Understanding Plant Vacuolar Trafficking from a Systems Biology Perspective

SYSTEMS BIOLOGY APPLIED TO PLANT VACUOLAR TRAFFICKING: INTEGRATE AND CONQUER

One of the most striking characteristics of plant development is the integration of external stimuli to trigger organ-specific developmental responses. This encompasses environmental aspects of development such as stress tolerance, spatial aspects such as the gravitropic and light responses, and temporal aspects such as flowering time. In the last century, plant researchers identified many signal transduction pathways responsible for specific developmental programs; however, little is known about the mechanisms that allow the integration of those processes at the subcellular level. For instance, hormones, receptors, and signal intermediates may be located in different subcellular compartments, and at specific developmental stages the cell has to deal with several transport issues: how do enzymes get to their proper compartments, how are intermediates transported between compartments, and how are metabolites transported across distinct organellar membranes? A significant number of such transported molecules are targeted to the vacuole either directly or via endomembrane trafficking. Thus, an understanding of vacuolar trafficking lies at the core of dissecting events that are central to cellular strategies for plant development, growth, and viability (Surpin and Raikhel, 2004; Zouhar and Rojo, 2009).

In vegetative tissues, the central vacuole occupies much of the cell volume and is involved in cellular responses to environmental and biotic stresses, storage of metabolic products, digestion of cytoplasmic constituents, and the maintenance of cell turgor, the driving force for cell expansion (Bassham and Raikhel, 2000). Membrane and soluble proteins, as well as metabolites, are imported into the vacuole in a process that requires the coordinated action of biochemically distinct subcellular organelles coupled to sophisticated vesicle-mediated endomembrane trafficking systems (Sanderfoot and Raikhel, 2002; Hanton et al., 2007; Hwang and Robinson, 2009). To uncover the molecular mechanisms of sorting, targeting, and vesicle fusion and to elucidate the network of proteins involved in different developmental programs, reductionist approaches based on the dissection of endomembrane dynamics into discrete steps have been widely used. Thus, the combination of conventional genetics, molecular biology, and confocal microscopy has identified a multitude of structural and regulatory proteins as well as targeting and retention signals. This has resulted in the definition of two main vacuolar trafficking pathways: (1) the secretory pathway, in which the bulk flow transport of proteins directed to the plasma membrane is separated from receptor-mediated trafficking to the vacuole in the Golgi stacks; and (2) the endocytic pathway, which represents a counter flow of cargo and membrane proteins internalized from the plasma membrane to the vacuole. Aside from these default pathways, several signal-mediated mechanisms have been identified that redirect specific proteins from both pathways and serve as recycling mechanisms (Sanderfoot and Raikhel, 2002; Jurgens, 2004; Schellmann and Pimpl, 2009). Therefore, in the observe-explain-predict-control-design framework of plant vacuolar trafficking research, the reductionist approach has provided abundant information regarding the “observe” and “explain” phases, but it is limited in its ability to “predict” and “control” the rapidly growing number of developmental phenotypes resulting from vacuolar trafficking defects. The failure of traditional approaches to integrate subcellular processes into whole plant developmental programs emphasizes the need for new experimental approaches with quantitative outputs and predictive capabilities.

In recent years, systems biology, a discipline that seeks to understand and predict the quantitative features of multicomponent biological systems (Way and Silver, 2007), has emerged as an important tool to explain the complex operations of vacuolar protein trafficking and its integration at the molecular, tissue, and organism levels. The transformation of vacuolar trafficking research, initially a qualitative discipline, into a fully quantitative, theory-rich science requires serious adjustments in terms of experimental design and the implementation of education to teach future cell biologists not only basic biological phenomena and principles but also a combination of mathematical and computational skills. The early success of multidisciplinary approaches using the expanded Arabidopsis (Arabidopsis thaliana) toolbox (live cell imaging, chemical genomics, deep sequencing), combined with large-scale international projects such as genome sequencing and mutant collections, illustrate that systems biology applied to Arabidopsis will continue being centrally important to answer biologically relevant questions in plant science. In the specific context of vacuolar trafficking research, Arabidopsis will help us to understand the dynamics of vacuolar transport, the nature of the vesicles and cargoes being transported, and the developmental and tissue specificity of the vacuolar trafficking pathways. This knowledge is required to fully integrate vacuolar trafficking in the broader context of plant development. In this essay, we will try to forecast some of the challenges facing systems biology as applied to vacuolar trafficking research in plants (Fig. 1).

