Genomics:GTL Awardee Workshop VII
Bethesda, Maryland,
February 8–11, 2009
Project Goals: The overall goal of the Center for Molecular and Cellular Systems (CMCS) is to provide a capability for generating high quality protein-protein interaction data from a variety of energy- and environment-relevant microbial species.
An Integrative Strategy for the Determination of the Modular Structure of Functional Networks of Rhodopseudomonas palustris
William R. Cannon2* (William.Cannon@pnl.gov), Don Daly,2 Mudita Singhal,2 Lee Ann McCue,2 Ronald Taylor,2 Dale A. Pelletier,1 Gregory B. Hurst,1 Denise D. Schmoyer,1 Jennifer L. Morrell‑Falvey,1 Brian S. Hooker,2 Chongle Pan,1 W. Hayes McDonald,1,3 Michelle V. Buchanan,1 and H. Steven Wiley2
1Oak Ridge National Laboratory, Oak Ridge, Tenn.; 2Pacific Northwest National Laboratory, Richland, Wash.; and 3Vanderbilt University, Nashville, Tenn.
Protein-protein interactions were used as the foundation for an integrative approach for determining the modular structure of Rhodopseudomonas palustris cellular networks. R. palustris, a metabolically versatile anoxygenic phototrophic bacterium, is the current target for the Genomics:GTL Center for Molecular and Cellular Systems (CMCS). Our analyses have focused on protein interactions observed under differing conditions for nitrogen metabolism in which either NH4+ (fixed nitrogen) or N2 serve as the primary source of nitrogen.
We have developed an approach where multiple sources of information are integrated to provide a more comprehensive perspective of the cellular networks than can be provided by protein interactions alone. The interaction data are integrated with functional clues from operon structure and with manual and automated analysis of functional subsystems from molecular machines and cellular processes. Transcriptional regulatory elements are then overlaid on the protein-interaction maps to provide an integrative perspective on the cellular machines.
Using this approach, we have reconstructed a wide ranging, catalogue of protein complexes and interactions involved in a diverse and rich set of cellular processes including nitrogen fixation, electron transfer, photosynthesis, protein synthesis, ATP synthesis, central metabolism, fatty acid synthesis and uncharacterized processes.
Analysis of the functional network associated with nitrogen fixation identified a number of hubs and bottlenecks. Of the five proteins that are both hubs and bottlenecks, the three proteins that have functional annotations are all involved in electron transfer or synthesis of electron transfer proteins. The two unknown proteins, in addition, appear to also be related to electron transfer processes via their interaction partners. We have also compared the global interaction network and individual subnetworks with those inferred from protein interaction data gathered in other bacteria, such as E. coli and the nitrogen-fixing soil bacterium Mesorhizobium loti. Of the comparable interactions studied between E. coli and R. palustris, 20% are shared between the two species. Our comparative approach both improves the functional annotation of interaction networks of non-model species, and reveals fundamental architectural principles of the biochemical networks of microbes.
An Imaging-Based Assay with High Sensitivity for Confirming and Characterizing Protein Interactions
Jennifer L. Morrell-Falvey* (morrelljl1@ornl.gov), A. Nicole Edwards, Jason D. Fowlkes, Robert F. Standaert, Dale A. Pelletier, Mitchel J. Doktycz, and Michelle V. Buchanan
Oak Ridge National Laboratory, Oak Ridge, Tenn.
Identifying and characterizing protein interactions are essential for understanding and modeling cellular networks. Several methods exist to assay protein interactions; however, none are known to provide both confirmation of protein interactions and simultaneous quantification of biophysical parameters (binding strengths and association/dissociation rates) in vivo. We are currently developing an approach that combines an imaging-based protein interaction assay with a fluorescence photobleaching and recovery technique (iFRAP), and computer simulations to provide a facile, general method for quantifying protein binding affinities in vivo. This protein interaction assay relies on the co-localization of two proteins of interest fused to DivIVA, a cell division protein from Bacillus subtilis, and green fluorescent protein (GFP). We have modified this imaging-based assay to facilitate high-throughput applications by constructing new vectors encoding N- and C-terminal DivIVA or GFP molecular tag fusions based on site-specific recombination technology and have determined the range of binding affinities that can be detected using this assay. The sensitivity of the assay was defined using a well-characterized protein interaction system involving the eukaryotic nuclear import receptor subunit, Importin α (Impα) and variant nuclear localization signals (NLS) representing a range of binding affinities. Using this system, we demonstrate that the modified co-localization assay is sensitive enough to detect protein interactions with Kd values that span over four orders of magnitude (1nM to 15µM). Moreover, the spatial confinement of the interacting proteins should also enable measurements of binding constants in vivo using iFRAP. Initial experiments demonstrate the anticipated decay of fluorescence at the cell poles indicative of a binding interaction. The statistical variance is reported as a function of Kd.
Functional Characterization of Protein Complexes and Cellular Systems in Rhodopseudomonas palustris using Stable Isotopic Labeling and Quantitative Proteomics
Chongle Pan1* (panc@ornl.gov), W. Judson Hervey IV,1,2 Michael S. Allen,1,3 Dale A. Pelletier,1 Gregory B. Hurst,1 and Michelle V. Buchanan1
1Oak Ridge National Laboratory, Oak Ridge, Tenn.; 2Graduate School of Genome Science and Technology, University of Tennessee-Oak Ridge National Laboratory, Oak Ridge, Tenn.; and 3University of North Texas, Denton, Tex.
