Genomes to Life Contractor-Grantee Workshop II
February 29-March 2, 2004, Washington, D.C.
Genomics:GTL Program Projects
Shewanella Federation
20
Global and Physiological Responses to Substrate Shifts in Continuous and Controlled Batch Cultures of Shewanella oneidensis MR-1
Jim Fredrickson (jim.fredrickson@pnl.gov), Alex Beliaev, Bill Cannon, Yuri Gorby, Mary Lipton, Peter Liu, Margie Romine, Richard Smith, and Harold Trease
Pacific Northwest National Laboratory, Richland, WA
Collaborating Shewanella Federation Team Leaders: Carol Giometti (Argonne NL); Eugene Kolker (BIATECH); Ken Nealson (USC); Monica Riley (MBL); Daad Saffarini (UW-M); Jim Tiedje (MSU), and Jizhong Zhou (Oak Ridge NL)
Shewanella oneidensis MR-1 is a facultative g-Proteobacterium with remarkable metabolic versatility in regards to electron acceptor utilization; it can utilize O2, nitrate, fumarate, Mn, Fe, and S0 as terminal electron acceptors during respiration. This versatility allows MR-1 to efficiently compete for resources in environments where electron acceptor type and concentration fluctuate in space and time. The ability to effectively reduce polyvalent metals and radionuclides, including solid phase Fe and Mn oxides, has generated considerable interest in the potential role of this organism in biogeochemical cycling and in the bioremediation of contaminant metals and radionuclides. The entire genome sequence of MR-1 has been determined and high throughput methods for measuring gene expression are being developed and applied. This project is part of the Shewanella Federation, a multi-investigator and cross-institutional consortium formed to achieve a systems level understanding of how S. oneidensis MR-1 senses and responds to its environment.
Electron Acceptor Responses. To define the networks of genes responding to metal electron acceptors, mRNA expression patterns of cells reducing fumarate were compared to those reducing nitrate, thiosulfate, DMSO, TMAO, and several forms of Fe(III) and Mn(III) using whole-genome arrays of S. oneidensis. Analysis of variance performed on the complete dataset identified over 1600 genes displaying significant expression changes across different metal-reducing conditions. Two principal components accounted for 78% of the variability within the multiple-electron- acceptor dataset and were represented by genes displaying specific response to metals. Hierarchical clustering revealed a high degree of similarity in mRNA relative abundance levels was displayed for all the metal-reducing conditions; all clustering separately from the inorganic electron acceptors. Interestingly, no significant differences in expression profiles were observed between solid and soluble metal acceptors. Only a few genes specific for any particular metal were identified. In contrast, K-means clustering identified a group of over 150 genes displaying highly specific up-regulation under all metal-reducing conditions. Among those, we identified putative transporters, outer membrane components, as well as two electron transfer proteins (flavodoxin and a c-type cytochrome). Further work will be aimed at differentiating cells responses to divalent metal cations (i.e., reduction products) and the oxidized form of the metals; and functional characterization of the differentially regulated genes.
Response to O2Concentrations. Autoaggregation occurs in Shewanella oneidensis MR-1 cultures growing at high O2 concentrations in the presence of Ca2+ ions. Despite the potential environmental importance of this phenomenon, little is known about the mechanisms inducing aggregate formation and subsequent impacts on cells inside the aggregates. In an effort to elucidate these mechanisms and identify processes associated with O2-induced autoaggregation in S. oneidensis, a comparative analysis using DNA microarrays was performed on samples grown under different O2 tensions in the presence and absence of Ca2+. Although, when compared to O2-limited conditions, both flocculated and unflocculated cells displayed some similarities in gene expression in response to elevated levels of O2, including genes involved in cell envelope functions and EPS/LPS production, autoaggregation had a significant impact on gene expression in MR-1. Direct comparison of aggregated versus nonagreggated cells grown under 50% dissolved O2 tension (DOT) revealed remarkable differences in mRNA patterns between these two states. The nonaggregated cells displayed significant increase of mRNA levels of genes involved in aerobic energy metabolism, amino acid and cofactor biosynthesis, as well as chemotaxis and motility. In contrast, genes putatively involved in anaerobic metabolism (fumarate and polysulfide reductases, and Ni/Fe hydrogenase), cell attachment (type IV pilins and curli), and transcription regulation (rpoS, spoIIAA) were upregulated under 50% DOT aggregated conditions. Notably, a gene cluster encoding outer membrane proteins and cytochromes (mtrDEF) also displayed up to 7-fold increase in mRNA levels in aggregated cells. Although further studies are required for resolution, we speculate that autoaggregation in S. oneidensis MR-1 may serve as a mechanism to facilitate reduced O2 tensions within aggregate, leading to the expression of anaerobic genes under bulk aerobic conditions.
