Roles of Active-Site Aromatic Residues in Cold Adaptation of Sphingomonas glacialis Esterase EstSP1.

The aromatic amino acids, Tyr or Trp, which line the active-site walls of esterases, stabilize the catalytic His loop via hydrogen bonding. A Tyr residue is preferred in extremophilic esterases (psychrophilic or hyperthermophilic esterases), whereas a Trp residue is preferred in moderate-temperature esterases. Here, we provide evidence that Tyr and Trp play distinct roles in cold adaptation of the psychrophilic esterase EstSP1 isolated from an Arctic bacterium Sphingomonas glacialis PAMC 26605. Stern-Volmer plots showed that the mutation of Tyr191 to Ala, Phe, Trp, and His resulted in reduced conformational flexibility of the overall protein structure. Interestingly, the Y191W and Y191H mutants showed increased thermal stability at moderate temperatures. All Tyr191 mutants showed reduced catalytic activity relative to wild-type EstSP1. Our results indicate that Tyr with its phenyl hydroxyl group is favored for increased conformational flexibility and high catalytic activity of EstSP1 at low temperatures at the expense of thermal stability. The results of this study suggest that, in the permanently cold Arctic zone, enzyme activity has been selected for psychrophilic enzymes over thermal stability. The results presented herein provide novel insight into the roles of Tyr and Trp residues for temperature adaptation of enzymes that function at low, moderate, and high temperatures.

Introduction

Enzymes from psychrophiles display high catalytic activity at low temperatures when compared to homologous enzymes from mesophiles and thermophiles.[1,2] For cold-adapted enzymes, the challenge is to ensure adequate molecular movement required for catalysis at low temperatures via the acquisition of flexible structure, particularly near the active site.[2−5] Regions distal to the active site have been shown to influence enzyme catalysis.[6−8] Structural adaptations include reduced intramolecular interactions, loop extension, and increased active-site accessibility.[9−11] The psychrophilic nature of an enzyme can also be strongly influenced by minor variations in amino acids.[12] Cold-adapted enzymes generally exhibit increased glycine content, but reduced arginine, proline, and acidic amino acid contents.[13,14] Accompanying kinetic adaptations include reduced Gibbs free energy of activation (ΔG‡), low enthalpy (ΔH), and reduced substrate-binding affinity;[10] however, these characteristics are not universal for all cold-adapted enzymes.[15] As a result of these adaptations, the enzymes become heat-labile at moderate temperatures.

Despite being more flexible than other parts of the enzyme, the active site of cold-adapted enzymes should maintain the correct conformation and orientation of catalytic residues.[16] Discrete adjustments of amino acid residues in the catalytic site wall are also responsible for adaptation of proteins to low or high temperature.[17] Esterases are hydrolase enzymes that prefer shorter-chain esters of less than 10 carbons, and their catalytic residues are generally Ser, His, and Asp residues.[18] Previous studies have shown that the aromatic amino acid residues, Trp or Tyr, line the active-site walls of cold-adapted, mesophilic, and hyperthermophilic esterases, playing an important role in stabilization of the catalytic His loop via hydrogen bonding as well as maintenance of their catalytic activity at each working temperature.[19−22] Multiple sequence alignment showed that Trp is highly conserved in moderate-temperature esterases (psychrotrophic or mesophilic esterases), whereas Tyr is preferred in the corresponding position of hyperthermophilic esterases and some mesophilic esterases (Figure a, upper panel). Boyineni et al. showed that Trp208 of a psychrotrophic esterase, EstK, is responsible for most of the Trp fluorescence generated by the five Trp residues in EstK, suggesting a hydrophobic environment of the Trp208 residue.[19] The mutation of Trp208 to Tyr in EstK showed 12-fold increased catalytic efficiency at 40 °C, as well as increased catalytic site thermal stability via a strengthened hydrogen bond to Asp308, whereas the other parts of the enzyme were denatured at elevated temperatures.[19] Similarly, the mutation of Trp187 to Tyr in a mesophilic esterase, rPPE, was shown to increase the catalytic efficiency of the enzyme.[20] Conversely, the mutation of Trp187 to His in rPPE has been shown to induce structural rearrangements, such as relocation of Asp287 (equivalent to Asp308 in EstK) and spatial adjustment for substrate binding, resulting in improved catalytic efficiency.[20,21] However, the mutation of Tyr182 to a bulky Trp in a hyperthermophilic esterase, EstE1, caused detrimental effects on catalytic activity and thermal stability of the enzyme.[22] Additionally, the mutation of Tyr182 to a similar-sized Phe had little effect on EstE1 thermal stability, indicating that the hydrogen bond involving Tyr182 is critical for stabilization of the catalytic His loop but not for thermal stability of the enzyme, as the overall structure of EstE1 was already adapted to high temperatures.[22]

