth, is largely limiting for life in the deep sea. Autochtonous members of deep-sea microbial communities, outside of hydrothermal vents, can be identified by the dependence of their growth characteristics on high
hydrostatic pressure and low temperature.

The existence of phylogenetically related, depth-specific ecotypes (bathytypes) prompted us to sequence the first five genomes of psychropiezophilic (cold- and pressure-loving) bacterial isolates and compare them to the genomes of their closest non-piezophilic relatives.  The genomes of the deep bathytypes show a high ratio of rRNA operon numbers/genome size and remarkably large intergenic regions compatible with the hypothesis of an opportunistic (r-strategy) lifestyle with a high degree of gene regulation. Other hallmarks of all piezophilic genomes are the presence of a large number of genes involved in membrane unsaturation, such as delta-9 desaturases and PUFA gene clusters, and the absence of light-dependant photolyase genes. On the contrary, the comparison of the orthologous protein-coding portion of the genome shows relatively few amino acid differences between dramatically diverse bathytypes, suggesting that very few changes are sufficient for the evolution of pressure-adapted proteins.

Expansion/contraction of specific gene families with depth was investigated by classifying and comparing the predicted ORFs from the diverse bathytypes according to the NCBI COG (Cluster of Orthologous Group) database. The shallow bathytypes were enriched in genes for energy production and conversion, while the deep bathytypes had a higher percentage of genes involved in cell motility and secretion, intracellular trafficking, secretion, translation, ribosomal structure, and DNA replication and repair. However, the last trend is biased by the presence in this COG functional category of mobile genetic elements, previously shown to be over-represented in metagenomic studies of deep-sea water samples. In fact, removing transposases and integrases from the analysis, the trend disappeared. The ecological significance of these findings will be discussed.

Linlin Li and Forest Rohwer
San Diego State University
linlinli99013@gmail.com

Metagenomic analysis of the microbial community in an aquaculture system
Microbes in aquaculture systems have great influence on the mortality rate, feed conversion ratios, growth rates, and susceptibility to pathogens of the cultivated organisms. The microbial community also contributes to the environmental impacts of aquaculture. Using tangential flow filtering and pyrosequencing, we analyze the viral and microbial communities in a freshwater aquaculture system for the first time.

Most of the resulting sequences are not related to anything previously reported. The recognized viruses mostly belong to myoviridae, siphoviridae, and podoviridae. Sequences of Herpesviruses have good abundance, and some fish pathogenic viruses are also discovered. The composition and diversity of the whole microbial communities are assessed by metagenomic and 16s rDNA-based sequence analysis.

Weizhong Li and Adam Godzik
Calit2/UCSD and Burnham Institute for Medical Research
liwz@sdsc.edu

Clustering analyses of global ocean proteomes
The recent Sorcerer II Global Ocean Sampling (GOS) expedition provided the largest example of a metagenomics study performed so far. This study acquired 17 million new predicted Open Reading Frames (ORFs), which significantly altered the landscape of current protein space. Such mega-sets of sequences, by their sheer size, defy any conventional methods for analysis. In this study, we describe a strategy for quick analyses of the sequence diversity and internal structure of such mega-sets, which requires only about 1% computation time of conventional methods.

We performed a hierarchical clustering analysis on the GOS ORFs using the newly developed CD-HIT algorithm and identified 34,803 clusters of more than 20 non-redundant ORFs. We also used the same clustering strategy on the NCBI non-redundant (NR) protein databases and identified 14,216 families. We found that, of the 34,803 GOS clusters, 22,377 have significant similarities to known proteins by BLASTP. With PDB-BLAST and FFAS, which are more sensitive homology detection methods, we could identify an additional 4,252 clusters having remote but statistically significant similarities to known proteins. The remaining 8,174 clusters were identified as potential novel protein families.

K. Mavromatis, N. Ivanova, K. Barry, H. Shapiro, E. Goltsman, A. McHardy, I. Rigoutsos, A. Salamov, F. Korzeniewski, M. Land, A. Lapidus, I. Grigoriev, P. Hugenholtz, and N. Kyrpides
DOE Joint Genome Institute
NNIvanova@lbl.gov

On the fidelity of processing metagenomic sequences using simulated datasets
In an effort to evaluate methods used to process metagenomic sequences, we have constructed three simulated datasets of increased complexity by combining Sanger sequencing reads from a selection of 113 isolate genomes sequenced at the DOE Joint Genome Institute. These datasets were designed to resemble real metagenomic datasets in terms of complexity and relative phylogenetic position of the organisms. Reads were randomly sampled from the selected genomes to match the read depth of their corresponding populations in the metagenomic datasets. Sampled reads were then assembled using three separate programs (Phrap, Arachne and JAZZ). Assembled contigs were binned using three different methods (oligonucleotide frequency, pattern discovery, best blast hit), and genes were predicted using two gene finding pipelines (fgenes, Critica/Glimmer).