Figure 1.

Systems biology is an integrative tool in vacuolar trafficking research. The systems biology toolbox acquires information from the subcellular, cellular, tissue, and organism levels to generate quantitative models with predictive capabilities. These models are the building blocks for the integration of vacuolar trafficking in higher order developmental programs. A, Micrography of vesicle trafficking to the vacuole (V) through multivesicular bodies (MVBs). The subcellular level will identify components of vacuolar trafficking pathways. B, Accumulation of anthocyanins in a cell labeled with the tonoplast δ-TIP:GFP marker. The cellular level will integrate different vacuolar transport pathways (arrows and vesicle colors in the cell scheme). C, Accumulation of anthocyanins in δ-TIP:GFP-labeled vacuoles in cotyledons. The tissue-specific level will integrate information (red arrow) from different cells within a cell type. D, A semidominant mutation on a Rab GTPase family member derived from a Vac2 T-DNA-mutagenized population (Rosado et al., 2010) induces the secretion of the vacuole-targeted cargo CLV3:CTPP and causes late flowering and floral meristem developmental defects (A. Rosado and N.V. Raikhel, unpublished data). The organism level will integrate information (red arrows) from different tissues (different cell colors) and will connect vacuolar trafficking with higher order developmental programs.

DYNAMICS OF VACUOLAR TRAFFICKING: A SNAPSHOT IS NOT ENOUGH

Visualization techniques are at the heart of cell biology, and microscopes have been one of the most fundamental tools in our laboratories. The historical observations made by Hooke and van Leeuwenhoek in the 17th century allowed the former to coin the term “cell” (Hooke, 1678) and marked the start of plant microscopy. In the early 1900s, the development of histological dyes and contrast enhancing techniques, such as phase contrast and differential interference contrast, showed a more detailed architecture of internal compartments of the cell and began what we now consider the advent of imaging technologies. The use of chemical fixation techniques that preserved cellular structures and improved the optical transparency of tissues and the development of electron microscopy greatly enhanced the spatial resolution of images, and some of the most remarkable cell biology images during the past 40 years were micrographs representing snapshots of fixed cells with detailed information on the substructure of intracellular compartments and vesicles. As dramatic and instructive as these snapshots of cellular activity were, they failed to capture the intrinsically dynamic nature of cellular activities such as vacuolar trafficking. In recent years, this impasse was partially superseded in plants by the introduction of genetically encoded fluorophores, such as GFP (Chalfie et al., 1994), and the development of live cell imaging techniques that permitted systematic studies of protein localization in vivo. In the present, the final constraint for plant visualization techniques is the attainable spatial resolution imposed by the diffraction limit of light, which does not permit far-field light microscopy to resolve objects less than 150 to 200 nm (Abbe, 1873). Yet, many intracellular trafficking processes take place within these diffraction-limited distances, and the development of superresolution systems that allow visualization below the 150-nm limit will be essential for future plant vacuolar trafficking research.

