Multiscale Metabolic Modeling: Dynamic Flux Balance Analysis on a Whole-Plant Scale

Plant metabolism is characterized by a unique complexity on the cellular, tissue, and organ levels. On a whole-plant scale, changing source and sink relations accompanying plant development add another level of complexity to metabolism. With the aim of achieving a spatiotemporal resolution of source-sink interactions in crop plant metabolism, a multiscale metabolic modeling (MMM) approach was applied that integrates static organ-specific models with a whole-plant dynamic model. Allowing for a dynamic flux balance analysis on a whole-plant scale, the MMM approach was used to decipher the metabolic behavior of source and sink organs during the generative phase of the barley (Hordeum vulgare) plant. It reveals a sink-to-source shift of the barley stem caused by the senescence-related decrease in leaf source capacity, which is not sufficient to meet the nutrient requirements of sink organs such as the growing seed. The MMM platform represents a novel approach for the in silico analysis of metabolism on a whole-plant level, allowing for a systemic, spatiotemporally resolved understanding of metabolic processes involved in carbon partitioning, thus providing a novel tool for studying yield stability and crop improvement.Plants are of vital significance as a source of food (Grusak and DellaPenna, 1999; Rogalski and Carrer, 2011), feed (Lu et al., 2011), energy (Tilman et al., 2006; Parmar et al., 2011), and feedstocks for the chemical industry (Metzger and Bornscheuer, 2006; Kinghorn et al., 2011). Given the close connection between plant metabolism and the usability of plant products, there is a growing interest in understanding and predicting the behavior and regulation of plant metabolic processes. In order to increase crop quality and yield, there is a need for methods guiding the rational redesign of the plant metabolic network (Schwender, 2009).Mathematical modeling of plant metabolism offers new approaches to understand, predict, and modify complex plant metabolic processes. In plant research, the issue of metabolic modeling is constantly gaining attention, and different modeling approaches applied to plant metabolism exist, ranging from highly detailed quantitative to less complex qualitative approaches (for review, see Giersch, 2000; Morgan and Rhodes, 2002; Poolman et al., 2004; Rios-Estepa and Lange, 2007).A widely used modeling approach is flux balance analysis (FBA), which allows the prediction of metabolic capabilities and steady-state fluxes under different environmental and genetic backgrounds using (non)linear optimization (Orth et al., 2010). Assuming steady-state conditions, FBA has the advantage of not requiring the knowledge of kinetic parameters and, therefore, can be applied to model detailed, large-scale systems. In recent years, the FBA approach has been applied to several different plant species, such as maize (Zea mays; Dal’Molin et al., 2010; Saha et al., 2011), barley (Hordeum vulgare; Grafahrend-Belau et al., 2009b; Melkus et al., 2011; Rolletschek et al., 2011), rice (Oryza sativa; Lakshmanan et al., 2013), Arabidopsis (Arabidopsis thaliana; Poolman et al., 2009; de Oliveira Dal’Molin et al., 2010; Radrich et al., 2010; Williams et al., 2010; Mintz-Oron et al., 2012; Cheung et al., 2013), and rapeseed (Brassica napus; Hay and Schwender, 2011a, 2011b; Pilalis et al., 2011), as well as algae (Boyle and Morgan, 2009; Cogne et al., 2011; Dal’Molin et al., 2011) and photoautotrophic bacteria (Knoop et al., 2010; Montagud et al., 2010; Boyle and Morgan, 2011). These models have been used to study different aspects of metabolism, including the prediction of optimal metabolic yields and energy efficiencies (Dal’Molin et al., 2010; Boyle and Morgan, 2011), changes in flux under different environmental and genetic backgrounds (Grafahrend-Belau et al., 2009b; Dal’Molin et al., 2010; Melkus et al., 2011), and nonintuitive metabolic pathways that merit subsequent experimental investigations (Poolman et al., 2009; Knoop et al., 2010; Rolletschek et al., 2011). Although FBA of plant metabolic models was shown to be capable of reproducing experimentally determined flux distributions (Williams et al., 2010; Hay and Schwender, 2011b) and generating new insights into metabolic behavior, capacities, and efficiencies (Sweetlove and Ratcliffe, 2011), challenges remain to advance the utility and predictive power of the models.Given that many plant metabolic functions are based on interactions between different subcellular compartments, cell types, tissues, and organs, the reconstruction of organ-specific models and the integration of these models into interacting multiorgan and/or whole-plant models is a prerequisite to get insight into complex plant metabolic processes organized on a whole-plant scale (e.g. source-sink interactions). Almost all FBA models of plant metabolism are restricted to one cell type (Boyle and Morgan, 2009; Knoop et al., 2010; Montagud et al., 2010; Cogne et al., 2011; Dal’Molin et al., 2011), one tissue or one organ (Grafahrend-Belau et al., 2009b; Hay and Schwender, 2011a, 2011b; Pilalis et al., 2011; Mintz-Oron et al., 2012), and only one model exists taking into account the interaction between two cell types by specifying the interaction between mesophyll and bundle sheath cells in C4 photosynthesis (Dal’Molin et al., 2010). So far, no model representing metabolism at the whole-plant scale exists.Considering whole-plant metabolism raises the problem of taking into account temporal and environmental changes in metabolism during plant development and growth. Although classical static FBA is unable to predict the dynamics of metabolic processes, as the network analysis is based on steady-state solutions, time-dependent processes can be taken into account by extending the classical static FBA to a dynamic flux balance analysis (dFBA), as proposed by Mahadevan et al. (2002). The static (SOA) and dynamic optimization approaches introduced in this work provide a framework for analyzing the transience of metabolism by integrating kinetic expressions to dynamically constrain exchange fluxes. Due to the requirement of knowing or estimating a large number of kinetic parameters, so far dFBA has only been applied to a plant metabolic model once, to study the photosynthetic metabolism in the chloroplasts of C3 plants by a simplified model of five biochemical reactions (Luo et al., 2009). Integrating a dynamic model into a static FBA model is an alternative approach to perform dFBA.In this study, a multiscale metabolic modeling (MMM) approach was applied with the aim of achieving a spatiotemporal resolution of cereal crop plant metabolism. To provide a framework for the in silico analysis of the metabolic dynamics of barley on a whole-plant scale, the MMM approach integrates a static multiorgan FBA model and a dynamic whole-plant multiscale functional plant model (FPM) to perform dFBA. The performance of the novel whole-plant MMM approach was tested by studying source-sink interactions during the seed developmental phase of barley plants.     