Here we present how the simulated metagenomic datasets can be used to evaluate the quality of each step in the process. We explore the contribution of the complexity in the population distribution and the various algorithms used to the quality of processing of the metagenomic sequences and the derived biological conclusions, and identify the optimum combination of methods coupled to the corresponding environmental complexity.

Annika Mosier and Christopher Francis
Stanford University
annikam@stanford.edu

Use of ammonia monooxygenase (amoA) genes as a molecular marker for characterization of estuarine ammonia-oxidizing archaeal and bacterial communities
Environmental genomic analyses have recently yielded profound new insights into the metabolism of non-thermophilic crenarchaeota, which are abundant in many terrestrial and marine environments. Venter et al.[1] first suggested that archaea may be capable of chemoautotrophic nitrification after finding a unique ammonia monooxygenase gene (amoA) on an archaeal-associated scaffold in the Sargasso Sea shotgun sequencing dataset.  This suggestion was further corroborated when a Sargasso Sea-like amoA homolog and a 16S rRNA gene from soil crenarchaeota were found on the same metagenomic fragment[2].  The recent cultivation of a novel, ammonia-oxidizing, marine crenarchaeota[3] revealed the first direct evidence for nitrification within the Archaeal domain and definitively linked the novel amo genes to this chemoautotrophic metabolism.  Finally, metagenomic analysis of the uncultivated sponge-symbiont, Cenarchaeum symbiosum, revealed further insight into crenarchaeal pathways for ammonia oxidation and carbon assimilation[4]. 

Together, these discoveries challenge the accepted view of the global nitrogen cycle and validate the need for further research on microbial diversity and function.  In particular, it is imperative to reexamine the microbial communities involved in ammonia oxidation in marine and estuarine sediments where this process plays a pivotal role in the cycling and removal of nitrogen. 

Using phylogenetic analyses of amoA gene sequences, we examined the distribution and diversity of ammonia-oxidizing archaea (AOA) and bacteria (AOB) in San Francisco Bay, the largest estuary on the west coast of the United States.  The highly impacted bay effectively connects two estuaries with varying physical-chemical characteristics to the Pacific Ocean.  We recovered archaeal and bacterial amoA genes from 11 sites distributed throughout the bay, spanning the northern and southern estuaries and the central region where they connect to the ocean.  Richness estimates varied considerably across all sites examined, with archaeal amoA estimates being generally higher than bacterial amoA.  Archaeal amoA sequences were phylogenetically diverse and grouped within previously described sediment and soil/sediment clusters.  Several sequences were closely related to the only cultivated AOA, Nitrosopumilus maritimus.  Both the archaeal and bacterial amoA sequences showed significant regional specificity.  Distinct populations exist in the northern and southern estuaries and sequences from the northern-most and southern-most sites clustered independently.  The central region of the bay appears to contain a blend of the northern and southern estuary ammonia-oxidizing communities.  Overall, this molecular data significantly expands our understanding of estuarine ammonia-oxidizing communities and suggests that AOA and AOB may respond similarly to environmental gradients within the San Francisco Bay estuary.

1Venter, J.C., Remington, K., Heidelberg, J.F., Halpern, A.L., Rusch, D., Eisen, J.A., Wu, D.Y., Paulsen, I., Nelson, K.E., Nelson, W., et al. (2004) Science 304, 66-74.
2Treusch, A.H., Leininger, S., Kletzin, A., Schuster, S.C., Klenk, H.P., Schleper, C. (2005) Environmental Microbiology 7, 1985-1995.
3Könneke, M., Bernhard, A.E., de la Torre, J.R., Walker, C.B., Waterbury, J.B. , Stahl, D.A. (2005) Nature 437, 543-546.
4Hallam, S.J., Mincer, T.J., Schleper, C., Preston, C.M., Roberts, K., Richardson, P.M., DeLong, E.F. (2006) PLoS Biology 4, 520.

