SEARCH

SEARCH BY CITATION

Keywords:

  • Trehalose;
  • Trehalose-6-phosphate synthase;
  • Arabidopsis;
  • embryo maturation

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Despite the recent discovery that trehalose synthesis is widespread in higher plants very little is known about its physiological significance. Here we report on an Arabidopsis mutant (tps1), disrupted in a gene encoding the first enzyme of trehalose biosynthesis (trehalose-6-phosphate synthase). The tps1 mutant is a recessive embryo lethal. Embryo morphogenesis is normal but development is retarded and stalls early in the phase of cell expansion and storage reserve accumulation. TPS1 is transiently up-regulated at this same developmental stage and is required for the full expression of seed maturation marker genes (2S2 and OLEOSN2). Sucrose levels also increase rapidly in seeds during the onset of cell expansion. In Saccharomyces cerevisiae trehalose-6-phosphate (T-6-P) is required to regulate sugar influx into glycolysis via the inhibition of hexokinase and a deficiency in TPS1 prevents growth on sugars (Thevelein and Hohmann, 1995). The growth of Arabidopsis tps1–1 embryos can be partially rescued in vitro by reducing the sucrose level. However, T-6-P is not an inhibitor of AtHXK1 or AtHXK2. Nor does reducing hexokinase activity rescue tps1–1 embryo growth. Our data establish for the first time that an enzyme of trehalose metabolism is essential in plants and is implicated in the regulation of sugar metabolism/embryo development via a different mechanism to that reported in S. cerevisiae.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Sugars play a pivotal role in plants acting both as carbon currency and metabolic signals, controlling many aspects of plant growth and development in response to changes in nutritional status (Koch, 1996; Smeekens, 2000). The recent discovery that genes encoding enzymes responsible for trehalose (α-d-glucopyranosyl-[1,1]-α-d-glucopyranoside) biosynthesis exist in many higher plants has fuelled interest in the physiological role of this pathway (Goddijn and Smeekens, 1998; Goddijn and van Dun, 1999). Trehalose is a non-reducing disaccharide sugar that is widely distributed in nature (Elbein, 1974). The biosynthesis of trehalose involves the formation of trehalose-6-phosphate from glucose-6-phosphate and UDP-glucose by the enzyme trehalose-6-phosphate synthase (TPS) followed by dephosphorylation to trehalose by trehalose-6-phosphate phosphatase (TPP) (Cabib and Leloir, 1958). Trehalose functions as a stress protection metabolite and as a storage carbohydrate in many organisms (Goddijn and van Dun, 1999).

Although trehalose is present in a few desiccation tolerant species a general role for this sugar in angiosperms has been questioned. Failure to detect trehalose led to the suggestion that the majority of higher plants have lost the ability to produce it (Crowe et al., 1992). However, trehalase, an enzyme responsible for the breakdown of trehalose to glucose has been detected in many plants (Muller et al., 1995). Recently, the inhibition of trehalase using Validamycin A resulted in the detection of trace levels of trehalose in tobacco (Nicotiana tabacum) and potato (Solanum tuberosum) (Goddijn et al., 1997). Furthermore, Arabidopsis thaliana TPS, TPP and trehalase genes have been identified by complementation of Saccharomyces cerevisiae mutants (Blazquez et al., 1998; Vogel et al., 1998; Muller et al., 2001) and homologous genes have now been cloned from a number of other plant species. These data suggest that the synthesis of trehalose may in fact be ubiquitous among angiosperms but that the levels are generally extremely low (Goddijn and Smeekens, 1998; Goddijn and van Dun, 1999).

In some yeasts, including S. cerevisiae, TPS1 plays a critical role in the regulation of sugar metabolism (Thevelein and Hohmann, 1995). Glycolysis operates via an autocatalytic (or ‘turbo’) principle in which ATP is consumed to drive the catabolism of glucose before it is replenished by subsequent metabolism (Teusink et al., 1998). As a consequence of this design, it has been argued that when the supply of glucose increases abruptly glycolysis is predisposed to use ATP faster than it can be generated, causing metabolism to stall. This leads to a phenomenon termed substrate-accelerated death (Teusink et al., 1998). TPS1 is required to regulate the influx of glucose into glycolysis, thereby preventing a fatal metabolic imbalance (Teusink et al., 1998). The mechanism by which TPS1 controls glycolysis in yeasts is not fully understood but the predominant site of action is believed to be the initial enzymatic step, catalysed by hexokinase (HXK) (Thevelein and Hohmann, 1995). T-6-P is a potent inhibitor of S. cerevisiae HXKII in vitro (Blazquez et al., 1993) providing a potential means of biochemical regulation. However, restoration of wild-type T-6-P levels in tps1Δ only partially rescues glycolytic function suggesting that the TPS1 gene product is also required for the correct control of glycolysis (Bonini et al., 2000; Noubhani et al., 2000).

