Introduction

Sucrose is the major product of photosynthesis in plants. The carbon of CO2, fixed via the Benson–Calvin cycle in photosynthetic tissues, is exported from chloroplasts as triose phosphates and converted into hexose phosphates in the cytoplasm. A major portion of the hexose phosphates is used for sucrose synthesis. As a non-reducing disaccharide, sucrose is chemically inert when in contact with proteins. This characteristic, together with its high solubility in aqueous solution and its ability to retain a high free energy of hydrolysis in its glycosidic bond (Avigad and Dey 1997), make sucrose a superior molecule for transporting fixed carbon from photosynthetic source tissues to sink tissues, a major energy-provider for metabolism, and carbon substrates for the synthesis of biomolecules in many plants. At present, only four enzymes are known to participate directly in sucrose synthesis and cleavage. These are sucrose-phosphate synthase (SPS), sucrose-phosphate phosphatase (SPP), invertase, and sucrose synthase (SuS). SPS coupled with SPP is responsible for the irreversible sucrose synthesis from UDP-glucose and fructose 6-phosphate. Invertase irreversibly hydrolyzes sucrose into glucose and fructose, while SuS catalyzes the reversible conversion of sucrose and UDP into fructose and UDP-glucose. The activities of these enzymes are finely tuned through regulation of their gene expression at different levels and modulation of enzymatic activities by metabolites (Winter and Huber 2000). Each enzyme has individual and important roles in plant growth and development (Koch 2004). In this study, we focus on SuS, which is usually abundant in sink tissues and performs activities that have been linked to many physiological processes including cellulose and starch biosynthesis (Fujii et al. 2010; Munoz et al. 2005; Song et al. 2010; Tang et al. 2009), phloem transport (Koch et al. 1992; Wang et al. 1994), establishment of nitrogen fixation, and responses to abiotic stress (Barrero-Sicilia et al. 2011; Curatti et al. 2006; Gordon et al. 1999; Hatmi et al. 2014; Wang et al. 2014, 2015).

SuS has been widely investigated in different plants, and was recently found in several bacteria (Curatti et al. 2000; Diricks et al. 2015; Porchia et al. 1999; Wu et al. 2015). However, the structure–function relationships of the enzyme were seldom reported until the crystal structures of Arabidopsis sucrose synthase 1 (AtSus1) were resolved (Zheng et al. 2011). The AtSus1 structure reveals the features involved in the active site and interactions with cellular targets. In this study, to gain insight into the amino acid residues involved in substrate binding and catalysis of rice sucrose synthase (RSuS), the amino acid sequences of several glycosyltransferases that use UDP-glucose or ADP-glucose as the glucosyl donor were aligned to find conserved regions. The functions of residues in a conserved motif E-X7-E were then studied by site-directed mutagenesis, substrate binding assays, and homology modeling. The results of this study provide new insights regarding the substrate binding mechanism of this enzyme and its role in controlling metabolic flux.

Materials and methods

Plasmid construction and site-directed mutagenesis

Plasmid pPICZA-RSuS3 was constructed by inserting the coding region of rice RSuS3 cDNA into the XhoIX/SacII site of the expression vector pPICZA. This plasmid was then used for expression of the wild-type recombinant RSuS3 in Pichia pastoris and as a template in site-directed mutagenesis, which was performed using the QuickChange site-directed mutagenesis kit (Stratagene, Agilent Technologies, Santa Clara, CA, USA) and mutagenic primer pairs (Supplementary Table 1).