Figure 1

(a) Partial multiple sequence alignment of esterases. Cold-adapted esterases; EstK (Pseudomonas mandelii)[23,24] and Q3K919 (Pseudomonas fluorescens). Mesophilic esterases; rPPE (Pseudomonas putida)[21,25] and Est8 (metagenome-derived).[26] Hyperthermophilic esterases; EstE1 (thermal environmental sample)[27,28] and Est2 (Alicyclobacillus acidocaldarius).[29] Psychrophilic esterases; EstSP1 (Sphingomonas glacialis PAMC 26605),[30] EstS (Shewanella halifaxensis),[31] and EstG (Sphingobium sp.).[32] (b) Detailed view of the active site of EstSP1. The loop containing the catalytic residues His284 and Asp254 is shown stabilized by the hydrogen bond of Tyr191 with Ile287.

Interestingly, there was an unusual presence of Tyr at the corresponding position of psychrophilic esterases, including EstSP1 (S. glacialis PAMC 26605),[30] EstS (S. halifaxensis),[31] and EstG (Sphingobium sp.)[32] (Figure a, lower panel). The 314-amino acid EstSP1 originating from the Arctic bacterium S. glacialis PAMC 26605 has an α/β hydrolase structure and a catalytic triad that consists of Ser162, Asp254, and His284 residues.[30] EstSP1 showed substrate preference for esters of short-chain fatty acid as well as dialkyl phthalates (C2–C6).[30] We hypothesized that Tyr191 located in the active-site wall of EstSP1 plays a distinct role in the cold adaptation of the enzyme. The structural model of EstSP1 based on the crystal structure of a mesophilic esterase Est8 (PDB ID: 4YPV)[26] indicated that Tyr191 stabilizes the catalytic His loop in EstSP1 via a hydrogen bond with Ile287 (Figure b), similar to the mechanism of active-site stabilization previously shown for EstK, rPPE, and EstE1.[16,19−22] We utilized site-directed mutagenesis to generate mutations of Tyr191 to form aromatic amino acids with different characteristics (Phe and Trp), as well as a small aliphatic amino acid (Ala) and a charged amino acid (His). Although His is a basic amino acid, the imidazole group of His has an aromatic characteristic at all pH values.[33] Here, we provide evidence that a subtle adjustment of Tyr or Trp in the active-site wall of esterase controls the level of activity and thermal stability in a reciprocal manner; hence, Tyr is preferred for its high catalytic activity of EstSP1 at permanently cold environments of the Arctic zone at the expense of thermal stability, whereas Trp is preferred for maintenance of thermal stability at the expense of catalytic activity. This study provides a striking example of the structural basis of protein evolution and adaptation to a permanently cold environment.

Results

Site-Directed Mutagenesis and Protein Expression

To elucidate the role of Tyr191 lining the active-site wall of EstSP1 in cold adaptation, EstSP1 mutants were constructed by site-directed mutagenesis using the estSP1 gene as a template, including aliphatic amino acid Ala (Y191A), aromatic amino acids Phe and Trp (Y191F and Y191W), or charged amino acid His (Y191H). The wild-type (WT) and mutant EstSP1 enzymes with a C-terminal six His-tag were expressed in Escherichia coli BL21 (DE3) as soluble proteins and purified using nickel-chelate affinity chromatography and anion-exchange chromatography to homogeneity. The mutant proteins appeared on an sodium dodecyl sulfate (SDS) gel as a single band of approximately 35 kDa in size (data not shown), as previously shown for WT EstSP1.[30]

Apparent Optimum Temperature

The apparent optimum activities of WT and mutant EstSP1 enzymes were measured in the temperature range of 4–60 °C. The WT showed maximum activity at 40 °C and pH 8.0 (Figure ). The Y191A, Y191F, and Y191W mutants showed similar apparent optimum temperatures to the WT of approximately 40 °C (Figure ). Surprisingly, Y191H had a thermophilic characteristic with a shift of apparent optimum temperature to 90 °C during 3 min of incubation, at which temperature, the WT lost its activity (Figure ). However, Y191H showed the lowest enzymatic activity below 40 °C compared to the other mutants (Figure ).

Figure 2

Apparent optimum temperature of WT and mutant EstSP1 enzymes. The optimum temperature was determined from 4 to 80 °C in reaction buffer D (100 mM Tris·HCl, pH 8.0, 100 mM NaCl, and 5% glycerol). Error bars represent mean ± S.D., and the points represent an average of at least three independent experiments.