Ryan S. Mueller, I. Naili, B. Lim, F. Yildiz, F. Azam, and D.H. Bartlett
Scripps Institution of Oceanography, UCSD
rmueller@ucsd.edu

Transcriptome response of Vibrio cholerae to the signaling molecule indole
Biofilms formed by the cholera-causing bacterium, Vibrio cholerae, have been suggested to be critical for survival in both its natural aquatic environment and during human infection by increasing its resistance to acid and chlorine stresses and protozoan predation. To find novel genes responsible for biofilm formation in two environmental strains of Vibrio cholerae, a transposon mutagenesis screen for mutants attenuated in biofilm formation was performed. This screen revealed that the tryptophanase gene, tnaA, is essential for proper biofilm formation in both of these strains. Similar to other tryptophanase-containing bacteria, it was found that exogenous addition of indole, a breakdown product of the tryptophanase reaction, to these V. cholerae strains could restore biofilm formation to wild-type levels. Additionally, lacZ fusion studies have discovered that indole directly affects the expression levels of genes involved in vibrio polysaccharide (VPS) production, the major component of the extracellular matrix of V. cholerae biofilms.

To uncover additional genes whose expression is also governed by tryptophanase activity, microarray experiments with the tnaA mutant and wild-type strains in the presence and absence of exogenous indole were performed. The results of these experiments show that tryptophanase activity also has a role in modulating the expression of many other genes, including some involved in motility, virulence, and grazing protection. Our results suggest that tryptophanase activity may govern the production of biofilms in V. cholerae by directly regulating VPS production and motility, in addition to regulating factors that may aid in survival from various environmental and host-induced stresses.

Marc Mußmann, Fen Z. Hu2, Michael Richter1, Dirk deBeer1, André Preisler1, Bo Barker Jørgensen1, Rudolf Amann1, Frank Oliver Glöckner1, Benjamin Janto2, Justin Hogg2, Robert Boissy2, Roger Lasken2, Paul Stoodley2, and Garth Ehrlich2
1Max Planck Institute for Marine Microbiology, Bremen, Germany
2Center for Genomic Sciences, Allegheny-Singer Research Institute, Pittsburgh, PA
mmussman@mpi-bremen.de

Genomic analysis of single filaments of uncultured Beggiatoa, key bacteria in global sulphur and nitrogen cycling
Marine surface sediments in coastal zones, hydrothermal vents, and methane seeps are frequently covered by mats of Beggiatoa and related, filamentous sulphur-oxidizing bacteria. These large, conspicuous organisms efficiently oxidize hydrogen sulphide by respiring either oxygen or nitrate, which they store in intracellular vacuoles. Here, we present a unique approach to access the genome of uncultured Beggiatoa by combining whole genome amplification, massively parallel pyrosequencing, and optical genome mapping. Incomplete DNA assemblies from a single filament and optical mapping indicate a genome size larger than 8 Mb. Encoded pathways confirmed the putative chemolithoautotrophic physiology of Beggiatoa; alternative energy yielding pathways are also present.

Based on our findings, we propose a mechanism for the exceptional accumulation of nitrate in the vacuoles. Comparative genomics between filamentous Cyanobacteria and Beggiatoa identified homologous genes encoding proteins possibly involved in cell adhesion or gliding and in the synthesis of secondary metabolites. The potential to synthesize biologically active compounds may imply a larger ecological significance of Beggiatoa in benthic habitats than assumed. The first look into the genome of these uncultured, filamentous sulphur-oxidizing bacteria substantially deepens our understanding of their evolution and the sulphur and nitrogen cycles in marine sediments.

Priya Narasingarao and Max M. Häggblom
Rutgers, State University of New Jersey
npriya@eden.rutgers.edu

Anaerobic bacterial respiration of selenium oxyanions
Selenium is a naturally occurring element in the Earth’s crust. Selenate and selenite are the mobile toxic oxyanions that gain entry into water systems due to oxidation-reduction reactions. Microorganisms have the capability to use selenate as a terminal electron acceptor by reducing it to selenite and elemental selenium through the process known as dissimilatory selenate reduction (DSeR). The diversity of microorganisms catalyzing this reaction is largely unknown. The overall objective of this paper is to gain an in-depth understanding of anaerobic biotransformation of selenium in the environment (with emphasis on respiration of selenium oxyanions) and elucidate the microorganisms catalyzing this process. Here we demonstrate the enrichment and isolation of selenate reducing bacteria from a wide range of sediments.

