How will systems biology Impact the general public?
Model organisms: why yeast?
State-of-the-art in yeast systems biology
Need for joining forces: the birth of YSBN
A European Initiative: the YSBN Coordination Action
System biology aims at a quantitative understanding of biological systems by analysing the relationships among their many components and by using mathematical models to describe their interactions and further predict their behaviour. This definition is not sufficiently complete, as describing the mission of a field does not characterize its tools, methodology and approach. What is commonly referred to as the ‘systems approach’ to problem-solving uses both computational and experimental tools and is composed of a number of well-defined steps: (a) creating a model of the system based on the information available through the definition of its components (high-throughput data generation, databases, literature and other sources of information); (b) testing the model by using it for prediction of, for example, a given perturbation of the system and evaluating this by experiment; and (c) making use of the resulting information to improve the model until realistic predictions are achieved. The ‘desk–bench–desk loop’ is concluded when a model makes predictions that match the biological reality of the analysed system, or the model has been used to gain new insight into the biological system that would otherwise have been difficult to obtain due to the complexity of the system.
Analogies have been used to illustrate the interconnectivity of the elements characterizing a network (or system): Leroy Hood compares a car and its components to a system32, Bernhard Palsson often uses the analogy of urban traffic and the creation of traffic jams to describe crucial modules in metabolic pathways15, and Yuri Lazebnik compared the functioning of an old transistor radio to a signal transduction pathway. Lazebnik critically reached the conclusions that a biologist who does not use a systems approach to research (integrating biological information with computational modelling) is as lost in science as he/she would be in attempting to repair the old radio49, 73. Along another line, Zoltán Oltvai represents the complexity of biological systems as a scale-free hierarchical pyramid, and thereby shows that the underlying principles can be related to non-biological networks58.
With the emergence of systems biology, there is a paradigm shift in biological research from a reductionist to a systems approach. However, the first attempts to look at cells as ‘systems’ can be traced back to at least 1968, when Ludwig von Bertalanffy conceptualized the first views of organisms as physical systems68. Similarly, in 1978, in their textbook Growth of the Bacterial Cell, Maaløe and co-workers quantified the metabolism of an Escherichia coli cell and they proposed concepts for calculating the ATP expenditure in the various metabolic pathways36. Furthermore, in 1998, a textbook definition of yeast physiology coincided with what we presently refer to as yeast systems biology69. If yeast physiology is defined as aiming at understanding the growth and metabolism of yeast cells through the analysis of its essential life cycle mechanisms (feeding, metabolizing, reproducing and dying), and biotechnology is defined as the science exploiting yeast physiology for industrial application, then systems biology can be seen as novel approach to quantitative physiology55. Then, is ‘systems biology’ a modern term redefining old concepts? It has been stated that systems biology is the merger between the development of molecular biology and the historically parallel development of quantitative biochemistry (e.g. non-equilibrium thermodynamics)71. According to this definition, systems biology combines engineering and molecular biology to design novel regulatory modules, in order to gain insights into biological processes or to improve the production of certain metabolites4, 55. The introduction of physics, microbial physiology and molecular biology techniques marks an important shift, as this has provided a very powerful toolbox that allowed the study of molecular events at both the molecular and at the systems level, and so formed the basis for the development of systems biology as we know it today. Other key advances driving systems biology include genome sequencing (human, yeast, worm), the advent of the Internet, high-throughput experimental platforms and the developments within bioinformatics. The introduction of whole-genome sequencing and the subsequent analysis of the ‘omes’ have allowed the study of interactions between many different cellular pathways at the same time. Most probably these techniques were the very toolbox that Maaløe and Von Bertalanffy were lacking 30 years ago for consolidating their theories through the study of specific molecular processes in living cells. Why did the ‘systems approach’ not develop in parallel to the development of molecular biology? Perhaps it is because, with the advent of the ‘molecular era’, the scientific community focused into the analysis of specific components and partly disregarded the connections among the pieces of the puzzle. It was only a few ‘scientific generations’ later that the key to further progress was ‘rediscovered’ in the holistic view of the system and the move from molecular to modular biology was formalized29. The parallel progress of fundamental research and systems biology is essential for addressing quantitative descriptions of how the different pathways interact with one another. It is absolutely essential to score the relative importance of the many different connections within the cell, which remains a significant challenge to the scientific community.