Expression and purification of the recombinant proteins

Pichia pastoris strain X-33 was transformed with plasmid pPICZA-RSuS3 or plasmids derived from it by site-directed mutagenesis. The transformed strains were grown on BMGY medium [0.1 M potassium phosphate, pH 6.0, 1 % yeast extract, 2 % peptone, 1.34 % yeast nitrogen base (YNB), 4 × 10−5 % biotin, and 1 % glycerol] at 30 °C until the A600 reached 2–6. The cells were harvested by centrifugation, suspended in BMMY medium (0.1 M potassium phosphate, pH 6.0, 1 % yeast extract, 2 % peptone, 1.34 % YNB, 4 × 10−5 % biotin, and 0.5 % methanol) to an A600 of 1.0, and then grown at 30 °C for 96 h. After centrifugation, the cells were suspended in breaking buffer (50 mM Tris–HCl, pH 7.5, 1 mM EDTA, and 5 % glycerol), then disrupted with a French Press. The cell lysate was centrifuged at 15,000×g for 15 min. The proteins in the resultant supernatant were precipitated by 60 % saturation of ammonium sulfate followed by centrifugation. The precipitates were dissolved in Tris-7.5 buffer (50 mM Tris–HCl, pH 7.5) and dialyzed against the same buffer. After dialysis, the enzyme solution was applied onto a DEAE-Sephacel column (2.6 × 15 cm). The column was first eluted with Tris-7.5 buffer, then with a linear gradient of NaCl (0–0.3 M in Tris-7.5 buffer). The fractions containing SuS activity were collected and applied onto a Sephacryl-S300 column (2.6 × 90 cm) that was pre-equilibrated with Tris-7.5 buffer. The fractions containing SuS activity were collected and stored at 4 °C.

SuS activity assay

The sucrose cleavage activity of SuS was measured in PB-6.4 (50 mM sodium phosphate, pH 6.4) buffer containing 100 mM sucrose and 0.1 mM UDP. The produced UDP-glucose was determined by a UDP-glucose dehydrogenase coupling method (Su 1977). The sucrose synthesis activity of SuS was measured in 50 mM Tris-7.5 buffer containing 25 mM fructose, 10 mM UDP-glucose, and 15 mM MgCl2, after which the product sucrose was determined by the anthrone method (Van Handel 1968).

Substrate binding assay

The purified recombinant RSuS3 proteins were incubated with individual substrate containing the corresponding radio-labeled substrate at 37 °C for 1 h. The radio-labeled substrates used in this study included 14C(U)-sucrose (PerkinElmer, San Jose, CA, USA), 14C(U)-UDP-glucose (PerkinElmer), 14C(U)-fructose (American Radiolabeled Chemicals, Inc., St. Louis, MO, USA), and 3H-UDP (American Radiolabeled Chemicals, Inc.). The reaction mixtures were then applied onto Sephadex G-25 columns (0.5 × 10 cm). The protein content and radioactivity of each fraction eluted with PB-7.0 (50 mM phosphate buffer, pH 7.0) were determined by the method of Bradford (1976) and scintillation counting, respectively.

Homology modeling and substrate docking

The molecular modeling of RSuS3 and all simulations were performed using the Discovery Studio 3.5 (DS) program (BioVia, San Diego, CA, USA). The crystal structure of Arabidopsis AtSus1 (PDB ID 3S28) was used as the template (Zheng et al. 2011). A homology model corresponding to the amino acid residues E30 to E810 of RSuS3 was generated using the “Modeller” in DS. The structure was fine-tuned by energy minimization with CHARMM. Protein-substrate docking was carried out with the receptor-ligand docking tool (CDOCKER) in DS.

Results

Sequence alignment of glycosyltransferases in families GT3, GT4, and GT5

SuS is a member of the glycosyltransferase family 4 (GT4) in the Carbohydrate Active Enzymes database (http://www.cazy.org) (Cantarel et al. 2009). According to the stereochemistry of substrates and products, members of GT4 belong to the retaining glycosyltransferases. SPS, starch synthase (SS), and glycogen synthase (GS), belonging to GT3, GT4, or GT5, are also retaining glycosyltransferases and use UDP-glucose or ADP-glucose as a substrate. Multiple sequence alignment of SuS, SPS, SS, and GS from different species showed that several regions and residues were conserved in those enzymes (Supplementary Fig. 1). The most conserved region, corresponding to amino acid residues F671 to 700 of RSuS3, contained an invariant Glu (E678 in RSuS3) and three invariant Gly (Fig. 1). The E-X7-E motif (Cid et al. 2000), corresponding to E678–E686 in RSuS3, was part of this region. As shown in Fig. 1, the first Glu residue in the E-X7-E motif was invariant, while the second Glu can be substituted by Gln or Tyr. The two Glu residues in this motif have been shown to participate in catalysis or to affect the enzyme activity of several enzymes in GT3, GT4, and GT5 (Cid et al. 2000; Kostova et al. 2003; Nichols et al. 2000; Yep et al. 2006).

