Introduction

Esterases (EC 3.1.1.1) and lipases (EC 3.1.1.3), commonly recognized as lipolytic enzymes, mainly catalyze the hydrolysis of short- and long-chain triacylglycerols, respectively, into fatty acids, diacylglycerol, monoacylglycerol, and glycerol (Arpigny and Jaeger 1999; Carrière et al. 1994). Most lipolylic enzymes belong to the α/β-hydrolase family, and possess similar structural and functional characteristics, including a catalytic triad, an α/β-hydrolase fold and independence of cofactors. Also, the catalytic triad is conserved and universally composed of a nucleophilic serine residue in the GXSXG motif (X refers to any residue), an acidic residue (glutamate or aspartate) and a histidine residue (Levisson et al. 2009). Lipolytic enzymes exist widely in animals, plants and microorganisms, involved in catalyzing transesterification, acidolysis, interesterification, alcoholysis and aminolysis in addition to the hydrolytic activity on triglycerides (Jaeger and Eggert 2002). They have become one of the most versatile biocatalysts due to their unique characteristics such as good chemioselectivity, regioselectivity, enantioselectivity, independence of cofactors and possessing broad substrates with optimum activities over a wide range of temperature (Buchholz et al. 2005). Currently, they have been broadly utilized in a mixed range of fields such as laundry detergents, food technology, pharmaceutical and chemical industries (Pandey et al. 1999).

Since extreme conditions (low/high temperatures, extreme pH, high salinity, detergents or organic solvents) have been found to restrict the catalytic ability of lipolytic enzymes, an increasing attention has been paid to explore some biotechnologically relevant enzymes from extremophilic microbes which live in such extreme environments such as permanently cold deep-sea, hot springs and oil-field waters (Levisson et al. 2009; Mander et al. 2012; Novototskaya-Vlasova et al. 2012; Tirawongsaroj et al. 2008) because the proteins produced by these microbes are believed to have evolved special structural and functional features to cope with these extremely adverse surroundings.

Currently, cold-adapted lipolytic enzymes have emerged as one of the most promising biocatalysts because, when compared with those from mesophiles or thermopiles, they display a much higher catalytic efficiency at a low or moderate temperature, and tend to be inactivated at a high temperature. These features of low thermostability and high activity at low temperatures can well meet the demands of some industrial applications including additives in laundry detergents (cold washing), additives in food process (fermentation, cheese manufacture, bakery, meat tenderizing), and environmental bioremediations (degradation of lipid wastes) (Joseph et al. 2008). However, so far only a few cold-adapted lipolytic enzymes have been discovered and characterized (Al Khudary et al. 2010; Fu et al. 2013; Novototskaya-Vlasova et al. 2012; Simkhada et al. 2012; Wang et al. 2012; Zhang and Zeng 2008).

Permanently cold habitats are good sources of cold-adapted lipolytic enzymes as well as other kinds of proteins, where the organisms and their proteins have high biochemical and physiological adaptation to cold conditions. In general, the polar regions are a constantly cold environment where many cold-adapted lipolytic proteins exist (Lo Giudice et al. 2006; Parra et al. 2008; Zhang et al. 2007). In addition, glaciers and mountain regions are also accessible habitats for cold-active lipase producing organisms (Joseph et al. 2008). Another potential source of cold-adapted proteins is deep-sea bacteria and sediments, where cold-active lipase producing strains are frequently isolated, including the Psychrobacter sp. wp37 (Zeng et al. 2004), Aeromonas sp. LPB 4 (Lee et al. 2003) and Psychrobacter sp. C18 (Chen et al. 2010).

In the present paper, we report on the cloning, expression and biochemical characterization of a cold-adapted and salt-tolerant esterase, obtained from the psychrotrophic bacterium Psychrobacter pacificensis (Gammaproteobacteria, Pseudomonadales, Moraxellaceae), which was isolated from sediments in the Gulf of Mexico. This work could add new a resource to the widely utilized cold-adapted esterases and provide some useful information regarding adaptability of proteins to cold environments and high salinity, as well as lay the foundation for its potential industrial applications.

