Puriﬁcation and Crystallographic Analysis of a Novel Cold-Active Esterase ( Ha Est1) from Halocynthiibacter arcticus

: This report deals with the puriﬁcation, characterization, and a preliminary crystallographic study of a novel cold-active esterase ( Ha Est1) from Halocynthiibacter arcticus . Primary sequence analysis reveals that Ha Est1 has a catalytic serine in G-x-S-x-G motif. The recombinant Ha Est1 was cloned, expressed, and puriﬁed. SDS-PAGE and zymographic analysis were carried out to characterize the properties of Ha Est1. A single crystal of Ha Est1 was obtained in a solution containing 10% ( w / v ) PEG 8000/8% ethylene glycol, 0.1 M Hepes-NaOH, pH 7.5. Diffraction data were collected to 2.10 Å resolution with P2 1 space group. The ﬁnal R merge and R p.i.m values were 7.6% and 3.5% for 50–2.10 Å resolution. The unit cell parameters were a = 35.69 Å, b = 91.21 Å, c = 79.15 Å, and β = 96.9 ◦ .


Introduction
Microbial esterase catalyzes the reaction of formation and hydrolysis of chemical bonds between hydroxyl and carboxylic acid groups, which could be used in the preparation of many biological products such as foods, flavors, cosmetics, drugs, and agrochemicals [1][2][3]. In addition, they are employed in the degradation of carbamates, pesticides, polymer-based plastics, and industrial wastes. These enzymes share the characteristic α/β hydrolase fold, conserved catalytic triad of Ser-His-Asp/Glu, catalytic strategies, substrate specificities, and a lack of cofactors. It has been shown that esterases perform catalytic reactions via a nucleophilic attack on the substrates [1,[4][5][6].
Cold-active enzymes show relatively high activity at low temperatures compared to their mesophilic or thermophilic proteins [7,8]. Due to high demands, there have been a lot of studies to characterize cold-active enzymes [9]. To date, cold-active esterases were identified from Paenibacillus sp. [10], Pseudomonas mandelii [11], Lactobacillus plantarum [12], and Bacillus halodurans [13]. However, limited information is still available on the structure and function of these enzymes [7][8][9].
Halocynthiibacter arcticus is a rod-shaped Gram-negative bacteria from the Arctic region, which could be a valuable resource for biotechnological applications [14,15]. Specifically, structural information of enzymes from this bacterium is largely unknown. Here, we report the identification, purification, and preliminary crystallographic analysis of a novel cold-active esterase (HaEst1) from H. arcticus. The recombinant enzyme was purified, characterized, and crystallized for structural studies. Considering that cold-active esterases have great importance for industrial applications, HaEst1 could be an interesting industrial enzyme with cold-active properties.

Chemicals and Columns
Nucleic acid modifying enzymes and DNA purification kits were purchased from New England BioLabs (Ipsdwich, MA, USA) and Intron Korea (Daejon, Korea). PD-10 column for dialysis and His-tag affinity column for purification were obtained from GE Korea (Seoul, Korea). All other reagents of high purity grade were obtained from Sigma-Aldrich (Goyang-si, Korea).

Gene Cloning and Protein Purification
Microbial culture of Halocynthiibacter arcticus (Korea Collection of Type Cultures, KCTC 42129) and purification of its chromosomal DNA were carried out as previously described [16]. The open reading frame of the HaEst1 gene was amplified and cloned into a pET-21a vector (pET-HaEst1) using the following primers with NheI and XhoI (forward primer: 5 -GTAACCGCTAGCATGACAGACCCACAG-3 , and reverse primer: 5 -GCTGACTCGAGTCAGAATTTCGCCCG-3 ). E. coli BL21(λDE3) cells were transformed and grown in LB medium at 18 • C with 1 mM isopropyl-β-D-1-thiogalactoside (IPTG) induction. Then, bacterial cells were centrifuged and resuspended in a cell disruption buffer (20 mM Tris-HCl pH 8.5, 150 mM NaCl, 20 mM imidazole). Following cell lysis by ultrasonication, supernatants were loaded onto a His-Trap affinity column. HaEst1 was eluted by a gradient of imidazole method (from 50 mM to 200 mM). The fractions were buffer-changed with PD-10 column. The final proteins were collected and stored at −20 • C.