Fig. 1
figure 1

Alignment of the amino acid sequences of the most conserved region among selected GT3, 4, and 5 glycosyltransferases (sucrose synthase, sucrose phosphate synthase, starch synthase, and glycogen synthase). The alignment was performed using ClustalW and the blosum62mt2 matrix. Identical residues are shown on a black background. The glycosyltransferase families are indicated on the left, and the E-X7-E motif is boxed. HsGy1: glycogen synthase 1 from Homo sapiens (P13807); HsGys2: glycogen synthase 2 from Homo sapiens (P54840); AtSuS1: sucrose synthase 1 from Arabidopsis thaliana (P49040); RSuS3: sucrose synthase from Oryza sativa (Q43009); BvSPS: sucrose phosphate synthase from Beta vulgaris (P49031); ZmSPS: sucrose phosphate synthase from Zea mays (P31927); BoSPS: sucrose phosphate synthase from Bambusa oldhamii (AAR16190); EcGlyA: glycogen synthase from Escherichia coli (P0A6U8); OsGbss1: granule-bound starch synthase from Oryza sativa (B7EHV0); PaGlgA: glycogen synthase from Pyrococcus abyssi (Q9V2J8)

Full size image

Effects of mutations of residues in the E-X7-E motif on the enzymatic activity of RSuS3

To gain insight into the role of the E-X7-E motif in RSuS, site-directed mutagenesis was performed. The roles of the two Glu residues, E678 and E686, were studied by single point mutation to either Asp or Gln. The role of F680 was examined using two mutants, F680S or F680Y. F680S was generated from an unexpected misincorporation of nucleotides during PCR amplification.

The wild-type and mutant recombinant RSuS3 proteins were expressed in P. pastoris and purified to essential homogeneity by ammonium sulfate fractionation, DEAE-Sephacel ion exchange, and Sephacryl S-300 gel filtration chromatographies (Fig. 2). The molecular masses of the wild-type and mutant recombinant proteins, which were estimated by gel filtration chromatography, were in the 336–375 kDa range (data not shown), showing that all of the mutant proteins maintained the tetrameric form, as did the wild-type RSuS3. The relative activities of the purified wild-type and mutant RSuS3 are shown in Table 1. Mutants E678D and E678Q lost almost all activities in both sucrose cleavage and sucrose synthesis, indicating the importance of the carboxyl group and the size of the side chain in this residue. The mutant E686D retained 34.9 and 37.9 % activity in sucrose cleavage and sucrose synthesis direction, respectively; however, the mutant E686Q was inactive, revealing that the contribution of E686 to the enzyme activity also relied on the acidic side chain, but its size may not be as critical as in E678. It is interesting to note that the mutant F680Y had 61 % activity in sucrose cleavage and retained full activity in sucrose synthesis, while F680S lost its activity in both directions. The hydroxyl group of Ser replacing F680 was less likely to account for the activity loss of F680S, since a hydroxyl group is present in a similar location in F680Y. This implied that it was the aromatic characteristics that were essential to the functions of F680.

Fig. 2
figure 2

SDS–polyacrylamide gel electrophoretic analysis of the recombinant wild-type and mutant RSuS3 proteins. The recombinant wild-type and mutant RuS3 proteins, purified by ammonium sulfate precipitation, DEAE-Sephacel ion exchange, and Sephacryl S-300 gel filtration chromatographies. The protein samples were separated on 10 % SDS-polyacrylamide gel, then stained with Coomassie blue R-250. Lane1, molecular mass markers; lanes 2–8, RSuS3, E678D, E678Q, F680S, F680Y, E686D and E686Q

Full size image
Table 1 Relative specific activities of the purified wild-type and mutant RSuS3
Full size table

