Plant Cell, Vol. 11, 707-726, April 1999, Copyright © 1999, American Society of Plant Physiologists
|
MEMBRANE STRUCTURE AND FUNCTION
|
The Dual Function of Sugar Carriers: Transport and Sugar Sensing
Sylvie Lalondea,
Eckhard Bolesb,
Hanjo Hellmanna,
Laurence Barkera,
John W. Patrickc,
Wolf B. Frommera, and
John M. Warda
a Center for Plant Molecular Biology (ZMBP), University of Tübingen, Auf der Morgenstelle 1, D-72076 Tübingen, Germany
b Institut für Mikrobiologie, Heinrich-Heine-Universität Düsseldorf, Universitätsstrasse 1, Geb. 26.12.01, D-40225 Düsseldorf, Germany
c Department of Biological Sciences, University of Newcastle, New South Wales 2308, Australia
Correspondence to:
Wolf B. Frommer, frommer{at}uni-tuebingen.de (E-mail), 49-7071-29-3287 (fax)
 |
INTRODUCTION |
|---|
Sucrose and its derivatives represent the major transport forms of photosynthetically assimilated carbon in plants. Sucrose synthesized in green leaves is exported via the phloem, the long-distance distribution network for assimilates, to supply nonphotosynthetic organs with energy and carbon resources. Sucrose not only functions as a transport metabolite but also contributes to the osmotic driving force for phloem translocation (mass flow) and serves as a signal to activate or repress specific genes in a variety of different tissues.
The long-distance transport of sucrose depends on a family of proteins that act as sucrose carriers. The analysis of transgenic plants impaired in sucrose transporter expression has demonstrated that sucrose transporter1 (SUT1) is essential for sucrose translocation in potato and tobacco. These results, together with the localization of SUT1 to sieve elements (SEs), indicate that phloem loading occurs in SEs by transmembrane uptake of sucrose directly from the apoplasm.
The sucrose transporters identified so far arise from a single gene family. Some of the newly identified members of the family are involved in specific functions, such as nutrition of developing seeds or pollen. Physiological and molecular studies show that sucrose transport is highly regulated at multiple levels of biological organization and in response to changing sucrose concentrations. Thus, one of the most exciting topics in the regulation of sucrose transport is signal perception. By analogy to yeast, in which members of the sugar transport family serve as sugar sensors, we propose that members of the plant sugar transporter family play a direct role in the signal transduction responsible for regulation of sugar transport and, thus, metabolism in general.
|
|
SUGAR TRANSPORT SYSTEMS IN HIGHER PLANTS |
|---|
Plant Anatomy: Vascular Tissues
Nonphotosynthetic tissues and organs, including the entire below-ground portion of the plant, need to be supplied with energy and fixed carbon. Sugars, synthesized in the mesophyll cells, serve as the major exported photosynthetic product. To accommodate long-distance transport of sugars from source (net exporting) to sink (net importing) organs, a vascular network—the phloem—has evolved in land plants. The most abundant compound in the phloem sap of most plant species is the disaccharide sucrose (Zimmermann and Ziegler 1975 
). For a minority of plant species, the principle translocated sugars fall into two main groups: the sugar alcohols (mannitol and sorbitol) and the raffinose series (raffinose, stachyose, and verbascose) (Zamski and Schnaffer 1996 ). In most cases in which such sugars predominate, however, sucrose is also present.
With the exception of a few well-studied species, our knowledge of phloem sap composition is limited to crude analyses (Zimmermann and Ziegler 1975 ) derived from stem incision experiments in which sugars present as storage compounds in stems may contaminate phloem sap samples. Therefore, accurate and less invasive techniques, such as in vivo NMR (Kockenberger et al. 1997 ) or positron-emitting tracer imaging system (Hayashi et al. 1997 ), will be required for a more accurate understanding of phloem sap composition.
The phloem of angiosperms consists of several types of cells that are closely associated with the xylem within the vascular bundle. The structure and development of the phloem has been reviewed recently (Sjolund 1997 ; Ward et al. 1998 ; Oparka and Turgeon 1999 , in this issue). The actual conduits in phloem consist of two ontogenetically related cell types: companion cells (CCs) and SEs. These two cells are highly modified and well interconnected by plasmodesmata. SEs, for instance, lose their nuclei, vacuoles, and many other organelles during maturation and form tubes of living cells connected by sieve pores. CCs, which are characterized by dense protoplasm, retain a nucleus and numerous mitochondria and are thought to provide functions essential for the survival of SEs via plasmodesmatal links (Lucas et al. 1993 ).
The complete physical pathway of sucrose transport from source to sink is not completely elucidated for any plant, and important differences may exist among species. From the point of sucrose synthesis in the mesophyll, the route to the SE involves several cell types: neighboring mesophyll cells, bundle-sheath cells, phloem parenchyma, and CCs. Cell-to-cell movement of sucrose is considered to occur via plasmodesmata from the point of synthesis up to the SE/CC complex which, in many species, is not well connected to the surrounding cells (Gamalei 1989 ). Two principal routes for the delivery of sucrose into the SE/CC complex have been proposed: (1) transporter-mediated export from mesophyll cells, apoplasmic diffusion through the cell-wall continuum, and subsequent carrier-mediated transport across the SE/CC plasma membrane; and (2) direct symplasmic cell-to-cell diffusion via plasmodesmata. The extent to which plants utilize either of these pathways and whether plants can switch between these pathways is a question currently under active investigation.
