J. Microbiol. Biotechnol. (2016), 26(12), 2087–2097 http://dx.doi.org/10.4014/jmb.1608.08049 Research Article Review jmb

Cloning, Expression, and Characterization of a Cold-Adapted Shikimate from the Psychrophilic Bacterium Colwellia psychrerythraea 34H Wahyu Sri Kunto Nugroho1, Dong-Woo Kim1, Jong-Cheol Han2, Young Baek Hur2, Soo-Wan Nam3, and Hak Jun Kim1*

1Department of Chemistry, Pukyong National University, Busan 48547, Republic of Korea 2Southeast Sea Fisheries Research Institute, National Fisheries Research and Development Institute, Tongyeong 53085, Republic of Korea 3Department of Biotechnology and Bioengineering, Dong-Eui University, Busan 47340, Republic of Korea

Received: August 24, 2016

Revised: September 2, 2016 Most cold-adapted possess higher Km and kcat values than those of their mesophilic Accepted: September 5, 2016 counterparts to maximize the reaction rate. This characteristic is often ascribed to a high structural flexibility and improved dynamics in the . However, this may be less convincing to cold-adapted metabolic enzymes, which work at substrate concentrations near

First published online Km. In this respect, cold adaptation of a shikimate kinase (SK) in the shikimate pathway from September 23, 2016 psychrophilic Colwellia psychrerythraea (CpSK) was characterized by comparing it with a

*Corresponding author mesophilic Escherichia coli homolog (EcSK). The optimum temperatures for CpSK and EcSK Phone: +82-51-629-5587; activity were approximately 30°C and 40°C, respectively. The melting points were 33°C and Fax: +82-51-629-5583; 45°C for CpSK and EcSK, respectively. The ΔG (denaturation in the absence of denaturing E-mail: [email protected] H2O agent) values were 3.94 and 5.74 kcal/mol for CpSK and EcSK, respectively. These results indicated that CpSK was a cold-adapted . However, contrary to typical kinetic data,

CpSK had a lower Km for its substrate shikimate than most mesophilic SKs, and the kcat was not increased. This observation suggested that CpSK may have evolved to exhibit increased substrate affinity at low intracellular concentrations of shikimate in the cold environment. Sequence analysis and homology modeling also showed that some important salt bridges were lost in CpSK, and higher Arg residues around critical Arg 140 seemed to increase flexibility for catalysis. Taken together, these data demonstrate that CpSK exhibits characteristics of cold adaptation with unusual kinetic parameters, which may provide important insights into the cold adaptation of metabolic enzymes. pISSN 1017-7825, eISSN 1738-8872 Keywords: Cold adaptation, shikimate kinase, psychrophile, Michaelis-Menten constant, Copyright© 2016 by The Korean Society for Microbiology turnover number and Biotechnology

Introduction temperatures because the viscosity of the solution increases, and molecular motion decelerates [9]. However, Psychrophiles manage to survive and proliferate in cold psychrophiles can maintain normal metabolic reaction environments, such as polar regions, deep oceans, and rates, even at extremely low temperatures, partly through alpine regions [38, 49], and have a maximal growth the functions of cold-adapted proteins [9, 44, 51]. Many temperature typically lower than 20°C [38]. Psychrophiles studies have shown that psychrophiles have developed have been found in a diverse range of microbial categories, various cold adaptation strategies for their enzymes, including bacteria, eucarya, and archaea [38]. According to mainly increasing conformational flexibility at the expense the Arrhenius equation, reaction rates slow down at low of thermal stability, to confer catalytic competence at low

December 2016 ⎪ Vol. 26⎪ No. 12 2088 Nugroho et al.

temperatures [12, 28, 33, 35, 36, 52, 56, 59]. Such adaptation thermal- and urea-induced stability, sequence analysis, and strategies at the primary sequence level include a higher homology modeling with mesophilic E. coli SK. frequency of Gly residues, lower frequency of Pro residues, lower Arg/(Arg + Lys) ratio, reduced number of hydrogen Materials and Methods bonds and electrostatic interactions, decreased hydrophobic core, and an increased surface hydrophobicity [9, 44, 51, 55]. Cloning, Expression, and Purification of C. psychrerythraea and The enhanced structural flexibility in cold-adapted E. coli SKs The aroK genes (Accession Nos. AAZ26032 and YP_026215 for enzymes usually results in higher Km and kcat values, thereby maximizing the overall reaction rate [14, 44, 51]. C. psychrerythraea and E. coli, respectively) were amplified using Hence, most cold-adapted enzymes have higher K values oligonucleotide primers: 5’-CAGCCATATGATGGCAGAAAA m ACGT-3’ (forward), and 5’-GAATTCGGATCCTTAAAAGTCGAG than mesophilic counterparts [14, 44, 51]. However, this AC-3’ (reverse) for CpSK, and 5’-CAGCCATATGATGGCAGA trend may be not be valid in some cold-adapted enzymes, GAAACG-3’ (forward) and 5’-GAATTCGGATCCTTAGTTGCT in particular the ones that function at low or near K m TTCC-3’ (reverse) for EcSK. The amplified products were digested substrate concentrations, such as intracellular enzymes in with NdeI and BamHI, and ligated into the pET28b vector. E. coli metabolic pathways [1, 6, 19, 48, 58]. BL21(DE3) harboring either the CpSK- or EcSK-expressing plasmid To investigate cold adaptation of intracellular metabolic was incubated in 5 ml of LB containing 50 μg/ml kanamycin at enzymes, we cloned, expressed, and characterized shikimate 37°C overnight with shaking. The culture was transferred to 1 L of (SKs, encoded by aroK; E.C. 2.7.1.71) from fresh LB medium containing 50 μg/ml kanamycin. When the psychrophilic Colwellia psychrerythraea 34H (CpSK) and optical density (OD600) of the culture reached 0.6, 0.4 mM isopropyl- Escherichia coli (EcSK). SK is an essential enzyme catalyzing β-D-1-thiogalactopyranoside was added to the culture media to the conversion of shikimate to shikimate-3-phosphate and induce the SKs. It should be noted that the BL21(DE3)-CpSK ATP to ADP (Fig. 1). SK is the fifth enzyme in the seven- culture was then shifted to a 15°C shaker and further incubated for 6 h, while BL21(DE3)-EcSK was induced at 37°C for 4 h. The step essential shikimate pathway, which contributes to cells were harvested by centrifugation at 12,000 ×g for 30 min at the production of aromatic amino acid precursors and 4°C and cell pellets were stored at -70°C prior to use. The metabolites, such as folate cofactors, ubiquinone, and recombinant SKs were purified by immobilized-nickel ion affinity vitamins E and K [23]. Since the shikimate pathway is not chromatography. One gram of cell pellet was resuspended in present in mammals but only present in microorganisms 10 ml of binding buffer consisting of 100 mM Tris-HCl (pH 7.5), and plants, enzymes in this pathway, including SK, have 0.25 M NaCl, and 5 mM imidazole. Freshly prepared phenylmethane become a target for the development of antimicrobial sulfonyl fluoride in isopropanol was added up to 1 mM in the agents, herbicides, and antiparasitic agents [7, 22, 47, 53]. suspension. The cell suspension was then lysed by sonication, and Here, we report the cold adaptation of metabolically essential the lysates were clarified by centrifugation at 22,000 ×g for 10 min shikimate kinase from psychrophilic C. psychrerythraea by at 4°C. The supernatant was then mixed with His-Bind agarose comparing its optimal temperature, kinetic parameters, resin (Elpis, Korea) pre-equilibrated with binding buffer at 4°C for