Patterns of Metabolite Changes Identified from Large-Scale Gene Perturbations in Arabidopsis Using a Genome-Scale Metabolic Network

Metabolomics enables quantitative evaluation of metabolic changes caused by genetic or environmental perturbations. However, little is known about how perturbing a single gene changes the metabolic system as a whole and which network and functional properties are involved in this response. To answer this question, we investigated the metabolite profiles from 136 mutants with single gene perturbations of functionally diverse Arabidopsis (Arabidopsis thaliana) genes. Fewer than 10 metabolites were changed significantly relative to the wild type in most of the mutants, indicating that the metabolic network was robust to perturbations of single metabolic genes. These changed metabolites were closer to each other in a genome-scale metabolic network than expected by chance, supporting the notion that the genetic perturbations changed the network more locally than globally. Surprisingly, the changed metabolites were close to the perturbed reactions in only 30% of the mutants of the well-characterized genes. To determine the factors that contributed to the distance between the observed metabolic changes and the perturbation site in the network, we examined nine network and functional properties of the perturbed genes. Only the isozyme number affected the distance between the perturbed reactions and changed metabolites. This study revealed patterns of metabolic changes from large-scale gene perturbations and relationships between characteristics of the perturbed genes and metabolic changes.Rational and quantitative assessment of metabolic changes in response to genetic modification (GM) is an open question and in need of innovative solutions. Nontargeted metabolite profiling can detect thousands of compounds, but it is not easy to understand the significance of the changed metabolites in the biochemical and biological context of the organism. To better assess the changes in metabolites from nontargeted metabolomics studies, it is important to examine the changed metabolites in the context of the genome-scale metabolic network of the organism.Metabolomics is a technique that aims to quantify all the metabolites in a biological system (Nikolau and Wurtele, 2007; Nicholson and Lindon, 2008; Roessner and Bowne, 2009). It has been used widely in studies ranging from disease diagnosis (Holmes et al., 2008; DeBerardinis and Thompson, 2012) and drug discovery (Cascante et al., 2002; Kell, 2006) to metabolic reconstruction (Feist et al., 2009; Kim et al., 2012) and metabolic engineering (Keasling, 2010; Lee et al., 2011). Metabolomic studies have demonstrated the possibility of identifying gene functions from changes in the relative concentrations of metabolites (metabotypes or metabolic signatures; Ebbels et al., 2004) in various species including yeast (Saccharomyces cerevisiae; Raamsdonk et al., 2001; Allen et al., 2003), Arabidopsis (Arabidopsis thaliana; Brotman et al., 2011), tomato (Solanum lycopersicum; Schauer et al., 2006), and maize (Zea mays; Riedelsheimer et al., 2012). Metabolomics has also been used to better understand how plants interact with their environments (Field and Lake, 2011), including their responses to biotic and abiotic stresses (Dixon et al., 2006; Arbona et al., 2013), and to predict important agronomic traits (Riedelsheimer et al., 2012). Metabolite profiling has been performed on many plant species, including angiosperms such as Arabidopsis, poplar (Populus trichocarpa), and Catharanthus roseus (Sumner et al., 2003; Rischer et al., 2006), basal land plants such as Selaginella moellendorffii and Physcomitrella patens (Erxleben et al., 2012; Yobi et al., 2012), and Chlamydomonas reinhardtii (Fernie et al., 2012; Davis et al., 2013). With the availability of whole genome sequences of various species, metabolomics has the potential to become a useful tool for elucidating the functions of genes using large-scale systematic analyses (Fiehn et al., 2000; Saito and Matsuda, 2010; Hur et al., 2013).Although metabolomics data have the potential for identifying the roles of genes that are associated with metabolic phenotypes, the biochemical mechanisms that link functions of genes with metabolic phenotypes are still poorly characterized. For example, we do not yet know the principles behind how perturbing the expression of a single gene changes the metabolic system as a whole. Large-scale metabolomics data have provided useful resources for linking phenotypes to genotypes (Fiehn et al., 2000; Roessner et al., 2001; Tikunov et al., 2005; Schauer et al., 2006; Lu et al., 2011; Fukushima et al., 2014). For example, Lu et al. (2011) compared morphological and metabolic phenotypes from more than 5,000 Arabidopsis chloroplast mutants using gas chromatography (GC)- and liquid chromatography (LC)-mass spectrometry (MS). Fukushima et al. (2014) generated metabolite profiles from various characterized and uncharacterized mutant plants and clustered the mutants with similar metabolic phenotypes by conducting multidimensional scaling with quantified metabolic phenotypes. Nonetheless, representation and analysis of such a large amount of data remains a challenge for scientific discovery (Lu et al., 2011). In addition, these studies do not examine the topological and functional characteristics of metabolic changes in the context of a genome-scale metabolic network. To understand the relationship between genotype and metabolic phenotype, we need to investigate the metabolic changes caused by perturbing the expression of a gene in a genome-scale metabolic network perspective, because metabolic pathways are not independent biochemical factories but are components of a complex network (Berg et al., 2002; Merico et al., 2009).Much progress has been made in the last 2 decades to represent metabolism at a genome scale (Terzer et al., 2009). The advances in genome sequencing and emerging fields such as biocuration and bioinformatics enabled the representation of genome-scale metabolic network reconstructions for model organisms (Bassel et al., 2012). Genome-scale metabolic models have been built and applied broadly from microbes to plants. The first step toward modeling a genome-scale metabolism in a plant species started with developing a genome-scale metabolic pathway database for Arabidopsis (AraCyc; Mueller et al., 2003) from reference pathway databases (Kanehisa and Goto, 2000; Karp et al., 2002; Zhang et al., 2010). Genome-scale metabolic pathway