Enhancement of the Brenneria sp. levansucrase thermostability by site-directed mutagenesis at Glu404 located at the “-TEAP-” residue motif

Accepted Manuscript Title: Enhancement of the Brenneria sp. levansucrase thermostability by site-directed mutagenesis at Glu404 located at the “-TEAP-” residue motif Authors: Wei Xu, Jiaying Peng, Wenli Zhang, Tao Zhang, Cuie Guang, Wanmeng Mu PII: DOI: Reference:

S0168-1656(18)30706-5 https://doi.org/10.1016/j.jbiotec.2018.11.021 BIOTEC 8315

To appear in:

Journal of Biotechnology

Received date: Revised date: Accepted date:

26 September 2018 17 November 2018 26 November 2018

Please cite this article as: Xu W, Peng J, Zhang W, Zhang T, Guang C, Mu W, Enhancement of the Brenneria sp. levansucrase thermostability by site-directed mutagenesis at Glu404 located at the “-TEAP-” residue motif, Journal of Biotechnology (2018), https://doi.org/10.1016/j.jbiotec.2018.11.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Enhancement of the Brenneria sp. levansucrase thermostability by site-directed mutagenesis at Glu404 located at the “-TEAP-” residue motif

State Key Laboratory of Food Science and Technology, Jiangnan University, Wuxi,

SC R

†

IP T

Wei Xu†, Jiaying Peng†, Wenli Zhang†, Tao Zhang†, Cuie Guang†, Wanmeng Mu†,§,*

Jiangsu, 214122, China.

International Joint Laboratory on Food Safety, Jiangnan University, Wuxi 214122,

U

§

Corresponding author.

M

*

A

N

China.

Address: State Key Laboratory of Food Science and Technology, Jiangnan University,

ED

Wuxi, Jiangsu, 214122, P. R. China.

Tel: (86) 510-85919161. Fax: (86) 510-85919161.

PT

Email address: [email protected]

CC E

Research Highlights 

Residue Glu404 had a significant influence on the LS thermostabilities.



Single mutants E404L, E404V, E404F, E404I and E404W exhibited an enhanced

A

thermostability and Tm value.



The LS thermostability could be increased via the B-factor-saturation mutagenesis at Glu404.



Increase the hydrophobic interaction within Glu404 could increase the Tm and thermostability of LS.

Abstract Levansucrase (EC 2.1.4.10, LS) has been used in the production of levan and levan-type fructooligosaccharides from sucrose; however, development of further application is restricted due to its poor thermostability. The LS from Brenneria sp.

IP T

EniD312 was engineered using a structure-guided approach. Residue Glu404 was located in the “-TEAP-” motif and varied among LSs with different thermostabilities.

SC R

Site-directed mutagenesis was performed in Glu404 and thermostability was evaluated

by measuring the half-life and structural melting temperature (Tm) of the wild-type LS

U

and its Glu404-mutant variants. The optimal temperature for the Glu404 mutants was

N

similar to that of the wild-type enzyme, however, the Tm of E404L mutant was

A

enhanced by 2.8 °C and the half-life was increased by 12.5- and 1.3- fold at 35 and

M

45 °C, respectively. The other mutants E404W, E404V, E404I, and E404F also

ED

showed a pronounced increase in Tm and thermostability. Finally, the improvement of thermostability of LS through mutation in Glu404 belonging to the “-TEAP”- motif

PT

could be ascribed to the change of microenvironment in the LS structure. The change

CC E

of the micro-environment mainly included the enhanced structural stability between two β-hairpins and the elevated hydrophobic interactions in the overall protein structure. This work proposes new insights into the thermostabilization mechanism of

A

other LSs.

Keywords: levan; levansucrase; site-directed mutagenesis; Glu404; thermostability

1. Introduction Sucrose is a common disaccharide composed of α-glucose and β-fructose by an α1β2 linkage. As cheap and abundant raw material, sucrose can be utilized in various fields. For example, it can be converted into functional carbohydrates, such as

IP T

isomaltulose, isomaltose (Mu et al., 2014), fructooligosaccharide (FOS) (RodriguezAlegria et al., 2010) and maltooligosaccharide (MOS) (Nilsson, 1991). In recent

SC R

years, biotechnologically-derived enzymes have been developed to produce glycan from sucrose, such as glucan and fructan (Fig. 1). The α- Glucan molecule is a

polysaccharide composed of D-glucose monomers, including dextran (Purama et al.,

N

U

2008), mutan (Firouzabadi et al., 2007), alternan (Cote et al., 2009) and reuteran

A

(Pijning et al., 2012), which have different glycosidic linkage types. The novel and

M

healthy dietary fiber made from glucan has gradually replaced the traditional highcalorie options (Moreno-Mendieta et al., 2017). In addition, β-glucans are important

ED

drug candidate for antifungal medications (Sikora et al., 2013).

PT

Fructan is a polymer consisting of fructose molecules, primarily levan and inulin (Van den Ende, 2013). The excellent water solubility of fructan enables its usage as a

CC E

novel micro hydrogel in the medicinal and food industry (Liu et al., 2017). In early 2001, levan-type fructan was approved by the European Food Safety Authority

A

(EFSA), Food and the Drug Administration (FDA), and the Food Standards in Australia and New Zealand (FSANZ) (Li et al., 2015). Industrial interests in levan and levan-type oligosaccharides have arisen due to their ability to accelerate growth of Bifidobacterium (Korakli et al., 2003), improve bowel condition, reduce fat absorption (Annarita et al., 2009), and balance blood sugar level (Yoo et al., 2004).

Because of this, levan has been widely used as emulsifier, stabilizer, surfactant, and humectant in the food, medical and chemical industries (Chen et al., 2014). Based on its definition in the CAZY database, LS (LS, EC 2.1.4.10) belongs to the glycoside hydrolases 68 (GH 68) family and “GH-J” clan (Kang et al., 2009).