Fig. 1. Coupled spectrophotometric assay for shkimate kinase.

J. Microbiol. Biotechnol. Cold-Adapted Shikimate Kinase 2089

30 min. The mixture was poured into the column and then while keeping the concentration of the other constant. washed with 2 volumes of washing buffer containing 100 mM Tris-HCl (pH 7.5), 0.25 M NaCl, and 60 mM imidazole. The Thermal- and Urea-Induced Denaturation recombinant SKs were recovered with an elution buffer The thermal denaturation curve for SKs was obtained by plotting containing 100 mM Tris-HCl (pH 7.5), 0.25 M NaCl, 500 mM changes in CD values at 220 nm as a function of temperature imidazole, and 5% glycerol. For circular dichroism (CD) and using a Chirascan Circular Dichroism Spectrometer (Applied fluorescence spectroscopy analysis, the buffer was exchanged to Photophysics Co., UK) equipped with a Peltier temperature 10 mM potassium phosphate (pH 7.4) with 5% glycerol by controller. A 200 μg/ml sample in 10 mM potassium phosphate ultrafiltration with an Ultra 15 Centrifugal filter (10,000 MWCO; buffer (pH 7.4) was placed in a quartz cuvette with a 2 mm Amicon, USA). Further experiments were carried out without pathlength (Hellma USA, Plainview, USA). Far-UV CD spectra removing the His-tags. were collected at 5°C, 7.5°C, 10°C, 12.5°C, 15°C, 20°C, 22.5°C, 25°C, 27.5°C, 30°C, 32.5°C, 35°C, 40°C, 45°C, 50°C, 55°C, 60°C, and Enzymatic Activity Assay 70°C with a 2 nm bandwidth. The sample was equilibrated for The enzymatic activity of SKs was assayed in the forward 5 min at each temperature. The temperature of the midpoint of the direction using a coupled assay (Fig. 1) [45]. In the coupled assay, transition, Tm, was calculated by fitting the CD values versus eventual oxidation of NADH to NAD+ was monitored at 340 nm temperature data on the basis of a least-squares analysis. through a series of (PK)- and lactate Urea denaturation of SKs was measured using a JASCO FP 6300 dehydrogenase (LDH)-catalyzed reactions. Since the two coupling spectrophotometer (Jasco, Japan). A 200 μg/ml sample of SK in enzymes functioned at approximately 25°C, we performed 10 mM potassium phosphate buffer (pH 7.4) was placed in a discontinuous coupled assays rather than continuous assays for cuvette with a 1 cm pathlength. The excitation wavelength was SKs to determine the optimum temperature for enzyme activity. 280 nm, and the emission wavelengths ranged from 300 to The SK reaction was carried out at various temperatures (5°C, 500 nm. Fluorescence spectra were collected at 25°C at varying 10°C, 15°C, 20°C, 25°C, 30°C, 35°C, 40°C, 50°C, and 60°C), and the urea concentrations (0.00, 0.20, 0.38, 0.57, 0.74, 0.91, 1.07, 1.23, 1.38, reaction product was then removed at different time points and 1.53, 1.67, 1.8, 1.94, 2.06, 2.19, 2.31, 2.42, 2.54, 2.65, 3.06, 3.51, 4.00, used for the coupled assay at 25°C. The assay mixture contained 4.51, 5.00, and 6.00 M) after a 10-min equilibration. The free 100mM Tris-HCl (pH 7.5), 25 mM shikimate, 25 mM ATP, and energy difference between the folded and unfolded states, ΔG, 15 mM MgCl . Each assay (100 μl) at different temperatures was and the free energy change of unfolding in H O, ΔG , were 2 2 H2O initiated by adding 1.25 μg of CpSK or EcSK and was stopped by estimated as described elsewhere [37]. adding 5 μl of 50% perchloric acid after 5, 10, or 20 min incubations. The reaction mixtures were then further incubated on ice for 3 min Sequence Alignment and Comparative Homology Modeling of and then were diluted with 400 μl of 100 mM Tris-HCl (pH 7.5). CpSK The recombinant SKs were removed by centrifugation of samples Multiple sequence alignments were performed using the T- at 12,000 ×g in an Amicon filter (10 kDa MWCO) for 3 min at 4°C. Coffee Web-based server [13] and rendered by ESPript [41]. A The flow-through fractions containing shikimate-3-phosphate and homology model of CpSK was generated using MODELLER ADP were neutralized using 8 μl of 30% KOH and used in the software with default options [57]. Its templates were unliganded following coupling reaction. One hundred microliters of each E. coli SK (PDB ID: 1KAG) and Mycobacterium tuberculosis SK (PDB flow-through fraction was mixed with 800 μl 2 mM PEP and 100 μl ID: 2IYW and 2IYQ). The best obtained model was evaluated of 2.5 mM NADH, and PK/LDH enzymes were added to start the using the Web-based program rampage (http://mordred.bioc.cam. reaction. Conversion of NADH to NAD+ was monitored at ac.uk/~rapper/rampage.php) to produce a Ramachandran plot. 3 340 nm. NADH (εNADH = 6.22 × 10 /M/cm) strongly absorbs light The three-dimensional structures were visualized using the at 340 nm, whereas NAD+ does not; therefore, the ADP and PyMol program [11]. The amino acid composition was analyzed shikimate-3-phosphate concentrations were considered equivalent using the COPid Web server (http://www.imtech.res.in/raghava/ to the decrease in NADH. copid/index.html) [31].

