Báo cáo khoa học: Serine hydroxymethyltransferase from Plasmodium vivax is different in substrate speciﬁcity from its homologues

Serine hydroxymethyltransferase from Plasmodium vivax is different in substrate speciﬁcity from its homologues Kittipat Sopitthummakhun1,, Somchart Maenpuen1,, Yongyuth Yuthavong2, Ubolsree Leartsakulpanich2 and Pimchai Chaiyen1

1 Department of Biochemistry and Center for Excellence in Protein Structure and Function, Faculty of Science, Mahidol University, Bangkok, Thailand 2 National Center for Genetic Engineering and Biotechnology, National Science and Technology Development Agency, Pathumthani, Thailand

cleavage of

retro-aldol

3-hydroxy

Keywords deoxythymidylate cycle; folate metabolism; Plasmodium vivax; pyridoxal-5-phosphate; serine hydroxymethyltransferase

Correspondence P. Chaiyen, Department of Biochemistry and Center for Excellence in Protein Structure and Function, Faculty of Science, Mahidol University, Rama 6 Road, Bangkok 10400, Thailand Fax: +66 2354 7174 Tel: +66 2201 5596 E-mail: scpcy@mahidol.ac.th U. Leartsakulpanich, National Center for Genetic Engineering and Biotechnology, National Science and Technology Development Agency, 113 Paholyothin Road, Pathumthani 12120, Thailand Fax: +66 2564 6707 Tel: +66 2564 6700 E-mail: ubolsree@biotec.or.th Website: http://www.sc.mahidol.ac.th/ chaiyen_p

The putative gene of Plasmodium vivax serine hydroxymethyltransferase (PvSHMT; EC 2.1.2.1) was cloned and expressed in Escherichia coli. The puriﬁed enzyme was shown to be a dimeric protein with a monomeric molecular mass of 49 kDa. PvSHMT has a maximum absorption peak at 422 nm with a molar absorption coefﬁcient of 6370 m)1Æcm)1. The Kd for binding of the enzyme and pyridoxal-5-phosphate was 0.14 ± 0.01 lm. An alternative assay for measuring the tetrahydrofolate-dependent SHMT activity based on the coupled reaction with 5,10-methylenetetrahydrofolate from E. coli was developed. PvSHMT uses a reductase (EC 1.5.1.20) ternary-complex mechanism with a kcat value of 0.98 ± 0.06 s)1 and Km values of 0.18 ± 0.03 and 0.14 ± 0.02 mm for l-serine and tetrahydro- folate, respectively. The optimum pH of the SHMT reaction was 8.0 and an Arrhenius’s plot showed a transition temperature of 19 (cid:2)C. Besides l-serine, PvSHMT forms an external aldimine complex with d-serine, l-alanine, l-threonine and glycine. PvSHMT also catalyzes the tetrahydro- amino acids. folate-independent Although l-serine is a physiological substrate for SHMT in the tetrahy- drofolate-dependent reaction, PvSHMT can also use d-serine. In the absence of tetrahydrofolate at high pH, PvSHMT forms an enzyme–quino- noid complex with d-serine, but not with l-serine, whereas SHMT from rabbit liver was reported to form an enzyme–quinonoid complex with l-serine. The substrate speciﬁcity difference between PvSHMT and the mammalian enzyme indicates the dissimilarity between their active sites, which could be exploited for the development of speciﬁc inhibitors against PvSHMT.

*These authors contributed equally to this work

(Received 9 February 2009, revised 17 April 2009, accepted 22 May 2009)

doi:10.1111/j.1742-4658.2009.07111.x

Introduction

world. In 2002, 515 (range 300–660) million episodes clinical Plasmodium falciparum malaria were of estimated [1]. Most of malaria cases are caused by

Malaria is a life-threatening disease caused by proto- zoan infection and poses one of the most widespread public health problems, especially in the developing

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4023

Abbreviations apoPvSHMT, apoenzyme of PvSHMT; E–Q, enzyme–quinonoid complex; holoPvSHMT, holoenzyme of PvSHMT; IPTG, isopropyl thio-b-D- galactoside; MTHF, 5,10-methylenetetrahydrofolate; MTHFR, 5,10-methylenetetrahydrofolate reductase; PLP, pyridoxal-5-phosphate; PvSHMT, Plasmodium vivax serine hydroxymethyltransferase; SHMT, serine hydroxymethyltransferase; THF, [6R,S]-tetrahydrofolate.

the investigation of

anti-malarial

compounds, both by

from P. vivax indeed encodes SHMT. An alternative assay for measuring the THF-dependent SHMT activ- ity based on a previously used method [20] was devel- oped and employed for the the puriﬁed biochemical and kinetic properties of enzyme. The results obtained revealed distinct differ- ences in substrate speciﬁcity between PvSHMT and mammalian enzymes, implying that PvSHMT is a possible target for developing speciﬁc inhibitors for anti-malarial therapy.

and

amino

purine,

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

Results and Discussion

for

the parasites.

