BIOTECHNOLOGICALLY RELEVANT ENZYMES AND PROTEINS

The quest for a thermostable sucrose phosphorylase revealssucrose 6′-phosphate phosphorylase as a novel specificity

Tom Verhaeghe & Dirk Aerts & Margo Diricks &

Wim Soetaert & Tom Desmet

Received: 21 November 2013 /Revised: 13 February 2014 /Accepted: 14 February 2014# Springer-Verlag Berlin Heidelberg 2014

Abstract Sucrose phosphorylase is a promising biocatalystfor the glycosylation of a wide range of compounds, but itsindustrial application has been hampered by the low thermo-stability of known representatives. Hence, in this study, theputative sucrose phosphorylase from the thermophileThermoanaerobacterium thermosaccharolyticum wasrecombinantly expressed and fully characterised. The enzymeshowed significant activity on sucrose (optimum at 55 °C),and with a melting temperature of 79 °C and a half-life of 60 hat the industrially relevant temperature of 60 °C, it is far morestable than known sucrose phosphorylases. Substrate screen-ing and detailed kinetic characterisation revealed however apreference for sucrose 6′-phosphate over sucrose. The enzymecan thus be considered as a sucrose 6′-phosphate phosphory-lase, a specificity not yet reported to date. Homology model-ling and mutagenesis pointed out particular residues (Arg134and His344) accounting for the difference in specificity. More-over, phylogenetic and sequence analysis suggest that glyco-side hydrolase 13 subfamily 18 might harbour even morespecificities. In addition, the second gene residing in the sameoperon as sucrose 6′-phosphate phosphorylase was identifiedas well, and found to be a phosphofructokinase. The concertedaction of both these enzymes implies a new pathway for thebreakdown of sucrose, in which the reaction products end upat different stages of the glycolysis.

Keywords Sucrose 6′-phosphate phosphorylase . Sucrosephosphorylase . Thermostability . Glycoside hydrolase familyGH13 . Sucrose metabolism

Introduction

Sucrose phosphorylase (SP, EC 2.4.1.7) is classified in glyco-side hydrolase family 13 and the sole specificity in subfamily18 (GH13_18) (Stam et al. 2006). Its closest phylogeneticneighbours are amylosucrase (EC 2.4.1.4) and sucrose hydro-lase (EC 3.2.1.48), both found in subfamily 4 (GH13_4).Sucrose phosphorylase catalyses the reversible phosphorolysisof sucrose into α-D-glucose 1-phosphate (Glc1P) and D-fruc-tose (Cantarel et al. 2009; Desmet and Soetaert 2011;Doudoroff 1943), but due to its promiscuous nature, SP canalso be used to glycosylate a variety of monosaccharides, sugaralcohols, phenolic compounds and even carboxylic acids (Aertset al. 2011b; Desmet et al. 2012; Goedl et al. 2010). Moreover,it uses a cheap, renewable and abundantly available donorsubstrate, i.e. sucrose, which makes SP a very interesting bio-catalyst for the production of α-D-glucose 1-phosphate and α-glycosides as industrial fine chemicals (Goedl et al. 2010; Skovet al. 2006). A commercial production process is already avail-able for 2-O-(α-D-glucopyranosyl)-sn-glycerol (trade nameGlycoin®, a moisturising agent in cosmetics) (Goedl et al.2008), and, recently, glycosylation efficiencies for phenoliccompounds (e.g. resveratrol) have been greatly improved (DeWinter et al. 2013).

Industrial carbohydrate conversions are preferably per-formed at 60 °C, mainly to avoid microbial contamination(Bruins et al. 2001; Eijsink et al. 2005; Vieille and Zeikus1996). Sucrose phosphorylases, however, have currently onlybeen isolated from mesophilic sources such as Leuconostocmesenteroides, Lactobacillus acidophilus and Bifidobacteriumadolescentis (Aerts et al. 2011b; Goedl et al. 2010). Among

Tom Verhaeghe and Dirk Aerts contributed equally to this work.

Electronic supplementary material The online version of this article(doi:10.1007/s00253-014-5621-y) contains supplementary material,which is available to authorized users.

T. Verhaeghe :D. Aerts :M. Diricks :W. Soetaert : T. Desmet (*)Centre for Industrial Biotechnology and Biocatalysis, Department ofBiochemical and Microbial Technology, Ghent University, CoupureLinks 653, 9000 Ghent, Belgiume-mail: tom.desmet@ugent.be

Appl Microbiol BiotechnolDOI 10.1007/s00253-014-5621-y

these, the SP from B. adolescentis is the most thermostable,with a half-life of 12 h at 60 °C (Cerdobbel et al. 2010). Sincethe limited thermostability of mesophilic isolates could hamperthe commercial exploitation, efforts have been made to im-prove their stability by means of immobilisation (Cerdobbelet al. 2010, 2011b; Goedl et al. 2007) and mutagenesis (Aertset al. 2013; Cerdobbel et al. 2011a; Fujii et al. 2006). Never-theless, the isolation of SP enzymes from thermophilic sourcescould still yield enzymes that are considerably more stable thanthe best variants available today.

A rich source of thermostable disaccharide phosphor-ylases are Thermoanaerobacter species, thermophilicbacteria able to grow on a wide range of carbohydrates.Several phosphorolytic enzymes have already been iden-tified in these organisms, but a sucrose phosphorylase isnotably missing (Chen et al. 2007; Maruta et al. 2002;Yamamoto et al. 2004). Therefore, we have exploredadditional sequence information in public databases toidentify SP enzymes from thermophilic organisms, andrecombinantly expressed a suitable candidate to determineits properties.

Materials and methods

Materials

Unless noted otherwise, all chemicals were obtainedfrom Sigma-Aldrich or Merck and were of the highestpurity. Sucrose and D-fructose were purchased at Sigma-Aldrich and were of ≥99.5 %pure (GC), while α-D-glucose 1-phosphate was produced in-house (DeWinter et al. 2011).

Sequence analysis

All full-length protein sequences classified in subfamilyGH13_18 were extracted from the CAZy database (http://www.cazy.org/; Cantarel et al. 2009) and aligned withClustalO (Sievers et al. 2011) using default parameters. Next,a phylogenetic tree was generated with MEGA5(Tamura et al. 2011) using the neighbour-joining algo-rithm (Saitou and Nei 1987) and default parameters. Forthe evaluation of the gene landscape and the organisa-tion of operons containing a (putative) sp gene, thecomplete genomes of L. mesenteroides ATCC 8293, L.acidophilus NCFM, B. adolescentis ATCC 15703, Strepto-coccus mutans UA159 and Thermoanaerobacteriumthermosaccharolyticum DSM 571 were retrieved from theNational Center for Biotechnology Information genomedatabase (http://www.ncbi.nlm.nih.gov/) and analysed,aided by data from the Prokaryotic Operon DataBase(Taboada et al. 2012).