IP T

Inulosucrase (IS, EC 2.4.1.9) and β-fructofuranosidase (EC 3.2.1.26) are two other members of GH 68, but LS is distinguishable from these two enzymes. IS specifically

SC R

catalyzes the formation of inulin and β-fructofuranosidase only generates oligo-

fructan with DP ≤ 5 (Pedezzi et al., 2014). Microbial levansucrase is an important

U

source for the production of levan from sucrose via transfer of fructosyl group to the

N

free polyfructose (fructan). LS can catalyze three reactions: hydrolysis,

A

transfructosylation, and polymerization, by accepting water, low molecular mass

M

fructooligosaccharides (FOS) and growing fructan as fructosyl acceptors, respectively

ED

(Li et al., 2015). However, previous studies showed that most LSs showed a low optimal temperature (<50 °C) and weak thermostability for levan production (Li et al.,

PT

2015). A few studies have reported the methods to improve LS thermostability. For

CC E

instance, immobilization of the Bacillus amyloliquefaciens LS onto glyoxyl agaroseIDA/Cu and glyoxyl agarose resulted in a 106-fold increase of the thermostability at 55 °C (Hill et al., 2016). Additionally, using cross-link reagents to conjugate the

A

Bacillus circulans LS increased its half-life from 130 min to 347 min at 50 °C by 1.7fold (El-Refai et al., 2009). However, the exact determinants responsible for LSs thermostability at the structural level are still unknown, even after the recent determination of four LSs crystal structures (Meng and Futterer, 2003).

Until now, LSs have been biochemically characterized from Gram-positive and Gram-negative bacteria, such as the Bacillus, Geobacillus, Lactobacillus, Weissel, Erwinia, Leuconostoc, Pseudomonas, Zymomonas, and Brenneria species (Li et al., 2015). Although most LSs displayed low optimal temperatures, some other LSs

IP T

showed a much higher thermostability. The LS from Bacillus sp. TH4-2 was stable up to 50 °C (Ben Ammar et al., 2002), and the G. stearothermophilus LS could retain all

SC R

initial activity after a 6 h incubation at temperatures ranging from 4 to 47 °C

(Inthanavong et al., 2013). According to previous studies and multiple sequence

U

alignment of LSs shown in Fig. 2, these LSs could be separated into two groups. LSs

N

belonging to Group I had a weak thermostability, whereas LSs belonging to Group II

A

were found to have a stronger thermostability, including the LSs from G.

M

stearothermophilus, L. reuteri LTH5448 (Ni et al., 2018), and G. diazotrophicus

ED

(Trujillo et al., 2001) (Table S2). Notably, a highly conserved motif “-TEAP-” was observed for all the Group I LSs, but for the Group II LSs, the glutamate residue in

PT

position 404 was replaced by residues such as leucine, tryptophan, and phenylalanine.

CC E

These residues were either hydrophobic residues or with a large side-chain. All of these results indicated that the introduction of hydrophobic residues in the 404 position might influence the thermostability of LS. Herein, based on the multiple

A

sequence alignment and structure information analysis, site-directed mutations in Glu404 were designed to evaluate the role of Glu404 in the thermostability of Brenneria sp. EniD312 LS.

2.1 Materials and methods 2.2 Chemicals, reagents, and strains Escherichia coli DH5α and BL21 (DE3) cells were purchased from Sangon Biological Engineering Technology and Services (Shanghai, China), and used as the

IP T

host for DNA manipulation and to express the target protein, respectively. Isopropylβ-D-1-thiogalactopyranosid (IPTG), ampicillin (Amp), and other chemicals of

SC R

analytical grade were purchased from Sangon Biological Engineering Technology & Services Co., Ltd (Shanghai, China). The reagents for protein electrophoresis were

U

obtained from Bio-Rad (Hercules, CA, USA). The Ni2+-chelating affinity

N

chromatography resin was purchased from GE (Maryland Heights, MO, USA). Luria-

A

Bertani (LB) medium consisted of 10 g/L NaCl, 10 g/L tryptone and 5 g/L yeast

M

extract. NdeI, and XhoI restriction endonuclease and primers were designed in Primer

ED

5.0 and synthesized by Shanghai Generay Biotech Co., Ltd (Shanghai, China). The gene sequence determination and the expression of Brenneria sp. EniD312 LS

PT

have been reported by Xu (Xu et al., 2018). In this article, the full length of the

CC E

encoding gene (GenBank accession No. EHD23269.1) was synthesized by Shanghai Generay Biotech Co., Ltd. (Shanghai, China). The synthesized gene was designed to contain an in-frame 6× histidine-tag sequence before the termination codon, and the

A

restriction enzyme sites NdeI and XhoI were designed at the 5′- and 3′-terminus. Then, the synthesized gene was fused into pET-22b (+) vector with the same restriction sites.

2.2 Preparation of the wild-type LS and its mutant variants After transformation and culturing, the cells were harvested by centrifugation (10000 × g, 20 min, 4 °C), and the pellets were washed twice with 50 mmol/L phosphate buffer containing 100 mmol/L NaCl (pH 6.0). The lysate was disrupted by

France) for 18 min (1 s sonication with 2 s breaks). Fast Protein Liquid

IP T

sonication using a Vibra-Cell TM72405 Sonicator (BioBlock Scientific, Illkirch,

Chromatography (FPLC, ÄKTA Purifier System, GE Healthcare) was utilized to

SC R

purify the recombinant enzyme. Firstly, the supernatants were filtered through a 50 × 0.45 μm water phase microporous membrane, and 15 mL crude enzyme was loaded

U

onto a 8.9 × 64 mm column containing Ni2+-chelating Sepharose Fast Flow resin (GE

N

Healthcare). The column was washed with deionized water by 5 column volume (CV)

A

at 1 mL/min to remove the 20% ethanol, and equilibrated with the fresh binding

M

buffer (50 mM sodium phosphate buffer, 500 mM NaCl, pH 6.5) at 0.8 mL/min by 10 CV. Secondly, the washing buffer (50 mM sodium phosphate buffer, 500 mM NaCl,

