Crystal structure of the cold-adapted haloalkane dehalogenase DpcA from Psychrobacter cryohalolentis K5

The structure of the haloalkane dehalogenase DpcA from the psychrophilic bacterium Psychrobacter cryohalolentis K5, an attractive enzyme for biotechnological applications, was solved at the atomic resolution of 1.05 Å. The enzyme possesses main and slot tunnels with the shortest lengths in comparison with other haloalkane dehalogenases. Structural comparisons show major differences in the region of the α4 helix of the cap domain, which is one of the determinants of the properties of the tunnels. The structural information on DpcA establishes a basis for understanding its catalytic properties and can guide the modification of this cold-adapted enzyme for various biotechnological applications.

Abstract

Haloalkane dehalogenases (HLDs) convert halogenated aliphatic pollutants to less toxic compounds by a hydrolytic mechanism. Owing to their broad substrate specificity and high enantioselectivity, haloalkane dehalogenases can function as biosensors to detect toxic compounds in the environment or can be used for the production of optically pure compounds. Here, the structural analysis of the haloalkane dehalogenase DpcA isolated from the psychrophilic bacterium Psychrobacter cryohalolentis K5 is presented at the atomic resolution of 1.05 Å. This enzyme exhibits a low temperature optimum, making it attractive for environmental applications such as biosensing at the subsurface environment, where the temperature typically does not exceed 25°C. The structure revealed that DpcA possesses the shortest access tunnel and one of the most widely open main tunnels among structural homologs of the HLD-I subfamily. Comparative analysis revealed major differences in the region of the α4 helix of the cap domain, which is one of the key determinants of the anatomy of the tunnels. The crystal structure of DpcA will contribute to better understanding of the structure–function relationships of cold-adapted enzymes.

1. Introduction  

Globally, technological progress has led to the spread and accumulation of hazardous compounds in water and soil. The more technologically active society is, the more humans search for new possibilities and strategies to protect the environment and eliminate toxic compounds. One such strategy involves enzymes that are capable of bioremediation. Microbial haloalkane dehalogenases (HLDs; EC 3.8.1.5) can be used to eliminate toxic haloalkanes, which are used as refrigerants, fumigants, herbicides and pesticides, as industrial solvents, as intermediates in various synthetic processes, as flame retardants and so on (Belkin, 1992 ▸). HLDs belong to the superfamily of α/β-hydrolases (Ollis et al., 1992 ▸). Based on phylogenetic analysis, HLDs can be divided into three subfamilies, HLD-I, HLD-II and HLD-III, which differ in the composition of the catalytic pentad and the anatomy of the cap domain (Chovancová et al., 2007 ▸). In the ESTHER database (Lenfant et al., 2013 ▸) HLD-I and HLD-III subfamilies are merged together due to their sister-group relationships and denoted as HLD1; HLD-II is denoted as HLD2. HLDs catalyze the hydrolytic cleavage of a carbon–halogen bond in a broad range of halogenated aliphatic compounds via an S_N2 mechanism. The cleavage step is followed by the formation of a covalent alkyl-enzyme intermediate, which is subsequently hydrolyzed by a molecule of catalytic water, yielding the corresponding alcohol, a halide anion and a proton (Janssen, 2004 ▸). HLDs have been used in bio­remediation, biosensing, the production of optically pure compounds and cellular imaging (Koudelakova et al., 2013 ▸; Nagata et al., 2015 ▸). However, the effective use of biocatalysts in these applications requires enzymes with high catalytic efficiency under the desired process conditions, which may include various temperatures, pH values and pressures and/or high salinity (Adrio & Demain, 2014 ▸; Govardhan, 1999 ▸; Schmid et al., 2001 ▸). Enzymes acting under extreme conditions can be isolated from extremophilic organisms (Gomes & Steiner, 2004 ▸; Adams et al., 1995 ▸; Niehaus et al., 1999 ▸; Russell, 1998 ▸).

