Characterization of a novel low-temperature-active, alkaline and sucrose-tolerant invertase

A glycoside hydrolase family 32 invertase from Bacillus sp. HJ14 was expressed in Escherichia coli. The purified recombinant enzyme (rInvHJ14) showed typical biochemical properties of low-temperature-active and alkaline enzymes: (i) rInvHJ14 was active and stable in the range of pH 7.0–9.5 with an apparent pH optimum of 8.0; (ii) rInvHJ14 was most active but not stable at 30–32.5 °C, with 19.7, 48.2 and 82.1% of its maximum activity when assayed at 0, 10 and 20 °C, respectively, and the Ea, ΔG* (30 °C), Km (30 °C) and kcat (30 °C) values for hydrolysis of sucrose by rInvHJ14 was 47.6 kJ mol−1, 57.6 kJ mol−1, 62.9 mM and 746.2 s−1, respectively. The enzyme also showed strong sucrose tolerance. rInvHJ14 preserved approximately 50% of its highest activity in the presence of 2045.0 mM sucrose. Furthermore, potential factors for low-temperature-active and alkaline adaptations of rInvHJ14 were presumed. Compared with more thermostable homologs, rInvHJ14 has a higher frequency of glycine residues and a longer loop but a lower frequency of proline residues (especially in a loop) in the catalytic domain. The catalytic pockets of acid invertases were almost negatively charged while that of alkaline rInvHJ14 was mostly positively charged.

Invertases have been isolated from a variety of sources, including bacteria 17 , yeast 12 , fungi 10 , higher plants 18 and animals 19 . Acid invertases have been the most extensively studied and isolated from various sources while much less is known about alkaline invertases which have been reported only from plants and cyanobacteria 2,4,5,20-24 . In this study, a new GH 32 invertase, designated InvHJ14, was discovered from Bacillus sp. HJ14. The enzyme was expressed in Escherichia coli and the recombinant enzyme was purified. The recombinant InvHJ14 (rInvHJ14) was tolerant to sucrose and showed typical properties of low-temperature-active and alkaline enzymes. The potential molecular adaptations to low-temperature and alkaline environments of the enzyme were presumed.

Results
Sequence analysis. A gene of 1,458 bp encoding putative invertase of 55.9 kDa was predicted from the draft genome of HJ14. In silico analysis indicated that the invertase lacked recognizable signal peptide sequence but was comprised of a catalytic domain of GH 32 from H33 to P339 and a C-terminal domain of GH 32 from K338 to Q480. The consensus pattern of GH 32, H-x(2)-[PV]-x(4)-[LIVMA]-N-D-P-N-[GA] (http://prosite.expasy. org/PS00609), was also detected in InvHJ14 (HLMPPVGLLNDPNG in Fig. 1). The residues D43, D162 and E220 are the predicted nucleophile, transition-state stabilizer and general acid/base catalyst, respectively ( Fig. 1) 11,25 .
A BLASTP search against NCBI protein database showed that InvHJ14 was most similar to a number of putative invertases translated from genome sequences. In the result, InvHJ14 was found to have the maximal sequence identity of 99.6% with the uncharacterized GH 32 invertase from Bacillus safensis JPL_MERTA8-2  13 , InvEH from E. haichowensis (AFV59227) 38 , InvSC from S. cerevisiae (4EQV) 27 , InvSL from S. lycopersicum (AAA34132) 39 , InvUB from an uncultured bacterium (ACN59531) 24 and InvIB from I. batatas (AAD01606) 28 . Asterisks show the putative active residues. The segment corresponding to V149-G155 of InvHJ14, which presents the difference of proline residues between InvHJ14 and more thermostable homologs, is framed. The amino acid residues N210 to F215 of InvHJ14, which build a longer loop in Fig. 2 Structure analysis. The acidic invertases InvSC from Saccharomyces cerevisiae (Accession No. or PDB ID: 4EQV), InvIB from Ipomoea batatas (AAD01606), InvTM from Thermotoga maritima (1UYP), InvBL from Bifidobacterium longum (3PIJ) and InvUB from an uncultured bacterium (ACN59531) exhibite optimal activities at pH 4.8, pH 5.0, pH 5.5, pH 6.2 and pH 6.5, respectively 13,24,[26][27][28] . The model of InvIB was built using the crystal structure of a GH 32 β -fructofuranosidase from Arabidopsis thaliana (PDB ID: 2QQV) as the template with a sequence similarity of 42% and a GMQE score of 0.63. Models of InvUB and InvHJ14 were built using the crystal structure of InvBL as the template (Fig. 2), with sequence similarities of both 36% and GMQE scores of 0.67 and 0.66, respectively. As shown in Fig. 3, catalytic pockets of these invertases presented different surface charges. The catalytic pockets of acidic InvSC, InvIB, InvTM, InvBL and InvUB were almost negatively charged, while the opposite was observed for InvHJ14.