ED

50 mM imidazole, pH 6.5) was applied to wash proteins that were nonspecifically bound to the resin, with the flow rate 1 mL/min by 5 CV. Lastly, the elution buffer (50

PT

mM sodium phosphate buffer, 500 mM NaCl, 500 mM imidazole, pH 6.5) was used to elute the bound target protein at 1 mL/min by 6 CV, and the protein fraction

CC E

detected by the HD-3 protein detector at UV 280 nm. The target protein fused with histidine-tag was pooled and dialyzed overnight against 50 mM sodium phosphate

A

buffer (pH 6.5) 4 times over 24 h. Coomassie Brilliant Blue R250 was used for protein staining, and protein concentration was determined by the method described by Bradford using bovine serum albumin as a standard (Bradford, 1976).

2.3 Molecular modelling

Referring to the method with a minor modification, the modelling was performed by submitting the amino acid sequences into the Phyre online server (http: //www.sbg. bio.ic.ac.uk/phyre2/html/page.cgi?id=index) and the quality of built model was verified by the SAVES online Server using its inbuilt modules (Xu et al., 2018). The

IP T

models were further optimized with Discovery Studio package (San Diego, CA, USA) using energy minimization to eliminate unreasonable contacts. The structures were

SC R

rendered with PyMOL Molecular Graphics and Visual System (https://pymol.org/2/).

U

2.4 Site-directed mutation

N

Site-directed mutagenesis of the Brenneria sp. EniD312 LS was achieved using

A

one-step PCR method. The primers for mutagenesis are listed in Table 1. The

M

recombinant plasmid containing the original gene encoding for Brenneria sp.

ED

EniD312 LS was used as the template. The routine PCR process was performed as follows: 1 cycle at 94 °C for 5 min, 35 cycles at 94 °C for 30 s, 58 °C for 1 min, and

PT

72 °C for 1 min, and 72 °C for last 10 min, using Taq Plus DNA polymerase. The

CC E

amplified PCR products were processed using the DpnI restriction enzyme, and the nucleotide sequences of these several mutant variants were verified by DNA sequencing by Sangon Biological Engineering Technology and Services (Shanghai,

A

China).

2.5 Effect of pH and temperature on enzyme activity. The optimum pH of the recombinant LS and mutants was determined by measuring

its relative activity with three buffer systems, including sodium acetate (50 mmol/L, pH 4.5 - 5.5), sodium phosphate (50 mmol/L, pH 6.0 - 7.0), and Tris-HCl (50 mmol/L, pH 7.5 - 9.0). The temperature dependence of the recombinant enzyme activity was measured at 30, 35, 40, 45, 50 and 55 °C in 50 mmol/L phosphate buffer

IP T

at pH 6.5. The reaction was initiated by the addition of the enzyme and was allowed

2.6 Effect of mutation on LS thermostability and activity

SC R

to proceed for 20 min. All experiments were independently performed in triplicate.

U

The LS activity was assayed using sucrose as the sole substrate by measuring total,

N

hydrolysis, and transfructosylation activities. Total activity was calculated by

A

determining the release of glucose from sucrose, and hydrolysis activity was

M

measured based on the release of fructose from sucrose. Then, the transfructosylation

ED

activity was calculated as the difference between total and hydrolysis activities by measuring the difference between the amount of glucose and fructose from sucrose. A

PT

reaction mixture of 1.5 mL included 200 g/L sucrose, 50 mmol/L sodium phosphate

CC E

buffer (pH 6.5), and purified enzyme with a final concentration of 5 μg/mL. The reactions were performed at 35 °C for 20 min and terminated by heating in boiling water for 20 min. One unit of total activity and hydrolysis activity were defined as the

A

amount of enzyme catalyzing the release of 1 μmol glucose and fructose per min, respectively. Unless stated otherwise, the LS activity was described as the total activity in this manuscript. Thermostability was examined by measuring the residual activity of LS after incubating the enzyme at various temperatures (35, 45, and

55 °C). The LS activity without incubation was set as 100%.

3. Results and discussion 3.1 Homology modelling

IP T

A previous study showed that the Brenneria sp. EniD312 LS displayed 87% and 71% identity with the LSs from B. goodwinii and E. amylovora, respectively (Xu et

SC R

al., 2018). However, the protein structure of the former has not yet been determined. Given this, the solved crystal structure of E. amylovora LS (PDB accession number:

U

4D47) was used as a template for modelling the structure of Brenneria sp. EniD312

N

LS (Wuerges et al., 2015). After several cycles, the optimal model was obtained and

A

showed in Fig. 3 (A). The built structure of Brenneria sp. EniD312 LS (cyan) and E.

M

amylovora LS (green) were very similar, and the root-mean-square deviation (RMSD)

ED

value between the two structures by superimposition was 0.350. All of these analyses

PT

suggested that the built 3D model was acceptable for further analysis.

CC E

3.2 Selection of the mutation sites The B-factor reflects the diffusion of the electron density when solving the protein

structure using X-ray diffraction spectra, and a higher B-factor value indicates lower

A

structure stability. Because of this, targeting residues with high B-factors has been recently proposed as a strategy for LS thermostabilization (Ge et al., 2018; Reetz et al., 2006). As the LS from Brenneria sp. EniD312 showed 71% identity with the E. amylovora LS, the B-factors of E. amylovora LS were calculated using the HotSpot

Wizard3.0 online server (https://loschmidt.chemi.muni.cz/hotspotwizard/program) (Bendl et al., 2016) and presented in PyMOL. The average B-factors for each residue in E. amylovora LS varied from 2 to 50 and were rendered in the colours blue, green, and red (Fig. 3B). Correspondingly, the four residue groups in Brenneria sp. EniD312

IP T

LS were presented and labelled as: group 1 – I15/G21/I22/K25/P27; group 2 – E105/Q106/G107/N108; group 3 – E395/D396; and group 4 – G401/G402/E404 (Fig.