Apoplasmic Transport
Plants that primarily utilize an apoplasmic phloem loading mechanism require the transmembrane transport of sucrose and other solutes into the phloem. For such plants, as schematized in Figure 1, we can predict a minimum of five sucrose transport activities along the translocation path. In leaves, the first transport step must be release of sucrose into the cell wall directly from the mesophyll cell (Figure 1, transporter 1). Mechanistically, this first transporter could be a facilitator or an antiporter. Such sucrose efflux systems have been described biochemically (see Delrot 1989 ; Laloi et al. 1993 ). Subsequently, at least one transporter is required for uptake into the phloem (Figure 1, transporter 2). These loading processes are required at various stages of development, such as during germination for sugar export from leaves or for mobilization of stored carbon, events that might require distinct transporters. Reuptake of sucrose along the translocation pathway is necessary to allow solute exchange with the phloem, for example, in stems (Figure 1, transporter 3; Minchin and Thorpe 1987 ).
View larger version (32K):
[in this window]
[in a new window]
|
Figure 1.
Long-Distance Sugar Transport by the Phloem.
From its point of synthesis in the mesophyll, sucrose may be loaded into the SE/CC complex either through plasmodesmata or via the apoplasm. The apoplasmic loading mechanism requires sucrose export (1) from the mesophyll or the vascular parenchyma and reuptake (2) into the SE/CC complex. Hydrostatic pressure drives phloem sap movement toward sink tissue. Passive leakage can take place along the path (indicated by wavy arrows). Reuptake (3) also occurs along the path of the phloem. Apoplasmic phloem or post-phloem unloading necessitates a sucrose exporter at the sink tissue (4). Import of sucrose and other solutes into sink tissue may occur through plasmodesmata or sucrose transporters (5). In addition to plasmodesmal and transporter-mediated uptake, cells in the sink may take up sucrose, subsequent to its hydrolysis by an apoplasmic invertase, as hexoses (6). The vacuolar transport system could consist of a H+/sucrose antiporter for uptake and a uniporter for release.
|
|
In sink tissue, unloading can occur either by transmembrane export of sucrose (Figure 1, transporter 4) or through plasmodesmata. Sucrose efflux transporters involved in phloem unloading have been postulated to function as facilitators or as proton antiporters (Walker et al. 1995 ). Sucrose in the apoplast of sink tissue can be taken up directly (Figure 1, transporter 5) or through the hydrolysis of sucrose into glucose and fructose by invertase followed by hexose uptake (Figure 1, transporter 6).
Symplasmic Transport
Figure 1 also schematizes the possibility that each transport process outlined above could alternatively take place symplasmically. Based on the systemic movement of plant virus in phloem, plasmodesmal connections between the SE/CC complex and surrounding cells are present even in Solanaceous species which are classified as type 2a (van Bel and Gamalei 1992 ). To achieve the pressure difference required for mass flow, the plasmodesmata must normally be closed. The systemic spread of silencing signals via the phloem (Voinnet et al. 1998 ) indicates that plasmodesmata can be gated open by endogenous plant factors. Symplasmic routes have been identified in the root by determination of dye coupling by using small cytoplasmic fluorescent dyes (Wright and Oparka 1997 ).
Interestingly, despite the presence of plasmodesmata along the entire length of the transport path, efflux occurs only in restricted regions, such as the region behind the root tip. It is possible that plants may utilize different mechanisms of phloem unloading in different tissues or may even be able to switch between apoplasmic and symplasmic mechanisms depending on growth conditions. Besides these plasma membrane transporters, uptake and release systems are required for subcellular compartments, especially the vacuole, which serves as a transient buffer for sugars and other metabolites (Figure 1; Marty 1999 , in this issue).
The Physics of Mass Flow
Uptake of sucrose into the SEs is believed to increase the hydrostatic pressure difference between the ends of the phloem conduits so as to drive the mass flow movement of the phloem sap. The velocity of translocation in the phloem is relatively high, ranging from 0.5 to 3 m hr-1 (Kockenberger et al. 1997 ). The phloem, then, contains an osmotic pump whose efficacy rests primarily on the cooperative functions of three types of plasma membrane proteins: (1) H+/sucrose symporters that accumulate osmotically active sucrose to molar concentrations in the phloem; (2) H+-ATPases that provide the energy necessary for active transport; and (3) water transporters, potentially aquaporins, that take up water derived from the xylem (for a discussion of aquaporins, see Chrispeels et al. 1999 , in this issue; for a discussion of H+-ATPases, see Sze et al. 1999 , in this issue). Osmolytes other than sucrose, such as potassium and amino acids, also contribute to the driving force of the sap flow, and the simultaneous withdrawal of osmolytes and water at sink tissues further increases hydrostatic pressure differences. An essential prerequisite for the resulting osmotic pump is that the SE/CC complex be osmotically isolated from neighboring cells. Specifically, if phloem cells are connected to neighboring mesophyll cells, such intercellular connections must be tightly regulated. Differences in the connectivity among different cell types in leaves are in fact discernable and result in the classification of plants as either apoplasmic or symplasmic loaders (van Bel 1993 ). Furthermore, evidence for the regulated opening of plasmodesmata has been provided (Oparka and Prior 1992 ; Oparka et al. 1994 ).
