Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
S. Kanaya, Department of Material and Life Science, Graduate School of Engineering, Osaka University, 2-1, Yamadaoka, Suita, Osaka 565-0871, Japan Fax: +81 6 6879 7938 Tel: +81 6 6879 7938 E-mail: firstname.lastname@example.org
Pro-Tk-SP from Thermococcus kodakaraensis consists of the four domains: N-propeptide, subtilisin (EC 220.127.116.11) domain, β-jelly roll domain and C-propeptide. To analyze the maturation process of this protein, the Pro-Tk-SP derivative with the mutation of the active-site serine residue to Cys (Pro-Tk-S359C), Pro-Tk-S359C derivatives lacking the N-propeptide (ProC-Tk-S359C) and both propeptides (Tk-S359C), and a His-tagged form of the isolated C-propeptide (ProC*) were constructed. Pro-Tk-S359C was purified mostly in an autoprocessed form in which the N-propeptide is autoprocessed but the isolated N-propeptide (ProN) forms a stable complex with ProC-Tk-S359C, indicating that the N-propeptide is autoprocessed first. The subsequent maturation process was analyzed using ProC-Tk-S359C, instead of the ProN:ProC-Tk-S359C complex. The C-propeptide was autoprocessed and degraded when ProC-Tk-S359C was incubated at 80 °C in the absence of Ca2+. However, it was not autoprocessed in the presence of Ca2+. Comparison of the susceptibility of ProC* to proteolytic degradation in the presence and absence of Ca2+ suggests that the C-propeptide becomes highly resistant to proteolytic degradation in the presence of Ca2+. We propose that Pro-Tk-SP derivative lacking N-propeptide (Val114-Gly640) represents a mature form of Pro-Tk-SP in a natural environment. The enzymatic activity of ProC-Tk-S359C was higher than (but comparable to) that of Tk-S359C, suggesting that the C-propeptide is not important for activity. However, the Tm value of ProC-Tk-S359C determined by far-UV CD spectroscopy was higher than that of Tk-S359C by 25.9 °C in the absence of Ca2+ and 7.5 °C in the presence of Ca2+, indicating that the C-propeptide contributes to the stabilization of ProC-Tk-S359C.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Pro-Tk-SP derivative with the single Ser359→Cys mutation
Tk-SP in a pro-form (Ala1-Gly640)
Tk-SP derivative with the single Ser359→Cys mutation
an active form (Val114-Val539) of subtilisin-like serine protease from Thermococcus kodakaraensis
Subtilisin (EC 18.104.22.168) is a serine protease with the catalytic triad consisting of Asp, His and Ser. It is commercially valuable enzyme and is widely used for industrial purposes, especially as an additive of detergents . Subtilisin is synthesized in a precursor form, termed prepro-subtilisin, and secreted to the external medium in an inactive pro-form, termed pro-subtilisin, in which a propeptide is attached to the N-terminus of the mature domain, with the assistance of a signal peptide . Pro-subtilisin is then matured to active protease molecule by three steps: folding of the subtilisin domain, autoprocessing of propeptide and degradation of propeptide, which proceed sequentially [2–5]. Propeptide functions not only as an intramolecular chaperone [6,7], but also as a strong inhibitor [8,9] of the cognate mature domain.
Thermococcus kodakaraensis is a hyperthermophilic archaeon, which grows most optimally at 90 °C . Its genome contains three genes encoding subtilisin-like serine proteases: Tk-SP (accession number YP184102), Tk-subtilisin (accession number BAB60701) and Tk-0076 (accession number YP182489) . The amino acid sequence identities between Pro-Tk-SP (pro-form of Tk-SP) and Pro-Tk-subtilisin (pro-form of Tk-subtilisin), between Pro-Tk-SP and Tk-0076, and between Pro-Tk-subtilisin and Tk-0076 are 29%, 28% and 34%, respectively. The structures and functions of Tk-SP [12,13] and Tk-subtilisin [14–20] have been well studied, whereas Tk-0076 remains to be characterized. Because Tk-SP and Tk-subtilisin are highly thermostable enzymes with an optimum temperature for activity in the range 90–100 °C and are highly resistant to heat, detergents and denaturants, they have great potential for biotechnological applications.