SC R

3C).

According to a previous study, the influence of loops on the overall protein stability

U

was complex, and properties of a loop could be significantly different from others due

N

to its position, flexibility, length, and micro-physicochemical environment (Ahmad et

A

al., 2012). Herein, group 1, 2 and 3 will not be discussed in detail. Group 4 consisted

M

of G401/G402/E404 and three residues located between two β-hairpins as showed in

ED

Fig. 6. Additionally, the residue Glu404 belonged to the “-TEAP-” motif and varied among LSs with different thermostabilities. Combined with the sequence analysis

PT

results, site-directed mutagenesis was performed in Glu404 to generate mutants and

CC E

investigate the thermostabilities between the wild-type LS and its mutants.

3.3 Effect of pH and temperature on enzyme activity.

A

The wild-type LS and its Glu404 mutant variants were all purified and used to

determine their optimal reaction conditions (Fig. S1). To determine the optimum temperature for the wild-type LS and its mutants, the relative activity was measured at pH 6.5 and varied temperature ranging from 30 °C to 55 °C with 5 °C intervals. The

optimum pH of the enzyme was measured at 45°C using three buffer systems at pH values ranging from pH 4.5 to pH 9.0. The results indicated that, all of the LS mutants had pH and temperature activity profiles similar to that of the wild-type LS from

IP T

Brenneria sp. EniD312 (Fig. 4 A and B).

3.4 Comparison of the activities and thermostabilities between the wild-type LS and

SC R

its mutant variants

All LS mutant variants in Glu404 displayed a decrease of the total activity compared

U

with that of the wild-type LS (Table 2). However, except for the E404K mutant, the

N

Km of other Glu404 mutants did not show a large derivation from those of the wide-

A

type LS. This indicated that the Glu404 residue did not participate in the sucrose

M

substrate binding, which was consistent with previous results (Xu et al., 2018).

ED

However, although the E404V, E404F, E404L, E404I and E404W mutants did not show an increase in total activity, the corresponding thermostability of mutants all

PT

displayed a clear enhancement compared with the wild-type LS (Fig. 4 C and D). The

CC E

mutant E→L in Glu404 resulted in a 12.5- and 1.3- fold increase of the half-life at 35 and 45 °C, respectively. In addition to E404L, the thermostability of E404V, E404I, and E404F mutants containing hydrophobic side-chains was simultaneously increased

A

to a different extent. Thus, we hypothesized that introducing hydrophobic residues in Glu404 would increase the thermostability of LS. Additionally, the E404W caused a prominent increase in thermostability despite of the weak hydrophobic index of tryptophan (Kyte and Doolittle, 1982) (Table S1).

3.5 Analysis of the thermostable behaviour for mutants As mentioned above, the residue Glu404 was selected for site-directed mutation due to its special location in the motif “-TEAP-”. Furthermore, to determine its special

obtained by submitting the built model to PROCHECK server

IP T

location, a more detailed wiring diagram of the secondary structure in the LS was

SC R

(http://servicesn.mbi.ucla.edu/PROCHECK/). As a result, from the residue Ala80 to Tyr425, approximately 15 β-hairpins were showed in the diagram for LS from

U

Brenneria sp. EniD312. (Fig. 5) Strikingly, although the short loop formed by the “-

N

TEAP-” motif was located away from the catalytic core (D68, D225 and E309), it

A

connected with 14th β-hairpin (S390-T400) and 15th β-hairpin (P406-Y425) of the LS

M

and hold an important potential “channel” for the product’s release (Xu et al., 2018).

ED

In our previous study, the protein denaturing temperature (Tm) was used as an assistant criteria for characterizing the protein structural stability, and a higher Tm

PT

might indicate a more stable protein (Yu et al., 2016). Herein, the Tm of the wild-type

CC E

LS and its five mutants with increased thermostability were detected and compared (Table 3). It was found that the E404V, E404F, and E404I mutants had a slight increase of the Tm value compared with that of the wild-type LS. Strikingly, the

A

largest increase measured approximately 2.8 °C and was obtained in E404L (Fig. S2). Although several loops located at the rim of LS active site funnel have proven significant in determining the levan chain length as well as the levansucrase catalytic properties, the exact determinants of LS for thermostability remained unclear

(Wuerges et al., 2015). Three points summarized how the thermostability was improved. Firstly, based on the previous studies, the appearance of tryptophan residue in the structural motif, especially in the short peptide could stabilize the β-hairpin conformations by increasing the proteolytic resistance (Cochran et al., 2001). Herein,

IP T

it might be inferred that the mutation Glu404→Trp404 resulted in an increase in the structural stability between the 14th β-hairpin (S390-T400) and 15th β-hairpin (P406-

SC R

Y425). Furthermore, the loop9 (E395-D396) in the 14th β-hairpin was speculated to be a slope of the “channel” affecting the transfructosylation activity of LS. An

U

increase in the structural stability of these two β-hairpins could indirectly enhance the

N

stability of the 9th loop and thus cause an enhancement of the thermostability.

