A novel nucleoside kinase from Burkholderia thailandensis

A member of the phosphofructokinase B-type family of enzymes Hiroko Ota1, Shin-ichi Sakasegawa1, Yuko Yasuda1, Shigeyuki Imamura1 and Tomohiro Tamura2,3

1 Asahi Kasei Pharma Corporation, Shizuoka, Japan 2 Research Institute of Genome-based Biofactory, National Institute of Advanced Industrial Sciences and Technology (AIST), Sapporo, Japan 3 Laboratory of Molecular Environmental Microbiology, Graduate School of Agriculture, Hokkaido University, Japan

Keywords adenosine kinase; Burkholderia thailandensis; inosine guanosine kinase; nucleoside kinase; phosphofructokinase B

Correspondence S. Sakasegawa, Asahi Kasei Pharma Corporation, Shizuoka 410-4321, Japan Fax: +81 0558 76 7149 Tel: +81 0558 76 8564 E-mail: sakasegawa.sb@om.asahi-kasei.co.jp

(Received 13 August 2008, revised 24 September 2008, accepted 29 September 2008)

The genome of the mesophilic Gram-negative bacterium Burkholderia thai- landensis contains an open reading frame (i.e. the Bth_I1158 gene) that has been annotated as a putative ribokinase and PFK-B family member. Nota- bly, although the deduced amino acid sequence of the gene showed only 29% similarity to the recently identified nucleoside kinase from hyperthermophilic archaea Methanocaldococcus jannaschii, 15 of 17 residues reportedly involved in the catalytic activity of M. jannaschii nucleoside kinase were conserved. The gene was cloned and functionally overexpressed in Rhodococcus erythropolis, and the purified enzyme was characterized biochemically. The substrate specificity of the enzyme was unusually broad for a bacterial PFK-B protein, and the specificity extended not only to purine and purine- analog nucleosides but also to uridine. Inosine was the most effective phos- phoryl acceptor, with the highest kcat ⁄ Km value (80 s)1Æmm)1) being achieved when ATP served as the phosphoryl donor. By contrast, this enzyme exhib- ited no activity toward ribose, indicating that the recombinant enzyme was a nucleoside kinase rather than a ribokinase. To our knowledge, this is the first detailed analysis of a bacterial nucleoside kinase in the PFK-B family.

detail both functionally and structurally. The latter was deemed to be a bifunctional 6-phosphofructokinase ⁄ NK because it showed 6-phosphofructokinase activity as well as NK activity. Other PFK-B family members include adenosine kinase (ADK, EC 2.7.1.20), inosine- fructokinase guanosine kinase (IGK, EC 2.7.1.73), (EC 2.7.1.4), 1-phosphofructokinase (EC 2.7.1.56) and RK (EC 2.7.1.15), to name a few. Although the amino acid sequence similarities of these proteins are < 30%, their 3D structures are similar [4–10]. MjNK also has low sequence similarity to the other PFK-B family enzymes, but showed the highest structural similarity to Escherichia coli RK [1].

Burkholderia thailandensis

Nucleoside kinases (NK, no EC number) catalyze the phosphorylation of a variety of nucleosides to the corre- sponding nucleoside 5¢-monophosphate in the presence of phosphate donors and divalent cations [1]. NKs belong to the phosphofructokinase B (PFK-B) family (Pfam PF00294), which is part of the ribokinase (RK)- like superfamily (Pfam CL0118). Recently, two archaeal PFK-B family member proteins, from Methanocaldo- coccus jannaschii and Aeropyrum pernix, have been biochemically characterized [2,3]. The former was deemed to be a NK (MjNK) because it catalyzed the phosphorylation of both purine and pyrimidine nucleo- sides with unusually broad substrate specificity, but showed extremely low activity toward fructose 6-phos- phate. The 3D structure of MjNK had already been solved [1], making it the first NK to be characterized in

is a b-proteobacterium closely related to two other species, Burkholderia mallei and Burkholderia pseudomallei. Whereas the latter two

doi:10.1111/j.1742-4658.2008.06716.x

FEBS Journal 275 (2008) 5865–5872 ª 2008 The Authors Journal compilation ª 2008 FEBS

5865

Abbreviations ADK, adenosine kinase; BthNK, Burkholderia thailandensis nucleoside kinase; IGK, inosine-guanosine kinase; IMPDH, inosine 5¢- monophosphate dehydrogenase dehydrogenase; MjNK, Methanocaldococcus jannaschii nucleoside kinase; NK, nucleoside kinase; PFK-B, phosphofructokinase B; RK, ribokinase.

cause

in archaeal members of

sequences

strepton. The recombinant enzyme was purified from a the transformed R. erythropolis as 40 L culture of described in Experimental procedures, and (cid:2) 190 mg purified recombinant enzyme was obtained. The overall recovery of enzyme activity was (cid:2) 13%. A typical result for purification of the recombinant BthNK from an extract of the R. erythropolis transformant is summa- rized in Table 1. When subjected to SDS ⁄ PAGE, the purified enzyme migrated as a single protein band that corresponded to 34 kDa (Fig. 2A), which was in good agreement with the calculated molecular mass (Table 2). In addition, gel filtration of the native enzyme revealed a molecular mass of 64 kDa (Fig. 2B), suggesting the protein exists as a homodimer (Table 2). This NK (BthNK) thus purified from R. erythropolis was used for all subsequent experiments.