Determination of Kinetic Parameters Results and Discussion To determine the kinetic parameters Km and kcat, we employed continuous coupled assays at 25°C, which was found to be the Expression and Purification of SKs optimal temperature for CpSK activity. The reaction was initiated To characterize the cold adaptation of CpSK, the by adding recombinant SKs (2.5 μg/ml) to the assay mixture psychrophilic C. psychrerythraea SK and mesophilic E. coli containing PK/LDH. Oxidation of NADH to NAD+ was monitored SK were expressed in E. coli. Notably, CpSK was soluble continuously at 340 nm using a BioDrop DUO spectrophotometer only at 15°C, up to 30% of total cellular protein, but was (BioDrop, UK). The steady-state kinetic parameters were determined from initial velocity by varying the concentrations of one substrate insoluble at 37°C. This implied that CpSK was likely to be

December 2016 ⎪ Vol. 26⎪ No. 12 2090 Nugroho et al.

Fig. 2. SDS-PAGE of CpSK (left panel) and EcSK (right panel) purification. Lane a, protein molecular mass markers; lane b, uninduced cell lysate; lane c, induced whole-cell lysate; lane d, supernatant of cell lysate; lane e, insoluble fraction of cell lysate; lane f, flow-through; lane g, wash fraction; lane h, first 1 ml purified protein fraction; lane i, second fraction of purified protein. Arrows indicate the SKs. thermally labile. The EcSK was soluble at 37°C, as reported time points, and the coupled reaction was carried out at previously [10]. The two recombinant SKs were purified 25°C, at which the activity of PK/LDH was highest. As using His-tags, and the total amounts of CpSK and EcSK presented in Fig. 3A, the overall temperature activity purified from a 1 L culture were approximately 1 and profile of CpSK was similar to that of EcSK, but shifted to a 3.6 mg, respectively. Based on the sequence of CpSK, the lower temperature. The optimum temperature of CpSK calculated molecular mass of CpSK was 19.49 kDa. SDS- activity was approximately 30°C, whereas that of mesophilic PAGE analysis showed that the recombinant CpSK migrated EcSK was approximately 40°C (Fig. 3A). A psychrophilic as expected. The purity of both enzymes was estimated to enzyme is classified as cold-adapted when its optimum be at least 95% (Fig. 2). temperature for activity is at or below 30°C [15, 20]. The 10°C shift in the optimum temperature demonstrated that Optimum Temperature of Activity CpSK appeared to be a cold-adapted enzyme. Moreover, The optimum temperature for activity of the two SKs CpSK retained about 40% of its maximal activity at 10°C, was determined by discontinuous coupled assays [46]. The the temperature near optimum growth of C. psychrerythraea, SK reaction was performed at various temperatures and whereas EcSK exhibited only 10% of its maximal activity at

Fig. 3. (A) Temperature dependence of activity, and (B) thermal denaturation curve.