Thermal Stability

To evaluate the effects of mutations on the thermal stability of EstSP1, enzymatic activity was measured at the apparent optimum temperature of WT (40 °C) after incubation at 4, 25, 40, and 60 °C for the indicated times. The WT started losing its thermal stability gradually at 25 °C after 20 min of incubation (Figure ), and its Tm value was 57.2 °C (Table ). In contrast, Y191A, Y191F, Y191H, and Y191W showed improved thermal stability at 25 °C for 2 h (Figure ). However, mutants Y191A and Y191F showed reduced thermal stability relative to the WT at 40 and 60 °C. Interestingly, Y191H and Y191W were more stable than WT, even at 40 °C. At 60 °C, WT lost almost all of its enzymatic activity after 2 h, but Y191W maintained more than 20% of its residual activity (Figure ).

Figure 3

Thermal stabilities of WT and mutant EstSP1 enzymes. Thermal stability was measured at the optimum temperature after incubation of enzymes at 4, 25, 40, and 60 °C for the indicated time. Data are expressed as the percentage of initial activity remaining after incubation at the indicated temperatures. The catalytic activity at the optimum temperature before incubation was considered 100%. Data shown reflect at least three independent experiments.

Table 1

Melting Temperature (Tm) and T50 Values of WT and Mutant EstSP1 Enzymes

	Tm (°C)	T50 (°C)
WT	57.2 ± 1.9	46.8 ± 0.2
Y191A	51.6 ± 2.6	34.5 ± 1.5
Y191F	54.7 ± 4.7	40.8 ± 1.1
Y191W	57.3 ± 0.3	49.6 ± 0.3
Y191H	58.2 ± 0.2	49.7 ± 0.2

Overall, EstSP1 proteins maintained thermal stability in the order of Y191H > Y191W > WT > Y191F > Y191A (Figure and Table ). The T50 values were in agreement with the thermal stability data (Table ).

Effects of Mutation on Conformational Flexibility

The conformational flexibility of WT and mutant EstSP1 enzymes was determined by acrylamide-induced quenching of Tyr and Trp fluorescence upon excitation at 280 nm. EstSP1 has three Trp and nine Tyr residues, with the Trp residues exposed on the surface. The Stern–Volmer plot indicated that all mutant EstSP1 enzymes exhibit reduced conformational flexibility of the overall protein structure relative to the WT (Figure a). Moreover, small-to-large mutations Y191H and Y191W resulted in larger conformational changes than amino acids with aliphatic group, Y191A, and the similar-sized Y191F. All of the substituted residues (Ala, Phe, Trp, and His) played a role in maintaining rigid conformations by providing good coverage and masking to produce hydrophobic pockets,[34] as the region was shown to be hydrophobic in EstK.[19] The hydropathy index values for the mutated amino acids support that Ala, Phe, and Trp participate in hydrophobic interactions, whereas the degree of hydrophobic interaction is less with Tyr because of its hydroxyl group.[35] The results indicate that the overall effect of the hydrogen bond of Tyr191 is to maintain conformational flexibility of enzyme catalysis, not to enhance protein thermal stability.

Figure 4

Conformational flexibility of WT and mutant EstSP1 enzymes. A Stern–Volmer plot was generated by recording the maximum fluorescence intensity in the presence of a range of acrylamide concentrations after excitation at 280 nm. (a) F0/F values were plotted against acrylamide concentration. F0, fluorescence intensity without acrylamide; F, fluorescence intensity with 0–0.5 M acrylamide. The curves are the average of three independent experiments. (b) Stern–Volmer plot obtained by subtracting the individual temperature variation in fluorescence quenching between 4 and 30 °C.

To further investigate the structural permeability of WT and mutants, dynamic acrylamide quenching was measured as shown in Figure b. The difference between quenching at 4 and 30 °C revealed that WT had a larger conformational change than all mutants as it has greater accessibility to acrylamide. Conversely, all mutants showed similar permeabilities at 4 and 30 °C up to 0.3 M acrylamide (Figure b). EstSP1 mutants underwent a discrete conformational change from 0.4 M acrylamide; nevertheless, all mutants exhibited more rigid overall structure than WT (Figure b).

CD Spectroscopy Analysis

To investigate the changes in the secondary structure of EstSP1, the CD spectra of the WT and mutants were measured at 4 and 40 °C, respectively (Figure ). As expected, the WT, Y191H, and Y191W maintained their secondary structure at 4 and 40 °C. Mutants Y191A and Y191F showed distinct changes in secondary structure at 40 °C relative to 4 °C, suggesting the conformation changed at 40 °C, which could be correlated with their loss of stability at 40 °C relative to the WT (Figure ).

Figure 5

The CD spectra were measured at 25 °C after incubating 0.38 mg mL−1 protein sample at 4 and 40 °C for 30 min buffer B (20 mM Tris·HCl, pH 7.5, 25 mM KCl, 0.1 mM ethylenediaminetetraacetic acid (EDTA), 5% glycerol).