Cloning and site-directed mutagenesis

The protein sequences of the sucrose 6′-phosphate phosphor-ylase (UniProt ID D9TT09) and phosphofructokinase(UniProt ID D9TT09) from T. thermosaccharolyticum wereback-translated to DNA sequences, codon optimised forEscherichia coli and chemically synthesised by GenScript(Piscataway, NJ, USA) (full sequences are given in Supple-mentary material). They were cloned into the constitutiveexpression vector pCXP34h (Aerts et al. 2011a) by means ofa Gibson assembly procedure (Gibson et al. 2008). First, thegenes and the vector backbone were amplified in a separatehigh-fidelity PCR (PfuUltra™ High-Fidelity, according to theprotocol of Roche) with the respective primers listed inTable S1. PCR products were treated with DpnI (Westburg)to remove template DNA, and were subsequently purifiedusing the Qiagen purification kit, checked on a 1 % agarosegel and the DNA concentration was measured with aNanodrop ND-1000 (Thermo Scientific). For the ligation ofthe two fragments, the Gibson assembly mix (20 μl) contain-ing 100 ng backbone and an equimolar amount of geneproduct was incubated for 1 h at 50 °C. Finally, the resultingexpression plasmids were transformed in E. coli CGSC 8974(Coli Genetic Stock Center, New Haven, CT, USA). Allconstructs were subjected to nucleotide sequencing (AGOWASequence Service, Berlin, Germany) to confirm that the liga-tion was correct and to exclude the presence of undesirablemutations.

Site-directed mutations were introduced with a modifiedtwo-stage megaprimer-based whole-plasmid PCR method(Sanchis et al. 2008) as previously reported (Verhaeghe et al.2013) using the primers described in Table S2.

Characterisation of sucrose 6′-phosphate phosphorylase

For enzyme production, 2 % of an overnight culture wasinoculated in 500-ml LB medium containing 100 μg/ml am-picillin in a 2-L shake flask and incubated at 37 °C withcontinuous shaking at 200 rpm for 8 h. The produced biomasswas harvested by centrifugation for 10 min at 10,000 rpm and4 °C, and the obtained cell pellets were frozen at −20 °C for atleast 4 h. For enzyme extraction and purification, cell pelletswere thawed and dissolved in 20-ml lysis buffer (300 mMNaCl, 10 mM imidazole, 0.1 mM PMSF and 50 mM sodiumphosphate buffer; pH 7.4) supplemented with lysozyme at afinal concentration of 1 mg/ml. This cell suspension wasincubated on ice for 30 min and sonicated three times for2.5 min (Branson Sonifier 250, level 3, 50 % duty cycle).The His6-tagged proteins were purified by Ni-NTA chroma-tography as described by the supplier (Qiagen), after whichthe buffer was exchanged to 50 mM MOPS pH 7.0 in aCentricon YM-30 (Millipore). The protein content wasanalysed by measuring the absorbance at 280 nm. The

Appl Microbiol Biotechnol

extinction coefficients for the His6-tagged proteins were cal-culated using the ProtParam tool on the ExPASy server (http://web.expasy.org/protparam/), as well as the theoreticalmolecular weight. Approximately 10 mg of protein wasobtained from 500 ml of culture medium. Purity andmolecular weight were verified by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE; 10 % gel).

Phosphorylase activity was measured in both directions ofthe equilibrium reaction, using discontinuous assays, as pre-viously described (Aerts et al. 2011b; Cerdobbel et al. 2010).The phosphorolytic and the synthetic reaction were monitoredbymeasuring the production ofα-D-glucose 1-phosphate withan enzymatic coupled assay (Koga et al. 1991; Silversteinet al. 1967) and the release of inorganic phosphate with thephosphomolybdate assay (Gawronski and Benson 2004), re-spectively. Hydrolytic activity was quantified with a discon-tinuous coupled assay using glucose oxidase and peroxidase(GOD-POD) (Blecher and Glassman 1962). Inactivation ofthe samples was achieved by either the acidic conditions of theassay solution (phosphomolybdate assay) or heating for 5 minat 95 °C (other assays).

The specificity toward different monosaccharides(synthesis) was analysed using 100-mM Glc1P as donor and200-mM acceptor, i.e. D-fructose 6-phosphate, D-fructose, D-psicose, D-tagatose, L-sorbose, D-glucose 6-phosphate, D-glu-cose, D-mannose, D-galactose, D-maltose and maltodextrinDE-15 (50 g/l), while phosphorolysis was assayed with 100-mM inorganic phosphate and 100-mM donor, i.e. sucrose 6′-phosphate, sucrose and trehalose 6-phosphate. All reactionswere monitored in 50-mM MES pH 6.5 at 55 °C, for 15 minwith sampling at regular intervals.

The optimum temperature was determined in the phospho-rolytic direction using 350-mM sucrose and 350-mM sodiumphosphate buffer at pH 6.5. The pH optimum was determinedwith acetate (pH 4.5), MES (pH 5–6.5) and MOPS (pH 7–8)buffer at a concentration of 50 mM and a temperature of55 °C. Reactions were again monitored for 15 min withsampling at regular intervals.

The apparent kinetic parameters for sucrose and Glc1P asdonor substrates and for inorganic phosphate and D-fructoseas acceptor substrates were determined at the optimal pH andtemperature. Michaelis–Menten curves were obtained using350 or 100 mM of co-substrate in the phosphorolytic direction(sucrose or inorganic phosphate, respectively) and 100 or200 mM of co-substrate in the synthetic direction (Glc1P orfructose, respectively). The parameters were calculated bynon-linear regression of theMichaelis–Menten equation usingSigma Plot 11.0.

The enzyme’s kinetic stability was examined by incubatingpurified protein (8.5 μg/mL) at 60 °C. At regular time inter-vals, the residual activity was measured and compared to theactivity of the untreated enzyme. The t50 value was calculatedfrom the equation obtained by fitting the experimental data to

the following equation: f=a.exp(−b.x) assuming first-orderinactivation kinetics. The enzyme’s melting temperature wasmeasured using differential scanning fluorimetry (DSF)(Niesen et al. 2007) in a Rotor-Gene Q cycler with HRMchannel (Qiagen). For optimal results, 10 μg of purifiedprotein was used with 1.25 μl SYPRO Orange (400× diluted)(Sigma-Aldrich) in 25 μL. The gain was optimised before thestart of the temperature increase. The temperature increasedfrom 35 to 95 °C, rising 1 °C/min. The fluorescent signal wasdetected at 510 nm with the green detection filter, and theexcitation occurred at 460 nm with an HRM lamp. Themelting temperature (Tm) was calculated from the maximumof the first derivative of themelt curve using the Rotor-Gene Qsoftware (Qiagen).

Production of α-D-glucose 1-phosphate was achievedstarting from 1M of sucrose in 1 or 0.2 M of phosphate bufferpH 7.0 and 0.5 mg/ml enzyme at 60 °C. Formation of α-D-glucose 1-phosphate and decrease of phosphate were measuredwith the assays described above in accordance with De Winteret al. (2011). Production of α-glucosyl glycerol was performedby adding sucrose, glycerol and enzyme in a final concentrationof 0.8 M, 2 M and 0.5 mg/ml, respectively, in 50-mM MOPSpH 7.0 at 60 °C. The reaction was monitored with the BCA-reducing sugars assay and GOD-POD, as well as HPLC, asdescribed by Goedl et al. (2008) and De Winter et al. (2012).

