Isolation of sucrose transporter cDNAs
Novel sucrose transporter genes were cloned from the seed coats of pea and bean by degenerate PCR using primers corresponding to two conserved regions of all known SUTs, GVQFGWA and DTDWMGK (Lalonde et al., 2004). An expected fragment of 800 bp was amplified and 40 degenerate PCR clones were sequenced. Six sucrose transporter genes were identified, three from each of pea and bean seed coats. For pea seed coats, the three genes included the previously identified, but functionally uncharacterized, PsSUT1 (Tegeder et al., 1999). Four new genes were isolated in full length by 3′ and 5′ rapid amplification of cDNA ends (RACE) PCR. For each gene, the absence of PCR-introduced mutations was verified by obtaining identical sequences from five independent clones.
Sequence analysis showed that the predicted proteins of the new genes bear all the hallmarks of the SUT family (Figure 1). These include the presence of 12 transmembrane domains and two copies of RXGR motifs, one in loop II/III and one in loop VIII/IX extended as [M/T]-[S]-x(2)-[V/M]-[E]-x-[L]-[G/C]-[R]-x(3)-[G/A] (Figure 1 and see Lemoine, 2000).
Figure 1. Comparison of the predicted amino acid sequences from PsSUF1, PsSUF4, PvSUT1 and PvSUF1 genes. Sequence alignment was performed by ClustalW algorithm. Black shading indicates identical residues and grey shading conserved residues. Transmembrane domains (TM) were predicted by the hidden Markov model. Annotated are TM domains, two R-X-G-R boxes (<<< with one extended as [MT]-[S]-x(2)-[V/M]-[E]-x-[L]-[G/C]-[R]-x(3)-[G/A]), C-termini phosphorylation sites (>>>), SQ/TQ-like motif (∧∧∧) and residues described in the text (*).
Download figure to PowerPoint
The existing nomenclature of SUTs neither rigorously reflects their phylogenetic relationships nor their transport functions. In this context, we adopted the preliminary measure of naming the newly discovered sucrose transporters following the clade system described by Lalonde et al. (2004). In addition, as all characterized SUTs/sucrose carriers (SUCs) function as sucrose/H+ symporters (Lalonde et al., 2004), the acronym of SUT was reserved for this function. This allows distinction from the newly discovered sucrose facilitators (SUFs; and see functional characterization of the transporters described later). Phylogenetic analysis revealed that one sucrose transporter, PsSUF4, was grouped in Clade II and was most closely related to the Lotus japonicus homolog, LjSUT4 (Genbank acc. AJ538041), with 73% identity at the amino acid level (Figure 2). The remaining sucrose transporters are clustered under Clade I (Figure 2), which contains SUT1s (Lalonde et al., 2004). These were PsSUF1 from pea, and PvSUT1 and PvSUF1 from bean seed coats. The nomenclature reflects their transport function as either symporters (SUT) or facilitators (SUF) and a phylogenetic clustering in Clade I with known SUT1s (Figure 2). A partial sequence of another sucrose transporter, putatively PvSUT3, was also clustered in Clade I. However, PvSUT3 was excluded from ongoing study because RT-PCR expression analyses demonstrated that its transcript was localized to flowers (data not shown). PsSUF1, PvSUT1 and PvSUF1 transporters grouped together outside the previously known Fabaceae SUT1 branch containing PsSUT1, GmSUT1 and VfSUT1 (see Figure 2). For instance, pea PsSUF1 is more closely related to bean PvSUF1 than its pea paralog, PsSUT1 (Figure 2). Thus a functional difference may exist between these two pea sucrose transporter paralogs and indeed the two sub-groups of Clade I.