TPS1 is also necessary for carbon catabolite repression of gene expression in S. cerevisiae (Thevelein and Hohmann, 1995), which is believed to operate through HXKII mediated sugar signalling (Entian and Frohlich, 1984). In plants sugars also act as global regulators of gene expression (Smeekens, 2000) and a similar sensing role has been proposed for HXK (Jang et al., 1997). Over-expression of a heterologous TPS gene from S. cerevisiae or Escherichia coli in plants results in significant morphological growth defects and altered metabolism (Goddijn et al., 1997; Romero et al., 1997). Exogenous trehalose also affects plant metabolism and gene expression (Muller et al., 1998; Wagner et al., 1986; Wingler et al., 2000). Such observations have led to speculation that trehalose synthesis might function in the regulation of plant growth and development (Goddijn and Smeekens, 1998; Goddijn and van Dun, 1999). In order to investigate whether a general physiological role exists for trehalose metabolism in plants we have isolated Arabidopsis mutants that are disrupted in the TPS1 gene (Genbank accession: Y08568) that was previously shown to complement the S. cereviseae tps1Δ mutant (Blazquez et al., 1998).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The tps1 mutant is embryo lethal

Two independent TPS1 mutant alleles (tps1–1 and tps1–2) were isolated from the Arabidopsis SLAT population, which carries a non-autonomous dSpm transposable element from maize (Tissier et al., 1999). tps1–1 was identified in the SINS (Sequence Insertion Site) database and tps1–2 was isolated by a PCR-based reverse genetic screen. The two mutants contain dSpm elements inserted in different positions, with the borders situated 581 and 36 bp 5′ of the start of translation in the second and first exon, respectively, of the TPS1 gene (Figure 1a). Both mutants were initially isolated as heterozygotes. Southern blotting confirmed that tps1–1 and tps1–2 each contained a single dSpm copy (not shown). TPS1 is located on the bottom arm of Chromosome 1 (approx. 125 cm).

image

Figure 1. Identification of two recessive embryo lethal tps1 mutants.

(a) Structure of the TPS1 gene showing the position of dSpm insertions in tps1–1 and tps1–2. Boxes and lines represent exons and introns, respectively. Nucleotide positions are relative to the translational start site. Primer positions are shown as arrows.

(b) Round and wrinkled seeds from a TPS1/tps1-1 plant segregating in a 3 : 1 ratio. A wrinkled seed is marked with an arrow. Scale bar is 300 μm.

(c) Developing tps1–1 seed segregating in a silique. A and N are arrested and normal embryos, respectively. Scale bar is 1 mm.

(d) Genotype determination of single A and N embryos by PCR. Primers 1 and 2 amplify a 0.8 Kb fragment from TPS1. Primers 1 and 3 amplify a 0.3 Kb fragment from tps1–1.

(e) Genotype determination of a tps1–1 revertant by PCR. The revertant was obtained by backcrossing the mutant to a line containing the transposase. F2 plants, which contained the transposon, were selected for seed that were not segregating. F3 plants were then tested using the combination of TPS1 primers described in (d) and BAR specific primers to confirm that the transposon had jumped. P is the parental F1 line and R is the revertant F3 line. PCR products spanning the former insertion site were then sequenced and found to be identical to wild-type TPS1 (not shown).

The tps1 mutants are recessive and embryo lethal. When segregating seed batches were inspected (Figure 1b), tps1–1 displayed a wrinkled seed phenotype in the ratio of 3 : 1 round to wrinkled (748 : 271 χ2 = 1.30). These wrinkled seeds failed to germinate. Seeds containing arrested embryos could be identified segregating in developing TPS1/tps1–1 siliques (Figure 1c). PCR performed on DNA from single excised embryos demonstrated that the phenotype and dSpm element co-segregated (Figure 1d). Complementation analysis confirmed that the two mutants were allelic (not shown). Revertants were isolated from tps1–1 by crossing the mutant into a line containing the transposase (Tissier et al., 1999). To identify revertants, seed from dSpm containing F2 plants were screened for a wrinkled phenotype. Out of 567 F2 plants analysed in this way, four produced seed that were not segregating. PCR analysis and sequencing of one such line (Figure 1e) confirmed that it was a revertant in which the transposon had jumped to a new location and the wild-type TPS1 open reading frame was restored. This line further demonstrates linkage between the insertion and phenotype.

