A novel cold-adapted and highly salt-tolerant esterase from Alkalibacterium sp. SL3 from the sediment of a soda lake

A novel esterase gene (estSL3) was cloned from the Alkalibacterium sp. SL3, which was isolated from the sediment of soda lake Dabusu. The 636-bp full-length gene encodes a polypeptide of 211 amino acid residues that is closely related with putative GDSL family lipases from Alkalibacterium and Enterococcus. The gene was successfully expressed in E. coli, and the recombinant protein (rEstSL3) was purified to electrophoretic homogeneity and characterized. rEstSL3 exhibited the highest activity towards pNP-acetate and had no activity towards pNP-esters with acyl chains longer than C8. The enzyme was highly cold-adapted, showing an apparent temperature optimum of 30 °C and remaining approximately 70% of the activity at 0 °C. It was active and stable over the pH range from 7 to 10, and highly salt-tolerant up to 5 M NaCl. Moreover, rEstSL3 was strongly resistant to most tested metal ions, chemical reagents, detergents and organic solvents. Amino acid composition analysis indicated that EstSL3 had fewer proline residues, hydrogen bonds and salt bridges than mesophilic and thermophilic counterparts, but more acidic amino acids and less hydrophobic amino acids when compared with other salt-tolerant esterases. The cold active, salt-tolerant and chemical-resistant properties make it a promising enzyme for basic research and industrial applications.

Esterases (EC 3.1.1.1) and lipases (EC 3.1.1.3) are a class of hydrolytic enzymes that catalyze the hydrolysis and transesterification of fatty acid esters 1 . These enzymes are grouped into the family of serine hydrolases and share some similar structural and functional characteristics, including a conserved catalytic triad (Ser-Asp/ Glu-His), a consensus sequence (Gly-x-Ser-x-Gly) around the active residue serine 2 , an α /β hydrolase fold and cofactor-independent activity 3 . Esterases differ from lipases mainly in terms of the kinetics and substrate specificity 4 . Esterases have classical Michaelis-Menten kinetics and prefer water-soluble short-chain fatty acid esters (C < 10), whereas lipases prefer water-insoluble substrates, typically long-chain triglycerides (C > 10). Based on the amino-acid sequences and fundamental biological properties, bacterial esterases and lipases have been grouped into eight families 5 . A new subfamily of hydrolytic/lipolytic enzymes designated as SGNH-hydrolase superfamily or subfamily has been proposed based on the oxyanion structure 6 . Enzymes in this subfamily typically have a GDSL motif including the active serine that is located near the N terminus, and exhibit multifunctional properties and regiospecificity.
Esterases and lipases are widely distributed in nature and can be found in animals, plants and microorganisms. Microbial esterases and lipases have been extensively studied due to their great potential for applications in the medical and pharmaceutical industries, detergents, synthesis of fine chemicals, and bioremediation 2,7 . The majority of esterases and lipases that are currently used in industry are obtained from mesophilic fungi or bacteria. However, those from extremophiles, also known as extremozymes, are attracting much attention due to their excellent performance under extreme physico-chemical conditions and superiority over their counterparts for industrial applications 8,9 . Lipolytic enzymes from thermophiles and hyperthermophiles, psychrophiles, Scientific RepoRts | 6:19494 | DOI: 10.1038/srep19494 halophiles, alkalophiles/acidophiles, and solvent-resistant microorganisms have been studied by using culture-dependent and -independent methods 8,10 . Among them, lipolytic enzymes from (halo)alkaliphiles are of interest for their application potentials in laundry detergents, finishing of fabrics, and pulp and paper industries 8 .
Soda lake represents a natural sodium carbonate/bicarbonate-buffered system and is one of the most stable alkaline environments 11 . It is characterized by high alkalinity (pH from 9.0 to 12.0) and moderately to extremely high salinity (up to saturation), and provides a unique habitat to harbor a rich diversity of (halo)alkaliphilic bacteria and archaea 11,12 . These microorganisms are excellent sources of novel genes, enzymes and biomolecules with biotechnological potentials. For example, two industrial cellulases from the Gram-positive isolates of soda lake Kenyan have been used in textile and laundry processes 13 ; and alkaline proteases, lipase, amylases, chitinases and caseinases have also been reported from the microorganisms isolated from soda lakes 14 . Until recently, few studies on the esterases and lipases from soda lake microorganisms have been reported 15,16 .