The HLD DpcA was isolated from the Gram-negative psychrophilic bacterium Psychrobacter cryohalolentis K5. DpcA exhibits the highest activity at 25°C and retains more than 25% of its activity at 5°C (Drienovska et al., 2012 ▸). This temperature optimum is much lower than those of other characterized HLDs, which generally have optimal activities at temperatures ranging from 40 to 50°C. DpcA possesses high enantioselectivity towards α-brominated esters and has a very narrow substrate specificity, with a preference for brominated propanes, butanes to hexanes (Drienovska et al., 2012 ▸). Its high activity at low temperatures and narrow substrate specificity make DpcA attractive for various biotechnological applications, particularly bioremediation and biosensing in a sub­surface environment. To gain deeper insight into the biochemical properties of DpcA, we performed X-ray structural analysis of the enzyme.

2. Materials and methods  

2.1. Gene synthesis, protein expression and purification  

The dpcA gene was artificially synthesized (Mr. Gene, Regensburg, Germany). The codon usage was adapted to the codon bias of Escherichia coli genes. The dpcA gene was subcloned into the expression vector pET-21b (Novagen, Madison, Wisconsin, USA) using the NdeI and BamHI restriction endonucleases (Takara, Japan) and T4 DNA ligase (Promega, Madison, USA). To overproduce C-terminally His-tagged DpcA in E. coli BL21(DE3) ArcticExpress cells, the cells containing the plasmid were cultured in Luria–Bertani medium at 37°C. When the culture reached an optical density of 0.6 at a wavelength of 600 nm, the temperature was lowered to 15°C and enzyme expression was induced with 0.5 mM IPTG. The cells were cultured overnight and were then harvested and disrupted by sonication using a Soniprep 150 ultrasonic disintegrator (Sanyo Gallenkamp PLC, Lough­borough, England). The supernatant was collected after centrifugation at 100 000g for 1 h. The crude extract was purified on a HiTrap Chelating HP 5 ml column charged with Ni²⁺ ions (GE Healthcare, Uppsala, Sweden). The His-tagged enzyme was bound to the resin in the presence of equilibration buffer (20 mM potassium phosphate pH 7.5, 0.5 M sodium chloride, 10 mM imidazole). Unbound and nonspecifically bound proteins were washed out with buffer containing 37.5 mM imidazole. The target enzyme was eluted with buffer containing 300 mM imidazole. The active fractions were pooled and were dialyzed overnight against 50 mM Tris–HCl pH 7.5. The enzyme was stored at 4°C in 50 mM Tris–HCl buffer prior to analysis. Macromolecule-production information is summarized in Table 1 ▸.

Table 1. Macromolecule-production information.

Source organism	P. cryohalolentis K5
DNA source	Synthetic DNA
Forward primer	GCGCATATGAAAATTCTGCGC
Reverse primer	GCGGGATCCTTAATGATGATG
Cloning vector	pMA
Expression vector	pET-21b
Expression host	E. coli BL21(DE3)
Complete amino-acid sequence of the construct produced	MKILRTPDSRFANLPDYNFDPHYLMVDDSEDSELRVHYLDEGPRDADPVLLLHGEPSWCYLYRKMIPILTAAGHRVIAPDLPGFGRSDKPASRTDYTYQRHVNWMQSVLDQLDLNNITLFCQDWGGLIGLRLVAENPDRFARVAAGNTMLPTGDHDLGEGFRKWQQFSQEIPQFHVGGTIKSGTVTKLSQAVIDAYNAPFPDESYKEGARQFPLLVPSTPDDPASENNRAAWIELSKWTKPFITLFSDSDPVTAGGDRIMQKIIPGTKGQAHTTIANGGHFLQEDQGEKVAKLLVQFIHDNPRHHHHHH

2.2. Crystallization, data collection and structure determination  

Crystallization of DpcA and data collection and processing for the crystal in a primitive monoclinic space group have been described previously (Tratsiak et al., 2013 ▸). X-ray diffraction data were collected to 1.05 Å resolution on the MX14.2 beamline operated by the Helmholtz-Zentrum Berlin (HZB) at the BESSY II electron-storage ring, Berlin-Adlershof, Germany (Mueller et al., 2015 ▸) at a wavelength of 0.978 Å. Crystal parameters and data-collection statistics are summarized in Table 2 ▸.