A

Secondly, the model for E404L, E404V, E404F, and E404L mutants were also built

M

and compared with that of the wild-type LS (Fig. 7). The residues D68, D225, and

ED

E309 were indentified as the nucleophile, transition stabilizer, and general acid/base in the Brenneria sp. EniD312 LS, respectively, and these three residues constituted a

PT

catalytic central core in the deep bottom of LS (Xu et al., 2018). When introducing

CC E

Leu404, Ile404, Phe404 and Val404 mutants, the protein hydrophobicity might be greatly increased according to the hydrophobic index (Table S1). In turn, all the hydrophobic side-chain stretched into the catalytic cleft. Generally, the protein hydrophobicity is

A

the main driving force for protein folding and increase the interplay of residues with hydrophobic side-chains inside the protein can alter the micro-physicochemical environment and increase protein stability (Chandler, 2005). Thus, we proposed that the introduction of hydrophobic residues in the 404 position increased the

hydrophobic micro-environment by altering the orientation of side-chains, and as a result, increased the thermostability. Thirdly, compared with the N-terminal and Cterminal residues of Brenneria sp. EniD312 LS, a hydrophobic cluster was strikingly located found between two sequences (L39-V59; I400-F423). The hydrophobic

IP T

stretch of amino acids with a repeating “-PXX-” motif has been mentioned in the IS from L. johnsonii NCC 533, but the function of hydrophobic interaction of N-

SC R

terminals and C-terminals in affecting the protein structure stability has not been

discussed yet (Anwar et al., 2008). The clamp-like pattern formed by the N-terminal

U

and C-terminal residues in maintaining the protein stability has not been thoroughly

N

investigated. On the whole, when Glu404 was altered to residues with high

A

hydrophobic indices, the hydrophobic interaction formed by the “clamp” was

M

strengthened, thereby altering the micro-physicochemical environment and enhancing

PT

4. Conclusions

ED

the overall protein structure (Fig. 8).

CC E

Notably, the five mutants E404L, E404V, E404F, E404I and E404W exhibited an enhanced thermostability and Tm in this study, suggesting that these mutants have the potential to provide a better performance than that of the wild-type LS in industrial

A

applications. Finally, the improvement of thermostability of LS through designing mutants in Glu404 belonging to the “-TEAP”- motif could be ascribed to the change in micro-environment in the LS structure. The change of the micro-environment mainly included the enhanced structural stability between two β-hairpins, and the elevated

hydrophobic interaction in the overall protein structure. However, this enhancement behaviour is not limited to hydrophobic interactions; it also includes interactions, such as hydrogen bonds, ionic bonds and disulfide bonds. These interactions should be further investigated. However, this is the first report about the enhancement of LS

IP T

thermostability and could be used as a guideline for the thermostabilization for other

SC R

LSs.

ACKNOWLEDGEMENTS

This work was supported by the Support Project of Jiangsu Province (No. 2015-

U

SWYY-009), and the Research Program of State Key Laboratory of Food Science and

N

Technology, Jiangnan University, Project No. SKLF-ZZA-201802 and SKLF-ZZB-

A

201814. Postgraduate Research & Practice Innovation Program of Jiangsu Province

ED

M

(KYCX18_1777).

SUPPORTING INFORMATION

PT

Table S1. Hydrophobic index and side-chain of 20 amino acids. Table S2. Comparison of the thermostabilites of LSs. Table S3. Tm value of all the Glu404 mutant. Figure S1.

CC E

SDS-PAGE analysis of the wild-type LS and its Glu404 variants. Fig. S2, Tm value of the wild-type LS and its E404I, E404W, E404F, E404V, and E404L variants.

A

AUTHOR CONTRIBUTIONS X.W., S.Y., W.Z., T.Z., C.G., and W.M. conceived and designed the experiments. X.W. and D.N. performed the experiments. X.W., S.Y., and W.M. analyzed the data. S.Y., X.W., and W.M. wrote the manuscript with contributions of all authors. All authors reviewed the manuscript.

References: Annarita, P., Hande, K., Bahar, G., Giuseppina, T., Giuseppina. P., Ebru. T.Ö., Barbara, N., 2009. High level synthesis of levan by a novel Halomonas species growing on defined media. Carbohydr. Polym 78, 651-657. Ahmad, S., Kumar, V., Ramanand, K. B., Rao, N. M., 2012. Probing protein stability

IP T

and proteolytic resistance by loop scanning: a comprehensive mutational analysis. Protein Sci 21, 433-446.

Anwar, M.A., Kralj, S., van der Maarel, M.J., Dijkhuizen, L., 2008. The probiotic

SC R

Lactobacillus johnsonii NCC 533 produces high-molecular-mass inulin from

sucrose by using an inulosucrase enzyme. Appl. Environ. Microbiol 74, 3426-3433. Ben Ammar, Y., Matsubara, T., Ito, K., Iizuka, M., Limpaseni, T., Pongsawasdi, P.,

U

Minamiura, N., 2002. Characterization of a thermostable levansucrase from

A

temperature. J. Biotechnol 99, 111-119.

N

Bacillus sp. TH4-2 capable of producing high molecular weight levan at high

M

Bendl, J., Stourac, J., Sebestova, E., Vavra, O., Musil, M., Brezovsky, J., Dambrosky, J., 2016. Hotspot Wizard 2.0: automated design of site-specific mutations and smart

ED

libraries in protein engineering. Nucleic Acids Res 44, 479-487. Bradford, M.M., 1976. A Rapid and Sensitive Method for the Quantification of

PT

Microgram Quantities of Protein Utilizing the Principle of Protein-Dye Binding. Anal. Biochem 72, 248-254.

CC E

Chandler, D., 2005. Interfaces and the driving force of hydrophobic assembly. Nature 437, 640-647.

Cote, G.L., Dunlap, C.A., Vermillion, K.E., 2009. Glucosylation of raffinose via

A

alternansucrase acceptor reactions. Carbohydr. Res 344, 1951-1959.

Chen, X., Gao, H., Ploehn, H., 2014. Montmorillonite–levan nanocomposites with improved thermal and mechanical properties. Carbohydr. Polym 101,

565-573.

Cochran, A.G., Skelton, N.J., Starovasnik, M.A., 2001. Tryptophan zippers: stable, monomeric beta -hairpins. Proc. Natl. Acad. Sci 98, 5578-5583. El-Refai, H.A., Abdel-Fattah, A.F., Mostafa, F.A., 2009. Enzymic synthesis of levan

and fructo-oligosaccharides by Bacillus circulans and improvement of levansucrase stability by carbohydrate coupling. World J. Microbiol. Biotechnol 25, 821-827. Firouzabadi, F., Kok-Jacon, G., Vincken, J., Ji, Q., Suurs, L., Visser, R., 2007. Fusion proteins comprising the catalytic domain of mutansucrase and a starch-binding domain can alter the morphology of amylose-free potato starch granules during biosynthesis. Transgenic Res 16, 645-656.