Kinetics and Thermodynamic Analysis

The kinetic parameters of the WT and mutants were determined as shown in Table . EstSP1 was found to have a Km value of 92.5 μM and a kcat value of 227 s–1. The Y191F mutant showed a decrease in kcat value to 119 s–1 and an increase in Km value to 247 μM, demonstrating that not only is the aromatic ring of Tyr essential but also the formation of a hydrogen bond via the Tyr phenyl hydroxyl group is critical for its catalytic activity. To our surprise, the Y191F mutant showed a drastic decrease in catalytic rate and substrate-binding affinity, resulting in lower catalytic efficiency (kcat/Km) at 40 °C relative to the WT. The Y191H mutant significantly decreased substrate-binding affinity (2441 μM), reducing the catalytic efficiency by 26-fold relative to the WT. All of the mutants showed significantly reduced substrate affinities as compared to WT, rendering them less efficient catalysts. Among the mutant enzymes, Y191F showed the highest catalytic efficiency of 0.48 μM–1 s–1, whereas Y191H showed the lowest catalytic efficiency value of 0.04 μM–1 s–1.

Table 2

Kinetic and Thermodynamic Parameters of WT and Tyr191 Mutants

	temp. (°C)	Km (μM)	kcat (s–1)	kcat/Km (μM–1 s–1)	ΔG‡ (kJ mol–1)	ΔH‡ (kJ mol–1)	TΔS‡ (kJ mol–1)
WT	20	70 ± 5.6	175 ± 6.3	2.49	59.2	20.2	–39.0
	40	92 ± 1.2	227 ± 3.8	2.46	62.7	20.0	–42.7
Y191A	20	286 ± 10	18 ± 1.0	0.06	64.8	25.9	–38.9
	40	247 ± 11	26 ± 3.2	0.11	68.3	25.8	–42.5
Y191F	20	54 ± 7.6	51 ± 1.2	0.95	62.2	25.2	–37.0
	40	249 ± 7.8	119 ± 1.0	0.48	64.4	25.0	–39.3
Y191H	20	752 ± 27	52 ± 7	0.01	ND	ND	ND
	40	2441 ± 75	90 ± 11	0.04	ND	ND	ND
Y191W	20	349 ± 23	14 ± 0.6	0.04	65.3	20.4	–44.9
	40	486 ± 31	22 ± 4.2	0.05	68.7	20.2	–48.5

The WT showed the lowest ΔG‡ values of 59.2 kJ mol–1 at 20 °C and 62.7 kJ mol–1 at 40 °C (Table ). The low ΔG‡ values of the WT can be attributed to high kcat values of 175 and 227 s–1 at 20 and 40 °C, respectively, which is usual for cold-adapted enzymes thriving at near-zero temperature. In contrast, the mutants exhibited lower catalytic rates and higher ΔG‡ values than the WT. However, the thermodynamic parameters of Y191H could not be determined. The WT and mutants showed negative TΔS‡ values and exhibited an apparent decrease in conformational disorder between the ES and ES‡ states at 20 °C. The change in entropy of the Y191W mutant was significantly larger than that of the WT, suggesting that Trp does not fit well in the space designated for the Tyr191 because of its bulkier side chain. In addition, the lack of a stabilizing hydrogen bond between Tyr191 and Ile287 further adds to disorder around the active site. Conversely, the TΔS‡ values of Y191A and Y191F were slightly decreased, indicating the compactness of the catalytic domain as expected for the small hydrophobic side-chain residues.

Overall, the mutants exhibited an increased enthalpy of activation (ΔH‡) relative to the WT, demonstrating that the mutations induced conformational change at lower temperature, reducing their catalytic rates.

Y191H Mutant

Although the Y191H mutant showed relatively higher thermal stability from 70 to 90 °C compared to WT (Figures and 3), the kinetic data demonstrated that the catalytic activity of Y191H remained much lower than that of WT (Table ). We speculated that Y191H might have acquired some additional intramolecular interactions. A structural model of Y191H based on the crystal structure of Est8 showed the formation of a salt bridge between His191 and Asp161 (Figure a). Moreover, the structural rotation of His191 in comparison to Tyr191 resulted in relocation of nearby residues, leading to the formation of a strong hydrophobic core near the catalytic site in the Y191H mutant. Salt bridges were shown to be the likely principal stabilizing force enhancing protein thermal stability.[36] Furthermore, salt bridges within thermally unstable regions were found to be relatively stronger and to undergo additional tightening at high temperatures.[37]

Figure 6

(a) Detailed view of salt bridge formation between His191 and Asp161, which is located next to catalytic residue Ser162 in Y191H mutant. (b) Temperature-induced unfolding for Y191H was measured upon excitation at 280 nm after incubation for 6 min at the indicated temperatures. The change in fluorescence intensities at different temperatures was shown as the ratio of fluorescence at 4 °C (F0) to fluorescence intensities at increasing temperatures (F).