Materials and methods

Materials

Substrates p-nitrophenyl acetate (pNPA), p-nitrophenyl butyrate (pNPB), p-nitrophenyl decanoate (pNPD), p-nitrophenyl caprylate (pNPC), p-nitrophenyl laulate (pNPL), and p-nitrophenyl palmitate (pNPP) were purchased from Sigma-Aldrich (St Louis, MO, USA). All the other chemicals used were of analytical grade and were commercially available.

Bacterial strains

Psychrobacter pacificensis was supplied by the Marine Culture Collection of China (the strain number: MCCC 1A06352). Strain was inoculated in the medium (peptone 1 %, yeast extract 0.5 %, NaCl 2 %) with shaking at 150 rpm, 28 °C for overnight. Escherichia coli strains DH5α and BL21 (DE3) were used for cloning and expression, respectively, and were cultivated at 37 °C in Luria–Bertani (LB) medium (yeast extract 0.5 %, NaCl 1 %, peptone 1 %).

Genomic library construction and screening

The genomic DNA of Psychrobacter pacificensis was extracted and partially digested by Sau3AI, before the resultant DNA fragments of 4–10 kb were purified and cloned into the vector pUC118. The E. coli DH5α cells were transformed with the recombinant plasmids and grown at 37 °C overnight on LB agar plates supplemented with 1 mM isopropyl-d-thiogalactopyranoside (IPTG) and 40 μg/ml of 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-gal) for white-blue screening. For identification of the lipase-active transformants, all the white clones were transferred into the lipase screening plates (LB agar plates plus 1 % tributyrin as the substrate) and grown at 37 °C overnight. Finally, the plates were incubated at 28 °C for a few days until the formation of clear halos around the clones with the lipolytic activity.

Sequence analysis

Nucleotide sequencing was performed by the Genscript Biotechnology Company (Nanjing, China). The open reading frame (ORF) and the promoter region in the obtained DNA fragments were analyzed by FGENSB and BPROM (http://linux1.softberry.com/berry.phtml), amino acid alignments by BLASTP (http://www.ncbi.nlm.nih.gov/), and Signal peptide by the SignaIP 4.0 Server (http://www.cbs.dtu.dk/services/SignalP/). The multiple sequence alignment was carried out using Multiply Sequence Alignment in DNAMAN program.

Cloning of est10 gene

The est10 gene was amplified using the forward primer EstF 5′-CGGGATCCATGAGTAACTATTT AGACGCCGTC-3′ (with BamHI restriction site underlined) and reverse primer EstR 5′-CGGAATT CTTAAGCCAAGACGGTGTTTAGCCA-3′ (with EcoRI restriction site underlined) using the Psychrobacter pacificensis genomic DNA as the template. Products were digested by BamHI and EcoRI and cloned into the expression vector pGEX-6p-1 (GE Healthcare Life Sciences, USA), and then the recombinant plasmids were transformed into the E. coli BL21 (DE3) cells. Two clones were sequenced and a clone with the correct sequence, harboring the recombinant plasmid pGEX-6p-1-est10, was selected for further protein expression and characterization experiments.