Samples were obtained from three water bodies in Chennai, India and the Meadowlands, NJ. Novel anaerobic dissimilatory selenate-reducing bacterial strains were isolated from these selenate-reducing enrichments. All these strains are phylogenetically distinct, belonging to various phyla in the bacterial domain. Based on 16S rRNA gene sequence analysis, strain AK4OH1 is classified as a new genus Sedimenticola selenatireducens, strain S5 as a new genus Selenospirillum indicus, strain KM as a new species Pelobacter selenogenes and a new selenate-respiring strain of Pseudomonas stutzeri. Strain S5 and Strain KM possess the unique ability to reduce selenate completely to elemental selenium as confirmed by XANES analysis. We have also successfully demonstrated the presence of a selenate reductase gene (serA) in strain AK4OH1 and P. stutzeri strain pn1. Using P. stutzeri strain pn1 as a model system, the effect of nitrate reduction on selenate reduction was examined. We found that selenate reduction was not inhibited by nitrate reduction, but there was a preferential utilization of nitrate even in selenate induced cultures. We have thus shown that there exists tremendous phylogenetic diversity of selenate-respiring bacteria in the environment, but it is still unknown why these diverse bacteria possess this conserved metabolic function. Further analysis of the biochemistry of the selenate reductase genes in other isolates will shed some light on the diversity of these bacteria in their natural environment.

Phil Papadopoulos, Mason Katz, and Brian Dunne
San Diego Supercomputer Center and Calit2
phil@sdsc.edu

The replicable grid for CAMERA infrastructure
The Community Cyberinfrastructure for Advanced Marine Microbial Ecology Research and Analysis (CAMERA) project is providing a replicable cyberinfrastructure of compute, database, portal, and visualization resources to enable the scientific community to use the vast and rapidly growing treasure of metagenomic information to accelerate understanding of biology and deliver novel biological solutions to important societal challenges in health care, energy, and the environment. The first instantiation of this infrastructure is being constructed at Calit2 with the initial purpose of hosting the J. Craig Venter Institute genomic databases (Global Ocean Survey), analysis tools, and portals. 

This infrastructure includes a medium-scale (512 CPUs) compute cluster, database servers, web farms, storage arrays (~200 TB), and 50 megapixel display walls, all interconnected over a 10-Gbps campus-wide network.  What’s novel about this approach is that the software stack (operating system to application) will be open and replicable by other groups to either extend the CAMERA Grid infrastructure or simply create personal mini-CAMERA systems.  Part of the key software technology for this is the SDSC Rocks cluster distribution, which allows non-experts to rapidly deploy compute and visualization clusters, and has a worldwide user base of several thousand scientists.  This poster details the existing and planned infrastructure for CAMERA, and highlights the methods to make this a completely replicable system.

Alexandra Purdy1, Robert Fieldhouse2, A. Rod Merrill2, and Douglas H. Bartlett1
1Scripps Institution of Oceanography, UCSD
2University of Guelph, ON, Canada
apurdy@ucsd.edu

A subtractive hybridization study of Vibrio cholerae environmental isolates reveals the presence of a novel toxin-like enzyme with ADP-ribosylation activity
Vibrio cholerae is a genetically diverse species that inhabits coastal oceans and estuaries, and only a small number of strains are responsible for cholera pandemics.  To uncover adaptations and survival mechanisms of environmental strains, a sequencing approach was taken.  Suppressive subtractive hybridization revealed the presence of gene segments in V. cholerae strains from southern California, but not present in the sequenced V. cholerae genome, that of O1 El T, or N16961.  A total of 0.2 Mb and 0.14 Mb of novel sequence was acquired from strains SIO and TP respectively, which provides a lower limit for the novel sequence in these strains.  Because this method does not ensure recovery of all novel sequences, a non-parametric estimator was used to show that as much as 0.7 - 0.8 Mb could be new sequences.  The power of this tool to uncover novel functional genes is demonstrated through the subsequent study of a new toxin-like gene with similarity to exotoxin A of Pseudomonas aeruginosa.  The enzyme has been over-expressed and purified, and it functions as an eEF-2-specific ADP-ribosylating enzyme. The role for this protein in the survival of these V. cholerae strains is currently under investigation using environmentally relevant microcosm studies.