The systems approach is already applied to many fields of biological sciences, from the description of the functioning of the heart56 to the characterization of yeast's glucose metabolism59, 42. There will be a trend towards a systems view of cellular functions based on two parallel routes: the understanding of the molecular mechanisms of individual processes and vice versa; the understanding of the whole system as a way to understand its components14. Even though systems biology is so widely applied, we will focus here on the community of scientists studying yeast and the network that was created in order to synergize and coordinate their efforts.
How will systems biology impact the general public?
By providing a holistic view of a given biological system, systems biology has already proved essential to applications like drug discovery, novel therapeutic methods and the decrease of market prices for many commercial products through the improvement of many industrial biotechnological processes59. Thus, yeast systems biology (and systems biology of other organisms) impacts on society in four ways (see Figure 1).
Drug discovery will be substantially facilitated through the use of complete mathematical models of living cells, particularly as model simulations may lead to identification of optimal drug targets. The identification of very specific targets for treating various diseases enables the design of efficient drugs causing fewer side-effects. The possibility of studying an organism at the systems level allows the recognition and avoidance of side-effects through the design of drugs that have the appropriate pharmaceutical properties (adsorption and pharmacokinetics) and the application regimes that lead to most specific effects33. The earlier the new chemical entity is validated in terms of its commercial value and potential risks, the more capital is saved in all subsequent steps of drug commercialization (target design, lead validation, pharmaceutical trials and risk assessment). This can potentially increase the output in terms of new chemical entities and optimize investment in pharmaceutical R&D. Even though our ability to characterize a human cell (including its complete network) is still in its infancy, systems biology-based approaches are already established for the analysis of selective networks and pathways to embody disease-relevant responses. Thus, Bioseek is a young drug discovery company, focused on inflammation diseases, that has developed a technological platform called BioMAP (biologically multiplexed activity profiling) for the identification and the optimization of novel pharmaceutical targets7, 6 (www.bioseekinc.com). Similarly, big pharmaceutical companies (Lilly Systems Biology—spun out from Eli Lilly, Johnson & Johnson, Novartis and NovoNordisk) have also introduced a systems biology approach to drug discovery51. In addition, there are also an increasing number of systems biology service companies that, whilst focusing on a specific area, consult for the big-pharma companies in optimizing their drug discovery process51. In addition to the above-mentioned application to pharmaceutical biotechnology, the comparison of normal biological systems (healthy states) and pathological (perturbed) states, provides an understanding of how different pathways interact within living cells and defines the phenotype of a ‘healthy state’. Only through the comparison of the two states is it possible to find ‘fragility’ (potential pharmaceutical targets) in a system as robust as cancer cells41. This process is also crucial to medical engineering used as combination treatment, where multiple processes are targeted to cure complex diseases, and to the development of individualized and preventive therapy. The latter application is likely to completely revolutionize the current conception of doctors and patients. Individualized medicine, defined as the approach to curing diseases that takes into account the individual's phenotypic profile, combined with prediction of the appropriate drug mixture through computer simulations, is expected to be the future treatment of many complex life-style diseases, e.g. diabetes and coronary-heart diseases. Looking into the future, the ability to view a system as complex as the human body to predict diseases from the study of a person's genome, proteome, metabolome and fluxome will start a new scientific era. The view of these advances as pure ‘science-fiction’ gains a further dimension if one takes into consideration the recent goals of a joint effort between the Institute of Systems Biology, Caltech and UCLA—the NanoSystems Biology Alliance. This alliance is working towards the merger of nanobiotechnology to systems biology by the creation of stacked nano-labs, integrated through microfluidics, that will allow researchers to carry out several ‘omes’ assays at once72. This has potential to bring about the development of portable micro-diagnostics. Whilst opening new horizons in medicine, these possibilities will introduce divergent ethical views into society by introducing a paradigmatic shift in the perception of healthcare72.