Protein expression and purification

The E. coli BL21 (DE3) cells, carrying the plasmid pGEX-6P-1-est10, were grown overnight at 37 °C in LB medium containing 100 μg/ml ampicillin. Subsequently, the culture was added to fresh LB medium (1:100 dilution) containing 100 μg/ml ampicillin and grown at 37 °C until the optical density at 600 nm reached approximately 0.6, when 0.5 mM IPTG was added. After another 10 h cultivation at 22 °C, the cells were collected by centrifugation at 6,000 rpm for 10 min, washed with 50 ml PBS buffer (0.8 % NaCl, 0.02 % KCl, 0.142 % Na2HPO4, 0.027 % KH2PO4; pH 7.4), resuspended in the same buffer, followed by the disruption through a French pressure cell, and then the centrifugation at 12000 rpm at 4 °C for 30 min and finally, the collection of the supernatant. The enzyme solution was then purified using glutathione S-transferase (GST) Gene Fusion System (GE Healthcare Life Sciences, USA) as described (Cao et al. 2008). The GST-Est10 fusion protein was bonded to glutathione affinity resins, and then Est10 was released from the bound GST tag by the 3C protease solution (10 U/μl, PreScission, Pharmacia). The quantification of the protein was determined using the Bradford reagent (Sigma, USA) with bovine serum albumin (BSA) as a standard (Bradford 1976) and the proteins were analyzed by sodium dodecyl sulfate-denatured polyacrylamide gel electrophoresis (SDS-PAGE).

Esterase activity assay

Standard lipolytic activity assay was performed using p-NPB (C4) as the substrate. The standard reaction mixture was composed of 2 μl of substrate (final concentration 1 mM), 196 μl of buffer (phosphate–citrate buffer, pH 7.5), and 2 μl of purified protein Est10. The reaction was started by adding the substrate to the premixed protein and buffer at room temperature (about 25 °C), and change in absorbance at 405 nm was recorded using Thermo Scientific Multiscan Spectrum within 2 min. The enzyme was replaced by the buffer in the case of a blank. One unit of esterase activity was defined as the release of 1 μmol of p-nitrophenol per min (p-NP as standard).

Characterization of Est10

The optimum esterase reaction temperature was determined from 0 to 70 °C under standard assay conditions. The buffer solutions were pre-incubated in the water bath for 5 min to reach the respective temperature before adding the proteins and substrates. For the thermostability study, the enzymes were incubated at a temperature ranging from 30 to 70 °C, followed by putting them on ice water for cooling, and the remaining activity was measured under standard conditions at 25 °C. The proteins treated at 4 °C were used as the control.

The optimal pH condition was determined in phosphate–citrate buffer (pH 4.0–8.0) and Tris–HCl buffer (pH 8.0–9.5). The esterase pH stability was determined by measuring the residual activity after incubating the enzyme for 24 h in phosphate–citrate (pH 2.2–8.0), and 200 mM Glycine–NaOH (pH 8.0–12.0).

The effect of various metal ions [using Hepes buffer pH 7.5, excluded of KCl and dithiothreitol (DTT)] and detergents on Est10 was assessed by adding them to the reaction mixture under various concentrations in the phosphate–citrate buffer (pH 7.5). The effects of salt concentration ranging from 0.5 to 5 M on the activity and stability of Est10 were also tested. The detergents resistance of Est10 was assayed by measuring the residual activity after incubating Est10 with different detergents Tween 20, Tween 80, Triton X-100, 3-[(3-Cholamidopropyl) dimethylammonio]-1-propanesulfonate) (CHAPS), sodium dodecyl sulfate (SDS) and cetyltrimethylammonium bromide (CTAB) under various concentrations (0.05–0.1 %) for 30 min at 4 °C.

The effect of various organic solvents including ethanol, isopropanol, ethanediol, propanediol, propanetriol, acetonitrile and dimethyl sulphoxide (DMSO) on Est10 was also estimated by determining the residual activity after 1-h incubation of the enzymes with the solvents at a different concentration (10 %, 20 %, 30 %).

The substrates p-nitrophenyl esters with variable chain length (C2–C16) were used for substrate-specificity experiment by measuring the amount of p-nitrophenol released.

The kinetic parameters K m and k cat were determined by measuring the absorbance at 405 nm with various concentrations (0.2–2 mM) of p-NPB (C4). The initial reaction velocities measured with various concentrations of the substrate were fitted to the Lineweaver–Burk transformation of the Michaelis–Menten equation allowing the calculation of K m and V max. And the k cat was calculated by the equation: k cat = V max/[E], where [E] means the total protein molecular weight in the reaction mixture.