A. Ramette, S. Boer, A. Boetius, and J.M. Tiedje
Max Planck Institute for Marine Microbiology
aramette@mpi-bremen.de

Making sense of ecological complexity
Future scientific efforts will focus on the acquisition of a large amount of data concerning the diversity of microbes in situ and the characterization of their physical and chemical environments in a spatio-temporal context. A current issue facing environmental microbiologists is the extraction and interpretation of useful ecological information out of such complex databases.

Here we present an analytical strategy aiming at revealing the most important trends in complex ecological databases where the patterns of variation in biological tables or in environmental data are determined alone or in combination. The use of simple, partial, or canonical ordination techniques along with indicator species analyses may provide a better way to extract ecological principles hiding in huge databases.

The usefulness of such approaches, which so far have been rarely used in microbial environmental genomics, is illustrated by two case studies, one dealing with terrestrial ecosystems and another with coastal sandy sediments. In both cases, multivariate data analyses identified the major trends in the datasets as well as the amount of meaningful biological variation that could be explained by environmental parameters, space, time, or combinations thereof.

Parke Rublee, Michael M. Marshall, and Vincent C. Henrich
University of North Carolina, Greensboro
rublee@uncg.edu

What does a metagenomics approach tell us about the “everything is everywhere” hypothesis?
The “everything microbial is everywhere” hypothesis proposed in the 20th century by Baas-Becking has received considerable attention in the decade following the publication of Findlay, et al., in 1996 (Global diversity and body size. Nature 383:132-133).  We generated eukaryotic and prokaryotic SSU rDNA libraries from three lakes: oligotrophic Toolik Lake in Alaska, and two temperate reservoirs in North Carolina (mesotrophic Lake Townsend and eutrophic City Lake).  We sequenced a sample of clones from the libraries and aligned them with each other and with GenBank entries.  Real-time PCR (Q-PCR) was used to test whether the clones found in one of the libraries could be detected in other lake samples.  In most cases, the abundance of target taxa was one to several orders of magnitude lower than in the source lake.  Nevertheless, the target microbial taxa were detected by Q-PCR, suggesting that the growth potential of individual populations is dictated by local environmental conditions.  The endemic presence of many microbial taxa in widely dispersed freshwater samples has implications for ecosystem assessment as well as ecosystem resistance and resilience.

Steven Smriga, O. Pantos, R. Edwards, F. Rohwer, E. Sala, S. Sandin, and F. Azam
Scripps Institution of Oceanography, UCSD
ssmriga@ucsd.edu

Exploring coral reef ecosystems along a human population gradient: Microbes in the Northern Line Islands
Anthropogenic stressors have been shown to affect many taxonomic groups in coral reef ecosystems.  Yet the effects of these stressors on coral reef microbial communities are not well understood, and microbial communities in “pristine” coral reefs remain poorly described.  To address these questions, we sampled coral reef waters along a gradient of human populations in the Northern Line Islands.  We measured important environmental factors and characterized several aspects of the microbial communities.  Preliminary results suggest that nutrients, primary productivity, and microbial abundances are higher in reef waters near human populations relative to a “pristine” reef.  Future analyses will address the potential relationships among microbes and macroeukaryotes within these coral reef ecosystems.

David Soergel1, David Tulga2, and Steven Brenner1,2,3
1Biophysics Graduate Group, University of California, Berkeley
2Department of Bioengineering, University of California, Berkeley
3Department of Plant and Microbial Biology, University of California, Berkeley
soergel@berkeley.edu

Interpreting metagenomic data using oligonucleotide signatures
Environmental shotgun sequences are accruing at an ever-increasing rate and promise to revolutionize our understanding of microbial community ecology. However, we presently lack many of the computational methods needed to interpret these data.

We are developing algorithms for understanding metagenomic data using modern methods in computational statistics, with emphasis on use of “genomic signatures” to identify the species origin of sequence reads.  Each microbial species has a surprisingly specific signature consisting of the relative frequencies of short oligonucleotides in its genome. Here, we develop and validate methods for extracting biological knowledge from distributions of oligonucleotides in metagenomic data sets.  From a given sample or set of samples, these methods may provide estimates of species numbers and abundances, strain heterogeneity, phylogenetic relationships, lateral gene transfer, and changes of all of these over time and space.

In addition to helping to answer basic questions about microbial evolution, these methods will have widespread practical applications, particularly in human health, biosensing, bioremediation, and improving metagenomic sampling efficiency.