Figure 2. Phylogenetic analysis of PsSUF1, PsSUF4, PvSUT1 and PvSUF1 in relation to other sucrose transporters. The tree was constructed based on maximum likelihood method using a program, TREE-PUZZLE5 (Schmidt et al., 2002). The branch length was computed using the JTT model with rate heterogeneity (Jones et al., 1992). The bar denotes the fraction of expected changes of amino acids described in the model. All sucrose transporter members are light-shaded into three Clades, Clade I–III with PsSUF1, PsSUF4, PvSUT1, PvSUF1 and the sucrose transporters from the legume family highlighted in dark shading. All sequences were obtained from NCBI (http://www.ncbi.nlm.nih.gov). Asarina barclaina: AbSUT1 (AF191024); Apium graveolens: AgSUT1 (AF063400); Alonsoa meridionalis: AmSUT1 (AF191025); Arabidopsis thaliana: AtSUC1 (At1g71880), AtSUC2 (At1g22710), AtSUC3(At2g02860), AtSUT4 (At1g09960), AtSUC5 (At1g71890), AtSUC6 (At5g43610), AtSUC7 (At1g66570), AtSUC8 (At2g14670) and AtSUC9 (At5g06170); Brassica oleracea: BoSUT1 (DQ020217), BoSUC2 (AY065840); Beta vulgaris: BvSUT1 (X83850); Citrus sinensis: CsSUT1 (AY098891), CsSUT2 (AY098894); Daucus carota: DcSUT2 (Y16768); DcSUT1 (Y16767); Euphorbia esula: EeSUT1 (AF242307); Glycine max: GmSUT1 (AJ563364); Hordeum vulgaris: HvSUT1 (AJ272309), HvSUT2 (AJ272308); Lycopersicon esculentum: LeSUT1 (X82275), LeSUT2 (AF166498), LeSUT4 (AF176950); Lotus japonicus: LjSUT4 (AJ538041); Nicotiana tabacum: NtSUT1 (X82276), NtSUT3 (AF149981); Oryza sativa: OsSUT1 (D87819), OsSUT2 (AY137242), OsSUT4(AB091673); Plantago major: PmSUC1 (X84379), PmSUC2 (X75764), PmSUC3 (AJ534442); Pisum sativum: PsSUT1 (AF109922); Ricinus communis: RcSCR1 (Z31561); Spinacia oleracea SoS21 (X67125); Solanum tuberosum: StSUT1 (X69165), StSUT2 (AY291289), StSUT4 (AF237780); Triticum aestivum: TaSUT1a (AF408842), TaSUT1b (AF408843), TaSUT1c (AF408844), TaSUT1d (AF408845); Vicia faba: VfSUT1 (Z93774); Vitis vinifera; VvSUC2 (AF439321), VvSUC27 (AF021810), VvSUC11 (AF021808), VvSUV12 (AF021809); Zea mays: ZmSUT1 (AB008464).
Download figure to PowerPoint
Sequence comparisons with all SUTs available in Genbank detected a number of major differences for the new Clade-I sub-group (Figure 2). These were as follows (see asterisks in Figure 1, unless otherwise stated): (1) the proline, conserved in loop VIII/IX of all known Clade-III/SUT2 members, was conserved across PsSUF1, PvSUT1 and PvSUF1; (2) the uncharged hydrophobic residue, comprising a leucine/isoleucine/valine in transmembrane domain (TM) X of most SUTs, was substituted by a phenylalanine; (3) three of the four cysteines conserved in SUTs (Lemoine, 2000) were detected in PsSUF1 and PvSUF1, with the fourth cysteine, conserved in loop TM II/TM III, replaced by either a serine or an asparagine; (4) unique substitutions were detected in PsSUF1, compared with all other SUTs – a methionine (loop V/VI) replaced either the positively charged lysine or histidine, an asparagine replaced the tyrosine/phenylalanine/leucine in TM II and another asparagine replaced either the positively charged lysine or arginine in the central loop; 5) in contrast to SUT1 from Clade I, a predicted protein kinase C phosphorylation site at the C-termini, conserved in most Clade III (SUT2) and Clade II (SUT4) members, was present in PsSUF1, PvSUF1 and PvSUT1 (Figure 1, indicated as >>>); 6) a SQ/TQ-like motif, Ser-Glu-Ser-Glu-Thr-Glu-Thr-Glu-Thr-Glu-Ser sequence located in the central cytoplasmic loop (loop VI/VII) of PsSUF1 (Figure 1, indicated as ∧∧∧). The SQ/TQ motif is considered to be a signature characteristic of DNA-dependent protein kinases that function as protein-protein interaction modules (e.g. Traven and Heierhorst, 2005).
Functional characterization of SUTs expressed in yeast
The cDNA of each sucrose transporter was cloned into a yeast vector (pDR196) and transformed into the yeast strain SUSY7/ura3, a complementation system for SUTs (Barker et al., 2000; Riesmeier et al., 1992). For each gene, six independent colonies were tested for complementation on sucrose plates. Compared to yeast transformed with empty vector pDR196 alone, expression of PsSUT1 and PvSUT1 allowed growth on a medium containing 2% sucrose as the sole carbon source (see representative growth of a single colony, each expressing a sucrose transporter, in Figure 3a). In contrast, verified incorporated sequences of PsSUF1, PsSUF4 and PvSUF1 (Figure 1 and for more details, see Experimental procedures) did not rescue SUSY7/ura3 on a 2% sucrose medium traditionally used for yeast complementation (e.g. Riesmeier et al., 1992; Weise et al., 2000). Interestingly, when medium sucrose concentration was raised to 4%, SUSY7/ura3 transformed with PsSUF1, PsSUF4 and PvSUF1 exhibited superior growth compared with the empty pDR196 vector (Figure 3b). This finding could prove to be a valuable modification in refining a functional screening system for low affinity sucrose transporters (see apparent Km values reported in Figure 4e,f).