Crystallization Method
For effective crystallization trials, purified HaEst1 (20.0 mg/mL) was used with crystallization kits including MCSG 1T~4T (Anatrace), JCSG-plus (Molecular Dimensions), and PGA Screen (Molecular Dimensions) [20,21]. The screening process was carried out using an automated crystallization robot (SPT Labtech, San Diego, CA, USA). Initial droplets contained 300 nL of protein solution mixed with 300 nL of reservoir solution by sitting-drop vapor-diffusion method in a 96-well plate. Various crystals of HaEst1 appeared in several conditions within a week. Diffraction-quality single crystals of HaEst1 were observed under JCSG-Plus #16 condition of 10% (w/v) PEG 8000/8% ethylene glycol, 0.1 M Hepes-NaOH, pH 7.5.

Data Collection and Processing
A single crystal of HaEst1 was removed and transferred to a cryo-protectant solution. After gentle soaking, the crystal was mounted on a synchrotron facility at beamline 5C of the Pohang Light Source (PAL, Pohang, Korea). X-ray diffraction data of HaEst1 were collected at 100 K using the Eiger X 9M detector (Dectris, Switzerland). For complete X-ray diffraction data collection, the cryo-cooled crystal was rotated throughout 360 • rotation with 1 • oscillation per frame. Finally, collected X-ray data were processed and indexed using HKL2000 (see Table 1). Sequences of HaEst1 and other enzymes were obtained from a public NCBI server, and multiple sequence alignment was prepared using ESPript.

