Intracellular sucrose communicates metabolic demand to sucrose transporters in developing pea cotyledons

Abstract

Mechanistic inter-relationships in sinks between sucrose compartmentation/metabolism and phloem unloading/translocation are poorly understood. Developing grain legume seeds provide tractable experimental systems to explore this question. Metabolic demand by cotyledons is communicated to phloem unloading and ultimately import by sucrose withdrawal from the seed apoplasmic space via a turgor-homeostat mechanism. What is unknown is how metabolic demand is communicated to cotyledon sucrose transporters responsible for withdrawing sucrose from the apoplasmic space. This question was explored here using a pea rugosus mutant (rrRbRb) compromised in starch biosynthesis compared with its wild-type counterpart (RRRbRb). Sucrose influx into cotyledons was found to account for 90% of developmental variations in their absolute growth and hence starch biosynthetic rates. Furthermore, rr and RR cotyledons shared identical response surfaces, indicating that control of transporter activity was likely to be similar for both lines. In this context, sucrose influx was correlated positively with expression of a sucrose/H⁺ symporter (PsSUT1) and negatively with two sucrose facilitators (PsSUF1 and PsSUF4). Sucrose influx exhibited a negative curvilinear relationship with cotyledon concentrations of sucrose and hexoses. In contrast, the impact of intracellular sugars on transporter expression was transporter dependent, with expression of PsSUT1 inhibited, PsSUF1 unaffected, and PsSUF4 enhanced by sugars. Sugar supply to, and sugar concentrations of, RR cotyledons were manipulated using in vitro pod and cotyledon culture. Collectively the results obtained showed that intracellular sucrose was the physiologically active sugar signal that communicated metabolic demand to sucrose influx and this transport function was primarily determined by PsSUT1 regulated at the transcriptional level.

Introduction

Most nutrients are imported into maternal seed tissues through the phloem and reach the symplasmically isolated filial tissues (endosperm/embryo) following their release to the seed apoplasm (Zhang et al., 2007). Import of sucrose, together with a spectrum of amino acids and amides, largely accounts for biomass gain of seed filial tissues (Patrick and Offler, 2001). Sugar and amino acid transporters, localized to filial cells that are juxtaposed to maternal tissues, take up these compounds from the seed apoplasm for subsequent symplasmic delivery to filial storage sites (Zhang et al., 2007). Demand for sugars and amino nitrogen compounds is set by biosynthetic capacities of processes responsible for their sequestration into fats, proteins, and starch (Borrás et al., 2004). How this metabolic demand is communicated from filial storage sites to phloem import of nutrients into seed maternal tissues is poorly understood. At least for developing seeds of grain legumes, nutrient demand by filial tissues appears to be sensed by osmotically driven alterations in turgor pressures of nutrient release cells located in their coats (Zhang et al., 2007; but see Wang and Fisher, 1994; van Dongen et al., 2001). Here enhanced nutrient uptake by filial tissues decreases the osmolality of the seed apoplasmic sap with a consequent rise in seed coat cell turgor. If seed coat turgor exceeds a set point, activities of transporters responsible for nutrient release increase to meet filial demand. Under conditions of sustained nutrient demand, the turgor set point decreases to drive higher rates of phloem import (Zhang et al., 2007).

A key element missing from the above turgor-homeostat model (Patrick, 1994) is the underlying mechanism that integrates storage product biosynthesis with activities of cotyledon transporters retrieving substrates from the seed apoplasmic space. The model predicts that nutrient uptake by cotyledons is regulated by rates of their intracellular consumption; a phenomenon consistent with observed sink-limited gains in seed biomass (Borrás et al., 2004). Here it is envisaged that intracellular pool sizes of nutrients inversely reflect activities of biosynthetic enzymes. These pools function as signals to regulate activities of cotyledon transporters at the transcriptional level through substrate de-repression of transporter gene expression. Tentative support for such a mechanism comes from the finding that expression of a sucrose/symporter (VfSUT1) was repressed by culturing Faba bean cotyledons for 3 d on medium containing elevated sucrose or glucose concentrations (Weber et al., 1997). However, the results from these studies (Weber et al., 1997) do not indicate whether the sugar signal was sucrose or its hydrolysis products and whether its action was mediated at the cotyledon plasma membranes or intracellularly. Moreover, the observed response may not be representative of in planta regulatory mechanisms.

An opportunity to discover in planta sugar regulation of transporter activity is offered by the near-isoline of a pea mutant (rrRbRb) with a lesion at the rugosus (r) locus encoding starch branching enzyme 1 (SBEI; Bhattacharyya et al., 1990). This insertion results in wrinkled seeds containing cotyledons with reduced starch (∼56%) and elevated sucrose levels (180%) compared with those of the round wild-type cotyledons (RRRbRb; Wang and Hedley, 1991). Indeed, rates of sucrose uptake by cotyledons from wrinkled and round seeded cultivars of pea are consistent with starch biosynthesis regulating sucrose transporter activity (Edwards and ap Rees, 1986). Net rates of in vitro sucrose uptake were found to be less for cotyledons of wrinkled seeds (Edwards and ap Rees, 1986) when excised cotyledons were incubated in media concentrations of sucrose at, or below, those of the seed apoplasmic space (i.e. <180 mM; Rosche et al., 2005). Whether this relationship is causal and mediated by de-repression of sucrose transporter gene expression awaits further study.

Using differences in starch biosynthetic capacities of rrRbRb and RRRbRb cotyledons of pea (Wang and Hedley, 1991), the hypothesis that intracellular sugars function as signals to co-ordinate sink demand with sucrose uptake by regulating sucrose transporter activities was explored. The weight of current evidence points to regulation of sucrose transporter activity at the transcriptional level (Weber et al., 1997; Chiou and Bush, 1998; Aoki et al., 1999; Barker et al., 2000; Matsukura et al., 2000; Vaughn et al., 2002; Ransom-Hodgkins et al., 2003; Yao et al., 2003) but translational (Conde et al., 2006) and post-translational regulation (Roblin et al., 1998) cannot be excluded. To provide context, experiments were first conducted to test for positive correlations between cotyledon biomass gain (as a surrogate measure of starch biosynthesis), sucrose transporter activities, and their expression levels. Sucrose uptake by pea cotyledons is mediated by the combined activities of a sucrose/proton symporter (PsSUT1) and two sucrose facilitators (PsSUF1 and PsSUF4; Zhou et al., 2007). Transport activities of sucrose facilitators were distinguished from those of the symporter by the latter's selective sensitivity to the membrane-impermeant sulphydryl reagent, diethypyrocarbonate (DEPC; Zhou et al., 2007). With this background, the role of intracellular sugars (sucrose and hexoses) as signals integrating metabolic demand with activities of sucrose transporters was investigated by testing in planta correlations and responses of excised pods and cotyledons cultured on media containing different sugar species.