In this study, a novel esterase gene (estSL3) was cloned from Alkalibacterium sp. SL3, which was isolated from the sediment of soda lake Dabusu. Sequence analysis suggested that EstSL3 belongs to the GDSL family and represents a member of the subfamily of SGNH hydrolases. The enzyme produced in E. coli showed cold-adapted, highly salt-tolerant and chemical/detergent/solvent-resistant properties. These properties make EstSL3 a good candidate for basic research and wide industrial applications.

Results
Strain identification. Based on the BLASTn analysis, the 16S rDNA sequence of strain SL3 (1515 bp) showed a nucleotide identity of 99.5% with that of Alkalibacterium sp. E-119 (FJ764767), 99.2% with A. pelagium strain NBRC 103242 (NR_114241), and 99.1% with A. pelagium strain T143-1-1 (NR_041574). Thus, strain SL3 was classified into the genus Alkalibacterium. The distance tree created by the neighbor-joining method also revealed the same classification (data not shown).
Gene cloning and sequence analysis of EstSL3. DNA fragments amplified through TAIL-PCR were assembled with the 3′ region, and an ORF of 636 bp, starting with ATG and terminating with TAA, was identified. The full-length gene (estSL3) encoded a polypeptide of 211 amino acid residues. No signal peptide was predicted. Its calculated molecular mass and theoretical isoelectric point were estimated to be 24.04 kDa and 5.28, respectively.
The deduced amino acid sequence of estSL3 showed the highest identity (69%) with a putative GDSL family lipase from Alkalibacterium sp. AK22 (WP_034300718), followed by the putative GDSL family lipases from Atopococcus tabaci (WP_028274330, 65% identity) and Lacticigenium naphtae (WP_035618581, 58% identity). A phylogenetic tree was constructed based on the amino acid sequences of EstSL3, its closest homologs and cold-adapted esterases retrieved from GenBank database. High bootstrap values separated these esterases into five major groups (Fig. 1). EstSL3 was closely related to the putative esterases from Alkalibacterium sp. AK22 (WP_034300718), A. tabaci (WP_028274330) and L. naphtae (WP_035618581), but was distant from other known cold-adapted esterases with functional verification.
Based on the multiple sequence alignment of EstSL3 and eight other esterases, five conserved regions of the GDSL family 6 were identified (Fig. S1). Three putative catalytic residues, Ser 15 , Asp 189 , and His 192 , were located in the conserved regions, and two residues, Gly 59 and Asn 88 , might form an oxyanion hole with Ser 15 . These results indicated that EstSL3 belongs to the SGNH hydrolase subfamily.
Modeled EstSL3 had a typical structure of SGNH hydrolases (Fig. 2), consisting of a single domain of five-stranded, parallel β -sheets, three helices at the convex side and two helices at the concave side of the sheet, and a short helix at the domain edge for ornament 6 . The catalytic triad of Ser 15 , Asp 189 and His 192 was located in a groove, and Gly 59 and Asn 88 together with Ser 15 formed the oxyanion hole. Expression and purification of rEstSL3. The gene fragment coding for the protein was expressed in E. coli BL21 (DE3). After induction with 0.5 mM IPTG at 30 °C for 12 h, significant esterase activity was detected after cell lysis. The crude enzyme was purified to electrophoretic homogeneity by Ni-affinity chromatography (Fig. S2). The purified rEstSL3 migrated as a single band of approximately 25 kDa on SDS-PAGE, which was identical to the calculated value (24.04 kDa). Three internal peptides obtained from LC-ESI-MS/MS, LTVLNRGIGGDSLKDLK, TDARVILMESFVLPYPKR and VGWRNDLDK, matched the deduced amino acid sequence of EstSL3 (Fig. S1), confirming that the purified enzyme was indeed EstSL3.