The specific activity of the mutants was determined using100-mM donor (Glc1P) and 50-mM acceptor (fructose orfructose 6-phosphate). Reaction conditions, sampling andsample treatment were the same as for the substrate specificitydetermination.

Characterisation of phosphofructokinase

The phosphofructokinase activity was analysed by high-performance anion exchange with pulsed amperometric detec-tion (HPAEC-PAD) (Dionex ICS-3000 Ion ChromatographySystem (Thermo Scientific), CarboPac PA20 column). All sam-ples were properly diluted and analysed at 30 °C with a fixedflow rate of 0.5 ml/min. Separation of D-fructose, D-fructose 6-phosphate and D-fructose 1,6-bisphosphate was achieved by a 5-min isocratic elution with 40 mM NaOH, followed by a linearincrease of sodium acetate from 0 to 500 mM over 9 min, afterwhich initial conditions were restore in 1 min and were main-tained for 4 min more. The phosphorylation reaction was per-formed with 1-mM ATP and 1-mM fructose or fructose 6-phosphate supplemented with 0.1 mM MgCl in 5 mM MESbuffer pH 6.5 at 55 °C. Samples were taken at regular timeintervals over 1 h, and inactivated at 95 °C for 5 min.

Homology modelling and docking

A homology model of sucrose 6′-phosphate phosphorylasefrom T. thermosaccharolyticum was generated with the

Appl Microbiol Biotechnol

molecular modelling program YASARA (Krieger et al. 2009;Krieger et al. 2002) using default parameters. Crystal struc-tures from the closely related sucrose phosphorylase fromB. adolescentis (40 % sequence identity and 61 % similarity)served as templates (pdb entries 1R7A, 2GDU and 2GDV).Sucrose 6′-phosphate was created with YASARA as well, anddocked in the generated model using the implemented VINAmodule (Trott and Olson 2010) (default parameters, except forthe number of runs which was increased to 50). Figures werecreated with PyMol v1.3 (Schrodinger 2010).

Nucleotide sequence accession numbers

The DNA sequences of the codon-optimised genes have beensubmitted to GenBank and are available under accessionnumbers KF914648 (sucrose 6′-phosphate phosphorylasefrom T. thermosaccharolyticum, TtSPP) and KF914649(Ttpfk, phosphofructokinase).

Results

Selection of a target sequence

To gain more insight in the genetic diversity of SP enzymes, aphylogenetic tree was constructed for all sequences (∼400)classified in family GH13_18 (Fig. 1). These originate from adiverse group of bacteria, such as lactic acid bacteria, soil andmarine bacteria, inhabitants of the gastrointestinal tract and

even cyanobacteria (Table S3). In Archaea or Eukaryota, incontrast, no SP genes have been identified yet. Interestingly,the phylogenetic tree was found to fall apart in two majorbranches, of which only one contains known SP enzymes(Aerts et al. 2011b; Goedl et al. 2010). From the around 150species harbouring a (putative) SP gene, 12 can be designatedas thermophilic (Table 1). Only the sequences from theThermoanaerobacterales Family III (>90 % identical) werefound in the clade with proven activity. Therefore, in thisstudy, the gene from T. thermosaccharolyticum was selectedas a candidate thermostable sucrose phosphorylase.

Substrate specificity and kinetic properties

The purified (putative) SP from T. thermosaccharolyticumdisplayed a single protein band on SDS-PAGE with an esti-mated molecular weight of 56 kDa (theoretical value,56.7 kDa; 488 amino acids), a typical value for SP enzymes(Goedl et al. 2010; Aerts et al. 2013). Activity measurementsshowed that the enzyme is indeed able to catalyse the revers-ible phosphorolysis of sucrose. Its optimal temperature foractivity was found to be 55 °C (Fig. S1), which is significantlyhigher than SP enzymes from L. mesenteroides and S. mutans(30–37 °C) (Goedl et al. 2010). It is, however, comparablewith the optimum of SP fromB. adolescentis (58 °C), which isremarkably stable for a mesophilic isolate (Cerdobbel et al.2010). The pH optimum was 6 and 6.5 for the synthetic andphosphorolytic reaction, respectively, and is in good agree-ment with that of other SP enzymes (Goedl et al. 2010)(Fig. S1). Furthermore, the apparent kinetic parameters havebeen determined for the donor as well as for the acceptorsubstrate in both directions of the reversible reaction (Table 2).The KM values for inorganic phosphate and α-D-glucose 1-phosphate are comparable with those of SP fromL. mesenteroides (Goedl et al. 2007; Schwarz and Nidetzky2006). For sucrose (77 mM) and D-fructose (42 mM) incontrast, KM values are significantly higher than for otherknown sucrose phosphorylases (1–15 and 8–22 mM, respec-tively; Aerts et al. 2011b). Hence, this low affinity mightimply that sucrose and fructose are not the natural substrates,and, accordingly, other substrates were evaluated (Table 3).Relative activities on monosaccharides like D-glucose, D-man-nose and D-galactose were, however, far below that of fructoseas acceptor. Maltose and maltodextrins in addition, which areacceptors for SP’s closest phylogenetic neighbouramylosucrase (Potocki de Montalk et al. 2000), also did notact as acceptors.

Alternatively, to guide the choice of substrate, the genelandscape around the sp gene from T. thermosaccharolyticumwas evaluated and compared with other organisms expressinga sucrose phosphorylase (Fig. 2). For the latter, sp genes oftenreside in an operon together with major facilitator superfamily(MFS) permeases (e.g. L. mesenteroides) or are accompanied

Fig. 1 Phylogenetic tree of all protein sequences classified in glycosidehydrolase family GH13_18 (known sucrose phosphorylases (emptycircle), thermophilic sources (filled circle))

Appl Microbiol Biotechnol

by multiple sugar metabolism (msm) transporter proteins andglycosidases (e.g. L. acidophilus and S. mutans). InT. thermosaccharolyticum, in contrast, a distinct gene land-scape was observed. The ttsp gene is located in an operonwitha putative phosphofructokinase, near genes that encode aphosphoenolpyruvate-dependent transport system (PTS). Un-like MFS permeases (passive transport) and msm transportsystems (active transport) that both bring in a sugar in itsnative form, the PTS system phosphorylates it during trans-port. Hence, the enzyme of T. thermosaccharolyticum mightbe active on phosphorylated sugars. Therefore, its ability touse sucrose 6′-phosphate (Suc6′P) and D-fructose 6-phosphate(Fru6P) as donor and acceptor substrate, respectively, wasinvestigated. It was found that the enzyme was indeed capableto perform these reactions, and kinetic characterisation re-vealed the preference for Fru6P over fructose and Suc6′P oversucrose (Table 2). KM values for Suc6′P and Fru6P were six

and three times lower, respectively, compared to their non-phosphorylated counterparts, and kcat values were slightlyhigher. The enzyme of T. thermosaccharolyticum can thus bedesignated as a sucrose 6′-phosphate phosphorylase (SPP;Fig. 3).