Food, chemical and pharmaceutical industries
The rapid development of efficient biotechnological processes will be made possible through metabolic engineering of tailor-made organisms as cell factories for the sustainable processing and production of food, feed ingredients, enzymes, advanced biopolymers, chemicals and pharmaceuticals. It was estimated that by the year 2010 industrial biotechnology (the production of industrial compounds by the use of biotechnology) will produce 10–20% of all chemicals and that, within the same time frame, the fine chemical segment of the market will turn to biotechnology in 60% of cases19. There are already a few examples of how genome-scale metabolic models have been used to identify targets for metabolic engineering61, 5. Most of these examples are relatively simple and could probably have been identified by classical means. However, as the models become more robust and include regulatory features, they will enable identification of fairly complex strategies which may be difficult to predict by classical means. An increasingly central aspect of the development of all improved bioprocesses is its economical and environmental sustainability. In this respect, the use of systems biology for the improved and eco-efficient use of renewable feedstock resources as bulk material for the production of chemicals, bioplastic or biofuels has a direct impact also on plant biotechnology. Several case studies of companies that are using industrial biotechnology (such as Novozymes, DuPont, BASF and DSM) show a resulting beneficial impact on both the environment and the cost of production19. Systems biology will clearly represent an important element and may well be a key enabling science of the future in the industrial biotech sector.
Understanding microbial metabolic networks will also reinforce microbial toxicology and microbial ecology, thus enabling more informed risk assessment in the application of new biological entities to the environment, as well as using concepts and methods from systems biology to improve biological waste treatment processes.
By looking at the above-mentioned contributions that systems biology is already bringing to society, the extent of its future impact becomes evident. The pending question is then, why is biotechnology estimated to be used in 60% of processes only by the year 2010 and why has this not happened before? What is delaying the fast development of this approach? As far as the industries are concerned, there is clearly an issue of proof-of-concept. In the case of academia and small start-up companies, there are other impeding issues, involving equipment costs and the need for multidisciplinary expertise. For these very reasons, the future of systems biology resides in collaborations among scientists and research centres.
Model organisms: why yeast?
The evolutionary conservation of biochemical pathways is the implicit basis for genetic model systems, such as the mouse (Mus musculus), the fruit fly (Drosophila melanogaster), the nematode (Coenorhabditis elegans), the zebrafish (Danio rerio), budding yeast (Saccharomyces cerevisiae) and fission yeast (Schizosaccharomyces pombe) and plants such as Arabidopsis thaliana, etc.
First of all, such model systems have allowed the elucidation of general scientific principles as well as development of generic experimental approaches suitable for studying any organism. As a more recent example, early microarray studies performed on S. cerevisiae led to first glimpses of the genomic dimension of transcriptional reprogramming23 as well as the further development of array technology and the bioinformatics tools to visualize the data18, 48. In addition, model systems are employed to identify and study key components and dynamic functions of metabolic, signalling or cell biological pathways and ultimately contribute towards identification of new drug targets or understand the molecular mechanisms underlying specific diseases.
The major current objectives of systems biology are the elucidation of principle properties of biological modules, to develop concepts for quantitative description of living cells as well as generating the computational tools for system analyses. This again requires model organisms in which detailed studies can be performed rapidly, accurately and reproducibly.
Due to the complexity of higher eukaryotic systems, it is appropriate to choose a relatively simple, but still relevant, model system for the development of systems biology technologies at the subcellular level. S. cerevisiae serves this purpose for numerous reasons:
S. cerevisiae has the best studied eukaryotic genome27, with genome sequences from several related yeasts being now available12, 16, 17, 40, 39.
Most high-throughput functional genomics techniques and the underlying bioinformatics for analysis of the transcriptome, proteome, metabolome and interactome were originally developed using yeast. This has resulted in databases (such as http://www.yeastgenome.org/; http://mips.gsf.de and numerous resources to which those link), that represent valuable, although not sufficient, sources for systems biology.
Genetic manipulations can be performed with greater ease than in any other cell, allowing truly genome-scale genetic analyses60, 66. Human (and other eukaryotic) genes can easily be expressed and studied in S. cerevisiae54.
Yeast constitutes the largest eukaryotic toolbox. A ‘bar-coded’ complete set of S. cerevisiae deletion mutants26 and comprehensive sets of open reading frames tagged for analysis of protein localization, protein levels and protein complex purification25, 34, 24 are available.