Mitchell L. Sogin1, Jan W. de Leeuw2, David J. Patterson1, Stefan Schouten2, Lucas J. Stal3,  Gerhard J. Herndl2, Susan Huse1, and Linda Amaral-Zettler1
1Josephine Bay Paul Center for Comparative Molecular Biology and Evolution, Marine Biological Laboratory, Marine Biological Laboratory, Woods Hole, MA
2Royal Netherlands Institute for Sea Research, Texel, The Netherlands
3Netherlands Institute of Ecology, Yerseke, The Netherlands
shuse@mbl.edu

ICoMM, the international census of marine microbes: Unveiling the ocean’s hidden majority
The International Census of Marine Microbes (ICoMM), a recent addition to the Census of Marine Life Program (CoML) seeks to determine what is known, what is unknown but knowable, and what may never be known about the biodiversity of marine microorganisms. ICoMM is a joint venture between The Royal Netherlands Institute for Sea Research (NIOZ) and the Marine Biological Laboratory (MBL) in Woods Hole. The ICoMM Secretariat at Woods Hole hosts the WEB site http://icomm.mbl.edu and the distributed database network MICROBIS. It has sponsored meetings for the four primary working groups (Benthic, Open Ocean and Coastal Systems, Technology, and Informatics and Data Management) and its Scientific Advisory Council. Through a series of workshops that engage the international community of marine microbiologists, ICoMM is forging a large-scale strategic plan to characterize microbial diversity in the sea through molecular approaches (high-throughput ribosomal tag sequencing, genomics, lipidomics, etc.) and build a cyberinfrastructure to index and organize the emerging body of information. The community of microbial oceanographers represented within ICoMM (>100 and growing) recognizes the enormity of the task at hand both in terms of total volume of the oceans (estimated to be 1-4 X 1018 m3) and marine sediments with a potential population of more than 1030 microbial cells. Although a complete census is most likely beyond our grasp, the scientific return will be considerable if the information is integrated with contextual information that can inform us about the interplay between microbe-mediated activities and oceanic processes.

Hikaru Suenaga, Tsutomu Ohnuki, and Kentaro Miyazaki
National Institute of Advanced Industrial Science and Technology (AIST)
suenaga-hikaru@aist.go.jp

Accessing microbial degradation pathways for aromatic compounds by a high throughput screening system for metagenomic libraries
Activity-based screening of metagenomic libraries has a large potential to discover entirely new classes of genes with novel and useful functions. However, the utility of currently used high-throughput screening (HTS) systems is limited; conventional visual screening on agar plate usually identifies only several positives out of ~10,000 clones screened. We considered that this inefficiency is due to the low expression level of foreign genes in E. coli and low detection sensitivity to enzyme activity.

In this study, we developed a liquid-based highly sensitive HTS system in 96-well format. Cells (transformants) were grown in a liquid medium, and enzyme activity was determined in a liquid assay solution. This system allowed us systematic tuning of all parameters necessary to enhance the detectable sensitivity to positive clones with ease. DNA extracted from activated sludge in a phenolic wastewater treatment plant was used for the construction of the metagenomic library using a fosmid vector. We screened for catechol 2,3-dioxygenase (C23O) following production of yellow pigment from catechol. By screening 960 wells (10 x 96-well plate), 91 positive clones were identified. Shotgun DNA sequencing was carried out for some of these clones, which revealed that some deduced amino acid sequences of retrieved C23O show low similarity to that of known related enzymes. In addition, not only the single enzyme but entire degradation pathways for various phenolic compounds were included in the positive fosmids. These clones can thus be useful in biotransformation of phenolic compounds in green chemistry/white biotechnology as well as bioremediation of pollutants.

Vera Tai, Ian Paulsen, and Brian Palenik
Scripps Institution of Oceanography, UCSD
vtai@ucsd.edu

Whole genome microarray analyses of Synechococcus-Vibrio interactions
Major metabolic pathways were affected in Synechococcus WH8102 when grown with a heterotrophic bacterium.  In co-cultures with Vibrio parahaemolyticus, Synechococcus may be experiencing phosphate stress since the expression of phosphate acquisition genes increases and alkaline phosphatase activity is higher than in monocultures.  Expression of cell wall synthesis genes are also up-regulated.  In contrast, gene expression of iron-rich proteins and that involved in detoxifying oxygen radicals was decreased.  This study demonstrates the impact of interspecific microbial interactions on the physiology of a major primary producer and the potential of microarray analyses to investigate changes in gene expression from microbial communities.