To investigate temperature-induced unfolding of the tertiary structures of WT and Y191H, fluorescence emission spectra were measured at various temperatures. The fluorescence emission spectra suggested that Y191H was resistant to heat and gradually unfolded from 40 to 70 °C with a constant increase in catalytic efficiency, whereas WT started unfolding at 40 °C and was approximately 60% unfolded at 70 °C with a drastic loss of catalytic activity (Figure b). Although Y191H exhibited the apparent maximum enzymatic activity at 90 °C (Figure ), the overall structure of Y191H was denatured (Figure and 6b). However, the Y191H mutant showed an increase in Tm value relative to WT.

Ile287 Mutants

To further elucidate the role of Tyr191, we mutated the counterpart amino acid Ile287 located at the loop holding the catalytic His284. Two more mutants, I287A and I287P, were constructed. The Ile backbone nitrogen atom was involved in the formation of the hydrogen bond between Tyr191 and Ile287, and the same bond remained for the I287A mutant (Figure b). The overall effect of mutation was more detrimental for I287P than I287A, with Pro causing major structural disruption. The optimum temperature of I287A was shown to be 40 °C, similar to that of WT, whereas I287P showed an optimum temperature range of 30–50 °C (Figure a). To compare their thermal stabilities, the esterase activities of WT and Ile287 mutants were measured at 40 °C following incubation at 4, 25, 40, and 60 °C for the indicated times. Mutation of Ile287 to Ala or Pro resulted in a loss of thermal stability (Figure b).

Figure 7

Ile287 mutants. (a) Apparent optimum temperature was determined from 4 to 80 °C in reaction buffer D (100 mM Tris·HCl, pH 8.0, 100 mM NaCl, and 5% glycerol). Error bars represent mean ± S.D., and the points represent an average of at least three independent experiments. (b) Thermal stability was measured at the optimum temperature after incubation at 4, 25, 40, and 60 °C for the indicated time. Data are expressed as the percentage of initial activity remaining after incubation at the indicated temperatures.

Discussion

The high catalytic activity at low temperatures, intrinsic flexibility, and thermal lability represent a central paradigm in cold adaptation of enzymes.[3,38−41]S. glacialis PAMC 26605 is a lichen-associated bacterium that thrives in the permanently cold Arctic zone.[42] EstSP1 showed an apparent optimum activity at 40 °C and approximately 25% activity at 4 °C (Figure ). EstSP1 was comparatively stable at 25 °C or below, but was quite unstable at temperatures over 40 °C (Figures and 3). These results reflect that EstSP1 evolved to be flexible and consequently more heat-labile to adapt to the permanently cold environment it inhabits.

The active-site wall of EstSP1 has an unusual presence of Tyr at position 191 (Figure a, lower panel), which plays a key role in establishment of its psychrophilic nature. Particularly, the phenyl hydroxyl group of Tyr191 is essential, not only for maintenance of proper conformation of the EstSP1 active site and enabling high catalytic activity, but also for preventing its structural rigidity. As shown in the Stern–Volmer plots, the mutation of Tyr191 to Ala, Phe, Trp, and His all led to the formation of a rigid structure (Figure ). Moreover, all mutant enzymes showed reduced catalytic rate (Table ). The replacement of Tyr191 with His in our study resulted in high apparent optimum temperature and increased thermal stability. Dou et al. showed that the mutation of Trp187 to His in rPPE caused spatial relocation of the residues,[21] but the effect of the mutation on the flexibility and stability of rPPE is unknown. However, we do not expect EstSP1 to evolve the additional unnecessary thermal stability required at high temperatures.

The role of a hydrogen bond in proteins existing in hydrophobic environments has been extensively investigated. Interleukin-1β (IL-1β) is a protein involved in inflammatory process regulation that has an internal hydrophobic cavity.[43] When Tyr157 of chicken IL-1β was mutated to Phe, a corresponding residue in human IL-1β, the Y157F mutant showed increased thermal stability with a 10 °C shift in Tm value, suggesting that the amino acid residues present around the internal hydrophobic cavity play important roles in maintenance of thermal stability rather than cavity size.[43,44] Similarly, Tyr125 near the active site of Aerococcus viridians lactate oxidase plays a subtle role in maintenance of the active-site microenvironment best suited for flavin reduction and oxidation steps.[45] Mutation of Tyr125 to Phe (Y125F) resulted in unfavorable participation in enzyme activity by generating a tighter hydrophobic association with the substrate and hence slower release of pyruvate.[45] However, for esterases, either Tyr or Trp is favored as both residues can participate in hydrogen bond formation along with their hydrophobic characteristics.

