- Open Access
Visualizing post genomics data-sets on customized pathway maps by ProMeTra – aeration-dependent gene expression and metabolism of Corynebacterium glutamicum as an example
© Neuweger et al; licensee BioMed Central Ltd. 2009
Received: 5 December 2008
Accepted: 23 August 2009
Published: 23 August 2009
The rapid progress of post-genomic analyses, such as transcriptomics, proteomics, and metabolomics has resulted in the generation of large amounts of quantitative data covering and connecting the complete cascade from genotype to phenotype for individual organisms. Various benefits can be achieved when these "Omics" data are integrated, such as the identification of unknown gene functions or the elucidation of regulatory networks of whole organisms. In order to be able to obtain deeper insights in the generated datasets, it is of utmost importance to present the data to the researcher in an intuitive, integrated, and knowledge-based environment. Therefore, various visualization paradigms have been established during the last years. The visualization of "Omics" data using metabolic pathway maps is intuitive and has been applied in various software tools. It has become obvious that the application of web-based and user driven software tools has great potential and benefits from the use of open and standardized formats for the description of pathways.
In order to combine datasets from heterogeneous "Omics" sources, we present the web-based ProMeTra system that visualizes and combines datasets from transcriptomics, proteomics, and metabolomics on user defined metabolic pathway maps. Therefore, structured exchange of data with our "Omics" applications Emma 2, Qupe and MeltDB is employed. Enriched SVG images or animations are generated and can be obtained via the user friendly web interface.
To demonstrate the functionality of ProMeTra, we use quantitative data obtained during a fermentation experiment of the L-lysine producing strain Corynebacterium glutamicum DM1730. During fermentation, oxygen supply was switched off in order to perturb the system and observe its reaction. At six different time points, transcript abundances, intracellular metabolite pools, as well as extracellular glucose, lactate, and L-lysine levels were determined.
The interpretation and visualization of the results of this complex experiment was facilitated by the ProMeTra software. Both transcriptome and metabolome data were visualized on a metabolic pathway map. Visual inspection of the combined data confirmed existing knowledge but also delivered novel correlations that are of potential biotechnological importance.
To obtain a complete understanding of the function of cells, it is important to identify the roles of genes and their products. The analysis of gene transcripts (transcriptomics) and proteins (proteomics) is accelerated through the use of microarrays, ultra fast sequencing, and mass spectrometry. Additionally, cells contain numerous other organic molecules not directly encoded in the DNA, the metabolites, which are critical for cell function. Knowledge about metabolites is crucial for an understanding of most cellular phenomena [1–3]. All of these "Omics" technologies are also known as post-genomics.
The aim of scientific data visualization is to display properties of a data set that help researchers to identify quickly its most important characteristics. In functional genomics and post-genomic techniques there are recurring visualization strategies that are generally favored by researchers. For example in metabolomics, a frequent way in which molecular biologists like to visualize data is through the use of metabolic pathway maps. For this purpose a several packages and tools have been implemented which will be presented in the next section in more detail. We will highlight the important features and limitations of existing approaches which led us to the decision to implement the web-based ProMeTra tool which is the focus of our work.
Existing Systems and Pathway repositories
Initially, there were databases such as KEGG  or the different realizations of MetaCyc  that store information about the structure of metabolic networks. These databases represent static knowledge of metabolic pathways of organisms from all three domains of life. The contained data have been collected and curated over the years of genomic research and can be presented using images of metabolic pathways linking metabolites and enzymes.
Several tools have been developed to visualize and analyze biological networks together with data obtained from functional genomics measurements. Most interesting in this context are tools that visualize experimental data in the form of biochemical networks. The authors of the VANTED system for advanced data analysis and visualization in the context of biological networks  presented a comprehensive review of existing pathway visualization and mapping tools such as Cytoscape , MapMan , KaPPA-View , PathwayExplorer , and the Viewer included in MetaCyc-related databases  such as AraCyc . They pointed out that often only two conditions can be compared. In experiments designed to provide the basis for simulation in systems biology this is of limited use since often changes in metabolite concentration or transcript levels can only be understood if time series experiments are conducted and analyzed. It is also stressed by the authors that most tools are limited to transcriptomics datasets and only Omics Viewer , Cytoscape, and MapMan are designed to also display metabolite or other data. A severe limitation of some of the existing tools is their dependency on static maps, i.e. the data is mapped onto predefined pictures. This might be appropriate if the tools are being developed for a single organism or metabolic pathway but in general it clearly limits the re-usability of the approach. We will present some of the main tools and their important features in more detail.
Celldesigner and SBML
SBML, the Systems Biology Markup Language, facilitates the description of models and enables their exchange between various simulation and analysis tools. The XML-based SBML is a free and open format distinguished to represent biochemical reaction networks via a clear notation system . The CellDesigner Software is a process diagram editor for visualization and modeling of biochemical networks and gene-regulation. As SBML-compliant Java software it enables the integration of SBW (Systems Biology Workbench) simulation modules . CellDesigner uses a human-readable diagrammatic representation and proposes a set of notations that enforces the established SBML notation . A metabolic pathway created in CellDesigner is a state transition diagram with complex node structure that represents vertexes, state nodes (SN) and transition nodes (TN) and edges between SN and TN (ST-Edge) or rather TN and SN (TS-Edge). A process diagram (PDN) is defined as PND = (SN, TN, ST-Edge, TS-Edge). Each SN has a graphical symbol, for example a protein or gene symbol. Additionally the nature of a reaction, such as catalysis or inhibition, is represented by a symbol for each type of TN . Whereas the CellDesigner pathways consist of well structured and strictly typed entities, users may not include additional descriptive graphical elements.
