Improving Thermostability and Catalytic Behavior of l-Rhamnose

Article pubs.acs.org/JAFC

Cite This: J. Agric. Food Chem. 2018, 66, 12017−12024

Improving Thermostability and Catalytic Behavior of L‑Rhamnose Isomerase from Caldicellulosiruptor obsidiansis OB47 toward D‑Allulose by Site-Directed Mutagenesis Ziwei Chen,† Jiajun Chen,† Wenli Zhang,† Tao Zhang,† Cuie Guang,† and Wanmeng Mu*,†,§ †

State Key Laboratory of Food Science and Technology, Jiangnan University, Wuxi, Jiangsu 214122, China International Joint Laboratory on Food Safety, Jiangnan University, Wuxi, Jiangsu 214122, China

§

Downloaded via AUSTRALIAN NATL UNIV on December 17, 2018 at 05:08:44 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: D-Allose, a rare sugar, is an ideal table-sugar substitute and has many advantageous physiological functions. LRhamnose isomerase (L-RI) is an important D-allose-producing enzyme, but it exhibits comparatively low catalytic activity on Dallulose. In this study, an array of hydrophobic residues located within β1−α1−loop were solely or collectively replaced with polar amino acids by site-directed mutagenesis. A group of mutants was designed to weaken the hydrophobic environment and strengthen the catalytic behavior on D-allulose. Compared with that of the wild-type enzyme, the relative activities of the V48N/ G59N/I63N and V48N/G59N/I63N/F335S mutants toward D-allulose were increased by 105.6 and 134.1%, respectively. Another group of mutants was designed to enhance thermostability. Finally, the t1/2 values of mutant S81A were increased by 7.7 and 1.1 h at 70 and 80 °C, respectively. These results revealed that site-directed mutagenesis is efficient for improving thermostability and catalytic behavior toward D-allulose. KEYWORDS: D-allose, L-rhamnose isomerase, catalytic behavior, thermostability, site-directed mutagenesis

■

Thermobacillus composti KWC4 L-RI,15 Bacillus subtilis WB600 L-RI,16 Dictyoglomus turgidum DSMZ 6724 L-RI,17 Bacillus pallidus Y25 L-RI,18 Thermotoga maritima ATCC 43589 L-RI,19 Caldicellulosiruptor saccharolyticus ATCC 43494 20 L-RI, Mesorhizobium loti L-RI,21 Thermoanaerobacterium saccharolyticum NTOU1 L-RI,22 and Pseudomonas stutzeri LRI.13 However, these investigations primarily focused on screening strains and property characterizations of L-RIs. These characterized L-RIs show ultralow catalytic activity on Dallulose, which limits the application of L-RIs in the industrial production of D-allose. Furthermore, few studies are available in regards to enhancing the thermostability and catalytic behavior by molecular modification, which are two important factors in the enzymatic production of D-allose. To date, the crystal structures and catalytic mechanisms of Escherichia coli LRI,23 Bacillus halodurans ATCC BAA-125 L-RI,24 and P. stutzeri L-RI have been resolved.25 In 2010, the effect of the Cterminal region and residue Ser329 of P. stutzeri L-RI, corresponding to Phe336 in E. coli L-RI, on substrate specificity was elaborated by Yoshida et al. with site-directed mutagenesis.26 To date, only this paper has focused on the sitedirected mutagenesis of L-RI. However, the variation of the specific activity of P. stutzeri L-RI on D-allulose and D-allose has not been further explored. Previously, wild-type L-RI from Caldicellulosiruptor obsidiansis OB47 was characterized in our laboratory.27 Although C. obsidiansis OB47 L-RI exhibits the highest catalytic activity

INTRODUCTION In the natural environment, only seven monosaccharides are common sugars that are abundant in nature, such as D-glucose and D-fructose. The majority of monosaccharides are rare sugars, which are defined by the International Society of Rare Sugars (ISRS) as monosaccharides and monosaccharides derivatives that barely appear in nature.1 D-Allose, an extensively studied rare sugar, has 80% relative sweetness to sucrose but is noncaloric and nontoxic.2,3 Therefore, D-allose is an ideal table-sugar substitute and food additive and is beneficial to weight loss. Also, D-allose displays many salutary physiological functions, such as antitumor, anticancer,4 cryoprotective,5 neuroprotective,6 antiosteoporotic,7 anti-inflammatory,8 antihypertensive,9 and immunosuppressant functions.10 More physiological functions and health benefits of Dallose have been reviewed in detail.11 D-Allose has huge application potential in the food-system, clinical-treatment, and health-care fields because of its remarkable physiological functions. However, chemical synthesis of D-allose has many disadvantages, including a low conversion rate, chemical pollution, and byproduct generation.12 The enzymatic production of D-allose has been widely investigated in recent years. L-Rhamnose isomerase (L-RI, EC 5.3.1.14), one type of aldose-ketose isomerase, catalyzes the conversion between Lrhamnose and L-rhamnulose. L-RI has a broad substrate spectrum and can also catalyze the isomerization of D-allulose and D-allose.13 L-RI, an important D-allose-producing enzyme, has been extensively studied. To date, from various microorganisms, more than ten L-RIs catalyzing the isomerization reaction between D-allulose and D-allose have been cloned and identified: Clostridium stercorarium ATCC 35414 L-RI,14 © 2018 American Chemical Society