Effects of E678, F680, and E686 mutations on the substrate binding of RSuS3

To investigate whether the mutations on E678, F680, and E686 affected substrate binding, recombinant wild-type and mutant proteins were subjected to gel filtration chromatography to test their binding preferences to the substrates in both synthesis and cleavage reactions. As shown in Fig. 3, when incubated with individual substrate, the wild-type RSuS3 was able to bind UDP and fructose but was unable to bind UDP-glucose and sucrose. The results showed that UDP and fructose are the leading substrates for the sucrose cleavage and sucrose synthesis reactions catalyzed by the enzyme, respectively. Table 2 shows the summary of the results of substrate binding assays for the mutant RSuS3 proteins incubated with individual substrate. All of the mutants preserved their UDP binding ability. Similar to the wild-type RSuS3, the mutants were unable to bind either sucrose or UDP-glucose in the absence of the corresponding leading substrate. However, the mutants E678D, E678Q, F680S and E686Q lost their fructose binding ability.

Fig. 3
figure 3

Substrate binding analysis of the wild-type recombinant RSuS3 proteins using Sephadex G25 gel filtration chromatography. Purified recombinant wild-type RSuS3 proteins were mixed with individual substrates containing the corresponding radio-labeled substrate, and then separated by Sephadex-G25 gel filtration chromatography. Protein concentrations in each fraction were determined by the Bradford method. The radioactivity (CPM) was determined by scintillation counting

Full size image
Table 2 Summary of the results of substrate binding assays of the wild-type and mutant RSuS3
Full size table

The kinetic parameters of F680Y and E686D, which possessed enzymatic activities, as well as the wild-type RSuS3 in the sucrose synthesis direction, were analyzed (Table 3). The k cat of F680Y and E686D was 1.01-fold higher and 4.4-fold lower than that of the wild-type RSuS3, respectively. The K m for fructose of F680Y and E686D were similar to that of the wild-type enzyme, but the K m for UDP-glucose of F680Y and E686D were 4-fold and 20-fold higher than that of the wild type, respectively, indicating that their affinity for UDP-glucose was decreased.

Table 3 Kinetics parameters of the recombinant wild-type and mutant RSuS3
Full size table

Homology modeling and substrate docking

To interpret the results derived from the site-directed mutagenesis experiments, a homology structure of RSuS3 was built using Arabidopsis AtSus1 as the template (Fig. 4a). The modeled region of RSuS3, E30-E810, corresponded to the amino acid residues E28–D808 of AtSus1 with a sequence identity of 70.7 % and a sequence similarity of 85.4 % (Supplementary Fig. 2). The modeled structure covering residues M280-T757 of RSuS3 displayed a classical GT-B folding structure with two Rossmann-fold domains as those shown in AtSus1 (Zheng et al. 2011). The N-terminal half (280–529) is designated with GT-BN, and the C-terminal half (530–757) is designated with GT-BC. The E-X7-E motif in RSuS3 (678–686) was in the GT-BC domain (Zheng et al. 2011). Based on knowledge gained from AtSus1 in a complex with UDP-glucose, the E-X7-E motif had potential interactions with UDP and UDP-glucose (Fig. 4b). The carboxylate oxygen of E678 of RSuS3 (E675 in AtSus1) acted as a hydrogen bond acceptor to bind the glucose moiety of UDP-glucose. F680 and G681 (F677 and G678 in AtSus1) also participated in the interaction with the glucose moiety of UDP-glucose. The oxygens of the carboxyl group of E686 (E683 in AtSus1) interacted with the ribose of UDP/UDP-glucose through hydrogen bonds. In addition, based on the crystal structure of AtSuS1 in a complex with fructose and UDP, the fructose in the β-furanose form was bound exclusively by residues of the GT-BN domain. It is interesting to note that these observations seemed not to completely agree with those observed in the substrate binding assays of the mutant RSuS3 proteins, which showed that the binding of fructose was affected by mutations at E678, F680, and E686. The modeled RSuS3 structure was further docked with β-fructopyranose (Fig. 5). The β-fructopyranose was employed due to the fact that fructose exists as an equilibrium mixture of β-fructopyranose (70 %), β-fructofuranose (23 %), and smaller amounts of cyclic α-anomers and the open chain in aqueous solutions (Mega et al. 1990). Figure 5 shows one of the docked poses of β-fructopyranose. Residues G306, Q307, R583, and E678 interacted with β-fructopyranose via hydrogen bonds. E678 also hydrogen-bonded with K588 and A679. Residues H441, K588, A679, and F680 provided electrostatic interaction with fructopyranose, while E686 did not directly interact with it. The results suggested that the binding of β-fructopyranose involved residues from both the GT-BN and the GT-BC domains, including those in the E-X7-E motif.