Materials and methods

Plant growth conditions

Wild-type (RR) and rugosus mutant (rr) Pisum sativum (pea) seed of the isogenic BC3-line (Hedley et al., 1994) were raised under glasshouse conditions of 20–26 °C by day and 15–17 °C by night, with day length extended to a 14 h photoperiod with tungsten incandescent lights. Four plants were grown in each 9.0 l pot containing a mixture of coarse sand, coconut fibre (Cocopeat), and perlite (3:1:1 by vol.). A controlled slow-release fertilizer (Nutricote, Yates, Chisso Asahi Fertiliser Co., Japan; 8 g l⁻¹) was applied at 70 and 140 d after sowing. Liquid fertilizer was applied each fortnight (Wuxal Liquid Foliar Nutrient, AgNova Technologies, Victoria, Australia; 6 ml l⁻¹). Flowers were date-tagged when their corollas fully opened co-incident with anthesis. Pods [10–35 days after flowering (DAF)] were harvested on ice and transferred to the laboratory where seeds and cotyledons were surgically removed for experimental investigations as described below.

Pod and cotyledon culture

Medium composition was a modified version of the Murashige and Skoog (MS; Murashige and Skoog, 1962) culture solution, with glycine, indole acetic acid (IAA), and kinetin omitted. Specified sugars, as the sole carbohydrate source, were supplied at given concentrations. Medium osmolality was adjusted with betaine to 490 mOsmol kg⁻¹ (Rosche et al., 2002) and medium pH adjusted to 5.8 with 1 M KOH. Filter-sterilized and double-strength modified MS medium was mixed with autoclaved double-strength bacteriological grade agar (1.6%, w/v) and poured into 9 cm diameter Petri plates under sterile conditions. Plates were sealed with Parafilm™ and stored at 4 °C until required for cotyledon culture. For pod culture, sterile single-strength modified MS medium was aliquoted into 10 ml autoclaved culture tubes fitted with non-absorbent cotton wool bungs. Two 10 ml tubes, containing medium, were then placed in an autoclaved 500 ml cylindrical culture vessel for support.

All materials and procedures for in vitro culture of cotyledons and pods were carried out using aseptic techniques. Harvested pods were surface-sterilized for 20 min in a sodium hypochlorite (1%, v/v) solution supplemented with SDS (0.01%, w/v), followed by three 5 min rinses with 600 ml of autoclaved distilled water. For pod culture, peduncles were re-cut with a sterile scalpel blade immediately prior to placing their cut ends in liquid culture medium (see above). Culture vessels, holding excised pods, were placed in controlled environmental conditions of 19 °C by day and 14 °C by night. Lighting (PAR of 170 μmol photons m⁻² s⁻¹ at the height of the culture vessel) was provided across a 14 h photoperiod.

For cotyledon culture, an incision was made along each sterilized pod wall, parallel to the median carpellary bundle, and exposed seeds were then severed at their funiculi. Cotyledons were surgically removed from seed coats with an incision made along the seed coat integument fusion line. Excised cotyledons were placed, adaxial epidermal side down, on agar culture medium at the rate of four cotyledons per plate. Plates were sealed with Parafilm™ and held in the dark at a constant temperature of 28 °C.

Growth analysis and estimates of seed development

Fresh weights of cotyledon pairs, with their embryonic axes removed, were determined immediately following excision. Their corresponding dry weights were obtained after oven drying for 2 d at 80 °C. Estimates for DAF were adjusted for seasonal variation in rates of seed development to a set of standard crop values. This was achieved using relative water content (RWC) of cotyledons as a reliable indicator of seed development from the following experimentally determined linear regressions:

[¹⁴C]Sucrose uptake

All solutions contained 1 mM CaCl₂ buffered to pH 5.5 with 5 mM MES-KOH and, unless specified otherwise, supplemented with 10 mM sucrose corresponding to the apparent Michaelis–Menten constant (K_m) value for sucrose transport into pea cotyledons (Tegeder et al., 1999). Solution osmolality was adjusted to 490 mOsmol kg⁻¹ with sorbitol. Uptake solutions were radiolabelled with [¹⁴C]sucrose (3.071×10⁵ Bq per 10 ml of buffer).

Seeds were excised and cotyledon pairs separated on ice. In order to reproduce the in vivo pathway of sucrose uptake into cotyledons through the abaxial epidermis (Tegeder et al., 1999; Rosche et al., 2002), the adaxial epidermal surfaces of excised cotyledons were smeared with Food Lube™ Grease (Rocol). Cotyledons were incubated in solutions held at 22 °C in a shaking water bath (60 oscillations). In those experiments where sucrose/H⁺ symport was distinguished from facilitated diffusion, sister cotyledons were pre-treated for 15 min in MES-KOH CaCl₂ buffer ±5 mM DEPC that selectively inhibits sucrose/H⁺ symport (Zhou et al., 2007). Unbound DEPC was removed by three successive 5 min rinses in MES-KOH CaCl₂ buffer. Pre-treated cotyledons (see above) or freshly excised cotyledons, used to determine total rates of sucrose uptake, were equilibrated for 20 min in sucrose buffer. Thereafter, surface solution was removed by gently blotting each cotyledon before immersing them in a [¹⁴C]sucrose solution for 10 min. Three 3 s washes in ice-cold sucrose buffer terminated radiolabel uptake. On removal of grease, a tissue cylinder was surgically excised from the centre of each cotyledon with a cork borer of known diameter. Any remaining [¹⁴C]sucrose located in the free space of the tissue cylinders was removed by a 3 min rinse with ice-cold sucrose buffer. Accumulated radiolabel was extracted in 80% aqueous ethanol at 80 °C for 18 h from which aliquots were radioassayed by liquid scintillation counting with automatic quench and background correction. Knowing the specific activity of the [¹⁴C]sucrose solution, rates of sucrose uptake were computed on a cotyledon surface area basis.

Semi-quantitative PCR

Upon removal from seed coats, one sister cotyledon of each cotyledon pair was immediately snap-frozen in liquid N₂. Total RNA was extracted from individual snap-frozen cotyledons using an RNeasy Plant Mini Kit (Qiagen), and reverse transcribed with Superscript reverse transcriptase and oligo(dT) (Invitrogen). Semi-quantitative PCR was performed within the linear range of amplification for each targeted fragment examined at 26 cycles with an annealing temperature of 55 °C. Gene-specific primers for PsSUF1 were CTTGCAGTTGGACTTGCCATGACTATAGTC and TCTTGACTAATGAAATCCACCCGCAATTGG, those for PsSUF4 were GAGGGGGAGGTTGAAATCTTGAGGAGAGG and GGACTGCCGATGTTCCTTGGTATCAC, and those for PsSUT1 were CATACCACAGATGTTTGTCTCTGTTTTGAG and AGTCCTGCTCTAGCAGGATTTGTATACAGC. A housekeeping gene, the glyceraldehyde-3-phosphate dehydrogenase gene from pea, PsGAPC1 (GenBank accession no. X73150), was amplified using primers ATAATTTGCTGCAGTACCCGTGTGGTTGAC and CCTTCCCCATTATTTTAAGGCATCACTCAC. PsGAPC1 expression levels served as internal positive controls for quantification of relative amounts of cDNA.