When assayed at 30 °C and using pNP-acetate as the substrate, purified rEstSL3 showed the highest activity at pH 9.0 among the pH tested, and remained 60% of the maximum activity between pH 7.5 and 9.0 (Fig. 3a). The rEstSL3 activity at pH higher than 9.0 was not tested due to the high self-degradation of pNP-acetate at alkaline pHs. The enzyme was stable over the neutral to alkaline pH range, retaining more than 80% of the initial activity after incubation in buffers ranging from pH 6.0 to 10.0 at 37 °C for 1 h (Fig. 3b). The thermal activity of purified rEstSL3 was apparently optimal at 30 °C when assayed at pH 9.0, and the enzyme retained greater than 50% of the maximum activity from 0 to 60 °C (Fig. 3c). At 0 °C, it exhibited 68% of the maximal activity. Without substrate, the enzyme was stable at 50 °C for more than 60 min, whereas at 55 °C and 60 °C, the half-lives of the enzyme were approximately 25 min and 2 min (Fig. 3d), respectively.
The effect of various metal ions and chemcial reagents on rEstSL3 activity was investigated ( Table 1). The activity of the purified rEstSL3 was enhanced by Li + , Na + , K + , Mg 2+ , Ni 2+ , Ca 2+ , β -mercaptoethanol and EDTA at both tested concentrations (1 and 5 mM), but inhibited by Ag + , Fe 2+ , Cu 2+ , Mn 2+ , Hg 2+ , Co 2+ , Zn 2+ , Cr 3+ and Pb 2+ at higher or both concentrations. rEstSL3 resistance against various detergents and organic solvents was also determined as shown in Table 2. As for the detergents we tested, the rEstSL3 activity was inhibited by tween-20 at 1-10% (v/v). In contrast, it was higly resisitent to tween-80, Triton X-100 and SDS, retaining more than 60% of the initial activity. On the other hand, The rEstSL3 activity was sensitive to all tested solvents. It was strongly inhibited by butanol, isobutanol, isoamyl, acetone and chloroform at concentrations of 20% (v/v), and slightly inhibited by methanol, ethanol, propanol, n-hexane, glycerol and acetonitrile. Purified rEstSL3 showed the highest activity in the presence of 2 M NaCl and retained greater than 98% esterase activity in the presence of 0.5-4.0 M NaCl (Fig. 4a). Moreover, purified rEstSL3 showed strong tolerance to high concentrations of NaCl, retaining more than 87% esterase activity after 2-h incubation with 5 M NaCl at 37 °C and pH 9.0 (Fig. 4b).

Discussion
Lipolytic enzymes from extremophiles have been attracting more and more attention due to their great application potentials in various industries 8 . Soda lake represent unique, saline and alkaline niche to harbor a microbial community that adapts to these extreme conditions 11 . Most of the microorganisms isolated from soda lakes have been described as halophiles and alkaliphiles, and diverse alkaline and salt-tolerant enzymes have been reported from them 12 . The lake Dabusu is a typical soda lake that has a high salinity and alkalinity 17 . Our previous studies have revealed the rich diversity of xylanase genes in the metagenome of soda lake Dabusu 18 , the source of thermophilic and salt-and alkaline-tolerant xylanase 19 . In this study, a novel esterase gene, estSL3, was cloned from Alkalibacterium sp.SL3, a strain isolated from the sediment of the lake Dabusu. EstSL3 has low similarities with  known sequences, which are all putative lipases of GDSL family or hypothetical proteins without functional verification. Multiple sequence alignment of EstSL3 and eight close homologs (Fig. S1) suggested that EstSL3 could be classified into the SGNH subfamily that is characterized with broad substrate specificity and regiospecificity 6 .  However, when EstSL3 was produced in E. coli, it exhibited a relatively narrow substrate spectrum. It showed the highest activity towards pNP-acetate and low or no activity towards pNP-esters with side chains longer than C2.
Considering the difference between lipase and esterase, EstSL3 represents an esterase instead of a true lipase. To the best of our knowledge, this is the first report of an esterase from the genus Alkalibacterium.