Embryo development arrests early in the phase of pattern formation and storage reserve deposition

The development of many higher plant embryos, including Arabidopsis, can be divided conceptually into three overlapping phases (Jurgens and Mayer, 1994). In the first phase cells divide and differentiate, establishing the pattern of the embryo. The second phase is characterised by growth (cell expansion) and the accumulation of storage reserves. In the final phase the embryo desiccates and enters a state of developmental arrest. In TPS1/tps1–1 the embryos from a single silique were essentially indistinguishable early in the morphogenic phase of development (Table 1). However, beyond the heart stage a segregation ratio emerged as approximately 25% of the seeds began to develop more slowly. Development was progressively retarded, eventually arresting at the torpedo stage (Figure 2a). This point in development approximately marks the transition between the first and second phases of embryo development, that is pattern formation and growth/storage reserve deposition (Jurgens and Mayer, 1994).

Table 1.  Segregation of embryos from single TPS1/tps1–1 siliques over the course of seed development
EmbryonicDays after pollination
Stage1234681012
  1. Embryonic stages are according to Jurgens and Mayer (1994). The data shown for single siliques are representative. The expected segregation ratio of developmentally arrested to normal embryos is approx. 0.25 and where observed is shown.

Quadrant57       
Octant1       
Dermatogen 4      
Globular 5223     
Heart  3132    
Torpedo   2911141215
Cotyledon    3941  
Mature      46 
Desiccation       48
Ratio0.220.250.210.24
image

Figure 2. Developing seeds produced by TPS1/tps1-1 heterozygotes. Wild-type (left) and tps1-1 seeds (right) are shown from the same silique.

(a) Mature Wt embryo versus arrested torpedo stage tps1-1 embryo. CE and EM are cellular endosperm and embryo, respectively. Scale bar is 100 µm. Samples are stained with toluidine blue.

(b) Carbohydrate in the endosperm of mature stage seeds. Starch grains (arrow) accumulate in the cellular endosperm. The embryo is not visible in the section shown in the right hand plate. Scale bar is 50 µm. Samples are stained using the periodic acid-Schiff's reaction.

(c) Aleurone (AL) in desiccation stage seeds. Gaps are indicated with an arrow. Scale bar is 20 µm. Samples are stained with toluidine blue. 

Download figure to PowerPoint

In contrast to the embryo, the endosperm of tps1–1 seeds developed normally up until seed maturation. During this period starch accumulated transiently in the endosperm cells of both the mutant and wild-type (Figure 2b). In desiccating tps1–1 seeds gaps were observed in the aleurone, a persistent layer of the endosperm. These cells appeared devoid of their endoplasm (Figure 2c). It is not clear whether this phenotype is directly associated with TPS1 function or if the cells are simply ruptured through mechanical stress as the seed coat collapses on the immature embryo.

TPS1 is up-regulated during seed development and is required for the expression of genes associated with seed storage reserve deposition

Expression of TPS1 in Arabidopsis was determined using semi-quantitative RT-PCR (Zhao et al., 1995). TPS1 mRNA was present at low levels in all the tissues analysed but was up-regulated during seed development, peaking at the cotyledon stage (Figure 3). Analysis of RNA from single excised embryos revealed that TPS1 transcripts were lacking from arrested torpedo stage tps1-1 (Figure 3c). Genes associated with the synthesis of seed storage protein (At2S2) and oil (AtOLEOSN2) were used as molecular markers for seed maturation (Conceicao and Krebbers, 1994; Kirik et al., 1996). Transcripts of both genes were suppressed in torpedo stage tps1-1 embryos relative to wild-type embryos of the same morphological stage, while actin (AtACT2) transcript levels were not significantly affected (Figure 3).

image

Figure 3. Semi-quantitative RT-PCR analysis of TPS1 expression. Data are representative of three separate experiments.

(a) Wild-type seed development. He, heart, To, torpedo, Co, cotyledon, De, desiccating, Dr, dry seed.

(b) Wild-type tissues. Se, seedling, Le, leaf, Ca, Cauline leaf, Ro, root, St, stem, Fl, flower, Si, silique.