Biochemical characterization. Purified rInvHJ14 had activity in the range of pH 7.0-9.5 with an apparent pH optimum of 8.0 (Fig. 4a). Over 85% of its initial activity was observed after incubating the enzyme for 1 h in buffers with pH 7.0-9.0 (Fig. 4b). The apparent optimal temperature for rInvHJ14 was in the range from 30 to 32.5 °C (Fig. 4c). The enzyme could retain 48.2 and 82.1% of its maximum activity when assayed at 10 and 20 °C, respectively ( Fig. 4c). At the temperature of 0 °C, invertase activity of 19.7% was even retained (Fig. 4c). Half-lives of the enzyme were approximately 30 and 10 min when assayed at 30 and 37 °C, respcetively (Fig. 4d).
Addition of 1.0 mM MnSO 4 and β -mercaptoethanol slightly enhanced the activity of rInvHJ14, but addition of CuSO 4 , HgCl 2 and SDS strongly or completely inhibited the enzyme activity. The presence of other metal ions and chemical reagents had little or no effect on the enzyme activity (Table S1).
The activity of rInvHJ14 increased with the increase in sucrose concentration from 14.6 mM to 876.4 mM (Fig. 5). Then, rInvHJ14 was inhibited by the substrate at concentration of higher than 876.4 mM (Fig. 5). rIn-vHJ14 preserved approximately 50% of its highest activity in the presence of 2045.0 mM sucrose (Fig. 5).
Kinetic and thermodynamic characterization. From the Arrhenius plots ( Fig. 6), the E a value for hydrolysis of sucrose by rInvHJ14 was 47.6 kJ mol −1 . The Q 10 value (20 °C and 30 °C) for the hydrolysis of sucrose by rInvHJ14 was 1.9. Other kinetic and thermodynamic values are summarized in Table 1.
Transglycosylation activity. Self-condensation or transglycosylation product was not observed for rIn-vHJ14 in this study (Fig. S3). Discussion rInvHJ14 shows typical properties of low-temperature-active enzymes as its optimal activity is 30-32.5 °C, it has high catalytic activity at 0 and 10 °C, and it is thermolabile at 30 °C or higher temperatures 15 . Compared with rInvHJ14, most invertases show properties of mesophilic or thermophilic enzymes as their optimal activities are at temperatures of equal to or higher than 45 °C, they are thermostable at 50 °C or higher than 50 °C, and they show no activity or less than 10% relative activity at 10 °C 2, 3,7,8,12,18,24,[28][29][30][31][32][33][34][35] . The kinetic and thermodynamic properties of rInvHJ14 further indicate that the invertase is a low-temperature-active enzyme. Many low-temperature-active enzymes have lower E a values than their more thermostable homologs because E a decreases with a decrease in Δ G * or a decrease in energy barriers which allow easy conformational changes of enzymes during catalysis at low temperatures 36,37 . The E a and Δ G * values for hydrolysis of sucrose by rInvHJ14 was 47.6 kJ mol −1 and 57.6 kJ mol −1 , respectively, which are lower than the values for some thermostable invertases, such as Saccharum officinarum L. invertase showing an optimal temperature of 55 °C with an E a value of 55.3 kJ mol −1 and a Δ G * value of 71.2 kJ mol −1 18 , Fusarium sp. invertase showing an optimal temperature of 50 °C  27 , InvIB from I. batatas (AAD01606) 28 , InvTM from T. maritima (1UYP) 13 , InvBL from B. longum (3PIJ) 26 and InvUB from an uncultured bacterium (ACN59531) 24 . Catalytic pockets are circled. Charge distribution on the surface was calculated at pH 7.0. Positive charges are depicted in blue and negative charges in red.