Of the numerous components required for long-distance photoassimilate transport (Figure 1), only proton-coupled sucrose and monosaccharide uptake transporters have been identified at the molecular level. It is unknown whether sucrose release is also proton coupled or if the release carriers are related in sequence to the uptake transporters (Ward et al. 1998 ). The sucrose binding protein, which in many cases colocalizes with the sucrose transporter, might be responsible for facilitated diffusion of sucrose (Overvoorde et al. 1996 ; Harrington et al. 1997 ) and may represent a second sucrose transporter class.
|
|
SUGAR TRANSPORTERS |
|---|
Monosaccharide Transporters
Monosaccharide transport activities have also been identified in a variety of plant species (Maynard and Lucas 1982 ; Getz et al. 1987 ; Gogarten and Bentrup 1989 ; Tubbe and Buckhout 1992 ). Substrates that are efficiently transported include ß-D-glucose, 3-O-methyl ß-D-glucose, 2-deoxy ß-D-glucose, 
-D-mannose, and ß-D-fructose, whereas ß-L-glucose and ß-D-ribose are poor substrates for transport (Gogarten and Bentrup 1989 ). Plant monosaccharide transporters were first cloned from Chlorella kessleri by exploiting their rapid induction following the addition of hexoses to the growth medium. Specifically, differential screening of cDNAs from autotrophic versus heterotrophic cells enabled the cloning of the first plant hexose transporter gene (HUP1; Sauer and Tanner 1989 ). In contrast to the yeast hexose transporters (HXTs), which function as uniporters, the C. kessleri hexose transporter is a symporter (Sauer et al. 1990a ; Aoshima et al. 1993 ). Despite this difference in the transport mechanism, yeast and C. kessleri transporter genes are homologous, encoding proteins composed of 12 putative membrane-spanning domains. At least three hexose transporters are known to exist in C. kessleri (Table 1). The functional proof that these genes encode hexose transporters derives from heterologous expression of the plant genes in yeast (Sauer et al. 1990a ).
The cloning of higher plant monosaccharide transporters was accomplished by heterologous hybridization (Sauer et al. 1990b ). The monosaccharide transporter family in Arabidopsis, as shown in Table 1, contains in excess of 26 genes, and multiple genes have been isolated from other species. STP1 was characterized by heterologous expression in Xenopus laevis oocytes and shown to function as a proton symporter (Aoshima et al. 1993 ; Boorer et al. 1994 ). The expression patterns of various plant monosaccharide transporters suggest that these membrane proteins function in hexose uptake in sink tissues (Sauer and Stadler 1993 ). Analysis of expression also shows that plant monosaccharide transporters are highly regulated, such as in response to pathogen infection or after wounding (Truernit et al. 1996 ), thus allowing flexible reallocation of fixed carbon. The expression of monosaccharide transporters in sink tissue also supports an apoplasmic mechanism for phloem or postphloem unloading (Sauer and Stadler 1993 ). In this way, plant sink tissues may acquire hexose in a manner similar to that employed by yeast that use an extracellular invertase to hydrolyze external sucrose. One requirement for this mechanism in plants is the coexpression of plant invertases and monosaccharide transporters (Ehness and Roitsch 1997 ).
The sugar permease family in yeast contains 34 genes (Andre 1995 ; Nelissen et al. 1997 ; reviewed in Boles and Hollenberg 1997 ), with 20 members belonging to the HXT subfamily (HXT1-17, GAL2, SNF3, and RGT2). The individual HXTs differ considerably with respect to their kinetic properties, with Km values for glucose uptake ranging from 1 mM for HXT7 to 100 mM for HXT1 (Reifenberger et al. 1997 ). The expression of HXT genes is regulated by glucose concentration, and the transporter homologs SNF3 and RGT2 function as sensors (see below).
Sucrose Transporters and SUT Genes
Sucrose transport activities have been described in a wide variety of systems (Maynard and Lucas 1982 ; Giaquinta 1983 ; reviewed in Bush 1993 ; Ward et al. 1998 ). To clone plant sucrose transporters, a yeast strain was generated that was unable to hydrolyze extracellular sucrose but was capable of metabolizing internal sucrose due to the presence of a plant-derived sucrose synthase activity (Figure 2). The strain was, furthermore, deficient in maltose utilization, a safeguard against side activities of the yeast-endogenous maltose transport systems. The resulting yeast cells grow inefficiently on media containing 0.5% sucrose as the sole carbon source and were consequently used to clone plant sucrose transporters from spinach and potato by functional complementation (Riesmeier et al. 1992 , Riesmeier et al. 1993 ).
View larger version (21K):
[in this window]
[in a new window]
|
Figure 2.
A Yeast System for Functional Cloning of Sucrose Transporters.
Wild-type yeast (Saccharomyces cerevisiae) utilizes sucrose primarily through activities of an extracellular invertase and hexose transporters in the plasma membrane (left). A yeast strain (SUSY7; Riesmeier et al. 1992 ) was constructed in which the cytosolic and extracellular invertases were genetically knocked out (middle). A plant sucrose synthase expressed in the cytosol allows growth on sucrose as the sole carbon source only if a sucrose transporter is expressed in the plasma membrane (right). The ability (+) or inability (-) of the yeast strains to grow on sucrose or glucose is shown. fruc, fructose; gluc, glucose; INV, invertase; suc, sucrose; SUC2, invertase gene; SUSY, sucrose synthase; SUT1, sucrose transporter 1 gene.
|
|
Detailed transport studies using radioactive tracers have subsequently allowed determination of kinetic properties, pH optima, inhibitor sensitivity, and substrate specificity of various SUT genes. All plant sucrose transporters identified so far are energy dependent and sensitive to protonophores, indicating that they function as proton symporters. The Km for sucrose, in all cases, was found to be in the range of 1 mM (Riesmeier et al. 1993 ; Lemoine et al. 1996 ). Similar results were obtained for the Arabidopsis sucrose transporters SUC1 and SUC2.