Biochemical characteristics of BthNK

that

to humans, serious health hazards species B. thailandensis is a non-pathogenic bacterium. The entire genomes of these bacteria have been analyzed in an effort to understand the genomic factors that con- tribute to this disparity [11]. The sequencing of the genome of B. thailandensis revealed that it harbors the Bth_I1158 gene, which putatively encodes a RK. Pro- teins belonging to the PFK-B family have a conserved glycine–glycine dipeptide [12], an NXXE motif [13] and an anion-hole sequence motif [13,14]. In addition, multiple sequence alignment revealed two additional consensus the PFK-B family. When we analyzed the Bth_I1158 gene, we found that it contained the archaeal consensus domains in addition to the PFK-B fingerprint motifs. Moreover, 15 of 17 amino acids reported to be involved in the catalytic activity of MjNK were con- served in the deduced sequence of Bth_I1158, despite the fact the amino acid sequence similarity between the two proteins is only 29% (Fig. 1).

In this study, we heterologously overexpressed the Bth_I1158 gene from B. thailandensis in Rhodococ- cus erythropolis and biochemically characterized the gene product. We found that the enzyme acted as a NK (BthNK) showing broad phosphorylation activity toward nucleosides in the presence of ATP and Mg2+. To our knowledge, BthNK is the first identified bacte- rial NK in the PFK-B family.

H. Ota et al. Nucleoside kinase from Burkholderia thailandensis

Results

Overproduction and purification of the NK from B. thailandensis (BthNK)

The biochemical properties of BthNK are summarized in Table 2. The activity of BthNK increased as the temperature was increased from 15 to (cid:2) 50 (cid:2)C; at 65 (cid:2)C, the activity plateaued, likely due to decomposi- tion of the substrate, cofactors, inosine 5¢-monophos- phate dehydrogenase dehydrogenase (IMPDH) or BthNK. Arrhenius plots were linear for temperatures between 20 and 45 (cid:2)C, and an activation energy of 64 kJÆmol)1 was calculated from the slope. In addition to ATP, ITP, TTP and GTP also served as effective phosphate donors. Enzyme activity also required the presence of divalent cations. Mn2+ was the most effi- cient, followed by Mg2+, Ni2+ and Co2+. Maximum activity was attained when the ATP ⁄ Mg2+ concentra- tion ratio was 0.5. The pH optimum for the enzyme

Table 1. Purification of nucleoside kinase from B. thailandensis heterologously produced in R. erythropolis.

Step Total activity (kU) Total protein (g) Yield (%) Specific activity (UÆmg)1)

18

The Bth_I1158 gene consists of 936 nucleotides encoding a deduced amino acid sequence of 312 residues with a calculated molecular mass of 34 kDa. After constructing a Bth_I1158 expression vector, pET21a(+) ⁄ BthNK, we transformed E. coli BL21(DE3) cells, but the transfor- mants barely grew either on solid or in liquid media. To improve growth, we evaluated several combinations of expression vectors, host cells and cultivation conditions. We found that R. erythropolis transformed with pTip QC2 ⁄ BthNK grew very well [15,16], and a large amount of polypeptide was produced after induction with thio-

Cell extract Q sep. BB DEAE sep. ff 1.6 0.63 0.21 3.9 0.19 0.09 0.16 1.1 100 39 13

FEBS Journal 275 (2008) 5865–5872 ª 2008 The Authors Journal compilation ª 2008 FEBS

5866

Fig. 1. Comparison of the amino acid sequence of the potential active sites of BthNK, a NK from B. thailandensis; MjNK, a NK from M. jann- aschii (PDB code 2c4e); and EcRK, a RK from E. coli (1rka). The lower case characters r, b, n, a and m indicate potential coordination sites for the nucleoside ribose, the nucleoside base, both the ribose and base of the nucleoside, ATP and Mg2+, respectively.

H. Ota et al. Nucleoside kinase from Burkholderia thailandensis

A

B

from Saccharomyces cerevisiae

was (cid:2) 6, and the enzyme activity was stable at pH values ranging from 6 to 10. In 100 mm potassium phosphate at pH 7.0, BthNK was fully stable for 20 min at 35 (cid:2)C, lost 60% of its activity at 60 (cid:2)C, and was completely inactivated by incubation at 70 (cid:2)C. Unlike ADKs or humans [17,18], this enzyme did not require the pres- ence of either thiol compounds or salt for stability.