and an E a value of 72.4 kJ mol −1 35 , and Oryza sativa invertase showing an E a value of 100.8 kJ mol −1 9 . The majority of low-temperature-active enzymes have higher K m and k cat values than their thermostable counterparts in order to decrease the activation free-energy barrier 36 . rInvHJ14 has higher K m (62.9 mM) and k cat (746.2 s −1 ) values than some thermostable invertases, such as Elsholtzia haichowensis invertase showing an optimal temperature of 70 °C with a K m value of 2.68 mM and a k cat value of 469.17 s −1 38 , Kluyveromyces marxianus invertase showing an optimal temperature of 50-55 °C with a K m value of 2.5 mM and a k cat value of 301 s −1 12 , and many microbial invertases with K m values of lower than 60 mM as reviewed previously 2 .
Catalytic domains of many low temperature-active enzymes present more number of glycine residues but less number of proline residues especially in loops than their more thermostable homologs because glycine residues   can provide localized chain mobility but proline residues restrict backbone rotations 15,16 . That will generate enhanced flexible region surrounding the catalytic site and finally allow catalysis of enzymes at low temperatures 15,16 . As shown in Fig. 1, InvHJ14 has more number or higher frequency of glycine residues but less number or lower frequency of proline residues in the catalytic domain from conserved HXXXXXXXXNDPNG to ECXXX than some mesophilic and thermophilic invertases. The number or frequency of glycine residues in the catalytic domain of InvHJ14 is approximately 2-fold compared with that of the invertase InvSC. On the contrary, the numbers or frequencies of proline residues in the catalytic domains of the invertases InvIB, InvUB, InvSL from Solanum lycopersicum (AAA34132) 39 and InvEH from E. haichowensis (AFV59227) 38 are around 2-fold compared with that of InvHJ14. And the numbers or frequencies of proline residues in the catalytic domains of the invertases InvSC and InvTM are approximately 1.5-fold compared with that of InvHJ14. Notably, proline residues in most mesophilic and thermophilic invertases increase significantly in the segment corresponding to V149-G155 of InvHJ14 as shown in Fig. 1. These amino acid residues in the segment build a loop near the predicted transition-state stabilizer D162 (Fig. 2). Consequently, the increased number of glycine residues combined with the decreased number of proline residues allow easier local conformational changes of InvHJ14 and finally contribute to low-temperature adaptation of the enzyme.
The amplitude of the movement between secondary structures can be increased by longer loop then result in the higher structure flexibility 16 . InvHJ14 has an increased number of amino acid residues from N210 to F215 compared with some mesophilic and thermophilic invertases (Fig. 1). These amino acid residues build a longer loop which is close to the catalytic E220 (Fig. 2). As such, the proper plasticity close to the catalytic site enables rInvHJ14 maintaining high catalytic activity at low temperatures accompanied by thermolability at intermediate and high temperatures.
rInvHJ14 is an alkaline invertase as it shows activity and stability at alkaline conditions. To our knowledge, alkaline invertases have been reported only from plants and cyanobacteria 2,4,5,20-24 . Most invertases produced by microorganisms worked optimally in the range of pH 2.9-6.2 2 . As such, the study is the first to report an alkaline invertase isolated from eubacteria. Compared with their acidic counterparts, alkaline enzymes have a larger positively charged surface to keep them active and stable under alkaline condition, such as the alkaline β -mannanase from Humicola insolens Y1 40 and α -galactosidase from Streptomyces sp. HJG4 37 . In this study, structure analysis found that the catalytic pockets of acidic InvSC, InvIB, InvTM, InvBL and InvUB were almost negatively charged while that of alkaline InvHJ14 was mostly positively charged. Therefore, a positively charged catalytic pocket is presumed to be a factor for alkaline adaptation of the GH 32 invertase rInvHJ14.