The SUT genes encode highly hydrophobic proteins. They consist of 12 membrane-spanning domains and are distantly related to the hexose transporter family found in many organisms, such as yeast and plants (reviewed in Ward et al., 1997; Rentsch et al. 1998 ). As for the monosaccharide transporters, sucrose transporters have been characterized by the two-electrode voltage clamp method in X. laevis oocytes (Boorer et al. 1996 ; Zhou et al. 1997 ). The stoichiometry of H+/sucrose cotransport was determined to be 1:1, consistent with stoichiometric estimates obtained earlier in plasma membrane vesicles from sugar beet leaves (Bush 1990 ; Slone et al. 1991 ). The SUT family thus corresponds to the protonophore-sensitive high-affinity component of sucrose-uptake kinetics measured in plants (Maynard and Lucas 1982 ).
|
|
CELLULAR LOCALIZATION OF SUT1 IN THE PHLOEM: CCs AND SEs |
|---|
Based on the transport mechanism (H+ symport) of SUT1, it was expected that the transporter would be involved in phloem loading and thus should be present at the plasma membrane of SE/CC complexes. In situ hybridization indeed demonstrates that SUT1 transcripts are phloem associated (Riesmeier et al. 1993 ), and the promoters of tomato SUT1 and Arabidopsis SUC2 genes direct the expression of reporter genes in leaf, stem, and root phloem (Truernit and Sauer 1995 ; A. Weise, B. Hirner, J.M. Ward, and W.B. Frommer, unpublished results). Hence, SUT1 (or SUC2 from Arabidopsis) might play a role not only in phloem loading but also in retrieval of sucrose leaking from sieve tubes along the translocation pathway (Figure 1, transporter 3).
To determine expression at the cellular level, immunolocalization studies have been applied to five species. In Plantago major and Arabidopsis, immunofluorescence with specific antibodies detects SUC2 in CCs (Stadler et al. 1995 ; Stadler and Sauer 1996 ). By contrast, immunolocalization using immunofluorescence and silver-enhanced immunogold staining assign SUT1 to the plasma membranes of enucleate SEs of tobacco, potato, and tomato (Kuhn et al. 1997 ). The differences in sucrose transporter localization observed in Arabidopsis and P. major compared with tomato, potato, and tobacco may be due to differences in loading mechanisms. However, the transporters studied in Arabidopsis and P. major do not appear to be orthologs of SUT1, which may remain to be identified in those species. This would be consistent with a stepwise manner of sucrose loading, whereby different carriers operate in CCs (SUC2) as opposed to SEs (SUT1); such a scenario is suggested by physiological analyses (Roeckl 1949 ).
In situ hybridization experiments at the electron microscope level corroborate the localization of SUT proteins to SEs in a remarkable manner. Specifically, Solanaceous SUT1 mRNA localizes mainly to SEs, primarily at the orifices of plasmodesmata (Kuhn et al. 1997 ). In addition, antisense inhibition of SUT1 expression using a CC-specific promoter (from the RolC gene) produces strong phenotypic effects in transgenic plants due to inhibition of sucrose export from leaves (Kuhn et al. 1996 ). These results indicate that transcription of the SUT1 gene occurs in CCs and, in conjunction with high turnover rates of mRNA and protein, provide strong evidence for targeting of plant endogenous mRNA, and potentially SUT1 protein, through the plasmodesmata that interconnect CCs and SEs.
These data may seem surprising at first sight; however, trafficking of RNA is known to occur in Drosophila and X. laevis during oogenesis and in neurons (St. Johnston 1995 ). In maize, moreover, mRNA and protein produced from the KNOTTED-1 gene are transported through plasmodesmata in apical meristems (Lucas et al. 1995 ). Similarly, microinjection studies in Cucurbita maxima suggest that an RNA binding protein identified in phloem sap of that species is capable of trafficking a number of mRNAs intercellularly, including the SUT1 mRNA (Xoconostle-Cazares et al. 1999 ). The transport of macromolecules such as mRNAs and proteins through the microchannels of plasmodesmata requires mechanisms of unfolding (Kragler et al. 1998 ; see also Lazarowitz and Beachy 1999 , in this issue). The function of plasmodesmata thus seems, in this respect and many others, to be highly similar to protein import into organelles such as plastids (see Keegstra and Cline 1999 , in this issue), mitochondria, and the endoplasmic reticulum lumen (Vitale and Denecke 1999 , in this issue).
As schematized in Figure 3, two potential pathways for the targeting of SUT1 can be postulated: either (1) mRNA is guided as part of a nucleoprotein complex along the cytoskeleton through the phloem plasmodesmata for subsequent translation; or (2) translation is effected in the CCs and the protein is deposited within the SE at the plasmodesmatal orifices via the endomembrane system (Overall and Blackman 1996 ).
View larger version (151K):
[in this window]
[in a new window]
|
Figure 3.
SUT1 Biosynthesis.
In Solanaceous species, SUT1 protein is located in the SE plasma membrane (PM), while transcription occurs in CCs. There are two possible pathways: SUT1 mRNA may be delivered through plasmodesmata (pd) by an RNA transport mechanism. Alternatively, SUT1 translation may take place in CCs and delivery to SEs may occur via the plasma membrane or endosomal membranes that are continuous through plasmodesmata. These pathways are not exclusive and could function in parallel. ER, endoplasmic reticulum; sER, sieve element reticulum.
|
|
Other phloem proteins alter the size exclusion limit for plasmodesmata and have been shown to move from cell to cell (Balachandran et al. 1997 ; Ishiwatari et al. 1998 ; see Lazarowitz and Beachy 1999 , in this issue). Various proteins have been identified in the phloem sap, some of which were found to be specific for SEs, namely, ß-amylase, glutaredoxin, thioredoxin, cyclophilin, protein kinases, ubiquitin, and P-protein PP2a (Schobert et al. 1995 ; Wang et al. 1995 ; Sjolund 1997 ; Szederkenyi et al. 1997 ). The role of most of these proteins in the phloem is unknown, and their origins and biosynthesis are not fully understood. The stability of proteins in the phloem also is generally unknown. The presence of ubiquitin in phloem sap may indicate that protein turnover is occurring. Further research is needed to determine both protein turnover rates in the phloem sap and protein import rates.