Structural modeling

To analyze the surface electrostatic potential of Est10, the homology model of Est10 was constructed using the software Molecular Operating Environment (MOE) as follows: (1) searched the homologous sequences of Est10 within the PDB data base using the PDB search under the Homology menu; (2) generated the homologous sequences with reasonable valve of Z (>6) and loaded them into the sequence editor; (3) aligned sequences of the loaded proteins and Est10 before displaying actual secondary structure; (4) chose the template with the highest Z value for homology model using Amber99 with a Reaction Field (R-field) treatment of solvation electrostatics to make a 3D structure. The superimposition of Est10 with the template, as well as the catalytic residues of both, was generated by the MOE through the “superpose”. Based on the model, the electrostatic potential surface of Est10 was displayed using the MOLCAD of the software SYBYL.

Nucleotide sequence accession numbers

The gene sequence of est10 was deposited in the GeneBank database with the accession number KC291403.

Results

Esterase gene clone and sequence analysis

The genomic library of Psychrobacter pacificensis was constructed in E. coli DH5α containing a total of 8000 clones and screened for clones of lipolytic activity on tributyrin agar plates. As a result, a clone forming a clear halo was selected.

Sequencing result of pEsc10 showed an inserted DNA fragment of 4799 bp. One predicted esterase open reading frame (ORF) gene (designated est10) of 672 bp, encoding 223 amino acids with a hypothetical molecular weight of 24,647 Da, was predicted by FGENSB. The predicted gene, est10, is predicted to start with an ATG codon and terminated in a TAA codon, with a 49.4 % GC content. The putative −35 box (TTGATA) and the putative −10 box (TCATAAAAT) were also identified. No signal peptide was found through SignaIP 4.0 Server upstream of est10, indicating that Est10 was potentially an intracellular esterase. The deduced amino acids of Est10 shared 92 % identities with the uncharacterized putative esterase from Psychrobacter sp. PAMC 21119 (ZP_10789976.1), and 88, 86 and 73 % with those from Psychrobacter arcticus 273-4 (AAZ19302.1), Psychrobacter cryohalolentis K5 (ABE75410.1) and Psychrobacter sp. 1501(2011) (EGK15152.1), respectively. The putative catalytic triad, composed of Ser121 (in the motif GFSQG), Asp173and His204, was revealed by the alignment of protein sequences among Est10 and other esterases of crystallographic structure from Pseudomonas aeruginosa (44 % identities from residue 20 to 222, PDB code 3CN7), Francisella tularensis (44 % identities from residue 17 to 218, PDB code 4F21), Pseudomonas fluorescens (43 % identities from residue 17 to 223, PDB code 1AUO) and Human Acyl Protein Thioesterase (35 % identities from residue 1 to 222, PDB code 1FJ2) (Fig. 1).

Fig. 1
figure 1

Alignment of the amino acid sequence of esterase from Psychrobacter pacificensis with other esterases of crystallographic structure. Est10, the esterase from Psychrobacter pacificensis (this study). PAC, the carboxylesterase from Pseudomonas aeruginosa. FTC, the carboxylesterase from Francisella tularensis. PFC, the carboxylesterase from Pseudomonas fluorescens. HAPC, Human Acyl Protein Thioesterase. The consensus GFSQG sequence containing the active-site serine residue is boxed. The amino acids forming the catalytic triad, Ser, Asp and His are indicated by a filled triangle

Full size image

Expression, purification and characterization of Est10

Results from purification of Est12 (Fig. 2) revealed that the recombinant fusion protein (about 51 kDa) which consists of Est10 (24.6 kDa) and a GST tag (26 kDa) was efficiently expressed as a soluble protein in the heterologous host E. coli BL21 (DE3). The purified Est10 (the GST tag was removed) showed a single band with an apparent molecular mass of about 25 kDa.