Figure 3. Rescue of yeast strain SUSY/ura3 transformed with specific SUT/SUFs on sucrose as the sole carbon source. (a) PsSUT1 and PvSUT1 supported cell growth on 2% sucrose. (b) PsSUF1, PsSUF4 and PvSUF1 supported cell growth on 4% sucrose. pDR196 is the empty vector. Single colonies, containing each sucrose transporter gene, were streaked onto sections of sucrose plates (ura−) followed by incubation at 30°C and observed after 10 days.
Download figure to PowerPoint
Figure 4. Characterization of sucrose transport function of SUT/SUFs, cloned from pea (a, c, e, g) and French bean (b, d, f, h), and expressed in yeast. (a–d) Time course for uptake from a 0.2 mm [14C]sucrose solution pH 5.5 (solid line, closed symbols) and by cells preloaded with 50 mm sucrose (broken line, open symbols). (e, f) Concentration-dependent uptake at pH 4.5. Inserted tables display apparent Km values. (g, h) pH dependent uptake from a 1 mm sucrose solution. Values are means ± SEs of at least 6 replicates per treatment. All experiments were energised with 10 mm glucose added 1 min prior to treatment, with the exception of PvSUF1, which was energised with 100 mm ethanol. (a, c, e, g) – PsSUT1; – PsSUF1; – PsSUF4; ×– pDR196. (b, d, f, h) – PvSUT1; – PvSUF1; ×– pDR196.
Download figure to PowerPoint
Time course studies of [14C]sucrose uptake by SUSY7/ura3 yeast, expressing PsSUT1, PsSUF1, PsSUF4, PvSUT1 or PvSUF1, showed that each transporter conferred higher rates of sucrose uptake than those cells transformed with empty vector alone (Figure 4a–d and note solid lines). These findings demonstrate that the new sucrose transporter genes, and the previously isolated PsSUT1 (Tegeder et al., 1999), encode functional sucrose transporters. Each transporter supported constant rates of [14C]sucrose uptake over the first four min of exposure to [14C]sucrose. These rates are commensurate with those supported by SUTs from other legume species including VfSUT1 (Weber et al., 1997) and GmSUT1 (Aldape et al., 2003). Thereafter, for PsSUF1, PsSUF4 and PvSUF1, uptake slowed between 6 and 10 min (Figure 4c,d). In contrast, uptake remained constant across the 10-min period for PsSUT1 and PvSUT1 (Figure 4a,b). Thus, all subsequent uptake experiments were conducted over a 4-min period following exposure to [14C]sucrose.
Concentration-dependent uptake of [14C]sucrose by transformed yeast demonstrated that all sucrose transporters exhibited saturating transport kinetics typical of a carrier function (Figure 4e,f). These sucrose uptake data were fitted by Eadie-Hofstee transformations to obtain estimates of apparent Km values that differed markedly between the two Clade I sub-groups (cf. Figure 2). Values obtained for the proposed new Clade I sub-group members were 5.5- (PvSUT1) to 66-(PsSUF1) fold higher than those found for PsSUT1 (Figure 4e,f). In addition, PsSUF4 exhibited an apparent Km value 4-fold larger than those reported for AtSUT4 and StSUT4 (Figure 4e and cf. Weise et al., 2000).
To determine whether sucrose was accumulated against its concentration gradient, comparisons were made between external (C0) and internal (Ci) sucrose concentrations. Assuming that 108S. cerevisiae cells have an intracellular volume of 5 μl (cf. Cirillo, 1989), internal sucrose concentrations were estimated from levels of [14C]sucrose accumulated after 10 min (6 min for PsSUF4 and see Figure 4a–d noting that 98 ± 1% of the accumulated 14C-label remained as [14C]sucrose). At a Co of 0.2 mm sucrose, SUSY7/ura3 cells expressing PsSUT1 and PvSUT1 achieved Ci/Co ratios in excess of one (Table 1a). This outcome indicates that sucrose was accumulated above its external concentration. In contrast, PsSUF1, PsSUF4 and PvSUF1 supported Ci/Co ratios approximating one, showing that equilibria between internal and external sucrose concentrations were reached by 10 min or earlier (Table 1a and cf. Figure 4c,d noting solid lines). The mitochondrial electron transport inhibitor, antimycin A, significantly inhibited [14C]sucrose transport by PsSUT1 and PvSUT1 (Table 1b) whereas PsSUF1, PsSUF4 and PvSUF1 transport functions were unaffected (Table 1b).