Tk-SP (Val114-Val539) is considered to be matured from Pro-Tk-SP (Ala1-Gly640) upon removal of the N-propeptide (Ala1-Ala113) and C-propeptide (Asp540-Gly640) because Tk-SP is purified from Escherichia coli cells overproducing Pro-Tk-SP . It requires neither Ca2+, nor propeptides for folding. Asp147, His180 and Ser359 form the catalytic triad of Tk-SP. The crystal structure of the active-site mutant of Pro-Tk-SP lacking the C-propeptide, ProN-Tk-S359A, has recently been determined . According to this structure, ProN-Tk-S359A consists of the N-propeptide, subtilisin domain and β-jelly roll domain (Fig. 1). The overall structure of ProN-Tk-S359A without β-jelly roll domain is similar to those of Pro-Tk-subtilisin  and bacterial subtilisin:propeptide complexes [21,22], except that it does not contain Ca2+ ions. This suggests that the N-propeptide of Pro-Tk-SP serves as a molecular chaperone and an inhibitor of the mature domain, as do propeptides of bacterial subtilisins [3–5] and Tk-subtilisin [14–18]. The β-jelly roll domain contains two Ca2+ ions. This domain is not required for folding but is required for hyperstability of Tk-SP in a Ca2+-bound form .
Tk-subtilisin is matured from Pro-Tk-subtilisin by three steps, as are bacterial subtilisins [3–5]. By contrast, Tk-SP is matured from Pro-Tk-SP through an intermediate form with the N- or C-propeptide . However, it remains to be determined whether the N- or C-propeptide is autoprocessed first because these propeptides are similar in size (13 kDa for N-propeptide and 12 kDa for C-propeptide) and the intermediate form is too unstable to be biochemically characterized. This intermediate form accumulates in E. coli cells upon overproduction of Pro-Tk-SP as a major recombinant protein, although it is rapidly matured to Tk-SP during sonication lysis of the cells and subsequent purification procedures, even at 4 °C . In addition, the role of the C-propeptide remains to be understood. It has been reported that the mutation of the catalytic serine residue to Cys arrests maturation of bacterial subtilisin  and Tk-subtilisin  at the step in which propeptide is autoprocessed and forms a complex with the mature domain. Therefore, it is expected that the Pro-Tk-SP derivative with the corresponding Ser359→Cys mutation, Pro-Tk-S359C, is matured very slowly, such that an intermediate form can be identified.
In the present study, we overproduced, purified and biochemically characterized Pro-Tk-S359C, its derivatives lacking the N-propeptide (ProC-Tk-S359C) and both propeptides (Tk-S359C), and a His-tagged form of the C-propeptide (ProC*). We show that the N-propeptide is autoprocessed first in the maturation process of Pro-Tk-S359C, although the C-propeptide is subsequently autoprocessed and degraded only in the absence of Ca2+. The C-propeptide was not autoprocessed in the presence of Ca2+, suggesting that Pro-Tk-SP derivative lacking N-propeptide (Val114-Gly640) (ProC-Tk-SP) is not an intermediate form but is the mature form of the enzyme. We also show that the C-propeptide contributes to the stabilization of ProC-Tk-S359C.
Results and Discussion
Autoprocessing of Pro-Tk-S359C
To examine whether the mutation of the active-site serine residue (Ser359) to Cys arrests the maturation process of Pro-Tk-SP at the step in which either the N- or C-propeptide is removed, Pro-Tk-S359C was constructed by greatly reducing the enzymatic activity of the mature domain of Pro-Tk-SP. The primary structure of this protein as well as those of its derivatives analyzed in the present study are schematically shown in Fig. 2. The isolated N- and C-propeptides are termed ProN and ProC, respectively. Upon overproduction of Pro-Tk-S359C, the 55 kDa protein, instead of Pro-Tk-S359C, accumulated in cells in a soluble form (data not shown). The purified protein was eluted from the gel-filtration column as a single peak. However, it gave three bands on SDS/PAGE with molecular masses of 66, 55 and 13 kDa, although the amount of the 66 kDa protein estimated from the intensity of the band visualized with Coomassie Brilliant Blue (CBB) staining was lower than that of the 55 kDa protein by more than 10-fold (Fig. 3, lane 1). The N-terminal amino acid sequences of the 55 and 13 kDa proteins extracted from the gel were determined to be VETE and PQKP, respectively, indicating that the N-terminal residues of the 55 and 13 kDa proteins are Val114 and Pro2, respectively. The molecular masses of the 66, 55 and 13 kDa proteins determined by MALDI-TOF MS were comparable to the calculated ones of Pro-Tk-S359C, Pro-Tk-S359C with the N-propeptide removed (ProC-Tk-S359C), and the isolated N-propeptide (ProN), respectively (Table 1). These results indicate that Pro-Tk-S359C was purified mostly in an autoprocessed form in which ProN and ProC-Tk-S359C form a stable complex (ProN:ProC-Tk-S359C). The amount of this complex purified from 1 L of culture as well as the molecular mass estimated from gel-filtration chromatography are summarized in Table 1. Autoprocessing of Pro-Tk-S359C into the ProN:ProC-Tk-S359C complex is initiated when Pro-Tk-S359C is synthesized in E. coli cells, although the subsequent maturation process is inhibited, probably because the enzymatic activity of ProC-Tk-S359C is sufficient to promote autoprocessing of the N-propeptide from Pro-Tk-S359C but is insufficient for subsequent degradation of ProN.