Table 2. Data-collection and refinement statistics.

Values in parentheses are for the highest resolution shell.

Crystal parameters and data-collection statistics†
Space group	P21
a, b, c (Å)	41.3, 79.4, 43.5
α, β, γ (°)	90.0, 95.0, 90.0
Wavelength (Å)	0.978
Resolution (Å)	50–1.05 (1.09–1.05)
No. of unique reflections	120315 (7363)
Multiplicity	5.7
Completeness (%)	92.2 (56.2)
R merge ‡ (%)	5.0 (30.7)
Average I/σ(I)	39.7 (3.4)
Overall B factor from Wilson plot (Å2)	16.72
Refinement statistics
Resolution range (Å)	20.2–1.05 (1.08–1.05)
Completeness (%)	92
No. of reflections, working set	114202
No. of reflections, test set	4716
Final R cryst § (%)	0.110
Final R free ¶ (%)	0.132
R.m.s. deviations
Bonds (Å)	0.012
Angles (°)	1.634
No. of non-H atoms
Protein	2465
Water	576
Heterogen atoms	28
Mean B value (Å2)	8.67
Ramachandran plot statistics††
Most favored (%)	95.35
Allowed (%)	4.65
PDB code	6f9o

^†As published in Tratsiak et al. (2013 ▸).

^‡ R _merge = , where I _i(hkl) is an individual intensity of the ith observation of reflection hkl and 〈I(hkl)〉 is the average intensity of reflection hkl with summation over all data.

^§ R = , where F _obs and F _calc are the observed and calculated structure factors, respectively.

^¶ R _free is equivalent to the R value but is calculated for 5% of the reflections that were chosen at random and omitted from the refinement process (Brünger, 1992 ▸).

^††As determined using the MolProbity server (Chen et al., 2010 ▸).

The structure of DpcA was determined by molecular replacement with MOLREP (Vagin & Teplyakov, 2010 ▸) from the CCP4 software suite (Winn et al., 2011 ▸) using the coordinates of HLD from Xanthobacter autotrophicus (DhlA; PDB entry 1b6g; Ridder et al., 1999 ▸) as a search model. Structure refinement of DpcA was carried out with REFMAC5 (Murshudov et al., 2011 ▸). The model was initially refined with isotropic atomic displacement parameters (ADPs); H atoms were later added in riding positions. Anisotropic models of ADPs were used in the final rounds of refinement. Manual building steps were conducted in Coot (Emsley et al., 2010 ▸). The model was optimized using the PDB-REDO web server (Joosten et al., 2014 ▸). The refinement statistics are shown in Table 2 ▸. The quality of the model was validated using the wwPDB validation server (Berman et al., 2003 ▸) and MolProbity (Chen et al., 2010 ▸; Hintze et al., 2016 ▸). Atomic coordinates and experimental structure factors have been deposited in the Protein Data Bank as entry 6f9o. All figures were prepared using PyMOL v.1.5 (Schrödinger). The structure-similarity search of the PDB was performed with DALI (Holm & Rosenström, 2010 ▸).

2.3. Analysis of tunnels in protein structures  

Crystal structures of the following HLD-I members were obtained from the PDB for tunnel analysis: DhlA (PDB entry 1b6g; Ridder et al., 1999 ▸), DppA (PDB entry 2xt0; Hesseler et al., 2011 ▸), DmrA (PDB entry 4mj3; Fung et al., 2015 ▸) and DccA (PDB entry 5esr; Carlucci et al., 2016 ▸). The newly obtained structure was used for DpcA (PDB entry 6f9o). All ligands and water molecules were removed, and all structures were aligned with the DpcA structure using PyMOL v.1.5 (Schrödinger). The tunnels were analyzed with the standalone version of CAVER 3.0.1 (Pavelka et al., 2016 ▸). Every atom was approximated by 12 spheres with radii corresponding to the smallest atom present in the analyzed protein structure. The starting point was specified by the catalytic nucleophile (Asp123 in DpcA) and the adjacent halide-stabilizing residue (Trp124 in DpcA). Transport tunnels were identified using a probe radius of 0.7 Å. Redundant tunnels were automatically removed from each structure, and tunnels were clustered using a threshold of 3.5. The active-site volumes were estimated by CASTp (Dundas et al., 2006 ▸).