As mentioned above, the spp gene is present in anoperon together with an enzyme annotated as a putativephosphofructokinase. In order to investigate the truespecificity of this enzyme, it was also recombinantlyexpressed, and activity measurements disclosed the pref-erence of Fru6P over fructose as substrate (Fig. S2),proving this enzyme to be a phosphofructokinase(Fig. 3). Indeed, for the same concentration of Fru6Pand fructose (1 mM), the former quickly gave rise tothe formation of fructose-1,6-bisphosphate, whereas thelatter did not result in product formation, even not uponprolonged incubation.

Table 2 Apparent kinetic parameters for sucrose 6′-phosphate phosphorylase from T. thermosaccharolyticum

Reaction Substrate KM (mM) kcat (s−1) kcat/KM (mM−1 s−1)

Phosphorolysis Sucrosea 76.5±10.3 66.2±4.5 0.9

Sucrose 6′-phosphatea 12.7±1.9 82.6±3.1 6.5

Phosphateb 6.9±0.6 59.3±3.1 8.6

Synthesis α-D-Glucose 1-phosphatec 15.6±1.8 14.4±0.7 0.9

D-Fructosed 41.6±5.3 16.9±1.3 0.4

D-Fructose 6-phosphated 15.1±2.3 24.2±1.8 1.6

a Initial rates were determined at 55 °C in 50 mM MES buffer pH 6.5 in the presence of 100 mM of inorganic phosphateb Initial rates were determined at 55 °C in 50 mM MES buffer pH 6.5 in the presence of 350 mM sucrose (phosphorolysis)c Initial rates were determined at 55 °C in 50 mM MES buffer pH 6.0 in the presence of 200 mM D-fructosed Initial rates were determined at 55 °C in 50 mM MES buffer pH 6.0 in the presence of 100 mM α-D-glucose 1-phosphate (synthesis)

Table 1 Thermophilic sources of GH13_18 sequences

Organism UniProt ID Growthrangea (°C)

Reference

Caldilinea aerophila (strain DSM 14535/JCM 11387/NBRC 104270/STL-6-O1) I0I189 50–60 (55) Sekiguchi et al. (2003)

Deferribacter desulfuricans (strain DSM 14783/JCM 11476/NBRC 101012/SSM1) D3PAG7 40–70 (60–65) Takai et al. (2003)

Geobacillus thermodenitrificans (strain NG80-2) A4ITA6 45–73 (65) Feng et al. (2007)

Marinithermus hydrothermalis (strain DSM 14884/JCM 11576/T1) F2NL32 50–72.5 (67.5) Sako et al. (2003)

Marinitoga piezophila (strain DSM 14283/JCM 11233/KA3) H2J6Y9 45–70 (65) Alain et al. (2002)

Meiothermus ruber (strain ATCC 35948/DSM 1279/VKM B-1258/21) (Thermus ruber) D3PQE0 35–70 (60) Tindall et al. (2010)

Meiothermus silvanus (strain ATCC 700542/DSM 9946/VI-R2) (Thermus silvanus) D7BAR0 40–65 (55) Sikorski et al. (2010)

Oceanithermus profundus (strain DSM 14977/NBRC 100410/VKM B-2274/506) E4U9R0 40–68 (60) Miroshnichenko et al. (2003)

Spirochaeta thermophila (strain ATCC 49972/DSM 6192/RI 19.B1) E0RTJ0 40–73 (66) Aksenova et al. (1992)

Spirochaeta thermophila (strain ATCC 700085/DSM 6578/Z-1203) G0GBS4 40–73 (66) Aksenova et al. (1992)

Thermoanaerobacterium thermosaccharolyticum (strain ATCC 7956/DSM 571/NCIB9385/NCA 3814) (Clostridium thermosaccharolyticum)

D9TT09 45–70 (60) Lee et al.(1993)

Thermoanaerobacterium thermosaccharolyticumM0795 L0IL15 45–70 (60) Lee et al. (1993)

Thermoanaerobacterium xylanolyticum (strain ATCC 49914/DSM 7097/LX-11) F6BJS0 45–70 (60) Lee et al. (1993)

a Optimal growth temperature between brackets

Appl Microbiol Biotechnol

Thermostability

To evaluate the thermostability of TtSPP, it was comparedwith SP from B. adolescentis (BaSP), the most stable SPrepresentative known to date (Cerdobbel et al. 2010). DSFwas used to determine the Tm, i.e. a parameter representing aprotein’s tendency to unfold as a function of temperature

(Polizzi et al. 2007). Melting temperatures were found to benearly identical for both enzymes, with a value of 78 °C(Fig. S3). Kinetic stability, reflecting the time a protein re-mains active before undergoing irreversible denaturation, wasexamined as well (Polizzi et al. 2007). To that end, the half-life(t50) was determined at the industrially relevant temperature of60 °C (Fig. S3). For TtSPP, it was found to be almost threetimes higher than for BaSP (60 h compared to 21 h, respec-tively). The enzyme from T. thermosaccharolyticum thus hascomparable melting characteristics as its most stable SP coun-terpart, but it is remarkably more stable over time.

This characteristic is particularly interesting for industrialprocesses, such as the production of α-D-glucose 1-phosphateor α-glucosyl glycerol. For that purpose, the enzyme’s loweraffinity for sucrose as glycosyl donor is not a problem, sincehigh substrate concentrations are typically applied. For exam-ple, one molar of sucrose and inorganic phosphate could bereadily converted to 0.7 M of α-D-glucose 1-phosphate(Fig. S4), which is the theoretical maximal yield (De Winteret al. 2011). Furthermore, a fivefold excess of sucrose ispreferentially used for convenient downstream processing, inwhich case the sucrose concentration remains saturatingthroughout the reaction (Fig. S4). In turn, 0.6 M of α-glucosyl glycerol could be produced from 0.8 M of sucroseand 2 M of glycerol, which are the optimal concentrationsaccording to Goedl et al. (2008; Fig. S5). At a temperature of60 °C, a glycosyl transfer yield of 75 % was achieved, com-parable to the 85 % reported for SP at 30 °C.

Structure–function relationships

In the phylogenetic tree, the enzymeofT. thermosaccharolyticumcan be found in between SP enzymes from lactic acid bacteria(LAB) and Bifidobacteria (Fig. 1), but nevertheless its

Table 3 Substrate specificity for sucrose 6′-phosphate phosphorylasefrom T. thermosaccharolyticum

Reaction Substrate Relative activity (%)

Phosphorolysisa Sucrose 6′-phosphate 100

Sucrose 51

Trehalose 6-phosphate 0

Synthesisb D-Fructose 6-phosphate 100

D-Fructose 62

D-Psicose n.s.