IP T

Ge L., Li, D., Wu, T., Zhao, L., Ding, G., Wang, Z.Z., 2018. B-factor-saturation

mutagenesis as a strategy to increase the thermostability of α-L-rhamnosidase from

SC R

Aspergillus terreus. J. Biotechnol 275, 17-23.

Hill, A., Karboune, S., Mateo, C., 2016. Immobilization and stabilization of levansucrase biocatalyst of high interest for the production of

U

fructooligosaccharides and levan. J. Chem. Technol. Biotechnol 91, 2440-2448.

N

Inthanavong, L., Tian, F., Khodadadi, M., Karboune, S., 2013. Properties of

A

Geobacillus stearothermophilus levansucrase as potential biocatalyst for the synthesis of levan and fructooligosaccharides. Biotechnol. Progress 29, 1405-1415.

M

Korakli, M., Pavlovic, M., Ganzle, M.G., Vogel, R.F., 2003. Exopolysaccharide and

ED

kestose production by Lactobacillus sanfranciscensis LTH2590. Appl. Environ. Microbiol 69, 2073-2079.

Kyte, J., Doolittle, R. F., 1982. A simple method for displaying the hydropathic

PT

character of a protein. J. Molecule Biol 157, 105-132. Kang, J., Kim, Y.M., Kim, N., Kim, D.W., Nam, S.H., Kim, D., 2009. Synthesis and

CC E

characterization of hydroquinone fructoside using Leuconostoc mesenteroides levansucrase. Appl. Microbiol. Biotechnol 83, 1009-1016.

A

Liu, Q., Yu, S.H., Zhang, T., Jiang, B., Mu, W.M. 2017, Efficient biosynthesis of levan from sucrose by a novel levansucrase from Brenneria goodwinii. Carbohydr. Polym 157, 1732-1740.

Li, R., Zhang, T., Jiang, B., Mu, W., Miao, M., 2015. Purification and characterization of an intracellular levansucrase derived from Bacillus methylotrophicus SK 21.002. Biotechnol. Appl. Biochem 62, 815-822. Li, W.J., Yu, S.H., Zhang, T., Jiang, B., Stressler, T., Fischer, L., Mu, W.M., 2015.

Efficient Biosynthesis of Lactosucrose from Sucrose and Lactose by the Purified Recombinant Levansucrase from Leuconostoc mesenteroides B-512 FMC. J. Agric. Food Chem 63, 9755-9763. Li, W.J., Yu, S.H., Zhang, T., Jiang, B., Mu, W.M., 2015. Recent novel applications of levansucrases. Appl. Microbiol. Biotechnol 99, 6959-6969. Mu, W.M., Li, W.J., Wang, X., Zhang, T., Jiang, B., 2014. Current studies on sucrose

IP T

isomerase and biological isomaltulose production using sucrose isomerase. Appl. Microbiol. Biotechnol 98, 6569-6582.

SC R

Moreno-Mendieta, S., Guillen, D., Hernandez-Pando, R., Sanchez, S., Rodriguez-

Sanoja, R., 2017. Potential of glucans as vaccine adjuvants: A review of the alphaglucans case. Carbohydr. Polym 165, 103-114.

U

Meng, G., Futterer, K., 2003. Structural framework of fructosyl transfer in Bacillus

N

subtilis levansucrase. Nat. Struct. Biol 10, 935-941.

A

Nilsson, K.G.I., 1991. Use of Glycosidases and Glycosyltransferases in the Synthesis of Complex Oligosaccharides and Their Glycosides. ACS Symp. Ser 466, 51-62.

M

Ni, D., Xu, W., Bai, Y., Zhang, W.L., Zhang, T., Mu, W.M., 2018. Biosynthesis of

ED

levan from sucrose using a thermostable levansucrase from Lactobacillus reuteri LTH5448. Int. J. Biol. Macromol 113, 29-37. Purama, R.K., Goyal, A., 2008. Identification, effective purification and functional

PT

characterization of dextransucrase from Leuconostoc mesenteroides NRRL B-640. Bioresource Technol 99, 3635-3642.

CC E

Pijning, T., Vujicic-Zagar, A., Kralj, S., Dijkhuizen, L., Dijkstra, B.W., 2012. Structure of the alpha-1,6/alpha-1,4-specific glucansucrase GTFA from

A

Lactobacillus reuteri 121. Acta Crystallogr., Sect. F: Struct. Biol. Cryst. Commun 68, 1448-1454.

Pedezzi, R., Fonseca, F.P.P., Santos Júnior, C.D., Kishi, L., Terra, W. R., HenriqueSilva, F., 2014. A novel beta-fructofuranosidase in Coleoptera: Characterization of a beta-fructofuranosidase from the sugarcane weevil, Sphenophorus levis. Insect Biochem. Mol. Biol 55, 31-38. Rodriguez-Alegria, M.E., Enciso-Rodriguez, A., Ortiz-Soto, M.E., Cassani, J., Olvera,

C., Munguia, A.L., 2010. Fructooligosaccharide production by a truncated Leuconostoc citreum inulosucrase mutant. Biocatal. Biotransform 8, 51-59. Robert, X., Gouet, P., 2014. Deciphering key features in protein structures with the new ENDscript server. Nucleic Acids Res 42, 320-324. Reetz, M.T., Carballeira, J.D., Vogel, A., 2006. Iterative Saturation Mutagenesis on the Basis of B Factors as a Strategy for Increasing Protein Thermostability. Angew.

IP T

Chem., Int. Ed 45, 7745-7751.

Sikora, P., Tosh, S.M., Brummer, Y., Olsson, O., 2013. Identification of high beta-

SC R

glucan oat lines and localization and chemical characterization of their seed kernel beta-glucans. Food Chem. 137, 83-91.