Table 1. Purification yields and molecular masses of the proteins.
Molecular mass estimated from:
Molecular mass determined by:
Gel filtration (kDa)
Calculated molecular mass (Da)b
a The amount of the protein purified from 1 L of culture. b Molecular mass calculated from the amino acid sequence. c Pro-Tk-S359C was purified together with the ProN:ProC-Tk-S359C complex and therefore its purification yield was estimated from the intensity of the band visualized with CBB staining after SDS/PAGE. d The ProN:ProC-Tk-S359C complex is eluted from the gel filtration column as a single peak but is separated into the 55 kDa (ProC-Tk-S359C) and 13 kDa (ProN) proteins by SDS/PAGE, for which the molecular masses were determined by MALDI-TOF MS.
68 475 ± 255
55 852 ± 222
12 631 ± 47
Pro-Tk-S359C purified as a mixture of the unautoprocessed and autoprocessed forms, but mostly in the autoprocessed form, was used as the ProN:ProC-Tk-S359C complex for further characterization. The far- and near-UV CD spectra of this complex were almost identical to those of the Pro-Tk-SP derivative with the Ser359→Ala mutation, Pro-Tk-S359A (data not shown), indicating that the conformation of Pro-Tk-S359C is not seriously changed by autoprocessing of the N-propeptide. The ProN:ProC-Tk-S359C complex was inactive at any temperature examined, indicating that ProN functions as an inhibitor of ProC-Tk-S359C, as do propeptides of bacterial subtilisins  and Tk-subtilisin [15,16].
Autoprocessing of ProC-Tk-S359C
Identification of the ProN:ProC-Tk-S359C complex as an intermediate form of the maturation process of Pro-Tk-S359C indicates that the N-propeptide is autoprocessed first. However, this complex may not be suitable for analyzing the subsequent maturation process because the enzymatic activity of the subtilisin domain of Pro-Tk-S359C is probably too low to degrade ProN. Therefore, ProC-Tk-S359C was constructed to examine whether the C-propeptide is autoprocessed from ProC-Tk-S359C (Fig. 2). Upon overproduction, ProC-Tk-S359C accumulated in cells in an insoluble form. It was solubilized in 20 mm Tris–HCl (pH 9.0) containing 8 m urea and 5 mm EDTA, purified in the presence of 8 m urea, and refolded by removing urea by dialysis. It gave almost a single band on SDS/PAGE (Fig. 3, lane 3). The amount of protein purified from 1 L of culture as well as the molecular mass of ProC-Tk-S359C estimated from gel-filtration chromatography are summarized in Table 1. ProC-Tk-S359C exists as a monomer in solution.
When ProC-Tk-S359C was incubated at 80 °C in 20 mm Tris–HCl (pH 7.5) for 1 h in the absence of Ca2+, the 44 kDa protein was generated with a yield of ∼ 80% (Fig. 3, lane 4). It was not autoprocessed into the 44 kDa protein at 4 °C (data not shown), probably because the enzymatic activity of the subtilisin domain of ProC-Tk-S359C is too low to be autoprocessed at this temperature. The molecular mass of the 44 kDa protein was determined to be 44 200 ± 202 Da by MALDI-TOF MS, which is comparable to that of Tk-S359C (44 220 Da) calculated from the amino acid sequence. This result indicates that the C-propeptide is autoprocessed from ProC-Tk-S359C. However, the isolated C-propeptide (ProC) was not detected as a band on SDS/PAGE (Fig. 3, lane 4). Instead, two peptides with molecular masses of 6360 and 4021 Da were detected, in addition to ProC-Tk-S359C and Tk-S359C, by MALDI-TOF MS. This result suggests that ProC is degraded by ProC-Tk-S359C or Tk-S359C immediately when it is released from ProC-Tk-S359C upon autoprocessing. Because the sum of the molecular masses of these two peptides are smaller than the molecular mass of ProC calculated from the amino acid sequence (11 780 Da), ProC is probably degraded at multiple sites. By contrast, the C-propeptide was not autoprocessed from ProC-Tk-S359C at all when ProC-Tk-S359C was incubated at 80 °C for 1 h in 20 mm Tris–HCl (pH 7.5) containing 10 mm CaCl2 (Fig. 3, lane 5).