Results and Discussion
An open reading frame encoding a novel cold-active esterase (HaEst1, locus tag: WP_039000957, 756 bp) was detected and obtained from the H. arcticus chromosome. HaEst1 has 252 amino acids with a pI of 5.21. Multiple sequence alignments of HaEst1 with three related proteins indicated that HaEst1 showed significant sequence identities with a putative hydrolase from Agrobacterium vitis (3LLC, 40.0%), a new family of carboxyl esterase with an OsmC domain from Rhodothermus marinus (5CML, 19.8%) [22], and a cinnamoyl esterase from Lactobacillus johnsonii LJ0536 (3PF8, 20.6%) [23] (Figure 1). Interestingly, HaEst1 showed substantial sequence similarity to an alpha/beta hydrolase domain-containing protein 10 (ABHD10) from Mus musculus (6NY9, 23.8%). Catalytic triad of Ser-His-Asp as well as G-x-S-x-G motif were also conserved in these two proteins. This ABHD10 was recently shown to be an S-depalmitoylase affecting reduction/oxidation homeostasis [24].
Highly conserved catalytic residues of Ser 101 , Asp 198 , and His 228 were identified, with Ser 101 located in a typical GXSXG motif. Sequence analysis revealed that HaEst1 has a high number of small amino acids such as Gly (11.9%) and Ala (9.9%). In addition, high percentages of Leu (9.1%) and Thr (7.9%) were also observed. The percentage of acidic amino acids (Asp + Glu) was larger than that of basic amino acids (Arg + Lys).
The recombinant HaEst1 was purified using an immobilized His-tag metal-binding column (Figure 2A). The enzymatic activity of HaEst1 was examined using 4-methylumbelliferyl (4-MU) acetate. As shown in Figure 2B, strong fluorescence due to hydrolysis reaction was observed for 4-MU acetate, although little hydrolysis reaction was carried out for control or 4-MU phosphate. The recombinant HaEst1 was purified using an immobilized His-tag metal-binding column (Figure 2A). The enzymatic activity of HaEst1 was examined using 4methylumbelliferyl (4-MU) acetate. As shown in Figure 2B, strong fluorescence due to hydrolysis reaction was observed for 4-MU acetate, although little hydrolysis reaction was carried out for control or 4-MU phosphate. The diffraction-quality crystals grew to final dimensions of 0.6 × 0.5 × 0.2 mm within three days at 297K (Figure 3), which were transferred to a paratone oil, cryo-protectant solution. The diffraction data set of HaEst1 was indexed to P21 space group with unit cell parameter of a = 35.69 Å , b = 91.21 Å , c = 79.15 Å , and  = 96.9°. The final data were processed using HKL2000 to 2.10 Å resolution with 99.9% completeness. The final Rmerge  The recombinant HaEst1 was purified using an immobilized His-tag metal-binding column (Figure 2A). The enzymatic activity of HaEst1 was examined using 4methylumbelliferyl (4-MU) acetate. As shown in Figure 2B, strong fluorescence due to hydrolysis reaction was observed for 4-MU acetate, although little hydrolysis reaction was carried out for control or 4-MU phosphate. The diffraction-quality crystals grew to final dimensions of 0.6 × 0.5 × 0.2 mm within three days at 297K (Figure 3), which were transferred to a paratone oil, cryo-protectant solution. The diffraction data set of HaEst1 was indexed to P21 space group with unit cell parameter of a = 35.69 Å , b = 91.21 Å , c = 79.15 Å , and  = 96.9°. The final data were processed using HKL2000 to 2.10 Å resolution with 99.9% completeness. The final Rmerge The diffraction-quality crystals grew to final dimensions of 0.6 × 0.5 × 0.2 mm within three days at 297 K (Figure 3), which were transferred to a paratone oil, cryo-protectant solution. The diffraction data set of HaEst1 was indexed to P2 1 space group with unit cell parameter of a = 35.69 Å, b = 91.21 Å, c = 79.15 Å, and β = 96.9 • . The final data were processed using HKL2000 to 2.10 Å resolution with 99.9% completeness. The final R merge and R p.i.m values were 8.1% and 3.5% for 50-2.10 Å resolution. The final data collection statistics are summarized in Table 1. Assuming two molecules of HaEst1 per asymmetric unit, Matthews coefficient (V M ) and solvent content were calculated to be 2.36 Å 3 /Da and 47.9% [25]. and Rp.i.m values were 8.1% and 3.5% for 50-2.10 Å resolution. The final data collection statistics are summarized in Table 1. Assuming two molecules of HaEst1 per asymmetric unit, Matthews coefficient (VM) and solvent content were calculated to be 2.36 Å 3 /Da and 47.9% [25]. We tried to solve the HaEst1 structure using the molecular replacement method usingMOLREP [26]. The crystal structure of a putative hydrolase from Agrobacterium vitis (PDB code 3LLC) was used as a search model for the cross-rotation search. For the crossrotation function calculation, we used data in the resolution range of 39.44-2.41 Å . The results of the cross-rotation showed that the highest peak height was above 7.49σ. The highest peak solution of the rotation function was used for the following translation function search. The solution model also gave a strong correlation coefficient value (above 0.52) in the translation function. Rigid body refinements and individual restrained Bfactor refinements were performed using REFMAC5 [27]. After these refinement steps, an interpretable electron density map was calculated, as shown in Figure 4. Model building and further refinement are now underway. Furthermore, substrate or product bound HaEst1 structure determination with site-directed mutagenesis experiments will be performed. Thus, a detailed structural analysis of HaEst1 and protein engineering results will be published in an upcoming research paper.  We tried to solve the HaEst1 structure using the molecular replacement method usingMOLREP [26]. The crystal structure of a putative hydrolase from Agrobacterium vitis (PDB code 3LLC) was used as a search model for the cross-rotation search. For the cross-rotation function calculation, we used data in the resolution range of 39.44-2.41 Å. The results of the cross-rotation showed that the highest peak height was above 7.49σ. The highest peak solution of the rotation function was used for the following translation function search. The solution model also gave a strong correlation coefficient value (above 0.52) in the translation function. Rigid body refinements and individual restrained Bfactor refinements were performed using REFMAC5 [27]. After these refinement steps, an interpretable electron density map was calculated, as shown in Figure 4. Model building and further refinement are now underway. Furthermore, substrate or product bound HaEst1 structure determination with site-directed mutagenesis experiments will be performed. Thus, a detailed structural analysis of HaEst1 and protein engineering results will be published in an upcoming research paper. and Rp.i.m values were 8.1% and 3.5% for 50-2.10 Å resolution. The final data collection statistics are summarized in Table 1. Assuming two molecules of HaEst1 per asymmetric unit, Matthews coefficient (VM) and solvent content were calculated to be 2.36 Å 3 /Da and 47.9% [25]. We tried to solve the HaEst1 structure using the molecular replacement method usingMOLREP [26]. The crystal structure of a putative hydrolase from Agrobacterium vitis (PDB code 3LLC) was used as a search model for the cross-rotation search. For the crossrotation function calculation, we used data in the resolution range of 39.44-2.41 Å . The results of the cross-rotation showed that the highest peak height was above 7.49σ. The highest peak solution of the rotation function was used for the following translation function search. The solution model also gave a strong correlation coefficient value (above 0.52) in the translation function. Rigid body refinements and individual restrained Bfactor refinements were performed using REFMAC5 [27]. After these refinement steps, an interpretable electron density map was calculated, as shown in Figure 4. Model building and further refinement are now underway. Furthermore, substrate or product bound HaEst1 structure determination with site-directed mutagenesis experiments will be performed. Thus, a detailed structural analysis of HaEst1 and protein engineering results will be published in an upcoming research paper.