(c) Expression of TPS1, 2S2 and OLEOSN2 in single tps1-1 and wild-type torpedo stage embryos. ACT2 is shown as a control. 

Download figure to PowerPoint

Sucrose levels increase markedly during the transition from pattern formation to growth/storage reserve deposition

In a variety of plants an increase in the supply of sucrose to the developing embryo has been observed, which coincides with an increase in sink strength as the embryo switches from pattern formation to growth and storage reserve accumulation. Recent studies have implicated the change in sucrose levels in the control of this developmental transition, through sugar-mediated regulation of gene expression (Wobus and Weber, 1999). Measurements of sugars during wild-type Arabidopsis seed development (Figure 4) revealed that sucrose levels are relatively low during pattern formation but increase markedly as the embryo enters the period of cell expansion and storage reserve deposition (torpedo to cotyledon). In contrast levels of glucose and fructose are relatively high during morphogenesis and subsequently decline.

image

Figure 4. Sugar levels in developing wild-type seeds. Values are the mean ± SE of five batches of 50 seeds. 

Download figure to PowerPoint

T-6-P levels in developing seeds and siliques containing cotyledon stage embryos were investigated using the method of Blazquez et al. (1994). T-6-P was found to be below the limit of detection in these tissues (determined to be approx. 50 µm using spiking controls). We were unable to measure metabolites in tps1-1 seed due to the technical difficulties of obtaining sufficient tissue of a confirmed genotype.

Reducing sucrose levels partially rescues tps1-1 embryo development in vitro

The null phenotype of S. cerevisiae tps1Δonly manifests itself when grown on high levels of sugars and this defect can be overcome by restricting the influx of glucose into glycolysis (Thevelein and Hohmann, 1995). To determine what affect a change in sucrose level has on tps1-1 embryo development, heart stage embryos were removed from a segregating silique and cultured in vitro (Hadfi et al., 1998). The maximum sucrose concentration in developing seeds was in the order of 200 mm (see Experimental procedures). Under normal culture conditions (340 mm sucrose) tps1-1 embryos arrest at the torpedo stage, while wild-type embryos undergo cell expansion (Figure 5). Reducing the level of sucrose did not significantly affect morphogenesis but did inhibit cell expansion in wild-type embryos. Relatively low levels of sucrose (30–90 mm) that are sub-optimal for wild-type embryo development, significantly enhanced cell expansion in tps1-1 embryos, partially rescuing the phenotype (Figure 5). In a control experiment high osmoticum (310 mm sorbitol) was provided to tps1-1 embryos in addition to 30 mm sucrose. The mean embryo length was 392 µm ± 12 SE (n = 7), which is not significantly different from the value obtained with 30 mm sucrose alone (P > 0.01) indicating that the stimulation of growth in tps1-1 is not due to low osmoticum. Together these data suggest that the arrested growth of tps1-1 embryos is linked to the increase in sucrose levels that occurs during the transition from pattern formation to expansion and storage reserve deposition.

image

Figure 5. In vitro culture of tps1–1 embryos. Heart stage embryos were cultured for 14 days on medium containing different compounds.

(a) Images of embryos cultured on different concentrations of sucrose. A heart stage embryo is shown as a zero time control. Scale bar is 100 μm.

(b) The mean length in µm ± SE (n = 3–18) of embryos cultured on various different compounds (in all but the top left plate the sucrose concentration was 180 mm). The genotype of single embryos was determined using PCR as in Figure 1(d). In a control experiment 310 mm sorbitol plus 30 mm sucrose yielded a mean tps1-1 embryo length of 392 µm ± 12 SE (n = 7).

The growth of tps1-1 embryos was not rescued in vitro by the provision of exogenous trehalose or T-6-P (5–20 mm and 1–5 mm, respectively) (Figure 5b). It has previously been demonstrated that trehalose is taken up by plant tissues and affects metabolism at similar concentrations to those used here (Wingler et al., 2000). However, it is unlikely that T-6-P is taken up, since sugar phosphates are not membrane permeable. Mutant embryos would not germinate precociously and develop into plants even after prolonged periods of culture while wild-type embryos could be induced to germinate by reducing the sucrose level in the medium to 30 mm (not shown).