Mature SEs are living cells despite the fact that they lack a nucleus and many other organelles (Sjolund 1997 ). SEs are thought to be dependent on CCs for many cellular functions, but the extent of this dependence cannot be determined until the actual metabolic capabilities of SEs are better understood. For example, it is not clear whether SEs contain plasma membrane proton pumps. In CCs, on the other hand, the expression of the H+-ATPase–encoding AtAHA3 gene has been demonstrated (DeWitt and Sussman 1995 ; see Sze et al. 1999 , in this issue). If SEs do in fact contain an H+-ATPase, then the supply of ATP to SEs through plasmodesmata would be presumably important. Consistent with this, phloem sap has been shown to contain concentrations of ATP in the range of 1 mM (Kluge et al. 1970 ). Additionally, the biogenesis of any ATPase in SEs may be dependent on a supply of mRNA or protein from CCs, by analogy to SUT1 biogenesis (Kuhn et al. 1997 ). On the other hand, if SEs do not contain a proton pump or if the proton motive force to drive phloem loading is generated only in CCs, then plasmodesmatal connections are important for the propagation of this driving force (for both electrical coupling and a pathway for protons) into SEs (van Bel 1996 ). In addition, trafficking of proteins, nucleic acids, and other molecules through plasmodesmata connecting CCs and SEs is important for long-distance signaling and to coordinate development at the whole-plant level in response to environmental factors, nutrients, and water availability (reviewed in Mezitt and Lucas 1996 ).
|
|
IN VIVO EVIDENCE OF SUT1 TRANSPORTER FUNCTION |
|---|
If sucrose transport mediated by SUT1 is essential for phloem loading, a reduction in transport activity should affect carbon partitioning and photosynthesis. In SUT1 antisense plants (Riesmeier et al. 1994 ; Kuhn et al. 1996 ), leaf sucrose and starch content is five- to 10-fold higher than in the wild type, and the increase in hexose content is even greater. In addition, SUT1 antisense plants grow at markedly retarded rates, producing crinkled leaves that exhibit chlorosis and accumulation of anthocyanins. Development of the phenotype depends on the length of the photoperiod and light intensity (Kuhn et al. 1996 ).
A similar accumulation of soluble carbohydrates occurs when petioles of potato leaves are cold girdled so as to block phloem translocation (Krapp et al. 1993 ). Enhanced partitioning into insoluble carbohydrates is also found in a number of studies in which heat girdling is used (Grusak et al. 1990 , and references therein). A direct comparison of antisense repression and cold girdling at the ultrastructural level demonstrates that both treatments equally evoke the accumulation of assimilates in all leaf tissues up to the SE/CC complex. However, microscopy reveals that antisense inhibition of loading produces a persistently high sugar level throughout the leaf, whereas cold girdling leads to localized patches containing high sugar and lipid levels (Schulz et al. 1998 ).
Efflux measurements of carbohydrates from excised leaves of antisense plants show strong reduction in phloem transport (Riesmeier et al. 1994 ). With less carbohydrate transport to sinks, the plants have reduced root growth and tuber yield, phenotypic qualities also observed in transgenic potato plants in which apoplasmic loading is prevented by the overexpression of a yeast invertase in cell walls of leaves (Heinecke et al. 1992 ; Riesmeier et al. 1994 ). In tobacco, antisense repression of SUT1 also leads to dramatic growth retardation and accumulation of carbohydrates in leaves (Burkle et al. 1998 ). The export of recently fixed 14CO2 is blocked to almost nondetectable levels even in plants inhibited to an intermediate degree, and stronger inhibition was found to be lethal. Thus, proton-coupled carriers seem to be indispensable for phloem loading at least in Solanaceous species.
Comparable effects were observed in potato plants in which SUT1 was expressed in antisense orientation under control of the CC-specific RolC promoter (Kuhn et al. 1996 ). It was, however, not possible to estimate the control coefficient of SUT1 because the actual amount of SUT1 protein and sucrose transport activity in the phloem was masked by low levels of SUT1 expression presumably in mesophyll cells (Lemoine et al. 1996 ). The effects observed are therefore in agreement with expected results that SUT1 transcription in CCs is essential for phloem loading.