In situ hybridization

Cotyledons were harvested on ice for cellular analysis of sucrose transporter expression by in situ hybridization. Cotyledons were fixed for 6 h at 4 °C in 4% paraformaldehyde buffered at pH 7.2 with 50 mM Na₂PO₄. Following dehydration in ethanol and substitution with tertiary-butyl alcohol (TBA), tissue was infiltrated with Paraplast-saturated TBA prior to embedding in 100% Paraplast (Sherwood Medical). Sections, 8 μm thick, were cut on a steel blade held at 4 °C and mounted on AES-coated slides (Digene Diagnostics). Digoxigenin (DIG)-labelled sense and antisense riboprobes were synthesized by in vitro transcription from PCR products with a T7 promoter sequence incorporated upstream or downstream of the SUT/SUF fragments (Roche). Probes of 200–270 bp were designed to target a specific unique sequence of each SUT/SUF. The primers binding to gene sequence regions were: GTCCCCGCGTTTAATATGGAAGAAG and GGGCTGAGTCCTGCTCTAGCAGGATTTG for PsSUT1, GGGCGGCCGTCATCAGTGGTGTATTAG and GGGCATGCCACAGAATGGAATTTGG for PsSUF1, and GGTTCTCAAAAGCCTAGAAACCCAG and GGTCATCCTATTGCAGAAAAGACGG for PsSUF4. In situ hybridization was performed as described by Harrington et al. (1997). Sections were viewed using a Zeiss Axiophot microscope equipped with standard FITC/rhodamine/DAPI filter sets, and photomicrographs were taken using an Olympus digital camera C-5050 Zoom.

Western blotting

Three peptides, with sequences LIAEKAVVTAEDGGSNGGMP, IESQSQTQTQTQSEPEQQVS, and SSGEPDAEAEGESGGSAEEA corresponding to amino acid residues 255–274 of PsSUT1, 251–270 of PsSUF1, and 254–273 of PsSUF4, respectively, were synthesized and used to raise antibodies in rabbit (Mimotopes Pty Ltd, Melbourne, Australia). In an attempt to target individual sucrose transporters specifically, the above peptides were mapped within the central cytoplasmic loop (between transmembrane helices VI and VII) of each protein, where very low homology exists between different SUT/SUFs (Zhou et al., 2007). The specificity of each antibody was confirmed by western blotting using microsomal fractions isolated from yeast that was transformed with individual SUT or SUF cDNA. The polyclonal anti-SUT/SUF was further purified by affinity chromatography using the specific peptide covalently coupled to Thiopropyl-Sepharose 6B gel prepared by Mimotopes Pty Ltd, Australia. The titres of the affinity-purified antibodies were determined by enzyme-linked immunosorbent assay (ELISA). Microsomal protein fractions were extracted from cotyledons according to Ripp et al. (1988). The final microsomal pellets were suspended in 25 mM TRIS-HCl pH 6.8 with 0.5% SDS to 10–15 mg ml⁻¹. Protein concentrations were determined by a bicinchoninic acid protein assay (Pierce). Samples were mixed (1:1) with 2× SDS–PAGE sample buffer, heated at 37 °C for 30 min, and spun down briefly before loading onto polyacrylamide gels. The protein extracts were resolved on 10% SDS–polyacrylamide gels and electroblotted onto nitrocellulose membranes followed by an immunodetection procedure as described by Tegeder et al. (2000). Colour development was performed using alkaline phosphatase-conjugated secondary antibody with Western blue (Promega) as substrate. Blots were also probed with pre-immune serum as negative controls.

Soluble sugar analysis

At harvest, each cotyledon was immediately weighed to determine fresh weight (FW) and then snap-frozen in liquid nitrogen. Thereafter, cotyledons were freeze-dried for 48 h and their dry weights (DWs) measured. Cellular water volume of each cotyledon was estimated as the difference between their FW and DW. Soluble sugars for each cotyledon were triple-extracted in distilled H₂O held at 60 °C for 2 h shaking at 140 rpm. Hexose (Boehringer Mannheim) and sucrose (Birnberg and Brenner, 1984) levels in the water extracts were determined using enzyme-coupled reactions leading to reduction of NAD, measured spectrophotometrically at 340 nm against a set of standard sugar concentrations. Sugar concentrations were estimated on the basis of the cellular water volume of each cotyledon.

Results

Rates of cotyledon growth and sucrose uptake are highly correlated across seed development

Absolute growth rates of RR and rr cotyledons were used as surrogate measures of their temporal shifts in demand for photosynthetically reduced carbon. Absolute growth rates were derived from fitted dry weight curves for cotyledons of each isoline across consecutive 2 d intervals. Following exhaustion of endosperm reserves at ∼10–12 DAF, absolute growth rates rose sharply in both isolines to reach maxima by 20–22 DAF (Fig. 1A). During this phase, absolute growth rates of the two isolines increasingly diverged such that maximal growth rates of RR cotyledons were some 40% greater than those of rr cotyledons by 20–22 DAF. Thereafter, absolute growth rates of both isolines declined as seeds approached maturity at 40 DAF. Rates of decline were more precipitous for RR cotyledons such that growth rates of both isolines converged from 28 DAF onwards (Fig. 1A). Similar trends in embryo dry weights of RR and rr isolines were reported by Hedley et al. (1986). However, there is a marked temporal difference between these two sets of data. The seeds examined here entered the major phase of dry weight accumulation 5 d earlier than those reported by Hedley et al. (compare Fig. 1A with Fig. 4D of Hedley et al., 1986). This is likely to reflect differences in plant growth temperatures during the photoperiod of 25±2.5 °C (present study) and 15±2 °C (Hedley et al., 1986).

Fig. 1.

Comparisons of absolute growth rates of, and in vitro sucrose fluxes into, RR (filled diamonds) and rr (open diamonds) cotyledons across the phase of seed fill. (A) Temporal changes in absolute growth rates of RR and rr cotyledons estimated at 2 d intervals from fitted curves of cotyledon dry weights. Absolute growth rates were fitted by third-order polynomials to give the following equations for RR and rr cotyledons, respectively: y= –0.055x²+2.363x–11.87, r²=1; and y= –0.031x²+1.431x–6.077, r²=1. (B) Temporal changes in in vitro fluxes of [¹⁴C]sucrose uptake by excised RR and rr cotyledons across their abaxial epidermal surfaces. Fluxes of [¹⁴C]sucrose uptake were determined from amounts of ¹⁴C label accumulated in a tissue disc, 5 mm (14–15 DAF) or 10 mm (16–35 DAF) in diameter, excised from each cotyledon following a 10 min immersion in a 10 mM [¹⁴C]sucrose solution. The sucrose fluxes were fitted to third-order polynomials to give the following equations for RR and rr cotyledons, respectively: y= –0.0004x³–0.0339x²+0.903x–6.660, r²=0.9024; and y= –0.0002x³+0.008x²–0.104x+0.874, r²=0.9805. (C) Relationship between [¹⁴C]sucrose fluxes and absolute growth rates of RR and rr cotyledons across the phase of seed fill. Fitted linear relationships between sucrose fluxes and absolute growth rates for cotyledons from each of the two isolines were found to be statistically identical (P <0.001). The combined data set fitted to a linear regression whereby y=9.271x+4.104, r²=0.803. Vertical bars represent standard errors of the means with a minimum of six replicates per treatment.