2.4. Docking studies  

Computational docking studies were performed using AutoDock Vina v.1.1.2 (Trott & Olson, 2010 ▸) as implemented in UCSF Chimera 1.3.1 (Pettersen et al., 2004 ▸). The crystal structure of DpcA without chloride ions, water molecules, alternative conformations of amino-acid side chains and other ligands was used as the receptor. 1-Bromohexane was used as a substrate for docking experiments. The substrate was built manually and the geometry was optimized using UCSF Chimera 1.3.1 (Pettersen et al., 2004 ▸). Hydrogens and charges were added to the ligand and the receptor. The docking searches were performed with an exhaustiveness of 8, ten modes and energy ranges of 3 kcal mol⁻¹. Searches were carried out over the whole molecule, allowing the ligand to be flexible. The results were examined based on the involvement of the ligand interactions in the active site of the DpcA structure and were visualized in PyMOL v.1.5 (Schrödinger).

3. Results and discussion  

3.1. Overall structure of DpcA  

The structure of DpcA was determined by molecular replacement and refined to the atomic resolution of 1.05 Å (Table 2 ▸). The crystal belonged to a primitive monoclinic space group and contained 40% solvent with one protein molecule in the asymmetric unit. All protein residues (1–309) were modeled into the electron-density map; the extra C-terminal residues (304–309) represent the C-terminal histidine tag. The side chains of 21 residues were modeled in alternative conformations, including that of the catalytic nucleophile Asp123. Electron-density maps also suggest a minor alternative main-chain conformation for regions 155–159 (forming a loop connecting β6 of the main domain and α4 of the cap domain) and 162–173 (forming the α4 helix). However, the quality of the maps did not allow modeling of these alternative conformations. Nonprotein electron density was explained by 576 water molecules, two chloride ions, one of which is in the enzyme active site, one sodium ion and five polyethylene glycol molecules.

The overall structure of the DpcA monomer, representing the biologically active unit, is depicted in Fig. 1 ▸(a). The structure of DpcA is similar to those of related members of the HLD-I subfamily and is composed of two domains: the main domain and the cap domain. The highly conserved main domain (residues 1–154 and 221–309) consists of an eight-stranded β-sheet with β2 lying in an antiparallel orientation with respect to the direction of the β-sheet. The β-sheets are surrounded by six α-helices (two and four on each side). The cap domain (residues 155–220), inserted between β-strand β6 and α-helix α8 of the main domain, acts as a lid covering the core domain. It is formed by four α-helices and five connecting loops.

Figure 1.

Crystal structure of the haloalkane dehalogenase DpcA from P. cryohalolentis K5. (a) The overall structure is shown in cartoon representation. The cap domain is colored cyan and the main domain red; the chloride anion bound in the active site is shown as a gray sphere. (b) Detail of the active site of DpcA. The catalytic pentad is represented by sticks; the chloride ion is shown as a gray sphere with coordinating interactions represented by black dashed lines. The 2F _o − F _c electron-density map for ions and interacting residues contoured at 1.5σ is shown in gray. (c) Transport tunnels identified in DpcA are shown in sphere representation; the main tunnel is colored blue and the slot tunnel green. (d) Predicted docking of 1-bromohexane as a substrate in the active site of DpcA; the molecule of 1-bromohexane is represented as green sticks.

3.2. DpcA active site  

The active site of DpcA is buried in a hydrophobic cavity located between the main and the cap domains. It contains a catalytic pentad typical of the HLD-I subfamily (Chovancová et al., 2007 ▸): Asp123, Trp124, Trp164, Asp250 and His270 (Fig. 1 ▸ b). The nonprotein electron density in the active site was interpreted as water molecules and a chloride anion. The chloride anion occupies the canonical halide-binding site and interacts with N^∊1 of Trp124 and Trp164, the two halide-stabilizing residues, at distances of 3.31 and 3.30 Å, respectively. Further coordination is provided by water molecule HOH106 at a distance of 3.07 Å (Fig. 1 ▸ b).