Cold-adapted enzymes usually have low temperature optima and retain high activity at 0 °C 20,21 . rEstSL3 demonstrated the typical characteristics of cold-adapted enzymes, including a temperature optimum of 30 °C and 68% activity even at 0 °C (Fig. 3). Over the past few years, a number of cold-adapted esterases have been cloned and characterized from cultured microorganisms [22][23][24][25] or by using metagenomic method [26][27][28][29] . Phylogenetic analysis ( Fig. 1) and sequence alignment (Fig. S1) indicated that EstSL3 is distantly related to those reported cold-adapted or cold-active esterases and has some novelty. Except EstPc from Psychrobacter cryohalolentis K5 T 22 , EstB from Alcanivorax dieselolei B-5(T) 25 and Lp_2631 from Lactobacillus plantarum 30 , EstSL3 showed greater or comparable performance under low-temperature conditions to other characterized counterparts, including Est10 and Est11 from psychrotrophic Psychrobacter pacificensis 23,24 , Est97 from an Arctic intertidal metagenomic library 27 and EstF from a deep-sea metagenomic library 26 . Although rEstSL3 owned the cold-adapted properties, it also had better thermostability than most cold-adapted esterases at high temperatures (50 °C). These characteristics may widen the application range of EstSL3 to the pharmaceutical, agricultural and chemical industries and so on.
Compared with mesophilic or thermophilic counterparts, cold-adapted enzymes usually employ several mechanisms to confer the enzymes high flexibility and catalytic efficiency at low temperatures. These factors include but are not limited to more glycine residues, fewer proline residues, a lower Arg/Arg + Lys ratio and a declined number of disulfide bridges and fewer weak interactions such as salt bridges, aromatic interactions and hydrogen bonds 20,21 . In the case of EstSL3, several characteristics adapting to cold-environment can be observed (Table 3). Firstly, it has fewer proline residues than cold active esterases EstPc 22 and Est97 27 , mesophilic esterase Axe2 31 and thermophilic esterase EstA 32 . Secondly, EstSL3 has fewer salt bridges. We found that all the cold-active esterases (EstSL3, Lip1Pc and Est97) have fewer salt bridges than mesophilic Axe2 and thermophilic EstA, and EstSL3 has even fewer salt bridges than cold-adapted Lip1Pc and Est97. Thirdly, fewer hydrogen bonds were found in EstSL3. The hydrogen bond percentage of EstSL3 is much lower than that of mesophilic Axe2, thermophilic EstA, and cold-active Lip1Pc and Est97. All of these factors in combination might lead to the enhanced flexibility but relatively increased stability of the structure of EstSL3 and consequently have a key role in maintaining its high catalytic activity at low temperature and great thermostability at high temperature.
Enzymes from bacteria isolated from soda lakes are usually capable of functioning at high pH and possibly high salt concentrations 12 . Although the activity of rEstSL3 at pH higher than 9 was not assayed due to the self-degradation of pNP-acetate at extreme alkaline pHs, EstSL3 could be regarded as an alkaline esterase based on the results we obtained (Fig. 3). Another noteworthy characteristic of rEstSL3 is the stability with increased NaCl concentration (Fig. 4). It is confirmed that an increase in the number of charged amino acids, especially acidic residues at the protein surface, can confer halophilic proteins enhanced activity at high salt concentrations 33 . We found that charged amino acids constitute about 30.3% (64/211) of EstSL3, with higher percentage of acidic residues (16.6%) than basic residues (13.7%). EstSL3 also has higher percentage of acidic residues when compared with other salt-tolerant esterases ( Table 4). The surface electrostatic potential analysis suggested that the charged amino acids of EstSL3 are distributed on the surface of the protein (Fig. 2). Moreover, EstSL3 has lower percentage of hydrophobic amino acids. These factors might make EstSL3 to form a solvation shell that keeps the protein surface hydrated and thus highly tolerant to salt.