J. Microbiol. Biotechnol. Cold-Adapted Shikimate Kinase 2091

the same temperature. This observation is consistent with at 20°C [32]. As shown in Fig. 3, the unfolding showed previous results from cold-adapted enzymes [9, 15, 20]. In a cooperative transition for both enzymes. The urea addition, CpSK, like EcSK, lost its activity dramatically concentrations required for half-maximal denaturation when the temperature increased above the optimum were 1.5 and 2.5M for CpSK and EcSK, respectively, temperature owing to its heat lability, as described below. indicating that mesophilic enzymes are more resistant to urea denaturation than psychrophilic counterparts. Linear Thermal- and Urea-Induced Unfolding extrapolation of the ΔG for the sharp transition (inset in High catalytic activity at low temperature is generally Fig. 3) presented ΔG values, indicative of the conformational H2O achieved by increasing the flexibility of cold-adapted stability in the absence of denaturant, of 3.94 and 5.74 kcal/ enzymes. However, the flexibility of these enzymes results mol for CpSK and EcSK, respectively. This indicated that in low stability [9, 44, 51]. Thermal unfolding of both the structural stability of CpSK was comparable to that of enzymes was followed at 220 nm in the temperature range DdSK, the ΔG of which was about 4.06 kcal/mol, but H2O of 5–70°C using far-UV CD spectroscopy [21] (Fig. 2B). The lower than that of EcSK [3]. The thermal- and urea-induced thermal transition showed sharp curves with Tm values of unfolding results clearly demonstrated that the stability of 33°C and 45°C for CpSK and EcSK, respectively. The both enzymes correlated with the physiological temperature difference in the unfolding temperature was 12°C, which and that CpSK was more heat labile than mesophilic SKs. was very close to the optimum temperature shift between psychrophilic and mesophilic enzymes shown above. Thus, Kinetic Parameters both enzymes were inactivated as they unfolded, suggesting The observed Km and kcat values for CpSK at 25°C are that CpSK was cold-adapted globally. Additionally, the Tm given in Table 1, together with those of mesophilic SKs. of CpSK was lower than those of other mesophilic SKs: The Km value for shikimate was lower for CpSK than for 39.7°C for DdSK [29] and 47°C for HpSK [5]. Krell et al. [31] other SKs, including EcSK, except for HpSK, which was also showed that substitution of Lys15 with Met in DdSK quite unexpected in psychrophilic enzymes since the increased the Tm by 4.2°C, which introduced a novel enhanced flexibility in the active site typically leads to electrostatic interaction between Arg11 and Asp32 in the reduced substrate affinity (higher Km) [44, 51]. However, mutant enzyme. Similarly, the thermal lability of CpSK the Km value for ATP was comparable to that of other SKs, compared with that of EcSK may result partly from the with the exception of DdSK. In psychrophilic enzymes, the substitution of Lys86 in E. coli with Ile in C. psychrerythraea, turnover number (kcat) increases to counterbalance the which forms an ion pair with Glu139 in the X-ray crystal slower reaction rates imposed by low temperatures [44, 51]. structure of EcSK [29]. This was not the case in CpSK, as shown in Table 1, whose

To compare the structural stability of the enzymes, urea- kcat appeared to not have been improved dramatically. Thus, induced unfolding was monitored by intrinsic fluorescence based on the paradigm that typical cold-adapted enzymes

Fig. 3. Urea-induced unfolding (right panel) of CpSK (filled circle) and EcSK (filled square). The left panel exhibits the linear extrapolation of Δ in the transition region.

December 2016 ⎪ Vol. 26⎪ No. 12 2092 Nugroho et al.

Table 1. Kinetic parameters of shikimate kinases. K (μM) k (/s) k /K (/M/s) Enzyme m cat cat m Reference Shikimate ATP Shikimate ATP Shikimate ATP EcSKa > 20,000-----[26] EcSKa 362 500 55.86 122.51 1.54 × 105 2.45 × 105 This study EcSKb II 200 160 27.1 - 1.36 × 105 - [26] DdSKc 310 620 35 35 1.3 × 105 0.56 × 105 [29] MtSKc 650 112 60 44 0.9 × 105 5.4 × 105 [43] HpSKc 60 101 7.3 - 1.36 × 105 -[4] CpSK100 153 28.8 32.80 2.88 × 5 2.1410 × 105 This study aE. coli shikimate kinase I encoded by aroK. bE. coli shikimate kinase II encoded by aroL. cDdSK, Dickeya dadantii shikimate kinase; MtSK, Mycobacterium tuberculosis shikimate kinase; and HpSK, Helicobacter pylori shikimate kinase.

have higher Km and kcat than thermostable mesophilic residues are displayed in green in Fig. 5C and included counterparts [44], CpSK was exceptional; however, several R57/R58/R60, R116/R117/R120, R132/R136/R140, D33/ reports have also indicated that the Km values of some cold- D34/D34, G79-81/G79-81/G81-83, and F48/F49/F49 (in adapted enzymes are lower than those of mesophilic H. pylori, M. tuberculosis, and C. psychrerythraea, respectively) enzymes [1, 2, 7, 22, 24]. The lower Km of CpSK for shikimate [5, 18]. SK has three domains: the CORE domain composed may account for the low intracellular concentration of of five parallel β-sheets (residues 4–34, 64–114, and 152– shikimate in C. psychrerythraea in the cold environment. 172) and the P-loop (residues 11–18), involved in nucleotide

However, the catalytic efficiency kcat/Km of CpSK was binding; the LID domain (residues 115–130); and the higher than those of the other SKs (including EcSK) listed shikimate binding (SB) domain (residues 34–63; Fig. 5B). in Table 1, indicating that CpSK was catalytically optimized Substrate binding is known to induce large conformational to function at a low temperature to a certain degree. changes in the LID domain (Fig. 5C), which closes over the SB domain for catalysis [5, 43]. The LID domain varies in Sequence Analysis and Comparative Homology Modeling length and amino acid composition between SKs, except Multiple sequence alignment (Fig. 4) using the T-Coffee for the conserved Arg residue (R116/R117/R120 in H. pylori, Web-based server [13] showed that CpSK shared 73% M. tuberculosis, and C. psychrerythraea, respectively) [5, 17, identity with EcSK, 46% identify with Acinetobacter 18, 43]. Variable regions, including the LID region, which is baumannii SK (AbSK), 42% identity with Coxiella burnetii SK displayed in gray (Fig. 5B) in the CpSK homology model, (CbSK), 38% identity with M. tuberculosis SK (MtSK), 35% contained residues 68–75, 86–97, 114–132, and 153–159. identity with Aquifex aeolicus SK (AaSK), 33% identity with Despite the high sequence identity and concomitant Dickeya dadantii (formerly known as Erwinia chrysanthemi) conservation of critical residues, CpSK had a lower optimum SK (DdSK), and 29% identity with SKs from Campylobacter temperature (Fig. 2A), increased heat lability (Fig. 2B), and jejuni (CjSK), Helicobacter pylori (HpSK), and Bacteroides reduced structural stability (Fig. 3) than the mesophilic thetaiotaomicron (BtSK). Except for hyperthermophilic AaSK, EcSK. To elucidate the cold-adaption mechanisms of all SKs originated from mesophilic gram-negative bacteria. CpSK in detail, we attempted to crystallize CpSK; however, Based on Ramachandran plot analysis (data not shown), we were unsuccessful. Thus, primary sequence analysis most residues of CpSK were in the favored and allowed and homology modeling were performed to further explore regions (total 169 residues (98%)), whereas only one the cold-adaptation mechanisms of CpSK. Homology residue (R119) was in the outlier region. This implies that models of CpSK were generated using unliganded EcSK the homology model was of good quality based on the (PDB ID: 1KAG) and MtSK (PDB ID: 2IYW) and the dihedral angles of all the residues. The homology model of MtSK·shikimate·ADP structure (PDB ID: 2IYQ) as templates. CpSK was superimposed with nine other SKs (Fig. 5A). As Generally, global or local flexibility in cold-adapted shown in Figs. 4 and 5, the alignment showed that residues enzymes contributes to increasing catalytic activity at low involved in substrate binding and catalysis were well temperature at the expense of their stability [9, 44, 51]. This conserved throughout the sequence [5, 17, 18, 43]. Conserved flexibility was manifested in the amino acid composition of