S. cerevisiae can be cultivated under controlled conditions, allowing for studies of high reproducibility (e.g.2, 62, 70).
Yeast is a paradigm for all eukaryotic cells. Most fundamental cellular processes are conserved from S. cerevisiae to humans and often have first been discovered in yeast. Many human disease genes have homologues in yeast53.
The yeast research community encompasses far more than 1000 laboratories and involves many world-leading scientists. The community is well organized, open for exchange of data and material, and has a demonstrated record in joint achievements.
Yeast is an excellent organism for training of candidates in cell biology, bioinformatics and systems biology.
Thirty percent of human genes involved in the development of diseases have functional homologues in yeast22. Therefore, yeast serves as an important model organism for resolving the molecular mechanisms governing, for example, various metabolic diseases47, cancer46 and Parkinson's disease64.
A recent review by Mager and Winderickx presents several examples of how yeast, bearing human cDNA for the expression of specific proteins, has played a fundamental role in the elucidation of important medicinal processes such as those of apoptosis control, anti-prion assays and two important neurodegenerative diseases (Alzheimer's and Parkinson's disease)52. Yeast has served as a key model organism, not only in fundamental studies of molecular mechanisms that are important for human health, but also in developing key technologies that are currently used in cutting-edge genomics research. Therefore, it is clearly understandable why yeast is widely recognized as a touchstone model in post-genomic research8, 57.
State of the art in yeast systems biology
Systems biology presently has two mainstream directions: (a) data-driven: identifying, building and analysing networks (modules) as an extension to the interpretation of genome-wide data; (b) module-driven: building of detailed kinetic (dynamic) models of cellular pathways or modules based on a variety of different qualitative, but most importantly quantitative, data.
The feasibility of generating genome-wide datasets has prompted computational modelling to interpret such data and build cell networks. Examples include the galactose metabolic network35, yeast protein interaction networks37, 67, 24, 30, 1, the yeast transcription network28, 50, the network orchestrating yeast filamentous growth63, the yeast genetic interaction network66, 75 or reconstructions of yeast metabolism20, 21.
Building detailed kinetic models has a history within physiology and metabolic analysis, also in yeast. Kinetic models that simulate processes over time currently focus on well-defined cellular modules or pathways. A classical subject for metabolic modelling is yeast glycolysis65. A similar contribution to the field is a model of the budding yeast cell cycle9, 11, 10. Signalling pathways subject to modelling include the pheromone response pathway45, 74 as well as yeast osmoregulation44. A recent remarkable contribution to a central question of systems biology, i.e. quantitative differences between individual cells, was presented by Roger Brent's group when studying the pheromone response13.
In order for the models to describe realistically the dynamics of cellular processes, such as the flow of metabolites or information through pathways, quantitative data, time courses and spatial information are required. While qualitative data have been useful to define pathways and describe functions, they are largely inadequate to describe processes. To obtain quantitative data and to capture subtle changes is straightforward for relative levels of certain metabolites, proteins or mRNA, but constitutes a challenge for the quantification of absolute levels of biomolecules. Such absolute levels have recently been approached for RNA (Takashi Ito, personal communication) and protein25. Ultimately, one would like to measure (or predict) AIMS-4D [amount, interactions, modifications, spatial movements at each XYZ + T (time coordinate) for every component]. Hence, technology development is needed both in computational systems biology and in data acquisition. In addition, a close collaboration between experimentalists and theoreticians is required for modelling to be driven by biological questions, for models to take into account biological knowledge beyond data, and to enable iterative experimental model testing and optimization.
Eventually, both mainstreams of systems biology should come together to build kinetic and spatial models of the entire cell. This can be achieved by connecting detailed models of cell modules using whole-cell networks as templates. This long-term vision will require substantial efforts as well as coordination both for data collection and computational analysis. This is one of the reasons why the Yeast Systems Biology Network (YSBN) has been established.