Trehalose-6-phosphate is not an inhibitor of Arabidopsis HXK1 and 2 activities

The predominant mechanism through which TPS1 controls sugar influx into glycolysis in yeasts is believed to be that of HXK inhibition by T-6-P (Blazquez et al., 1993; Thevelein and Hohmann, 1995). In Arabidopsis two HXK genes have so far been characterised (Jang et al., 1997). To determine whether T-6-P inhibits HXK activity in Arabidopsis assays were performed on crude extracts from leaves of wild-type plants and transgenic plants constitutively over-expressing AtHXK1 or AtHXK2 (Figure 6). Although T-6-P levels are likely to be less than 50 µm in cotyledon stage Arabidopsis seeds, T-6-P was not a significant inhibitor of both glucose- and fructose-dependent HXK activity even at 5 mm (Figure 6). The concentration of hexose used as a substrate (2 mm) was saturating and is less than that estimated to be in developing seeds at the cotyledon stage (approx. 15 mm). In contrast S. cerevisiae HXKII was strongly inhibited (Figure 6). Kinetic analysis was performed using glucose as a substrate and the inhibition of AtHXK1 and AtHXK2 found not to be statistically significant (P > 0.01) while S. cerevisiae HXKII was inhibited competitively with a Ki value of 45 µm (Table 2). In control experiments it was confirmed that crude extracts from Arabidopsis did not effect T-6-P inhibition of S. cerevisiae HXKII activity suggesting that T-6-P is not inactivated or degraded in the extract (not shown).

image

Figure 6. A comparison of the affect of T-6-P on the activity of S. cerevisiae HXKII with that on Arabidopsis HXK1 and HXK2. Values are the mean ± SE of measurements made of four separate protein extracts from leaves and are expressed as a percentage of the control. S. cerevisiae HXKII was from Sigma-Aldrich Co. Transgenic lines contained 34-and 17-fold over-expression of HXK activity for sense-AtHXK1 and sense-AtHXK2, respectively. Expression was driven by the CaMV 35S promoter.

(a) Using 2 mm glucose as substrate.

(b) Using 2 mm fructose as substrate.

Download figure to PowerPoint

Table 2.  Kinetic analysis of T-6-P inhibition of Arabidopsis HXK1 and 2
 S. cerevisiaeArabidopsis sense-AtHXK  
PropertyHXKIIWild typeAtHXK1AtHXK2
  1. Glucose was used as a substrate. Km and Ki values were derived from double reciprocal plots. Data for S. cerevisiae best fitted a competitive model of inhibition. Values are the mean ± SE of results from four separate plots. NS = inhibition not significant (P > 0.01).

Km (µM)167 ± 2377 ± 3168 ± 1181 ± 13
Competitive Ki (µM)45 ± 6NSNSNS

Inhibition of hexokinase activity does not rescue tps1–1 embryo growth

In S. cerevisiae reducing HXK activity by disruption of HXKII rescues tps1Δ growth on glucose (Thevelein and Hohmann, 1995). To investigate whether a decrease in HXK activity can rescue the tps1-1 phenotype in Arabidopsis antisense suppression and inhibitor experiments were performed. The tps1-1 mutant was crossed into transgenic lines containing antisense AtHXK2, which suppresses both AtHXK1 and AtHXK2 transcripts (Jang et al., 1997). Total HXK activity was reduced by 60–80% in the leaves of TPS1/tps1–1 F1 plants (Table 3). However, F2 seed segregated with a 3 : 1 round to wrinkled ratio demonstrating that embryo development was not rescued (Table 3). The CaMV 35S promoter that drove anti-AtHXK2 expression is active during embryo maturation but at reduced levels compared with leaves (Dormann et al., 2000). We cannot exclude the possibility that the level of suppression is insufficient to rescue tps1-1. However in S. cerevisiae only a 30–40% reduction in HXK activity is required to rescue tps1Δ growth on glucose (Hohmann et al., 1999). Developing tps1-1 embryos were also cultured in vitro in the presence of the HXK inhibitor glucosamine (Wiese et al., 1999). Wild-type embryo growth was reduced with increasing inhibitor concentration (20–200 mm). However, growth of tps1-1 embryos was not rescued by this treatment (Figure 5b).

Table 3.  Effect of AtHXK2 antisense suppression on tps1–1 seed development
Cross (♂×♀)Activity in F1F2 segregation ratio (R:W)χ2
  1. Anti-sense AtHXK2 has previously been shown to suppress both AtHXK1 and AtHXK2 transcripts (Jang et al., 1997). Two independent 35S::AtHXK2 antisense lines were used (A225B7–5 and A225B10–5). F1 heterozygous TPS1/tps1–1 plants containing the anti-AtHXK2 construct were selected and HXK assays were performed on crude extracts from the leaves. Values are the mean ± SE from four separate extracts. The ratio of round to wrinkled seeds (R:W) was determined in the F2 generation. Deviations are not significantly different from the expected 3 : 1 ratio (P > 0.5).