Purification and characterization of ProC*
To examine whether ProC-Tk-S359C is susceptible to autoprocessing in the absence of Ca2+ but is resistant to autoprocessing in the presence Ca2+ as a result of stabilization of the C-propeptide, a His-tagged form of ProC (ProC*) was constructed (Fig. 2). ProC* has a N-terminal three residues extension, a 11 residues insertion between a His-tag and ProC, and a C-terminal 26 residues extension, all of which are derived from the pET28a vector. Upon overproduction, ProC* accumulated in cells in a soluble form and purified to give a single band on SDS/PAGE (Fig. 4, lane 1). The amount of the protein purified from 1 L of culture as well as the molecular mass of ProC* estimated from gel-filtration chromatography are summarized in Table 1. ProC* exists as a monomer in solution.
To compare the susceptibility of ProC* to proteolytic degradation in the presence of Ca2+ with that in the absence of Ca2+, ProC* was incubated with Tk-SP at 80 °C for 30 min in the presence or absence of Ca2+ at an enzyme to substrate ratio of 1 : 100 (w/w) and subjected to 17% SDS/PAGE. As shown in Fig. 4, ProC* was almost fully degraded by Tk-SP in the absence of Ca2+ (lane 3), whereas it was mostly kept intact in this condition in the presence of Ca2+ (lane 4). The 14 kDa protein generated upon digestion of ProC* with Tk-SP in the presence of Ca2+ (lane 4) was identified as the ProC* derivative with the N-terminal His-tag removed by determination of its amino acid sequence. These results suggest that the C-propeptide of ProC-Tk-S359C is not autoprocessed in the presence of Ca2+ as a result of its stabilization by Ca2+ binding.
To examine whether conformation of ProC* is changed by Ca2+ binding, the far-UV CD spectrum of ProC* was measured in the presence and absence of Ca2+. The spectrum of ProC* gave a broad trough with a single minimum at ∼ 220 nm, which was accompanied by a shoulder at 210 nm, in the presence of Ca2+, whereas it gave a broad trough with a single minimum at ∼ 215 nm in the absence of Ca2+ (Fig. 5). The former and latter spectra are typical of proteins with high α-helical and β-sheet contents, respectively . These results suggest that the α-helical content of ProC* is increased by Ca2+ binding.
Preparation of Tk-S359C
The finding that ProC-Tk-S359C is not autoprocessed further in the presence of Ca2+ suggests that ProC-Tk-SP, instead of Tk-SP, represents a mature form of Pro-Tk-SP. Consequently, the question arises as to whether the C-propeptide is important for activity and/or stability of ProC-Tk-SP. To answer this question, it is necessary to prepare Tk-S359C, which lacks both of the N- and C-propeptides (Fig. 2), and compare its activity and stability with those of ProC-Tk-S359C. Tk-S359C has been shown to accumulate in E. coli cells in a soluble form upon overproduction . It was purified to give a single band on SDS/PAGE in the present study (Fig. 3, lane 2). The amount of protein purified from 1 L of culture as well as the molecular mass of Tk-S359C estimated from gel-filtration chromatography are summarized in Table 1. Tk-S359C exists as a monomer in solution.
The far-UV CD spectrum of Tk-S359C is similar to that of ProC-Tk-S359C in the absence of Ca2+ (Fig. 5). These spectra are similar to each other also in the presence of Ca2+ (data not shown). These results indicate that the conformation of ProC-Tk-S359C is not seriously changed on removal of the C-propeptide. The far-UV CD spectra of ProC-Tk-S359C measured in the presence and absence of Ca2+ also resemble each other (Fig. 5). Alteration of the secondary structure content of ProC-Tk-S359C on removal of the C-propeptide or by binding of the Ca2+ ion(s) to the C-propeptide is probably too small to be detected by the far-UV CD spectrum of ProC-Tk-S359C because the Tk-S359C domain of ProC-Tk-S359C is approximately four-fold larger than the C-propeptide in size and is more structured than the C-propeptide.