With the advent of time-lapse microscopy, we have witnessed inventive solutions that have successfully overcome the diffraction limit of light and allow the observation of single molecules in animal cell samples. These techniques are based on the combination of multiphoton technology, photoactivatable or photoswitchable fluorophores, quantum dots, and high-accuracy single-molecule localization and have enabled the visualization of cellular features with previously unimagined detail. For example, stimulated emission-depletion microscopy has allowed real-time tracking of single vesicles and has revealed the anatomy and dynamics of protein clusters in animal cells. Fluorescence photoactivated localization microscopy has allowed the visualization of single molecules in multiple colors. And stochastic optical reconstruction microscopy and advances in electron tomography have allowed the simultaneous visualization of microtubules and clathrin-coated pits with 20- to 30-nm lateral resolution (for review, see Fernández-Suárez and Ting, 2008). Despite the success of superresolution techniques in animal systems, their application to plants has been largely unsuccessful due to the intrinsic properties of plant tissues, including (1) a cuticular layer that reduces the diffusion of membrane-permeable dyes, (2) autofluorescent cell walls with enzymatic activities that generate background signal and deplete the activity of organelle-specific dyes based on intracellular esterase activities, (3) refractive differences across cell structures that cause light scattering and resolution losses, (4) wavelength absorption properties of cell walls and plastids that limit the use of multiphoton technology due to heat generation and photodamage, and (5) lack of standard cell shapes and the inability to culture plant tissues in monolayers. On the other hand, an important advantage of plants compared with animal systems is the robustness and reliability of the GFP-based markers. Thus, GFP-based markers in plants can be used for time resolution fluorescence microscopy, in which differences in fluorescence decay help to define properties of the local environment of the marker, such as pH or complex formation, and for particle dynamics and tracking by using photobleaching techniques or GFP variants such as transient photoswitchable markers. Therefore, to observe plant vacuolar trafficking processes using superresolution, plant visualization tools have to be specifically designed, and desirable advances for plant applications should include fixation techniques that allow the homogenization of the refractive index while maintaining the integrity of the cells, a new set of fluorophores and membrane-permeable dyes that allow diffusion through the cell wall and do not depend on esterases for their activation, and brighter sources such as solid-state lasers and light-emitting diodes that allow the use of a wide range of wavelengths to minimize photodamage and background signal. Examples of vacuolar trafficking questions that will greatly benefit from the implementation of superresolution techniques in plants range from the study of polar gradients of auxins generated by endocytosis and recycling of PINFORMED auxin transporters according to environmental or developmental triggers (Geldner et al., 2003; Dhonukshe et al., 2007) to the study of dynamic processes at the plasma membrane that are important for immune reactions upon attack by pathogenic microorganisms (Robatzek et al., 2006) and the quantification of trafficking vesicles such as endosomes, exosomes, and multivesicular bodies subjected to dynamic changes in response to environmental signals (Robinson et al., 2008).

In the future, the combination of plant-friendly visualization techniques, GFP-based markers, and bioinformatic tools suited for the acquisition, processing, and quantification of imaging data will enable real-time access to subcellular quantitative dynamics in live plant cells. The new optical systems will continue improving the spatial resolution, and soon we will reach the goal of video-rate imaging of live cells with molecular (1–5 nm) resolution.

COATS AND CARGOES IDENTIFICATION: NOT ALL VESICLES WERE MADE EQUAL

Secretion, endocytosis, and transport to the lytic and storage vacuoles are highly coordinated features of the plant cell. These intracellular transport processes are facilitated by small coated vesicles (100 nm or less in diameter) that preferentially package soluble proteins sharing a common destination (Robinson et al., 2005). During the last decade, extensive research on the structure of coat proteins, the identification of proteins and complexes that regulate vesicle docking and fusion, and the mechanism by which vesicle membranes ensure the selective packaging and targeting of the cargo molecules has been performed in plants (for review, see Sanderfoot and Raikhel, 2002; Carter et al., 2004; Jurgens, 2004; Surpin and Raikhel, 2004; Lipka et al., 2007). Plant cells contain the three major types of vesicles (COPI, COPII, and clathrin coated) also found in other eukaryotic cells and share the major molecular components in vesicle-mediated protein transport (Hwang and Robinson, 2009). However, plant cells generally contain more isoforms of the coat proteins, docking complexes (SNAREs), and regulatory proteins (ARF-GTPases), which reflects the complexity of plant-specific transport pathways, such as the routes to the prevacuolar compartments, storage vacuoles, and cell wall (Sanderfoot, 2007; Nielsen et al., 2008). Although genomic information about the multiple isoforms of vesicle components is available in Arabidopsis, we still lack detailed knowledge regarding their cell type- and tissue-specific expression patterns as well as their specific association with lipid raft microdomains. In addition, basically nothing is known about the identity of the vesicle cargoes, the specificity of the regulatory mechanisms of transport, and the differential roles of vesicle populations in plant cell development. For these reasons, the characterization of the full set of biologically meaningful binary and ternary docking complexes, the elucidation of the lipid composition of the docking sites, the description of the mechanisms controlling protein complex assembly/disassembly, and the identification of specific cargoes are important areas of future vacuolar trafficking research.