KEGG Pathways and KEGG Markup Language
A major component of KEGG, the Kyoto Encyclopedia of Genes and Genomes, is the PATHWAY database which represents most of the known metabolic pathways . The database is continuously updated and consists of a collection of graphical diagrams, the so called pathway maps. In these maps, a box represents an enzyme and a circle a metabolic compound. The manually drawn and annotated pathway maps represent knowledge about the metabolism, genetic information processing, and cellular processes.
The KEGG Markup Language (KGML) is an XML-based exchange format and contains computerized information about graphical objects and their relations in the KEGG pathways. In KGML a pathway element is the root element that specifies one graph object. The nodes of the graph object are represented by the entry elements, whereas the relation and reaction elements specify the edges http://www.genome.jp/kegg/docs/xml/. An entry element contains information about a node of the pathway, like id, name and type. The relation element specifies a relationship between two proteins or protein and compound, which is indicated by an arrow. The reaction element describes the chemical reaction between substrates and a products http://www.genome.jp/kegg/docs/xml/. XML-files, which are defined by the KGML schema, can be downloaded from ftp://ftp.genome.jp/pub/kegg/xml/map.
KaPPA-View is a web-based tool and was developed to represent quantitative data for individual transcripts as well as metabolites on plant metabolic pathway maps. The aim of the system is to support the generation of hypotheses of gene function in the metabolic pathways through an intuitive visualization of the transcripts and metabolites. The system uses SVG vector graphic images for the representation of the biochemical pathways and the experimental datasets that are mapped on the pathway representations .
MapMan is a user-driven tool that displays large datasets (e.g. gene expression data from Arabidopsis thaliana Affymetrix arrays) onto diagrams of metabolic pathways or other processes. It has been developed specifically for data generated in Arabidopsis thaliana experiments measuring transcript or metabolite levels. The visualizations focus is on the display of experimental data in hierarchical and pre-defined pathway maps.
The functionality of KaPPA-View and MapMan can be accessed via web-applications. In general, a tendency to provide sophisticated analysis methods for functional genomics experiments and datasets through web-based frameworks can be observed. Recent examples are the Babelomics project  or the DAVID  database and analysis tools focussing on e.g. the functional profiling of genome scale experiments. The advantage of web-based analysis tools compared to stand-alone applications is the ease of updates and the possibility to rapidly release new features. Apart from recent web-browsers no additional software needs to be installed by the user.
To summarize, tools such as KaPPA-View or MapMan focus on a limited set of organisms and user defined pathway maps for other organisms or related strains are not supported. Additionally, the means to visualize metabolic pathway information is usually limited by the underlying pathway model as can be seen in the CellDesigner and KEGG pathways. Informative legends or additional user definable graphical elements that explain details are in general not supported.
Whereas most of the aforementioned tools allow to directly upload files with numerical results from "Omics" experiments in simple text-based files (CSV, TSV) or spreadsheets, the support to directly access "Omics" databases containing experimental results via Web Services is to our knowledge not well established. Data integration using Web Services is an elegant method to connect heterogeneous "Omics" frameworks and we will explore this approach for the following functional genomics experiment which combines transcriptomics and metabolomics measurements.
Corynebacterium glutamicum and lysine production
The Gram-positive soil bacterium Corynebacterium glutamicum is widely used for the production of industrially important amino acids. L-glutamate (1.5 million tons) and L-lysine (850,000 tons) are the major products and the amino acid market is growing at an annual rate of 7% [21, 22]. The high importance of L-lysine in animal nutrition led to extensive research and optimization of L-lysine production strains in the last decades. An important step towards optimized L-lysine production was the development of a strain having a feedback-deregulated aspartokinase by selecting for resistance against the L-lysine analogue, S-(2-aminoehtyl)-cysteine . Later, it was found that a single amino acid exchange led to the feedback deregulation . From that time on, rational strain improvement replaced the classical mutational approaches. For this strategy, it is necessary to understand not only single enzyme reactions, but to understand global metabolic regulation. The first step along this path was the sequencing of the whole genome of C. glutamicum ATCC 13032 . This allowed for the comparison of the wildtype sequence to gene sequences obtained from classically derived producer strains, the basis for the pioneering study of Ohnishi et al. . The authors showed that the introduction of only three genes from a producer strain obtained by chemical mutagenesis, each carrying a single mutation, into the wildtype strain led to a tremendous increase in L-lysine production. In fact, the production yield of the recombinant strain was better than that of the original producer, since the recombinant strain does only carry the beneficial mutations and grows faster than the original strain, therefore producing similar amounts of L-lysine in a shorter fermentation period.
The three mutated genes genes are the already mentioned lysC, pyc, and hom. The mutation in lysC results in the expression of a feedback-deregulated aspartokinase. Likewise, the mutated pyc encodes a pyruvate carboxylase with increased activity, resulting in an improved supply of oxaloacetate. Finally, the homoserine dehydrogenase derived from the mutated hom allel is less active, i.e. a leaky mutation, decreasing flux of the L-lysine precursor aspartate-β-semialdehyde into the threonine, isoleucine and methionine biosynthetic pathways.
The complete genome sequence was also essential to develop methods for genome-wide high-throughput analysis techniques like transcriptome analysis with DNA-microarrays [27, 28] and proteome analysis by two-dimensional gel electrophoresis coupled with peptide mass fingerprinting [29, 30]. The genome sequence also helped in deriving metabolic models that supported metabolomics with HPLC-MS or GC-MS [31, 32] and fluxomics, a combination of 13C-tracer experiments, isotopomer modeling, and metabolite balancing [33, 34]. It is regarded important for process and strain optimization to use data sets on the global physiological state of the cell during the production process not only using one, but several techniques.
The problem with this strategy is not only to analyze a process with all available techniques, but to interpret the data in relation to each other. The tool ProMeTra supports this process by a combined display of gene expression and metabolome data sets on a chosen metabolic or other pathway of biological relevance.