The superior quality of electron-density maps at atomic resolution allowed us to identify two alternative conformations for the side chain of the catalytic nucleophile Asp123 (Fig. 1 ▸ b). These conformations differ only in a small shift of the terminal carboxyl-group position by 0.54 Å (distance of C^γ atoms). Both alternative conformations retain a similar pattern of polar interactions with neighboring residues and water molecules: Asp123 O^δ1 forms hydrogen bonds to N^∊2 of the catalytic base His280 and to water molecule HOH106. Asp123 O^δ2 forms a hydrogen bond to the main-chain amino group of the halide-stabilizing Trp124 N (Fig. 1 ▸ b).

DpcA is the first HLD-I enzyme in which two alternative conformations of the nucleophile residue were observed. The individual conformations of Asp123 in DpcA are not unique; similar conformations of nucleophile residues have previously been observed for other members of HLD-I with known structures. The major conformation A (with occupancy of 0.6) is similar to the conformation of Asp124 in the DmrA structure. Conformation B (with occupancy of 0.4) has the same position as in the DccA, DhlA and DppA structures. Both conformations of Asp123 in DpcA are probably competent to bind substrate. The reason for the simultaneous presence of both conformations in DpcA might be a higher flexibility of the active site of DpcA. High conformational flexibility of the active site has been reported as a structural adaptation in psychrophilic enzymes (Georlette et al., 2003 ▸).

DpcA contains two tunnels connecting the buried enzyme active site to the surrounding solvent, and these were analyzed with CAVER (Chovancova et al., 2012 ▸; Pavelka et al., 2016 ▸; Fig. 1 ▸ c). The main tunnel is formed by Glu55, Gln122, Asp123, Trp124, Gly125, Leu127, Asn147, Thr148, Met149, Gly160, Phe161, Trp164, Thr179, Gly183, Phe212, Pro213, Val216, Asp250, Val252, Thr253, His280 and Phe281; the bottleneck residues are Glu55, Gln122, Asp123, Trp124, Gly125, Leu127, Asn147, Thr148, Met149, Phe161, Val216, Val252, Thr253 and His280. The smaller and shorter slot tunnel that allows solvent entry to the catalytic site is formed by Glu55, Gln122, Asp123, Trp124, Gly125, Leu127, Asn147, Thr148, Met149, Pro151, Phe161, Val216, Phe246, Asp250, Val252, Thr253, Ala254, Gly256 and His280; the bottleneck residues are Leu127, Asn147, Thr148, Met149, Phe246, Val252 and Thr253.

Attempts to obtain a structure with substrate bound to the active site failed. Although numerous co-crystallization as well as crystal-soaking experiments with various substrates were performed, no interpretable electron-density map for a ligand was located in the active site of the enzyme. We thus used the DpcA crystal structure in docking experiments. The selected substrate, 1-bromohexane, was placed into the enzyme active site with a docking energy of −4.7 kcal mol⁻¹ (Fig. 1 ▸ d). The 1-bromohexane molecule is well coordinated by two halide-stabilizing residues, Trp124 and Trp164. The bromide is coordinated by N^∊1 of Trp124 at a distance of 3.65 Å and N^∊1 of Trp164 at a distance of 3.80 Å. The side chain of residue Asp123 is in a good position for the nucleophilic attack: the distance between the C1 atom of the substrate and O^δ2 of Asp123 is 3.28 and 3.60 Å for conformations A and B, respectively.