Enzymatic activities of ProC-Tk-S359C and Tk-S359C
The specific activities of ProC-Tk-S359C, Tk-S359C and Tk-SP were determined to be 0.015 ± 0.001, 0.077 ± 0.006 and 410 ± 25 units·mg−1, respectively, at 80 °C in the presence of 10 mm CaCl2 using azocasien as a substrate. ProC-Tk-S359C is not autoprocessed and is kept intact in this condition. These results indicate that ProC-Tk-S359C is less active than Tk-S359C by approximately five-fold and Tk-S359C is less active than Tk-SP by 5300-fold. However, a possibility that a weak enzymatic activity of ProC-Tk-S359C reflects incomplete refolding of the protein cannot be ruled out because Tk-S359C and Tk-SP are folded in vivo, whereas ProC-Tk-S359C is refolded in vitro. Therefore, ProC-Tk-S359C was incubated at 80 °C for 1 h in the absence of Ca2+ for autoprocessing of the C-propeptide prior to assay. The specific activity of ProC-Tk-S359C with this heat treatment was determined to be 0.0087 ± 0.001 units·mg−1. Assuming that 80% of ProC-Tk-S359C is converted to Tk-S359C in this condition, the specific activity of Tk-S359C matured from ProC-Tk-S359C is calculated to be 0.011 units·mg−1, which is lower than (but comparable to) that of ProC-Tk-S359C without heat treatment. This result suggests that the C-propeptide is not important for activity of ProC-Tk-S359C. However, further studies are required to clarify why the specific activity of Tk-S359C matured from ProC-Tk-S359C is lower than that of Tk-S359C folded in vivo by six- to seven-fold.
It should be noted that ProC-Tk-S359C and Tk-S359C could be overproduced in E. coli, whereas ProC-Tk-SP and Tk-SP could not (T. Foophow, Osaka University, Japan, personal communication). ProC-Tk-SP and Tk-SP are probably cytotoxic to E. coli cells as a result of their protease activities, although they lose their cytotoxicity by the Ser359→Cys mutation as a result of the reduction of protease activity, as noted above.
Stabilities of ProC-Tk-S359C and Tk-S359C
Thermal denaturation of ProC-Tk-S359C and Tk-S359C was analyzed by monitoring the change in CD values at 222 nm as the temperature was increased. Thermal denaturation of these proteins was analyzed either in the presence of 2.5 m guanidine hydrochloride (GdnHCl) and in the absence of CaCl2 or in the presence of 6 m GdnHCl and 10 mm CaCl2. The GdnHCl concentration was increased to 6 m in the presence of Ca2+ because both proteins are greatly stabilized in the presence of Ca2+, such that they are not fully denatured even at 95 °C in the presence of 2.5 m GdnHCl. The GdnHCl concentration was decreased to 2.5 m in the absence of Ca2+ because Tk-S359C is fully denatured even at 4 °C in the presence of 6 m GdnHCl and the absence of Ca2+. The thermal denaturation curves of ProC-Tk-S359C and Tk-S359C measured in 20 mm Tris–HCl (pH 7.5) containing 2.5 m GdnHCl in the absence of CaCl2 are shown in Fig. 6A, and those measured in the same buffer containing 6 m GdnHCl and 10 mm CaCl2 are shown in Fig. 6B. The midpoints of the transition of these thermal denaturation curves (Tm) are summarized in Table 2. The Tm value of ProC-Tk-S359C is higher than that of Tk-S359C by 25.9 °C in the absence of CaCl2 and 7.5 °C in the presence of 10 mm CaCl2, indicating that the C-propeptide of ProC-Tk-S359C contributes to the stabilization of the protein by 25.9 °C in Tm in the absence of Ca2+ and 7.5 °C in Tm in the presence of Ca2+.
Table 2. Thermal stabilities of ProC-Tk-S359C and Tk-S359C.
a Thermal denaturation of the protein was analyzed in 20 mm Tris–HCl (pH 7.5) containing 2.5 m GdnHCl in the absence of CaCl2 or in the same buffer containing 6 m GdnHCl in the presence of 10 mm CaCl2 by monitoring the change in CD values at 222 nm. The experiment was carried out at least twice and the errors from the average values are indicated. b ΔTm = Tm(ProC-Tk-S359C) –Tm(Tk-S359C) (58.0 and 80.1 °C in the absence and presence, respectively, of 10 mm CaCl2).
58.0 ± 0.83
80.1 ± 0.81
83.9 ± 1.2
87.6 ± 1.1
It should be noted that the Tm value of Tk-S359A is 58.8 °C in the presence of 1 m GdnHCl and the absence of Ca2+ and 88.3 °C in the presence of 1 m GdnHCl and 10 mm CaCl2 . These values should be considerably decreased if the GdnHCl concentration were increased to 2.5 or 6 m. However, the Tm value of Tk-S359C is 58.0 °C in the presence of 2.5 m GdnHCl and the absence of Ca2+ and 80.1 °C in the presence of 6 m GdnHCl and 10 mm CaCl2. These results indicate that Tk-S359C is more stable than Tk-S359A. Further mutational and structural studies will be required to understand the mechanisms by which the Ser359→Cys mutation stabilizes the protein.