J. Microbiol. Biotechnol. Cold-Adapted Shikimate Kinase 2093

Fig. 4. Multiple sequence alignment of shikimate kinases. The critical residues involved in substrate binding and catalysis are indicated (filled circle) at the bottom of the alignment. Abbreviations are as follow: Acinetobacter baumannii (AbSK, ALX97864.1), Bacteroides thetaiotaomicron (BtSK, ALJ44421.1), Coxiella burnetii (CbSK, AML54232.1), Campylobacter jejuni (CjSK, ALK80993.1), Dickeya dadantii (formely known as Erwinia chrysanthemi) (DdSK, KGT98987.1), E. coli (EcSK, YP_026215), Helicobacter pylori (HpSK, EIE29871.1), Mycobacterium tuberculosis (MtSK, AMP29152.1), and Aquifex aeolicus (AaSK, WP_010881430.1). The alignments were generated with the T-Coffee Web-based server and rendered by ESPript. psychrophilic proteins [25, 37, 44]: clustering of Gly residues, versus 3.97%), and total charged residues (38.44% versus lower frequency of Pro residues in loops, a lower Arg/(Arg 33% versus 26.7%) and lower Gly content (6.77% versus + Lys) ratio, fewer hydrophobic residues thus leading to a 7.54% versus 10.8%) than EcSK and MtSK, respectively. lower (Ile + Leu)/(Ile + Leu +Val) ratio, and fewer total Limiting the analysis to the variable regions, the proportion charged residues. However, comparative analysis of the of Arg residues, which are likely to increase thermal stability entire sequence of CpSK against mesophilic SKs was not to a greater degree than Lys residues since their guanidine consistent with many of features in cold adaptation (data group can form more ionic interactions in psychrophilic not shown). For example, CpSK had higher contents of Arg proteins [45], was 25% in the LID region, compared with (9.9% versus 8.6% versus 11.9%), Ile (9.3% versus 5.78% 12.5% and 6% for EcSK and MtSK, respectively. The high

December 2016 ⎪ Vol. 26⎪ No. 12 2094 Nugroho et al.

Fig. 5. Structure of the CpSK model. (A) The CpSK homology model (purple) superimposed to AbSK (pink, PDB ID: 4Y0A), BtSK (cyan, PDB ID: 3VAA), CbSK (green, PDB ID: 3TRF), CjSK (yellow, PDB ID: 1VIA), DdSK (blue, PDB ID: 2SHK), EcSK (gray, PDB ID: 1KAG), HpSK (orange, PDB ID: 1ZUH), MtSK (brown, PDB ID: 2IYQ), and AaSK (magenta, PDB ID: 2PT5). (B) Superimposed structures of open and closed CpSK models generated using unliganded and liganded MtSK structures, highlighting large movement of the LID domain. Domains of the open and closed structures are displayed in different colors with increasing thickness: CORE domain (gray), shikimate binding (SB) domain (blue), and nucleotide binding (NB) domain (orange) with A and B motifs (yellow). (C) Conserved (in green) and variable (in gray) regions are mapped on the CpSK model with substrate binding residues as sticks. content of Arg in the LID domain may have increased local residues in EcSK (i.e., E69 and E146) forming salt bridges flexibility and catalysis, similar to the observation that with R42 and R91, respectively, were replaced with Asp in cold-adapted uracil-DNA glycosylase has a higher Arg/Lys CpSK. The R42 and R91 residues in CpSK were conserved. ratio and that the surface Arg residues of this enzyme are The salt bridges formed by Asp, which has one methylene thought to confer flexibility [34, 50]. group less, may be weaker, thereby decreasing the Consistent with several previous studies [9, 44, 51], we structural stability of the enzyme [16, 30, 54]. Another salt found that the reduced number of salt bridges caused by bridge found in E. coli, K86-E139, was odd in that a similar substitution may not only contribute to thermal stability ionic pair was not found in other SK structures [4, 42, 43]. but also to substrate binding (Km) in CpSK. Two Glu E139 is next to the essential shikimate binding residue

Table 2. Acidic to basic amino acid ratios in SKs. Acidic amino acids Basic amino acids Acidic to SKsa Asp (D) Glu (E) Total Arg (R) His (H) Lys (K) Total basic ratio CpSK13 19 32 17 4 7 28 1.143 AbSK61723153 13310.742 BtSK9 19 28 14 3 8 25 1.120 CbSK113149176320.438 CjSK10 17 27 6 6 23 35 0.771 DdSK515201541201.000 EcSK10 25 35 18 3 11 32 1.094 HpSK616228314250.880 MtSK9 9 18 21 3 5 29 0.621 AaSK72229100 18281.036 aCpSK, Colwellia psychrerythraea; AbSK, Acinetobacter baumannii; BtSK, Bacteroides thetaiotaomicron; CbSK, Coxiella burnetii; CjSK, Campylobacter jejuni; DdSK, Dickeya dadantii; EcSK, E. coli; HpSK, Helicobacter pylori; MtSK, Mycobacterium tuberculosis; AaSK, Aquifex aeolicus.