Fig. 2
figure 2

SDS-PAGE analysis of the overexpression and purification of the Psychrobacter pacificensis esterase Est10 produced in E. coli BL21 (DE3). Lane M standard molecular weight marker, Lane 1 purified Est10, Lane 2 uninduced cell lysate of E. coli BL21 (DE3) harboring pGEX-6p-1-est10, Lane 3 IPTG-induced cell lysate of E. coli BL21 (DE3) harboring pGEX-6p-1-est10. The fusion protein GST-Est10 is indicated

Full size image

Est10 activity peaked around 25 °C and remained more than 80 % of the maximum activity from 15 to 40 °C. The relative activity of Est10 was about 55.0 % at 0 °C, indicating it is a cold-adapted esterase (Fig. 3a). After incubation for 2 h, Est10 lost less than 20 % activity at 30 and 40 °C. After incubation at 50 °C for 2 h, the activity of Est10 decreased to 34.2 % of the original activity. The enzyme lost about 80 % of activity after being treated for 20 min at 60 and 70 °C (Fig. 3b).

Fig. 3
figure 3

Effect of pH and temperature on enzyme activity and stability. a The effect of temperature on enzyme activity. The activity was determined at a different temperature ranging from 0 to 70 °C at pH 7.5 for 5 min. The specific activity (4.8 U/mg) at 25 °C was defined as 100 %. b temperature stability. The residual activity was measured under 25 °C after the enzyme incubation at different temperatures (30 °C, filled circle; 40 °C, filled square; 50 °C, filled triangle; 60 °C, filled inverted triangle; 70 °C, filled diamond) at a interval of 20 min in 2 h. The specific activity (4.9 U/mg) without incubation was defined as 100 %. c effect of pH on enzyme activity. The esterase activity was assayed at 25 °C for 5 min. Buffers were phosphate–citrate (pH 5.0–8.0) and Tris–HCL (pH 8.0–9.5). The specific activity (6.1 U/mg) under pH 7.5 was defined as 100 %. d pH stability. The residual activity was determined after incubation of Est10 at different pH buffer, phosphate–citrate (pH 4.0–8.0) and Glycine-NaOH (pH 8.0–12.0), for 24 h at 4 °C. The specific activity (4.5 U/mg) after being treated with pH 8.0 buffer was defined as 100 %

Full size image

The optimal pH of Est10 was around pH 7.5, and about 80 and 60 % of its maximum activity were observed at pH 7.0 and pH 8.0, respectively (Fig. 3c). After incubation for 24 h, Est10 was found to be stable over pH 7.0 to 10.0 and retain even 40 % of the maximum activity at pH 12.0 (Fig. 3d), indicating its relative stability in alkaline solution.

The presence of 0.05 % non-ionic detergents (Tween 20, Tween 80, CHAPS, Triton X-100) resulted in higher relative activity compared to measurements at 0.1 % (Table 1). Incubation with 0.05 and 0.1 % CTAB reduced the activity to less than 30 % while SDS almost inactivated it. Stability in non-ionic detergents (activity after 0.5 h in 0.05 or 0.1 %), showed that Est10 activity increased about twofold compared to the control while stability in ionic detergents generally decreased activity (except for 0.1 % CTAB).

Table 1 Effect of detergents on the activity and stability of Est10
Full size table

The effects of different metal ions, ethylene diamine tetraacetic acid (EDTA), urea, DTT and phenylmethanesulfonyl fluoride (PMSF) were also tested (Table 2). Esterase activity was below the detectable limit at 1 mM of the serine inhibitor PMSF, which reacts with the proposed serine residue in the catalytic triad Ser-Asp-His of Est10. When raised to 5 mM, Ba2+, Ca2+, urea and EDTA could stimulate the activity while other chemicals, including the heavy metal ions Cu2+, Zn2+, Sr2+ and Mg2+ showed little or inhibitory effects. Particularly, the addition of 5 mM DTT and Fe3+ exerted the strongest negative effect, but still 69.8 and 75.1 % of the original activity could be observed, respectively. Interestingly, the Est10 activity was significantly stimulated by the salt ranging from 0.5 to 4 M and peaked around 140 % with 1 or 2 M NaCl (Fig. 4). Moreover, even at the concentration of 5 M NaCl, the activity of Est10 was still robust, reaching 81.8 % of its original activity. The effect of NaCl on activity was confirmed again by the stability test, in which the overall activity of Est10 incubated with 0.5–5 M NaCl for 6.5 h was enhanced with the highest 164.5 % activity at 3 M and even 126.4 % activity at 5 M.