databases have been built for several plant species (Mueller et al., 2005; Zhang et al., 2005, 2010; Urbanczyk-Wochniak and Sumner, 2007; May et al., 2009; Dharmawardhana et al., 2013; Monaco et al., 2013, 2014; Van Moerkercke et al., 2013; Chae et al., 2014; Jung et al., 2014). Efforts have been made to develop predictive genome-scale metabolic models using enzyme kinetics and stoichiometric flux-balance approaches (Sweetlove et al., 2008). de Oliveira Dal’Molin et al. (2010) developed a genome-scale metabolic model for Arabidopsis and successfully validated the model by predicting the classical photorespiratory cycle as well as known key differences between redox metabolism in photosynthetic and nonphotosynthetic plant cells. Other genome-scale models have been developed for Arabidopsis (Poolman et al., 2009; Radrich et al., 2010; Mintz-Oron et al., 2012), C. reinhardtii (Chang et al., 2011; Dal’Molin et al., 2011), maize (Dal’Molin et al., 2010; Saha et al., 2011), sorghum (Sorghum bicolor; Dal’Molin et al., 2010), and sugarcane (Saccharum officinarum; Dal’Molin et al., 2010). These predictive models have the potential to be applied broadly in fields such as metabolic engineering, drug target discovery, identification of gene function, study of evolutionary processes, risk assessment of genetically modified crops, and interpretations of mutant phenotypes (Feist and Palsson, 2008; Ricroch et al., 2011).Here, we interrogate the metabotypes caused by 136 single gene perturbations of Arabidopsis by analyzing the relative concentration changes of 1,348 chemically identified metabolites using a reconstructed genome-scale metabolic network. We examine the characteristics of the changed metabolites (the metabolites whose relative concentrations were significantly different in mutants relative to the wild type) in the metabolic network to uncover biological and topological consequences of the perturbed genes.     

The Endocytosis of Cellulose Synthase in Arabidopsis Is Dependent on μ2, a Clathrin-Mediated Endocytosis Adaptin

Clathrin-mediated endocytosis (CME) is the best-characterized type of endocytosis in eukaryotic cells. Plants appear to possess all of the molecular components necessary to carry out CME; however, functional characterization of the components is still in its infancy. A yeast two-hybrid screen identified μ2 as a putative interaction partner of CELLULOSE SYNTHASE6 (CESA6). Arabidopsis (Arabidopsis thaliana) μ2 is homologous to the medium subunit 2 of the mammalian ADAPTOR PROTEIN COMPLEX2 (AP2). In mammals, the AP2 complex acts as the central hub of CME by docking to the plasma membrane while concomitantly recruiting cargo proteins, clathrin triskelia, and accessory proteins to the sites of endocytosis. We confirmed that μ2 interacts with multiple CESA proteins through the μ-homology domain of μ2, which is involved in specific interactions with endocytic cargo proteins in mammals. Consistent with its role in mediating the endocytosis of cargos at the plasma membrane, μ2-YELLOW FLUORESCENT PROTEIN localized to transient foci at the plasma membrane, and loss of μ2 resulted in defects in bulk endocytosis. Furthermore, loss of μ2 led to increased accumulation of YELLOW FLUORESCENT PROTEIN-CESA6 particles at the plasma membrane. Our results suggest that CESA represents a new class of CME cargo proteins and that plant cells might regulate cellulose synthesis by controlling the abundance of active CESA complexes at the plasma membrane through CME.Cellulose microfibrils, as the major load-bearing polymers in plant cell walls, are the predominant component that enforces asymmetric cell expansion (Green, 1962). In higher plants, cellulose is synthesized by multimeric rosettes, which are also referred to as cellulose synthase complexes (CSCs; Kimura et al., 1999). Genetic and coimmunoprecipitation studies have indicated that CELLULOSE SYNTHASE1 (CESA1), CESA3, and CESA6-like (CESA6, CESA2, CESA5, and CESA9) isoforms are constituents of CSCs during primary cell wall synthesis (Persson et al., 2005; Desprez et al., 2007; Persson et al., 2007; Wang et al., 2008), whereas CESA4, CESA7, and CESA8 are implicated in the cellulose synthesis of secondary cell walls (Taylor et al., 1999, 2003; Brown et al., 2005). Knowledge about cellulose synthesis has recently been enhanced by the development of a system whereby the dynamics of CESA can be imaged in living cells (Paredez et al., 2006; Desprez et al., 2007). In agreement with earlier transmission electron microscopy studies in which rosettes were visualized in Golgi cisternae, vesicles, and at the plasma membrane (Haigler and Brown, 1986), fluorescent protein tagging of CESA has identified CESA localization at the plasma membrane, in Golgi bodies, and in small intracellular compartments (Paredez et al., 2006; Desprez et al., 2007; Crowell et al., 2009; Gutierrez et al., 2009; Gu et al., 2010; Lei et al., 2012; Li et al., 2012b).Assuming that cellulose synthesis occurs solely at the plasma membrane, the trafficking of CSCs to and from the plasma membrane may act as a significant regulatory mechanism. Although the mechanistic details of CESA trafficking are lacking, live cell imaging has shown that CESA localizes to various subcellular compartments. A subset of CESAs colocalize with markers of the trans-Golgi network (TGN)/early endosome (EE), an organelle that is part of both the secretory and endocytic pathways in Arabidopsis (Arabidopsis thaliana; Dettmer et al., 2006; Lam et al., 2007; Crowell et al., 2009, 2010; Viotti et al., 2010). CESAs also localize to microtubule-associated cellulose synthase compartments (MASCs) and small CESA-containing compartments (SmaCCs). The exact function of SmaCCs/MASCs is unknown, but it has been proposed that SmaCCs/MASCs might result from the internalization of CSCs or might act in the delivery of CSCs to the plasma membrane (Crowell et al., 2009, 2010; Gutierrez et al., 2009).Clathrin-mediated endocytosis (CME) has been shown to be a major endocytic pathway in Arabidopsis (Holstein, 2002; Samaj et al., 2005; Dhonukshe et al., 2007; Kleine-Vehn and Friml, 2008; Chen et al., 2011; Beck et al., 2012; Wang et al., 2013), although there is also evidence of clathrin-independent endocytosis mechanisms (Bandmann and Homann, 2012). The function of many CME proteins has been extensively characterized in mammals (McMahon and Boucrot, 2011), and homologs of many CME components are encoded by the Arabidopsis genome, including multiple copies of clathrin H chain and clathrin light chain (CLC), all four