Maturation process of Pro-Tk-SP
In the present study, we show that the C-propeptide is autoprocessed from ProC-Tk-S359C at 80 °C in the absence of Ca2+ but is not autoprocessed from it in the presence of Ca2+ even at 80 °C. Because Tk-SP is synthesized as Prepro-Tk-SP, which contains a putative N-terminal signal sequence, it is likely that Tk-SP is secreted to the external medium in a proform termed Pro-Tk-SP, similar to bacterial subtilisins . The source organism of Tk-SP, T. kodakaraensis, was isolated from sediments and seawater samples from a solfatara at a wharf of Kodakara Island (Kagoshima, Japan) , suggesting that the concentration of the Ca2+ ions in its growing environment is similar to that in seawater (10 mm). Therefore, the maturation process of Pro-Tk-SP might be terminated at the step in which the N-propeptide is autoprocessed and degraded in a natural environment, as are those of bacterial pro-subtilisins [3–5] and Pro-Tk-subtilisin [14,15,19,20].
We have previously shown that Pro-Tk-SP accumulates in the cytoplasm of E. coli cells in three soluble forms with molecular masses of 65, 55 and 44 kDa . The 65, 55 and 44 kDa proteins are Pro-Tk-SP, Pro-Tk-SP derivative with N-terminal truncation, and Pro-Tk-SP derivative with N- and C-terminal truncations, respectively. Because the 44 kDa protein is enzymatically active and is as stable as Tk-subtilisin, this protein has been regarded as a mature form of Pro-Tk-SP and designated as Tk-SP. However, the concentration of free Ca2+ ion in the cytoplasm of E. coli, which is reported to be 0.6 μm , is probably too low to prevent C-terminal truncation. ProC-Tk-SP would be produced as a mature form, if a system in which Pro-Tk-SP is folded and matured in the presence of ≥ 1 mm Ca2+ were developed.
Role of C-domain
Identification of ProC-Tk-S359C as a mature form of Pro-Tk-S359C suggests that the C-terminal region autoprocessed in the absence of Ca2+ does not function as a propeptide but exists as one of the domains of ProC-Tk-S359C. In this section, we have changed the terminology for this region from ‘C-propeptide’ to ‘C-domain’. The C-domain is not important for activity but is important for the stability of ProC-Tk-SP. We have previously shown that the β-jelly roll domain contributes to the stabilization of Tk-SP only in the presence of Ca2+ because the stability of Tk-S359A is reduced by 29.5 °C in Tm on removal of this domain in the presence of Ca2+ but is not seriously changed by such a removal in the absence of Ca2+ . The β-jelly roll domain contains two Ca2+ binding sites. Binding of the Ca2+ ions to these sites is responsible for stabilization of Tk-SP, and probably for stabilization of ProC-Tk-SP. Similarly, the C-domain contributes to the stabilization of ProC-Tk-SP by 7.5 °C in Tm in the presence of Ca2+. Binding of the Ca2+ ion(s) to the C-domain may be responsible for the stabilization of ProC-Tk-SP. These results suggest that ProC-Tk-SP acquires high stability by attaching not only the β-jelly roll domain, but also the C-domain. However, the reason why the C-domain contributes to the stabilization of ProC-Tk-SP more significantly in the absence of Ca2+ than in the presence of Ca2+ remains to be clarified because the C-domain in a Ca2+-free form is less stable than that in a Ca2+-bound form. Crystallographic studies of the active-site mutants of Pro-Tk-SP and ProC-Tk-SP are in progress that aim to determine the role of the C-domain.
Database searches indicate that subtilisin-like serine proteases from Pyrococcus furiosus DSM 3638 (accession number NP579399), Thermococcus onnurineus NA1 (accession number YP002307742), Thermococcus sp. AM4 (accession number ZP_04880580), Thermococcusgammatolerans EJ3 (accession number YP002958734) and Bacillus megaterium DSM 319 (accession number YP_003595559) also contain the C-domains, which show amino acid sequence identities of 88%, 82%, 73%, 58 and 31%, respectively, to that of ProC-Tk-SP. All of them also contain the β-jelly roll domain between the subtilisin domain and the C-domain, and all of their source organisms, except for B. megaterium, are hyperthermophiles. Therefore, these regions probably contribute to the stabilization of these proteases as well. These hyperthermophilic proteases probably adapt to high-temperature environment by attaching not only the β-jelly roll domain, but also the C-domain to their C-termini.