Carbon Metabolism. In contrast to the wide array of electron acceptors reduced by S. oneidensis, this organism is relatively limited in regards to utilization of multicarbon substrates for anaerobic respiration. Earlier studies indicated that MR-1 can utilize formate as a sole source of carbon and energy under anaerobic condition. A hypothetical amalgam pathway for lactate metabolism that included the elements of serine-isocitrate cycle for formate utilization was previously proposed by K. Nealson and colleagues. The availability of whole-genome sequence allowed compilation of a possible pathway for formate assimilation in S. oneidensis. MR-1 is predicted to possess a number of putative enzymes including pyruvate formate-lyase (PFL) that may allow for the assimilation of both exogenous and lactate-derived formate through a modification of a serine-isocitrate pathway. Three independent lines of evidence of methylotroph-like metabolism of lactate in S. oneidensis MR-1 are provided: lactate is stoichiometrically converted to acetate in O2-limited or lactate-excess anaerobic chemostat cultures and no formate is present in the supernatant; the amount of PFL protein increased 2.5-fold under O2 limited growth compared to fully aerobic growth; and relative abundance of mRNA from genes encoding key enzymes of the proposed pathway for formate assimilation including isocitrate lyase, malate synthase, and serine hydroxymethyl-transferase increased under O2-limitation compared to aerobic growth. Moreover, biomass yield from O2-limited or anaerobic chemostat cultures of MR-1 grown under excess lactate indicated that formate was utilized as the sole source of carbon. These results suggest the presence of an unusual mechanism of carbon metabolism of lactate in S. oneidensis with formate as the key element of the intermediary carbon metabolism under O2-depleted and/or lactate excess conditions.
Protein Secretion. In many bacteria, translocation of key respiratory enzymes is mediated by the twin arginine translocation (TAT) machinery. Analysis of the N-terminal sequences of MR-1 proteins for conserved TAT leader properties revealed 30 candidates. This list includes 11 genes predicted to be responsible for utilization of formate and H2 as electron donors or for catalyzing the terminal step in electron transfer to nitrate, nitrite, and sulfur-containing substrates. Various uncharacterized proteases and putative redox active proteins were among the remaining 19 candidates. In order to investigate the role of secreted proteins in MR-1 respiratory metabolism, two mutants were constructed: a tatC gene deletion mutant and a transposon mutant of the terminal branch of type II general secretion pathway (gspD). The TAT mutant was unable to reduce metals with formate or H2 as an electron donor. Although cells could grow with lactate and fumarate, there was a longer lag phase relative to the wild type. As expected, the ability to grow on nitrate and DMSO was abolished. Reduction of technetium (TcO4-) with lactate was severely impaired in the TAT mutant, likely due to mislocalization of one or more hydrogenases. Hydrogenase have reported to be involved in Tc reduction in other bacteria. The TAT system is a key cellular machine in MR-1 that is essential for its diverse energy metabolism.
A comprehensive proteomic approach was applied to identify candidate GSP proteins including those potentially involved in metal-reduction. Steady state O2-limited wild type and gspD mutant cells of MR-1 were sampled from bioreactors for mass spectral proteomics and metal reduction activities. Proteome analysis of whole cells, cell fractions, membrane vesicles, and extracellular proteins revealed the mislocalization of several hypothetical proteins in the mutant compared to the wild type, as well as OmpW. OmpW and one hypothetical exhibited increased expression in cells incubated with various metals. The localization of MtrA, MtrB, and MtrC, proteins previously implicated in metal reduction, were unaffected according to proteome and western blot analyses. These results demonstrate that while the GSP is necessary for efficient metal reduction in these cells, several key electron transfer proteins essential for Fe(III) and Mn(IV) reduction were not mislocalized. By applying a combination of controlled cultivation integrated with proteome measurements genes that are candidate secreted proteins, including those involved in metal transformation, can be identified.