In addition, the effects of metal ions, detergents and organic solvents on rEstSL3 activity (Tables 1 and 2) were also studied. EstSL3 is resistant to most tested chemicals, including Li + , Na + , K + , Ni 2+ , Mg 2+ , Ca 2+ , β -mercaptoethanol, EDTA, and low concentrations of tween-20, tween-80, Triton X-100, methanol, ethanol, propanol, Acetonitrile, glycerol and n-hexane. Most of the reported cold-active esterases are strongly inhibited by SDS, an anionic detergent that causes strong protein denaturation. For example, Est10 from P. pacificensis only retained 2.8% activity in the presence of 1% SDS 23 while the EstPc was completely inhibited by 0.05% SDS 22 . In  Table 3. Factors affecting the stability and flexibility of EstSL3 and other cold active mesophilic or thermophilic counterparts.
contrast, rEstSL3 retained 68.5% activity at the presence of 1% SDS, and had greater SDS-resistance over other cold-active esterases.
In conclusion, a novel esterase gene estSL3 was cloned from a soda lake isolate, Alkalibacterium sp. SL3, and successfully expressed in E. coli. rEstSL3 is a cold-adapted, highly salt-tolerant enzyme with greater thermostability over most cold-active esterases. Moreover, the enzyme was slightly activated by several metal ions and detergents and showed tolerance towards SDS and organic solvents. All these enzymatic properties make EstSL3 a good candidate for basic research and broad industrial applications. Microorganism isolation. The sediment sample was collected from soda lake Dabusu, which is located in the southwest of Qian'an County, Jilin Province, China. The lake Dabusu has a salinity of 62.34 g L −1 to 347.34 g L −1 and a pH of 10 to 11 17 . Strain SL3 was isolated from the sediment sample as described by our previous study 19 . The taxon of the strain was identified by the 16S rDNA sequence PCR-amplified using primers 27F and 1492R (Table S1).

Strains
Gene cloning of the full-length esterase gene (estSL3). Genomic DNA was extracted from strain SL3 using the Omega genomic DNA isolation kit following the manufacturer's instructions. The 3′ end of the esterase gene was obtained when we amplified the flanking regions of a xylanase gene by using thermal asymmetric interlaced (TAIL)-PCR 34 . Then three specific primers were designed to obtain the flanking region of 5′ end (Table S1). The PCR products were excised, purified, and ligated into vector pMD 18-T. The recombinant vector was then transformed into E. coli DH5α and sequenced by Invitrogen (Carlsbad, CA, USA). The full-length esterase gene was designated as estSL3.
Sequence and phylogenetic analysis. The open reading frame (ORF) was identified by using the Vector NTI 10.3 (InforMax, Gaithersburg, MD, USA). The signal peptide sequence was predicted using SignalP (http:// www.cbs.dtu.dk/services/SignalP/). The DNA and protein sequence similarities were assessed by using the BLASTn and BLASTp programs (http://www.ncbi.nlm.nih.gov/BLAST/), respectively. Multiple sequence alignments were performed with ClustalW (http://www.ebi.ac.uk/Tools/clustalw2/). A phylogenetic tree, including deduced EstSL3 and its closest homologs, was constructed using the neighbor-joining (NJ) algorithm in MEGA 4.0 35 . Confidence for the tree topology was estimated using the bootstrap values based on 1,000 replicates.
Putative structure analysis. The protein sequence of EstSL3 was submitted to I-TASSER 36 (http:// zhanglab.ccmb.med.umich.edu/I-TASSER/), which represents one of the best CASP 37 wining ab initio protein folding and three-dimensional (3D) structure prediction servers. The acetylxylan esterase Axe2 31 (PDB: 3w7v) from Geobacillus stearothermophilus was identified as a homologous template for EstSL3 modeling with the global sequence identity of 35%. Based on the estimated RMSD value and TM-score of 2.8 ± 2.1 Å and 0.90 ± 0.06, respectively, EstSL3 was classified as an 'Easy' target by I-TASSER and the predicted model is therefore reliable 38 . The predicted model and surface electrostatic potential were visualized via Pymol with the assistance of APBS plugin. Prediction of disulfide bridges, salt bridges (distances < 3.2 Å), and hydrogen bonds were performed as described by Zhou et al. 39 . Two cold-adapted esterases LipPc 22 and Est97 27 , a mesophilic esterase Axe2 31 as well as a thermophilic esterase EstA 32 were selected for comparison with EstSL3.  Expression and purification of EstSL3 in E. coli. The full-length gene of estSL3 was amplified by PCR using the expression primers (Table S1), and cloned into the NcoI-HindIII site of pET-28a (+ ). The recombinant plasmid, pET-estSL3, was transformed into E. coli BL21 (DE3) competent cells. Positive transformants harboring the recombinant plasmid (pET-estSL3) were identified by PCR and further confirmed by DNA sequencing. The cells were grown in LB medium containing 100 μ g mL −1 of ampicillin at 37 °C to an A 600 of 0.6. Protein expression was induced by addition of IPTG at a final concentration of 0.5 mM at 30 °C for 12 h. Esterase activities of the cell pellet were assayed as described below.