3.3. Comparison of DpcA with structural homologs  

A structural comparison with structures deposited in the Protein Data Bank using DALI (Holm & Rosenström, 2010 ▸) identified four HLD-I subfamily members as structural homologs of DpcA: DhlA from Xanthobacter autotrophicus GJ10 (Ridder et al., 1999 ▸; PDB entry 1b6g; r.m.s.d. of 1.6 Å for superposition of 294 residues), DccA from Caulobacter crescentus (Carlucci et al., 2016 ▸; PDB entry 5esr; r.m.s.d. of 1.3 Å for superposition of 304 residues), DppA from Plesiocystis pacifica (Hesseler et al., 2011 ▸; PDB entry 2xt0; r.m.s.d. of 1.5 Å for superposition of 293 residues) and DmrA from Mycobacterium rhodesiae JS60 (Fung et al., 2015 ▸; PDB entry 4mj3; r.m.s.d. of 0.7 Å for superposition of 303 residues). The superimposition of the HLD-I structures with DpcA revealed that the main structural differences occur in the cap domain, mostly in the position of the α4 helix (Fig. 2 ▸ a).

Figure 2.

Structural comparison of the haloalkane dehalogenase DpcA with related HLD-I members. (a) The structure of DpcA (the main domain is in gray and the cap domain is in black) superposed with those of DhlA from X. autotrophicus GJ10 (PDB entry 1b6g, raspberry), DccA from C. crescentus (PDB entry 5esr, forest green), DppA from P. pacifica (PDB entry 2xt0, yellow) and DmrA from M. rhodesiae JS60 (PDB entry 4mj3, marine). (b) Comparison of the tunnels in HLD-I subfamily members. The main (blue) and slot (green) tunnels were identified with CAVER 3.0.1 (Pavelka et al., 2016 ▸). The orientation of the tunnels corresponds to the orientation of the HLD-I structures presented in (a).

DpcA is the first psychrophilic HLD-I enzyme with known structure; we thus looked for structural features underlying its cold adaptation. In general, compared with their mesophilic counterparts, psychrophilic enzymes are often characterized by higher flexibility, which allows better interaction with substrates. The tertiary structures of psychrophilic enzymes tend to have fewer stabilizing interactions, longer and more hydrophilic loops, a higher glycine content and a lower proline and arginine content (Tronelli et al., 2007 ▸).

The amino-acid compositions of the psychrophilic DpcA and 14 mesophilic HLDs have previously been compared (Drienovska et al., 2012 ▸) and showed no significant differences. The DpcA structure has mean values for Leu, Gly, Pro and Arg content and a slightly higher Asp content (9% in comparison to a maximum of 8% in 14 mesophilic HLDs). DpcA has an increased content of Lys and Asn, but it has a low Ala content in comparison to mesophilic HLDs from different subfamilies. Comparison of the crystal structure of DpcA with the structures of other mesophilic enzymes from the HLD-I family did not uncover major structural differences that could contribute to the cold adaptation of DpcA, such as longer hydrophobic loops. We did notice an increased flexibility of the cap domain, especially the regions 155–159 (forming a loop) and 162–173 (forming α4). The high B factors and high quality of the electron-density maps indicate high dynamic disorder in this region, and even the presence of a minor alternative conformation of helix α4, which was not modeled. Cap-domain helix α4 covers the active site of the enzyme and contains the halide-stabilizing residue Trp164. Flexibility in this region thus directly affects the active-site flexibility. It is in line with previous observations that adaptive mutations favoring active-site flexibility are located outside the catalytic center (Feller & Gerday, 2003 ▸).

We compared the active site, the volume of the main and slot tunnels, and the active-site cavity volume of DpcA with those of the four structural homologs (Fig. 2 ▸ b, Table 3 ▸).

Table 3. Comparison of the properties of the haloalkane dehalogenase DpcA with those of its closest structural homologs.