Furthermore, enzymatic hydrolysis of sucrose is preferable in the production of high-quality inverted syrups compared with acidic hydrolysis 41 . A high sucrose content; even as high as 2 M, is used in the industrial application 41 . As such, the identification of an invertase that is active in the presence of 2 M sucrose is relevant. It has been reported that the commercial invertase from S. cerevisiae displayed approximately 30% of its highest activity at 2 M sucrose 42 . The invertases from a metagenomic library and Aspergillus niger show approximately 50% 24 and 30% 10 of their highest activities at 1 M sucrose, respectively. In the presence of 2 M sucrose, the yeast Candida guilliermondii MpIIIa invertases INV3a-N and INV3a-D present nearly 50% and 10% of their highest activities, respectively 7 . Similar to INV3a-N, rInvHJ14 remained approximately 50% of its highest activity in the presence of 2045.0 mM sucrose. The big differences between INV3a-N and rInvHJ14 are that INV3a-N is an acidic and thermophilic invertase 7 . Together, rInvHJ14 may be a potential candidate for enzymatic hydrolysis of sucrose in the production of high-quality inverted syrups.
In conclusion, a GH 32 invertase was isolated from an eubacteria-Bacillus sp. HJ14. Biochemical characterization revealed that the enzyme was a novel low-temperature-active, alkaline and sucrose-tolerant invertase. These characteristics make the invertase suitable for versatile applications, especially in the food industry. Moreover, molecular characterization of the enzyme contributes to a better understanding of the structure-function relationships of low-temperature-active and alkaline invertases.

Materials and Methods
Strains, vectors and reagents. The details of isolation, identification and preservation of Bacillus sp. HJ14 were described in our previous study 43  d-fructose and d-xylose were purchased from Sigma− Aldrich (St. Louis, MO, USA). Wheat flour arabinoxylan (P-WAXYM) was purchased from Megazyme (Wicklow, Ireland). Stachyose and sucralose were purchased from Tokyo Chemical Industry (Tokyo, Japan). Fructooligosaccharides set including kestose, nystose and fructofuranosylnystose was purchased from Wako Pure Chemical (Osaka, Japan). Levan from Zymomonas mobilis was purchased from Advanced Technology & Industrial (Hong Kong, China), Isopropyl-β -d-1-thiogalactopyranoside (IPTG) was purchased from Amresco (Solon, OH, USA). Silica gel G plates were purchased from Haiyang (Qingdao, China). All chemicals were of analytical grade.
Gene cloning and sequence analysis. The genome of HJ14 was sequenced on a Hiseq 2000 sequencer (Illumina) and genomic data was analyzed on a NF supercomputing server (Inspur, Shandong, China) in our lab. Other details are as described in our previous study 43 . Genes from the draft genome of HJ14 were predicted by the tools GeneMark.hmm 2.8 (http://exon.gatech. edu/GeneMark/gmhmm2_prok.cgi) combined with local BLAST 2.2.25 44 . The identity values of DNA and protein sequences were obtained from the BLASTN and BLASTP (http://www.ncbi.nlm.nih.gov/BLAST/) results, respectively. The signal peptide and catalytic domain were predicted by SignalP (http://www.cbs.dtu.dk/services/ SignalP/) and InterPro online tool (http://www.ebi.ac.uk/interpro/), respectively. Structure analysis. Tertiary structures of invertases were predicted by homology modeling using SwissModel (http://swissmodel.expasy.org/). Spatial distributions of electrostatic potential of tertiary structures were calculated using the Discovery Studio 2.5 software (Accelrys, San Diego, CA, USA).
Heterologous expression. The InvHJ14-encoding gene (invHJ14) was amplified by PCR using Pyrobest DNA polymerase and the primer set rinvHJ14EF (ATGACAACTACTGACGCAGC) and rinvHJ14ER (TTGATCTGCTTCTCTTTGTAAAT), and then the resulting PCR product was added base A at the 5′ terminal using rTaq DNA polymerase. The recombinant plasmid, designated pEASY-invHJ14, was constructed with invHJ14 and pEASY-E2 vector through T-A ligation and transformed into E. coli BL21 (DE3) competent cells. The positive transformant harboring pEASY-invHJ14 was identified by PCR analysis and confirmed by DNA sequencing. The overnight culture of positive transformant was inoculated (1:100 dilutions) into fresh Luria-Bertani medium containing 100 μ g mL −1 ampicillin, and incubated at 37 °C. Upon reaching an OD 600nm of 0.7, 0.7 mM IPTG was added to the broth to induce enzyme expression at 20 °C for further 20 h.