21
Integrated Analysis of Gene Functions and Regulatory Networks Involved in Anaerobic Energy Metabolism of Shewanella oneidensis MR-1
Jizhong Zhou1 (zhouj@ornl.gov), Dorothea K. Thompson1, Matthew W. Fields1, Timothy Palzkill2, James M. Tiedje3, Kenneth H. Nealson4, Alex S. Beliaev5, Ting Li1, Xiufeng Wan1, Steven Brown1, Dawn Stanek1, Weimin Gao1, Feng Luo1, Jianxin Zhong1, Liyou Wu1, Barua Soumitra1, Crystal B. McAlvin1, David Yang1, Robert Hettich1, Nathan VerBerkmoes1, Yuri Gorby5, Richard Smith5, Mary Lipton5, and James Cole3
1Oak Ridge National Laboratory, Oak Ridge, TN; 2Baylor College of Medicine, Houston, TX; 3Michigan State University, East Lansing, MI; 4University of Southern California, Los Angeles, CA; and 5Pacific Northwest National Laboratory, Richland, WA
Collaborating Shewanella Federation Team Leaders: Jim Fredrickson, Pacific Northwest National Laboratory, Richland, WA; Carol Giometti, Argonne National Laboratory, Argonne, IL; Eugene Kolker, BIATECH, Bothell, WA; and Monica Riley, Marine Biological Laboratory, Woods Hole, MA
Shewanella oneidensis MR-1, a facultatively anaerobic g-proteobacterium, possesses remarkably diverse respiratory capacities. In addition to utilizing oxygen as a terminal electron acceptor during aerobic respiration, S. oneidensis can anaerobically respire various organic and inorganic substrates, including fumarate, nitrate, nitrite, thiosulfate, elemental sulfur, trimethylamine N-oxide (TMAO), dimethyl sulfoxide (DMSO), Fe(III), Mn(III) and (IV), Cr(VI), and U(VI). However, the molecular mechanisms underlying the anaerobic respiratory versatility of MR-1, however, remain poorly understood. In this project, we have integrated genomic, proteomic and computational technologies to study energy metabolism of this bacterium from a systems-level perspective.
Molecular Responses to Anaerobic Growth with Different Electron Acceptors. To define the repertoire of MR-1 genes responding to different terminal electron acceptors, transcriptome profiles were examined in cells grown with fumarate, nitrate, thiosulfate, DMSO, TMAO, ferric citrate, ferric oxide, manganese dioxide, colloidal manganese, and cobalt using DNA microarrays covering ~99% of the total predicted protein-encoding open reading frames in S. oneidensis. Total RNA was isolated from cells grown anaerobically for 3.5 hours in the presence of different electron acceptors and compared to RNA extracted from cells grown under fumarate-reducing conditions (the reference condition). More than 1600 genes display significant expression changes across different metal-reducing conditions. Real-time PCR analysis for some selected genes showed that microarray-based quantitation is highly accurate. Hierarchical cluster analysis indicated that genes showing differential expression under metal-reducing conditions generally clustered together, whereas genes showing differences in mRNA abundance levels under non-metal respiratory conditions clustered together. Interestingly, no significant differences in expression profiles were observed between solid and soluble metal acceptors. Only a few genes specific for any particular metal were identified. In contrast, a group of over 150 genes displaying highly specific up-regulation under all metal-reducing conditions were identified, including, putative transporters, outer membrane components, as well as two electron transfer proteins (flavodoxin and a c-type cytochrome). In addition, a number of genes, of which 35-55% encoded hypothetical proteins, were uniquely induced or repressed in response to a single electron acceptor. This work has yielded numerous candidates for targeted mutagenesis and represents an important step towards the goal of characterizing the anaerobic respiratory system of S. oneidensis MR-1 on a genomic scale.
Phage Display. Along with mass spectrometry, two-hybrid system and protein arrays, phage display is another powerful technique for studying protein-ligand interactions. The first key step of mapping protein interactions was to clone all protein-coding ORFs to allow exogeneous expression of its protein for functional analysis. We have spent tremendous efforts in cloning genes into an universal vector. Progress as of the close of 2003 is that 1,691 genes were cloned while no clones were obtained for 174 genes. Additionally, a random phage display library utilizing “shotgun” cloning of sheared S. oneidensis genomic DNA has been constructed.
Genetic Mutagenesis. One of the most powerful ways to define the function of a gene is to turn the gene off or change the expression by replacing the normal gene with a mutated counterpart. We have successfully modified and utilized vector systems for homologous recombination in S. oneidensis MR-1. Currently, our laboratory is interested in understanding transcriptional gene regulation in S. oneidensis. We are targeting approximately 220 annotated transcription factors (TFs) for knockout mutatgenesis. Analysis of the S. oneidensis genome sequence suggests that insertional mutagenesis is appropriate for only 78 of these TFs as they are transcribed in their own operon. We have also systematically knocked out some of the genes involved in metal reduction as revealed by microarray analysis discussed above.
Numerous other MR-1 genes have been successfully inactivated using a PCR-based, in-frame deletion mutagenesis strategy. Our current collection of deletion mutants includes those strains with mutations in etrA, arcA, fur, crp, fur/etrA, etrA/crp, rpoH (sigma-32), ompR, envZ, oxyR, cya1-3 (adenylate cyclases), and many others. Microarray-based gene expression profiling has been used to analyze a number of these mutant strains. For example, we have employed whole-genome DNA microarrays, large-scale proteomic analysis using liquid chromatography-mass spectrometry (LC-MS), and computational motif discovery tools to define the S. oneidensis Fur regulon. Using this integrated approach, we identified 9 probable operons (containing 24 genes) and 15 individual ORFs of either unknown function (SO0447-48-49, 0798-97, 0799, 1188-89-90, 2039, 3025, 3027, 3406-07-08, 3062, 3344, 4700, 4740) or annotated as encoding transport and binding proteins (ftn, bfd, SO1111-12, 1482, 1580, feoAB, alcA-3031-32, 3669-68-67, tonB1-exbB1-exbD1, viuA, irgA, 4743) that are predicted to be direct targets of Fur-mediated repression based on their up-regulated expression profiles in a fur deletion mutant and the presence of potential Fur-binding sites in their upstream regulatory regions. This study suggests, for the first time, a possible role of 4 operons and 8 ORFs of unknown function in iron metabolism.