Fig. 4.

Schematic representation of PsSUT1 (A, D), PsSUF1 (B, E), and PsSUF4 (C, F) transcript distribution in RR (A, B, C) and rr (D, E, F) cotyledons at 20 DAF detected by in situ hybridization with DIG-labelled riboprobes. Increased shading density indicates higher transcript abundance.

Overall fluxes of [¹⁴C]sucrose uptake were determined through the abaxial surfaces of cotyledons excised from the two isolines across a developmental window identical to that of the growth study. Essentially, temporal profiles of sucrose uptake by cotyledons mirrored those found for their absolute growth rates (Fig. 1B). However, the onset of differences in sucrose uptake fluxes between the two isolines was delayed by 4 d to 14 DAF onward compared with absolute growth rates (Fig. 1B versus A). This might be related to a higher proportion of soluble sugars present in the endosperm being hexoses (Rosche et al., 2005). These sugars are likely to be taken up preferentially into cotyledons from the endosperm by functional hexose transporters prior to expression of sucrose transporters following depletion of endosperm reserves (Harrington et al., 2005).

To test whether the present in vitro estimates of instantaneous rates of sucrose uptake by cotyledons reflect sucrose fuelling biomass gain, a linear regression analysis was performed on the two independent data sets (Fig. 1C). Slopes and intercepts of fitted linear relationships between rates of absolute growth and sucrose uptake for the two isolines were found to be statistically identical (P <0.001). Thus, these data were plotted as a single population as shown in Fig. 1C. The composite linear regression demonstrates that estimates of in vitro fluxes of sucrose uptake account for 90% of any temporal variation in absolute growth rates of rr and RR cotyledons.

Relationship between sucrose uptake fluxes and transporter expression levels

One sucrose/H⁺ symporter (PsSUT1) and two sucrose facilitators (PsSUF1 and PsSUF4) are expressed in developing pea cotyledons (Zhou et al., 2007). The relationship between symporter and facilitator expression to overall fluxes of sucrose uptake by RR and rr cotyledons was explored at three stages of cotyledon development (i.e. 16, 20, and 26 DAF, and see Fig. 2). Expression levels of PsSUT1 and sucrose uptake fluxes were positively correlated (r=0.87). The slope of this linear relationship was significantly greater than zero (P <0.05, and see Fig. 2A). In contrast, expression of the two genes encoding sucrose facilitators, PsSUF1 and PsSUF4, was negatively correlated with fluxes of sucrose uptake and, in both cases, exhibited statistically significant slopes (Fig. 2B, C). RR and rr cotyledon data sets fitted the same linear relationships for PsSUT1 and PsSUF4 (compare Fig. 2A and C). However, in the case of PsSUF1, the linear relationship differed significantly between RR and rr cotyledons for intercepts (P <0.001) but not slopes (Fig. 2B). Expression of PsSUF1 in rr cotyledons would appear to have a more negative impact on sucrose uptake fluxes than comparable expression levels by PsSUF1 in RR cotyledons.

Fig. 2.

Relationship between sucrose influx into, and relative expression levels of (A) PsSUT1, (B) PsSUF1, and (C) PsSUF4 in, RR (filled diamonds) and rr (open diamonds) cotyledons harvested at 16, 20, and 24 DAF. Transcript levels of each sucrose transporter were determined from total RNA extracted from one sister cotyledon. The relative expression, by semi-quantitative RT-PCR, of each sucrose transporter was normalized to expression of the housekeeping gene, PsGAPC1 (GenBank accession no. X73150). The corresponding sister cotyledon was used to estimate flux of [¹⁴C]sucrose uptake from a 10 mM sucrose solution. Except for relative expression of PsSUF1, the data sets for RR and rr cotyledons were combined. All data sets fitted linear relationships, the slopes of which were found to be significantly different from zero. The probability value for the slope is presented for each fitted line along with their associated regression coefficient as follows: PsSUT1, y=0.009x+0.465, r²=0.764, P <0.05; PsSUF1 RR, y= –0.007x+1.173, r²=0.997, P <0.05; PsSUF1 rr, y= –0.011x+0.742, r²=0.937, P <0.05; PsSUF4, y= –0.007x+1.05, r²=0.775, P <0.01. Vertical and horizontal bars represent standard errors of means derived from at least six replicates per treatment.

Cellular localization of transporter expression

Differences in gene expression levels of symporters and facilitators between RR and rr cotyledons (Fig. 2) could result from differing expression levels in the same cell population or differing cell numbers expressing these genes. To distinguish these possibilities, cellular localization of gene expression was visualized in RR and rr cotyledons at one developmental stage (20 DAF) using in situ hybridization. Serial cross-sections from the same cotyledon were probed for cells expressing each sucrose transporter gene. Photomicrographic images of probed (sense and antisense riboprobes) cotyledon sections were collected. These images were used to construct montages of transects extending from the abaxial to adaxial epidermis of each cotyledon (see Fig. 3). Summaries of observed (Fig. 3) spatial expression of each sucrose transporter gene are presented diagrammatically in Fig. 4.

Fig. 3.

Photomicrographs of the cellular localization of PsSUT1, PsSUF1, and PsSUF4 transcripts in transverse sections of RR (A, B) and rr (C, D) cotyledons detected by in situ hybridization of DIG-labelled sense (A, C) and antisense (B, D) riboprobes. Each image is a montage of photomicrographs to provide a complete transverse section of a cotyledon from abaxial to adaxial epidermis. Differences in montage lengths reflect the combined influence of variations in cotyledon size and transection location. Cotyledons are at 20 DAF. Ab, abaxial epidermis; Sp, storage parenchyma; Ad, adaxial epidermis. Bar = 40 μm.