We previously showed that the active site of EstK is constrained via a stability and flexibility trade-off.[16] Specifically, the hydrogen bond of Trp208 and Asp308 was crucial in maintaining the thermal stability of WT EstK, although the mutation of Asp308 to Ala exhibited enhanced substrate affinity and catalytic rate via enlargement of the active site.[16] Although both Trp and Tyr can participate in hydrogen bond formation, Trp is more hydrophobic than Tyr and therefore contributes to enhanced thermal stability, which in turn negatively affects binding specificity with the substrate. Our results suggest that distinct roles of Tyr and Trp for catalytic activity and thermal stability are also applied to other esterases adapted to elevated temperatures. As shown in Table and Figure , Tyr is favored for high catalytic activity of WT EstSP1 (Figure a) and WT EstE1 (Figure d)[22] as well as EstK W208Y[19] and rPPE W187Y,[20] whereas Trp is favored in WT EstK (Figure b)[19] and WT rPPE (Figure c).[20] Recently, Pereira et al. have characterized a mesophilic esterase, Est8, having higher catalytic efficiency against short-chain esters with mild thermal stability.[26] Sequence alignment showed the presence of Tyr near the catalytic site of Est8 (Figure a, upper panel), further supporting the idea that the presence of Tyr can be attributed to high enzymatic activity.[44]

Table 3

Comparison of Kinetic Parameters of EstSP1, EstK, rPPE, and EstE1

esterase		amino acid		Km (μM)	kcat (s–1)	kcat/Km (μM–1 s–1)	references
EstSP1	psychrophilic	WT	Y191	92.5 ± 1.2	227 ± 3.8	2.46	this study
		mutant	Y191W	486 ± 31	22.3 ± 4.2	0.05	this study
EstK	psychrotrophic	mutant	W208Y	287 ± 32	134 ± 7	0.47	Boyineni et al.,[19] Truongvan et al.[16]
		WT	W208	1183 ± 160	49 ± 5	0.04	Truongvan et al.[16]
rPPE	mesophilic	mutant	W187Y	3 × 105	844	356.6	Chen et al.[20]
		WT	W187	1.46 × 105	8.90	6.1 × 10–5	Ma et al.[25]
EstE1	thermophilic	WT	Y182	921 ± 80	1270 ± 156	1.4	Truongvan et al.[22]
		mutant	Y182W	180 ± 45	19 ± 3	0.1	Truongvan et al.[22]

Figure 8

View of the active sites. (a) Psychrophilic esterase EstSP1; (b) psychrotrophic esterase EstK; (c) mesophilic esterase rPPE; (d) hyperthermophilic esterase EstE1.

Nevertheless, compared to our previous studies on EstK and EstE1,[16,19,22] we found that EstSP1 has adopted the same strategy as hyperthermophilic EstE1 rather than cold-adapted EstK, showing that cold-adapted enzymes could be specifically categorized into two sets of enzymes, i.e., psychrophilic and psychrotrophic. Furthermore, it gives insight that instead of global flexibility the localized flexibility around active site is sufficient to render psychrophilic characters, which is evident from the mutational and comparative studies of EstSP1. The mutation of Tyr to a bulky Trp in EstSP1 was accommodated, but not in hyperthermophilic esterase EstE1 as it disrupted both the thermal stability and catalytic activity of EstE1 Y182W.[22] It is very likely that the requirements of active-site stabilization and high catalytic activity under respective physiological conditions are the primary factors involved in the evolution of the EstSP1 structure. Kovacic et al. showed that fine-tuned atomic interactions in psychrophilic EstS by structural elements of Loop10 (having corresponding Tyr residue) contributed significantly to the flexibility.[31]

Selective pressure of temperature provides evolutionary response with a defined set of rules that can impart desirable properties in the adapted protein. Previous studies showed that site-specific evolution of enzymes leaves distinct signatures on an organism’s genome that are influenced by a dynamic relationship between structural and functional constraints.[46] The tendency of certain amino acids to fine-tune atomic interactions and functions at particular temperatures can lead to preferences for substitution of one amino acid for another. For example, the cold-adapted organisms may replace Arg residues with Lys to promote protein flexibility and optimal protein function.[47,48] Cold-adapted Archaea have a higher content of uncharged amino acids, such as Gln and Thr, and lower contents of hydrophobic amino acids, such as Leu.[49] Systematic analysis of proteins adapted to different temperatures showed that thermophilic proteins have a significantly higher frequency of Tyr compared to mesophilic proteins.[22,50,51]

Our results are consistent with recent predicted trends from computational and statistical analyses that indicate that organisms living in extreme environments (psychrophiles and thermophiles) have similar possibilities of combining adaptation features and environmental constraints.[52,53] These findings are concordant with the prediction of origin of the last universal common ancestor, which assumes that the bacterial ancestor is thermophilic; therefore, psychrophilic esterases retained some features of their thermophilic ancestors.[54] Our study strongly supports the idea that enzyme activity and conformational flexibility are closely correlated, which is consistent with the hypothesis that the psychrophilic enzyme EstSP1 evolved as a result of divergent evolution from psychrotolerant esterases. On the basis of the results of this and previous studies, we suggest that Tyr is favored in the active-site wall of extremophilic enzymes (both psychrophilic esterases and hyperthermophilic esterases) to stabilize the catalytic His loop for optimum activity at extreme temperatures, whereas Trp is preferred in psychrotolerant and mesophilic enzymes to ensure stability. EstSP1 can therefore serve as a model system for understanding enzyme evolution and molecular mechanisms of cold adaptation. These findings can lead to the development of protein engineering strategies for design of more efficient proteins under extreme conditions.