It remains unclear whether antisense repression also affects other members of the SUT gene family. A complete analysis of the role of individual members of the gene family will require approaches such as the use of "knockout" mutants in Arabidopsis. The potential of this approach for studying transport has been elegantly demonstrated in the case of potassium channels (Krysan et al. 1996 ; Gaymard et al. 1998 ; Hirsch et al. 1998 ; see also Chrispeels et al. 1999 , in this issue). Identification of "knockout" mutants in sugar transport is in progress in several research laboratories.
|
|
SUCROSE TRANSPORTERS IN PHLOEM AND POST-PHLOEM UNLOADING |
|---|
As detailed above, sucrose and hexoses have to be imported into sink tissues in roots, pollen, seeds, and elsewhere. SUT1/SUC2 expression has indeed been found in these tissues (Riesmeier et al. 1993 ; Truernit and Sauer 1995 ). However, the direct function of such phloem-associated H+/sucrose symporters in sink tissues remains unclear. In several plant species, such as tomato, tobacco, potato, Arabidopsis, P. major, Ricinus communis, Vicia faba, carrot, rice, and pea, additional sucrose transporter genes have been identified, as listed in Table 2 (Gahrtz et al. 1994 , Gahrtz et al. 1996 ; Sauer and Stolz 1994 ; Harrington et al. 1997 ; Hirose et al. 1997 ; Burkle et al. 1998 ; Shakya and Sturm 1998 ; Tegeder et al. 1999 ). Several of these genes/proteins show highly specific expression patterns. For example, in Plantago, SUC1 is expressed in young ovules (Gahrtz et al. 1996 ), and sucrose transporter transcripts can be detected in the transfer cells of cotyledons from Vicia seeds and pea (Table 2; Harrington et al. 1997 ; Weber et al. 1997 ; Tegeder et al. 1999 ). Interestingly, these carriers are also expressed in source leaves, indicating a dual function in both phloem loading in leaves and in seed import.
Other members of the SUT family are required in unloading zones, such as pollen, ovules, and roots. A pollen-specific sucrose transporter has been identified in tobacco (R. Lemoine, L. Barker, L. Bürkle, C. Kühn, M. Regnacq, C. Gaillard, S. Delrot, and W.B. Frommer, unpublished data). Within sink tissues, sucrose transporters could function in direct transport into sink cells or in sucrose retrieval. This latter function could control sink strength. Osmotic regulation, especially that involved in regulation of phloem or postphloem unloading, has been discussed in detail by Patrick 1997 . Carriers in source tissue for efflux from mesophyll, carriers in sink tissues for efflux from phloem and postphloem, and carriers involved in transient storage in vacuoles have yet to be identified (Figure 1).
|
|
SENSING MECHANISMS IN SUGAR TRANSPORT |
|---|
A common misconception is that transport processes, especially sugar transport in higher plants, are constitutive and that biosynthetic activities in the source and catabolic activities in the sink are the key factors controlling allocation of carbohydrates. Carbohydrate export rate is increased 10-fold in plants overexpressing pyruvate decarboxylase (Tadege et al. 1998 ), indicating a large potential for upregulation. Transporters, however, are located in strategic positions along metabolic pathways, and thus an effective regulatory mechanism would be to control uptake and efflux directly. Indeed, there is strong evidence that sugar transport regulates the distribution of assimilates within the plant through various macromolecular signaling events.
In the simplest sensing scenario, cells would contain only an intracellular receptor for a sugar metabolite. Such cells, whether located in source or sink tissues, could modulate metabolic processes, such as photosynthesis or carrier gene expression. However, due to intracellular metabolism, the effector cells involved in these processes must be able to differentiate between biosynthesis and transport, and therefore both intra- and extracellular concentrations of sugars need to be sensed. Furthermore, provided that the cell has a spectrum of carriers of varying affinity and capacity for sugar, extracellular sensors, as shown in Figure 4, can adjust sugar uptake to match requirements, for example, by inducing high-affinity uptake systems at low external concentration.
View larger version (26K):
[in this window]
[in a new window]
|
Figure 4.
Model for Metabolite Sensing by a Combination of Internal and Membrane-Bound Receptors.
In addition to an internal receptor (see text), membrane-bound receptors are used to sense external sugar concentrations. Such sensors trigger a signaling cascade regulating transporter biogenesis and insertion or degradation of proteins at the plasma membrane. In this scenario, low sugar concentrations activate a sensor to induce sugar transporter genes (preferentially high affinity). This enables more efficient sugar uptake. If, on the other hand, internal sugar concentration becomes too high, the intracellular sensor may either repress transporter transcription or trigger inactivation via endocytosis and degradation of the transporter, thus reducing the influx. Induction of low-affinity/high-capacity transporters could increase uptake at higher external concentrations.
|
|
In principle, multiple sensors could be utilized, some for high-affinity responses and others for adaptation to high-flux requirements. Because the plasma membrane has a limited capacity for protein content, this signal may at the same time lead to an increase in the turnover of low-affinity systems. Increased turnover of transporters can also be controlled via an internal sensor, thus decreasing import if intracellular concentrations exceed the requirements.
Although such regulatory networks could also be effective in unicellular organisms like yeast or algae, intercellular transport in higher plants is more complicated. This is because at least two cellular activities—export from one cell and import into the adjacent cell—have to be integrated. Little is known about the molecular mechanisms involved in sugar sensing in relation to transport in higher plants. Yeast could thus serve as a model to help uncover the regulatory networks in higher plants.
|
|
YEAST AS A MODEL OF SUGAR SIGNALING |
|---|
The yeast Saccharomyces cerevisiae contains a large spectrum of >200 integral membrane proteins, many of which are clearly involved in transmembrane solute transport. For example, yeast contains >20 permeases for amino acid transport (Andre 1995 ; Nelissen et al. 1997 ) and >20 permeases for sugar transport (Andre 1995 ; Boles and Hollenberg 1997 ). The redundancy of transport systems suggests that complex regulatory networks are absolutely necessary to control the uptake of nutrients in response to a rapidly changing external environment. As shown in Figure 5, yeast has developed a two-pronged regulatory system to ensure coordination between the supply of sugars from the environment and the enzymatic machinery of the cells: (1) the extracellular concentration of sugars is sensed and sugar transport activity is regulated accordingly; (2) sugar transport activity determines the flux of sugars into the cell, subsequently generating intracellular signals for further regulatory processes.