Received: Revised: Accepted: Published: 12017

September 18, 2018 October 26, 2018 October 29, 2018 October 29, 2018 DOI: 10.1021/acs.jafc.8b05107 J. Agric. Food Chem. 2018, 66, 12017−12024

Article

Journal of Agricultural and Food Chemistry

8000g for 10 min. Then, the collected cells were washed twice using distilled water and stored at −20 °C. The purification procedures for the wild-type enzyme and mutants were accomplished in a cold room. The pelleted cells were suspended in cell-lysis buffer (pH 8.0) and disrupted by sonication for 16 min (on 1 s, off 2 s) with a Scientz-II D ultrasonic homogenizer (Scientz Biotechnology). The cellular lysates were centrifuged at 10 000g for 15 min to remove the cell fragments, and then the supernatant was collected and filtered through a 0.45 μm water-phase filter. Fast protein liquid chromatography (FPLC, Ä KTA Purifier System, GE Healthcare) was used for purification of the wild-type enzyme and mutants. The column was washed using five column volumes (CV) of ultrapure water at a 1 mL/min flow rate. The filtrate was loaded on a Ni2+-chelating Sepharose Fast Flow resin column (8.9 × 64 mm, GE Healthcare). The column was pre-equilibrated with 12 CV of fresh binding buffer (50 mM Tris-HCl and 500 mM NaCl, pH 8.0) at a 0.6 mL/min flow rate. Afterward, the unbound and unwanted enzymes were eliminated from the resin column using 6 CV of washing buffer (50 mM Tris-HCl, 500 mM NaCl, and 50 mM imidazole; pH 8.0) at a 1 mL/min flow rate. Lastly, the target proteins were eluted using 6 CV of elution buffer (50 mM Tris-HCl, 500 mM NaCl, and 500 mM imidazole; pH 8.0) at 1 mL/min. The fractions displaying catalytic activity were pooled and dialyzed against 50 mM Tris-HCl buffer (pH 8.0) containing EDTA for 12 h to remove metal ions. Subsequently, the protein was dialyzed against Tris-HCl buffer (pH 8.0) to remove EDTA. The protein concentrations were measured by the Lowry method using bovine serum albumin as the reference.37 The purities and molecular weights of wild-type enzyme and mutants were checked using 12% sodium dodecyl sulfate−polyacrylamide-gel electrophoresis (SDS-PAGE) with Coomassie brilliant blue R250 staining. Isomerization Activity. The catalytic activity of the wild-type enzyme and mutants toward D-allulose was determined by assaying the formation of D-allose. Under optimal reaction conditions, the enzymatic reactions were implemented at 85 °C with 50 mM N-(2hydroxyethyl)piperazine-N-3-propanesulfonic acid (HEPPS) buffer (pH 8.0) containing 40 mM D-allulose, 1 mM Co2+, and 0.05 μM purified enzyme. After 10 min, the reaction mixture was boiled for 15 min to stop the enzyme reaction. The catalytic activity of the wildtype enzyme and mutants toward L-rhamnose was investigated by assaying the accumulated amount of L-rhamnulose. The reaction conditions of catalytic activity acting on L-rhamnose are the same as those acting on D-allulose. The concentration of D-allose was determined by high-performance liquid chromatography (HPLC) with a refractive index (IR) detector (2414, Waters) and a Ca2+-ligand-exchange column (6.5 × 300 mm, Sugar-Pak 1, Waters Corporation). The column was eluted with ultrapure water at a column temperature of 85 °C with a flow rate of 0.4 mL/min. The concentration of L-rhamnulose was determined using a modified cysteine−sulfuric acid−carbazol method.38 First, 500 μL reaction mixture was diluted and added to 100 μL of 1.5% cysteine hydrochloride solution. After blending, 3 mL of 75% sulfuric acid and 100 μL of ethanol-carbazole were added in turn. Then, the generated mixture was immediately incubated at 60 °C for 10 min, and the absorbance was promptly measured at 540 nm. One unit of enzyme activity was defined as the amount of enzyme catalyzing the generation of 1 μmol of monosaccharide per minute at 85 °C and pH 8.0. Mutation for Thermostability. To investigate the thermostability of the mutants, the half-lives (t1/2) and melting temperatures (Tm) were measured. For t1/2 determination, the wild-type enzyme and mutants were preincubated at 70 and 80 °C. The samples were discontinuously withdrawn at specific times, and the residual activity was later assayed at pH 8.0 and 85 °C. The initial activity without incubation was set as 100%. A differential-scanning calorimeter (Nano DSC III, TA Instruments) equipped with a Platinum Capillary Cell was used for Tm value determination. After vacuum degasification (635 mmHg), the dialyzed buffers and proteins were loaded into reference and sample cells, respectively. The scanning was carried out at 3 atm of air

(19.4 U/mg) on D-allulose compared with other reported LRIs, such as T. composti KWC4 L-RI (1.7 U/mg) and T. maritima ATCC 43589 L-RI (1.1 U/mg),19,28 it still cannot remotely meet the needs of industrial production of D-allose. In this work, we built a model of C. obsidiansis OB47 L-RI on the basis of the B. halodurans ATCC BAA-125 L-RI structure. Then, we rationally designed site-directed mutagenesis on the grounds of reported structural information on L-RIs to further improve thermostability and catalytic activity of C. obsidiansis OB47 L-RI on D-allulose, which is conducive to the industrial production of D-allose.