D-Tagatose 8

L-Sorbose 31

α-D-Glucose 1-phosphate n.s.

D-Glucose 6-phosphate n.s.

D-Glucose n.s.

D-Mannose n.s.

D-Galactose n.s.

D-Maltose n.s.

Maltodextrin DE-15 n.s.

n.s. not significantly higher than the hydrolytic side activity (4 % com-pared to wild-type activity)a All reactions were performed at 55 °C in 50mMMES buffer pH 6.5 and100 mM donor and 100 mM acceptor (inorganic phosphate)b All reactions were performed at 55 °C in 50mMMES buffer pH 6.5 and100 mM donor (α-D-glucose 1-phosphate) and 200 mM acceptor

Fig. 2 Gene landscape ofdifferent organisms expressing asucrose phosphorylase (MFSmajor facilitator superfamilypermease, msm multiple sugarmetabolism, PTSphosphoenolpyruvate-dependentsugar phosphotransferase system,pfkB phosphofructokinase, SPsucrose phosphorylase; genenames and identifiers wereretrieved from the NCBI genomedatabase)

Appl Microbiol Biotechnol

specificity differs. To gain more insight in which residuesaccount for this difference, a multiple sequence alignment(MSA) was performed of several SP representatives togetherwith TtSPP (Fig. 4a and Fig. S6). The acceptor site of theformer mainly constitutes of loops A and B, and a number ofresidues are highly conserved and involved in specificity(Mueller and Nidetzky 2007; Verhaeghe et al. 2013). TheMSA shows that in TtSPP, a His is present at position 344(loop A) whereas all SP enzymes have a Tyr, a residueinvolved in phosphate binding. Furthermore, the NL motif atthe beginning of the loop is replaced by GF. In loop B, twoconserved motifs can be found in SP enzymes, i.e. YRPRP

(Bifidobacteria) and YKRKD (lactic acid bacteria). In TtSPP,however, this loop is one residue shorter and it has a distinctmotif, i.e. FLRR.

Docking of sucrose 6′-phosphate in the homologymodel ofTtSPP suggested that Arg134 is important for binding of thephosphate group (Fig. 4b). As Lys301 and Arg135 are also inclose proximity and therefore could fulfil this function as well,mutagenesis was performed to indentify the role of theseresidues (Fig. 4b). Alanine mutants were created and thespecific activities on fructose and fructose 6-phosphate werecompared (Table 4). Additionally, His344 was mutated to Tyr,the residue present in SP enzymes.

R135A and K301A displayed similar activities as the wild-type enzyme. R134A and H344Y, on the other hand, showed adecreased ratio of activity on phosphorylated fructose overfructose, proving their role in binding of the phosphate group.Their lower activity on fructose might indicate that they arealso involved in binding of the donor substrate (Glc1P) or thatthey are simply necessary for the overall activity.

Discussion

Specificities in GH13 subfamily 18

The gene from T. thermosaccharolyticum was predicted toencode a sucrose phosphorylase, and the enzyme was indeedactive on sucrose. Kinetic analysis, however, revealed its

Fig. 3 Reactions catalysed bysucrose 6′-phosphatephosphorylase (SPP) andphosphofructokinase (pfk) fromT. thermosaccharolyticum

Fig. 4 Multiple sequence alignment (MSA) of the acceptor site loops ofseveral representative sucrose phosphorylases (SP) and sucrose 6′-phos-phate phosphorylase (SPP) (loop A and B, starting at position 340 and132, respectively (TtSPP numbering)) and docking of sucrose 6′-phos-phate into the homology model of Thermoanaerobacteriumthermosaccharolyticum SPP (BaSP Bifidobacterium adolescentis SP;BlSP Bifidobacterium longum SP; LmSP Leuconostoc mesenteroidesSP; SmSP Streptococcus mutans SP; LaSP Lactobacillus acidophilusSP; TtSPP T. thermosaccharolyticum SPP; residue with grey backgroundconserved in SP enzymes; boxed residue conserved within subgroup ofSP enzymes (lactic acid bacteria or Bifidobacteria); bold residues in-volved in specificity, confirmed by mutagenesis)

Table 4 Specific activity of wild-type sucrose 6′-phosphate phosphory-lase from T. thermosaccharolyticum and alanine mutants (55 °C, 50 mMMES pH 6.5, 100 mM donor (α-D-glucose 1-phosphate) and 50 mMacceptor (D-fructose or D-fructose 6-phosphate))

Enzyme ActFru6P (U/mg) ActFru (U/mg) Ratio ActFru6P/ActFru

WT 20.7 10.1 2.0

R134A 0.5 0.7 0.7

R135A 15.5 7.8 2.0

K301A 21.2 8.6 2.5

H344Y 4.4 4.3 1.0

Appl Microbiol Biotechnol

preference for phosphorylated sucrose and more specificallysucrose 6′-phosphate (β-D-Fru6P-(2→1)-α-D-Glc). The en-zyme thus represents a novel specificity, i.e. sucrose 6′-phos-phate phosphorylase, for which no EC number is yet avail-able. Furthermore, it was shown that TtSPP is able to accom-modate the extra phosphate group due to some differences inthe acceptor site loops compared to SP enzymes. Severalresidues in these loops are highly conserved in SP enzymesand their importance in substrate recognition has already beendemonstrated (Mueller and Nidetzky 2007; Verhaeghe et al.2013). Hence, the presence of different residues might beindicative for diverging specificities. Multiple sequence align-ment of all GH13_18 sequences disclosed some distinct mo-tifs for different branches of the phylogenetic tree (Fig. 5).Sequences from Thermoanaerobacterium xylanolyticum and

Clostridium saccharoperbutylacetonicum (SPP branch) sharethe same motifs as TtSPP and are most likely also sucrose 6′-phosphate phosphorylases. The closely related sequencesfrom Marinobacter adhaerens and Thalassolituus oleivorans(unknown branch, n=3) on the other hand differ from both SPand SPP enzymes, and particularly the aspartate in loop Awhich is replaced by an alanine. In SP enzymes, this residue isvery important for binding of the fructose moiety and substi-tution to alanine led to a tremendous loss of activity (Muellerand Nidetzky 2007; Verhaeghe et al. 2013). Therefore, it islikely that these enzymes are not sucrose phosphorylases. Forthe branch that does not contain any characterised representa-tives (unknown, n=222), major differences are observed com-pared to the remainder of the tree. In SP enzymes, the histidinenear the general acid/base catalyst is crucial for sucrose

Fig. 5 Sequence logos of theacceptor site representing thefrequency of each residue in amultiple sequence alignment ofall sequences classified in familyGH13_18, for different branchesof the phylogenetic tree (LABlactic acid bacteria; A/B loopcontaining the general acid/basecatalyst; asterisks residuesinvolved in specificity, confirmedby mutagenesis)

Fig. 6 Different pathways foruptake and catabolism of sucrosevia a sucrose 6′-phosphatephosphorylase, b sucrosephosphorylase and c sucrose 6-phosphate hydrolase (PTSphosphoenolpyruvate-dependentsugar phosphotransferase system;SPH sucrose 6-phosphatehydrolase, SP sucrosephosphorylase, SPP sucrose 6′-phosphate phosphorylase, fkfructokinase, pfkphosphofructokinase, Suc6Psucrose 6-phosphate (β-d-Fru-(2→1)-α-d-Glc 6-P), Suc6′Psucrose 6′-phosphate (β-d-Fru 6-P-(2→1)-α-d-Glc))

Appl Microbiol Biotechnol

phosphorylase activity (Verhaeghe et al. 2013), but here anasparagine is present. In addition, loop A is two residuesshorter and only shares its tyrosine with sucrose phosphory-lases. From all these findings, it can be concluded that theGH13 subfamily 18 might be more diverse than expected.Although it cannot be excluded that the different acceptor sitearchitectures yield the same specificity, it is very likely thatnext to sucrose phosphorylases and sucrose 6′-phosphatephosphorylases still other specificities are present.