Substrate specificity and kinetic examination of BthNK

dehydrogenas lactate

Fig. 2. (A) SDS ⁄ PAGE analysis of NK from B. thailandensis. Approximately 5 lg of protein were loaded onto a 5–25% gradient gel. After electrophoresis, the gel was stained with Coomassie Bril- liant Blue R250. (B) Molecular mass analysis of the native NK. The molecular mass standard (open circles, Oriental yeast, Osaka): gutamate dehydrogenase (290 kDa), (142 kDa), enolase (67 kDa), myokinase (32 kDa) and cytochrome (12.4 kDa).

Table 2. Biochemical and kinetic properties of nucleoside kinase from B. thailandensis. V, velocity (lmolÆmin)1Æmg)1).

Parameter

Molecular mass

Native enzyme (kDa) Calculated (Da) 64 33 994 64

Arrhenius activation energya (kJÆmol)1, 20–45 (cid:2)C) Substrate specificity (% of V ) 100, 49, 24, 12, 10, 8, 2

Inosine > Adenosin > Guanosine > Mizoribine, 2-deoxy-adenosine, Uridine > Rivabirin Cation specificityb (% of V ) 100, 90, 86, 85

Mn2+ > Mg2+ > Ni2+, Co2+ Phosphate donorsc (% of V ) 200, 130, 125, 100, 45, 40 ITP > TTP, GTP > ATP > CTP, UTP

The ATP-dependent reactivity of sugars, sugar-phos- phates, and nucleosides (all at 1 mm) were tested in assays making use of pyruvate kinase and lactate dehy- In the presence of d-ribose, drogenase at 37 (cid:2)C. l-ribose, fructose, glucose, 2-deoxy-glucose, xanthosine, galactose, xylose, N-acetyl glucosamine, fructose-6- phosphate, mannose-6-phosphate and ribose-5-phos- phate, BthNK activity was below the detection limit; however, significantly higher levels of activity were observed in the presence of nucleosides. Of the nucleo- sides tested, inosine was the most preferred substrate, followed by adenosine, guanosine, mizoribine, 2-deoxy- adenosine, uridine and rivabirin. By contrast, no activ- ity was seen with either cytidine or thymidine (Table 2). This substrate specificity was confirmed by HPLC anal- ysis. We suggest the reaction product of inosine phos- phorylation with BthNK is IMP, as enzymatic activity was detected in a coupled assay with IMPDH. In addi- tion, HPLC analyses in which the retention times of various products of BthNK-catalyzed nucleoside phos- phorylation reactions were compared with standards revealed the products to be the corresponding nucleo- side 5¢-monophosphates. Figure 3 shows an example of the phosphorylation of mizoribine by BthNK. Two sig- nals corresponding to mizoribine 5¢-monophosphate and ADP, respectively, were observed after the reac- tion. Moreover, when a sample was injected together

pH optimumd pH stabilitye Thermostabilityf 5.5–6.5 6–10 (cid:2) 50 (cid:2)C

A

B

a Activation energy was calculated from the slopes of the linear parts in the Arrhenius plot (20–45 of 15–65 (cid:2)C). b The divalent cat- ion requirement was examined by exchanging Mg2+ (20 mM) for alternative cations. c The phosphoryl donor specificity was tested by replacing ATP (10 mM) for alternative donors. d To cover the whole testing pH range, the following buffer systems overlap each other were used for the pH optimum determination; citrate-NaOH (pH 4.5–6.0), potassium phosphate (pH 6.0–7.5) and Tris ⁄ HCl (pH 7.0–9.0). e Aliquots of 0.5 mg enzyme per mL in the same buffer systems used to determine the pH optimum and glycine ⁄ NaOH (pH 9.5–11.0) buffer (all at 100 mM) were incubated for 3 h at 37 (cid:2)C. After incubation, the remaining activities were analyzed. f Aliquots of 0.5 mg enzyme per mL in 100 mM potassium phos- phate (pH 7.0) were incubated in sealed tubes for 20 min at tem- peratures 0–80 (cid:2)C. After incubation, all tubes were rapidly cooled in an ice bath and analyzed for activity.

FEBS Journal 275 (2008) 5865–5872 ª 2008 The Authors Journal compilation ª 2008 FEBS

5867

Fig. 3. HPLC analyses of mizoribine phosphorylation activity. (A) The enzyme-free control, (B) after the reaction.

with authentic mizoribine 5¢-monophosphate and ADP, the observed peaks were enhanced.

the parameters for inosine and ATP in an IMP-forming assay using a coupling reaction catalyzed by IMPDH. Initial velocities of inosine phosphorylation with ATP were determined by running the reaction with several fixed levels of ATP and varying the concentration of inosine; the ATP ⁄ Mg2+ ratio was maintained at 0.5 throughout the experiment. Under these conditions, BthNK showed typical Michaelis–Menten kinetics for both inosine and ATP, and the kinetic parameters were estimated using two-substrate kinetic analysis (Fig. 4, Table 3) [19].