■

MATERIALS AND METHODS

Strains, Reagents, and Chemicals. E. coli DH5α and BL21 (DE3) strains were purchased from Sangon Biotech Company, Ltd. The plasmid harboring the C. obsidiansis OB47 L-RI gene was constructed in our previous work.27 The reagents used for sitedirected mutagenesis of wild-type L-RI gene were obtained from Generay Biotech Company, Ltd. Isopropyl-β-D-1-thiogalactopyranoside (IPTG) for induction was from Sigma. The Ni2+-chelating affinity-chromatography resin was provided by GE. Electrophoretic reagents were obtained from Bio-Rad. Other chemicals were from Sinopharm Chemical Reagent or Sigma. Molecular Modeling and Docking. Homology modeling of the three-dimensional structure of the wild-type enzyme and the mutants was conducted by the SWISS-MODEL online server (http://www. expasy.ch/swissmod/SWISS-MODEL.html) using the crystal structure of B. halodurans ATCC BAA-125 L-RI (PDB ID: 3P14) as a template (55% sequence identity).29−31 The model-energy minimization was performed using the Discovery Studio package. The accuracies of the wild-type enzyme and mutant models were examined by the SAVES server.32,33 The stereochemical quality was verified by Procheck with its Ramachandran-plot module.34 The obtained models were delineated and presented with the Pymol Molecular Graphics Software. The L-rhamnose and D-allulose models were constructed by the GlycoBioChem PRODRG2 online server (http://davapc1.bioch. dundee.ac.uk/cgi-bin/prodrg/submit.html). The ligand-energy minimization was implemented by Chem3D Pro14.0.35 L-Rhamnose and D-allulose were used as ligands. Correspondingly, the wild-type enzyme and mutant models were used as acceptors for docking. The docking procedure was executed by the Autodock 4.2 software package.36 The obtained models were further handled by a battery of programs in the AutoDock Tool, such as programs that added hydrogen atoms, calculated charge, and removed water molecules. Site-Directed Mutagenesis. Site-directed mutagenesis of C. obsidiansis OB47 L-RI was manipulated by one-step PCR methods using a TaKaRa MutantBEST Kit (TaKaRa). The mutants were divided into two groups: one group included S81A, S81Q, S88R, V421I, and I343A for enhancing the thermostability, and the other group included V48N, G59N, G60T, G62T, I63N, I101N, F335C, F335S, V48N/G59N/I63N, and V48N/G59N/I63N/F335S for improving the catalytic activity toward D-allulose. The recombinant plasmid containing the wild-type C. obsidiansis OB47 L-RI gene was used as the template. All primers used for mutagenesis are shown in Table S1. After PCR amplification, the obtained products were digested and purified by DpnI. The gene sequences of the various mutants were verified by Sangon Biotech Company, Ltd. Heterologous Expression and Purification. The mutant plasmids were introduced into E. coli DH5α and BL21 (DE3) for gene cloning and enzyme overexpression, respectively. The recombinant BL21 strains were cultivated in Luria−Bertani medium (LB, 10 g/L tryptone, 10 g/L NaCl, and 5 g/L yeast extract) with ampicillin (100 μg/mL) at 37 °C and 200 rpm. The IPTG was added to the LB medium at a final concentration of 1 mM until the OD600 was 0.5− 0.7. The recombinant cells were induced for 6 h at 28 °C. After that, the induced recombinant cells were harvested by centrifugation at 12018

DOI: 10.1021/acs.jafc.8b05107 J. Agric. Food Chem. 2018, 66, 12017−12024

Article


Figure 1. C. obsidiansis OB47 L-RI structural model and alignment with B. halodurans ATCC BAA-125 L-RI (PDB ID: 3P14). (A) Dimer model of C. obsidiansis OB47 L-RI. The α-helix, β-strand, and random coil are colored cyan, magenta, and salmon, respectively. (B) Monomer superimposition of C. obsidiansis OB47 L-RI (cyan) and B. halodurans ATCC BAA-125 L-RI (warm pink). pressure from 25 to 100 °C with a heating rate of 1 °C/min after preequilibration for 10 min. The DSC data of the wild-type enzyme and mutants were analyzed using the TA Instruments Nano Analyze software. The two-state scaled model was selected in the fitting process after baseline-corrected fitting. Production of D-Allose from D-Allulose. The production of Dallose from D-allulose was implemented in 5 mL reaction mixtures containing 50 mM HEPPS buffer (pH 8.0), 1 mM Co2+, and 5 μM purified enzyme. The initial substrate concentration was 25 g/L Dallulose. Considering the thermostability and productivity, the conversation temperature was set at 60 °C. The reaction mixture was inactivated and detected by HPLC at the given times to determine the concentration of D-allose. All experiments were implemented in triplicate.