Enzyme	DpcA	DhlA	DccA	DppA	DmrA
PDB code	6f9o	1b6g	5esr	2xt0	4mj3
Resolution (Å)	1.05	1.15	1.5	1.95	1.7
Catalytic pentad	Asp123, Trp124, Trp164, Asp250, His280	Asp124, Trp125, Trp175, Asp260, His289	Asp123, Trp124, Trp163, Asp249, His279	Asp123, Trp124, Trp163, Asp249, His278	Asp123, Trp124, Trp164, Asp250, His280
Sequence identity to DpcA (%)	—	40	58	43	57
R.m.s.d. for superposition with DpcA (Å)/No. of aligned Cα atoms	—	1.6/294	1.3/304	1.5/293	0.7/303
Substrate preference	C3–C6	C2–C3	C3–C6	C3–C5	C4–C5
	Monosubstituted, disubstituted, brominated	Terminally substituted, chlorinated, brominated	Monosubstituted, disubstituted, brominated, chlorinated	Monosubstituted, brominated	Monosubstituted, disubstituted, brominated
	Drienovska et al. (2012 ▸)	Janssen et al. (1985 ▸, 1989 ▸)	Carlucci et al. (2016 ▸)	Hesseler et al. (2011 ▸)	Fung et al. (2015 ▸)
Active-site cavity volume† (Å3)	435	363	869	441	1063
Main tunnel radius/length‡ (Å)	1.7/8.0	0.7/15.4	1.5/10.8	1.3/9.8	1.9/9.1
Slot tunnel radius/length‡ (Å)	0.9/7.9	0.9/16	1.5/12.3	1.2/9.1	1.9/11.6

^†Active-site volumes were calculated with CASTp (Dundas et al., 2006 ▸) using a 1.4 Å radius probe; the calculation of cavity volumes with opening to the solvent is approximate and is dependent on the termination of the calculation at the protein surface.

^‡Average bottleneck radius and average tunnel length were determined using CAVER (Pavelka et al., 2016 ▸).

The active-site cavity volume of DpcA is 435 ų, which is comparable to that of DppA. DhlA has a smaller active-site cavity, while the cavities of DccA and DmrA are 2.0 and 2.5 times larger than that of DpcA, respectively. DpcA has the shortest and most open main tunnel (with an average length of 8.0 Å and an average bottleneck radius of 1.7 Å). The DpcA slot tunnel is the shortest among the HLD-1 representatives, and its radius is comparable to that of DhlA and 1.3–2 times narrower than those of DppA, DccA and DmrA.

4. Conclusions  

In this study, we performed an X-ray structural analysis of DpcA from P. cryohalolentis K5, the first biochemically characterized HLD of psychrophilic origin (Drienovska et al., 2012 ▸). This enzyme is a promising candidate for environmental applications such as biosensing at the subsurface environment (Bidmanova et al., 2010 ▸, 2016 ▸), as it exhibits maximal activity at 25°C and is specific towards longer brominated substrates such as 1-bromobutane, 1-bromohexane and 1,3-dibromopropane (Drienovska et al., 2012 ▸). Our atomic resolution DpcA structure enhances the available structural information on enzymes belonging to the HLD-I subfamily. The overall structural similarity of DpcA to other HLD-I enzymes is quite high (r.m.s.d. values of 0.7–1.6 Å). However, we identified substantial differences in the architecture of the active site and access tunnels. Compared with other structurally characterized HLD-I members, DpcA possesses tunnels with the shortest length. Interestingly, the main tunnel of DpcA is relatively open (1.7 Å), while its slot tunnel is very narrow (0.9 Å). This may indicate that the main tunnel can be explored by substrate and alcohol product molecules, while the other tunnel is used for halide ions or water molecules. Moreover, comparative analysis revealed major differences in the region of the α4 helix of the cap domain, which is one of the determinants of the properties of the tunnels. The short and open tunnel of DpcA and the medium-size volume of the active site explain the preference of DpcA for C3–C6 halogenated substrates, but it also might be owing to the cold adaptation of the enzyme. Better accessibility and a favorable substrate fit into the active site might help to improve catalysis at low temperatures.

In conclusion, the DpcA structural information reported here establishes a basis for understanding its enzymatic properties and may guide the modification of this cold-adapted enzyme for various biotechnological applications.

Supplementary Material

PDB reference: DpcA from Psychrobacter cryohalolentis K5, 6f9o

Funding Statement

This work was funded by Ministerstvo Školství, Mládeže a Tělovýchovy grants CZ.1.05/2.1.00/01.0024, CZ.1.05/2.1.00/01.0001, and CZ.02.1.01/0.0/0.0/15_003/0000441. Grantová Agentura České Republiky grant 17-24321S.