As an application example, we present the combined display of transcript abundance and metabolite pool data obtained from different time points of a batch-fermentation of the L-lysine production strain C. glutamicum DM1730. C. glutamicum DM1730 has the mutations pyc P458S, hom V59A, lysC T311I, and Δpck introduced into a wildtype genetic background . Although a cultivation in a fermenter reduces respectively abolishes many stresses like shifting pH and temperature by stringently controlling these parameters, there are stress parameters that can not be avoided. Among these is low oxygen stress that appears at high cell densities. Here, we introduced this stress on purpose by switching off oxygen supply. Analysis of the time course data with the help of ProMeTra gave new insights into the physiology of C. glutamicum under L-lysine fermentation and low oxygen stress conditions.
Researchers can access all the preprocessing and visualization functionality of ProMeTra via the web interface. Apart from a recent web browser that supports SVG images directly (e.g. Firefox or Safari) or via specialized plugins (e.g. the Adobe™ SVG viewer for Microsoft Internet Explorer™), no additional software needs to be installed.
Upload of own datasets is possible via CSV formatted files or Microsoft Excel spreadsheets. Details of the supported data formats and the organization of Excel files can be found in the online documentation. The uploaded data files in Excel or CSV format are only stored during a ProMeTra session and are automatically deleted afterwards to ensure the privacy of experimental data. In contrast to the temporarily stored data files, user defined pathway images enriched with information on the presented genes, transcripts, proteins or metabolites can be stored on the ProMeTra server persistently. Every user can decide if his pathways are made public and can also delete and update the uploaded pathway images via the ProMeTra web interface. Information on the pathway maps are stored in an object relational database on the server. User defined SVG pathway maps can be generated using the free Inkscape software available at http://www.inkscape.org, the online documentation and the user manual of ProMeTra contain further information on how to install the software and how to generate customized pathway maps.
The core of ProMeTra is an object oriented API that provides access to the pathway maps and the experimental data sets. The main classes are DataFactory, Element and Color. Subclasses of the interface DataFactory are responsible for retrieving experimental data from supported data sources. Based on the numerical range of the experimental data, a mapping of various color gradients (e.g. red-yellow-green) is computed by instances of the Color class. The functionality to enrich annotated SVG elements in pathway or genome maps is encapsulated in the Element class. It provides XML parser functionality to access and extend the DOM tree of any SVG image. The Element class inherits all methods of the XML::DOM::Element class and adds animation and coloring methods. Here, the Decorator design pattern was applied in order to attach additional responsibilities to SVG objects and sustain modularity.
Use of Web Services
We have already shown the successful use of web services to connect heterogeneous software frameworks in functional genomics  and also presented the advantages of a tight integration via the BRIDGE layer . The recently established MeltDB, Emma 2, and Qupe systems provide functional genomics datasets through the standardized and interoperable approach of web services. MeltDB and Emma 2 employ SOAP-based web service written in Perl which provide access to normalized quantitative data from metabolomics and transcriptomics experiments. Qupe offers Java-based and WSDL specified methods to obtain the pre-processed experimental datasets originating from quantitative proteomics experiments. ProMeTra is the first web-based system to make use of this functionality and integrates these datasets in one system. For researchers that do not have the possibility to analyze their data using the described web-based systems, we also provide the aforementioned CSV and Excel data import via the ProMeTra web interface.
Visualization and Animation features
ProMeTra supports SVG images that have been extended by annotations for genes, proteins or metabolites. The images in the open and user readable data format SVG can be uploaded to the web-server via the ProMeTra web interface. We already provide a set of customized pathway images for the industrial amino acid producer Corynebacterium glutamicum which is used in the following application example. Metabolic pathways can either be designed and submitted by the user or can be converted via ProMeTra functionality from SBML files defined in CellDesigner. We therefore developed a SBML-to-SVG converter that already includes the mapping of the elements to the KEGG compound database and includes annotated gene locus tags. The mapping of numerical experimental data such as concentrations and ratios is done through a color encoding and rectangles in the SVG image representing genes, proteins or metabolites are subdivided. Therefore the DOM tree of the SVG image is extended by ProMeTra. Child elements are added to the respective rectangles which preserves the original user defined layout.
The number of experimental factors that can be reasonably mapped on a Pathway Map element is limited by its size. For datasets with large numbers of experimental conditions or factors, ProMeTra supports the color animation feature of SVG images. Therefore the background color of an element changes over time which results in an animated SVG image. This feature of SVG images can be visualized in the Opera or Microsoft Internet Explorer web browsers.
ProMeTra offers different color gradients to encode the values of the submitted datasets. Further color gradients can easily be defined with the flexible ProMeTra API. If discrete values instead of M-Value ratios are submitted to the ProMeTra system, the color gradients are computed on the fly ranging from the maximal and minimal values found in the datasets.
It has been pointed out that the representation of "Omics" data on metabolic pathways is most intuitive to the researcher but we also address other concepts of visualization in ProMeTra. We have therefore developed functionality that transforms annotated bacterial genomes present at the NCBI genome repository into so called GenomeMaps. GenBank  files of the available replicons are parsed using BioPerl and SVG images (the GenomeMaps) are generated automatically. A GenomeMap represents each annotated coding sequence of a replicon as rectangle in a grid. The order of the rectangles is determined by the chromosomal position of the stop codon of the respective coding sequence and the rectangles are labeled by the associated locus tag or the gene name if present. The grid is filled row after row starting at the top left position for the first gene after the origin of replication. GenomeMaps have been generated for more than 400 bacterial genomes and are available through the ProMeTra web application. An example of a GenomeMap of the chromosome of C. glutamicum will be presented in the following application example.