Physiological role of sucrose 6′-phosphorylase

Sucrose 6′-phosphate phosphorylase not only represents anew functionality, it also implies a new pathway for thedegradation of sucrose (Fig. 6a). The higher in vivo concen-trations of the substrates compared to the products, especiallyinorganic phosphate (Goedl et al. 2010), indeed favour phos-phorolysis (breakdown) over synthesis. Note that, analoguesto sucrose phosphorylase (Fig. 6b), this is an energy-savingstep as no ATP is required (Reid and Abratt 2005). Thecatabolic role is further supported by the presence of a phos-phofructokinase gene in the same operon. The pfk enzyme cannamely directly process the Fru6P produced by SPP. Accord-ingly, a complete pathway is established in which sucrose 6′-phosphate is first phosphorolytically split by SPP into Glc1Pand Fru6P, where after the latter is phosphorylated by pfkyielding Fru1,6PP (Fig. 6a). The Glc1P on the other handcan be readily converted to Glc6P by the action of phospho-glucomutase present in bacteria. Finally, both reaction prod-ucts end up as intermediates in the glycolytic pathway.

Although the pathway itself is straightforward, it remainsunclear where the sucrose 6′-phosphate originates from. Cur-rently, the only knownway in bacteria to obtain Suc6′P is by theaction of sucrose 6′-phosphate synthase, which belongs to gly-cosyltransferase family 4 (GT4) and uses UDP-Glc and Fru6Pas substrates (Lutfiyya et al. 2007). T. thermosaccharolyticumhas a few gene products classified in this family (Table S4) andsince these enzymes are not characterised yet, they mightentail sucrose 6′-phosphate synthase activity. However, asT. thermosaccharolyticum is known for the degradation ratherthan synthesis of sugars (O-Thong et al. 2008), another possi-bility would be that Suc6′P is generated during uptake ofsucrose. Phosphorylation of sucrose during translocation isnormally performed by a sucrose-specific PTS system, whichattaches a phosphate to the six positions of the glucose moiety(Reid and Abratt 2005). Thereafter, it is further degraded viasucrose 6-phosphate hydrolase (Fig. 6c). Here, in contrast,sucrose is phosphorylated on the fructose moiety. Among theT. thermosaccharolyticum genes annotated as PTS compo-nents, none of them seem to be sucrose specific (Table S4).In contrary, complete PTS systems are predicted from themannose/fructose/sorbose family (Transporter ClassificationDatabase; www.tcdb.org). This family is known for its wide

range of specificities, and its members often accept a widerange of sugars (Barabote and Saier 2005). It is thereforepossible that sucrose enters via one of these PTS systems andbecomes phosphorylated on its fructose moiety. Future re-search, however, should provide further prove for this newroute for breakdown of sucrose.

Potential applications

Next to its interesting in vivo role, TtSPP also has somepotential advantages in terms of practical applications. Al-though sucrose 6′-phosphate is its natural substrate, sucrosecan be readily used as alternative glycosyl donor. This allowsα-D-glucose 1-phosphate to be produced in an identical way asreported for SP, albeit with a biocatalyst that is more stable.Additionally, the production of α-glucosyl glycerol can nowbe performed at 60 °C, whereas 30 °C was previously usedwith SP. Of course, TtSPP also offers the possibility to convertsucrose 6′-phosphate to D-fructose 6-phosphate and vice versa,although the market of these compounds probably is rathersmall. Nevertheless, none of these applications require proteinimmobilisation, thanks to the enzyme’s excellent stability.

Acknowledgments The authors wish to thank the Special ResearchFund (BOF) of Ghent University (MRP “Ghent Bio-Economy”), as wellas the EC (FP7-project “Novosides”, grant agreement nr. KBBE-4-265854) for financial support.

References

Aerts D, Verhaeghe T, De Mey M, Desmet T, Soetaert W (2011a) Aconstitutive expression system for high-throughput screening. EngLife Sci 11(1):10–19

Aerts D, Verhaeghe T, Joosten HJ, VriendG, SoetaertW, Desmet T (2013)Consensus engineering of sucrose phosphorylase: the outcome re-flects the sequence input. Biotechnol Bioeng 110(10):2563–2572

Aerts D, Verhaeghe TF, Roman BI, Stevens CV, Desmet T, Soetaert W(2011b) Transglucosylation potential of six sucrose phosphorylasestoward different classes of acceptors. Carbohydr Res 346(13):1860–1867

Aksenova HY, Rainey FA, Janssen PH, Zavarzin GA, Morgan HW(1992) Spirochaeta thermophila sp. nov, an obligately anaerobic,polysaccharolytic, extremely thermophilic bacterium. Int J SystBacteriol 42(1):175–177

Alain K, Marteinsson VT, Miroshnichenko ML, Bonch-OsmolovskayaEA, Prieur D, Birrien JL (2002) Marinitoga piezophila sp nov., arod-shaped, thermo-piezophilic bacterium isolated under high hy-drostatic pressure from a deep-sea hydrothermal vent. Int J Syst EvolMicrobiol 52:1331–1339

Barabote RD, Saier MH (2005) Comparative genomic analyses of thebacterial phosphotransferase system.MicrobiolMol Biol Rev 69(4):608

Blecher M, Glassman AB (1962) Determination of glucose in the pres-ence of sucrose using glucose oxidase; effect of pH on absorptionspectrum of oxidized o-dianisidine. Anal Biochem 3(4):343–352

Bruins ME, Janssen AEM, Boom RM (2001) Thermozymes and theirapplications—a review of recent literature and patents. ApplBiochem Biotechnol 90(2):155–186

Appl Microbiol Biotechnol

Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V,Henrissat B (2009) The Carbohydrate-Active EnZymes database(CAZy): an expert resource for glycogenomics. Nucleic Acids Res37:D233–D238

Cerdobbel A, De Winter K, Aerts D, Kuipers R, Joosten HJ, Soetaert W,Desmet T (2011a) Increasing the thermostability of sucrose phos-phorylase by a combination of sequence- and structure-based muta-genesis. Protein Eng Des Sel 24(11):829–834