J. Microbiol. Biotechnol. Cold-Adapted Shikimate Kinase 2095

R140 (corresponding to R132 and R136 in H.pylori and References M. tuberculosis, respectively). Other SKs form salt bridges further away from critical Arg (e.g., E86-K138 in H.pylori), 1. Ásgeirsson B, Cekan P. 2006. Microscopic rate-constants for maintaining the conformational flexibility of the shikimate substrate binding and acylation in cold-adaptation of binding region (F49, R60, R120, and R140). Considering its trypsin I from Atlantic cod. FEBS Lett. 580: 4639-4644. low affinity for shikimate, this bond seems to decrease the 2. Bentahir M. 2000. Structural, kinetic, and calorimetric flexibility in the shikimate binding region in EcSK. In characterization of the cold-active from the antarctic Pseudomonas sp. TACII18. J. Biol. Chem. CpSK, this residue was substituted with Ile and did not 275: 11147-11153. form any salt bridges in the homology model, restoring the 3. Cerasoli E, Kelly SM, Coggins JR, Boam DJ, Clarke DT, elasticity of the shikimate binding regions, leading to a Price NC. 2002. The refolding of type II shikimate kinase higher Km, but it may reduce the thermal stability of CpSK. from Erwinia chrysanthemi after denaturation in urea. Eur. J. Ile is thought to confer thermal stability to the protein; Biochem. 269: 2124-2132. however, CpSK has a high content of Ile. Four valine 4. Cheng W-C, Chang Y-N, Wang W-C. 2005. Structural basis residues, 10, 47, 99, and 150, in EcSK were substituted by for shikimate-binding specificity of Helicobacter pylori shikimate Ile. However, in the model, V150 to Ile may decrease the kinase. J. Bacteriol. 187: 8156-8163. stability of CpSK due to the substitution of neighboring 5. Cheng W-C, Chen Y-F, Wang H-J, Hsu K-C, Lin S-C, Chen Met169 to Arg in CpSK. Overall, CpSK had a lower (Ile + T-J, et al. 2012. Structures of Helicobacter pylori shikimate Leu)/(Ile + Leu +Val) ratio than those from other SKs, with kinase reveal a selective inhibitor-induced-fit mechanism. the exception of the SKs of E. coli and D. dadantii. In PLoS One 7: e33481. 6. Collins T, Meuwis M-A, Stals I, Claeyssens M, Feller G, addition, another characteristic of psychrophilic enzymes Gerday C. 2002. A novel family 8 xylanase, functional and is the presence of more highly acidic residues (Asp and physicochemical characterization. J. Biol. Chem. 277: 35133- Glu) than basic residues (Arg, His, and Lys) compared with 35139. mesophilic and thermophilic enzymes. Acidic amino acids 7. Coracini JD, de Azevedo WF. 2014. Shikimate kinase, a are thought to increase the interactions between enzymes protein target for drug design. Curr. Med. Chem. 21: 592-604. and solvents [40]. Although the differences in acidic to 8. D’Amico S, Claverie P, Collins T, Georlette D, Gratia E, basic amino acid ratios were not significant (Table 2), CpSK Hoyoux A, et al. 2002. Molecular basis of cold adaptation. contained the highest acidic to basic amino acid residue Philos. Trans. R. Soc. Lond. B Biol. Sci. 357: 917-925. ratio among all the SKs compared in this study. Like many 9. D’Amico S, Collins T, Marx J-C, Feller G, Gerday C. 2006. other cold-adapted proteins, CpSK had a larger solvent Psychrophilic microorganisms: challenges for life. EMBO accessible area than the other SKs (data not shown) [44]. Rep. 7: 385-389. In summary, CpSK shares 73% sequence identity with its 10. DeFeyter RC, Pittard J. 1986. Purification and properties of shikimate kinase II from Escherichia coli K-12. J. Bacteriol. E. coli homolog. CpSK was found to be a cold-adapted 165: 331-333. enzyme with an optimum temperature of 30°C. Additionally, 11. DeLano WL. 2002. The PyMOL Molecular Graphics System. CpSK showed high thermal liability and structural instability Accessible at http://www.pymol.org. compared with mesophilic SKs. Notably, the lower Km of 12. De Santi C, Tedesco P, Ambrosino L, Altermark B, Willassen CpSK for shikimate, contrary to many previous studies, N-P, de Pascale D. 2014. A new alkaliphilic cold-active indicated that this psychrophile had evolved an adaptation esterase from the psychrophilic marine bacterium Rhodococcus strategy to improve substrate affinity for metabolically sp.: functional and structural studies and biotechnological essential enzymes, substrates for which are typically present potential. Appl. Biochem. Biotechnol. 172: 3054-3068. at low concentrations intracellularly. We strongly believe 13. Di Tommaso P, Moretti S, Xenarios I, Orobitg M, Montanyola that further structural and mutational studies of this enzyme A, Chang J-M, et al. 2011. T-Coffee: a Web server for the may provide insights not only into the cold adaptation of multiple sequence alignment of protein and RNA sequences metabolic enzymes, but also for the engineering of enzymes using structural information and homology extension. Nucleic Acids Res. 39: W13-W17. to improve their activity at low temperatures. 14. Feller G, Gerday C. 1997. Psychrophilic enzymes: molecular basis of cold adaptation. Cell. Mol. Life Sci. 53: 830-841. Acknowledgments 15. Feller G, Narinx E, Arpigny JL, Aittaleb M, Baise E, Genicot S, et al. 1996. Enzymes from psychrophilic organisms. FEMS This work was supported by the National Fisheries Microbiol. Rev. 18: 189-202. Research and Development Institute (Grant No. R2016002). 16. Feller G, Zekhnini Z, Lamotte-Brasseur J, Gerday C. 1997.