Wt X TPS1/tps1–1100 (%)647 : 2300.64
A225B7–5 X TPS1/tps1–137.4 ± 2.1549 : 1930.35
A225B10–5 X TPS1/tps1–121.6 ± 1.3639 : 2341.42

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

In this study we have demonstrated that TPS1, which catalyses the first step of trehalose biosynthesis, is essential for Arabidopsis embryo development, establishing for the first time that this pathway plays a critical (and possibly generic) role in higher plants. Our data are consistent with a requirement for T-6-P, trehalose and/or the TPS1 gene product. Growth of tps1 embryos arrests during the transition from pattern formation to growth/storage reserve deposition. Sucrose levels in the seed increase dramatically at the same stage. This increase has been observed in other species and is thought not only to be necessary to support the rapid synthesis of seed storage reserves, but also to play a role in triggering the developmental switch via sugar signalling (Wobus and Weber, 1999).

The S. cerevisiae tps1Δ mutant also only manifests a null phenotype when grown on high levels of sugars (Thevelein and Hohmann, 1995). This defect can be overcome by restricting the influx of sugars into glycolysis (Thevelein and Hohmann, 1995). The observation that reducing the availability of sucrose (but not osmoticum) to tps1-1 embryos in vitro can partially rescue growth is consistent with the ‘yeast model’ and suggests that TPS1 might prevent a sucrose-induced metabolic imbalance that would otherwise result in arrested growth.

The mechanism by which TPS1 controls glycolysis in yeasts is not fully understood but substantial evidence suggests that both T-6-P and the TPS1 protein regulate HXK activity (Bonini et al., 2000; Noubhani et al., 2000). In S. cerevisiae HXKII is the primary site of entry of carbon skeletons from glucose and fructose into glycolysis and is therefore the logical target for TPS1/T-6-P mediated regulation. However, T-6-P is not an inhibitor of AtHXK1 and AtHXK2 activity in vitro. Similar findings have been reported for the effect of T-6-P on HXK activity from spinach (Spinacia oleracea) leaf extracts (Wiese et al., 1999). Furthermore, although mutations in HXKII can rescue growth of tps1Δ on glucose in S. cerevisiae (Thevelein and Hohmann, 1995), our attempts to rescue tps1-1 embryo growth in vivo using AtHXK2 antisense suppression and in vitro using the inhibitor glucosamine failed. Together these lines of evidence suggest that TPS1 exerts its dramatic effect on embryo metabolism/development via a different mechanism than that involving T-6-P inhibition of hexokinase in some yeasts. The possibilities remain that the TPS1 protein its self is required and/or that T-6-P or trehalose act on different targets.

The TPS1 protein might act via direct interaction with hexokinase or other important proteins, as has been proposed in yeasts (Bonini et al., 2000; Noubhani et al., 2000). In this context it is interesting to note that TPS has been identified as a 14 : 3 : 3 binding protein, along with key regulatory enzymes of plant primary metabolism such as nitrate reductase and sucrose phosphate synthase (Moorhead et al., 1999). Recent findings show that the binding of 14 : 3 : 3 s to some proteins is dependent on the nutritional status of the plant cell and determines their stability (Cotelle et al., 2000). Interaction with 14 : 3 : 3 s may provide one possible means by which TPS1 is regulated in Arabidopsis in response to changing sugar supply, thereby coupling carbohydrate status to the regulation of metabolism and development.

T-6-P or trehalose may target proteins other than hexokinase. In plants sugars can enter glycolysis via more than one pathway. In many seeds during the early stages of embryo development sucrose is broken down mainly by invertase producing glucose and fructose (Wobus and Weber, 1999). However invertase activity declines as the sink status of the embryo increases and during the phase of storage reserve deposition sucrose is predominantly metabolised by sucrose synthase to form fructose and UDP-glucose. These metabolites are subsequently metabolised by fructokinase and UDP-glucose pyrophosphorylase, bypassing HXK (Sturm and Tang, 1999). The significance of glucose as a substrate for glycolysis might therefore be minor in developing embryos at the stage in which tps1-1 growth arrests. To investigate the possibility that T-6-P or trehalose might affect other enzyme activities associated with sucrose catabolism we have assayed hexokinase, fructokinase, invertase, sucrose synthase and UDP-glucose pyrophosphorylase in crude extracts from developing cotyledon stage embryos of Brassica napus, a close relative of Arabidopsis. However, no significant effect was observed (not shown).