To purify the His-tagged recombinant protein (rEstSL3), the cells were harvested by centrifugation (12,000 × g, 4 °C for 10 min) and washed with sterile distilled water. The cells were then resuspended in sterilized ice-cold buffer (20 mM Tris-HCl, 0.5 M NaCl, pH 7.6) and disrupted by sonication (6 s, 160 W) on ice. The crude enzyme was collected (12,000 × g for 10 min at 4 °C) and loaded onto a Ni 2+ -NTA agarose gel column. The purified enzyme was washed with a linear imidazole gradient of 20-300 mM in Tris-HCl buffer (20 mM Tris-HCl, 500 mM NaCl, pH 7.6).
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was used to determine the purity and apparent molecular mass of rEstSL3. The protein concentration was determined by the Bradford method 40 , using bovine serum albumin as a standard. The identity of the purified enzyme was verified by liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS).
Enzyme assay. Esterase activity assay was performed with pNP-esters as the substrates, and the production of p-nitrophenol was measured at 405 nm 41 . Reactions containing 0.1 mL of enzyme solution, 0.1 mL of 10 mM substrate and 1.8 mL of 50 mM Tris-HCl (pH 9.0) were incubated at 30 °C for 5 min. One unit (U) of esterase activity was defined as the amount of enzyme that released 1 μ mol of p-nitrophenol per minute. The extinction coefficients of p-nitrophenol were measured at different pHs (ranged from 3,400 to 19,600 M -1 cm -1 ) and used to correct the results. All reactions were performed in triplicate. The controls were reaction systems with addition of thermo-inactivated EstSL3.
Biochemical characterization. The substrate specificity of purified rEstSL3 was determined in 50 mM Tris-HCl buffer (pH 9.0) containing 1 mM of pNP-esters (C2-C16) under standard conditions (30 °C and 5 min). The stock substrate solutions (10 mM) were prepared by dissolving pNP-esters in pure acetonitrile.
The optimal pH for esterase activity of the purified rEstSL3 was determined at 30 °C in buffers with pH ranging from 4.0 to 9.0. The stability of purified rEstSL3 at different pH values was estimated by incubating the enzyme solution in various buffers at 37 °C for 1 h without substrate. The remaining activity was measured in Tris-HCl buffer (pH 9.0) at 30 °C for 5 min. The initial activity of rEstSL3 was set as 100%. The buffers used were McIlvaine buffer (0.2 M Na 2 HPO 4 , 0.1 M citric acid) for pH 4.0-8.0, 0.1 M Tris-HCl for pH 8.0-9.0 and 0.1 M glycine-NaOH for pH 9.0-12.0.
The optimal temperature for purified rEstSL3 activity was determined over the range of 0-70 °C in Tris-HCl buffer (pH 9.0). Thermostability of rEstSL3 was determined by measuring the residual activities after pre-incubation of the enzyme in Tris-HCl buffer (pH 9.0) at 50, 55 and 60 °C for various periods.
The K m , V max , and k cat values of rEstSL3 were determined in Tris-HCl buffer (pH 9.0) containing 0.05-5 mM pNP-acetate at 30 °C for 5 min. The K m and V max were determined from a Lineweaver-Burk plot using the non-linear regression computer program GraFit (Erithacus, Horley, Surrey, UK).
The effect of NaCl on the purified rEstSL3 activity was determined at 30 °C in Tris-HCl buffer (pH 9.0) containing 0.5-4.0 M NaCl. To examine its resistance to salt, rEstSL3 was incubated with 4 M of NaCl at 37 °C for 1 h, and the residual enzyme activities were measured.