Construction of plasmids
The pET25b derivatives for overproduction of Pro-Tk-S359C and Tk-S359C were constructed previously . For construction of the pET25b derivative for overproduction of ProC-Tk-S359C, the gene encoding ProC-Tk-S359C was amplified by PCR with a combination of primer 1 (5′-GGCCTTATCATATGGTTGAGACCGAGG-3′) and primer 2 (5′-GGCCTTGGATCCTCACCCGTAGTAAAC-3′), where the NdeI and BamHI sites for primers 1 and 2, respectively, are underlined. The pET25b derivative for overproduction of Pro-Tk-S359C was used as a template. The resultant DNA fragment was digested with NdeI and BamHI, and ligated into the NdeI-BamHI sites of pET25b. For construction of the pET28a derivative for overproduction of a His-tagged form of ProC (ProC*), the gene encoding ProC* was amplified by PCR with a combination of primer 3 (5′-GGCCTTTATCATATGGACGAGAAGACC-3′), where the NdeI site is underlined, and primer 2. The pET25b derivative for overproduction of Pro-Tk-SP was used as a template. The resultant DNA fragment was digested with NdeI and BamHI, and ligated into the NdeI-BamHI sites of pET28a. All DNA oligomers for PCR were synthesized by Hokkaido System Science (Hokkaido, Japan). PCR was performed for 25 cycles using a thermal cycler (Gene Amp PCR System 2400; Applied Biosystems, Tokyo, Japan) and KOD DNA polymerase (Toyobo Co., Ltd, Kyoto, Japan). The DNA sequences of the genes encoding Pro-Tk-S359C, ProC-Tk-S359C and ProC* were confirmed using an ABI Prism 310 DNA sequencer (Applied Biosystems).
Overproduction and purification
Tk-SP was overproduced in E. coli and purified as described previously . Pro-Tk-S359C, ProC-S359C, Tk-S359C and ProC* were overproduced in E. coli using the transformants of E. coli BL21-codonPlus(DE3)-RIL (Stratagene, La Jolla, CA, USA) with the pET25b derivatives (for Pro-Tk-S359C, ProC-S359C and Tk-S359C) or pET28a derivative (for ProC*), as described previously for Tk-SP . All of the subsequent purification procedures were carried out at 4 °C. Pro-Tk-S359C and Tk-S359C were purified as described previously for Pro-Tk-S359A , except that the HiTrap Q (GE Healthcare) column chromatography was performed at pH 9.0 in the presence of 5 mm EDTA (for both proteins), the protein was eluted from this column by linearly increasing the NaCl concentration from 0 to 2 m (for Pro-Tk-S359C), and the protein was not precipitated by ammonium sulfate prior to column chromatography (for Tk-S359C).
For purification of ProC-Tk-S359C, cells were suspended in 20 mm Tris–HCl (pH 8.0), disrupted by sonication and centrifuged at 30 000 g for 30 min. The pellet was collected and dissolved in 20 mm Tris–HCl (pH 9.0) containing 8 m urea and 5 mm EDTA (buffer A). After removing insoluble materials by centrifugation at 30 000 g for 30 min, the resultant solution was applied to a column (5 mL) of Hitrap Q (GE Healthcare) equilibrated with buffer A. The protein was eluted from the column by linearly increasing the NaCl concentration from 0 to 1.0 m in buffer A (20 column volumes in total). The fractions containing the protein were pooled, dialyzed against 20 mm Tris–HCl (pH 9.0) for refolding, and applied to a column (5 mL) of Hitrap Q equilibrated with the same buffer. The protein was eluted from the column by linearly increasing the NaCl concentration from 0 to 2.0 m in the same buffer (20 column volumes in total). The fractions containing the protein were pooled, dialyzed against 20 mm Tris–HCl (pH 9.0), concentrated using the Centricon (Millipore, Billerica, MA, USA) ultracentrifugation system, and loaded onto a Sephacryl S-200HR column (GE Healthcare) equilibrated with the same buffer containing 150 mm NaCl. The fractions containing the purified protein were collected and dialyzed against 20 mm Tris–HCl (pH 7.5).