subunits of the heterotetrameric ADAPTOR PROTEIN COMPLEX2 (AP2) complex, dynamin-related proteins, and accessory proteins such as AP180 (Holstein, 2002; Chen et al., 2011); however, many CME components have yet to be characterized in plants.It has been suggested that CME might also function in controlling cell wall metabolism. For example, dividing and growing cells internalize cross-linked cell wall pectins, which might allow for cell wall remodeling (Baluska et al., 2002, 2005; Samaj et al., 2004). Moreover, the importance of endocytosis for cell wall morphogenesis is apparent from the functional characterization of proteins involved in CME. A dynamin-related protein, DRP1A, plays a significant role in endocytosis and colocalizes with CLC (Collings et al., 2008; Konopka and Bednarek, 2008). Defective endocytosis in RADIAL SWELLING9 (rsw9) plants, which contain a mutation in DRP1A, results in cellulose deficiency and defects in cell elongation (Collings et al., 2008). A mutation in rice, brittle culm3 (bc3), was mapped to the dynamin-related gene OsDRP2A, which has been proposed to function in CME. The brittle-culm phenotype in this mutant was attributed to cellulose deficiency (Xiong et al., 2010). Although the abundance of OsCESA4 was also altered in bc3, it remains unclear whether the cellulose deficiency of either bc3 or rsw9 results directly from perturbations in CESA trafficking.To identify proteins involved in the regulation of cellulose biosynthesis, a yeast two-hybrid (Y2H) screen was performed in which the central domain of CESA6 (CESA6CD) was used as bait to screen an Arabidopsis complementary DNA library for potential interaction partners of CESA6 (Gu et al., 2010; Gu and Somerville, 2010). The Y2H screen identified μ2 as a putative interaction partner of CESA6CD. The mammalian homolog of μ2 is the medium subunit of the tetrameric AP2, which acts as the core of the CME machinery by docking to the plasma membrane while concomitantly recruiting cargo proteins, clathrin triskelia, and accessory proteins to the sites of endocytosis (Jackson et al., 2010; McMahon and Boucrot, 2011; Cocucci et al., 2012). In this study, we provide evidence that μ2 plays a role in CME in Arabidopsis, that CESAs are a new set of CME cargo proteins, and that plant cells might regulate cellulose synthesis by controlling the abundance of CSCs at the plasma membrane through CME. To our knowledge, this study is the first to show the affect of an AP2 complex component on endocytosis in plants and the first to visualize an AP2 complex component in living plant cells. Furthermore, our data suggest that the role of AP2 in plants may differ from what has been shown in animals.     

Cavitation Resistance in Seedless Vascular Plants: The Structure and Function of Interconduit Pit Membranes

Plant water transport occurs through interconnected xylem conduits that are separated by partially digested regions in the cell wall known as pit membranes. These structures have a dual function. Their porous construction facilitates water movement between conduits while limiting the spread of air that may enter the conduits and render them dysfunctional during a drought. Pit membranes have been well studied in woody plants, but very little is known about their function in more ancient lineages such as seedless vascular plants. Here, we examine the relationships between conduit air seeding, pit hydraulic resistance, and pit anatomy in 10 species of ferns (pteridophytes) and two lycophytes. Air seeding pressures ranged from 0.8 ± 0.15 MPa (mean ± sd) in the hydric fern Athyrium filix-femina to 4.9 ± 0.94 MPa in Psilotum nudum, an epiphytic species. Notably, a positive correlation was found between conduit pit area and vulnerability to air seeding, suggesting that the rare-pit hypothesis explains air seeding in early-diverging lineages much as it does in many angiosperms. Pit area resistance was variable but averaged 54.6 MPa s m⁻¹ across all surveyed pteridophytes. End walls contributed 52% to the overall transport resistance, similar to the 56% in angiosperm vessels and 64% in conifer tracheids. Taken together, our data imply that, irrespective of phylogenetic placement, selection acted on transport efficiency in seedless vascular plants and woody plants in equal measure by compensating for shorter conduits in tracheid-bearing plants with more permeable pit membranes.Water transport in plants occurs under tension, which renders the xylem susceptible to air entry. This air seeding may lead to the rupture of water columns (cavitation) such that the air expands within conduits to create air-vapor embolisms that block further transport. (Zimmermann and Tyree, 2002). Excessive embolism such as that which occurs during a drought may jeopardize leaf hydration and lead to stomatal closure, overheating, wilting, and possibly death of the plant (Hubbard et al., 2001; Choat et al., 2012; Schymanski et al., 2013). Consequently, strong selection pressure resulted in compartmentalized and redundant plant vascular networks that are adapted to a species habitat water availability by way of life history strategy (i.e. phenology) or resistance to air seeding (Tyree et al., 1994; Mencuccini et al., 2010; Brodersen et al., 2012). The spread of drought-induced embolism is limited primarily by pit membranes, which are permeable, mesh-like regions in the primary cell wall that connect two adjacent conduits. The construction of the pit membrane is such that water easily moves across the membrane between conduits, but because of the small membrane pore size and the presence of a surface coating on the membrane (Pesacreta et al., 2005; Lee et al., 2012), the spread of air and gas bubbles is restricted up to a certain pressure threshold known as the air-seeding pressure (ASP). When xylem sap tension exceeds the air-seeding threshold, air can be aspirated from an air-filled conduit into a functional water-filled conduit through perhaps a large, preexisting pore or one that is created by tension-induced membrane stress (Rockwell et al., 2014). Air seeding leads to cavitation and embolism formation, with emboli potentially propagating throughout the xylem network (Tyree and Sperry, 1988; Brodersen et al., 2013). So, on the one hand, pit membranes are critical to controlling the spread of air throughout the vascular network, while on the other hand, they must facilitate the efficient flow of water between conduits (Choat et al., 2008; Domec et al., 2008; Pittermann et al., 2010; Schulte, 2012). Much is known about such hydraulic tradeoffs