Marked differences in spatial gene expression patterns were apparent for each sucrose transporter between RR and rr cotyledons (Fig. 4). Expression patterns of PsSUT1 and PsSUF1 were identical within each isoline (Fig. 4). Here all cells of RR cotyledons exhibited detectable levels of expression, with the strongest expression levels localized to the abaxial epidermis and several underlying cell layers. Further inward, expression declined substantially through progressive layers of storage parenchyma cells (Fig. 3A versus B; Fig. 4A, B). In contrast, detectable transcript levels of PsSUT1 and PsSUF1 were localized to abaxial epidermal cells and two underlying cell layers in rr cotyledons (Fig. 3C versus D; Fig. 4D, E). In addition, PsSUT1 and PsSUF1 appeared to exhibit greater expression levels per cell in RR compared with rr cotyledons (Figs 3, 4). Expression of PsSUF4 was restricted to epidermal and one underlying cell layer that completely circumscribed rr cotyledons but, in RR cotyledons, was confined to their abaxial region (Fig. 3A versus B; C versus D; Fig. 4C, F). Moreover, higher densities of PsSUF4 transcript were detected in rr compared with RR cells (Figs 3, 4) as found for its relative temporal expression levels (Fig. 2). Thus, relative expression levels of each transporter gene (Fig. 4) reflect a composite of transcript densities per cell and cell numbers with detectable levels of transcript.

Sucrose uptake fluxes were inversely related to pool sizes of intracellular sugars

Under the present growth conditions, differences in intracellular concentrations of sucrose and hexoses between RR and rr cotyledons were apparent between 16 and 24 DAF (data not shown, but see Fig. 5). Intracellular concentrations of sucrose ranged from 150 mM to 300 mM and of hexoses from 2 mM to 7 mM (Fig. 5). This outcome was used to discover if there was any link between intracellular concentrations of sucrose and hexoses in RR and rr cotyledons and their overall fluxes of sucrose uptake.

Fig. 5.

Relationship between sucrose influx into RR (filled diamonds) and rr (open diamonds) and intracellular concentrations of (A) sucrose and (B) hexoses across the phase of seed fill. Sucrose influx estimated by determining accumulation of ¹⁴C label following cotyledon exposure to a 10 mM [¹⁴C]sucrose solution for each sister cotyledon. Corresponding sister cotyledons were used to measure intracellular sugar concentrations in protoplasmic saps extracted from freeze-dried cotyledons. The RR and rr data sets were combined and fitted with third-order polynomials to yield the following equations for RR and rr cotyledons, respectively: y=3×10⁻⁵x²–0.014x+2.484, r²=0.669; and y=0.032x²–0.417x+1.887, r²=0.882. Vertical and horizontal bars represent standard errors of means derived from a minimum of six replicates per treatment.

Data for RR and rr cotyledons were combined on the assumption that sucrose transporter activities are regulated by identical processes (see Fig. 1C and legend). These combined data sets demonstrate that fluxes of sucrose uptake and intracellular concentrations of sucrose (Fig. 5A) and hexoses (Fig. 5B) are related inversely by curvilinear relationships. An apparent dependence of sucrose transport activities upon intracellular sugar concentrations declined at progressively higher concentrations. These relationships were fitted by second-order polynomials showing that sucrose uptake fluxes were correlated more strongly with intracellular concentrations of hexoses (r=0.94) than sucrose (r=0.82).

Relationship between sucrose transporter expression and pool sizes of intracellular sugars

Tests were carried out to determine possible relationships between concentrations of intracellular sugars and transcript levels of PsSUT1, PsSUF1, and PsSUF4 across the developmental window of 16–24 DAF (see Fig. 1B). Using similar assumptions to those described for compiling Fig. 5, data sets for RR and rr cotyledons were combined (Fig. 6). Relationships between transcript levels and concentrations of intracellular sugars differed substantially between the sucrose transporters (Fig. 6). Expression of the symporter, PsSUT1, fitted a linear relationship and was negatively correlated with increasing concentrations of intracellular sucrose (r=0.89) and hexoses (r=0.94 and see Fig. 6A and D, respectively). In contrast, the pea sucrose facilitators were related positively with (PsSUF4; Fig. 6C, F), or independent of (PsSUF1; Fig. 6B, E), intracellular sugar concentrations. Expression of PsSUF4 exhibited a linear relationship, and intracellular concentrations of sucrose and hexoses accounted for 75% and 94%, respectively, of the observed variation in its transcript levels (see Fig. 6 legend).

Fig. 6.

Relationship between relative expression of PsSUT1 (A, D), PsSUF1 (B, E), and PsSUF4 (C, F) and intracellular concentrations of sucrose and hexoses in RR (filled diamonds) and rr (open diamonds) cotyledons harvested at 16, 20, and 24 DAF. Transcript levels of each sucrose transporter determined from total RNA extracted from one sister cotyledon. The relative expression, by semi-quantitative RT-PCR, of each sucrose transporter was normalized to expression of the PsGAPC1gene. Corresponding sister cotyledons were used to measure intracellular sugar concentrations in protoplasmic saps extracted from freeze-dried cotyledons. The RR and rr data sets were combined and fitted by linear regressions. The slope of each fitted line was tested for significance from zero. Equations for the straight lines and their associated regression coefficients and probability values for statistically significant slopes are as follows. PsSUT1 versus sucrose: y= –0.160x+58.138, r²=0.800, P <0.01. PsSUT1 versus hexose: y= –10.591x+74.353, r²=0.870, P <0.01. PsSUF1 versus sucrose: y=0.0367x+4.210, r²=0.383, P >0.05. PsSUF1 versus hexose: y= –3.393x+29.833, r²=0.326, P >0.05. PsSUF4 versus sucrose: y=0.241x–9.245, r²=0.569, P <0.05. PsSUF4 versus hexose: y=18.833x–47.748, r²=0.874, P <0.05. Vertical and horizontal bars represent standard errors of means derived from a minimum of six replicates per treatment.

In vitro cotyledon culture demonstrated that intracellular sucrose is the probable signal

Pod culture was used to approximate in vivo supply of all nutrients, except sucrose, to developing seeds. To this end, peduncles of excised pods were immersed in MS culture medium with or without 200 mM sucrose. After 4 d of culture, intracellular sucrose concentrations in RR cotyledons were reduced substantially (data not shown) in both treatments but significantly more so for cotyledons of pods cultured in the absence of sucrose (Table 1). In contrast, and unexpectedly, concentrations of intracellular hexoses (both glucose and fructose) were identical between treatments (Table 1). Sucrose transport by treated RR cotyledons was partitioned between DEPC-sensitive (symporter) and -insensitive (facilitator) components (Zhou et al., 2007). Overall fluxes of sucrose uptake were 2-fold higher for cotyledons harvested from explants cultured in the absence of sucrose from the culture medium (Table 1). This response primarily resulted from a substantial increase in transport activity of the DEPC-sensitive component with a small increase in the DEPC-insensitive component (Table 1).

Table 1.