Purification and identification of recombinant enzyme. E. coli cells were collected by centrifugation at 10,000 × g at room temperature for 5 min. The resulting cell pellet was resuspended in sterilized, ice-cold buffer A (20 mM Tris-HCl, 0.5 M NaCl, pH 7.2), and sonicated on ice with 100 short bursts for 4 s each at a power output of 150 W. The suspension was centrifuged at 10,000 × g at room temperature for 5 min then the precipitation was removed. The supernatant was collected and passed over Ni 2+ -NTA agarose gel columns previously equilibrated with buffer A. Stepwise elution was carried out with buffer A containing 0, 20, 40, 60, 80, 100, 200, 300 and 500 mM imidazole, in that order. Eluted fractions containing enzyme activity were collected and stored at − 20 °C until used for subsequent analyses.
For determination of protein purity after Ni 2+ -NTA purification, eluted fractions were loaded for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 12% running gels. The purified protein in the SDS-PAGE gel was identified using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) performed by Tianjin Biochip (Tianjin, China). The protein concentration was determined with Bradford method using bovine serum albumin as the standard 45 .
Enzyme assay and substrate specificity. The sucrose hydrolysis reactions involved incubations at 37 °C for 10 min with a reaction mixture containing 100 μ L of appropriately diluted rInvHJ14 and 900 μ L of McIlvaine buffer (pH 8.0) containing 0.5% (w/v) sucrose. The amount of reducing sugars released was determined by the 3,5-dinitrosalicylic acid (DNS) reagent at 540 nm 46 . One activity unit (U) was defined as the amount of enzyme that produces 1.0 μ mol equimolar mixture of glucose and fructose per minute under the above assay conditions unless otherwise noted.
Substrate specificity assay was performed at 37 °C in McIlvaine buffer (pH 8.0) using 0.5% (w/v) sucralose, raffinose, stachyose, levan, inulin, starch, birchwood xylan, beechwood xylan, wheat flour arabinoxylan or fructooligosaccharides mixture as substrate instead of sucrose. Enzyme catalysis on p-nitrophenyl-β -d-xylopyranoside was assayed using the method described in the previous study 43 . Biochemical characterization. Biochemical characterization of purified rInvHJ14 was determined using sucrose as substrate unless otherwise noted.
The effects of pH and temperature on rInvHJ14 were determined at pH 7.0-10.0 and 0-50 °C, respectively. pH and temperature stabilities of rInvHJ14 were studied respectively by keeping the enzyme solution at 37 °C for 60 min at different pH values (pH 5.0-11.0) and by keeping the enzyme solution at pH 8.0 for 1-60 min at different temperatures (30-50 °C). Residual activities of these pre-incubated samples were then determined under standard assay conditions. The buffers used were McIlvaine buffer for pH 5.0-8.5 and 0.1 M glycine-NaOH for pH 9.0-11.0.
Scientific RepoRts | 6:32081 | DOI: 10.1038/srep32081 In order to determine whether the activity of rInvHJ14 was inhibited by sucrose at high final concentrations, the enzyme was incubated in the reaction mixture with 14.6-2045.0 mM sucrose. Enzyme activity was then determined under standard assay conditions. Kinetic and thermodynamic characterization. K m , V max and k cat values of purified rInvHJ14 were determined using 1.5-36.5 mM sucrose as substrate in McIlvaine buffer (pH 8.0) at 10, 20 and 30 °C. The data were plotted according to the Michaelis-Menten method using the computer software GraphPad Prism (GraphPad Software, San Diego, CA, USA).
The activation energy (E a ), free energy of activation (Δ G * ), enthalpy of activation (Δ H * ), entropy of activation (Δ S * ) and temperature coefficient (Q 10 ) for rInvHJ14 were calculated using the equations as described in previous study 47 . Transfructosylating reactions. In order to determine whether rInvHJ14 possessed self-condensation activity, the enzyme (approximately 0.5 U mL −1 ) was incubated in the reaction mixture with 146 or 584 mM sucrose at 20 °C and pH 8.0 for 17 h. Transglycosylation reactions were carried out at 20 °C and pH 8.0 for 17 h using 146 mM sucrose as the donor with d-galactose and d-xylose as the acceptors at 555 mM and 666 mM, respectively. Transglycosylation products were analyzed by thin layer chromatography (TLC) as previously described 37 .
Accession number. Nucleotide sequence of invHJ14 was deposited in GenBank under the accession number KT943473.