Table 2 Effect of various reagents on the activity of Est10
Full size table
Fig. 4
figure 4

Effect of NaCl on the activity (filled circle) and stability (filled square) of Est10. Filled circle activity was measured in pH 7.5 hepes buffer with a different concentration of NaCl at 25 °C. The specific activity (5.0 U/mg) without NaCl was defined as 100 %. Filled square after incubating Est10 in pH 7.5 hepes buffer with different concentrations of NaCl at 4 °C for 6.5 h, the relative activity was determined under the standard conditions. The specific activity (4.8 U/mg) after incubation with 0 M NaCl was defined as 100 %

Full size image

Stability in organic solvents is an important property of protein catalysts used in organic synthesis reactions. For this reason, after being treated with various organic solvents for 1 h, the remaining activity of Est10 was determined. As shown in Fig. 5a, incubation with 10–30 % DMSO, ethanediol, methanol, propanediol, glycerine, 10 and 20 % ethanol, or 10 % isopropanol resulted in greater than 85 % activity. However, the activity was greatly reduced by 10 % n-butanol. At a concentration of 30 % n-butanol, acetonitrile, ethanol and isopropanol, little or no activity could be detected.

Fig. 5
figure 5

Substrate specificity and the effect of organic solvents on the activity of Est10. a Effect of the organic solvents on Est10 activity. White bars, gray bars and black bars, respectively, represent the residual activity after 1-h pre-incubation of the enzymes in 10, 20 and 30 % of each organic solvent. The specific activity (5.1 U/mg) without organic solvent was defined as 100 %. b Substrate specificity of Est10. The relative activities to different substrates were determined in the presence of each p-nitrophenyl ester (C2–C16, 2 mM) with the specific activity (6.9 U/mg) to the optimal substrate pNPB (C4) as 100 %

Full size image

The substrate-specificity assay indicated that p-nitrophenyl esters with short-and medium-chain fatty acids were the optimum substrates for Est10 (Fig. 5b), because Est10 displayed the highest activity toward pNPB (C4). Its kinetic parameters K m, k cat and V max were measured to be 0.7667 mM, 3.92 s−1 and 9.54 μm min−1 mg−1, respectively. Only 7 % activity was observed against pNPL (C12) and no activity toward pNPP (C16).

Structural modeling

To uncover the surface electrostatic potential of Est10, we generated a 3D structural model using the software Molecular Operating Environment (MOE). According to the alignment result, 3 best structural templates were obtained, with the E value ranging from 1.9e −41 for the best hit to 1.0e −2 for the most distant hit. Based on the best template of carboxylesterase Pa3859 (PDB code 3CN9, 40 % identities to Est10 for the full length) from Pseudomonas aeruginosa, the final model (Fig. 6a) covering residue 8–222 of Est10 was constructed and well superimposed onto the structure of Pa3859, with a small rmsd of 0.703 Å (Fig. 6b). The overall quality of the model was estimated through the evaluation of Ramachandran plot by MOE. The result of Ramachandran plot showed that 213 of the total modeled 215 residues were located in the favorable region of the plot except for 2 residues (Lys17, Ala149), indicating the overall good quality of the model (data not shown). Electrostatic potential calculation of Est10 surface was undertaken through SYBYL indicating that the surface exhibited a high negative electrostatic potential (Fig. 6d).