favored regions, the additionally allowed regions, and the generously allowed regions, respectively. Moreover, amino acid residues were scarcely located in disallowed regions. The monomer structures of C. obsidiansis OB47 L-RI (cyan) and B. halodurans ATCC BAA-125 L-RI (warm pink) were superimposed (Figure 1B) with a 0.120 root-mean-square-deviation (RMSD) value. The result of superimposition indicated that the structures of C. obsidiansis OB47 L-RI and template were very similar. All of these results revealed that the acquired 3D models were applicable and could be used for further structural analysis. The structural arrangement of C. obsidiansis OB47 L-RI is the (β/α)8-barrel conformation (Figure 1), which is composed of alternating connections between eight α-helices (α1−α8) and eight β-strands (β1−β8). Moreover, the C. obsidiansis OB47 L-RI structure has additional α-helical domains (α0, α9, α10, α11, and α12) and an extended flexible loop, which are similar to other resolved L-RI structures and may contribute to the association between subunits and the combined action with active sites in the catalytic center, respectively.23,25 It is thought that the (β/α)8-barrel conformation of L-RIs is the most widespread and stable fold in characterized and resolved enzymes. Having intrinsic stability, this (β/α)8-barrel structure can serve as the core scaffold for molecular modification aiming at thermostability and catalytic activity. Notably, the flexible loop above the catalytic core, which determines substrate specificity, is the most attractive candidate for improving catalytic activity on D-allulose by site-directed mutagenesis, according to the reported E. coli L-RI structural information. Effect of Mutation on Catalytic Behavior. In 2000, Korndörfer et al. reported that in the E. coli L-RI structure, β1−α1−loop, which is a flexible loop domain consisting of a series of hydrophobic residues (Asp52−Arg78), is probably in charge of the recognition of substrates. This β1−α1−loop of E. coli L-RI is similar to a lid or switch partly covering the catalytic pocket to control the entry of L-rhamnose. Furthermore, this β1−α1−loop, together with several nonconserved hydrophobic residues (Ile105, Tyr106, and Phe336), creates a hydrophobic region encompassing the substrate of the C6methyl group. It revealed that E. coli L-RI prefers L-rhamnose with a C6-methyl group over substrates with a C6-oxhydryl group, such as D-allose and D-allulose. Particularly, in E. coli LRI, V53, I67, and I105 (corresponding to V48, I63, and I101 in

■

RESULTS AND DISCUSSION Expression and Purification. After overexpression, the wild-type enzyme and mutants were purified. As shown in Figure S1, an approximately 48 kDa band was visible in the SDS-PAGE gel, which was in agreement with the theoretical molecular weight. This result manifested that the folding of the mutant proteins was correct and that expression and purification were not affected by mutagenesis. Structural Modeling. Homology modeling is the most effective method to predict unresolved protein structures. Homology modeling is based on two principles: the first is that the protein three-dimensional structure is exclusively determined by the amino acid sequence and can be theoretically inferred from the primary sequence; and the second point is that a protein’s three-dimensional structure is highly conserved over the course of protein evolution. The B. halodurans ATCC BAA-125 L-RI shared 55% sequence identity with C. obsidiansis OB47 L-RI and thus was chosen as the template for homology modeling. The models of wild-type (as shown in Figure 1A) and variant C. obsidiansis OB47 L-RI were constructed on the basis of the crystal structure of B. halodurans ATCC BAA-125 L-RI (PDB ID: 3P14) using the SWISS-MODEL server. After the energy minimization, the quality of the obtained model was evaluated using the VERIFY-3D procedure from the SAVES server. The results of VERIFY-3D revealed that 94.98% of the amino acid residues possessed an average 3D−1D score ≥0.2 in 3D−1D structural compatibility, which was considerably greater than the minimal quality requirement (80%). The Ramachandran plot (Figure S2) exhibited that 90.3, 9.0, and 0.7% of the amino acid residues were located in the most 12019


Article


Figure 2. (A) Planform and (B) lateral view of the surface model of the wild-type enzyme. Surface and cartoon models are depicted in cyan and green, respectively. β1−α1−loop (maroon) embraces the catalytic pocket (yellow line) and partly covers the catalytic tunnel (green). (C,D) Superimposition of the surface and cartoon models by (C) part and (D) whole transparency.

C. obsidiansis OB47 L-RI) have a hydrophobic stacking interaction and a significant effect on the recognition of substrate.23 In 2007, a similar β1−α1−loop (Gly60−Arg76) was found in the P. stutzeri L-RI, which exhibits broad substrate specificity. However, the difference is that β1−α1−loop of P. stutzeri L-RI covers the adjacent subunit molecule in connection with substrate binding. In addition, β1−α1−loop of P. stutzeri L-RI only forms a hydrophobic interaction with the substrate instead of a hydrophobic pocket, which results in slight recognition for the C6 position. This can explain why P. stutzeri L-RI has a broader substrate specificity than E. coli LRI.25 In E. coli L-RI, Phe336 (corresponding to Phe335 in C. obsidiansis OB47 L-RI), in the vicinity of conserved residues, has a significant impact on the substrate specificity.23 However, this site is a hydrophilic serine in D. turgidum DSMZ 6724 LRI, M. loti L-RI, and P. stutzeri L-RI (S329) and a hydrophilic cysteine in Caldilinea aerophila L-RI (a hypothetical L-RI in GenBank, NCBI number: WP_014435274.1; Figure S3). Furthermore, to investigate the effect of S329 in P. stutzeri LRI on substrate specificity, Yoshida et al. designed four