We then carried out kinetic analyses of the reactions toward these substrates by measuring the production of ADP coupled with pyruvate kinase and lactate dehy- drogenase. Reactions with ATP and Mg2+, guanosine, mizoribine, 2-deoxy-adenosine, uridine and rivabirin followed Michaelis–Menten kinetics under our assay conditions, and the apparent Km and Vmax values were estimated successfully (Table 3). With inosine and adenosine, however, concentrations > 0.1 mm led to a reduction in enzyme activity. We therefore re-examined

H. Ota et al. Nucleoside kinase from Burkholderia thailandensis

Discussion

Table 3. Kinetic parameters of nucleoside kinase from B. thailand- ensis.

)1)

Substrate Km (mM) Vmax (lmolÆmin)1Æmg)1) Kcat ⁄ Km (s)1ÆmM

18 18 80 16

The Bth_I1158 gene from B. thailandensis was cloned and functionally overexpressed in R. erythropolis. The E. coli transformants harboring expression vectors for Bth_I1158, e.g. pET21a(+) ⁄ BthNK, barely grew on solid or in liquid media, even before induction, proba- bly due to disturbance of intracellular nucleoside levels by the expressed protein. When the gene was expressed in R. erythropolis, however, (cid:2) 190 mg of the Bth_I1158 gene product was purified per 270 g wet mass of cell. Thus the recombinant protein expression system using R. erythropolis proved to be a profitable ‘biofactory’ and appears superior to E. coli bacterial systems in such cases. To date, a number of proteins that could not be expressed in E. coli have been func- tionally expressed using this protocol (including unpublished data) [15,16,20].

a Michaelis–Menten constants for inosine and ATP were estimated using two-substrate kinetic analysis (Fig. 4) [19]. The activities were determined by coupling the IMP-forming reaction with the reaction catalyzed by IMPDH. The inosine and ATP concentrations ranged from 0.133 to 0.666 and 0.333 to 1.333 mM, respectively. The MgCl2 concentration was twice the ATP concentration. b The appar- ent Vmax and Km values were examined by measuring the produc- tion ADP with pyruvate kinase and lactate dehydrogenase (see Experimental procedures). Substrate concentrations ranged from 0.0015 to 0.01 mM for adenosine and 0.05 to 1.00 mM for the other substrates.

The purified Bth_I1158 gene product was found to be a NK (BthNK). The recombinant protein had a native molecular mass of 64 kDa (Fig. 2B) and a subunit size of 34 kDa (Fig. 2A), indicating it to be a typical PFK- B family enzyme in terms of both its subunit size and its homodimeric structure. The quaternary structures of

Inosinea ATPa Adenosineb Guanosineb Mizoribineb 2-Deoxy-adenosineb Uridineb Ribavirinb 0.19 0.20 0.87 1.7 0.83 0.12 0.15 0.63 0.02 0.46 0.87 1.1 1.7 0.79 6.05 0.24 0.04 0.84 0.28 0.08

A

B

C

D

FEBS Journal 275 (2008) 5865–5872 ª 2008 The Authors Journal compilation ª 2008 FEBS

5868

Fig. 4. (A) Double-reciprocal plots of activity versus inosine concentration at 0.333 mM (open circles), 0.500 mM (open triangles), 0.666 mM (open rectangles), 1.000 mM (filled circles) and 1.333 mM (filled triangles) of ATP. (B) Double-reciprocal plots of activity versus ATP concen- tration at 0.133 mM (open circles), 0.200 mM (open triangles), 0.250 mM (open rectangles), 0.500 mM (filled circles) and 0.666 mM (filled triangles) of inosine. (C) Double-reciprocal plots of the apparent Vmax versus ATP (filled circles) or inosine (open circles) concentrations. Data were taken from (A) and (B). (D) Effect of the inorganic phosphate (Pi) concentration on the activity of BthNK. The activities were examined under standard assay conditions while varying the potassium phosphate (pH 7.0) concentration.

respective preferences

ADK and MjNK [2,21,22]. Both E. coli RK and human ADK are absolutely dependent on the presence of inor- ganic phosphate (Pi) for activity [26], although Pi stimu- lated little or no activity by M. tuberculosis ADK [21]. When we examined the effect of Pi on BthNK activity, we found that although Pi is not essential, it does stimu- late BthNK activity (Fig. 4D). For half-maximal activ- ity, (cid:2) 70 mm K2HPO4 was required. We also compared for phosphate donors the between BthNK and MjNK. BthNK preferred ITP, TTP and GTP, with which it respectively showed 2.0-, 1.3- and 1.2-fold greater activity than it did with ATP. TTP being a better substrate than ATP, considering the fact that ATP is present in cells at a maximum concen- tration of (cid:2) 2 mm, whereas the TTP concentration may be (cid:2) 100-times lower. By contrast, MjNK exhibited its highest level of activity with ATP and showed less than half as much activity with ITP or GTP, and < 10% as much activity with UTP or CTP [2].