Responses to various types of stresses can be identified in microbial cells at the protein level using quantitative comparative proteomics. Proteins can be metabolically labeled with stable isotopes, such as 15N and 13C, by growing a microorganism of interest in a medium enriched in those stable isotopes. Stable isotope labeling enables large-scale accurate protein quantification by mass spectrometry, referred to as quantitative proteomics. Here, quantitative proteomics was used to characterize protein complexes and cellular systems in Rhodopseudomonas palustris.
Stable isotope labeling can be used to evaluate performance of affinity-tagging strategies for studies of protein-protein interactions both at the level of the protein complex, and at the level of the proteome. Affinity-purified protein complexes are often accompanied by background, non-specific proteins. In this study, authentic interacting proteins of a model complex, DNA-dependent RNA polymerase (RNAP), were successfully distinguished from artifactual co-isolating proteins by the isotopic differentiation of interactions as random or targeted (I-DIRT) method (A. J. Tackett et al. J. Proteome Res. 2005, 4 (5), 1752-1756). To investigate broader effects of bait protein production on bacterial metabolism, we compared proteomes from strains harboring the plasmid that encodes an affinity-tagged subunit (RpoA) of the RNAP complex with the corresponding wild-type strains using stable isotope metabolic labeling. Expression of plasmid-encoded bait protein significantly induced the expression of several proteins involved in amino acid biosynthesis.
Cellular systems function not only via a physical interaction network but also within a regulatory network. In results from the Center for Molecular and Cellular Systems, observations of protein-protein interactions among a putative anti-σ factor RPA4224, an extracytoplasmic function (ECF) σ factor RPA4225, and the predicted response regulator RPA4223 led us to study this system further. We characterized a global stress regulon controlled by RPA4225 in R. palustris using quantitative proteomics. Changes in expression of several genes resulting from overproduction of RPA4225 were further verified by quantitative PCR. Furthermore, most of the strongly up-regulated proteins revealed a conserved binding motif, which we also found in the promoters of over 150 genes, including general stress proteins. These data suggest that RPA4225 controls a global stress regulon that may be conserved among several members of α-Proteobacteria.
These studies showcased the biological insights one can obtain using stable isotope labeling and quantitative proteomics. In addition, new methods based on stable isotope labeling are under development at the Center for Molecular and Cellular Systems for protein absolute quantification and complex dissociation kinetics.
Protein-Protein Interactions in Rhodopseudomonas palustris at the Genomics:GTL Center for Molecular and Cellular Systems
Dale A. Pelletier1* (pelletierda@ornl.gov), Kevin K. Anderson,2 William R. Cannon,2 Don S. Daly,2 Brian S. Hooker,2 H. Steven Wiley,2 Lee Ann McCue,2 Chongle Pan,1 Manesh B. Shah,1 W. Hayes McDonald,1 Keiji G. Asano,1 Gregory B. Hurst,1 Denise D. Schmoyer,1 Jenny L. Morrell‑Falvey,1 Mitchel J. Doktycz,1 Sheryl A. Martin,1 Mudita Singhal,2 Ronald C. Taylor,2 and Michelle V. Buchanan1
1Oak Ridge National Laboratory, Oak Ridge, Tenn.; and 2Pacific Northwest National Laboratory, Richland, Wash.
The overall goal of the Center for Molecular and Cellular Systems (CMCS) is to provide a capability for generating high quality protein-protein interaction data from a variety of energy- and environment-relevant microbial species. The CMCS approach combines expression of affinity tagged proteins, affinity purification of interacting proteins, and tandem mass spectrometric identification of these proteins (Pelletier et al., J. Proteome Research 2008, 7, 3319-3328).
We have recently completed the characterization of soluble protein-protein interactions in Rhodopseudomas palustris. These results are the first large-scale protein-protein interaction results of this type in an organism other than a model system such as E. coli or yeast. The protein-protein interactions from the metabolically versatile R. palustris provide insights into microbial processes of high relevance to DOE missions, including the ability to produce hydrogen, to degrade lignin monomers, to perform photosynthesis, and to fix nitrogen.
As of early December 2008, nearly 1200 R. palustris genes have been cloned as Gateway entry vectors, and approximately 1060 expression clones for a dual affinity tag (6-His/V5) have been produced. Over 800 affinity-tagged bait proteins have been expressed, affinity purified, and subjected to mass spectrometry (MS) analysis to identify interacting proteins. Criteria for choosing these bait proteins from among the >4800 in the R. palustris predicted proteome included predicted location in the cytosol, and previous detection at medium to high levels in proteomics measurements (VerBerkmoes et al., J. Proteome Research 2006, 5, 287-298). Quantitative estimates of confidence in putative bait-prey interactions identified from the MS analysis are obtained using the statistical tool BePro3 (Sharp et al., J. Proteome Research 2007, 6, 3788-3795). From thousands of putative interactions, a few hundred survive as candidates for further study, based on preliminarily chosen BePro3 threshold values. Integration with additional data, including comparative genomics, transcriptomics, operon structure, and regulatory networks, that aid in interpretation and functional annotation of novel interactions is described in the poster "An Integrative Strategy for the Determination of the Modular Structure of Functional Networks of Rhodopseudomonas palustris" by Cannon et al. Results of the protein-protein interaction survey in R. palustris are available through the publicly accessible Microbial Protein-Protein Interaction Database (MiPPI.ornl.gov).
*Presenting Author