Chemostat Growth Studies with MR-1 Mutant Strains. Using the growth facility at PNNL, Shewanella oneidensis etrA and arcA deletion strains and the parental strain were each grown in chemostats in continuous culture for 410 hours. The growth conditions were altered from an aerobic steady state, to a microoxic steady state and to an anaerobic steady state to examine the contribution of each regulator in S. oneidensis. Samples were collected at each steady state for organic acid, proteome, cytochrome and transcriptome analyses. Samples were also harvested at 0, 5, 10, 20, 30, 40, 50, 60, 90, 120, and 150 minutes after transition from aerobic to microoxic steady states for mRNA and protein analysis.
Elucidation of the Functions of a Conserved Hypothetical Protein. Whole-genome sequence analyses of a variety of microorganisms indicated that 30-60% of the identified genes encode functionally unknown proteins. Defining the functions of hypothetical proteins is a great challenge. Integrated approaches for systematic study of their functions are needed. As a first attempt, an in-frame deletion mutant was generated for the conserved hypothetical protein of 592 amino acids, SO1377. Physiological analysis showed that this mutant was very sensitive to hydrogen peroxide, showed slow growth rate under aerobic condition but not anaerobic conditions, and had higher spontaneous mutation rates. Microarray analysis revealed that numerous genes are affected by this mutation. Computational analyses of secondary and tertiary structure also revealed that the protein could have potential functions in formation of protein complexes at the inner bacterial cell membrane, ATP/GTP binding, nucleotide binding, protein transport and molecular chaperone. Overall, our results suggested this gene could be involved in iron homeostasis and oxidative damage protection in S. oneidensis MR-1.
Molecular Basis of Stress Responses. Other work related to Shewanella focuses on the elucidation of the molecular basis of bacterial adaptive responses to various environmental stresses, namely, heat stress, cold stress, high salt, low/high pH, oxidative stress, and metal toxicity. These studies employ primarily global gene expression profiling using cDNA/oligonucleotide microarrays and targeted gene mutagenesis. The initial manuscripts for two of these studies (heat shock and salt stress) have already been written and have been submitted for publication or are close to being submitted. In the study on oxidative stress, the effect of H2O2-induced oxidative stress on the gene expression profiles of S. oneidensis wild type and mutant strains was investigated. Microarray analysis of the wild type cells indicated significant changes in the expression levels of numerous genes that are known or have not been previously described to be involved in the oxidative stress responses of other bacterial species. Among these are the alkyl hydroperoxide reductase (Ahp) gene, the catalase (Kat) gene, the stress response DNA-binding protein (dps) gene, and the genes involved in the TonB transport systems. In addition, a LysR family transcriptional regulator showed immediate yet transient upregulation in response to H2O2 treatment, suggesting the hypothesis that it regulates H2O2 stress responses in Shewanella oneidensis. Sequence comparison and computational modeling predicted the gene to be the potential analog of the E. coli OxyR gene. Yet phenotype characterization of the deletion mutant of the gene revealed interesting responses toward various oxidative stimuli. Global expression profiling of the mutant indicated that the LysR regulator indeed controlled some of the genes that had been reported to belong to the OxyR regulon in E. coli, but it also regulated many uncharacterized genes.
In another study, we examined the response of S. oneidensis to high levels of heavy metals to better understand the repertoire of genes and regulatory mechanisms enabling heavy metal resistance. MR-1 was able to grow in LB medium with strontium (Sr2+) concentrations as high as 180 mM, but showed substantial growth inhibition at levels above 180 mM. S. oneidensis resistance to 180 mM Sr was examined using DNA microarrays. Transcriptome profiles were generated from mid-exponential phase bacteria grown in the presence of Sr2+ and compared to profiles from MR-1 cultured to the same growth phase in the absence of strontium. The stress response of S. oneidensis to a shock addition of 180 mM Sr was also examined after 5, 30, 60 and 90 minutes using microarrays. Siderophore biosynthesis and iron uptake genes were highly induced (up to 622 fold) and a siderophore biosynthetic mutant was more sensitive to strontium, suggesting that siderophore production plays an integral role in the ability of S. oneidensis to mediate strontium resistance.