The need for joining forces: the birth of YSBN
Systems biology is, by definition, multidisciplinary. Only through close collaboration between experimental and theoretical scientists in various laboratories is it possible to fully exploit the large range of methods necessary to describe the complete dynamic operation of biological systems and to generate quantitative data and mathematical models taking into account time courses and spatial information. This is exemplified by a recent study of the HOG pathway in S. cerevisiae, where dynamic modelling of the signal transduction pathway was experimentally validated and subsequently used to gain new insight into the system43 So far, no single laboratory can encompass all competences and costs of a fully integrated systems biology institute, and only through coordination of activities in different laboratories will it be possible to substantially progress the field. It is widely perceived within the systems biology community that good contacts and coordination are essential if systems biology wants to become the driver of innovation and sustainability73. Despite the extensive focus on systems biology by funding agencies in the USA and Japan there is, as yet, no multi-laboratory effort on yeast systems biology like the Japanese effort on the bacterium E. coli.
The need for synergetic efforts in the field of yeast systems biology was put on the agenda in November 2003 at the International Conference on Systems Biology in St. Louis. However, in July 2003, during the XXI International Conference on Yeast Genetics and Molecular Biology in Gothenburg, the idea of formally joining forces to set up a global network establishing the important goals and objectives for the yeast systems biology community started taking shape. In 2004, a white paper defining the goals, priorities and direction of the ‘alliance’ was produced as a joint effort of the community31 and the proposition of applying for a EU Coordination Action for YSBN arose as a follow-up of EUSYSBIO, a Specific Support Action (SSA) focused on systems biology. Despite the lack of official funding for the initiative, this point in time signed the official coining of YSBN as an international ‘virtual team’ of researchers wanting to promote systems biology with the yeast S. cerevisiae as a model system. The current academics that have expressed interest in YSBN are listed in Table 1. Since June 2003, the community has already organized a number of YSBN-sponsored activities, with the scope of disseminating information on yeast systems biology and training and facilitating the inter-laboratory exchange of students (see Figure 2). As of 1 November 2005, the European branch of YSBN received substantial financial support via funding from the EU sixth framework programme in the form of a Coordination Action (CA). Among many other activities, there will be the first international YSBN-sponsored conference, which will take place in Helsinki in June 2006 (http://issy25.vtt.fi/).
Table 1. Institutions and individuals having expressed an interest in YSBN
Mike Cherry (SGD), Pat Brown, Ron Davis, Guri Giaever, Donna Bowe
Institute for Systems Biology, Seattle
John Aitchison, Tom Galitski
The Systems Biology Institute, Tokyo
Stefan Hohmann, Anders Blomberg
Chalmers University of Technology
Olle Nerman, Mats Jirstrand
Uwe Sauer, Ruedi Aebersold, Igor Stagljar
Medical University, Vienna
Technical University of Denmark
Jens Nielsen, Søren Brunak, Rasmus Wernersson
Virginia Polytechnic Institute and State University
Pedro Mendes, John Tyson
Max-Planck Institute for Molecular Genetics, Berlin
University of Manchester
Steve Oliver, Douglas Kell, Hans Westerhoff
University of Milano, Bicocca
Lilia Alberghina, Marco Vanoni
University of Stuttgart
University of Tokyo
Takashi Ito, Shinichi Morishita, Yoshikazu Ohya
Budapest University of Technology and Economics
David Botstein, Kara Dolinski
Merja Penttilä, Matej Oresic
University of Basel
University of Stellenbosch
University of California at San Diego
Bernard Palsson, Trey Ideker
Delft University of Technology
Jack Pronk, J. J. Heijnen, W. M. van Gulik, W. A. van Winden
University of Toronto
Charlie Boone, Tim Hughes, Brenda Andrews, Mike Tyers, Jack Greenblatt, Jack Greenblatt
Fritz Roth, Erin O'Shea, George Church, Frank Gibbons
Institute Curie, Paris
Alain Nicolas, Emmanuel Barillot
University of Amsterdam
Joost Teixeira de Mattos
University of New South Wales
Ian Dawes, Mark Temple
University of Graz
University of Bordeaux
Michel Aigle, Helian Boucherie, Macha Nikolski
Vrije Universiteit, Amsterdam
Barbara Backer, Hans Westerhoff
University of Karlsruhe
University of Washington
Washington University, St. Louis
The Blueprint Initiative
University of Newcastle
University of Helsinki
Marja Makarow, Risto Renkonen
University Pompeu Fabra, Barcelona
Jordi Villà i Freixa
University of Valencia
José Perez Ortin
University of Essex
M. V. Metodiev
Lunds Tekniska Höjskola
University of the Lleida
A European initiative: the YSBN coordination action
The Yeast Systems Biology Network (YSBN) Coordination Action (CA) is funded by the European Commission within its FP6 Programme, under the thematic area ‘Life sciences, genomics and biotechnology for health’, and it involves all of the YSBN European Partners plus two Scandinavian start-up companies (InNetics AB and Fluxome Sciences A/S) (Figure 3). The overall objectives of the coordination action are to structure European systems biology research on S. cerevisiae, by coordinating research activities in laboratories with different expertise, and to build an infrastructure that can be extended to other organisms. These goals will be achieved through:
1.The creation of standards for the documentation of transcriptome, metabolome, interactome, locasome, fluxome and other relevant quantitative data.