Intracellular sugar levels of, and sucrose uptake by, RR pea cotyledons excised from pods cultured on medium with or without 200 mM sucrose

Parameter measured	Sucrose concentration (mM) added to MS medium for pod culture
	0	200
Intracellular concentration (mM) of:
Sucrose	14.1±1.7	62.1±11.7
Glucose	3.0±0.1	3.0±0.3
Fructose	0.3±0.1	0.3±0.0
Transporter activity (μmol cm−2 h−1)
Sucrose facilitators	0.12±0.01	0.07±0.01
Sucrose/proton symporters	0.44±0.04	0.17±0.01

Pods were cultured under sterile conditions for 4 d following their harvest at 14 DAF. Intracellular sugar concentrations were determined in extracted protoplast saps from freeze-dried cotyledons. Influx of [¹⁴C]sucrose uptake across the abaxial epidermis was determined for sister cotyledons pre-treated with ±5 mM DEPC to estimate activities of sucrose/proton symporters (total flux minus DEPC-insensitive flux) and sucrose facilitators (DEPC insensitive flux). Means ±SE of four replicates per treatment.

These findings were examined further by culturing excised RR cotyledons for 2 d on different MS media each containing a specified sugar species at a concentration of 300 mM. The sugar species were sucrose, palatinose, glucose, and fructose. Palatinose is a membrane-impermeant sucrose analogue that, in some plant systems, is perceived by putative sucrose receptors embedded in plasma membranes (Atanassova et al., 2003, and references cited therein). Intracellular sucrose and hexose concentrations were identical for cotyledons cultured on palatinose and in the absence of any sugar, and were significantly lower than those concentrations found in cotyledons cultured on any of the remaining sugars (Table 2). Not surprisingly, intracellular sucrose concentrations were some 3–6 times higher in cotyledons cultured on a sucrose medium compared with cotyledons exposed to the other sugar treatments. Differences in intracellular hexose concentrations were less marked between treatments, with the highest concentrations found in cotyledons exposed to the corresponding hexose (Table 2).

Table 2.

Intracellular sugar levels of, and sucrose uptake by, 18 DAF RR pea cotyledons cultured for 2 d on specified soluble sugar species

Parameter measured	300 mM sugar added to MS culture medium
	Sucrose	Palatinose	Glucose	Fructose	No sugar
Intracellular concentration (mM) of:
Sucrose	36.8±1.8	9.2±1.2	12.4±0.8	12.1±1.1	6.4±0.7
Glucose	8.2±0.3	5.3±0.3	12.0±0.7	7.8±0.7	4.4.±0.2
Fructose	6.1±0.3	3.2±0.2	4.7±0.3	9.7±0.6	2.6±0.5
Transporter activity (μmol cm−2 h−1) of:
Sucrose facilitators	0.45±0.09	0.44±0.13	0.45±0.10	0.50±0.12	0.40±0.09
Sucrose/proton symporters	0.16±09	0.33±08	0.33±10	0.27±0.05	0.31±0.08

Intracellular sugar concentrations were determined in protoplast saps extracted from freeze-dried cotyledons. Influx of [¹⁴C]sucrose uptake across the abaxial epidermis was determined for sister cotyledons pre-treated with ±5 mM DEPC to estimate activities of sucrose/proton symporters (total flux minus DEPC-insensitive flux) and sucrose facilitators (DEPC insensitive flux). Means ±SE of four replicates per treatment.

Transport rates by sucrose symport and facilitated diffusion in cultured cotyledons were distinguished by the known selective sensitivity of symporter function to the membrane-impermeant sulphydryl reagent, DEPC (Zhou et al., 2007). Facilitator activity (DEPC-insensitive) appeared to be positively related to intracellular fructose levels. In contrast, symporter activity was depressed substantially (∼3-fold) by elevated intracellular levels of sucrose and was identical for the remaining sugar treatments including palatinose (Table 2). These sugar effects on transporter activity were explored further at transporter transcript and protein levels. Consistent with symporter activity (Table 2), PsSUT1 transcript and protein levels were significantly depressed in cotyledons cultured on sucrose. In contrast, transcript and protein levels detected in cotyledons exposed to the other sugar treatments were not significantly different from cotyledons cultured in the absence of sugars (Fig. 7A and D, respectively). These data point to an intracellular sucrose signal acting on sucrose symporter activity through control at the transcriptional level. Expression levels of the sucrose facilitators in response to intracellular sugars are less clear. Broadly, PsSUF4 transcript levels responded positively to all sugars whilst PsSUF1 expression responded similarly except that fructose had no effect (Fig. 7C and B, respectively). However, these effects were not detected at the protein level (Fig. 7F and E, respectively).

Fig. 7.

Effect of culturing RR cotyledons on 300 mM concentrations of specific sugars on transcript (A, B, C) and protein (D, E, F) levels of PsSUT1 (A, D), PsSUF1 (B, E), and PsSUF4 (C, F). Cotyledons were excised at 18 DAF and aseptically cultured for 2 d separately on medium containing a specific sugar. Thereafter, total RNA and protein were extracted separately from paired batches of sister cotyledons for determination of transporter transcript and protein levels. The relative expression, by semi-quantitative RT-PCR, of each sucrose transporter was normalized to expression of the PsGAPC1gene. The abundance of SUT/SUF protein was determined by western blotting with affinity-purified antibody. A representative immunoblot of specific SUT/SUF protein for each sugar treatment is shown on top of each corresponding histogram. A total of 70 μg of microsomal protein extract was loaded on each lane. The band intensity of immunoblots was analysed and normalized to the level of PsSUT1 from one replicate of the glucose treatment (this replicate was set as 100) by the Quantity One software in GelDoc system (BioRad). Vertical bars represent standard errors of means derived from three replicate measures for each sugar treatment. Suc, sucrose; Glu, glucose; Fru, fructose; Pal, palatinose; NS, no sugar added to the medium.

Discussion

The current study is set in the context of biomass gain by developing seeds being sink limited during their storage phase (Borrás et al., 2004). In developing pea cotyledons, sink limitation is imposed by processes responsible for starch and storage protein accumulation (Wang and Hedley, 1991). Imported sucrose contributes carbon skeletons for the biosynthesis of these storage compounds (Patrick and Offler, 2001). To manipulate sink demand of pea cotyledons independently of sucrose supply, advantage was taken of a reduced capacity for starch biosynthesis conferred by an insertion in the r locus that knocks out one isoform of starch branching enzyme (Bhattacharyya et al., 1990). A reduced demand for carbon skeletons (Hedley et al., 1986) by impaired starch biosynthesis in rrRbRb cotyledons was verified based on measures of absolute growth rates as surrogate measures of metabolic demand in a sink-limited system (see Fig. 1A). The metabolic lesion was associated with elevated intracellular levels of sucrose and hexoses (Fig. 5), indicating that slowed sucrose metabolism was the primary effect regulating a reduced biomass gain by the cotyledons (Fig. 1A). These observations provide a framework to explore the hypothesis that cotyledon sink demand for carbon skeletons regulates activities of sucrose transporters by transcriptional control mediated by intracellular sugars. The analysis begins by exploring contributions of sucrose transporters to cotyledon biomass gain.