Network Modelling. Understanding the regulatory interactions between thousands of genes in a given organism from massive time-course microarray data is one of the most challenging tasks in the field of microbial functional genomics. Currently, the inference of such genetic interaction networks is harmed by the dimensionality problem because the number of genes in a genome far exceeds the number of measured time points due to high cost of measurements. It is essential to develop powerful computational tools to extract as much biological information as possible from ambiguous expression data containing noise. Different from existing methods, we are developing a computational method based on random matrix theory. We are using the matrix of pair-wise correlation to identify connections between genes. In contrast to other network identification methods, the threshold for defining network links is determined automatically and self-consistently based on the data structure itself. We have applied this method to identify regulatory networks in yeast based on the massive available microarray data. The identified gene interactions were very consistent with our knowledge, suggesting that this method is very useful for network identification. We are now further testing this method based on microarray data from Shewanella, Deinococcus, yeast, worm, fly and human.
22
Profiling Shewanella oneidensis Strain MR-1: Converting Hypothetical Genes into Real, Functional Proteins
Eugene Kolker (ekolker@biatech.org), Samuel Purvine, Alex F. Picone, Natali Kolker, and Tim Cherny
BIATECH, Bothell, WA
Collaborating Shewanella Federation Teams: J. Fredrickson, M. Romine, Y. Gorbi, A. Beliaev, B. Cannon (PNNL); R. Smith, G. Anderson, K. Auberry, M. Lipton, D. Elias (PNNL); J. Tiedje, X. Qiu, J. Cole (MSU); K. Nealson, S. Tsapin (USC); M. Riley, M. Serres (MBL); C. Giometti, G. Babnigg (ANL); J. Zhou, D. Thompson (ORNL).
Other Collaborating Teams: E. Koonin, M. Galperin, K. Makarova (NCBI); C. Lawrence, L.-A. McCue (WC); B. Palsson, A. Raghunathan, N. Price (UCSD); H. Heffelfinger, J. Timlin (SNL); J. Yates, W. Zhu (Scripps).
The progress in genome sequencing has led to a rapid accumulation in GenBank submissions of uncharacterized “hypothetical” proteins. These proteins, which have not been experimentally characterized and whose functions cannot be deduced from simple sequence comparisons alone, now comprise approximately one third of the public databases. That is, despite significant progress in the experimental research, this so called “70% hurdle” still holds, with every new genome bringing novel unknown proteins numbering in the hundreds or even thousands. Being very complex and fascinating in numerous aspects of its behavior and responses, Shewanella oneidensis strain MR-1 (SO) presents an even greater challenge, as over half of its predicted genes are considered hypothetical. If past performance in experimental characterization of new proteins from Escherichia coli K-12, roughly 25 per year, is of any predictive power, it will take many decades before the biological function of all these (SO) proteins is discovered.
Expression profiling of SO cells under multiple growth conditions done by the Shewanella Federation consortium was performed. Among the performed experiments are continuous and controlled batch cultures of SO cells under a variety of different environmental conditions. These include electron acceptors and substrates, and limitations of thereof, such as O2, Ca2+, and Pi-limitations and UV-radiation stresses. Special emphasis was placed on robust, reproducible, and statistically validated results, rather than optimizing coverage of the expressed gene and protein contents for the above conditions. Earlier studies of SO presented a baseline of over 4,600 predicted genes with approximately 2,350 hypothetical ones.
SO gene profiling resulted in conservative estimation of over 4,000 expressed genes, including identification of over 1,900 hypothetical genes. Protein profiling experiments conservatively estimated approximately 1,550 expressed proteins with approximately 500 hypothetical ones. Using a combination of transcriptomic and proteomic approaches as well as statistical and computational methods, this analysis confidently identified over 450 hypothetical genes that were expressed in cells both as genes and proteins. In an attempt to understand the functions of these proteins, we used a variety of publicly available analysis tools. This resulted in exact or general functional assignments for over 200 hypothetical proteins. Accurate functional annotation of uncharacterized proteins calls for an integrative approach, combining expression studies with extensive computational analysis and curation, followed by the directed experimental verification.