Peter J. Turnbaugh, Ruth E. Ley, Michael A. Mahowald, Vincent Magrini, Elaine R. Mardis, and Jeffrey I. Gordon
Washington University, Saint Louis
turnbapj@genetics.wustl.edu

Comparative metagenomic studies of the gut microbiomes of lean and obese mice
The worldwide epidemic of obesity is stimulating efforts to identify host and environmental factors that affect energy balance and can be used as novel therapeutic targets. Studies of germ-free and colonized mice indicate that the intestinal microbial community plays an important role in regulating energy balance. Comparisons of genetically obese ob/ob mice and their lean ob/+ and +/+ litter mates indicate that obesity is associated with a pronounced change in the relative abundance of the two dominant bacterial divisions in the distal gut, the Bacteroidetes and the Firmicutes. To determine whether these changes in gut microbial ecology contribute to obesity, we performed a comparative metagenomic study of microbial community gene content in the distal guts (ceca) of ob/ob, ob/+, and +/+ litter mates. In silico characterization of the functional attributes of the cecal microbiomes of these mice, together with follow-up biochemical analyses, indicate that the ob/ob microbiome has an increased capacity to harvest energy and thus may be a contributing factor to host adiposity. Our study in mice demonstrates the feasibility and utility of applying comparative metagenomics approaches, and the 454 GS-20 pyrosequencer, to mouse models of human physiologic or pathophysiologic states to understand the complex interplay between host genetics, microbial community gene content, and the biological properties of the resulting superorganism.

Taku Uchiyama1 and Kazuya Watanabe2
1National Institute of Advanced Industrial Science and Technology
2Marine Biotechnology Institute
uchiyama-taku@aist.go.jp

An improved inverse PCR scheme for metagenome walking
Inverse PCR has been used for the recovery of genome regions franking a known sequence, although its application to metagenome walking is limited due to inefficient amplification from genome fragments at low copy numbers. Here we present an improved inverse PCR scheme that enables walking of a low-copy-number fragment in an environmental metagenome. Our scheme includes the following steps; (i) inverse PCR in which one primer is connected to an affinity tag (e.g., biotin), (ii) affinity purification of amplicons for removing the background metagenome, and (iii) nested PCR to recover target franking regions (inverse affinity nested PCR, IAN-PCR).

A model experiment was conducted in which dilutions of the Ralstonia eutropha genome were mixed with the Escherichia coli genome, and genome regions franking a Ralstonia phenol-hydroxylase gene fragment were recovered by inverse PCR. It was shown that the franking regions were recovered by standard inverse PCR when the ratio of the Ralstonia genome to the Escherichia genome was greater than 1/10, and they were recovered by inverse PCR coupled to nested PCR when the ratio was greater than 1/1000. In contrast, the franking regions were recovered by IAN-PCR at 1/100,000 and greater. The utility of IAN-PCR was also demonstrated by recovering franking regions of PCR-amplified putative chitinase gene fragments from a groundwater metagenome. Since the natural genetic diversity is supported by rare sequences, we suggest that IAN-PCR facilitates access to undiscovered diverse genes in the environment.

Yusuke Unno1,2, Jun Wasaki3, Takuro Shinano3, and Mitsuru Osaki1
1Graduate School of Agriculture, Hokkaido University
2JSPS
3CRIS, Hokkaido University

White lupin’s rhizosphere soil microorganisms analyzed by using a metagenomic approach
One of the subjects in modern agriculture is the depletion of phosphate rock, material of phosphorus fertilizer. White lupin (Lupinus albus L.) possesses a particularly high organic acid-releasing ability and shows more vigorous growth under low soil P condition than other plant species. Previous study indicated that some microorganisms living in the rhizosphere soil might contribute to lupin's ability to absorb P from low P soil (Unno et al., 2005). Rhizosphere soil, the soil in the immediate vicinity of roots, is characterized by high microbial densities (Hiltner, 1904); stimulation of microbial growth by roots is commonly known as the rhizosphere effect, however, the details of this effect on the functional diversity of soil microorganisms are still unclear because more than 99% of microbial species in nature are refractory to cultivation in present laboratory growth conditions. Recently, the results of comparative analysis of environmental metagenomes were reported, and it is suggested that this approach can be an extremely valuable tool for characterization of complex microbial communities. Our study aimed to introduce a metagenomic approach for a pilot-scale analysis of the rhizosphere effect on functional diversity of soil microorganisms. 