Results and Discussion
Fermentation of the strain C. glutamicum DM1730 under different aeration conditions
Offline variables of the fermentation
Visualization of single analysis experiments – transcriptome analysis
Up regulation of the nar operon under oxygen-limiting conditions
In aerobic bacteria, oxygen is required as exogenous electron acceptor in respiration. The aerobic electron transfer chain in C. glutamicum is branched, one branch operates via menaquinone and the other via cytochrome . Under low-oxygen conditions, the anaerobic electron transfer is processed via nitrate respiration by nitrate reductase NarGHJI [44, 46]. The genes of the nar operon comprise (in the direction of transcription) a putative nitrate/nitrite transporter (NarK), a respiratory nitrate reductase enzyme (NarGHJI) , and a transcriptional regulator of the whole operon (ArnR). This regulator acts as transcriptional repressor of the nar operon under aerobic conditions . Here, the expression analysis of the nar operon revealed an increased transcription at t4 and t5 and a possible co-transcription with arnR. It was also proposed that arnR is co-transcribed with narKGHJI under anaerobic conditions , although the gene has its own promoter.
Down regulation of the atp operon after oxygen depletion
The eight gene atp operon of C. glutamicum encodes the subunits of the ATP synthetase that uses ATP to build up a proton gradient and can synthesize ATP by using this gradient . The atp operon is less transcribed under low oxygen conditions (t4 and t5), correlating with a lowered energy demand in the absence of growth. This phenomenon was also observed by Inui et al.  under oxygen deprivation conditions. Since ATPase hydrolyses ATP under non-respiratory conditions, cells save energy by reducing ATPase gene expression.
Combined metabolite pools and transcript abundances
Although we were aware of the fact that neither transcript levels precisely predict enzyme activities nor metabolite pools do this for fluxes, a high number of correlations could be identified that correspond with actual knowledge on bacterial metabolism.
Lactate consumption and production varies under different oxygen levels
One of the physiological consequence of oxygen depletion for C. glutamicum is demonstrated by production and secretion of L-lactate. Lactate production is mediated by the assimilatory lactate dehydrogenase encoded by the ldh gene . After aeration is switched off, transcription levels of the ldh gene were increasing, coherent with the increasing amount of external L-lactate (Figure 3) . The increasing lactate pool sizes were correlated with increasing internal pool sizes of succinate (t3, t4 and t5). Inui et al. postulated that under oxygen deprivation conditions the oxidative arm of the tricarboxylic acid cycle (TCA) is downregulated (gltA, sucB, and sucCD) and mdh (malate dehydrogenase) is upregulated, resulting in a high succinate pool. Downregulation of gltA (citrate synthase) leads to an accumulation of pyruvate, which is converted to lactate via lactate dehydrogenase (ldh) . Both enzymes, Mdh and Ldh, may regenerate NAD+ to compensate both, downregulation of the oxidative arm of the TCA and the loss of energy-regenerating respiration.
After aeration is switched on again, transcript levels of the ldh gene decreased and at the same time (t5 and t6), those of the gene encoding the dissimilatory lactate dehydrogenase lldA  were increasing. This is consistent with the consumption of external lactate that is metabolized by LldA at the end of the fermentation (Figure 3). Lactate is co-utilized with glucose, a capability of C. glutamicum already reported . The observed correlation between lactate utilization and upregulation of the glyoxylate pathway remains unclear. In C. glutamicum, carbon sources which enter the metabolism downstream of pyruvate use the glyoxylate pathway as anaplerotic reaction . This is true for single substrate utilization as well as for co-utilization. A reason for the correlation might be that the TCA has to be refilled after downregulation of the oxidative arm of the TCA.
Effects on carbon metabolism
In carbon metabolism, drastic differences relative to t1 appeared when all carbon sources had been consumed (t6). Again, several correlations between transcript level and metabolite pool ratios were observed. After glucose was consumed (t6), the transcript level of the gene encoding the glucose-specific enzyme II of the phosphoenol pyruvate (PEP) phosphotransferase system ptsG was lower due to the fact that its transcriptional repressor SugR [54, 55] was upregulated (data not shown). Under these conditions cells normally operate gluconeogenesis, which is not possible in DM1730, because the gene for the gluconeogenetic enzyme pyruvate carboxykinase (Pck) was deleted. Under glucose depletion, the gene encoding pyruvate kinase (pyk) was found downregulated (t6), possibly to avoid efflux of PEP into the TCA. Probably, due to the fact that at the same time the genes of the two PEP-consuming enzymes PtsG and Pyk displayed lower transcript levels, PEP itself accumulated.
Observed variations in production rate of L-lysine
It was apparent that the L-lysine precursor D, L-diaminopimelate accumulated in the cell and the internal L-lysine pool rose after aeration was switched off for a longer time. Diaminopimelate is not only the precursor for L-lysine but also for peptidoglycan biosynthesis, which is used in cell wall production. Since biomass formation and apparently cell wall synthesis almost stopped during low-oxygen conditions, diaminopimelate accumulated (t3, t4 and t5) due to decreased consumption. These higher internal pools stayed almost constant, even after aeration is switched on again and were reflected by higher extracellular L-lysine concentrations (Figure 2). The transcript levels of lysA (diaminopimelate decarboxylase) and lysE (L-lysine exporter) varied over the fermentation process in an inverse manner. The L-lysine exporter LysE is transcriptionally regulated by LysG . L-lysine is the positive effector of LysG, acting as a sensor for internal L-lysine concentrations. Here, lysE was found upregulated at t5, correlating with a high internal L-lysine pool. It was interesting to note that the lysA gene encoding diaminopimelate carboxylase, the final step in lysine synthesis, showed an inverse expression behavior. The reason for this is unclear since no transcriptional regulation of the argS-lysA operon  is known to date.