Sucrose transporter activity is linked with biomass gain

Temporal changes in absolute growth rates of cotyledons exhibited typical bell-shaped curves. Absolute growth rates of RRRbRb cotyledons exceeded those of rrRbRb cotyledons following depletion of endosperm reserves until the approach of seed maturation from 30 DAF onward (Fig. 1A). The differences in absolute growth rates correspond to the major phase of storage product accumulation (Weber et al., 2005), consistent with differing capacities for starch biosynthesis between the two genotypes (Wang and Hedley, 1991; Hedley et al., 1994).

Sucrose, released from seed coats, is taken up from the seed apoplasm by abaxial epidermal transfer cells, subepidermal cells, and one or two outer layers of underlying storage parenchyma cells of pea cotyledons (Tegeder et al., 1999; Rosche et al., 2002). That these transporters play a central role in carbon accumulation by cotyledons was demonstrated by their activities accounting for 90% of temporal variation in biomass gain by both rr and RR cotyledons (Fig. 1C). Similar relationships have been reported for French bean cotyledons exhibiting genotypic differences in their growth rates (Tegeder et al., 2000) and for temporal profiles of broad bean cotyledons (Harrington et al., 1997). Importantly for the present study, this finding strongly suggests that processes regulating the relationship between sucrose uptake by, and growth of, cotyledons are identical for RR and rr isolines across their development. The rr data occupied lower levels of the common response surface compared with that for RR cotyledons. In this context, increased storage protein accumulation by rr cotyledons (Perez et al., 1993), presumably initiated by elevated intracellular sucrose levels (Rosche et al., 2005), must have exerted little influence relative to reduced starch biosynthesis on overall cotyledon biomass gains.

Since identified differences in starch biosynthesis between the pea isolines (Wang and Hedley, 1991) contribute to observed variation in their cotyledon growth rates (Fig. 1A), the strong correlation (Fig. 1C) provides compelling evidence that fluxes of sucrose uptake are tightly regulated by metabolic demand for sucrose. The extrapolated intercept of 4.1 mg d⁻¹ growth rate, when sucrose uptake is zero (Fig. 1C), provides an estimate of the averaged contribution by other nutrients to cotyledon biomass gain. Thus for wild-type cotyledons, growing at their maximal rates (i.e. 13.8 mg d⁻¹; see Fig. 1B), sucrose uptake is estimated to contribute 70% to biomass gain, a value comparable with independent measures of this contribution (Patrick and Offler, 2001). Thus the claimed functional significance of sucrose transporters in determining cotyledon growth rates is consistent with an enhanced biomass gain exhibited by pea cotyledons expressing a potato sucrose symporter in their storage parenchyma cells (Rosche et al., 2002).

Deduced functional roles of known cotyledon sucrose transporters from their expression profiles

To date three sucrose transporters are known to be expressed in developing pea cotyledons. These are a proton symporter (PsSUT1) and two sucrose facilitators (PsSUF1 and PsSUF4; Zhou et al., 2007). The symporter could account for the saturable component of sucrose uptake and the facilitators the non-saturable component (Zhang et al., 2007; Zhou et al., 2007). Potential contributions by the symporter and two facilitators to total sucrose uptake were deduced from their relative levels of expression (Fig. 2).

The positive relationship between transcript levels of PsSUT1 and sucrose transport (Fig. 2A) suggests a primary regulatory role for PsSUT1 in sucrose uptake by cotyledons. Similar conclusions have been drawn for developing cotyledons of Faba (Weber et al., 1997), broad (Harrington et al., 1997) and French (Tegeder et al., 2000) bean. In contrast, expression levels of the sucrose facilitator genes (PsSUF1 and PsSUF4) correlated negatively with sucrose uptake activities (Fig. 2B, C). These negative relationships mirror that found for a sucrose-binding protein (a putative sucrose facilitator; Overoorde et al., 1996) expressed in French bean cotyledons (Tegeder et al., 2000). Interestingly, PsSUT1 and PsSUF4 transcript levels share identical response surfaces in rr and RR cotyledons (Fig. 2A). This is consistent with expression of each transporter responding to the same regulatory signal(s) in the cotyledons of both isolines (see also Fig. 1C). This does not appear to apply to regulation of PsSUF1 expression that exhibited an isoline-specific relationship (Fig. 2B) and hence this gene's product may play a lesser role in regulating sucrose uptake. Overall, higher fluxes of sucrose uptake in RR compared with rr cotyledons are linked with higher expression of PsSUT1 and lowered expression of PsSUF4.

The physiological significance of whole cotyledon expression depends upon cellular localization patterns, with cells proximal to the abaxial surface accounting for 70% of sucrose influx (Tegeder et al., 1999; Rosche et al., 2002). In this context, the highest transcript densities of each transporter gene were found to co-localize to several concentric layers of abaxial cells (Figs 3, 4). Within an isoline, PsSUT1 and PsSUF1 expression exhibited comparable spatial distribution patterns. Greater expression levels in RR cotyledons (Fig. 2A, B) are accounted for by higher transcript densities in the abaxial cell layers, with lower levels of expression extending into the underlying storage parenchyma cells (Figs 3, 4A–C). In contrast, PsSUF4 expression was restricted to outer cell layers and appeared to be most strongly expressed in cells of rr cotyledons (Fig. 4C, F). Together, these observations suggest that the depressed sucrose transport activity by rr cotyledons (Fig. 1C) probably results from lower symporter and higher facilitator (PsSUF4) expression in outer cell layers and an absence of capacity to retrieve sucrose leaked from storage parenchyma cells (Rosche et al., 2002).

Sugars regulate sucrose transporter activity at the transcriptional level

Altered intracellular levels of soluble sugars resulting from a block in starch biosynthesis imposed by the rugosus mutation, r (Bhattacharyya et al., 1990), were used to test the hypothesis that sugar signals functioned to co-ordinate the sugar flux into starch biosynthesis with sucrose transporter activities retrieving sucrose from the seed apoplasm.

Sucrose uptake activity by, and intracellular sucrose and hexose concentrations of, cotyledons of RR and rr peas exhibited inverse relationships, with rr cotyledons exhibiting lower fluxes of sucrose uptake at elevated intracellular sugar levels (Fig. 5, and see Edwards and ap Rees, 1986). Previously, such dose–response relationships have been reported only for extracellularly delivered sugars in two experimental systems—sugar beet leaves (Chiou and Bush, 1998; Vaughn et al., 2002) and vegetative peach-tree buds (Maurel et al., 2004). In sugar beet leaves, sucrose delivered in the transpiration stream was shown to exhibit a selective and concentration-dependent inhibition of sucrose symporter activities (Chiou and Bush, 1998; Vaughn et al., 2002). Similarly, sorbitol uptake activity of vegetative peach-tree buds was progressively slowed when bud tissues were pre-treated with increasing concentrations of glucose or fructose (Maurel et al., 2004). The dose–response relationships approached linearity for sugar beet leaves up to an external sucrose concentration of 250 mM (Chiou and Bush, 1998; Vaughn et al., 2002). In contrast, curvilinear relationships were found for glucose inhibition of sorbitol uptake by peach buds saturating at 3 mM glucose (Maurel et al., 2004). Similar curvilinear kinetics were discovered for intracellular sugars in pea cotyledons where inhibition saturated at 6 mM hexose and 250 mM sucrose (Fig. 5).