Fig. 4
figure 4

Theoretical structure of RSuS3 constructed by homology modeling using the structure of Arabidopsis AtSuS1 as a template. a Theoretical structure of RSuS3 (E30–E810) based on the known structure of AtSuS1. The GT-BN domain is colored by blue; the GT-BC domain is colored by red while the E-X7-E motif is colored by yellow. b The theoretical structure of RSuS3 was structurally aligned with AtSuS1 and only the residues at and around the substrate binding site are shown. The homology modeling was produced with ESyPred3D and graphed using PyMOL

Full size image
Fig. 5
figure 5

The binding model between RSuS3 and fructopyranose. The molecular modeling of RSuS3 and all simulations were performed using the Discovery Studio 3.5 program. The crystal structure of Arabidopsis AtSuS1 (PDB: 3S28) was used as the template. After sequence alignment, the RSuS3 homology model was performed, followed by energy minimization. Protein-substrate docking was carried out with the receptor-ligand docking tool (CDOCKER). a A 2D diagram is showing the amino acid residues that are in close proximity with fructopyranose. b The amino acids or ligands that form the hydrogen bonds are listed

Full size image

Recently, the crystal structure of SuS from the non-photosynthetic bacterium Nitrosomonas europaea was reported (Wu et al. 2015). Figure 6 shows the bacterial SuS structure docked with β-fructopyranose. The residues E663 and F665, corresponding to E678 and F680 in RSuS3, and other residues in the E-X7-E motif, also participated in β-fructopyranose binding, further supporting the results of this study.

Fig. 6
figure 6

The binding model between NeSuS and fructopyranose. The crystal structure of Nitrosomonas europaea NeSuS1 (PDB: 4RBN) performed as receptor, and protein-substrate docking was carried out with the receptor-ligand docking tool (CDOCKER). a A 2D diagram is showing the amino acid residues that are in close proximity with fructopyranose. b The amino acids or ligands that form the hydrogen bonds are listed

Full size image

Discussion

E678, F680, and E686 are essential for the enzymatic activity of RSuS3

The E-X7-E motif is conserved in the GT3, 4, and 5 families of glycosyltransferases, which adopt GT-B folding and catalyze reactions using a retaining mechanism (Abdian et al. 2000; Cid et al. 2000). Divergent roles of the two conserved Glu residues in the E-X7-E motif have been reported for different enzymes (Absmanner et al. 2010; Nichols et al. 2000; Yep et al. 2006). For example, they are both involved in catalysis in human muscle glycogen synthase (a member of GT3) as a nucleophile and a general acid/base catalyst, respectively (Cid et al. 2000). However, they are not essential for the enzyme activity of yeast bifunctional α-1,3- and -1,6-mannosyltransferase, which belongs to GT4 (Kampf et al. 2009). In this study, we have demonstrated that the two Glu residues, E678 and E686, as well as F680 in the E-X7-E motif of RSuS3, were important for the enzyme function because a single amino acid mutation could cause the activity loss.