We have created the web-based ProMeTra application that is able to visualize and combine data from complex functional genomics experiments. The user-friendly web interface allows researchers to easily visualize and integrate their datasets on pathway maps using the established SVG graphics format. Additional information and graphical content can be added to the vector-based images using drawing software (i.e. CorelDraw™, Inkscape). Unlike other tools such as Omics-Viewer  or MapMan , ProMeTra is designed to support user designed and annotated pathway maps. This is useful as metabolic pathways which are e.g. available in the KEGG database do not exactly represent the genetic content of the organism under study or genetically modified organisms are analyzed. Furthermore, experimental data stored in functional genomics applications such as MeltDB, Emma 2 or Qupe can be accessed directly via web services. Besides, ProMeTra supports simple CSV and Microsoft Excel files as data input formats. In contrast to commercially available packages such as MetaCore™ http://www.genego.com/metacore.php by GeneGo and Ingenuity Pathways™ http://www.ingenuity.com/products/pathways_analysis.html which also provide sophisticated metabolic pathway maps and functionality to map experimental data, ProMeTra uses open standards such as SBML, SVG and Web Services, provides free access to the functionality and offers several public pathway maps. Similar to ProMeTra, the commercially available systems contain a set of predefined pathway maps and allow users to generate their own pathway maps. They also allow to visualize quantitative experimental data from e.g. metabolomics or transcriptomics measurements. Nonetheless, the MetaCore and Ingenuity Pathways system can only visualize results of multiple experiments, time points and dosages through animated graphics according to the systems manuals. This is a limitation since animated pathway visualizations can not be used in publications or on posters. A comprehensive overview of e.g. the progress of a fermentation experiment as presented in Figure 5 can therefore not be generated using the two commercial systems.
The API of the ProMeTra system has an object oriented, modular design. The popular Design Patterns have been used in the creation of the application and allow to easily extend the system to include new data sources and visualization methods. With the use of the MVC approach, we have created a versatile and extendable web interface.
Through the conversion of CellDesigners SBML pathways, we provide means to use the pathway mapping functionality with existing pathway maps and we plan to extend this by conversion of the pathway maps from the KEGG database to our enriched SVG format. In summary, ProMeTra offers a flexible and extendible approach for the analysis, visualization, and integration of functional genomics datasets. Since ProMeTra can access complex experiments processed in MeltDB, Emma 2, and Qupe, we have a system that allows the experienced researcher to focus on the interpretation of the experimental results in an intuitive and visual way rather than on the time consuming conversion of vast tabular data into hopefully meaningful results.
The visualization of transcript abundances from the application example on ProMeTra GenomeMaps confirmed known transcription units and provides an intuitive genome wide overview on transcriptomics datasets. The function of ProMeTra to visualize transcript abundances and metabolite pool deviations onto user-defined pathway maps verified that coherence between transcript level and metabolite pool sizes exist. The visualizations of the application example confirmed existing knowledge and spurred new insights in gene regulation and the corresponding phenotype of cells. In this application example, we used data from two functional genomics techniques, namely transcriptomics and metabolomics. ProMeTra is by design not limited to these data-sources. The pathway map that we employed for this study contained identifiers for the compounds and the transcripts. An extended version of this pathway map that also contains identifiers for the proteins of C. glutamicum could easily be generated if additional experimental data becomes available.
Availability and requirements
ProMeTra is publicly available at http://prometra.cebitec.uni-bielefeld.de. The project info page at http://www.cebitec.uni-bielefeld.de/groups/brf/software/prometra_info/index.html provides further information. We have set up a wiki page together with a user manual in PDF format which details the work flow of a typical ProMeTra analysis, both is available at http://www.cebitec.uni-bielefeld.de/groups/brf/software/wiki/ProMeTraWiki. Researchers can access the system using the public prometra account without further registration. To review the generated SVG images, a recent web browser which supports the SVG image format is needed. For Microsoft Internet Explorer, a SVG plugin is required wich is freely available from Adobe.