23
Systems Biology of Shewanella oneidensis MR-1: Physiology and Genomics of Nitrate Reduction, the Radiation Stress Response, and Bioinformatics Applications
James M. Tiedje (tiedjej@msu.edu), James R. Cole, Claribel Cruz-Garcia, Joel A. Klappenbach, and Xiaoyun Qiu
Michigan State University, East Lansing, MI
Collaborating Shewanella Federation Team Leaders: Jim Fredrickson, Margie Romine, Yuri Gorby (PNNL); Eugene Kolker (BIATECH); and Jizhong Zhou (ORNL)
The Stress Response: Effects of Ultraviolet Radiation. Successful application of Shewanella oneidensis MR-1 in bioremediation applications may necessitate cellular tolerance to toxic levels of pollutants and damage-inducing radiation. Solar ultraviolet radiation (UVR) is perhaps the most mutagenic agent to which many organisms are exposed due to its abundance. We systematically investigated the stress response in MR-1 following exposure to UVC, UVB and UVA radiation. MR-1 showed extremely high sensitivity to both far- and near-UV with a D37 value (UVC) of 5.6% relative to E. coli K12. Photoreactivation conferred a significantly increased survival rate to MR-1 in both UVB and UVC irradiated cells: as much as 177- to 365-fold and 11- to 23-fold survival increase after UVC and UVB irradiation respectively. A significant UV mutability to rifampin resistance was detected in both UVC and UVB treated cells. Different gene expression profiles were observed after UVC, UVB and UVA treatments. More than 300 genes were up-regulated after UVA exposure whereas only about 100 genes were induced after UVC exposure. Although the SOS response occurred in all three treatments, the induction of key genes in the SOS regulon (e.g. recA, lexA, polB etc.) was most robust in response to UVC. Genes that are involved in protection from oxidative damage showed an increased expression level in both UVB and UVA treatments. Unexpectedly, we did not observe induction of genes encoding nucleotide excision repair (NER) components (e.g. uvrA, uvrB and uvrD) in either UVB or UVC treatments. We were also unable to identify any potential SOS box upstream of uvrA, uvrB and uvrD. Complementation of Pseudomonas aeruginosa UA11079 (uvrA-) with uvrA of MR-1 increased the UVC resistance of this strain more than three orders of magnitude, indicating the functionality of UvrA in repairing UVR-induced DNA damage. Using RT-PCR, we detected transcripts of uvrA, uvrB and uvrD from MR-1 in both UVR treated and untreated sample at equivalent levels, indicating that component genes of NER are constitutively expressed. Loss of the damage inducible NER system may contribute to the high sensitivity of this bacterium to UVR.
Aerobic and Anaerobic Nitrate Reduction. Nitrate is often found as a co-contaminant in metal and radionuclide contaminated groundwater, and understanding the response of S. oneidensis MR-1 to these compounds is critical to effective bioremediation applications. S. oneidensis MR-1 is capable of dissimilatory nitrate reduction to ammonia during both aerobic and anaerobic growth. Nitrate is reduced by S. oneidensis MR-1 in a stepwise manner from nitrate > nitrite > ammonia. Complete reduction of nitrate precedes initiation of nitrite reduction, a process controlled by thermodynamics. Genome analysis supports this physiology: S. oneidensis MR-1 possesses genes for a single nitrate reductase (NapA) and a single nitrite reductase (NfrA) that catalyze the reduction of nitrate to ammonia in two enzymatic steps. Expression of napA and nrfA, measured via quantitative PCR, is maximal in anaerobic batch cultures initiated with >0.5 mM nitrate. A decrease in napA and nrfA gene expression with increasing concentrations of nitrate occurs in E. coli, indicating an alterative regulatory system is operating in S. onedensis MR-1. The expression of narP/narQ (NO3/NO2 sensor/response regulator) was constant in cultures fed up to 10 mM nitrate. During aerobic growth conditions with and without nitrate, napA and narP/narQ were equivalently expressed, indicating constitutive expression of nitrate reductase activity independent of the presence oxygen or nitrate.
Nitrite accumulates during lactate-dependent nitrate reduction, and was found to inhibit growth in both aerobic and anaerobic batch culture. Decreased growth rates during chemostat culture did not alleviate nitrite toxicity due to the stepwise reduction of nitrate to ammonia. Anaerobic nitrate reduction was limited to the oxidation of lactate and pyruvate as sole carbon and energy sources - other sugars and carboxylic acids (and hydrogen) did not support growth. Growth limitation due to nitrite toxicity during aerobic batch culture was assessed using whole-genome microarrays. Significant up-regulation of genes encoding heat shock and DNA repair proteins that are associated with oxidative stress occurred in the presence of nitrite. Nitrite also resulted in the significant down-regulation of many genes involved in iron acquisition, possibly as a mechanism of reducing DNA damage induced by hydroxyl radicals generated via intracellular iron oxidation. The capacity for nitrate reduction of S. oneidensis MR-1 may therefore be limited by its ability to mediate oxidative damage induced by nitrite accumulation.