December 2016 ⎪ Vol. 26⎪ No. 12 2096 Nugroho et al.

Enzymes from cold-adapted microorganisms - the class C 2004. Cold-active esterase from Psychrobacter sp. Ant300: beta-lactamase from the antarctic psychrophile Psychrobacter gene cloning, characterization, and the effects of Gly → Pro Immobilis A5. Eur. J. Biochem. 244: 186-191. substitution near the active site on its catalytic activity and 17. Fucile G, Garcia C, Carlsson J, Sunnerhagen M, Christendat stability. Biochim. Biophys. Acta 1696: 59-65. D. 2011. Structural and biochemical investigation of two 31. Kumar M, Thakur V, Raghava GPS. 2008. COPid: composition Arabidopsis shikimate kinases: the heat-inducible isoform is based protein identification. In Silico Biol. 8: 121-128. thermostable. Protein Sci. 20: 1125-1136. 32. Lakowicz JR. 2006. Principles of Fluorescence Spectroscopy. 18. Gan J, Gu Y, Li Y, Yan H, Ji X. 2006. Crystal structure of Springer US, Boston, MA. Mycobacterium tuberculosis shikimate kinase in complex with 33.Lee YS, Bok HJ, Lee JH, Choi YL. 2014. A cold-adapted and an ATP analogue. Biochemistry 45: 8539- carbohydrate esterase from the oil-degrading marine bacterium 8545. Microbulbifer thermotolerans DAU221: gene cloning, purification, 19. Genicot S, Feller G, Gerday C. 1988. Trypsin from antarctic and characterization. J. Microbiol. Biotechnol. 24: 925-935. fish (Paranotothenia magellanica forster) as compared with 34. Leiros I, Moe E, Lanes O, Smalås AO, Willassen NP. 2003. trout (Salmo gairdneri) trypsin. Comp. Biochem. Physiol. B 90: The structure of uracil-DNA glycosylase from Atlantic cod 601-609. (Gadus morhua) reveals cold-adaptation features. Acta Crystallogr. 20. Georlette D, Bentahir M, Claverie P, Collins T, D’Amico S, D Biol. Crystallogr. 59: 1357-1365. Delille D, et al. 2002. Physics and Chemistry Basis of 35. Li S, Yang X, Zhang L, Yu W, Han F. 2015. Cloning, expression Biotechnology. Springer Netherlands, Dordrecht. and characterization of a cold-adapted and surfactant-stable 21. Greenfield NJ. 2006. Using circular dichroism collected as a alginate from marine bacterium Agarivorans sp. L11. J. function of temperature to determine the thermodynamics Microbiol. Biotechnol. 25: 681-686. of protein unfolding and binding interactions. Nat. Protoc. 36. Maiangwa J, Ali MSM, Salleh AB, Rahman RNZRA, Shariff 1: 2527-2535. FM, Leow TC. 2015. Adaptational properties and applications 22. Han C, Zhang J, Chen L, Chen K, Shen X, Jiang H. 2007. of cold-active lipases from psychrophilic bacteria. Extremophiles Discovery of Helicobacter pylori shikimate kinase inhibitors: 19: 235-247. bioassay and molecular modeling. Bioorg. Med. Chem. 15: 37. Metpally RPR, Reddy BVB. 2009. Comparative proteome 656-662. analysis of psychrophilic versus mesophilic bacterial species: 23. Herrmann KM, Weaver LM. 1999. The shikimate pathway. insights into the molecular basis of cold adaptation of Annu. Rev. Plant Physiol. Plant Mol. Biol. 50: 473-503. proteins. BMC Genomics 10: 11. 24. Hoyoux A, Jennes I, Dubois P, Genicot S, Dubail F, François 38. Moyer CL, Morita RY. 2007. Psychrophiles and psychrotrophs. JM, et al. 2001. Cold-adapted beta-galactosidase from the eLS DOI: 10.1002/9780470015902.a0000402.pub2. Antarctic psychrophile Pseudoalteromonas haloplanktis. Appl. 39. Pace CN, Treviño S, Prabhakaran E, Scholtz JM. 2004. Environ. Microbiol. 67: 1529-1535. Protein structure, stability and solubility in water and other 25. Jahandideh M, Barkooie SMH, Jahandideh S, Abdolmaleki P, solvents. Philos. Trans. R. Soc. Lond. B Biol. Sci. 359: 1225- Movahedi MM, Hoseini S, et al. 2008. Elucidating the protein 1234; discussion 1234-5. cold-adaptation: investigation of the parameters enhancing 40. Paredes DI, Watters K, Pitman DJ, Bystroff C, Dordick JS. protein psychrophilicity. J. Theor. Biol. 255: 113-118. 2011. Comparative void-volume analysis of psychrophilic 26. Kaufmman S (ed.). 1987. Methods in Enzymology, Vol. 147: and mesophilic enzymes: structural bioinformatics of Metabolism of Aromatic Amino Acids and Amines. Academic psychrophilic enzymes reveals sources of core flexibility. Press; Elsevier, Amsterdam. BMC Struct. Biol. 11: 42. 27. Kim S-Y, Hwang KY, Kim S-H, Sung H-C, Han YS, Cho Y. 41. Robert X, Gouet P. 2014. Deciphering key features in protein 1999. Structural basis for cold adaptation: sequence, structures with the new ENDscript server. Nucleic Acids Res. biochemical properties, and crystal structure of malate 42: W320-W324. dehydrogenase from a psychrophile Aquaspirillium arcticum. 42. Romanowski MJ, Burley SK. 2002. Crystal structure of the J. Biol. Chem. 274: 11761-11767. Escherichia coli shikimate kinase I (AroK) that confers 28. Kim YO, Park IS, Nam BH, Kim DG, Jee YJ, Lee SJ, et al. sensitivity to mecillinam. Proteins 47: 558-562. 2014. A novel esterase from Paenibacillus sp. PBS-2 is a new 43. Rosado LA, Vasconcelos IB, Palma MS, Frappier V, member of the β-lactamase belonging to the family VIII Najmanovich RJ, Santos DS, et al. 2013. The mode of action lipases/esterases. J. Microbiol. Biotechnol. 24: 1260-1268. of recombinant Mycobacterium tuberculosis shikimate kinase: 29. Krell T, Maclean J, Boam DJ, Cooper A, Resmini M, kinetics and thermodynamics analyses. PLoS One 8: e61918. Brocklehurst K, et al. 2001. Biochemical and X-ray crystallographic 44. Siddiqui KS, Cavicchioli R. 2006. Cold-adapted enzymes. studies on shikimate kinase: the important structural role of Annu. Rev. Biochem. 75: 403-433. the P-loop lysine. Protein Sci. 10: 1137-1149. 45. Siddiqui KS, Poljak A, Guilhaus M, De Francisci D, Curmi 30. Kulakova L, Galkin A, Nakayama T, Nishino T, Esaki N. PMG, Feller G, et al. 2006. Role of lysine versus arginine in