Thus far, not many components of vesicles targeted to the vacuoles have been identified using reductionist approaches, although a large number of genes are presumably involved in these processes (Rojo and Denecke, 2008). The limitations of those approaches for the identification of vacuolar trafficking components are well known. Vacuoles are essential organelles, and mutations in numerous components of the vacuolar trafficking pathway result in gametophyte or embryo lethality (Rojo et al., 2001; Sanderfoot et al., 2001), and loss-of-function mutations in Arabidopsis frequently fail to produce observable phenotypes due to gene redundancy (Surpin and Raikhel., 2004). To overcome these limitations, the integration of the genomic information with chemical genomics approaches, which allows instantaneous, reversible, tunable, and conditional control of phenotypes in Arabidopsis, provides a strategic advantage and a powerful high-throughput tool to unravel the vacuolar trafficking gene network (Robert et al., 2008; Hicks and Raikhel, 2009). Bioactive chemicals also permit transport processes to be slowed or arrested, permitting the identification of the mostly unknown vesicle cargoes via biochemical and imaging approaches (Robert et al., 2008). In the future, the large-scale use of chemical-genomics screens will also require a deep knowledge of synthetic chemistry to create chemical libraries of compounds that can be absorbed but not metabolized by plants, the design of high-throughput methods to identify chemicals that specifically modify vacuolar trafficking components, and the development of additional computer modeling techniques to identify the interaction sites between a chemical and its cognate partner(s). A striking example of the importance of chemical genomic approaches in plant research was recently illustrated by the use of pyrabactin, a new growth inhibitor discovered in an abscisic acid (ABA)-related chemical genetic screen (Park et al., 2009). The combination of chemical screens, genetics, transcriptomic, computational modeling, and physiological evidence allowed the identification of the first family of ABA receptors in Arabidopsis (Ma et al., 2009; Park et al., 2009). These studies not only revolutionized the study of stress-related vesicular trafficking processes (with ongoing efforts focused on the identification of putative vesicles containing the ABA receptors in different stress conditions) but also have enormous implications in the biotechnology and crop development fields.

As important as the identification of the specific coats and vesicle populations is the determination of their transported cargoes, including proteins, hormones, or secondary metabolites. In this context, the development of methods for the isolation of vesicle subpopulations, and the characterization of their proteome and metabolome, will shed light on the mechanisms that trigger specific vesicle transport as well as their biological role in plants. In recent years, proteomic approaches have been successfully used to characterize the protein constituents of highly purified organelles, including mitochondria, chloroplasts, the Golgi apparatus, and vacuoles (for review, see Pan et al., 2005), and recent experiments indicate that proteomics analysis can be used for protein identification in isolated vesicles (G. Drakakaki and N.V. Raikhel, unpublished data). However, the analysis of specific vesicles by proteomic methods is still developing and presents a set of challenges that will have to be addressed in the future. First, the identification and sorting of cargoes requires the development of purification methods that are vesicle specific. Second, proteins cannot be amplified and vesicles are difficult to purify in large quantities; therefore, proteomic analyses have to be both more sensitive and quantitative. Third, depletion methods have to be developed to eliminate the most abundant proteins that mask less abundant proteins that may have critical roles. Fourth, proteomic analyses have to be coupled with metabolomic and lipidomic analyses to determine which metabolites share transport machinery with protein cargoes, which metabolites follow independent trafficking pathways, and where in the cell those cargoes are delivered. Finally, the integration of proteomic and metabolomic data beyond the current situation (where analysis still relies mainly on human expert interpretation) will require the creation of and access to highly standardized heterogeneous external databases that store and classify in a comprehensive manner diverse knowledge surrounding genomic sequences, proteins, lipids, metabolites, and phenotypes.