mutants: S329F, S329 K, S329L and S329A. The results showed that the kcat/Km of S329F acting on D-allose was distinctly lower. To summarize, this site together with β1−α1− loop creates the hydrophobic catalytic environment that possibly has an enormous effect on recognition of substrate according to the L-RI structural information that has been verified. The possible location of β1−α1−loop (Asp47−Agr73) in the structure model of C. obsidiansis OB47 L-RI was determined by sequence alignment and structural analysis. As shown in Figure 2A, the surface model of C. obsidiansis OB47 L-RI β1−α1−loop (maroon) above the central tunnel (green) is also similar to a lid and may be involved in the recognition of substrate. A hydrophobic cavity between β1−α1−loop and the catalytic pocket (yellow) can be clearly observed from the lateral view (Figure 2B). The surface models superimposed with the cartoon model are presented in Figure 2C,D. To improve the catalytic activity of C. obsidiansis OB47 L-RI on Dallulose (C6-oxhydryl group), to make it appropriate for the industrial production of D-allose, a group of mutants was designed by weakening the hydrophobic environment created 12020


Article

Journal of Agricultural and Food Chemistry by the β1−α1−loop. In C. obsidiansis OB47 L-RI, a group of hydrophobic residues comprising four continuous glycines (G59, G60, G61, and G62) may exert a strong hydrophobic interaction. Therefore, five hydrophobic residues (V48, G59, G60, G62, and I63) located within β1−α1−loop and two subsidiary residues, I101 and F335 (corresponding to Ile105 and F336, respectively, in E. coli L-RI), were selected as the mutation sites and replaced with hydrophilic residues. Hence, eight single-point mutants (V48N, G59N, G60T, G62T, I63N, I101N, F335C, and F335S) and two multiple mutants (V48N/ G59N/I63N and V48N/G59N/I63N/F335S) were designed to test catalytic behavior. The catalytic activities of wild-type enzyme and mutants toward L-rhamnose and D-allulose were determined at optimal reaction conditions as previously described,27 and the activity of the wild-type enzyme was set as 100%. As shown in Table 1,

31.7%, respectively. Moreover, the relative activities of all mutants acting on L-rhamnose were visibly decreased. In particular, the relative activities of V48N/G59N/I63N and V48N/G59N/I63N/F335S were increased by 105.6 and 134.1% on D-allulose and decreased by 38.5 and 39.4% on Lrhamnose, respectively. This may suggest that multiple mutation sites exert a synergistic effect. This result is largely consistent with the design principle of mutation. The cartoon models of the wild-type enzyme and V48N/G59N/I63N/ F335S mutant are shown in Figure 3A,B, respectively. Furthermore, to elucidate vividly the variation of the catalytic pocket, their surface models are presented in Figure 4. After the residues of V48 and F335 were substituted with N48 and S335, respectively, their positions were closer to the substrate, and the catalytic pocket had partly shrunk. The shrinking of the catalytic pocket enhanced the hydrophilic environment and the interaction with the substrate of the C6-oxhydryl group. After G59 was replaced with N59, the position of the residue shifted to the inside from the edge of the catalytic pocket. However, when the I63 was replaced with N63, the side chain of the N63 residue diverged the central tunnel. This finding could explain why the relative activity of I63N toward Dallulose was not increased. Effect of Mutation on Thermostability. To enhance the thermostability of C. obsidiansis OB47 L-RI, the PDB file of the model was uploaded to the Hotspot Wizard 3.0 online server (https://loschmidt.chemi.muni.cz/hotspotwizard/), which can automatically establish mutation libraries and design sitespecific mutations for protein stability according to the amino acid frequency and evolutionary information from three large databases.39 The server recommended many sites that may alter the thermostability of L-RI. By the analysis of the C. obsidiansis OB47 L-RI structural model, it was observed that two residues, S81 and S88, located in the α1 region may generate interplay with V421 and I343 located in the α1 and α8 regions, respectively. Thus, five mutants, S81A, S81Q, S88R, V421I, and I343A, were designed for further studies.

Table 1. Relative Activities of the Wild-Type Enzyme and Mutants towards L-Rhamnose and D-Allulose relative activity (%) enzyme wild type V48N G59N G60T G62T I63N I101N F335C F335S V48N/G59N/I63N V48N/G59N/I63N/F335S

L-rhamnose

100.0 49.9 86.0 86.5 70.4 40.8 87.4 69.6 59.1 38.5 39.4

± ± ± ± ± ± ± ± ± ± ±

2.2 0.8 0.4 0.6 0.6 0.8 1.1 1.6 1.4 0.7 0.5

D-allulose

100.0 168.6 161.4 87.1 136.1 94.5 136.8 187.4 131.7 205.6 234.1

± ± ± ± ± ± ± ± ± ± ±

1.8 1.6 1.2 0.6 1.4 1.1 1.2 2.0 2.1 2.4 2.2

compared with the wild-type enzyme, the relative activities of V48N, G59N, G62T, I101N, F335C, and F335S acting on Dallulose were increased by 68.6, 61.4, 36.1, 36.8, 87.4, and

Figure 3. Residue distributions of the 48, 59, 63, and 335 positions of (A) wild-type enzyme and (B) V48N/G59N/I63N/F335S mutant. These residues are presented as stick models. 12021


Article


Figure 4. Surface models of (A) wild-type enzyme and (B) V48N/G59N/I63N/F335S mutant. The residues of the 48, 59, 63, and 335 positions are colored magenta, orange, blue, and green, respectively.