Materials and Methods

Materials

S. glacialis PAMC 26605 was kindly provided by Polar and Alpine Microbial Collection (PAMC) of Korea Polar Research Institute (Incheon, South Korea).[42,55] The strain formerly known as Sphingomonas sp. PAMC 26605 was named S. glacialis PAMC 26605 in the EzTaxon database based on its 16S rRNA homology to S. glacialis C16yT (99.78%).[42,56,57] An EZchange site-directed mutagenesis kit was obtained from Enzynomics (Daejeon, South Korea), whereas pET28a(+) vector was purchased from Novagen (Madison, WI). Additionally, HisTrap and CaptoQ columns were purchased from GE Healthcare (Piscataway, NJ), p-nitrophenyl butyrate (pNPB) was acquired from Sigma (St. Louis, MO), and SYPRO orange dye was procured from Life Technologies (Carlsbad, CA). All other reagents were obtained from Sigma, unless otherwise noted.

Construction of Expression Plasmid and Site-Directed Mutagenesis

The WT estSP1 gene (GenBank ID: WP_010186968.1) was subcloned into a TA vector in our previous study.[30] The mutations in EstSP1 were introduced using an EZchange site-directed mutagenesis kit according to the manufacturer’s instructions. Four EstSP1 mutants at the Tyr191 position (Y191A, Y191F, Y191H, and Y191W) and two EstSP1 mutants at the Ile287 position (I287A and I287P) were generated using the WT estSP1 gene as a template. The primers used for site-directed mutagenesis are listed in Table S1. The resulting PCR products were treated with Dpn I to destroy all plasmid templates, leaving only mutant DNA plasmid. The Dpn I-treated PCR product was then transformed into E. coli DH5α. TA plasmids with the desired mutation were isolated and confirmed by DNA sequencing, after which the mutant estSP1 genes were amplified from the TA vector and subcloned into a pET28a(+) vector.

Protein Expression and Purification

Recombinant WT and mutant EstSP1 proteins were purified to homogeneity as previously described.[30] Briefly, E. coli BL21 (DE3) transformed with pET28a(+) vector bearing the WT or mutant estSP1 gene was grown in a shaking incubator at 37 °C until an OD600 of 0.6–0.8 was attained. After induction with 1 mM isopropyl β-d-1-thiogalactopyranoside, cells were grown for an additional 6 h at 30 °C, after which they were harvested at 12 000g for 5 min. Following cell lysis by sonication, the crude enzyme was purified to electrophoretic homogeneity by nickel-chelate affinity chromatography with a 1 mL HisTrap column using a linear gradient of imidazole (20–500 mM) in buffer A (20 mM Tris·HCl, 100 mM M NaCl, 20 mM imidazole, 5% glycerol, pH 7.5), followed by anion-exchange chromatography with a 5 mL CaptoQ column using a linear gradient of KCl (25–1000 mM) in buffer B (20 mM Tris·HCl, 25 mM KCl, 0.1 mM EDTA, 5% glycerol, pH 7.5). The purity of the enzyme was verified by sodium dodecyl sulfate electrophoresis, and the protein concentration was determined by the Bradford method. The purified enzyme was frozen in liquid nitrogen and stored at −80 °C.

Enzyme Assay

Enzyme activity was determined by measuring the amount of p-nitrophenol formed by hydrolysis of pNPB by 100 pmol of the enzyme. The concentration of p-nitrophenol was measured with a Shimadzu UV-1800 spectrophotometer at 400 nm. One enzyme unit releases 1 μM p-nitrophenol per minute at 25 °C from pNPB. The reaction was conducted in 1 mL of buffer C (100 mM Tris·HCl, pH 7.5, 100 mM NaCl) at 25 °C for 3 min. The background rate of pNPB hydrolysis was subtracted for each reading.

Biochemical Characterization

The apparent optimum temperature was determined by conducting the reaction at 4–80 °C in reaction buffer D (100 mM Tris·HCl, pH 8.0, 100 mM NaCl, 5% glycerol). The thermal stability of WT and mutant EstSP1 enzymes was determined by measuring the residual enzyme activity after exposing the enzyme separately to different temperatures (4, 25, 40, 60, and 80 °C). Residual activity was measured in buffer D every 20 min up to 2 h. The enzyme activity before incubation was considered to be 100%.