Trujillo, L., Arrieta, J., Dafhnis, F., Garcı́a, J., Valdés, J., Tambara, Y., Pérez, M.,

U

Hernández, L., 2001. Fructo-oligosaccharides production by the Gluconacetobacter

A

Enzyme Microbiol. Technol 28, 139-144.

N

diazotrophicus levansucrase expressed in the methylotrophic yeast Pichia pastoris.

Wuerges, J., Caputi, L., Cianci, M., Boivin, S., Meijers, R., Benini, S., 2015. The

M

crystal structure of Erwinia amylovora levansucrase provides a snapshot of the

ED

products of sucrose hydrolysis trapped into the active site. J. Struct. Biol 191, 290298.

Xu, W., Ni, D.W., Yu, S.H., Zhang, T., Mu, W.M., 2018. Insights into hydrolysis

PT

versus transfructosylation: mutagenesis studies of a novel levansucrase from Brenneria sp. EniD312. Int. J. Biol. Macromole 116, 335-345.

CC E

Yoo, S.H., Yoon, E.J., Cha, J., Lee, H.G., 2004. Antitumor activity of levan polysaccharides from selected microorganisms. Int. J. Biol. Macromol 34, 37-41.

A

Yu, S.H., Wang, X., Zhang, T., Jiang, B., Mu, W., 2016. Probing the Role of Two Critical Residues in Inulin Fructotransferase (DFA III-Producing) Thermostability from Arthrobacter sp. 161MFSha2.1. J. Agr. Food Chem 64, 6188-6195.

Van den Ende, W., 2013. Multifunctional fructans and raffinose family oligosaccharides. Front. Plant Sci 4, 1-11.

Figure Legends

Fig. 1 Summary of the fructan and glucan biosynthesis from sucrose by microbial fructosyl transferases and glucosyl transferases. The EC number for each enzyme was

IP T

labelled.

SC R

Fig. 2 Multiple sequence alignment of LS and their homologs. Amino acid sequence of LS from Brenneria. sp EniD312 (GeneBank accession No: EDH23269.1) was

U

aligned with LSs from Brenneria goodwinii (CPR14579.1), Erwinia amylovora

N

(CAA52972.1), Leuconostoc mesenteroides B-512 FMC (AAT81165.1), Rahnella

A

aquatilis JCM-1683 (AAK14974.1), Rahnella aquatilis ATCC33071 (AAC36458.1),

M

Bacillus megaterium DSM319 (ADF38395.1), Bacillus licheniformis 8-37-0-1

ED

(AGZ16261.1), Bacillus licheniformis RN-01 (ACI15886.1), Bacillus amyloliquefaciens (ACD39394.1), Bacillus subtilis (CAA26513.1), Geobacillus

PT

stearothermophilus ATCC 12980 (AAB97111.1) Leuconostoc citreum CW28

CC E

(AA025086.1), Leuconostoc mesenteroides NRRL B-512 (AAY19523.1), Weissella confusa MBFCNC-2, Lactobacillus reuteri 121 (AAO14168.1), Lactobacillus sanfranciscensis TMW 1.392 (CAD48195.1), Lactobacillus reuteri LTH 5448

A

(AJ812736) and Gluconacetobacter diazotrophicus SRT4 (AAB36606.1). The strictly conserved residues were shown on a red background, and the highly conserved residues are shown in red type and boxed in blue. The alignment was performed using ESPript (Robert and Gouet, 2014).

Fig. 3 (A) A superimposition of the built model for Brenneria. sp EniD312 LS (cyan) using the crystal structure of E. amylovora LS (PDB ID: 4D47) (green) as a template. (B) B-factor analysis of the crystal structure for E. amylovora LS and with values

IP T

between 2 to 50 and color in blue, green, and red. The model was shown in the cartoon model. (C) Selected sites in Brenneria. sp EniD312 LS model according to

SC R

the high B-factor values of E. amylovora LS and shown in spheres, including the

group 1, I15/G21/I22/K25/P27 (magenta); group 2, E105/Q106/G107/N108 (dark

U

blue); group 3, E395/D396 (green) and group 4, G401/G402/E404 (cyan). The rest of

A

N

the carbon skeleton was shown in cartoon and rendered with grey.

M

Fig. 4 Effect of temperature (A) and pH (B) on the activtiy of wild-type LS and its

PT

(C) and 45 °C (D).

ED

mutants; comparsion of the thermostability of the wild-type LS and its mutants at 35

CC E

Fig. 5 Wiring diagram of 15 pairs of potential β-hairpins in the Brenneria. sp EniD312 LS model and the potential interaction between two β-strands in every β-hairpin was presented with double-headed arrow in purple. For each β-hairpin, residues

A

consistuting two β-strands were labelled as the single letter and delineated with blue and green respectively, while the others consistuting the loop were coloured by red. Additionally, the residues in group 2 (E105/Q106/G107/N108), group 3 (E395/D396) and group 4 (G401/G402/T403/E404/A405) were circled with a hollow circle in

black. The formed loop by G401/G402/T403/E404/ was showed by a black arc.

Fig. 6 A cartoon model of the Brenneria. sp EniD312 LS model. The loop formed by residues in the N-terminal was delineated with dark blue while the “-TEAP-” loop

IP T

was magneta. Two β-hairpins were rendered with red, and the potential channel for substrate entry and product release was shown by a black arrow. Hydrophobic

SC R

residues L52, V53, F57 and V59 located in the N-terminal were labelled in the model. Additionally, the substrate sucrose was shown in the stick model, with the carbon

N

U

selekton marked with yellow and the hydroxyl group with red.

A

Fig. 7 Change in the active pockets after introducing a mutation in the 404 position.