In the sucrose synthesis direction, the effects of the mutations on the enzymatic activity of RSuS3 were partly due to changes in enzyme-substrate interactions as revealed by the substrate binding assays (Table 2). Homology modeling and substrate docking analysis showed that both E678 and F680 participated in interactions with β-fructopyranose. Based on the presence of different tautomeric forms of fructose, the results did not contradict those observed in the AtSuS1–UDP–fructose complex, which showed that fructose in the β-furanose form is firmly bound within a pocket formed exclusively by residues from the GT-BN domain (Zheng et al. 2011). Binding of β-fructopyranose, the predominant form of fructose in aqueous solutions, to the enzyme would be followed by tautomerization of the pyranose form to the furanose form because the configuration of fructose in sucrose is the β-furanose form. Therefore, it is reasonable to have some different residues involved in the initial binding of fructose in the pyranose form and the stabilization of fructose in the furanose form in the active site. E678 not only contacted with β-fructopyranose directly but also stabilized the amino acids involved in the hydrogen bonding with β-fructopyranose (Fig. 5). Mutations in this position would result in disruption of the network for stabilization of β-fructopyranose. However, its role in catalysis could not be excluded. This residue is highly conserved in the retaining GT-B glycosyltransferases and was recently proposed to be critical for catalysis in not only SuS but in all of the retaining GT-B glycosyltransferases (Wu et al. 2015).

Unlike the E678 residue that interacted with β-fructopyranose through the side chain, the side chain of F680 residue faced away from the substrate binding pocket and was not directly involved in the electrostatic interaction between this residue and fructopyranose. However, the results derived from kinetic analysis, and substrate binding assay on F680Y and F680S showed that as long as the aromatic group of the side chain was present, the affinity for fructose and the enzymatic activity would not be significantly changed, which highlights the importance of the side chain of this residue. The aromatic side chain of F680 has both polar and nonpolar properties and therefore it probably plays a role in guiding fructose into its binding site. Despite no significant change in the enzyme’s affinity for fructose, the affinity for UDP-glucose was affected by the replacement of F680 with Tyr (Table 3). F680 might also have a role in stabilizing the residues that participated in UDP-glucose binding in accurate positions.

In contrast to E678 and F680, E686 did not build direct interactions with fructose in the modeled RSuS3 structure docked with β-fructopyranose, but it did form a hydrogen bond with the ribose moiety of UDP/UDP-glucose according to the AtSus1 structure. This observation was consistent with the results of a kinetic analysis of E686D, which showed that the affinity for UDP-glucose, but not for fructose, was affected (Table 3). Nevertheless, replacement of E686 with Gln resulted in prevention of fructose binding. The causes for the loss of fructose binding remain unclear but the result, together with the result for F680S, suggested that the binding of fructose to the enzyme was sensitive to changes in the local environment.

In the sucrose cleavage direction, all mutants of E678, F680, and E686 retained their ability to bind the leading substrate, UDP, but the sucrose cleavage activity was affected. Among the three residues, E678 was indispensable in enzymatic activity. This indicated an essential role of E678 in catalysis, as is suggested in the sucrose synthesis direction. The other possible cause of the activity loss of the E678 mutants was that sucrose binding was affected. This was confirmed by substrate binding assays performed in the presence of both UDP and sucrose; the binding of sucrose to the inactive mutants E678D and E678Q was not observed (data not shown). This residue also participated in the binding of the glucose moiety of UDP-glucose and the binding of fructopyranose, as mentioned above. We propose that E678 plays an important role in the stabilization of six-membered ring hexose-containing substrates, which include sucrose, UDP-glucose, and fructose in the pyranose form.

Differing from the sucrose synthesis direction, the Phe-to-Tyr mutation at F680 resulted in a reduction of the sucrose cleavage activity of the enzyme (Table 2). Residues E678–L682 in the E-X7-E motif were in a loop structure linking a β-sheet, followed by a short α-helix structure. Mutations occurring at F680 might cause a change in local conformation, which may in turn affect the substrate orientation and/or catalytic reaction. Although the side chain of E686 participated in stabilizing the ribose moiety of UDP according to the AtSus1 structure, a single mutation at this position would not cause the collapse of interactions between UDP and the enzyme because many residues contribute to the stabilization of UDP. However, mutations occurring at E686 may result in improper positioning of UDP in the active site, so that subsequent catalytic reaction is thus hindered.

The leading substrates of RSuS3-catalyzed reaction in both directions are the aglycone

This study’s other important finding derived from the substrate binding assay was that RSuS3 employed UDP and fructose as the leading substrates in the sucrose cleavage and synthesis reactions, respectively. Both UDP and fructose can be viewed as the aglycone part, or as the acceptor for the glycone part, in a glycosyltransferase-catalyzed reaction. To our knowledge, this is the first report showing that the acceptor substrate binds first in both forward and reverse reactions for enzymes belonging to the GT-B fold family. Our results excluded not only a random mechanism but also a ping-pong mechanism, which might employ sucrose or UDP-glucose as leading substrates for RSuS3-catalyzed reactions.