Sugars are known to induce a number of genes associated with storage reserve synthesis in plants (Koch, 1996; Smeekens, 2000). A recent study has shown that exogenous trehalose can influence the expression of the ADP-glucose pyrophosphorylase gene ApL3 and enhance starch accumulation in Arabidopsis (Wingler et al., 2000). The suppression of genes associated with storage product deposition in tps1-1 embryos also implicates trehalose metabolism in sugar signalling. However we were unable to rescue tps1-1 embryo growth in vitro using exogenous trehalose. It is not clear whether gene expression is controlled directly or indirectly by trehalose. In S. cerevisiae TPS1 is necessary for carbon catabolite repression of gene expression, but trehalose metabolism is not thought to be involved directly in sugar signalling (Thevelein and Hohmann, 1995).

TPS1 is up-regulated in maturing seeds. This finding is consistent with microarray data, which shows that TPS1 transcripts are about 4- to 6-fold more abundant in developing seeds compared with leaves and roots (Girke et al., 2000). TPS1 is also expressed at low levels in all the Arabidopsis tissues analysed. A search of the Arabidopsis genomic database revealed approx. 11 putative TPS genes, of which at least eight are expressed, as indicated by the existence of ESTs. In addition there are also approx. 11 putative TPP genes of which at least six are expressed. Only TPS1, TPPA and TPPB have thus far been functionally characterised (Blazquez et al., 1998; Vogel et al., 1998). It is probable that these genes exhibit different spatial and temporal patterns of expression. Trehalose metabolism may play important physiological roles at additional stages in the plant's life cycle to embryo development.

In conclusion we have demonstrated for the first time that an enzyme of the trehalose biosynthetic pathway is essential for Arabidopsis embryo development (and therefore potentially higher plants in general). It is not yet established whether T-6-P, trehalose or the TPS1 protein is required. Furthermore the precise mechanism by which embryo growth is regulated remains to be elucidated. TPS1 appears to play a vital role in responding to the increase in sucrose supply, which accompanies the onset of embryo maturation. However, importantly our data indicate that the mechanism is fundamentally different from that reported in many yeasts.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Isolation of tps1

Genomic DNA pools from an En/Spm transposon tagged Arabidopsis thaliana (ecotype Col0) mutant population, consisting of approx. 48 000 lines were screened by PCR (Tissier et al., 1999) for a knockout in TPS1. Gene specific primers were used in combination with the dSpm border primers; dSpm1 and dSpm8 (Tissier et al., 1999). Two mutant alleles (tps1-1 and tps1-2) were identified (tps1-1 in the SINS database of Sequence Insertion Sites) and the insertion sites mapped by sequencing PCR products spanning the borders. Revertants in tps1-1 were obtained by backcrossing to an En1 line (Wisman et al., 1998) containing active copies of the transposase. F2 TPS1/tps1-1 plants containing dSpm were then selected by Basta resistance and revertants identified by screening for normal seed batches (not displaying a 3 : 1 round to wrinkled ratio). Excision of dSpm and restoration of the TPS1 wild type open reading frame was confirmed by PCR and sequencing of the former insertion site (Tissier et al., 1999).