The t1/2 values of the wild-type enzyme and mutants were determined at 70 and 80 °C. As shown in Table 2, in contrast

to those of the wild-type enzyme, the t1/2 values of S88R, V421I, and I343A were obviously lower at 70 and 80 °C. Interestingly, the t1/2 values of S81A were enhanced to 34.1 and 5.6 h, but the t1/2 values of S81Q were dramatically reduced at 70 and 80 °C. The structural stability of S81A was further investigated by Nano-DSC (Figure S4). Compared with that of the wild-type enzyme, the Tm value of S81A was increased by approximately 3 °C. As illustrated in Figure 5A, the valine of the 421 position contains two methyl groups that are closest to the S81 residue, which contains a hydroxyl located in the α-helix of the C-terminus. Thus, when serine-81 was replaced with a hydrophobic residue of alanine containing a methyl group, a hydrophobic interaction was formed between alanine-81 and valine-421, which contributed to strengthening the locking force of the overall structure and thereby enhancing the structural thermostability (Figure 5B).

Table 2. Thermostability of C. obsidiansis OB47 L-RI Mutants half-life (t1/2, h) enzyme

70 °C

80 °C

wild type S81A S81Q S88R I343A V421I

26.4 34.1 6.5 5.7 4.7 13.6

4.5 5.6 1.8 2.0 1.9 4.6

Figure 5. Location of S81 and V421 in a cartoon model of (A) wild-type enzyme and (B) V48N/G59N/I63N/F335S mutant. The hydrophobic interaction is represented using red dotted lines. 12022



■

Bioconversion of D-Allulose to D-Allose. The production of D-allose was investigated in 5 mL reaction mixtures containing 25 g/L D-allulose using the wild-type enzyme and the mutant V48N/G59N/I63N/F335S. As shown in Figure 6,

Article

AUTHOR INFORMATION

Corresponding Author

*Tel.: (86) 510-85919161. Fax: (86) 510-85919161. E-mail: [email protected]. ORCID

Tao Zhang: 0000-0003-3194-2028 Wanmeng Mu: 0000-0001-6597-527X Funding

This work was supported by the Support Project of Jiangsu Province (No. 2015-SWYY-009); the Research Program of the State Key Laboratory of Food Science and Technology, Jiangnan University (Nos. SKLF-ZZA-201802 and SKLF-ZZB201814); and the National First-Class Discipline Program of Food Science and Technology (No. JUFSTR20180203). Notes

The authors declare no competing financial interest.

■

(1) Izumori, K. Bioproduction strategies for rare hexose sugars. Naturwissenschaften 2002, 89 (3), 120−124. (2) Iga, Y.; Nakamichi, K.; Shirai, Y.; Matsuo, T. Acute and subchronic toxicity of D-allose in rats. Biosci., Biotechnol., Biochem. 2010, 74 (7), 1476−1478. (3) Mooradian, A. D.; Smith, M.; Tokuda, M. The role of artificial and natural sweeteners in reducing the consumption of table sugar: A narrative review. Clin. Nutr. Espen. 2017, 18, 1−8. (4) Noguchi, C.; Kamitori, K.; Hossain, A.; Hoshikawa, H.; Katagi, A.; Dong, Y.; Sui, L.; Tokuda, M.; Yamaguchi, F. D-Allose inhibits cancer cell growth by reducing GLUT1 expression. Tohoku J. Exp. Med. 2016, 238 (2), 131. (5) Sui, L.; Nomura, R.; Dong, Y.; Yamaguchi, F.; Izumori, K.; Tokuda, M. Cryoprotective effects of D-allose on mammalian cells. Cryobiology 2007, 55 (2), 87. (6) Gao, D.; Kawai, N.; Nakamura, T.; Lu, F.; Fei, Z.; Tamiya, T. Anti-inflammatory effect of D-allose in cerebral ischemia/reperfusion injury in rats. Neurol. Med-Chir. 2013, 53 (6), 365−374. (7) Yamada, K.; Noguchi, C.; Kamitori, K.; Dong, Y.; Hirata, Y.; Hossain, M. A.; Tsukamoto, I.; Tokuda, M.; Yamaguchi, F. Rare sugar D-allose strongly induces thioredoxin-interacting protein and inhibits osteoclast differentiation in Raw264 cells. Nutr. Res. (N. Y., NY, U. S.) 2012, 32 (2), 116−123. (8) Shinohara, N.; Nakamura, T.; Abe, Y.; Hifumi, T.; Kawakita, K.; Shinomiya, A.; Tamiya, T.; Tokuda, M.; Keep, R. F.; Yamamoto, T.; Kuroda, Y. D-Allose attenuates overexpression of inflammatory cytokines after cerebral ischemia/reperfusion injury in gerbil. J. Stroke. Cerebrovasc. 2016, 25 (9), 2184−2188. (9) Kimura, S.; Zhang, G. X.; Nishiyama, A.; Nagai, Y.; Nakagawa, T.; Miyanaka, H.; Fujisawa, Y.; Miyatake, A.; Nagai, T.; Tokuda, M.; Abe, Y. D-Allose, an all-cis aldo-hexose, suppresses development of salt-induced hypertension in Dahl rats. J. Hypertens. 2005, 23 (10), 1887−1894. (10) Hossain, M. A.; Wakabayashi, H.; Goda, F.; Kobayashi, S.; Maeba, T.; Maeta, H. Effect of the immunosuppressants FK506 and D-allose on allogenic orthotopic liver transplantation in rats. Transplant. Proc. 2000, 32 (7), 2021−2023. (11) Chen, Z.; Chen, J.; Zhang, W.; Zhang, T.; Guang, C.; Mu, W. Recent research on the physiological functions, applications, and biotechnological production of D-allose. Appl. Microbiol. Biotechnol. 2018, 102, 4269−4278. (12) Herber, R. R.; Maher, G. F.; Arnold, E. C.; Lorsbach, T. W. Preparation of high purity D-allose from D-glucose. U.S. Patent 5433793, 1995. (13) Leang, K.; Takada, G.; Fukai, Y.; Morimoto, K.; Granström, T. B.; Izumori, K. Novel reactions of L-rhamnose isomerase from Pseudomonas stutzeri and its relation with D-xylose isomerase via