The T50 value, which is the temperature at which 50% of the enzyme activity is lost during a fixed incubation period, was determined by incubating the aliquots of enzymes in buffer D for 20 min at various temperatures. Residual activity was determined and expressed as percentage of the initial activity measured at 40 °C.

Enzyme Kinetics and Thermodynamic Analysis

The catalytic rate constant (kcat) and Michaelis–Menten constant (Km) were determined by measuring the reaction rate of WT and mutants in substrate (pNPB) at concentrations ranging from 0.1 to 0.5 mM and 40 °C for 3 min. The kinetic parameters (Vmax and Km) were estimated by a Lineweaver–Burk plot with the GraphPad Prism software (San Diego, CA). The kcat parameter was calculated using the equation kcat = Vmax/[ET], and kcat values at various temperatures were calculated to construct an Arrhenius plot (ln kcat vs 1/T). The activation energy (Ea) of pNPB hydrolysis was determined from the slope of the Arrhenius plot. Thermodynamic parameters (Gibbs free energy of activation (ΔG‡), enthalpy (ΔH‡), and entropy (ΔS‡)) were calculated as described by Lonhienne et al.[58] using the following equationsIn the above equations, kB is the Boltzmann constant (1.3805 × 10–23 J K–1), h is the Plank constant (6.6256 × 10–34 J s), k is the catalytic rate constant, and R is the gas constant (8.314 J mol–1 K–1).

Protein Thermal Shift Analysis

Thermal shift analysis was conducted on an Applied Biosystems StepOnePlus Real-Time PCR instrument using SYPRO orange as a dye.[59] A thermal denaturation curve was recorded from 25 to 99 °C using a fixed protein concentration (5 μM) added to buffer C and SYPRO orange dye in a total volume of 20 μL. The melting temperature (Tm), which is the temperature at which 50% of the protein is unfolded, was determined using the protein thermal shift analysis software from Applied Biosystems.

Fluorescence Spectroscopy

The fluorescence emission spectra of WT and mutant EstSP1 enzymes were measured using a Scinco FS-2 fluorescence spectrometer at 25 °C and an excitation wavelength of 280 nm. Acrylamide-dependent fluorescence quenching was determined in the presence of increasing concentrations of acrylamide (0–0.5 M) with 0.5 μM of the enzyme in buffer B. Quenching data were plotted using the GraphPad Prism software and presented as the ratio of intrinsic fluorescence intensity (F0) to the fluorescence intensity with 0–0.5 M acrylamide (F). Temperature-dependent unfolding was determined by measuring the intrinsic fluorescence of the protein samples by incubating at different temperatures (4, 10, 20, 30, 40, 50, 60, 70, and 80 °C) for 6 min. The changes in the fluorescence intensities at various temperatures were shown as the ratio of fluorescence at 4 °C (F0) to the fluorescence intensities at increasing temperatures (F).

Circular Dichroism (CD) Spectroscopy

The CD spectra were analyzed to estimate the protein secondary structure and stability using a JASCO J-715 spectropolarimeter at 25 °C at Korea Basic Science Institute (Ochang, South Korea). The protein sample concentration was 0.38 mg mL–1 in buffer B. Before measurement, samples were incubated at the indicated temperature (4 or 40 °C). The spectra were then plotted as residual ellipticity (mdeg) versus wavelength (nm) using the GraphPad Prism software.

Sequence Analysis and Homology Modeling

Multiple sequence alignment was conducted using Clustal Omega.[60] To gain insight into the three-dimensional structure of WT and mutant EstSP1, a homology model was generated using the Swiss-Model server (http://swissmodel.expasy.org/) with the crystal structure of Est8 (PDB ID: 4YPV) as a template.[26] Multiple sequence alignment between EstSP1 and Est8 showed 45% sequence identity with catalytic residues consisting of Ser, His, and Asp (Figure S1a). A conserved GXSXG motif containing a catalytic Ser residue and an HGGG motif involved in oxyanion hole formation were also conserved. The Ramachandran plot statistics validated by the PROCHECK program showed that more than 87% of the amino acids in the WT and mutant models were in the most favored regions, whereas less than 11% were in additional allowed regions, only 0.4% of residues were in generously allowed regions, and 1.6% were in disallowed regions (Figure S1b and Table S2), signifying good quality of the structural models. The predicted structure was visualized by the Chimera software.[61] All structural models were submitted to the Protein Model Database (PMDB),[62] and the accession numbers are listed in Table S3. Substrate-binding pocket volumes were calculated using the Computed Atlas of Surface Topography of proteins (CASTp) database.[63]