HN would like to thank the International Graduate School in Bioinformatics and Genome Research for providing financial support. MP, TB and AH acknowledge the SysMap project financed by the BMBF and Evonik-Degussa GmbH (grant 0313704). MD acknowledges financial support by the BMBF (grant 0313805A 'GenoMik-Plus'), SA received financial support from the BMBF in the frame of the QuantPro initiative (grant 0313812). The authors further wish to thank the BRF team for expert technical support.
- Weckwerth W: Metabolomics in systems biology. Annu Rev Plant Biol. 2003, 54: 669-689. 10.1146/annurev.arplant.54.031902.135014View ArticlePubMedGoogle Scholar
- Fernie AR, Trethewey RN, Krotzky AJ, Willmitzer L: Metabolite profiling: from diagnostics to systems biology. Nat Rev Mol Cell Biol. 2004, 5 (9): 763-769. 10.1038/nrm1451View ArticlePubMedGoogle Scholar
- Kell DB: Metabolomics and systems biology: making sense of the soup. Curr Opin Microbiol. 2004, 7 (3): 296-307. 10.1016/j.mib.2004.04.012View ArticlePubMedGoogle Scholar
- Verhoeckx KCM, Bijlsma S, de Groene EM, Witkamp RF, Greef van der J, Rodenburg RJT: A combination of proteomics, principal component analysis and transcriptomics is a powerful tool for the identification of biomarkers for macrophage maturation in the U937 cell line. Proteomics. 2004, 4 (4): 1014-1028. 10.1002/pmic.200300669View ArticlePubMedGoogle Scholar
- Broeckling CD, Huhman DV, Farag MA, Smith JT, May GD, Mendes P, Dixon RA, Sumner LW: Metabolic profiling of Medicago truncatula cell cultures reveals the effects of biotic and abiotic elicitors on metabolism. J Exp Bot. 2005, 56 (410): 323-336. 10.1093/jxb/eri058View ArticlePubMedGoogle Scholar
- Ogata H, Goto S, Sato K, Fujibuchi W, Bono H, Kanehisa M: KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 1999, 27: 29-34. 10.1093/nar/27.1.29PubMed CentralView ArticlePubMedGoogle Scholar
- Caspi R, Foerster H, Fulcher CA, Kaipa P, Krummenacker M, Latendresse M, Paley S, Rhee SY, Shearer AG, Tissier C, Walk TC, Zhang P, Karp PD: The MetaCyc Database of metabolic pathways and enzymes and the BioCyc collection of Pathway/Genome Databases. Nucleic Acids Res. 2008, D623-D631. 36 DatabaseGoogle Scholar
- Junker BH, Klukas C, Schreiber F: VANTED: a system for advanced data analysis and visualization in the context of biological networks. BMC Bioinformatics. 2006, 7: 109- 10.1186/1471-2105-7-109PubMed CentralView ArticlePubMedGoogle Scholar
- Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003, 13 (11): 2498-2504. 10.1101/gr.1239303PubMed CentralView ArticlePubMedGoogle Scholar
- Thimm O, Bläsing O, Gibon Y, Nagel A, Meyer S, Krüger P, Selbig J, Müller LA, Rhee SY, Stitt M: MAPMAN: a user-driven tool to display genomics data sets onto diagrams of metabolic pathways and other biological processes. Plant J. 2004, 37 (6): 914-939. 10.1111/j.1365-313X.2004.02016.xView ArticlePubMedGoogle Scholar
- Tokimatsu T, Sakurai N, Suzuki H, Ohta H, Nishitani K, Koyama T, Umezawa T, Misawa N, Saito K, Shibata D: KaPPA-view: a web-based analysis tool for integration of transcript and metabolite data on plant metabolic pathway maps. Plant Physiol. 2005, 138 (3): 1289-1300. 10.1104/pp.105.060525PubMed CentralView ArticlePubMedGoogle Scholar
- Mlecnik B, Scheideler M, Hackl H, Hartler J, Sanchez-Cabo F, Trajanoski Z: PathwayExplorer: web service for visualizing high-throughput expression data on biological pathways. Nucleic Acids Res. 2005, W633-W637. 33 Web ServerGoogle Scholar
- Karp PD, Riley M, Paley SM, Pellegrini-Toole A: The MetaCyc Database. Nucleic Acids Res. 2002, 30: 59-61. 10.1093/nar/30.1.59PubMed CentralView ArticlePubMedGoogle Scholar
- Mueller LA, Zhang P, Rhee SY: AraCyc: a biochemical pathway database for Arabidopsis. Plant Physiol. 2003, 132 (2): 453-460. 10.1104/pp.102.017236PubMed CentralView ArticlePubMedGoogle Scholar
- Paley SM, Karp PD: The Pathway Tools cellular overview diagram and Omics Viewer. Nucleic Acids Res. 2006, 34 (13): 3771-3778. 10.1093/nar/gkl334PubMed CentralView ArticlePubMedGoogle Scholar
- Hucka M, Finney A, Sauro HM, Bolouri H, Doyle JC, Kitano H, Arkin AP, Bornstein BJ, Bray D, Cornish-Bowden A, Cuellar AA, Dronov S, Gilles ED, Ginkel M, Gor V, Goryanin II, Hedley WJ, Hodgman TC, Hofmeyr JH, Hunter PJ, Juty NS, Kasberger JL, Kremling A, Kummer U, Novère NL, Loew LM, Lucio D, Mendes P, Minch E, Mjolsness ED, Nakayama Y, Nelson MR, Nielsen PF, Sakurada T, Schaff JC, Shapiro BE, Shimizu TS, Spence HD, Stelling J, Takahashi K, Tomita M, Wagner J, Wang J, : The systems biology markup language (SBML): a medium for representation and exchange of biochemical network models. Bioinformatics. 2003, 19 (4): 524-531. 10.1093/bioinformatics/btg015View ArticlePubMedGoogle Scholar
- Funahashi A, Morohashi M, Kitano H, Tanimura N: CellDesigner: a process diagram editor for gene-regulatory and biochemical networks. BIOSILICO. 2003, 1 (5): 159-162. 10.1016/S1478-5382(03)02370-9.View ArticleGoogle Scholar
- Kitano H, Funahashi A, Matsuoka Y, Oda K: Using process diagrams for the graphical representation of biological networks. Nat Biotechnol. 2005, 23 (8): 961-966. 10.1038/nbt1111View ArticlePubMedGoogle Scholar
- Al-Shahrour F, Carbonell J, Minguez P, Goetz S, Conesa A, Trraga J, Medina I, Alloza E, Montaner D, Dopazo J: Babelomics: advanced functional profiling of transcriptomics, proteomics and genomics experiments. Nucleic Acids Res. 2008, W341-W346. 36 Web ServerGoogle Scholar