Table 1. Sucrose concentration ratios, relative rates of sucrose efflux and effects of inhibitors and competing sugars on sucrose uptake by pea and bean sucrose transporter genes expressed in SUSY7/ura3 yeast
|Estimated Ci/Co ratioa|| 1.81|| 1.03|| 0.73|| 8.28|| 1.18|
|Inhibitor (% control)b|
| 10 μm Antimycin A|| 40*||110||110|| 66*|| 94|
| 50 μm CCCP|| 69*||105|| 92|| 39*||102|
| 1.5 mm DEPC|| 29*||106|| 92|| 59*||100|
|Sucrose remaining (% of preloaded 14C)c|| 92|| 52*|| 39*|| 90|| 42*|
|Competing sugar (% control)d|
| 10 mm glucose||127*||103|| 95||124*|| 59*|
| 10 mm fructose||128*|| 98|| 86||137*|| 66*|
| 10 mm maltose|| 72*||102|| 90|| 58*|| 94|
| 10 mm mannitol|| 96||111|| 97|| 95|| 90|
| 10 mm raffinose||101||104||104|| 93|| 91|
| 10 mm palatinose|| 99||100|| 88|| 98|| 70*|
The prospect of sucrose transport utilizing an ATPase-generated H+ gradient was explored by manipulating the pmf with the protonophore, carbonyl cyanide (CCCP). Treatment with CCCP led to a significant inhibition of [14C]sucrose transport by PsSUT1 and PvSUT1 but had no effect on PsSUF1, PsSUF4 or PvSUF1 (Table 1b). PsSUT1 and PvSUT1 transport activities were strongly pH dependent as shown by progressive increased rates of sucrose uptake when media pH was decreased below 6.0 (Figure 4g) and 8.0 (Figure 4h) respectively. PvSUT1 exhibited a distinct pH optimum at pH 5.0. In contrast, rates of sucrose transport mediated by PsSUF1, PsSUF4, and PvSUF1 were independent of medium pH (Figure 4g,h). Collectively, these data demonstrate that an inward-directed pmf across the yeast plasma membrane is required only for sucrose uptake by PsSUT1 and PvSUT1. As an added marker for sucrose/H+ symporter function (Lalonde et al., 2004), sucrose transporter-transformed yeast was treated with diethylpyrocarbonate (DEPC) which inhibits sucrose/H+ symporter activity by binding with the highly conserved histidine-65 (Lemoine, 2000; Lu and Bush, 1998). PsSUT1 and PvSUT1 were the only transporters that displayed DEPC inhibition of their transport activity (Table 1b).
To test whether the newly-discovered sucrose tranporters supported sucrose efflux, relative amounts of [14C] sucrose lost from [14C]sucrose preloaded yeast cells were determined following a 10-min wash in buffer (Table 1c and noting that 98 ± 1% of the accumulated 14C-label remained as [14C]sucrose). These data demonstrate that PsSUF1, PsSUF4 and PvSUF1 support high rates of sucrose efflux in comparison to the proton symporters, PsSUT1 and PvSUT1. The latter rates of efflux corresponded to those found for yeast cells transformed with empty vector alone (Table 1 caption). The strongest evidence that substrate influx (cf. Figure 4e,f) and efflux are carrier- and not channel-mediated is that, for a carrier, substrate efflux temporarily drives influx against the substrate's electrochemical gradient – a phenomenon referred to as counter transport (Stein, 1986). To test this proposition, transformed yeast cells were preloaded with a high (50 mm) concentration of [12C]sucrose prior to resuspending them in a 0.2 mm [14C] sucrose medium. Under these conditions, if the expressed sucrose transporter functions as a carrier, a transient unidirectional influx of [14C]sucrose will result by counter-transport until equilibrium between [12C] and [14C]sucrose is restored. Consistent with this model, yeast cells expressing PsSUF1, PsSUF4 or PvSUF1, exhibited transient higher uptake rates of [14C]sucrose at two min compared to those exposed to a preload solution containing an equimolar concentration of non-permeating sorbitol. This difference in influx was dissipated by 4 to 10 min (Figure 4c,d noting broken lines). In contrast, a transient burst in [14C]sucrose uptake was not detected for sucrose preloaded cells transformed with PsSUT1 or PvSUT1 (Figure 4a,b noting broken lines).