DNA extraction, RNA extraction and PCR analysis

Genomic DNA and total RNA were extracted using the Gentra PUREgene and PUREscript kits from Flowgen (Lichfield, Staffordshire, UK) following the manufacturer's protocols. Where DNA and RNA were extracted from single embryos the embryos were excised using fine forceps, transferred to 0.5 ml tubes, rinsed in water and frozen in liquid nitrogen. 50 μl of lysis buffer was then added (from the appropriate kit) and the embryo crushed using a pipette tip. The volumes in subsequent steps of the extraction were scaled down accordingly. 0.1 μg of glycogen was included before adding isopropanol to aid DNA or RNA precipitation. The pellet was resuspended in 5 μl of water. For genotyping single embryos PCR was performed with the whole DNA sample using primers dSpm8, TPS1 1 and 2 (5′-cgtgtgagaggcaagaagttagg-3′ and 5′-gacgaagggaatggtgtatggag-3′) in a single reaction (Figure 1d). The presence of dSpm in tps1-1 revertants was tested by PCR using primers to BAR (Figure 1e). The synthesis of single stranded cDNA was carried out using the Reverse-iT kit from Advanced Biotechnologies Ltd. (Epsom, Surrey, UK). For cDNA synthesis from single embryos the reaction volume was 10 μL. Semi-quantitative RT-PCR was described by Zhao et al. (1995). The amount of template and the number of cycles were optimised so that product abundance could be compared by agarose gel electrophoresis within the linear phase of amplification. Details of the PCR method were taken from Mai et al. (1998). The primers used were TPS1 and 2, 2S2 (5′-atgatctttccgtggaggtg-3′ and 5′-gccgtagcgatgagtttgat-3′), OLEOSN2 (5′-gcgtgtgcatgtagaccgta-3′ and 5′-ggctcatgggtctcagtcat-3′) and ACT2 (5′-gttgggatgaaccagaagga-3′ and 5′-cttacaatttcccgctctgc-3′). The identity of PCR products was confirmed by sequencing. Southern analysis was performed as previously described (Eastmond et al., 2000) except that the filter was probed using a BAR specific probe.

Microscopy

Siliques were fixed overnight in 4% paraformaldehyde, 25 mm KPO4 buffer pH 7.0, 0.1% Tween-20, washed in buffer, dehydrated through a graded ethanol series to 95% ethanol, and embedded in JB-4 glycol methacrylate resin (Polysciences, Warrington, PA, USA). 5 µm sections were cut in ribbons using glass knives made from microscope slides on a Microm HM 335E rotary microtome, according to Ruzin (1999). Samples were stained in 0.1% toluidine blue in 25 mm KPO4 buffer pH 5.5, or using the periodic acid-Schiff's reaction. Stained samples were mounted under coverslips in DPX and photographed on a Zeiss Axiophot (Oberkochen, Germany) microscope using Fujichrome 64TII slide film (Fuji Photo Film (UK) Ltd., London, UK). Images were digitized with a Nikon CoolscanII film scanner (Nikon UK Ltd., Kingston Upon Thames, Surrey, UK) and processed in Adobe Photoshop (Adobe Systems UK, Uxbridge, Middlesex, UK).

Metabolite measurements and assays

Seeds were excised from a single silique and frozen in liquid nitrogen. Sugars were extracted three times with 80% (v/v) ethanol at 80°C for 20 min. The extracts were pooled, evaporated to dryness and the pellet resuspended in water. Glucose, fructose and sucrose levels were determined enzymatically as described by Stitt et al. (1989). T-6-P was extracted from developing seeds or siliques and the levels determined according to Blazquez et al. (1994). The detection limit for T-6-P was approx. 50 µm. Spiking controls determined that the recovery of all metabolites was between 86 and 104%. Concentrations were estimated based on a seed volume of 40 nL derived from Figure 2(a), assuming that 20% is accessible to the metabolite. Hexokinase activity was determined according to Wiese et al. (1999) using rosette leaves from wild type (Col0) and transgenic Arabidopsis plants containing 35S::AtHXK1 and 35S::AtHXK2 constructs (Jang et al., 1997). Assays were performed in a 1-ml volume using a UV4 Unicam UV/Vis spectrometer.

Embryo culture in vitro

Siliques from TPS1/tps1-1 plants were surface-sterilised in 1.5% sodium hypochlorite and 0.1% Tween 80 for 10 min. After rinsing in sterile water heart stage embryos were removed using fine forceps under sterile conditions. Embryos were cultured at 21°C (20 µmol m−2s−1PPFD) on 0.8% (w/v) agarose containing Gamborg's B5 basal medium (Sigma-Aldrich Co., Poole, Dorset, UK), 30–340 mm sucrose and 2.7 mm glutamine (Hadfi et al., 1998). T-6-P, trehalose, sorbitol and glucosamine were also included in some experiments.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We wish to thank Jyan-Chyun Jang and Jen Sheen for providing us with Arabidopsis sense-and antisense-AtHXK2 lines (225B3–4, A225B7–5 and A225B10–5) and constructs to generate sense-AtHXK1 lines, which were made by Susanna Boxall. Analysis of tps1-1 was funded by the Biotechnology and Biological Sciences Research Council via the Genome Analysis of Agriculturally Important Traits grant 17/GAT9139 to I.A.G.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References