- Huang DW, Sherman BT, Lempicki RA: Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009, 4: 44-57. 10.1038/nprot.2008.211View ArticleGoogle Scholar
- Hermann T: Industrial production of amino acids by coryneform bacteria. J Biotechnol. 2003, 104 (1–3): 155-172. 10.1016/S0168-1656(03)00149-4View ArticlePubMedGoogle Scholar
- Leuchtenberger W, Huthmacher K, Drauz K: Biotechnological production of amino acids and derivatives: current status and prospects. Appl Microbiol Biotechnol. 2005, 69: 1-8. 10.1007/s00253-005-0155-yView ArticlePubMedGoogle Scholar
- Nakayama K, Araki K: Process for producing L-lysine. US Patent. 3708395. 1973Google Scholar
- Kalinowski J, Cremer J, Bachmann B, Eggeling L, Sahm H, Pühler A: Genetic and biochemical analysis of the aspartokinase from Corynebacterium glutamicum. Mol Microbiol. 1991, 5 (5): 1197-1204. 10.1111/j.1365-2958.1991.tb01893.xView ArticlePubMedGoogle Scholar
- Kalinowski J, Bathe B, Bartels D, Bischoff N, Bott M, Burkovski A, Dusch N, Eggeling L, Eikmanns BJ, Gaigalat L, Goesmann A, Hartmann M, Huthmacher K, Krämer R, Linke B, McHardy AC, Meyer F, Möckel B, Pfefferle W, Pühler A, Rey DA, Rückert C, Rupp O, Sahm H, Wendisch VF, Wiegräbe I, Tauch A: The complete Corynebacterium glutamicum ATCC 13032 genome sequence and its impact on the production of L-aspartate-derived amino acids and vitamins. J Biotechnol. 2003, 104 (1–3): 5-25. 10.1016/S0168-1656(03)00154-8View ArticlePubMedGoogle Scholar
- Ohnishi J, Mitsuhashi S, Hayashi M, Ando S, Yokoi H, Ochiai K, Ikeda M: A novel methodology employing Corynebacterium glutamicum genome information to generate a new L-lysine-producing mutant. Appl Microbiol Biotechnol. 2002, 58 (2): 217-223. 10.1007/s00253-001-0883-6View ArticlePubMedGoogle Scholar
- Hüser AT, Becker A, Brune I, Dondrup M, Kalinowski J, Plassmeier J, Pühler A, Wiegräbe I, Tauch A: Development of a Corynebacterium glutamicum DNA microarray and validation by genome-wide expression profiling during growth with propionate as carbon source. J Biotechnol. 2003, 106 (2–3): 269-286. 10.1016/j.jbiotec.2003.08.006View ArticlePubMedGoogle Scholar
- Wendisch VF, Bott M, Kalinowski J, Oldiges M, Wiechert W: Emerging Corynebacterium glutamicum systems biology. J Biotechnol. 2006, 124: 74-92. 10.1016/j.jbiotec.2005.12.002View ArticlePubMedGoogle Scholar
- Burkovski A: Proteomics of Corynebacterium glutamicum: essential industrial bacterium. Methods Biochem Anal. 2006, 49: 137-147.PubMedGoogle Scholar
- Hansmeier N, Chao TC, Pühler A, Tauch A, Kalinowski J: The cytosolic, cell surface and extracellular proteomes of the biotechnologically important soil bacterium Corynebacterium efficiens YS-314 in comparison to those of Corynebacterium glutamicum ATCC 13032. Proteomics. 2006, 6: 233-250. 10.1002/pmic.200500144View ArticlePubMedGoogle Scholar
- Krömer JO, Fritz M, Heinzle E, Wittmann C: In vivo quantification of intracellular amino acids and intermediates of the methionine pathway in Corynebacterium glutamicum. Anal Biochem. 2005, 340: 171-173. 10.1016/j.ab.2005.01.027View ArticlePubMedGoogle Scholar
- Plassmeier J, Barsch A, Persicke M, Niehaus K, Kalinowski J: Investigation of central carbon metabolism and the 2-methylcitrate cycle in Corynebacterium glutamicum by metabolic profiling using gas chromatography-mass spectrometry. J Biotechnol. 2007, 130 (4): 354-363. 10.1016/j.jbiotec.2007.04.026View ArticlePubMedGoogle Scholar
- Drysch A, Massaoudi ME, Mack C, Takors R, de Graaf AA, Sahm H: Production process monitoring by serial mapping of microbial carbon flux distributions using a novel Sensor Reactor approach: II-(13)C-labeling-based metabolic flux analysis and L-lysine production. Metab Eng. 2003, 5 (2): 96-107. 10.1016/S1096-7176(03)00005-3View ArticlePubMedGoogle Scholar
- Yang TH, Wittmann C, Heinzle E: Respirometric 13C flux analysis, Part I: design, construction and validation of a novel multiple reactor system using on-line membrane inlet mass spectrometry. Metab Eng. 2006, 8 (5): 417-431. 10.1016/j.ymben.2006.03.001View ArticlePubMedGoogle Scholar
- Seibold G, Auchter M, Berens S, Kalinowski J, Eikmanns BJ: Utilization of soluble starch by a recombinant Corynebacterium glutamicum strain: growth and lysine production. J Biotechnol. 2006, 124 (2): 381-391. 10.1016/j.jbiotec.2005.12.027View ArticlePubMedGoogle Scholar
- Gamma E, Helm R, Johnson R, Vlissides J: Design Patterns. Elements of Reusable Object-Oriented Software. 1995, Addison WesleyGoogle Scholar
- Neuweger H, Albaum SP, Dondrup M, MarcusPersicke , Watt T, Niehaus K, Stoye J, Goesmann A: MeltDB -A software platform for the analysis and integration of metabolomics experiment data. Bioinformatics. 2008, 24 (23): 2726-2732. 10.1093/bioinformatics/btn452View ArticlePubMedGoogle Scholar
- Dondrup M, Albaum SP, Griebel T, Henckel K, Jünemann S, Kahlke T, Kleindt CK, Küster H, Linke B, Mertens D, Mittard-Runte V, Neuweger H, Runte KJ, Tauch A, Tille F, Pühler A, Goesmann A: EMMA 2-a MAGE-compliant system for the collaborative analysis and integration of microarray data. BMC Bioinformatics. 2009, 10: 50- 10.1186/1471-2105-10-50PubMed CentralView ArticlePubMedGoogle Scholar