Cerdobbel A, De Winter K, Desmet T, Soetaert W (2011b) Sucrosephosphorylase as cross-linked enzyme aggregate: improved thermalstability for industrial applications. Biotechnol J 5(11):1192–1197

Cerdobbel A, Desmet T, De Winter K, Maertens J, Soetaert W (2010)Increasing the thermostability of sucrose phosphorylase bymultipoint covalent immobilization. J Biotechnol 150(1):125–130

Chen S, Liu J, Pei H, Li J, Zhou J, Xiang H (2007) Molecular investiga-tion of a novel thermostable glucan phosphorylase fromThermoanaerobacter tengcongensis. Enzyme Microb Technol 41(3):390–396

De Winter K, Cerdobbel A, Soetaert W, Desmet T (2011) Operationalstability of immobilized sucrose phosphorylase: continuous produc-tion of alpha-glucose-1-phosphate at elevated temperatures. ProcessBiochem 46(10):2074–2078

De Winter K, Soetaert W, Desmet T (2012) An Imprinted Cross-LinkedEnzyme Aggregate (iCLEA) of sucrose phosphorylase: combiningimproved stability with altered specificity. Int J Mol Sci 13(9):11333–11342

DeWinter K, Verlinden K, Kren V, Weignerova L, Soetaert W, Desmet T(2013) Ionic liquids as cosolvents for glycosylation by sucrosephosphorylase: balancing acceptor solubility and enzyme stability.Green Chem 15(7):1949–1955

Desmet T, Soetaert W (2011) Enzymatic glycosyl transfer: mechanismsand applications. Biocatal Biotransform 29(1):1–18

Desmet T, Soetaert W, Bojarova P, Kren V, Dijkhuizen L, Eastwick-FieldV, Schiller A (2012) Enzymatic glycosylation of small molecules:challenging substrates require tailored catalysts. Chem-a Europ J18(35):10786–10801

Doudoroff M (1943) Studies on the phosphorolysis of sucrose. J BiolChem 151(2):351–361

Eijsink VGH, Gaseidnes S, Borchert TV, van den Burg B (2005) Directedevolution of enzyme stability. Biomol Eng 22(1–3):21–30

Feng L,WangW, Cheng J, Ren Y, Zhao G, Gao C, TangY, LiuX,HanW,Peng X, Liu R,Wang L (2007) Genome and proteome of long-chainalkane degrading Geobacillus thermodenitrificans NG80-2 isolatedfrom a deep-subsurface oil reservoir. Proc Natl Acad Sci U S A 104(13):5602–5607

Fujii K, Iiboshi M, Yanase M, Takaha T, Kuriki T (2006) Enhancing thethermal stability of sucrose phoshphorylase from Streptococcusmutans by random mutagenesis. J appl glycosci 53:91–97

Gawronski JD, Benson DR (2004) Microtiter assay for glutamine syn-thetase biosynthetic activity using inorganic phosphate detection.Anal Biochem 327(1):114–118

Gibson DG, Benders GA, Axelrod KC, Zaveri J, Algire MA, Moodie M,Montague MG, Venter JC, Smith HO, Hutchison CA (2008) One-step assembly in yeast of 25 overlapping DNA fragments to form acomplete synthetic Mycoplasma genitalium genome. Proc NatlAcad Sci U S A 105(51):20404–20409

Goedl C, Sawangwan T, Mueller M, Schwarz A, Nidetzky B (2008) Ahigh-yielding biocatalytic process for the production of 2-O-(alpha-D-glucopyranosyl)-sn-glycerol, a natural osmolyte and useful mois-turizing ingredient. Angew Chem Int Ed 47(52):10086–10089

Goedl C, Sawangwan T, Wildberger P, Nidetzky B (2010) Sucrosephosphorylase: a powerful transglucosylation catalyst for synthesisof alpha-D-glucosides as industrial fine chemicals. BiocatalBiotransform 28(1):10–21

Goedl C, Schwarz A,Minani A, Nidetzky B (2007) Recombinant sucrosephosphorylase from Leuconostoc mesenteroides: characterization,

kinetic studies of transglucosylation, and application of immobilisedenzyme for production of alpha-D-glucose 1-phosphate. JBiotechnol 129(1):77–86

Koga T, Nakamura K, Shirokane Y, Mizusawa K, Kitao S, Kikuchi M(1991) Purification and some properties of sucrose phosphorylasefrom Leuconostoc mesenteroides. Agric Biol Chem 55(7):1805–1810

Krieger E, Joo K, Lee J, Lee J, Raman S, Thompson J, TykaM, Baker D,Karplus K (2009) Improving physical realism, stereochemistry, andside-chain accuracy in homology modeling: four approaches that per-formed well in CASP8. Proteins-Struct Funct Bioinforma 77:114–122

Krieger E, Koraimann G, Vriend G (2002) Increasing the precision ofcomparative models with YASARA NOVA—a self-parameterizingforce field. Proteins Struct Funct Genet 47(3):393–402

Lee YE, Jain MK, Lee CY, Lowe SE, Zeikus JG (1993) Taxonomicdistinction of saccharolytic thermophilic anaerobes: description ofThermoanaerobacterium xylanolyticum gen. Nov., sp. Nov., andThermoanaerobacterium saccharolyticum gen. Nov., sp. Nov.; re-classification of Thermoanaerobium brockii, Clostridiumthermosulfurogenes, and Clostridium thermohydrosulfuricume100-69 as Thermoanaerobacter brockii comb. Nov.,Thermoanaerobacterium thermosulfurigenes comb. nov., andThermoanaerobacter thermohydrosulfuricus comb. nov., respec-tively; and transfer of Clostridium thermohydrosulfuricum 39E toThermoanaerobacter ethanolicus. Int J Syst Bacteriol 43(1):41–51

Lutfiyya LL, Xu N, D’Ordine RL, Morrell JA, Miller PW, Duff SM(2007) Phylogenetic and expression analysis of sucrose phosphatesynthase isozymes in plants. J Plant Physiol 164(7):923–933

Maruta K, Mukai K, Yamashita H, Kubota M, Chaen H, Fukuda S,Kurimoto M (2002) Gene encoding a trehalose phosphorylase fromThermoanaerobacter brockii ATCC 35047. Biosci BiotechnolBiochem 66(9):1976–1980

Miroshnichenko ML, L’Haridon S, Jeanthon C, Antipov AN, KostrikinaNA, Tindall BJ, Schumann P, Spring S, Stackebrandt E, Bonch-Osmolovskaya EA (2003) Oceanithermus profundus gen.nov., sp nov., a thermophilic, microaerophilic, facultativelychemolithoheterotrophic bacterium from a deep-sea hydrothermalvent. Int J Syst Evol Microbiol 53:747–752

Mueller M, Nidetzky B (2007) Dissecting differential binding of fructoseand phosphate as leaving group/nucleophile of glucosyl transfercatalyzed by sucrose phosphorylase. FEBS Lett 581(20):3814–3818