Figure 6. Production of D-allose using the V48N/G59N/I63N/F335S mutant and wild-type enzyme. The conversion reactions were carried out at 60 °C and pH 8.0 with 1 mM Co2+, 5 μM purified enzyme, and 25 g/L D-allulose as substrate. The experiments were conducted in three replicates, and data points are the means ± standard deviations.

the isomerization reactions of the mutant and wild-type enzymes approached equilibrium at 16 and 24 h, respectively, with an approximately 32% conversion ratio. It was observed that the mutant V48N/G59N/I63N/F335S exhibits higher catalytic efficiency than the wild-type enzyme under the same reaction conditions. Moreover, D-altrose, a potential byproduct, has not been detected by HPLC analysis in reaction mixtures of the wild-type enzyme or in reaction mixtures of the mutant (data not shown), which simplifies the separation and purification and is better for the industrial production of Dallose. Compared with the wild-type enzyme, the mutant V48N/G59N/I63N/F335S has a better catalytic behavior in the industrial production of D-allose. D-Ribose-5-phosphate isomerase from Thermotoga lettingae TMO converts D-allulose to D-allose with a rate of 32%, but it exhibits a low productivity. D-Galactose-6-phosphate isomerase from Lactococcus lactis and glucose-6-phosphate isomerase from Pyrococcus f uriosus produce D-allose with 25 and 32% conversion rates but with detectable byproducts.11 Compared with those D-alloseproducing enzymes, the C. obsidiansis OB47 L-RI displays a larger application potential.

■

REFERENCES

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jafc.8b05107. SDS-PAGE analysis of mutants, Ramachandran plot of the C. obsidiansis OB47 L-RI model, multiple-sequence alignment of various L-RIs, Nano-DSC analysis of the wild-type enzyme and V48N/G59N/I63N/F335S mutant, and primers for site-directed mutagenesis (PDF) 12023