View larger version (39K):
[in this window]
[in a new window]
|
Figure 5.
Sugar Signal Transduction in Saccharomyces cerevisiae.
HXT-type transporters mediate glucose uptake. Hexokinase PII (HXK2) functions in cytosolic sugar signaling, resulting in transcriptional regulation of enzymes and transporters necessary for sugar utilization. The SNF3 (high affinity) and RGT2 (low affinity) glucose receptors sense the external glucose concentration. These are homologous to HXT-type transporters but contain a C-terminal extension that functions in signaling. The SNF3 and RGT2 signaling pathways affect transcriptional regulation of high-affinity and low-affinity/high-capacity hexose transporter genes, respectively. See text for details.
|
|
In S. cerevisiae, multiple transport systems for glucose are regulated at the transcriptional level in response to the external concentration of glucose. For example, HXT2 and HXT7 serve as high-affinity glucose transporters and are induced only by low levels of glucose but repressed at high levels, whereas HXT1 functions as a low-affinity transporter and is induced only by high concentrations of glucose (Ozcan and Johnston 1995 ; Boles and Hollenberg 1997 ). Consequently, sensors of extracellular glucose that respond not only to the kind of carbon source in the medium but also to its concentration are required. Once inside the cell, glucose is phosphorylated by three different kinases: HXK1, HXK2, and GLK1. Finally, it is converted through the glycolytic pathway mainly into ethanol. By contrast to the uptake of glucose, the regulation of the intracellular glucose concentration, phosphorylation, and subsequently, the regulation of flux through glycolysis must be controlled by intracellular signals (Boles et al. 1997 ). Furthermore, the expression of genes for the utilization of alternative carbon sources like sucrose or galactose, and genes involved in gluconeogenesis, must be shut off in the presence of sufficient amounts of the preferred carbon source, glucose. This is achieved through a mechanism known as glucose (or carbon) catabolite repression (Ronne 1995 ; Gancedo 1998 ).
The glucose signal that triggers induction of hexose transporter genes is generated by the hexose sensors SNF3 and RGT2. On the other hand, the signal that triggers glucose repression is somehow connected to the kinase activity of HXK2 (Ma et al. 1989 ; Rose et al. 1991 ). In principle, there are two possibilities for sensory proteins to detect signaling molecules. First, sensors might act as receptors, binding the triggering molecule (e.g., glucose) and transducing the signal via other proteins. Second, sensors might behave like enzymes or transporters and undergo structural changes so as to monitor the presence or absence of the triggering compound directly. Flux measurements by such a sensor protein would involve the recognition of the velocity of the enzymatic reaction as the ratio of active to free enzyme.
|
|
SUGAR SENSORS: A NEW PERSPECTIVE OF TRANSPORTERS |
|---|
Cell Surface Sugar Sensors
In Escherichia coli, it has long been known that glucose sensing is mediated through glucose phosphorylation during transport by the phosphotransferase system (Postma et al. 1993 ). However, no related proteins have yet been detected at the plasma membrane of eukaryotic cells. Glucose sensing in yeast seems to involve plasma membrane proteins that resemble glucose transporters but additionally possess large cytoplasmic signaling domains (Ozcan et al. 1996a ). A similar mechanism has been discovered recently for amino acid sensing (Didion et al. 1998 ; Iraqui et al. 1999 ) and might be a general phenomenon not restricted to yeast.
In yeast, SNF3 appears to be a sensor of low levels of glucose and mainly regulates expression of high-affinity glucose transporters, whereas RGT2 appears to be a sensor of high glucose concentrations that regulates expression of low-affinity glucose transporters. Additionally, SNF3 is required at high levels of glucose for repression of the high-affinity transporters HXT2, HXT6, and HXT7 (Liang and Gaber 1996 ; Vagnoli et al. 1998 ). Dominant mutations in both RGT2 and SNF3 that lead to the generation of signals in the absence of glucose have been identified (Ozcan et al. 1996a ). Despite the homology of SNF3 and RGT2 to glucose permeases, these transmembrane proteins do not seem to be able to mediate significant glucose transport (Liang and Gaber 1996 ; Ozcan et al. 1998 ). The large C-terminal extension of SNF3 (303 amino acids) contains two nearly identical repeats of 25 amino acids. One of these repeats is also present in the 218–amino acid C-terminal extension of RGT2 and in RAG4 (GenBank accession number
Y14849) from Kluyveromyces lactis, which contains a 250–amino acid C-terminal extension and controls the expression of the low-affinity glucose transporter RAG1 (Chen et al. 1992 ). In Neurospora crassa, the glucose sensor RCO3 contains a 119–amino acid C-terminal extension that is dissimilar to that of SNF3, RGT2, and RAG4 (Madi et al. 1997 ).
From mutational analyses, yeast glucose sensors appear to function as two interacting domains (Ozcan et al. 1998 ; Vagnoli et al. 1998 ): the C-terminal extensions that are required for the transmission of the glucose signal, and the membrane-spanning domain necessary for glucose recognition. Because the glucose transporters HXT1 and HXT2 can be converted into glucose sensors by fusion to the SNF3 C terminus, it is tempting to speculate that the glucose sensors act as glucose transporters with a very low glucose transport capacity (i.e., too low to support growth of mutants that lack glucose transporters), which transduce the glucose signal by a conformational shift during glucose transport.