J. Microbiol. Biotechnol. Cold-Adapted Shikimate Kinase 2097

enzyme cold-adaptation: modifying lysine to homo-arginine 53. Sutton KA, Breen J, MacDonald U, Beanan JM, Olson R, stabilizes the cold-adapted α-amylase from Pseudoalteramonas Russo TA, et al. 2015. Structure of shikimate kinase, an in haloplanktis. Proteins 64: 486-501. vivo essential metabolic enzyme in the nosocomial pathogen 46. Simithy J, Gill G, Wang Y, Goodwin DC, Calderón AI. 2015. Acinetobacter baumannii, in complex with shikimate. Acta Development of an ESI-LC-MS-based assay for kinetic Crystallogr. D Biol. Crystallogr. 71: 1736-1744. evaluation of Mycobacterium tuberculosis shikimate kinase 54. Szilágyi A, Závodszky P. 2000. Structural differences activity and inhibition. Anal. Chem. 87: 2129-2136. between mesophilic, moderately thermophilic and extremely 47. Simithy J, Reeve N, Hobrath JV, Reynolds RC, Calderón AI. thermophilic protein subunits: results of a comprehensive 2014. Identification of shikimate kinase inhibitors among survey. Structure 8: 493-504. anti-Mycobacterium tuberculosis compounds by LC-MS. 55. Tang MAK, Motoshima H, Watanabe K. 2014. Cold adaptation: Tuberculosis (Edinb.) 94: 152-158. structural and functional characterizations of psychrophilic 48. Simpson BK, Haard NF. 2011. Purification and characterization and mesophilic . Protein J. 33: 313-322. of trypsin from the Greenland cod (Gadus ogac). 1. Kinetic 56.Wang G, Wang Q, Lin X, Bun Ng T, Yan R, Lin J, et al. and thermodynamic characteristics. Can. J. Biochem. Cell Biol. 2016. A novel cold-adapted and highly salt-tolerant esterase 62: 894-900. from Alkalibacterium sp. SL3 from the sediment of a soda 49. Singh P, Singh SM, Dhakephalkar P. 2014. Diversity, cold lake. Sci. Rep. 6: 19494. active enzymes and adaptation strategies of bacteria inhabiting 57. Webb B, Sali A. 2014. Comparative protein structure modeling glacier cryoconite holes of High Arctic. Extremophiles 18: using MODELLER. Curr. Protoc. Bioinformatics 47: 5.6.1- 229-242. 5.6.32. 50.Smalås AO, Leiros HK, Os V, Willassen NP. 2000. Cold 58. Xu Y, Feller G, Gerday C, Glansdorff N. 2003. Metabolic adapted enzymes. Biotechnol. Annu. Rev. 6: 1-57. enzymes from psychrophilic bacteria: challenge of adaptation 51. Struvay C, Feller G. 2012. Optimization to low temperature to low temperatures in ornithine carbamoyltransferase from activity in psychrophilic enzymes. Int. J. Mol. Sci. 13: 11643- Moritella abyssi. J. Bacteriol. 185: 2161-2168. 11665. 59. Zanphorlin LM, de Giuseppe PO, Honorato RV, Tonoli 52. Su H, Mai Z, Yang J, Xiao Y, Tian X, Zhang S. 2016. CCC, Fattori J, Crespim E, et al. 2016. Oligomerization as a Cloning, expression and characterization of a cold-active strategy for cold adaptation: structure and dynamics of the and organic solvent-tolerant lipase from Aeromicrobium sp. GH1 β-glucosidase from Exiguobacterium antarcticum B7. Sci. SCSIO 25071. J. Microbiol. Biotechnol. 26: 1067-1076. Rep. 6: 23776.

December 2016 ⎪ Vol. 26⎪ No. 12