- Neuweger H, Baumbach J, Albaum S, Bekel T, Dondrup M, Hüser AT, Kalinowski J, Oehm S, Pühler A, Rahmann S, Weile J, Goesmann A: CoryneCenter – an online resource for the integrated analysis of corynebacterial genome and transcriptome data. BMC Syst Biol. 2007, 1: 1752-0509. 10.1186/1752-0509-1-55.View ArticleGoogle Scholar
- Goesmann A, Linke B, Bartels D, Dondrup M, Krause L, Neuweger H, Oehm S, Paczian T, Wilke A, Meyer F: BRIGEP-the BRIDGE-based genome-transcriptome-proteome browser. Nucleic Acids Res. 2005, W710-W716. 33 Web ServerGoogle Scholar
- Benson DA, Karsch-Mizrachi I, Lipman DJ, Ostell J, Wheeler DL: GenBank. Nucleic Acids Res. 2008, D25-D30. 36 DatabaseGoogle Scholar
- Keilhauer C, Eggeling L, Sahm H: Isoleucine synthesis in Corynebacterium glutamicum: molecular analysis of the ilvB-ilvN-ilvC operon. J Bacteriol. 1993, 175 (17): 5595-5603.PubMed CentralPubMedGoogle Scholar
- Inui M, Murakami S, Okino S, Kawaguchi H, Vertès AA, Yukawa H: Metabolic analysis of Corynebacterium glutamicum during lactate and succinate productions under oxygen deprivation conditions. J Mol Microbiol Biotechnol. 2004, 7 (4): 182-196. 10.1159/000079827View ArticlePubMedGoogle Scholar
- Takeno S, Ohnishi J, Komatsu T, Masaki T, Sen K, Ikeda M: Anaerobic growth and potential for amino acid production by nitrate respiration in Corynebacterium glutamicum. Appl Microbiol Biotechnol. 2007, 75 (5): 1173-1182. 10.1007/s00253-007-0926-8View ArticlePubMedGoogle Scholar
- Bott M, Niebisch A: The respiratory chain of Corynebacterium glutamicum. J Biotechnol. 2003, 104 (1–3): 129-153. 10.1016/S0168-1656(03)00144-5View ArticlePubMedGoogle Scholar
- Nishimura T, Vertès AA, Shinoda Y, Inui M, Yukawa H: Anaerobic growth of Corynebacterium glutamicum using nitrate as a terminal electron acceptor. Appl Microbiol Biotechnol. 2007, 75 (4): 889-897. 10.1007/s00253-007-0879-yView ArticlePubMedGoogle Scholar
- Nishimura T, Teramoto H, Vertès AA, Inui M, Yukawa H: ArnR, a novel transcriptional regulator, represses expression of the narKGHJI operon in Corynebacterium glutamicum. J Bacteriol. 2008, 190 (9): 3264-3273. 10.1128/JB.01801-07PubMed CentralView ArticlePubMedGoogle Scholar
- Barriuso-Iglesias M, Barreiro C, Flechoso F, Martín JF: Transcriptional analysis of the F0F1 ATPase operon of Corynebacterium glutamicum ATCC 13032 reveals strong induction by alkaline pH. Microbiology. 2006, 152 (Pt 1): 11-21. 10.1099/mic.0.28383-0View ArticlePubMedGoogle Scholar
- Inui M, Suda M, Okino S, Nonaka H, Puskás LG, Vertès AA, Yukawa H: Transcriptional profiling of Corynebacterium glutamicum metabolism during organic acid production under oxygen deprivation conditions. Microbiology. 2007, 153 (Pt 8): 2491-2504. 10.1099/mic.0.2006/005587-0View ArticlePubMedGoogle Scholar
- Okino S, Suda M, Fujikura K, Inui M, Yukawa H: Production of D-lactic acid by Corynebacterium glutamicum under oxygen deprivation. Appl Microbiol Biotechnol. 2008, 78 (3): 449-454. 10.1007/s00253-007-1336-7View ArticlePubMedGoogle Scholar
- Georgi T, Engels V, Wendisch VF: Regulation of L-lactate utilization by the FadR-type regulator LldR of Corynebacterium glutamicum. J Bacteriol. 2008, 190 (3): 963-971. 10.1128/JB.01147-07PubMed CentralView ArticlePubMedGoogle Scholar
- Stansen C, Uy D, Delaunay S, Eggeling L, Goergen JL, Wendisch VF: Characterization of a Corynebacterium glutamicum lactate utilization operon induced during temperature-triggered glutamate production. Appl Environ Microbiol. 2005, 71 (10): 5920-5928. 10.1128/AEM.71.10.5920-5928.2005PubMed CentralView ArticlePubMedGoogle Scholar
- Wendisch VF, de Graaf AA, Sahm H, Eikmanns BJ: Quantitative determination of metabolic fluxes during coutilization of two carbon sources: comparative analyses with Corynebacterium glutamicum during growth on acetate and/or glucose. J Bacteriol. 2000, 182 (11): 3088-3096. 10.1128/JB.182.11.3088-3096.2000PubMed CentralView ArticlePubMedGoogle Scholar
- Engels V, Wendisch VF: The DeoR-type regulator SugR represses expression of ptsG in Corynebacterium glutamicum. J Bacteriol. 2007, 189 (8): 2955-2966. 10.1128/JB.01596-06PubMed CentralView ArticlePubMedGoogle Scholar
- Gaigalat L, Schlüter JP, Hartmann M, Mormann S, Tauch A, Pühler A, Kalinowski J: The DeoR-type transcriptional regulator SugR acts as a repressor for genes encoding the phosphoenolpyruvate:sugar phosphotransferase system (PTS) in Corynebacterium glutamicum. BMC Mol Biol. 2007, 8: 104- 10.1186/1471-2199-8-104PubMed CentralView ArticlePubMedGoogle Scholar
- Bellmann A, Vrljić M, Pátek M, Sahm H, Krämer R, Eggeling L: Expression control and specificity of the basic amino acid exporter LysE of Corynebacterium glutamicum. Microbiology. 2001, 147 (Pt 7): 1765-1774.View ArticlePubMedGoogle Scholar
- Oguiza JA, Malumbres M, Eriani G, Pisabarro A, Mateos LM, Martin F, Martín JF: A gene encoding arginyl-tRNA synthetase is located in the upstream region of the lysA gene in Brevibacterium lactofermentum: regulation of argS-lysA cluster expression by arginine. J Bacteriol. 1993, 175 (22): 7356-7362.PubMed CentralPubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.