including MjNK,

sequence

the anion-hole motif

is

the active forms of RK-like superfamily proteins are monomeric to tetrameric, though most PFK-B family enzymes are described as monomeric or homodimeric enzymes [10]. BthNK showed activity toward purine and purine-analog nucleosides (inosine, adenosine, gua- nosine, mizoribine, 2-deoxy-adenosine and rivabirin); no activity was detected toward either cytidine or thy- midine. In addition, although the catalytic efficacy (kcat ⁄ Km) was 286-fold lower than that of using inosine as a substrate, BthNK showed some basal activity toward uridine (Table 3). Among PFK-B family enzymes, ADKs from S. cerevisiae [17], Mycobacte- rium tuberculosis [21] and Toxoplasma gondii [5] have defined activities toward adenosine. Currenty known enzymes with relatively broad substrate specificities include E. coli IGK, mammalian ADK, MjNK and 6-phosphofructokinase ⁄ NK from A. pernix. The phos- phate acceptors of E. coli IGK [22], human ADK [18] and rabbit liver ADK [23] are all purine nucleosides or purine nucleoside analogs. Only the PFK-B family enzymes MjNK and the 6-phosphofructokinase ⁄ NK from A. pernix show purine and pyrimidine kinase activities [2]. MjNK utilizes cytidine, guanosine and inosine, but exhibits little activity toward adenosine, 2-deoxy-adenosine and thymidine. Although ADK, IGK and deoxy-cytidine kinase have all been isolated from various cellular extracts, an NK-like BthNK had not yet been purified [24,25]. To our knowledge, there- fore, BthNK is the first example of a bacterial NK from the PFK-B family that catalyzes both purine and pyrimidine nucleosides. Although the physiological sig- nificance of BthNK function is currently unknown, its broad activity toward nucleosides suggests it may play a role in nucleoside metabolism. The Km values of BthNK (at 37 (cid:2)C) towards inosine, guanosine and uri- dine were 6–10-fold greater than the values observed for MjNK at 50 (cid:2)C and half that observed for MjNK toward 2-deoxy-adenosine [2]. At 37 (cid:2)C, MjNK showed little or no activity [2]. Measurements of ADP showed that BthNK activity was inhibited by excess inosine or adenosine (> 0.1 mm). A similar property was observed for human ADK under the same assay condi- tions [18]. Interestingly, BthNK activity was not inhib- ited by inosine in a coupled assay with IMPDH (Fig. 4), and a Vmax of 18 lmolÆmin)1Æmg)1 and Km values of 0.15 and 0.63 mm toward inosine and ATP, respectively, were calculated. Thus, the inhibition of ADK by inosine likely reflects the experimental assay conditions rather than the properties of the enzyme. The ATP ⁄ Mg2+ ratio plays a critical role in regulating the activity of PFK-B family enzymes. The activity of BthNK was maximum when the ATP ⁄ Mg2+ ratio was 0.5, which is similar to E. coli IGK, M. tuberculosis

The amino acid alignment showed that although BthNK possesses < 30% sequence similarity with the sub- PFK-B family enzymes, strate-, cofactor- and divalent cation-binding sites are highly conserved (Fig. 1) [1]. The highly conserved sequences include three consensus motifs found in the RK-like superfamily and the PFK-B family. One of the fingerprint regions of the RK-like superfamily is the glycine–glycine dipeptide (G47, G48 in BthNK; Fig. 1) [12]. This dipeptide is believed to play an important role in forming the closed conformation of the enzyme and in subsequent substrate sequestra- tion [27]. The NXXE sequence is another motif (N192– E195 in BthNK; Fig. 1) that is highly conserved in the RK-like superfamily and has been identified as impor- tant for the binding of both Mg2+ and ATP [13]. The most highly conserved region among the superfamily members (G250–D253 in BthNK; Fig. 1), which helps to neutralize negative charges accumulated during phosphate group transfer [13,14]. Two additional consensus sequences were found in BthNK and MjNK (Fig. 1). The A103–F118 region of BthNK corresponds to the lid domain of MjNK, and the D166–L171 sequence in BthNK corre- sponds to a region that includes D160 and Q163 in MjNK and is involved in both Mg2+ and ribose bind- ing in the active site [1]. Thus 15 of 17 residues reported to be involved in the catalytic activity of MjNK are conserved in BthNK. The amino acid sequence of BthNK is highly conserved in Burkholderia spp., with 80–99% homology. Given that BthNK showed no activity toward the tested sugars, but strong activity toward the nucleosides, we propose that these gene products function as NKs, even though the genes are

FEBS Journal 275 (2008) 5865–5872 ª 2008 The Authors Journal compilation ª 2008 FEBS

5869

H. Ota et al. Nucleoside kinase from Burkholderia thailandensis

annotated as ADK, RK and sugar kinase. Hansen et al. showed that PFK-B paralogous sequences might have all evolved from an ancestral sequence because several archaeal PFK-B sequences were found among bacterial or bacterial and eukaryotic PFK-B sequences in the phylogenetic tree [2]. BthNK would be the first bacterial NK to be functionally characterized in this family. The findings of this study will therefore contrib- ute to our understanding of the evolution of PFK-B family enzymes and further their characterization.