in the pit membranes of woody plants, but comparatively little data exist on seedless vascular plants such as ferns and lycophytes. Given that seedless vascular plants may bridge the evolutionary transition from bryophytes to woody plants, the lack of functional data on pit membrane structure in early-derived tracheophytes is a major gap in our understanding of the evolution of plant water transport.In woody plants, pit membranes fall into one of two categories: the torus-margo type found in most gymnosperms and the homogenous pit membrane characteristic of angiosperms (Choat et al., 2008; Choat and Pittermann, 2009). In conifers, water moves from one tracheid to another through the margo region of the membrane, with the torus sealing the pit aperture should one conduit become embolized. Air seeding occurs when water potential in the functional conduit drops low enough to dislodge the torus from its sealing position, letting air pass through the pit aperture into the water-filled tracheid (Domec et al., 2006; Delzon et al., 2010; Pittermann et al., 2010; Schulte, 2012; but see Jansen et al., 2012). Across north-temperate conifer species, larger pit apertures correlate with lower pit resistance to water flow (r_pit; MPa s m⁻¹), but it is the ratio of torus-aperture overlap that sets a species cavitation resistance (Pittermann et al., 2006, 2010; Domec et al., 2008; Hacke and Jansen, 2009). A similar though mechanistically different tradeoff exists in angiosperm pit membranes. Here, air seeding reflects a probabilistic relationship between membrane porosity and the total area of pit membranes present in the vessel walls. Specifically, the likelihood of air aspirating into a functional conduit is determined by the combination of xylem water potential and the diameter of the largest pore and/or the weakest zone in the cellulose matrix in the vessel’s array of pit membranes (Wheeler et al., 2005; Hacke et al., 2006; Christman et al., 2009; Rockwell et al., 2014). As it has come to be known, the rare-pit hypothesis suggests that the infrequent, large-diameter leaky pore giving rise to that rare pit reflects some combination of pit membrane traits such as variation in conduit membrane area (large or small), membrane properties (tight or porous), and hydrogel membrane chemistry (Hargrave et al., 1994; Choat et al., 2003; Wheeler et al., 2005; Hacke et al., 2006; Christman et al., 2009; Lee et al., 2012; Plavcová et al., 2013; Rockwell et al., 2014). The maximum pore size is critical because, per the Young-Laplace law, the larger the radius of curvature, the lower the air-water pressure difference under which the contained meniscus will fail (Jarbeau et al., 1995; Choat et al., 2003; Jansen et al., 2009). Consequently, angiosperms adapted to drier habitats may exhibit thicker, denser, smaller, and less abundant pit membranes than plants occupying regions with higher water availability (Wheeler et al., 2005; Hacke et al., 2007; Jansen et al., 2009; Lens et al., 2011; Scholz et al., 2013). However, despite these qualitative observations, there is no evidence that increased cavitation resistance arrives at the cost of higher r_pit. Indeed, the bulk of the data suggest that prevailing pit membrane porosity is decoupled from the presence of the single largest pore that allows air seeding to occur (Choat et al., 2003; Wheeler et al., 2005 Hacke et al., 2006, 2007).As water moves from one conduit to another, pit membranes offer considerable hydraulic resistance throughout the xylem network. On average, r_pit contributes 64% and 56% to transport resistance in conifers and angiosperms, respectively (Wheeler et al., 2005; Pittermann et al., 2006; Sperry et al., 2006). In conifers, the average r_pit is estimated at 6 ± 1 MPa s m⁻¹, almost 60 times lower than the 336 ± 81 MPa s m⁻¹ computed for angiosperms (Wheeler et al., 2005; Hacke et al., 2006; Sperry et al., 2006). Presumably, the high porosity of conifer pits compensates for the higher transport resistance offered by a vascular system composed of narrow, short, single-celled conduits (Pittermann et al., 2005; Sperry et al., 2006).Transport in seedless vascular plants presents an interesting conundrum because, with the exception of a handful of species, their primary xylem is composed of tracheids, the walls of which are occupied by homogenous pit membranes (Gibson et al., 1985; Carlquist and Schneider, 2001, 2007; but see Morrow and Dute, 1998, for torus-margo membranes in Botrychium spp.). At first pass, this combination of traits appears hydraulically maladaptive, but several studies have shown that ferns can exhibit transport capacities that are on par with more recently evolved plants (Wheeler et al., 2005; Watkins et al., 2010; Pittermann et al., 2011, 2013; Brodersen et al., 2012). Certainly, several taxa possess large-diameter, highly overlapping conduits, some even have vessels such as Pteridium aquilinum and many species have high conduit density, all of which could contribute to increased hydraulic efficiency (Wheeler et al., 2005; Pittermann et al., 2011, 2013). But how do the pit membranes of seedless vascular plants compare? Scanning electron micrographs of fern and lycopod xylem conduits suggest that they are thin, diaphanous, and susceptible to damage during specimen preparation (Carlquist and Schneider 2001, 2007). Consistent with such observations, two estimates of r_pit imply that r_pit in ferns may be significantly lower than in angiosperms; Wheeler et al. (2005) calculated r_pit in the fern Pteridium aquilinum at 31 MPa s m⁻¹, while Schulte et al. (1987) estimated r_pit at 1.99 MPa s m⁻¹ in the basal fern Psilotum nudum. The closest structural analogy to seedless vascular plant tracheids can be found in the secondary xylem of the early-derived vesselless angiosperms, in which tracheids possess homogenous pit membranes with r_pit values that at 16 MPa s m⁻¹ are marginally higher than those of conifers (Hacke et al., 2007). Given that xylem in seedless vascular plants is functionally similar to that in vesselless angiosperms, we expected convergent r_pit values in these two groups despite their phylogenetic distance. We tested this hypothesis, as well as the intrinsic cavitation resistance of conduits in seedless vascular plants, by scrutinizing the pit membranes of ferns and fern allies using the anatomical and experimental approaches applied previously to woody taxa. In particular, we focused on the relationship between pit membrane traits and cavitation resistance at the level of the individual conduit.     