Transcript abundance of known sucrose transporters in pea cotyledons (Zhou et al., 2007) was used as a surrogate for their activities to dissect out transporter-specific responses to intracellular sugar concentrations. The findings revealed a spectrum of responses of transporter expression to sugar concentrations, and these were identical for intracellular sucrose and hexoses (Fig. 6). Expression of PsSUT1 exhibited a linear decrease in transcript abundance across a 5-fold range of increasing intracellular sugar concentrations (Fig. 6A, D). Sugar repression of similar magnitudes has been reported for VfSUT1 in Faba bean cotyledons cultured for several days on medium containing high (150 mM) concentrations of sucrose or glucose (Weber et al., 1997) and for BvSUT1 of sugar beet leaves exposed to 250 mM sucrose (Chiou and Bush, 1998; Vaughn et al., 2002). Expression of CitSUT1 was lowered in citrus leaf discs floated on solutions containing 100 mM glucose or sucrose (Yao et al., 2003). In contrast, expression of StSUT1 was unaffected by placing EDTA-treated petioles of excised potato leaves in a 100 mM sucrose solution (Barker et al., 2000). The latter finding could repay revisiting as the EDTA treatment might have resulted in sucrose export matching import through the xylem transpiration stream such that leaf sucrose levels remained unchanged. Overall, the current body of evidence indicates that expression of sucrose symporters belonging to the SUT1 clade (Lalonde et al., 2004) responds negatively to increasing concentrations of sucrose (Chiou and Bush 1998; Vaughn et al., 2002) and/or its hexose moieties (Weber et al., 1997; Yao et al., 2003).

Transcript abundance of pea sucrose facilitators was either unresponsive to altered cotyledon sugar concentrations (PsSUF1; Fig. 6B, E) or positively correlated with increasing sugar concentrations (PsSUF4; Fig. 6D, F). The latter relationship has been observed by sucrose symporters belonging to the Clade 2 symporter family. These transporter genes appear to be specifically up-regulated by sucrose in leaves of maize (ZmSUT1; Aoki et al., 1999) and tomato (LeSUT2; Barker et al., 2000), but in rice embryos (OsSUT1; Matsukura et al., 2000) expression responds equally to sucrose and glucose. In contrast, when citrus leaf discs were floated on sugar solutions, CitSUT2 expression remained unaltered (Yao et al., 2003).

The dual responses of sucrose transporter expression in Faba bean (Weber et al., 1997) or pea (present study) cotyledons to sucrose and hexoses could result from a number of causes. Glucose could be the active signal (Rolland et al., 2006) and the link with sucrose is simply as a substrate supplying glucose moieties through intracellular hydrolysis by invertase or sucrose synthase. Alternatively, sucrose could be the active signal (Rolland et al., 2006) and correlation with hexose concentrations reflects the sucrose pool size available for hydrolysis (present study) or externally supplied glucose is converted into sucrose through the SPS pathway (Weber et al., 1997). Moreover, sugars could be sensed at cotyledon plasma membranes or intracellularly (Rolland et al., 2006). To distinguish these possibilities, and to verify causality of observed correlations between intracellular sugar concentrations and sucrose transporter activities, in vitro studies were undertaken with RR peas.

Pod explants provided evidence that sucrose was the probable sugar signal. Here enhanced sucrose uptake by cotyledons was linked with lowered intracellular sucrose concentrations whilst intracellular hexose concentrations remained unaltered (Table 1). Furthermore, in large measure, reduced sucrose uptake by RR cotyledons was accounted for by a commensurate reduction in DEPC-sensitive transport (Table 1) that is mediated by sucrose symporters (Zhou et al., 2007). Identical responses were obtained by culturing excised cotyledons on a range of sugar species, verifying that sucrose functioned as a regulatory signal of symporter activity (Table 2); an effect clearly mediated by repression of PsSUT1 transcription (Fig. 7A, D) by intracellular sucrose concentrations (Table 2). This response is consistent with findings reported for BvSUT1 in sugar beet leaves (Chiou and Bush 1998; Vaughn et al., 2002; Ransom-Hodgkins et al., 2003) but contrasts with a putative regulatory control of CitSUT1 expression in citrus leaves through the hexokinase pathway (Yao et al., 2003). Sucrose repression of symporter activity was accompanied by a small stimulatory effect of sugars (sucrose and/or hexoses) on sucrose facilitator activity (Table 2 and note DEPC-insensitive responses to the various sugar species) and on transcript levels (Fig. 7B, C). However, these effects were not reflected at a protein level (Fig. 7E, F) for which there is no obvious explanation. Significantly, conclusions drawn from the present in vitro studies are consistent with the in planta correlative observations (Figs 5, 6). This suggests that the regulatory mechanisms proposed above are likely to operate within intact pea cotyledons.

The sucrose-specific responses detected in the pod and cotyledon culture experiments could have resulted from sucrose interacting with receptors localized at the plasma membrane or intracellularly (Rolland et al., 2006). These two possibilities were distinguished by culturing cotyledons on medium containing palatinose, a non-metabolizable and non-transportable analogue of sucrose (Atanassova et al., 2003, and references cited therein). A comparable concentration of palatinose was found not to exert any inhibitory effect on sucrose uptake (Table 2) or PsSUT1 expression (Fig. 7A, D) by RR cotyledons. This suggests that the sucrose signal responsible for controlling sucrose transporter activity is unlikely to be sensed at the plasma membrane as found for part of the complex regulating expression of VvHT1 in cell suspension cultures (Anatossova et al., 2003). There is only fragmentary information accounting for the intracellular sensing and transduction of sucrose signals. Elements of this signal cascade may include protein phosphorylation/dephosphorylation (Ransom-Hodgins et al., 2003) and an SNF-like kinase (Tiessen et al., 2003).

Conclusions

Metabolic demand for sucrose by starch and protein synthesis in developing pea cotyledons appears to be transmitted to cotyledon sucrose symporters as an intracellular sucrose signal that negatively regulates sucrose symporter activity of PsSUT1 at the transcriptional level. These sucrose signals must be distinguished from those regulating the activity of the starch biosynthetic machinery that, in contrast to sucrose transporters, responds positively to sucrose levels reaching the cotyledons (Weber et al., 2005). The latter sucrose signals might be sensed at the cotyledon plasma membranes (Tiessen et al., 2003).