H. Ota et al. Nucleoside kinase from Burkholderia thailandensis

Experimental procedures

Materials

system comprised of E. coli

A pET vector strain BL21(DE3) and plasmid DNA pET21a(+) was obtained from Novagen (Madison, WI, USA), and expression trials in the E. coli were performed according to manufacturer’s instructions. Mizoribine (also known as Bredinine(cid:3)) and glucose-6-phosphate dehydrogenase were obtained from Asahi Kasei Pharma (Tokyo, Japan). pTip QC2 vector and R. erythropolis strain L-88 has been developed in our labo- ratory [15,16].

2 days at 30 (cid:2)C in 40 L of liquid Luria–Bertani medium containing 1 mgÆL)1 thiostrepton and 38 mgÆL)1 chloram- phenicol. The cells, which expressed the NK from B. thai- landensis, were harvested by centrifugation, suspended in 20 mm Tris ⁄ HCl (pH 8.0), and then disrupted by ultrasoni- cation on ice (Cell Disruptor; Branson, CT, USA). The cell debris was removed by centrifugation for 20 min at 15 000 g, after which the supernatant was loaded onto a Q Sepharose Big Beads column (1.5 L, BPG100; GE Health- care, Tokyo, Japan) pre-equilibrated with 10 mm Tris ⁄ HCl then washed with 10 mm (pH 8.5). The column was Tris ⁄ HCl (pH 8.5), and the protein was eluted with a linear gradient of 0–0.5 m KCl. The fractions containing NK activity were collected and desalted by passage through a Sephadex G-25 Superfine (GE Healthcare) column pre- equilibrated with 10 mm potassium phosphate (pH 7.0). The resultant protein solution was loaded onto a DEAE Sepharose Fast Flow (250 mL, XK50; GE Healthcare) col- umn pre-equilibrated with 10 mm potassium phosphate (pH 7.0), and NK activity was eluted with a linear gradient of 0–0.5 m KCl. The active fractions were pooled, concen- trated using a 30 kDa centrifugal filter device (Millipore, Bedford, MA, USA), and desalted by passage through a Sephadex G-25 Superfine column pre-equilibrated with 10 mm potassium phosphate (pH 7.0). The entire operation was performed at room temperature ((cid:2) 25 (cid:2)C).

Cloning of the Bth_I1158 gene from B. thailandensis

Determination of the BthNK activity

two enzymatic reactions:

Unless otherwise specified, BthNK activity was assayed by measuring the formation of NADH in a coupled reaction that consisted of (a) BthNK- dependent phosphorylation of inosine to IMP with the dep- hosphorization of ATP to ADP, and (b) oxidation of IMP to xanthosine 5¢-monophosphate with reduction of NAD to NADH by IMP dehydrogenase (EC 1.1.1.205). IMPDH was prepared from Bacillus subtilis as described previously [28]. The reaction mixture contained 30 mm potassium phosphate (pH 7.0), 10 mm ATP, 20 mm MgCl2, 5 mm ino- sine, 5 mm NAD, 50 mm KCl, 1 mm dithiothreitol and 5 UÆmL)1 IMPDH in a total volume of 150 lL. The reac- tion was initiated by the addition of 5 lL ((cid:2) 0.2 mgÆmL)1) of enzyme solution. The rate of NADH generation was fol- lowed spectrophotometrically at 340 nm with d = 1 cm and e340 = 6.2 mm)1Æcm)1. One unit (U) of enzyme is defined as the amount of enzyme that catalyzed the phos- phorylation of 1 lmole inosine per min at pH 7.0 and 37 (cid:2)C.

PCR carried out with the following oligonucleotide primers were used to amplify the B. thailandensis (E264, DSMZ 13276) Bth_I1158 gene: 5¢-TTTCATATGGCTACGCTG ATTTGCGGTTCG-3¢ (sense) and 5¢-TTTGGATCCTCAC TTCGGACGATAGCCGAACGC-3¢ (antisense). The prim- ers contained unique NdeI and BamHI restriction sites, respectively (underlined). The amplified 1 kb fragment was digested with NdeI and BamHI and then ligated into pTip QC2 vector which had been linearized with the same restric- tion enzymes to generate pTip QC2 ⁄ BthNK. E. coli strain DH5a was transformed with pTip QC2 ⁄ BthNK and spread onto an Luria–Bertani–agar plate containing 50 mgÆL)1 ampicillin, after which positive colonies were selected using the colony PCR method using an Insert Check Ready Blue kit (Toyobo, Tokyo, Japan). A positive transformant was then cultured in liquid Luria–Bertani medium containing 50 mgÆL)1 ampicillin, and the plasmid pTip QC2 ⁄ BthNK carrying Bth_I1158 was isolated. pET21a(+) ⁄ BthNK was constructed using the same procedure used for pTip QC2 ⁄ BthNK.