ORM Expression Alters Sphingolipid Homeostasis and Differentially Affects Ceramide Synthase Activity

Sphingolipid synthesis is tightly regulated in eukaryotes. This regulation in plants ensures sufficient sphingolipids to support growth while limiting the accumulation of sphingolipid metabolites that induce programmed cell death. Serine palmitoyltransferase (SPT) catalyzes the first step in sphingolipid biosynthesis and is considered the primary sphingolipid homeostatic regulatory point. In this report, Arabidopsis (Arabidopsis thaliana) putative SPT regulatory proteins, orosomucoid-like proteins AtORM1 and AtORM2, were found to interact physically with Arabidopsis SPT and to suppress SPT activity when coexpressed with Arabidopsis SPT subunits long-chain base1 (LCB1) and LCB2 and the small subunit of SPT in a yeast (Saccharomyces cerevisiae) SPT-deficient mutant. Consistent with a role in SPT suppression, AtORM1 and AtORM2 overexpression lines displayed increased resistance to the programmed cell death-inducing mycotoxin fumonisin B₁, with an accompanying reduced accumulation of LCBs and C16 fatty acid-containing ceramides relative to wild-type plants. Conversely, RNA interference (RNAi) suppression lines of AtORM1 and AtORM2 displayed increased sensitivity to fumonisin B₁ and an accompanying strong increase in LCBs and C16 fatty acid-containing ceramides relative to wild-type plants. Overexpression lines also were found to have reduced activity of the class I ceramide synthase that uses C16 fatty acid acyl-coenzyme A and dihydroxy LCB substrates but increased activity of class II ceramide synthases that use very-long-chain fatty acyl-coenzyme A and trihydroxy LCB substrates. RNAi suppression lines, in contrast, displayed increased class I ceramide synthase activity but reduced class II ceramide synthase activity. These findings indicate that ORM mediation of SPT activity differentially regulates functionally distinct ceramide synthase activities as part of a broader sphingolipid homeostatic regulatory network.Sphingolipids play critical roles in plant growth and development as essential components of endomembranes, including the plasma membrane, where they constitute more than 40% of the total lipid (Sperling et al., 2005; Cacas et al., 2016). Sphingolipids also are highly enriched in detergent-insoluble membrane fractions of the plasma membrane that form microdomains for proteins with important cell surface activities, including cell wall biosynthesis and hormone transport (Cacas et al., 2012, 2016; Perraki et al., 2012; Bayer et al., 2014). In addition, sphingolipids, particularly those with very-long-chain fatty acids (VLCFAs), are integrally associated with Golgi-mediated protein trafficking that underlies processes related to the growth of plant cells (Bach et al., 2008, 2011; Markham et al., 2011; Melser et al., 2011). Furthermore, sphingolipids function through their bioactive long-chain base (LCB) and ceramide metabolites to initiate programmed cell death (PCD), important for mediating plant pathogen resistance through the hypersensitive response (Greenberg et al., 2000; Liang et al., 2003; Shi et al., 2007; Bi et al., 2014; Simanshu et al., 2014).Sphingolipid biosynthesis is highly regulated in all eukaryotes. In plants, the maintenance of sphingolipid homeostasis is vital to ensure sufficient sphingolipids for growth (Chen et al., 2006; Kimberlin et al., 2013) while restricting the accumulation of PCD-inducing ceramides and LCBs until required for processes such as the pathogen-triggered hypersensitive response. Serine palmitoyltransferase (SPT), which catalyzes the first step in LCB synthesis, is generally believed to be the primary control point for sphingolipid homeostasis (Hanada, 2003). SPT synthesizes LCBs, unique components of sphingolipids, by catalyzing a pyridoxal phosphate-dependent condensation of Ser and palmitoyl (16:0)-CoA in plants (Markham et al., 2013). Similar to other eukaryotes, the Arabidopsis (Arabidopsis thaliana) SPT is a heterodimer consisting of LCB1 and LCB2 subunits (Chen et al., 2006; Dietrich et al., 2008; Teng et al., 2008). Research to date has shown that SPT is regulated primarily by posttranslational mechanisms involving physical interactions with noncatalytic, membrane-associated proteins that confer positive and negative regulation of SPT activity (Han et al., 2009, 2010; Breslow et al., 2010). These proteins include a 56-amino acid small subunit of SPT (ssSPT) in Arabidopsis, which was recently shown to stimulate SPT activity and to be essential for generating sufficient amounts of sphingolipids for pollen and sporophytic cell viability (Kimberlin et al., 2013).Evidence from yeast and mammalian research points to a more critical role for proteins termed ORMs (for orosomucoid-like proteins) in sphingolipid homeostatic regulation (Breslow et al., 2010; Han et al., 2010). The Saccharomyces cerevisiae Orm1p and Orm2p negatively regulate SPT through reversible phosphorylation of these polypeptides in response to intracellular sphingolipid levels (Breslow et al., 2010; Han et al., 2010; Roelants et al., 2011; Gururaj et al., 2013; Muir et al., 2014). Phosphorylation/dephosphorylation of ORMs in S. cerevisiae presumably affects the higher order assembly of SPT to mediate flux through this enzyme for LCB synthesis (Breslow, 2013). In this sphingolipid homeostatic regulatory mechanism, the S. cerevisiae Orm1p and Orm2p are phosphorylated at their N termini by Ypk1, a TORC2-dependent protein kinase (Han et al., 2010; Roelants et al., 2011). The absence of this phosphorylation domain in mammalian and plant ORM homologs brings into question the nature of SPT reversible regulation by ORMs in other eukaryotic systems (Hjelmqvist et al., 2002).Sphingolipid synthesis also is mediated by the N-acylation of LCBs by ceramide synthases to form ceramides, the hydrophobic backbone of the major plant glycosphingolipids, glucosylceramide (GlcCer) and glycosyl inositolphosphoceramide (GIPC). Two functionally