MicroPlateDB – a LIMS for Quality Control and Data Archiving Microplate Data. In a multi-investigator research effort, such as the Shewanella Federation, open-access to research data and protocols is critical to genomics-level investigations involving tools such as microarrays and proteomics. We have continued development of an internet-browser accessible laboratory information management system (LIMS), the ‘MicroPlateDB’, for tracking and archiving data generated during microarray construction. The LIMS is structured with the laboratory microplate as the central data type and the contents of plates are combined during virtual “reactions” as they are carried out in the. Customization of the LIMS is controlled by a set of basic data tables containing information on microplate types, contents, and how contents of microplates are combined and stored during laboratory procedures. User-level permissions control LIMS access and allow a project manager to specify the ability to view and/or modify data on an individual basis. With a login name and password, an investigator performing microarray studies can access the LIMS to find a gene of interest in the plate used to print the array, including the concentration, size, and gel-resolved quality of the PCR-product. The user also has the ability to ‘drill-down’ through a set of hyperlinked microplate graphic representations to track a PCR-product from a spot on the microarray to the primers and template used to create that product. Enhancements currently under development include user-defined searching and an open-source version for public release. The LIMS was initially customized to track process information obtained during the production of a PCR-based DNA microarray for S. oneidensis MR-1 and has also been chosen for use in several other microarray construction projects, including the ORNL Deinococcus radiodurans microarray.
24
Development and Application of Optical Methods for Characterization of Protein-Protein Interactions in Shewanella oneidensis MR-1
Natalie R. Gassman1* (ngassman@chem.ucla.edu), Achillefs N. Kapanidis1, Nam Ki Lee1, 4, Ted A. Laurence1,5, Xiangxu Kong1, and Shimon Weiss1,2,3
*Presenting author
1Dept. of Chemistry and Biochemistry, University of California, Los Angeles, CA; 2Dept. of Physiology, David Geffen School of Medicine, University of California, Los Angeles, CA; 3California NanoSystems Institute, University of California, Los Angeles and Santa Barbara, CA; 4Seoul National University, Seoul, Korea; and 5Lawrence Livermore National Laboratory, Livermore, CA
The increased availability of microbial genomes has increased the drive to elucidate the complex biological networks these microbes utilize to adapt to extreme environmental conditions. Regulation, cell adhesion, and respiration networks, which accomplish bioremediation of metals and radionuclides, are of particular interest to the Genomics:GTL program. Shewanella oneidensis MR-1, a dissimiliatory metal reducing bacterium with the ability to utilize a large variety of electron acceptors, is ideal for bioremediation applications. While widely studied, the mechanism by which MR-1 utilizes these electron acceptors remains unclear. Understanding these complex respiration mechanisms requires the characterization of post-translational macromolecular interactions. We are investigating these numerous protein-protein interactions in MR-1 by developing novel optical methods for their detection and characterization.
_VPBKGD.JPG)
Figure 1. Analysis of a protein-DNA interaction using ALEX-FAMS. A. E-S his togram for D-only, A-only, and D-A species with different RD-A. B. Model of CAP-DNA complex and labeling scheme. Acceptor (red) was placed on DNA, and the donor was placed on 2 possible sites on CAP (green arrows). C. E-S histogram of A-containing species, DNAA. D. E-S histogram of DNAA incubated with CAPd and 0.2 mM cAMP. E. Single molecule sorting allows calculation of biomolecular constants; ALEX-based titration of DNA with CAP in the presence (filled circles) or absence of cAMP (open circles).
We have expanded on a current single-molecule fluorescence spectroscopy (SMFS) method, single-pair Förster resonance energy transfer (spFRET), to measure stochiometry and interactions. spFRET uses a single laser to probe the transfer of excitation energy from a donor (D) fluorophore to a complementary acceptor (A) fluorophore of an interacting pair, yielding a D-A distance sensitive value E, which acts as a “spectroscopic ruler” for the 1-10 nm scale. While an excellent qualitative indicator of molecular interactions, the limited dynamic range of the FRET ruler precludes the measurement of interactions between large macromolecules and/or multimeric complexes. By using an alternating-laser excitation (ALEX) scheme, we have expanded the spFRET technique to report on structure, dynamics, stoichiometries, local environment and molecular interactions. This is accomplished by obtaining D-excitation and A-excitation–based observables for ach single molecule by rapidly alternating between D-excitation and A-excitation lasers. This scheme probes directly both FRET donors and acceptors present in a single diffusing complex and recovers distinct emission signatures for all species involved in interactions by calculating two fluorescence ratios: the FRET efficiency E, a distance-based ratio which reports on conformational status of the species, and a new, distance-independent stoichiometry-based ratio, S, which reports on the association status of the species. Two-dimensional histograms of E and S allow virtual sorting of single molecules by conformation and association status (Fig. 1A). ALEX is a homogeneous, “mix-and-read” assay, where interacting species are combined and optical readouts report simultaneously on their association status and conformational status. The potential applications of this methodology are extensive and characterization of known protein-DNA interactions, Escherichia coli catabolite activator protein (CAP) with DNA (Fig. 1B-D), has illustrated the method’s robust nature.