The substrate-binding order of RSuS3 in the sucrose synthesis direction could be interpreted by the modeled RSuS3 structure established in this study. As mentioned above, E678 and several residues in the GT-BC domain participated in the binding of β-fructopyranose (Fig. 5). These residues also contributed to the stabilization of UDP-glucose according to the AtSus1 structure (Zheng et al. 2011). This suggested that a “UDP-glucose-first” binding would prevent the entrance of fructose. On the other hand, an RSuS3 loaded with β-fructopyranose can allow the following UDP-glucose to be placed in the correct position because the β-fructopyranose would be tautomerized to a furanose form and stabilized in a binding pocket contributed by the residues in the GT-BN domain. It is reasonable to predict that RSuS can distinguish fructopyranose from glucopyranose contained in UDP-glucose; however, the mechanism remains unclear.

Unlike the AtSus1 structure in complex with UDP and fructose or with UDP-glucose, the N. europaea SuS structure, solved without any substrate, is in an open form. It was concluded that fructose binding induces a local conformational change, which in turn promotes the interactions between the two GT-B domains and stabilizes a closed-form structure (Wu et al. 2015). Based on the results of fructopyranose docking analyses, we proposed that binding and tautomerization of the fructopyranose molecule contributed to the induction of local conformational change, which caused the β-fructofuranose to be positioned in the correct site and allowed the second substrate, UDP-glucose, to enter the active site. The presence of two distinct forms of SuS proteins provides further avenues for structure–function studies in RSuS3, in particular, how RSuS3 distinguishes its leading substrates in both directions.

Sequential substrate binding implies a role of RSuS in response to metabolic status

The discovery that fructose and UDP were the leading substrates for sucrose synthesis and cleavage, respectively, of RSuS3 suggested that the activation of the enzyme in vivo is controlled by the availability of fructose and UDP. It also had important implications for RSuS’s physiological role in response to metabolic status. First, in sink tissues, the unloaded sucrose that is transported via sucrose transporter (SUT) on the plasma membrane into the cytosol of the sink cells can be cleaved by cytoplasmic (alkaline/neutral) invertase and/or SuS, and/or can be transported by SUT on the tonoplast into vacuoles for storage, depending on the metabolic status of a given sink tissue in a given plant species. SuS, which is usually abundant in sink tissues, may compete with invertase and SUT for the same substrate sucrose. It has been shown that the UDP levels might be one of the factors that restrict the rate of sucrose degradation in potato tubers (Bologa et al. 2003). Therefore, using UDP as the leading substrate could control the levels of sucrose bound by SuS and thus adjust the proportion of sucrose channeled into the SuS pathway according to the metabolic demand. On the other hand, although it is generally accepted that sucrose in plants is mainly synthesized by SPS, and that SuS is involved in sucrose cleavage in sink cells to provide UDP-glucose and fructose for utilization, the involvement of SuS in sucrose synthesis in vivo has been reported (Geigenberger and Stitt 1993; Nguyen-Quoc and Foyer 2001). Both SPS and SuS use UDP-glucose for sucrose synthesis. UDP-glucose is also a specific substrate for cellulose synthase (CesA), UDP-glucose pyrophosphorylase (UGPase), and UDP-glucose dehydrogenase. Therefore, using fructose as the leading substrate could ensure that SuS binds UDP-glucose only when the SuS pathway is required for the metabolic demand. Moreover, it has been suggested that the reaction of membrane-associated SuS is coupled to the CesA (Amor et al. 1995), and both SuS and UGPase are likely to be metabolically linked, especially in sink tissues (Kleczkowski et al. 2004). Direct interactions between SuS and CesA and between SuS and UGPase have been demonstrated in vitro (Chen 2010). The direction of reaction catalyzed by RSuS can be controlled by the availability of fructose and UDP, depending on the metabolic status of a tissue.