Niesen FH, Berglund H, Vedadi M (2007) The use of differential scan-ning fluorimetry to detect ligand interactions that promote proteinstability. Nat Protoc 2(9):2212–2221

O-Thong S, Prasertsan P, Karakashev D, Angelidaki I (2008)Thermophilic fermentative hydrogen production by the newly iso-lated Thermoanaerobacterium thermosaccharolyticum PSU-2. Int JHydrogen Energy 33(4):1204–1214

Polizzi KM, Bommarius AS, Broering JM, Chaparro-Riggers JF (2007)Stability of biocatalysts. Curr Opin Chem Biol 11(2):220–225

Potocki de Montalk G, Remaud-Simeon M, Willemot R-M, Sarçabal P,Planchot V, Monsan P (2000) Amylosucrase from Neisseriapolysaccharea: novel catalytic properties. FEBS Lett 471(2–3):219–223

Reid SJ, Abratt VR (2005) Sucrose utilisation in bacteria: genetic orga-nisation and regulation. Appl Microbiol Biotechnol 67(3):312–321

SaitouN,NeiM (1987) The neighbor-joiningmethod—a newmethod forreconstructing phylogenetic trees. Mol Biol Evol 4(4):406–425

Sako Y, Nakagawa S, Takai K, Horikoshi K (2003) Marinithermushydrothermalis gen. nov., sp nov., a strictly aerobic, thermophilicbacterium from a deep-sea hydrothermal vent chimney. Int J SystEvol Microbiol 53:59–65

Sanchis J, Fernandez L, Carballeira J, Drone J, Gumulya Y, HobenreichH, Kahakeaw D, Kille S, Lohmer R, Peyralans J, Podtetenieff J,Prasad S, Soni P, Taglieber A, Wu S, Zilly F, Reetz M (2008)Improved PCR method for the creation of saturation mutagenesis

Appl Microbiol Biotechnol

libraries in directed evolution: application to difficult-to-amplifytemplates. Appl Microbiol Biotechnol 81(2):387–397

Schrodinger L (2010) The PyMOL Molecular Graphics System, version1.3r1. 1.3r1 edn. Schrodinger, New York, USA

Schwarz A, Nidetzky B (2006) Asp-196 →Ala mutant of Leuconostocmesenteroides sucrose phosphorylase exhibits altered stereochemi-cal course and kinetic mechanism of glucosyl transfer to and fromphosphate. FEBS Lett 580(16):3905–3910

Sekiguchi Y, Yamada T, Hanada S, Ohashi A, Harada H, Kamagata Y(2003) Anaerolinea thermophila gen. nov., sp nov and Caldilineaaerophila gen. nov., sp nov., novel filamentous thermophiles thatrepresent a previously uncultured lineage of the domain bacteria atthe subphylum level. Int J Syst Evol Microbiol 53:1843–1851

Sievers F, Wilm A, Dineen D, Gibson TJ, Karplus K, Li W, Lopez R,McWilliam H, Remmert M, Soeding J, Thompson JD, Higgins DG(2011) Fast, scalable generation of high-quality protein multiplesequence alignments using Clustal Omega. Mol Syst Biol 7

Sikorski J, Tindall BJ, Lowry S, Lucas S, Nolan M, Copeland A, Del RioTG, Tice H, Cheng J-F, Han C, Pitluck S, Liolios K, Ivanova N,Mavromatis K, Mikhailova N, Pati A, Goodwin L, Chen A,Palaniappan K, Land M, Hauser L, Chang Y-J, Jeffries CD, RohdeM, Goeker M, Woyke T, Bristow J, Eisen JA, Markowitz V,Hugenholtz P, Kyrpides NC, Klenk H-P, Lapidus A (2010)Complete genome sequence of Meiothermus silvanus type strain(VI-R2(T)). Stand Genomic Sci 3(1):37–46

Silverstein R, Voet J, Reed D, Abeles RH (1967) Purification and mech-anism of action of sucrose phosphorylase. J Biol Chem 242(6):1338–1346

Skov LK, Mirza O, Sprogoe D, Van der Veen BA, Remaud-Simeon M,Albenne C, Monsan P, Gajhede M (2006) Crystal structure of theGlu328Gln mutant of Neisseria polysaccharea amylosucrase incomplex with sucrose and maltoheptaose. Biocatal Biotransform24(1–2):99–105

Stam MR, Danchin EGJ, Rancurel C, Coutinho PM, Henrissat B (2006)Dividing the large glycoside hydrolase family 13 into subfamilies:

towards improved functional annotations of alpha-amylase-relatedproteins. Protein Eng Des Sel 19(12):555–562

Taboada B, Ciria R, Martinez-Guerrero CE, Merino E (2012) ProOpDB:Prokaryotic Operon DataBase. Nucleic Acids Res 40(D1):D627–D631

Takai K, Kobayashi H, Nealson KH, Horikoshi K (2003) Deferribacterdesulfuricans sp nov., a novel sulfur-, nitrate- and arsenate-reducingthermophile isolated from a deep-sea hydrothermal vent. Int J SystEvol Microbiol 53:839–846

Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011)MEGA5: Molecular Evolutionary Genetics Analysis using maxi-mum likelihood, evolutionary distance, and maximum parsimonymethods. Mol Biol Evol 28(10):2731–2739

Tindall BJ, Sikorski J, Lucas S, Goltsman E, Copeland A, Del Rio TG,Nolan M, Tice H, Cheng J-F, Han C, Pitluck S, Liolios K, IvanovaN, Mavromatis K, Ovchinnikova G, Pati A, Faehnrich R, GoodwinL, Chen A, Palaniappan K, Land M, Hauser L, Chang Y-J, JeffriesCD, Rohde M, Goeker M, Woyke T, Bristow J, Eisen JA,Markowitz V, Hugenholtz P, Kyrpides NC, Klenk H-P, Lapidus A(2010) Complete genome sequence of Meiothermus ruber typestrain (21(T)). Stand Genomic Sci 3(1):26–36

Trott O, Olson AJ (2010) Software news and update AutoDock Vina:improving the speed and accuracy of docking with a new scoringfunction, efficient optimization, and multithreading. J ComputChem 31(2):455–461

Verhaeghe T, Diricks M, Aerts D, Soetaert W, Desmet T (2013) Mappingthe acceptor site of sucrose phosphorylase from Bifidobacteriumadolescentis by alanine scanning. J Mol Catal B Enzym 96:81–88

Vieille C, Zeikus JG (1996) Thermozymes: identifying molecular deter-minants of protein structural and functional stability. TrendsBiotechnol 14(6):183–190

Yamamoto T,Maruta K,Mukai K, Yamashita H, Nishimoto T, KubotaM,Fukuda S, KurimotoM, TsujisakaY (2004) Cloning and sequencingof kojibiose phosphorylase gene from Thermoanaerobacter brockiiATCC35047. J Biosci Bioeng 98(2):99–106

Appl Microbiol Biotechnol