Fig. 6
figure 6

The homology model and the surface electrostatic potential of Est10. a The model of Est10 constructed by homology model through MOE. α-helices are shown in red, β-strands in yellow and the active site in green. b The superimposition between model of Est10 (green) and the crystal structure of carboxylesterase Pa3859 from Pseudomonas aeruginosa (yellow). c Superimposition of the catalytic triad between model of Est10 (green) and Pa389 (yellow). d The surface electrostatic potential of Est10 obtained by SYBYL. The negative and positive electrostatic potentials are indicated by blue and red, respectively. The right image is the 180° rotated view of the left one

Full size image

Discussion

In this study, we isolated and characterized a cold-adapted, and salt-tolerant esterase (Est10) screened from the genomic library of a psychrotrophic bacterium Psychrobacter pacificensis. Many other cold-adapted lipolytic enzymes were also identified through biotechnology-related methods including PCR amplification, and the genomic or metagenomic library construction (Al Khudary et al. 2010; Fu et al. 2013; Jeon et al. 2009; Kulakova et al. 2004; Novototskaya-Vlasova et al. 2012) whereas a large number of putative lipolytic enzymes in those sequenced psychrotrophic bacteria such as Psychrobacter cryohalolentis, Psychrobacter phenylpyruvicus and Psychrobacter arcticus, remain to be characterized.

As shown in Fig. 5b, Est10 preferentially hydrolyzed short-chain, water-soluble substrates rather than long-chain, hydrophobic substrates indicating that Est10 is an esterase instead of a true lipase (Arpigny and Jaeger 1999).

Our test also suggests that Est10 retained considerable activity after being treated with 20 % ethanol, 30 % methanol and 30 % DMSO (Fig. 5a). Likewise, a previous study reported that solvent-tolerant esterase from Arthrobacter nitroguajacolicus had 120 % of the original activity after incubation with 30 % methanol for 20 min (Schütte and Fetzner 2007). In fact, the tolerance to DMSO and other organic solvents is attractive in organic synthesis (Klibanov 2001).

Interestingly, Est10 displayed significant cold adaptation in terms of the optimum activity at around 25 °C and the substantially high activity (55.0 %) at 0 °C (Fig. 3a). The optimum temperature 25 °C was similar to that of lipases from Streptomyces sp. CS628 (30 °C) (Mander et al. 2012), Psychrobacter cryohalolentis K5T (35 °C) (Novototskaya-Vlasova et al. 2012), and Pseudoalteromonas arctica (25 °C) (Al Khudary et al. 2010). Meanwhile, Est10 also showed a relatively higher residual activity (55.0 %) at 0 °C compared with most characterized cold-adapted lipolytic enzymes such as the esterase from Psychrobacter sp. Ant300 (Kulakova et al. 2004), and the lipase from Pseudomonas sp. 7323 (Zhang and Zeng 2008). Besides, Est10 was considerably stable below 40 °C (higher than the room temperature). These unique characteristics (low optimal temperature, high activity at 0 °C, high stability under room temperature) make Est10 potentially suitable not only in a few applications, e.g., pharmaceutical industries, agriculture and chemical industries (Panda and Gowrishankar 2005), but also in those particularly requiring cold conditions such as biosynthesis of fragile pharmaceutical compounds, avoiding adverse side-effects and conserving energy (Joseph et al. 2008).

According to the previous studies, compared with lipolytic enzymes from mesophilic or thermophilic organisms, cold-adapted esterases revealed some possible common features: a low ratio of Arg/(Arg + Lys), a low proportion of proline residues, a small hydrophobic core, a very small number of salt bridges and of aromatic–aromatic interactions (Joseph et al. 2008). Also, the number of proline and arginine residues (restricted backbone rotations, and formed multiple hydrogen bonds and salt bridges, respectively) was reduced, while clusters of small glycine residues were formed to offer localized chain mobility. A smaller number of ion pairs and weak of charge-dipole interactions in α helices were usually observed, which could contribute to the poor thermostability (Feller and Gerday 2003). For all the above-mentioned factors, the higher degree of flexibility was usually related to the decreased stability, i.e., a delicate balance between activity and stability could often be observed. As for Est10, due to the lack of structure information, only the amino acid composition was analyzed and the ratio of Arg/(Arg + Lys) was found to be 0.50, lower than that of another cold-adapted esterase rEst97 (0.56) from a metagenomic library and a thermophilic esterase Este1 (0.65) from thermal environmental metagenomic library (Byun et al. 2007; Fu et al. 2013). Also the total percentage (16.14 %) of small residues Gly (7.17 %) and Ala (8.97 %) is relatively high compared with the low proportion of Arg (3.14 %) and Pro (5.83 %), which could provide the increasing plasticity and flexibility of the structure. These primary structural characteristics may explain the cold adaptation of Est10, but to best illustrate this issue, a 3D structure is needed to obtain more accurate details like the intramolecular interactions.