Overproduction and purification of the NK from B. thailandensis

A second BthNK assay was based on a coupled spectro- photometric determination of ADP using phosphoenolpyru- vate, pyruvate kinase (PK; Sigma-Aldrich, St Louis, MO, USA), and lactate dehydrogenase (Sigma-Aldrich). BthNK the substrates coupled catalyzes the phosphorylation of with the hydrolysis of ATP to ADP. PK transfers a phos- phate group from phosphoenolpyruvate to ADP, yielding

R. erythropolis strain L-88 was transformed with pTip QC2 ⁄ BthNK. The transformants were then cultivated for

FEBS Journal 275 (2008) 5865–5872 ª 2008 The Authors Journal compilation ª 2008 FEBS

5870

2 Hansen T, Arnfors L, Ladenstein R & Scho¨ nheit P

(2006) The phosphofructokinase-B (MJ0406) from Met- hanocaldococcus jannaschii represents a nucleoside kinase with a broad substrate specificity. Extremophiles 11, 105–114.

3 Hansen T & Scho¨ nheit P (2001) Sequence, expression, and characterization of the first archaeal ATP-depen- dent 6-phosphofructokinase, a non-allosteric enzyme related to the phosphofructokinase-B sugar kinase family, from the hyperthermophilic crenarchaeote Aeropyrum pernix. Arch Microbiol 177, 62–69.

4 Campobasso N, Mathews II, Begley TP & Ealick SE

pyruvate and ATP. Lactate dehydrogenase catalyzes the conversion from pyruvate to lactate with concomitant con- version from NADH to NAD. The reaction mixture con- tained 50 mm potassium phosphate (pH 7.0), 50 mm KCl, 5 mm ATP, 10 mm MgCl2, 5 mm phosphoenolpyruvate, 0.15 mm NADH, 5 UÆmL)1 PK and 5 UÆmL)1 lactate dehydrogenase in a total volume of 150 lL. The reaction was initiated by the addition of 5 lL (0.4–8.4 mgÆmL)1, depends on the substrates) of enzyme solution. The oxida- tion of NADH linked to ADP generation was determined from the change in absorbance at 340 nm at 37 (cid:2)C. Protein levels were determined with the Bradford dye-binding method using the BioRad protein assay kit. Bovine serum albumin served as the assay protein standard.

(2000) Crystal structure of 4-methyl-5-beta-hydroxyeth- ylthiazole kinase from Bacillus subtilis at 1.5 A˚ resolu- tion. Biochemistry 39, 7868–7877.

5 Darling JA, Sullivan WJ Jr, Carter D, Ullman B &

Molecular mass determination

Roos DS (1999) Recombinant expression, purification, and characterization of Toxoplasma gondii adenosine kinase. Mol Biochem Parasitol 103, 15–23.

6 Mathews II, Erion MD & Ealick SE (1998) Structure of human adenosine kinase at 1.5 A˚ resolution. Biochemis- try 37, 15607–15620.

Molecular mass of the native enzyme was analyzed by gel filtration on a TSK gel column (G3000SWXL; Tosoh, Tokyo, Japan) pre-equilibrated with 50 mm potassium phosphate buffer (pH 7) containing 0.2 m NaCl. The flow rate was 0.4 mLÆmin)1.

HPLC analysis

7 Reddy MC, Palaninathan SK, Shetty ND, Owen JL, Watson MD & Sacchettini JC (2007) High resolution crystal structures of Mycobacterium tuberculosis adeno- sine kinase: insights into the mechanism and specificity of this novel prokaryotic enzyme. J Biol Chem 282, 27334–27342.

8 Sigrell JA, Cameron AD, Jones TA & Mowbray SL

(1998) Structure of Escherichia coli ribokinase in com- plex with ribose and dinucleotide determined to 1.8 A˚ resolution: insights into a new family of kinase struc- tures. Structure 6, 183–193.

9 Spychala J, Datta NS, Takabayashi K, Datta M, Fox

The substrate specificity was confirmed by HPLC analysis. An Asahipak GS-320HQ column (Shodex, Tokyo, Japan) was used, and an isocratic elution was carried out using 200 mm sodium phosphate (pH 3.0) with a flow rate of 0.5 mLÆmin)1. The absorbance of the column effluent was monitored at 260 nm. The reaction mixture contained 30 mm potassium phosphate (pH 7.0), 1 mm substrates, 1 mm ATP and 2 mm MgCl2. The reaction was started by adding 25 lg of enzyme per mL. After 60 min at 37 (cid:2)C, the reaction was stopped by ultrafiltration (Millipore ultrafree- MC 10,000 NMWL filter unit), after which the mixture was injected onto the column.

IH, Gribbin T & Mitchell BS (1996) Cloning of human adenosine kinase cDNA: sequence similarity to micro- bial ribokinases and fructokinases. Proc Natl Acad Sci USA 93, 1232–1237.

10 Zhang Y, Dougherty M, Downs DM & Ealick SE

H. Ota et al. Nucleoside kinase from Burkholderia thailandensis

Acknowledgements

(2004) Crystal structure of an aminoimidazole riboside kinase from Salmonella enterica: implications for the evolution of the ribokinase superfamily. Structure 12, 1809–1821.

11 Kim HS, Schell MA, Yu Y, Ulrich RL, Sarria SH,

We thank Prof. Dr Satoshi Shuto of Faculty of Phar- maceutical Sciences Division, Hokkaido University, Japan for his generous gift of the mizoribine 5¢-mono- phosphate and Prof. Dr Haruhiko Sakuraba of Faculty of Agriculture, Kagawa University, Japan for his scientific advice.