Article

Journal of Agricultural and Food Chemistry substrate specificity. Biochim. Biophys. Acta, Gen. Subj. 2004, 1674 (1), 68−77. (14) Seo, M. J.; Choi, J. H.; Kang, S. H.; Shin, K. C.; Oh, D. K. Characterization of L-rhamnose isomerase from Clostridium stercorarium and its application to the production of D-allose from D-allulose (D-psicose). Biotechnol. Lett. 2018, 40, 325−334. (15) Xu, W.; Zhang, W.; Tian, Y.; Zhang, T.; Jiang, B.; Mu, W. Characterization of a novel thermostable L-rhamnose isomerase from Thermobacillus composti KWC4 and its application for production of D-allose. Process Biochem. 2017, 53, 153−161. (16) Bai, W.; Shen, J.; Zhu, Y.; Men, Y.; Sun, Y.; Ma, Y. Characteristics and kinetic properties of L-rhamnose isomerase from Bacillus subtilis by isothermal titration calorimetry for the production of D-allose. Food Sci. Technol. Res. 2015, 21 (1), 13−22. (17) Kim, Y. S.; Shin, K. C.; Lim, Y. R.; Oh, D. K. Characterization of a recombinant L-rhamnose isomerase from Dictyoglomus turgidum and its application for L-rhamnulose production. Biotechnol. Lett. 2013, 35 (2), 259−264. (18) Poonperm, W.; Takata, G.; Okada, H.; Morimoto, K.; Granstrom, T. B.; Izumori, K. Cloning, sequencing, overexpression and characterization of L-rhamnose isomerase from Bacillus pallidus Y25 for rare sugar production. Appl. Microbiol. Biotechnol. 2007, 76 (6), 1297−1307. (19) Park, C. S.; Yeom, S. J.; Lim, Y. R.; Kim, Y. S.; Oh, D. K. Characterization of a recombinant thermostable L-rhamnose isomerase from Thermotoga maritima ATCC 43589 and its application in the production of L-lyxose and L-mannose. Biotechnol. Lett. 2010, 32 (12), 1947−1953. (20) Lin, C. J.; Tseng, W. C.; Fang, T. Y. Characterization of a thermophilic L-rhamnose isomerase from Caldicellulosiruptor saccharolyticus ATCC 43494. J. Agric. Food Chem. 2011, 59 (16), 8702− 8708. (21) Takata, G.; Uechi, K.; Taniguchi, E.; Kanbara, Y.; Yoshihara, A.; Morimoto, K.; Izumori, K. Characterization of Mesorhizobium loti Lrhamnose isomerase and its application to L-talose production. Biosci., Biotechnol., Biochem. 2011, 75 (5), 1006−1009. (22) Lin, C. J.; Tseng, W. C.; Lin, T. H.; Liu, S. M.; Tzou, W. S.; Fang, T. Y. Characterization of a thermophilic L-rhamnose isomerase from Thermoanaerobacterium saccharolyticum NTOU1. J. Agric. Food Chem. 2010, 58 (19), 10431−10436. (23) Korndörfer, I. P.; Fessner, W. D.; Matthews, B. W. The structure of rhamnose isomerase from Escherichia coli and its relation with xylose isomerase illustrates a change between inter and intrasubunit complementation during evolution. J. Mol. Biol. 2000, 300 (4), 917−933. (24) Doan, T. N.; Prabhu, P.; Kim, J. K.; Ahn, Y. J.; Natarajan, S.; Kang, L. W.; Park, G. T.; Lim, S. B.; Lee, J. K. Crystallization and preliminary X-ray crystallographic analysis of L-rhamnose isomerase with a novel high thermostability from Bacillus halodurans. Acta Crystallogr., Sect. F: Struct. Biol. Cryst. Commun. 2010, 66, 677−680. (25) Yoshida, H.; Yamada, M.; Ohyama, Y.; Takada, G.; Izumori, K.; Kamitori, S. The structures of L -rhamnose isomerase from Pseudomonas stutzeri in complexes with L-rhamnose and D-allose provide insights into broad substrate specificity. J. Mol. Biol. 2007, 365 (5), 1505−1516. (26) Yoshida, H.; Takeda, K.; Izumori, K.; Kamitori, S. Elucidation of the role of Ser329 and the C-terminal region in the catalytic activity of Pseudomonas stutzeri L-rhamnose isomerase. Protein Eng., Des. Sel. 2010, 23 (12), 919−927. (27) Chen, Z.; Xu, W.; Zhang, W.; Zhang, T.; Jiang, B.; Mu, W. Characterization of a thermostable recombinant L-rhamnose isomerase from Caldicellulosiruptor obsidiansis OB47 and its application for the production of L-fructose and L-rhamnulose. J. Sci. Food Agric. 2018, 98 (6), 2184−2193. (28) Xu, W.; Zhang, W. L.; Zhang, T.; Jiang, B.; Mu, W. M. LRhamnose isomerase and its use for biotechnological production of rare sugars. Appl. Microbiol. Biotechnol. 2016, 100 (7), 2985−2992.

(29) Arnold, K.; Bordoli, L.; Kopp, J.; Schwede, T. The SWISSMODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 2006, 22 (2), 195−201. (30) Kiefer, F.; Arnold, K.; Künzli, M.; Bordoli, L.; Schwede, T. The SWISS-MODEL Repository and associated resources. Nucleic Acids Res. 2009, 37, D387−D392. (31) Biasini, M.; Bienert, S.; Waterhouse, A.; Arnold, K.; Studer, G.; Schmidt, T.; Kiefer, F.; Cassarino, T. G.; Bertoni, M.; Bordoli, L.; et al. SWISS-MODEL: modelling protein tertiary and quaternary structure using evolutionary information. Nucleic Acids Res. 2014, 42, W252−W258. (32) Bowie, J. U.; Luthy, R.; Eisenberg, D. A method to identify protein sequences that fold into a known three-dimensional structure. Science 1991, 253 (5016), 164−170. (33) Lüthy, R.; Bowie, J. U.; Eisenberg, D. Assessment of protein models with three-dimensional profiles. Nature 1992, 356 (6364), 83. (34) Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J. M. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 1993, 26 (2), 283−291. (35) Xu, W.; Ni, D.; Yu, S.; Zhang, T.; Mu, W. Insights into hydrolysis versus transfructosylation: Mutagenesis studies of a novel levansucrase from Brenneria sp. EniD312. Int. J. Biol. Macromol. 2018, 116, 335−345. (36) Laederach, A.; Reilly, P. J. Specific empirical free energy function for automated docking of carbohydrates to proteins. J. Comput. Chem. 2003, 24 (14), 1748−1757. (37) Rosebrough, N. J.; Farr, S. A.; Randall, R. L.; Lowry, O. H. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 1951, 193 (1), 265. (38) Dische, Z.; Borenfreund, E. A new spectrophotometric method for the detection and determination of keto sugars and trioses. J. Biol. Chem. 1951, 192 (2), 583−587. (39) Bendl, J.; Stourac, J.; Sebestova, E.; Vavra, O.; Musil, M.; Brezovsky, J.; Damborsky, J. HotSpot Wizard 2.0: automated design of site-specific mutations and smart libraries in protein engineering. Nucleic Acids Res. 2016, 44, W479−W487.

12024


Improving Thermostability and Catalytic Behavior of l-Rhamnose

Recommend Documents