distinct classes of ceramide synthases occur in Arabidopsis, designated class I and class II (Chen et al., 2008). Class I ceramide synthase activity resulting from the Longevity Assurance Gene One Homolog2 (LOH2)-encoded ceramide synthase acylates, almost exclusively, LCBs containing two hydroxyl groups (dihydroxy LCBs) with 16:0-CoA to form C16 ceramides, which are used primarily for GlcCer synthesis (Markham et al., 2011; Ternes et al., 2011; Luttgeharm et al., 2016). Class II ceramide synthase activities resulting from the LOH1- and LOH3-encoded ceramide synthases are most active in the acylation of LCBs containing three hydroxyl groups (trihydroxy LCBs) with VLCFA-CoAs, including primarily C₂₄ and C₂₆ acyl-CoAs (Markham et al., 2011; Ternes et al., 2011; Luttgeharm et al., 2016). Class II (LOH1 and LOH3) ceramide synthase activity is essential for producing VLCFA-containing glycosphingolipids to support the growth of plant cells, whereas class I (LOH2) ceramide synthase activity is nonessential under normal growth conditions (Markham et al., 2011; Luttgeharm et al., 2015b). It was speculated recently that LOH2 ceramide synthase functions, in part, as a safety valve to acylate excess LCBs for glycosylation, resulting in a less cytotoxic form (Luttgeharm et al., 2015b; Msanne et al., 2015). Recent studies have shown that the Lag1/Lac1 components of the S. cerevisiae ceramide synthase are phosphorylated by Ypk1, and this phosphorylation stimulates ceramide synthase activity in response to heat and reduced intracellular sphingolipid levels (Muir et al., 2014). This finding points to possible coordinated regulation of ORM-mediated SPT and ceramide synthase activities to regulate sphingolipid homeostasis, which is likely more complicated in plants and mammals due to the occurrence of functionally distinct ceramide synthases in these systems (Stiban et al., 2010; Markham et al., 2011; Ternes et al., 2011; Luttgeharm et al., 2016).RNA interference (RNAi) suppression of ORM genes in rice (Oryza sativa) has been shown to affect pollen viability (Chueasiri et al., 2014), but no mechanistic characterization of ORM proteins in plants has yet to be reported. Here, we describe two Arabidopsis ORMs, AtORM1 and AtORM2, that suppress SPT activity through direct interaction with the LCB1/LCB2 heterodimer. We also show that strong up-regulation of AtORM expression impairs growth. In addition, up- or down-regulation of ORMs is shown to differentially affect the sensitivity of Arabidopsis to the PCD-inducing mycotoxin fumonisin B₁ (FB₁), a ceramide synthase inhibitor, and to differentially affect the activities of class I and II ceramide synthases as a possible additional mechanism for regulating sphingolipid homeostasis.     

CELLULOSE SYNTHASE-LIKE A2, a Glucomannan Synthase,Is Involved in Maintaining Adherent Mucilage Structure in Arabidopsis Seed

Mannans are hemicellulosic polysaccharides that are considered to have both structural and storage functions in the plant cell wall. However, it is not yet known how mannans function in Arabidopsis (Arabidopsis thaliana) seed mucilage. In this study, CELLULOSE SYNTHASE-LIKE A2 (CSLA2; At5g22740) expression was observed in several seed tissues, including the epidermal cells of developing seed coats. Disruption of CSLA2 resulted in thinner adherent mucilage halos, although the total amount of the adherent mucilage did not change compared with the wild type. This suggested that the adherent mucilage in the mutant was more compact compared with that of the wild type. In accordance with the role of CSLA2 in glucomannan synthesis, csla2-1 mucilage contained 30% less mannosyl and glucosyl content than did the wild type. No appreciable changes in the composition, structure, or macromolecular properties were observed for nonmannan polysaccharides in mutant mucilage. Biochemical analysis revealed that cellulose crystallinity was substantially reduced in csla2-1 mucilage; this was supported by the removal of most mucilage cellulose through treatment of csla2-1 seeds with endo-β-glucanase. Mutation in CSLA2 also resulted in altered spatial distribution of cellulose and an absence of birefringent cellulose microfibrils within the adherent mucilage. As with the observed changes in crystalline cellulose, the spatial distribution of pectin was also modified in csla2-1 mucilage. Taken together, our results demonstrate that glucomannans synthesized by CSLA2 are involved in modulating the structure of adherent mucilage, potentially through altering cellulose organization and crystallization.Mannan polysaccharides are a complex set of hemicellulosic cell wall polymers that are considered to have both structural and storage functions. Based on the particular chemical composition of the backbone and the side chains, mannan polysaccharides are classified into four types: pure mannan, glucomannan, galactomannan, and galactoglucomannan (Moreira and Filho, 2008; Wang et al., 2012; Pauly et al., 2013). Each of these polysaccharides is composed of a β-1,4-linked backbone containing Man or a combination of Glc and Man residues. In addition, the mannan backbone can be substituted with side chains of α-1,6-linked Gal residues. Mannan polysaccharides have been proposed to cross link with cellulose and other hemicelluloses via hydrogen bonds (Fry, 1986; Iiyama et al., 1994; Obel et al., 2007; Scheller and Ulvskov, 2010). Furthermore, it has been reported that heteromannans with different levels of substitution can interact with cellulose in diverse ways (Whitney et al., 1998). Together, these observations indicate the complexity of mannan polysaccharides in the context of cell wall architecture.CELLULOSE SYNTHASE-LIKE A (CSLA) enzymes have been shown to have mannan synthase activity in vitro. These enzymes polymerize the β-1,4-linked backbone of mannans or glucomannans, depending on the substrates (GDP-Man and/or GDP-Glc) provided (Richmond and Somerville, 2000; Liepman et al., 2005, 2007; Pauly et al., 2013). In Arabidopsis (Arabidopsis