Nierman WC & DeShazer D (2005) Bacterial genome adaptation to niches: divergence of the potential viru- lence genes in three Burkholderia species of different survival strategies. BMC Genomics 6, 174–187.

References

12 Bork P, Sander C & Valencia A (1993) Convergent evo- lution of similar enzymatic function on different protein folds: the hexokinase, ribokinase, and galactokinase families of sugar kinases. Protein Sci 2, 31–40.

1 Arnfors L, Hansen T, Scho¨ nheit P, Ladenstein R & Meining W (2006) Structure of Methanocaldococcus jannaschii nucleoside kinase: an archaeal member of the ribokinase family. Acta Crystallogr D Biol Crystallogr 62, 1085–1097.

13 Maj MC, Singh B & Gupta RS (2002) Pentavalent ions dependency is a conserved property of adenosine kinase from diverse sources: identification of a novel motif

FEBS Journal 275 (2008) 5865–5872 ª 2008 The Authors Journal compilation ª 2008 FEBS

5871

implicated in phosphate and magnesium ion binding and substrate inhibition. Biochemistry 41, 4059–4069.

14 Sigrell JA, Cameron AD & Mowbray SL (1999)

22 Mori H, Iida A, Teshiba S & Fujio T (1995) Cloning of a guanosine-inosine kinase gene of Escherichia coli and characterization of the purified gene product. J Bacte- riol 177, 4921–4926.

Induced fit on sugar binding activates ribokinase. J Mol Biol 290, 1009–1018.

15 Mitani Y, Meng X, Kamagata Y & Tamura T (2005) Characterization of LtsA from Rhodococcus erythropo- lis, an enzyme with glutamine amidotransferase activity. J Bacteriol 187, 2582–2591.

23 Miller RL, Adamczyk DL, Miller WH, Koszalka GW, Rideout JL, Beacham LM III, Chao EY, Haggerty JJ, Krenitsky TA & Elion GB (1979) Adenosine kinase from rabbit liver. II. Substrate and inhibitor specificity. J Biol Chem 254, 2346–2352.

24 Kawasaki H, Usuda Y, Shimaoka M & Utagawa T

16 Nakashima N & Tamura T (2004) A novel system for expressing recombinant proteins over a wide tempera- ture range from 4 to 35 degrees C. Biotechnol Bioeng 86, 136–148.

17 Lu XB, Wu HZ, Ye J, Fan Y & Zhang HZ (2003)

(2000) Development of an improved assay for purine nucleoside kinase activity in cell extracts and detection of inosine kinase activity in Brevibacterium acetylicum ATCC 953, related species, and Corynebacterium flac- cumfaciens ATCC 6887. Biosci Biotechnol Biochem 64, 761–766.

25 Yamada Y, Goto H & Ogasawara N (1983) Purine

Expression, purification, and characterization of recom- binant Saccharomyces cerevisiae adenosine kinase. Acta Biochim Biophys Sin 35, 666–670.

nucleoside kinases in human T- and B-lymphoblasts. Biochim Biophys Acta 761, 34–40.

18 Yamada Y, Goto H & Ogasawara N (1981) Adenosine kinase from human liver. Biochim Biophys Acta 660, 36–43.

19 Breitung J, Borner G, Scholz S, Linder D, Stetter KO

26 Park J, van Koeverden P, Singh B & Gupta RS (2007) Identification and characterization of human ribokinase and comparison of its properties with E. coli ribokinase and human adenosine kinase. FEBS Lett 581, 3211– 3216.

& Thauer RK (1992) Salt dependence, kinetic properties and catalytic mechanism of N-formylmethanofuran:tet- rahydromethanopterin formyltransferase from the extreme thermophile Methanopyrus kandleri. Eur J Bio- chem 210, 971–981.

27 Schumacher MA, Scott DM, Mathews II, Ealick SE, Roos DS, Ullman B & Brennan RG (2000) Crystal structures of Toxoplasma gondii adenosine kinase reveal a novel catalytic mechanism and prodrug binding. J Mol Biol 298, 875–893.

20 Kim WS, Shimazaki K & Tamura T (2006) Expression of bovine lactoferrin C-lobe in Rhodococcus erythropolis and its purification and characterization. Biosci Biotech- nol Biochem 70, 2641–2645.

28 Ota H, Yasuda Y, Sakasegawa S, Imamura S & Tam- ura T (2008) A novel enzymatic method for measuring mizoribine phosphate levels in serum. J Biosci Bioeng 5, in press.

21 Long MC, Escuyer V & Parker WB (2003) Identification and characterization of a unique adenosine kinase from Mycobacterium tuberculosis. J Bacteriol 185, 6548–6555.

FEBS Journal 275 (2008) 5865–5872 ª 2008 The Authors Journal compilation ª 2008 FEBS

5872

H. Ota et al. Nucleoside kinase from Burkholderia thailandensis