M

(A), (B), (C), and (D) are the active pockets with surface representation about the

ED

mutation E404F, E404I, E404V and E404W of Brenneria. sp EniD312 LS, respectively. The potential catalytic core formed by D68, D225 and E309 was

PT

highlighted with an elliptic area. All the hydrophobic residues were shown in the stick

CC E

model with the carbon seleketon in green, hydroxyl group in red, and amino group in blue, respectively.

A

Fig. 8 The “clamp-like” pattern formed by hydrophobic clusters in Brenneria. sp EniD312 LS. Partial hydrophobic residues in the N-terminal, including L39, V41, L52, V53, F57 and V59 were labelled in blue, whilst the residues in the C-terminal, including V408, L409, F410, F418, V419, L420, F423 and I428, were colored in red.

A ED

PT

CC E

IP T

SC R

U

N

A

M

A ED

PT

CC E

IP T

SC R

U

N

A

M

A ED

PT

CC E

IP T

SC R

U

N

A

M

A ED

PT

CC E

IP T

SC R

U

N

A

M

A ED

PT

CC E

IP T

SC R

U

N

A

M

Table 1. Primers of the site-directed mutagenesis for LS from Brenneria sp. EniD312. Mutations

Forward (5’→3’) E404V

GTATCGGAGGAACCGTCGCTCCCACGGTATTG

E404F

GTATCGGAGGAACCTTCGCTCCCACGGTATTG

E404G

ATCGGAGGAACCGGGGCTCCCACGGTATTG

E404D

ATCGGAGGAACCGACGCTCCCACGGTATTG

E404R

ATCGGAGGAACCAGGGCTCCCACGGTATTG

E404L

ATCGGAGGAACCTTGGCTCCCACGGTATTG

E404T

ATCGGAGGAACCACCGCTCCCACGGTATTG

E404H

ATCGGAGGAACCCACGCTCCCACGGTATTG

E404Q

ATCGGAGGAACCCAGGCTCCCACGGTATTG

E404M

ATCGGAGGAACCATGGCTCCCACGGTATTG

E404P

ATCGGAGGAACCCCCGCTCCCACGGTATTG

E404C

ATCGGAGGAACCCGCGCTCCCACGGTATTG

E404I

ATCGGAGGAACCATTGCTCCCACGGTATTG

E404S

ATCGGAGGAACCTCCGCTCCCACGGTATTG

E404N

ATCGGAGGAACCAACGCTCCCACGGTATTG

E404W

ATCGGAGGAACCTGGGCTCCCACGGTATTG

E404Y

ATCGGAGGAACCTACGCTCCCACGGTATTG

E404K

ATCGGAGGAACCAAAGCTCCCACGGTATTG

A

N

U

SC R

IP T

GTATCGGAGGAACCGCCGCTCCCACGGTATTG

A

CC E

PT

ED

M

Brenneria sp. EniD312 LS

E404A

Table 2. Library information about the enzyme activity and thermostability of the wild-type Brenneria sp. EniD312 LS and its variants.

PT

Km (mM)

4.6 ± 0.2 3.4 ± 0.3 3.7 ± 0.1 3.1 ± 0.2 3.3 ± 0.2 2.9 ± 0.3 4.0 ± 0.4 4.1 ± 0.1 4.2 ± 0.2 4.1 ± 0.3 4.0 ± 0.1 4.2 ± 0.2 4.1 ± 0.2 3.4 ± 0.2 3.7 ± 0.1 54 ± 5.2 16 ± 3.9 19 ± 4.4 6.2 ± 4.8 52 ± 2.3

2.1 ± 0.1 1.9 ± 0.3 1.8 ± 0.2 2.0 ± 0.4 1.9 ± 0.2 2.0 ± 0.3 2.0 ± 0.1 2.0 ± 0.5 2.0 ± 0.1 1.8 ± 0.2 1.9 ± 0.3 1.9 ± 0.4 2.0 ± 0.3 1.7 ± 0.1 1.8 ± 0.1 4.8 ± 0.4 3.3 ± 0.3 4.6 ± 0.6 5.3 ± 0.4 6.1 ± 0.8

305 ± 13 346 ± 21 355 ± 17 293 ± 24 286 ± 19 249 ± 18 377 ± 24 338 ± 41 296 ± 19 287 ± 23 118 ± 15 283 ± 23 273 ± 31 370 ± 28 296 ± 28 296 ± 21 318 ± 33 284 ± 41 309 ± 46 294 ± 39

A

N

U

SC R

42 ± 2.2 35 ± 2.6 35 ± 3.1 40 ± 3.1 35 ± 2.7 35± 1.9 34 ± 3.3 36 ± 2.4 38 ± 2.1 30 ± 3.7 33 ± 3.5 35 ± 2.8 34 ± 1.6 30 ± 1.6 33 ± 1.6 87 ± 6.6 73 ± 5.1 88 ± 5.6 63 ± 5.1 104 ± 8.6

IP T

t1/2 (45 °C) t1/2 (55 °C) (h) (min)

M

Wild-Type LS E404D E404A E404G E404H E404P E404R E404T E404S E404Y E404K E404Q E404M E404C E404N E404L E404I E404F E404V E404W

t1/2 (35 °C) (h)

ED

Enzyme

Total activity a (U/mg) 570 ± 10 517 ± 6 360 ± 6 330 ± 10 360 ± 7 320 ± 5 420 ± 8 420 ± 11 390 ± 10 410 ± 5 444 ± 7 370 ± 4 330 ± 9 300 ± 8 370 ± 8 450 ± 7 433 ± 8 359 ± 11 381 ± 9 413 ± 10

A

CC E

Table 3. Comparison of the Tm value for wild-type LS and its variants Brenneria sp. EniD312 LS Tm (°C) ΔTm a. Wild-type 50.0 E404V 51.4 +1.4 E404L 52.8 +2.8 E404F 51.5 +1.5 E404W 51.6 +1.6 E404I 51.5 +1.5 a. The ΔTm is the difference of the Tm between wild-type Brenneria sp. EniD312 LS and its variants.