The complex regulatory mechanisms governing the expression of genes involved in electron transport and energy generation in MR-1 provide a diverse array of protein-DNA and protein-protein interactions that are ideally suited for the ALEX method. One such regulatory mechanism, activated under environmental stress, is the two-component signaling cascade that initiates gene expression by the alternative sigma factor, s54. Transcriptional regulation is achieved through a cascade of protein-protein interaction that results in the interaction of a transcription regulator with the s54-RNA polymerase (RNAP) holoenzyme complex to initiate transcription. One example of this signaling cascade is the interaction of a nitrogen regulatory protein (NtrC) with s54-RNAP holoenzyme to initiate transcription of genes involved in nitrogen fixation in MR-1. Upon stimulus by environmental stress, a sensor protein autophosphorylates resulting in the downstream phosphorylation the transcriptional regulator, NtrC. An NtrC oligomeric form then binds upstream of the promoter region and via a looped DNA intermediate catalyzes the formation of the open transcription complex. Using the ALEX methodology, we can now examine the mechanistic process of gene regulation under stress conditions from the oligomerization of the transcription regulator to the activating interaction between NtrC and the s54-RNAP holoenzyme to initiation of transcription. Progress in protein expression, site-directed mutagenesis and fluorescence labeling of MR-1 NtrC, s54-and RNAP holoenzyme will be reported.
25
Annotation of Genes and Metabolism of Shewanella oneidensis MR-1
Margrethe Serres and Monica Riley (mriley@mbl.edu)
Marine Biological Laboratory, Woods Hole, MA
Annotation
Our continuing annotation of the genes and gene products of Shewanella oneidensis MR-1 is taking advantage of two seldom used sources of information on protein function: (1) functions of structural domains within protein sequences, and (2) the functions of paralogous groups, groups of genes that seem to have descended from the same ancestor and tend to retain related functions.
The Structural Classification of Proteins database (SCOP) has a section describing Superfamilies of structural domains. Using a HMM method, the presence and location of structural domains has been determined for some genomes. The data for S. oneidensis finds structural domains in 2570 of the proteins. The data has been scrutinized in particular for all open reading frames (ORFs) having no functional assignment. 366 of unknown ORFs could be assigned some information on function from this connection.
Many bacterial genomes have genes for proteins that appear to have arisen during evolution by duplication followed by divergence. Distantly but firmly related proteins have been assembled by collecting related sequences (determined by the Darwin algorithm) into groups by transitive relationships. Such groups of sequence similar proteins can include only distantly related members in that similarity to only one protein in the group is sufficient for inclusion. No protein is a member of more than one group. Using this approach, 408 paralogous groups were identified, ranging in size from 2 to 64 members per group.
Paralogous groups give some insight to the numbers of ancestral genes required to generate contemporary bacterial gene families. Also, since paralogous group member of known function show similarity of function (sometimes closely related, sometimes more distantly), any unknown members of a paralogous group can be assigned the common denominator of function for that group.
Metabolism
To survey the metabolic capabilities or S. oneidensis, we recorded similarities to the protein sequences of 50 fully sequenced organisms, again using the Darwin algorithm. The organisms with the largest number of “hits” were Vibrio cholera, Yersinia tuberculosis, Escherichia coli and Pseudomonas aeruginosa. Of these, biochemical information for gene products is far and away greatest for E. coli. Thus it was possible to assign putative enzyme function and pathway existence when homologs exist in E. coli, but not possible when there was similarity to a protein of unknown function in one of the other bacteria, no analogous gene in E. coli. Therefore broadly speaking, the metabolic capacities of S. oneidensis are similar to those of E. coli, but this is partly because more E. coli proteins have been characterized than in the other bacteria. Nevertheless, there were a few functions in for instance Pseudomonas aeruginosa such as part of the beta-ketoadipate pathway that seem to be present in S. oneidensis but not in E. coli. With further exploration of the biochemistry of organisms other than E. coli, more assignment of biochemical capability will be possible.
Biosynthetic capabilities for small molecule cofactors, carriers are largely intact. However in most cases where E. coli has two or more isozymes, S. oneidensis has only one. At present the broad picture for utilization of carbon sources involves very few 5 or 6 carbon sugars and sugar derivatives, rather evidence for the utilization of 3 or 2 carbon carbohydrates and organic acids. There is a defect in an essential enzyme of the glycolytic pathway. However the enzymes for utilization of some 6 carbon sugars, e.g. galactose are present and the Entner-Douderoff pathway seems to be present, completely adequate to serve 5 and 6 carbon substrates. Therefore it is not clear why many 5 and 6 carbon compounds are not utilized. This aspect bears experimental exploration.
Only some of the enzymes for reduction of organic terminal electron acceptors that are found in E. coli seem to be present in S. oneidensis. Formate metabolism is present. The well-known use of metal ions as electron acceptors could abrogate the need for using many organic acceptors. Many electron transfer intermediates are present, consistent with the unusual richness of energy transfer by this organism.