CELL AND TISSUE SPECIFICITY: WHERE VACUOLAR TRAFFICKING MEETS CONTEXT

Higher plants have developed multicellularity, and plant stem cells differentiate into a wide range of cell types and organs. In the past, vacuolar trafficking was analyzed from a cell-autonomous perspective, and the results obtained from an individual cell type were extrapolated to the whole system. For that reason, the conclusions of those approaches should be considered cautiously in terms of biological discovery, since they may be of limited value in studying development, organism responses to environment, or even cellular processes that require the coordinated action of different cell types. Now, it is known that specific vacuolar trafficking systems regulate fundamental processes such as plant nutrition, responses to environmental stresses, gravitropic and light responses, hormone transport and signaling, and plant defense responses in a tissue- and cell type-specific manner (Surpin and Raikhel, 2004; Robinson et al., 2005; Zouhar and Rojo, 2009). Therefore, the functional assignment of trafficking pathways and vacuole-targeted vesicles in different cell types and tissues are central to a mechanistic understanding of whole plant responses. In this context, the specificity of vacuole-targeted vesicles could be determined by changes in expression pattern and the subcellular localization of transport factors, the presence/absence of particular regulatory mechanisms, changes in the identity or state of the transported cargoes, or transient associations and interchange of cargoes in response to external stimuli.

In the future, approaches to determine the cell and tissue specificity of plant vacuolar trafficking pathways could exploit technologies such as fluorescence-activated cell sorting, a method for the separation of individual cell types based on GFP fluorescence (Bargmann and Birnbaum, 2010), or laser capture and microdissection systems that are currently used to obtain the gene expression profiles of individual cell types and tissues (Kerk et al., 2003). Coupled with these systems are advances in the ability to amplify mRNA populations from limited cell numbers, as is the case with apical and root stem cells and gametophytes (Honys and Twell, 2004; Mustroph et al., 2009). However, our dream microscope for plant tissue analysis will not need to separate and sort cell types and will combine the advantages of confocal microscopy with the ability to scan large samples. It would be fully automated and detect and classify phenotypes in whole seedlings in a tissue-specific manner, it would detect several probes simultaneously without damaging the samples, it would quantify and normalize against the total fluorescent signals of known standards, and it would visualize probes with high resolution. It would include advances in optical tomography that allowed three-dimensional reconstructions and time resolution fluorescence capabilities that allowed the tracking of dynamic processes. Finally, it would be a fully functional endoscopic instrument that would allow high-resolution imaging of subcellular compartments, avoiding the changes in the diffraction index. These combined advances will get us closer to the ideal situation in which scientists will choose to apply the microscope directly to a whole plant or a specific cell type and will analyze both tissue and subcellular performance in real time using superresolution imaging.

WHAT’S NEXT

In the near future, plant cell biologists will be able to analyze specific vesicles and organelle movement in real time, the advances in imaging technologies and algorithms will permit the recognition of complex intracellular phenotypes that include organelle shape, size, and velocity, and to rapidly assess the state of complex endomembrane trafficking networks in response to environmental challenges. We will also be able to perturb the interior and exterior of the cell in a controlled manner to analyze vacuolar trafficking behavioral changes and will simultaneously analyze many other aspects of plant growth and development to better understand vacuolar trafficking function in an integrated context. The use of truly multidisciplinary approaches such as imaging and quantitative analysis of cellular structures in living plants, large experimental data sets derived from the “omics” analyses, and inputs from the chemical, physical, engineering, mathematical, and computational sciences will allow the integration of the results into models and networks with predictive capabilities. These models will provide more quantitative and mechanistic answers to current vacuolar trafficking questions but will in turn generate a new set of unimaginable questions and challenges to be solved by future generations. As in other plant biology disciplines, a balance between high-throughput data and single-gene information will still be required to optimize efforts and deliver breakthrough knowledge. In this sense, systems biology will continue benefiting from reductionist approaches that will deliver basic building blocks for a better understanding of the complex field of plant vacuolar trafficking research. We hope to see this understanding reflected in future issues of Plant Physiology. Happy 25000!