Another protein (SSY1) that appears to function as an amino acid sensor has recently been identified in yeast. SSY1 shows the features typical of the SNF3/RGT2 pair: a transmembrane domain related to amino acid permeases, a long cytoplasmic extension (in this case located at the N terminus of the protein), a low transcription rate, and a low coding probability (Andre 1995 ; Didion et al. 1998 ; Iraqui et al. 1999 ). SSY1, rather than being a transporter, controls the transcription of genes encoding amino acid permeases and seems to respond to changes in the extracellular concentration of various amino acids (Didion et al. 1998 ; Iraqui et al. 1999 ). The amino acid sequence of the cytoplasmic extension of SSY1 shows no significant similarities to those of SNF3 and RGT2.
The only intermediate components so far known to be involved in glucose-induced signal transduction in yeast are the transcription factor RGT1 (Ozcan et al. 1996a ) and SCFGrr1, a ubiquitin protein ligase complex including the F-box protein GRR1, SKP1, and CDC53 (Li and Johnston 1997 ; Skowyra et al. 1997 ). RGT1 is a zinc cluster protein that binds directly to promoters of the HXT genes (Ozcan et al. 1996b ). It acts as a repressor of HXT genes in cells growing without glucose and as a transcriptional activator of HXT1 in cells growing on high levels of glucose. Repression of transcription by RGT1 is mediated by the SSN6-TUP1 complex, whereas RGT1-mediated induction is independent of these proteins (Figure 5).
The SCFGrr1 protein complex is required for regulation of RGT1 activity (Ozcan and Johnston 1995 ) mediating the signal generated by SNF3 to inhibit the RGT1 repressor function in response to low levels of glucose. It is also required for conversion of RGT1 into an activator triggered by RGT2 in the presence of high levels of glucose. Ubiquitination of RGT1 or its regulator may target it for protein degradation by the proteasome or directly affect its function. SCFGrr1 complex may be directly stimulated by glucose (Li and Johnston 1997 ) or may depend on glucose-activated kinase because target proteins must be phosphorylated to interact with GRR1 (Skowyra et al. 1997 ).
A third sensor protein, MEP2, was recently found in S. cerevisiae to be required for pseudohyphal differentiation in response to ammonium limitation (Lorenz and Heitman 1998 ). Under such conditions, diploid yeast cells differentiate into a filamentous, pseudohyphal growth form. MEP2 belongs to a family of three closely related ammonium permeases (Marini et al. 1997 ). However, unlike the other two members of this transporter family, MEP2 serves as both an ammonium transporter and a component of an ammonium sensor. mep2 mutant strains have no defects in growth rates or ammonium uptake, but no longer form pseudohyphal filaments on media containing limiting amounts of ammonium. Unlike the glucose sensor proteins, however, MEP2 is able to transport its substrate; and as compared with the other members of the ammonium permease family, MEP2 contains no significant amino acid extensions on either its N or C terminus.
Intracellular Sugar Sensing: Enzymes as Sensors
In the triggering reaction of glucose repression in yeast, HXK2 seems to play an important and probably more direct role. Mutations in HXK2 abolish glucose repression of invertase and other glucose-repressed genes (Entian 1980 ). The catalytic and regulatory functions of HXK2 are inseparable from glucose repression and inversely correlated to its sugar-phosphorylating activity (Ma et al. 1989 ; Rose et al. 1991 ). HXK2 exists in a dimeric–monomeric equilibrium that is affected by phosphorylation (Randez-Gil et al. 1998 ). In vivo, dephosphorylation of HXK2 is promoted upon addition of glucose and is dependent on protein phosphatase 1 (CID1/GLC7). A protein kinase involved in phosphorylation of HXK2 has not been found, and it is possible that phosphorylation of HXK2 could result from substrate-induced autophosphorylation (Fernandez et al. 1988 ). It has been reported that HXK2 has a weak protein kinase activity (Herrero et al. 1989 ; see above). Moreover, a hxk2 mutant that is not able to undergo phosphorylation can no longer transduce the glucose repression signal (Randez-Gil et al. 1998 ). These properties might classify HXK2 as an enzyme that additionally, through its enzymatic function, can act as a sensor protein. However, the actual mechanism representing the on–off switch of the glucose signal is not understood. The glucose repression signal finally inhibits the protein kinase SNF1/CAT1, a central element in the regulatory process (Figure 5) and highly conserved in eukaryotes. Glucose inhibition of SNF1 depends on protein phosphatase 1 (GLC7) and its targeting subunit REG1 (Ludin et al. 1998 ). In the absence of glucose, SNF1 relieves repression by the MIG-SSN6-TUP1 complexes (De Vit et al. 1997 ) but is also required for the operation of other transcription factors (Gancedo 1998 ).
In accordance with HXK2 being a sensor for glucose, the triggering of glucose repression in yeast is dependent on glucose uptake. However, the repression is not dependent on a specific hexose transporter protein; the glucose repression signal rather correlates with the extent of glucose influx into cells (Reifenberger et al. 1997 ). These observations fit very well with the two-pronged regulatory model described above: first, glucose transport activity is adjusted to the extracellular glucose concentration via SNF3 and RGT2; second, glucose transport activity limits provision of intracellular glucose, which then might act as a substrate for HXK2 as well as a global glucose signal (Figure 5). In addition, cross-talk between the two different processes might ensure feedback coordination. Such a function has recently been assigned to HXK2 which, in addition to carbon catabolite repression, seems also to be involved in glucose induction of HXT gene expression (Randez-Gil et al. 1998 ).
|
|
SUGAR-MEDIATED REGULATION OF SINK AND SOURCE GENES |
|---|
TOP
INTRODUCTION
SUGAR TRANSPORT SYSTEMS IN...