Rhizosphere soil samples were collected from the adhering soil to roots of rhizobox-grown white lupin plants, and bulk soil samples were obtained from unplanted pot. DNA samples were extracted from 0.5g wet weight of homogenized bulk soil by using ISOIL (Nippon gene). While in the case of rhizosphere soil as the obtained soil is lower than 0.5 g, soil DNA was amplified by using GenomiPhi (GE Healthcare) based on Multiple Displacement Amplification (MDA) technology after DNA was extracted by ISOIL. To check the quality of MDA on the extracted DNA sample, partial 16S rDNA fragments were amplified by using bacteria universal specific primers, then subjected to DGGE analysis. The result showed that MDA is feasible to obtain sufficient amount of DNA for metagenomic analysis with minimal amplification bias from a small community sample, such as the rhizosphere soil. For the analysis of metagenomes using complementary and less expensive methods, Suppressive Subtractive Hybridization (SSH) was performed using total genomic DNA isolated from rhizosphere soil as a tester that was previously treated with MDA, and bulk soil as a driver. After DNA sequence analysis of the SSH fragments, computer-assisted DNA-DNA comparison using FASTA program was performed. Sequences from 46.1% subtractive hybridization fragments exhibited similarity to Bacterial, 26.9% subtractive hybridization fragments exhibited similarity to Eukaryotic containing sequences of white lupin origin. 

A comparative metagenomic approach using MDA and SSH on limited samples represents an efficient methodology for identifying unique sequences present in one complex microbial community and will likely provide a valuable method for the analysis of rhizosphere microbial dynamics.

References; Hiltner (1904) Arb Deut Landw Gesell 98: 59-78, Unno et al., (2005) Environ Microbiol 7: 396-404.

Rebecca Vega Thurber and Forest Rohwer
San Diego State University
rvegathurber@gmail.com

Community changes in the coral holobiont after exposure to environmental stressors: using metagenomics for hypothesis testing
Unique communities of microbes (both Bacteria and Archaea) and viruses live on and within corals. The combination of coral animal, algal symbiont, and microbiota has been termed the coral holobiont. The abundance, structure, and activity of these associated taxa are hypothesized to play important roles in the health and homeostasis of the host.

This study aims to uncover these roles by comparing and contrasting the metagenomes of the coral holobiont under different conditions. Porites compressa corals were exposed to a variety of stressors known to cause demise, disease, and death of coral reefs - elevated temperature, increased nutrient concentration, dissolved organic carbon (DOC) loading, and reduced pH. Coral slurries were separated into microbial and viral fractions. The resulting genomic DNA was isolated and sequenced using 454 Life Sciences pyrosequencing. The metagenomes showed that increasing the water temperature results in a 67% reduction in the relative abundance of 18S rDNA sequences similar to zooxanthellae (i.e., the intracellular dinoflagellates). This reduction is consistent with previous studies of coral bleaching. DOC exposure led to an increase in the number of hits to ssDNA eukaryotic viruses and a reduction in herpes-like viruses. Overall, this study demonstrates that metagenomics can be used to evaluate and understand how entire communities change in response to environmental stress.

Linda Wegley, Rob Edwards, Beltran Rodriguez-Brito, Selina Liu, and Forest Rohwer
San Diego State University
lwegley@gmail.com

Metagenome of the microbial community inhabiting the coral porites astreoides
Coral-inhabiting microbes form specific associations with their host, but how or why these associations persist is unknown. Two questions that remain unresolved are how do these microbes benefit from living within the coral, and do these microbes offer protection or advantages for the coral? Current methods using 16S rDNA sequencing provide information on the microbial composition of the environment but do not provide information on the microbial functions. In contrast, metagenomic analyses provide a unique opportunity to describe all the taxonomic components, relative abundance and metabolic potential of any microbial community.

A fractionation procedure was used to separate the microbes from the Eukaryotic cells associated with the coral animal. Microbial DNA was sequenced using 454 pyrosequencing. This data showed that 16S comparisons to the Ribosomal Database revealed the presence of Proteobacteria and Firmicutes, which are commonly found on corals. Sequence similarities to Eukaryotic viruses suggest that corals may act as vectors for viral infection of other ocean inhabiting organisms (i.e. fish, insects, and birds). Subsystem comparisons against the SEED database suggests that microbes within coral reefs are fighting hard against virulence, stress responses, and RNA metabolism compared to microbes from the water surrounding coral reefs.