thaliana), nine CSLA genes have been identified; different CSLAs are responsible for the synthesis of different mannan types (Liepman et al., 2005, 2007). CSLA7 has mannan synthase activity in vitro (Liepman et al., 2005) and has been shown to synthesize stem glucomannan in vivo (Goubet et al., 2009). Disrupting the CSLA7 gene results in defective pollen growth and embryo lethality phenotypes in Arabidopsis, indicating structural or signaling functions of mannan polysaccharides during plant embryo development (Goubet et al., 2003). A mutation in CSLA9 results in the inhibition of Agrobacterium tumefaciens-mediated root transformation in the rat4 mutant (Zhu et al., 2003). CSLA2, CSLA3, and CSLA9 are proposed to play nonredundant roles in the biosynthesis of stem glucomannans, although mutations in CSLA2, CSLA3, or CSLA9 have no effect on stem development or strength (Goubet et al., 2009). All of the Arabidopsis CSLA proteins have been shown to be involved in the biosynthesis of mannan polysaccharides in the plant cell wall (Liepman et al., 2005, 2007), although the precise physiological functions of only CSLA7 and CSLA9 have been conclusively demonstrated.In Arabidopsis, when mature dry seeds are hydrated, gel-like mucilage is extruded to envelop the entire seed. Ruthenium red staining of Arabidopsis seeds reveals two different mucilage layers, termed the nonadherent and the adherent mucilage layers (Western et al., 2000; Macquet et al., 2007a). The outer, nonadherent mucilage is loosely attached and can be easily extracted by shaking seeds in water. Compositional and linkage analyses suggest that this layer is almost exclusively composed of unbranched rhamnogalacturonan I (RG-I) (>80% to 90%), with small amounts of branched RG-I, arabinoxylan, and high methylesterified homogalacturonan (HG). By contrast, the inner, adherent mucilage layer is tightly attached to the seed and can only be removed by strong acid or base treatment, or by enzymatic digestion (Macquet et al., 2007a; Huang et al., 2011; Walker et al., 2011). As with the nonadherent layer, adherent mucilage is also mainly composed of unbranched RG-I, but with small numbers of arabinan and galactan ramifications (Penfield et al., 2001; Willats et al., 2001; Dean et al., 2007; Macquet et al., 2007a, 2007b; Arsovski et al., 2009; Haughn and Western, 2012). There are also minor amounts of pectic HG in the adherent mucilage, with high methylesterified HG in the external domain compared with the internal domain of the adherent layer (Willats et al., 2001; Macquet et al., 2007a; Rautengarten et al., 2008; Sullivan et al., 2011; Saez-Aguayo et al., 2013). In addition, the adherent mucilage contains cellulose (Blake et al., 2006; Macquet et al., 2007a), which is entangled with RG-I and is thought to anchor the pectin-rich mucilage onto seeds (Macquet et al., 2007a; Harpaz-Saad et al., 2011, 2012; Mendu et al., 2011; Sullivan et al., 2011). As such, Arabidopsis seed mucilage is considered to be a useful model for investigating the biosynthesis of cell wall polysaccharides and how this process is regulated in vivo (Haughn and Western, 2012).Screening for altered seed coat mucilage has led to the identification of several genes encoding enzymes that are involved in the biosynthesis or modification of mucilage components. RHAMNOSE SYNTHASE2/MUCILAGE-MODIFIED4 (MUM4) is responsible for the synthesis of UDP-l-Rha (Usadel et al., 2004; Western et al., 2004; Oka et al., 2007). The putative GALACTURONSYLTRANSFERASE11 can potentially synthesize mucilage RG-I or HG pectin from UDP-d-GalUA (Caffall et al., 2009). GALACTURONSYLTRANSFERASE-LIKE5 appears to function in the regulation of the final size of the mucilage RG-I (Kong et al., 2011, 2013). Mutant seeds defective in these genes display reduced thickness of the extruded mucilage layer compared with wild-type Arabidopsis seeds.RG-I deposited in the apoplast of seed coat epidermal cells appears to be synthesized in a branched form that is subsequently modified by enzymes in the apoplast. MUM2 encodes a β-galactosidase that removes Gal residues from RG-I side chains (Dean et al., 2007; Macquet et al., 2007b). β-XYLOSIDASE1 encodes an α-l-arabinfuranosidase that removes Ara residues from RG-I side chains (Arsovski et al., 2009). Disruptions of these genes lead to defective hydration properties and affect the extrusion of mucilage. Furthermore, correct methylesterification of mucilage HG is also required for mucilage extrusion. HG is secreted into the wall in a high methylesterified form that can then be enzymatically demethylesterified by pectin methylesterases (PMEs; Bosch and Hepler, 2005). PECTIN METHYLESTERASE INHIBITOR6 (PMEI6) inhibits PME activities (Saez-Aguayo et al., 2013). The subtilisin-like Ser protease (SBT1.7) can activate other PME inhibitors, but not PMEI6 (Rautengarten et al., 2008; Saez-Aguayo et al., 2013). Disruption of either PMEI6 or SBT1.7 results in the delay of mucilage release.Although cellulose is present at low levels in adherent mucilage, it plays an important adhesive role for the attachment of mucilage pectin to the seed coat epidermal cells. The orientation and amount of pectin associated with the cellulose network is largely determined by cellulose conformation properties (Macquet et al., 2007a; Haughn and Western, 2012). Previous studies have demonstrated that CELLULOSE SYNTHASE A5 (CESA5) is required for the production of seed mucilage cellulose and the adherent mucilage in the cesa5 mutant can be easily extracted with water (Harpaz-Saad et al., 2011, 2012; Mendu et al., 2011; Sullivan et al., 2011).Despite all of these discoveries, large gaps remain in the current knowledge of the biosynthesis and functions of mucilage polysaccharides in seed coats. In this study, we show that CSLA2 is involved in the biosynthesis of mucilage glucomannan. Furthermore, we show that CSLA2 functions in the maintenance of the normal structure of the adherent mucilage layer through modifying the mucilage cellulose ultrastructure.