Substrate specificity of each transporter was determined by sucrose uptake in the presence of selected sugar species supplied at 10x greater concentration. Among tested sugars, maltose inhibited [14C]sucrose uptake mediated by PsSUT1 and PvSUT1 (Table 1d). Glucose and fructose stimulated sucrose uptake by PsSUT1 and PvSUT1, but interestingly these sugars and palatinose competed with sucrose for PvSUF1 (Table 1d). The latter observation is the subject of a separate investigation.
Expression analysis of sucrose transporters in pea and bean
Similar to expression of VfSUT1 (Weber et al., 1997) and PsSUT1 (Tegeder et al., 1999), transcripts of the newly discovered sucrose transporters were detected in all plant organs including source leaves, sink leaves, flowers, stem, roots, cotyledons and seed coats (Figure 5). The most marked features of the organ expression profiles were that PsSUF4 (Figure 5b) was expressed equally in all tissues including seed coats and cotyledons. In contrast, the remaining sucrose transporter genes exhibited organ specific expression. Thus PsSUF1 (Figure 5b) exhibited strongest expression in cotyledons with transcript levels some 3- to 4-fold greater than those detected in seed coats and roots with relatively low levels of expression in the remaining organs. In bean, PvSUF1 expressed most strongly in vegetative tissues (Figure 5a), particularly source leaves and stems, with low levels of expression detected in seed coats and cotyledons. This pattern contrasted with PvSUT1 (Figure 5a) that exhibited strongest expression in seed coats and weakest by three-fold in cotyledons, source leaves and roots. Overall, PsSUF1 and PvSUT1 showed the greatest seed specificity with PsSUF1 and PvSUT1 expression strongest in cotyledons and coats of developing pea and bean seeds respectively.
Figure 5. Expression pattern of specified SUT/SUF in plant organs. (a) PvSUT1 () and PvSUF1 (□) in specified organs of bean. (b) PsSUF1 () and PsSUF4 (□) in specified organs of pea. Relative expression, by semi-quantitative RT-PCR, of each sucrose transporter normalized to expression of bean and pea actin genes. Values are the means ± SE from three separate RNA extractions. ScL, source leaves; SkL, sink leaves; Fl, flower; St, stem; Rt, root; SC, seed coats; Cot, cotyledons.
Download figure to PowerPoint
Cellular localization of transcript was examined in bean and pea seed coats by hybridizing tissue sections (1 μm thick) with DIG-labelled RNA probes specific to each gene. For bean seed coats, PvSUT1 and PvSUF1 were localized to vascular bundles, ground and branch parenchyma cells (Figure 6b,c,e cf. a, d). In contrast, PsSUT1, PsSUF1 and PsSUF4 transcripts were localized to all coat tissues except the palisade and hypodermis (Figure 6f cf. g–i). In the thin-walled parenchyma and thin-walled parenchyma transfer cells the signal was located around the periphery of the cells, consistent with the highly vacuolated nature of these cell types (Patrick and Offler, 2001). Comparison of signal in these two tissues suggests stronger expression in the thin-walled parenchyma transfer cells bordering the seed coat/cotyledon interface. The apparently stronger signal in the chlorenchyma probably reflects their denser cytoplasm.
Figure 6. Cellular localization of SUT/SUF transcripts in seed coats of bean and pea. Light micrographs illustrating in-situ hybridization of DIG-labelled sucrose transporter sense (a, d, f) and anti-sense (b, c, e, g–i) riboprobes to coat transverse sections of developing seeds. (a–e) Bean seed coat sections treated with riboprobes to PvSUF1 (a, b, d, e) and PvSUT1 (c). Signal is localized in vascular, ground parenchyma inward of the vascular bundles and branch parenchyma tissues. (f–i) Pea seed coat sections treated with riboprobes to PsSUT1 (f, g), PsSUF1 (h) and PsSUF4 (i). Signal for all three genes occurs in all tissues except the palisade and hypodermis (g–i cf. f). Bar = 37.5 μm (a–c, f–i). Bar = 150 μm (d, e). bp, branch parenchyma; c, chlorenchyma; gp, ground parenchyma; h, hypodermis; p, palisade; tc, thin-walled parenchyma transfer cells; tp, thin-walled parenchyma; vb, vascular bundle.
Download figure to PowerPoint