Another noteworthy phenomenon was the enhanced activity and stability with increased NaCl concentration. Very few similar results were found among other cold-adapted lipolytic enzymes: the lipase from Pseudoalteromonas sp. NJ 70 retained only 2.6 % activity at 3 M NaCl (Wang et al. 2012), the esterase from Psychrobacter cryohalolentis K5T showed 179 % activity at 1.75 M NaCl (Novototskaya-Vlasova et al. 2012), but neither of them has reported activity at 4 or 5 M NaCl. As high salinity is likely to lower the water activity of the solution, it can mimic organic solvents. Since Est10 shows a certain degree of resistance to some organic solvents, it is quite possible to apply halotolerant Est10 in aqueous/organic and non-aqueous media (Sellek and Chaudhuri 1999). This unique property may also greatly enable massive production of Est10 without desalination.

In general, halophilic proteins come from halophilic organisms living in high saline environments such as hypersaline lakes, marine salterns and salt mines. From the perspective of the protein composition, halophilic enzymes have a large number of acidic amino acid residues at the protein surface, which acts to form a solvation shell keeping the protein surface hydrated (Fukuchi et al. 2003; Müller-Santos et al. 2009). Stabilization model is widely accepted and used to explain the halophilic adaptation, such as the tetrameric malate dehydrogenase (MalDH) from Haloarcula marismortui (Bonneté et al. 1993; Ebel et al. 1995; Zaccai et al. 1989). Another instance was that the salt-dependent esterase Hm EST from halophilic archaea Haloarcula marismortui showed optimal activity in the presence of 3 M KCl, due to the highly negatively charged protein surface by acidic amino acids (Müller-Santos et al. 2009). As for Est10, it does not come from halophilic bacteria, but from Psychrobacter sp., which could optimally thrive at 3 % NaCl (about 0.5 M, far more low than high salinity) and is unable to grow well at more than 8 % NaCl (about 1.37 M) (Maruyama et al. 2000). It is also not considered as a halophilic protein but as a salt-tolerant one because of its activity in the absence of NaCl (Fig. 4) (Madern et al. 2000). Est10 actually shares some features with a halophilic protein: (1) a higher percentage of the acidic amino acid (Asp + Glu; 11.21 %) than the basic amino acid (Arg + Lys; 6.28 %); (2) the occurrence of most of the 25 acidic amino acids (11 Asp and 14 Glu) on the protein surface, leading to a high negative surface electrostatic potential of Est10 (Fig. 6d) which is similar to the halophilic esterase Hm EST from Haloarcula marismortui (Müller-Santos et al. 2009) and may contribute to the stability of Est10 under high-salinity environment. Overall, Est10 is the first reported cold-adapted and non-halophilic esterase with considerable activity under high salinity (81.7 % activity at 5 M NaCl), though the detailed mechanism needs further research.

In conclusion, the esterase Est10 from Psychrobacter pacificensis has been proved to be a cold-adapted esterase and shows increased activity and extreme stability at a high concentration of salt, a good tolerance of organic solvents, and enhanced activity after pre-incubation with a low concentration of non-ionic detergents. The enzyme is also capable of hydrolyzing broad p-nitrophenyl esters (C2–C12). These unique features may enable Est10 to be a good candidate for further industrial applications under extreme conditions (low temperatures, high salt, high content of organic solvents).