2.The creation of a web resource (as a complete restructuring of the existing YSBN site: www.ysbn.org).
3.The creation of an S. cerevisiae database that will link the available experimental information, thus facilitating and accelerating data-mining for modelling.
4.The demonstration of how systems biology can be used in the design of novel cell factories and software tools for the user-friendly visualization of modelling results.
5.The organization of courses, conferences and workshops.
The goal of standardizing procedures for data documentation is both ambitious and largely debated within the systems biology community. The creation of standards for all main procedures involved in systems biology presents challenges and raises important topics of debate. It should be stressed that it is not the goal of YSBN to standardize the way experiments are done or which strains to be used (the use of different yeast strains in the community rather is a resource for discovery), but to develop the mechanisms for standards of data documentation and presentation that facilitates their use for modelling. A number of organizations have already started valuable initiatives towards this goal. For example, the MGED Society has, since 2001, created the MIAME standards for documentation of transcriptome analysis, to which the community has responded very positively. Since the generation of these standards, several high-impact scientific journals (all journals within the Nature group, plus Cell, EMBO Journal and The Lancet) have formally requested compliance to MIAME standards by all authors. Following this request, both the number of publications approved through MIAMExpress and the interest of the scientific community in these standards has constantly increased since 2002 (Figure 4). Similarly, the HUPO Proteomics Standard Initiative (PSI) has, since 2002, developed standards for proteome research38. Learning from these initiatives, YSBN will discuss an approach to standardize the documentation of experiments specifically regarding the procedures used for yeast studies. Similar questions will also be addressed in relation to database development and bioinformatics services. The consortium of the YSBN Coordination Action will provide bioinformatics services in standardized ways, as several participating groups participate in the EMBRACE consortium (European Model for Bioinformatics Research and Community Education), which addresses issues of resource sharing in the GRID context at the European level. A large number of research laboratories participate in the EMBRACE network, where the purpose is to integrate the major databases and software tools in bioinformatics, using existing methods and emerging GRID technologies. Several initiatives aimed at defining and testing protocols allowing for efficient resource sharing in bioinformatics are under way, in various stages of development. Similarly YSBN groups also participate in another European large-scale genome annotation effort, BioSapiens (www.biosapiens.info), which aims at biological data–structure integration. These links will ensure that the present project can take advantage of the newest developments, without unproductive duplication of effort. Another technical aspect of the effort in YSBN will be to contribute to the development of distributed annotation systems, where annotation generated in one country will remain physically in that country, but will be made accessible to genome and proteome annotation efforts elsewhere via seamless protocols for database–database communication. DAS (biodas.org), BioMoby and other similar initiatives will, over time, replace the conventional centralized database, where functional characterizations of macromolecular data are kept in one single repository (e.g. UniProt).
YSBN is a global activity, but there is no instrument to provide funding for a global network. Hence, funding of the EC CA is a great opportunity for the entire YSBN to drive its activities and scientific performance. YSBN hopes to serve as an integrated network for all interested researchers and to provide a service for the scientific community. As stated above, its goal is to drive systems biology and the understanding of fundamental biology by bringing together experimental and computational scientists in joint efforts
The authors would like to thank the European Commission for funding the YSBN CA contract no.LSHG-CT-2005-018942 and for their active interest in the project's activities. In addition, RM would like to thank Joel F. Moxley for his help with Figure 4.