For purification of ProC*, cells were suspended in 20 mm Tris–HCl (pH 7.5), disrupted by sonication and centrifuged at 30 000 g for 30 min at 4 °C. The supernatant was applied to a HiTrap Chelating HP column (5 mL) (GE healthcare) charged with Ni2+ ions. The protein was eluted from the column with a linear gradient of imidazole from 30 to 500 mm in 20 mm Tris–HCl (pH 7.5) at a flow rate of 2 mL·min−1. The fractions containing the protein eluted at an imidazole concentration of ∼ 500 mm were pooled, dialyzed against 20 mm Tris–HCl (pH 9.0), and applied to a Sephacryl S-200HR column (GE Healthcare) equilibrated with 50 mm Tris–HCl (pH 9.0) containing 100 mm NaCl. The fractions containing the purified protein were collected and dialyzed against 20 mm Tris–HCl (pH 7.5). Elution of all proteins was performed at a flow rate of 1 mL·min−1 for gel filtration column chromatography and 2 mL·min−1 for other column chromatographies.
The N-terminal amino acid sequence of the protein was determined by a Procise automated sequencer model 491 (Applied Biosystems). The protein concentration was determined from UV absorption of a 0.1% (1.0 mg·mL−1) solution using A280 of 1.80 for Pro-Tk-S359C, 1.99 for ProC-Tk-S359C, 1.84 for Tk-S359C and 2.58 for ProC*. These values were calculated by using absorption coefficients of 1526 m−1·cm−1 for tyrosine and 5225 m−1·cm−1 for tryptophan at 280 nm . The purity of the protein was analyzed by SDS/PAGE using a 17% (for ProC*) or 15% (for other proteins) polyacrylamide gel  followed by staining with CBB. The sample was precipitated by 10% (v/v) trichloroacetic acid (TCA) for complete inactivation prior to this analysis.
The molecular mass of the protein was determined by MALDI-TOF MS (Autoflex or Ultraflex, Bruker Daltonik GmbH, Bremen, Germany). Mass calibration was performed using protein calibration standard II (Bruker Daltonik GmbH). The molecular mass of the protein was also estimated by gel filtration column chromatography using TSK-GEL G2000SWXL (Tosoh Co., Tokyo, Japan). BSA (67 kDa), ovalbumin (44 kDa), chymotrypsinogen A (25 kDa) and RNase A (14 kDa) were used as standard proteins.
The far-UV (200–260 nm) CD spectra of the proteins were measured at 25 °C on a J-725 automatic spectropolarimeter of Japan Spectroscopic Co., Ltd. (Tokyo, Japan). The protein was dissolved in 20 mm Tris–HCl (pH 7.5) or the same buffer containing 10 mm CaCl2. The protein concentration and optical path length were 0.1 mg·mL−1 and 2 mm. The mean residue ellipticity (θ), which has the units of deg·cm−2·dmol−1, was calculated by using an average amino acid relative molecular mass of 110.
The enzymatic activity was determined at 80 °C by using azocasein (Sigma Chemical Co., St Louis, MO, USA) as a substrate. The reaction mixture (300 μL) contained 50 mm Tris–HCl (pH 7.5), 10 mm CaCl2 and 2% (w/v) azocasein. The enzymatic reaction was initiated by adding an appropriate amount of the enzyme and terminated by adding 200 μL of 15% (v/v) TCA. The reaction time was typically 30 min. After centrifugation at 15 000 g for 15 min, an aliquot of the supernatant (160 μL) was withdrawn, mixed with 40 μL of 2 m NaOH, and A440 was measured. One unit of enzymatic activity was defined as the amount of enzyme that increased the A440 value of the assay reaction mixture by 0.1 in 1 min. The specific activity was defined as the enzymatic activity per milligram of protein.
Thermal denaturation curves of the proteins were obtained by plotting the change in CD values at 222 nm against increasing temperature. The protein was dissolved in 20 mm Tris–HCl (pH 7.5) containing 2.5 m GdnHCl in the absence of CaCl2 or the same buffer containing 6 m GdnHCl and 10 mm CaCl2. When thermal denaturation of the protein was analyzed in the absence of Ca2+, the protein was dissolved in the buffer containing 10 mm EDTA and dialyzed against the buffer, which contains neither EDTA, nor Ca2+ prior to the analysis. The protein concentration and optical path length were 0.1 mg·mL−1 and 2 mm, respectively. The linear rate for temperature increase was ∼ 1.0 °C·min−1. The thermal denaturation processes of the proteins were reversible under these conditions. The thermal denaturation curves were normalized, assuming a linear temperature dependence of the base lines of native and denatured states. Tm was calculated from the resultant normalized curves on the basis of a least squares analysis.
This work was supported in part by a Grant (21380065) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and by an Industrial Technology Research Grant Program from the New Energy and Industrial Technology Development Organization (NEDO) of Japan. The study represents a portion of the dissertation submitted by Nitat Sinsereekul to Osaka University in partial fulfilment of the requirement for his PhD.