Almost two decades ago at the inception of our institute, the enormous growth in the field of genomics (the sequencing and study of entire genomes) had only just made it possible to study molecular biology using a systems approach. Our undertaking was to study the plant system at the molecular level.
All biological organisms are, in fact, systems within systems. They are composed of networks of interdependent components that integrate to form a unified whole. Through systems approaches, researchers demonstrate these linkages by modifying the level of one component and observing the communicated effects on others.
Plants adjust their growth in response to a multitude of exogenous and endogenous cues. To thrive, they optimize their use of resources to meet their needs for energy and biosynthetic building blocks. Their ability to grow depends entirely on their own photosynthetic and metabolic capacity, so they constantly adjust their rate of development to match their metabolic status.
We perform detailed analyses of these complex interactions – integrating systems data rather than reducing the system to the interactions of its parts.
Unbiased screens of genotyped populations have replaced the pathway-oriented, (single) gene-driven analyses of decades past. We are searching for novel genes controlling metabolism and growth (metabolic and biomass quantitative trait loci). We focus on the complex ways plants respond to abiotic stresses, such as nutrient deprivation, anoxia, etc., to develop a picture of the processes taking place. We gear our analyses towards unravelling regulatory processes and the signal molecules involved.
We are becoming increasingly competent at acquiring and analysing phenotype data (transcripts, proteins, enzymes, metabolites, fluxes, growth, photosynthetic rates, etc.), which is leading quite naturally to network-based analyses of the molecular processes underlying systems phenomena.
These network analyses form the basis for multiple hypotheses regarding the importance of single parameters. In this context, a major challenge is the development and implementation of strategies that allow the independent validation/falsification of hypotheses derived from networks.
We are using and developing genetic implements (e.g., inducible expression or repression systems) that allow us to disturb and subsequently monitor the plant system. In addition, we are using chemical genomics – a strategy employing libraries of small molecules (natural compounds, aptamers, or the products of combinatorial chemistry) with high-throughput screening – to identify compounds that act as positive or negative regulators of individual gene products, pathways, or cellular phenotypes. Finally, we are making small, systematic changes in gene expression (e.g., knockout heterozygotes) and monitoring system responses.
From genome structure to genome function and genome evolution
There are two basic types of genomics approaches: structural approaches (such as mapping, sequencing, comparisons between and among genomes, and studies of genome organisation) and functional approaches with which gene expression is analysed at transcriptome (mRNA), metabolome (metabolite), or proteome (protein) levels, often using high throughput technology.
A few years ago, our institute made a profound fundamental genomic contribution (in the framework of an international consortium) by establishing the first complete BAC-based physical map of the Arabidopsis genome (see Developmental Physiology and Genomics). This map aided large-scale sequencing efforts. Now that Arabidopsis has been mapped and sequenced, the challenge is rapid and accurate plant phenotyping at the molecular level. Every group in our institute applies genomics in the context of analysing altered plant systems.
We are also interested in understanding how plant genomes evolved. In addition to using genomics data to reconstruct the co-evolution of the three genomes of the plant cell (in the nucleus, the plastids and the mitochondria), we have begun to reproduce key processes in plant genome evolution in laboratory experiments. We have developed genetic screens allowing us to observe gene transfer between compartments in real time and to elucidate the molecular event involved in functional activation of successfully transferred genes.
In practice, we have based our research platform on four concepts: genetic diversity, gauntlets, phenotyping, and biomodeling. Genetic diversity is created using forward and reverse genetics and by taking advantage of naturally occurring variation. Phenotyping is the characterisation of observable plant traits. Gauntlets are ordered sets of plant growth conditions that bring about quantitative changes in phenotype. Computational biomodeling builds computer simulations of the plant system.
By altering genetic parameters in the plant system and analysing the effects of these perturbations, we gain insight into biological processes. Often, variation is brought about using transgenic technology or mutagenesis. We routinely carry out genetic modifications of the genomes in the nucleus and the chloroplast of both model species and crops and we also make use of naturally occurring genetic variation (e.g. recombinant inbred lines, near isogenic lines, and introgression lines).
We measure the effects of system perturbation by phenotyping. The phenotype is the observable properties of a plant (including morphology, development, and behaviour). The genotype, on the other hand, is the inherited instructions the plant carries, which may or may not be expressed. The phenotype arises from the interaction of the genotype with its environment. Expression profiling is a functional genomics approach that allows us to get an overview of phenotype by checking out where and when transcripts, proteins, and metabolites are expressed. For high-throughput transcript profiling, Affymetrix gene chip arrays are used or custom microarrays are produced by arraying expressed sequence tags (ESTs) or oligonucleotides on high-density membranes. The Transcript Profiling group manages these arrays, which are now also used to study translational regulation in the cytosol, the plastid and the mitochondrion (via polysome profiling).
We have developed a gas chromatography/mass spectroscopy-based, high throughput metabolite profiling approach with which more than 1000 leaf extract compounds and 400 phloem exudate compounds can be detected. Two of our research groups are profiling proteins in specific tissues or cell types. A system enabling searches for post-translational protein modifications is also in place.
Transcript, metabolite, and protein profiling have enormous resolving power but lack spatial resolution. To remedy this, we generate tissue and cell type-specific samples for analysis. Subcellular metabolite distributions are resolved using non-aqueous fractionation.
Plants are analysed at the whole plant level by looking at yields, photosynthetic rates, and so on, in growth chambers and during field trials. We often analyse plants kinetically by coupling precursor feeding to mass spectral analysis.We use confocal microscopy to analyse morphology, and employ various forms of electrophysiological analysis.
We have developed suites of high-throughput sensitive cycling tests for metabolites to process large numbers of samples from plants that have been passed through gauntlets (see below). We have established assays for sugars, amino acids, pyrophosphate, glycolysis intermediates, and several other compounds.
Many of these assays are based in a novel enzyme cycling system. They are adapted for multiwell plates and a pipetting robot and help us prioritise samples for expression profiling. Moreover, we are developing accurate high throughput enzyme assays that can be adapted to detect post-translational regulation.
Many interesting phenotypes are missed because they are environmentally conditional or subtle. To help us perceive such phenotypes, we have developed gauntlets – sets of well-defined growth conditions (either mimicking naturally occurring environmental stresses or forcing plants to cope with informative, non-natural situations) that result in quantitative phenotypic changes. Gauntlets set up so far include nitrogen supply, low sugar, micronutrients, low temperatures, different atmospheric conditions, and light cycles.
Data mining and biomodeling
The large volumes of data acquired through our functional genomic approaches is systematically analysed and mined. Two of our research groups focus exclusively on these research aspects (see the Bioinformatics and Computational Biology group profiles). The Bioinformatics group focuses expressly on developing informational systems that foster an infrastructure for the analytical methods established in our institute. Institute-wide, our research efforts promote advances in our biomodeling capabilities.
Although primarily developed for basic research, some of our techniques and tools for genetic engineering also offer great potential in plant biotechnology. Furthermore, we are interested in exploring the practical applicability of the knowledge we gain about plant metabolism and gene functions. We use nuclear and chloroplast transformation technologies to pursue proof-of-concept applications in metabolic engineering, molecular farming and engineering of disease resistances and stress tolerance.