Protein expression and puriﬁcation of recombinant PvSHMT

synthase

[4]. The

infection of P. falciparum and Plasmodium vivax. Thus far, chemotherapy remains the most effective way to control the disease; however, reports of drugs resis- tance in parasites necessitate efforts to ﬁnd new effec- random tive searches and by identifying and validating novel drug targets. The folate biosynthesis pathway has long been accepted to provide good targets for anti-malarial ther- apy because it comprises a metabolic process that is essential for parasite growth and is different from that the human host, and links to many pathways, of including acids pyrimidine syntheses. Blocking enzymes in this pathway would In Plasmodium be detrimental spp., only two bifunctional enzymes in the pathway, dihydrofolate and reductase-thymidylate hydroxymethyldihydropterin pyrophosphokinase-dihy- dropteroate synthetase, have been extensively investi- gated and validated as good targets for anti-malarial therapy [2]. Another enzyme in the plasmodial deoxy- thymidylate synthesis cycle, serine hydroxymethyltrans- ferse (SHMT; EC 2.1.2.1), has been much less studied. Only SHMT from P. falciparum was cloned and expressed in Escherichia coli, whereas the biochemical properties of the enzyme have not been studied [3]. Recently, the putative gene of SHMT from P. vivax (PvSHMT) was reported and the amino acid sequence of the enzyme was compared with those from various organisms results obtained showed that PvSHMT shares 42% sequence identity with that of the human enzyme, implying that it is possible to design speciﬁc inhibitors against this target.

The sequence of Pv100730, previously identiﬁed as PvSHMT [4], was cloned into pET17b expression vec- tor. The gene was placed under the regulation of the T7 promoter. Expression of PvSHMT was optimum when 1 mm isopropyl thio-b- d-galactoside (IPTG) was added into the culture with D600 (cid:2)1.0 and the culture was maintained at 16 (cid:2)C for the next 16 h. An inducible pro- tein with an approximate molecular mass of 49 kDa was detected on SDS ⁄ PAGE (Fig. 1). This expression level accounted for approximately 10% of the total soluble proteins in the crude extract. Other conditions, such as higher induction temperatures (at 25 or 37 (cid:2)C) or add- ing 1 mm IPTG at a different cell density, decreased the production of the soluble PvSHMT (data not shown). The recombinant PvSHMT was puriﬁed to approxi- mately 90% purity as judged by SDS ⁄ PAGE according to the protocol described in Experimental procedures. Puriﬁcation by DEAE-Sepharose chromatography efﬁ- ciently separated PvSHMT from other host proteins because the enzyme does not bind to the DEAE resin under the conditions used. A purity greater than 60% was achieved at this step. The results of the protein puri- ﬁcation are summarized in Table 1. The puriﬁed enzyme with a speciﬁc activity of 1.86 unitÆmg)1 protein was obtained with the recovery yield of 15.8%.

Biochemical characterization of recombinant PvSHMT

SHMT, one of the three enzymes in the deoxythy- midylate cycle [2], is a pyridoxal phosphate (PLP)-con- taining enzyme that catalyzes the conversion of serine and [6R,S]-tetrahydrofolate (THF) to glycine and 5,10- methylenetetrahydrofolate (MTHF) [5–8]. The enzyme is essential for cell replication; therefore, it is a well- known target for cancer chemotherapy [9]. The enzyme is classiﬁed as a member of the a-elimination and replacement group of PLP-dependent enzymes [10]. Besides catalyzing its main physiological reaction of SHMT activity, the enzyme is known to catalyze a broad range of reactions, including a retro-aldol cleav- age of 3-hydroxy-amino acids, transamination and decarboxylation [7,8]. Although SHMTs from many prokaryotes and eukaryotes have been studied [11–17], in-depth investigation has focused only on SHMTs from rabbit and sheep liver, as well as E. coli and Bacillus stearothermophilus [6,7,11,18,19].

The native molecular mass of PvSHMT is 72.4 kDa as determined by the gel ﬁltration method (Fig. 1). According to sequence and SDS ⁄ PAGE analyses, the subunit molecular mass is 49 kDa, suggesting that PvSHMT exists as a homodimer. This result is consis- tent with the previous prediction based on sequence analysis indicating that PvSHMT lacks the amino acids necessary for tetrameric stabilization [4]. Previous

In the present study, we describe high level expres- sion in E. coli, puriﬁcation and biochemical character- izations of PvSHMT, verifying that the gene Pv100730

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4024

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

M 1

2

3

4 A

indicates

kDa

in the dimeric form with native molecular masses of approximately 90 kDa [11,24]. This that PvSHMT is different from mammalian SHMTs with respect to oligomeric state of the enzyme.

97 66

PvSHMT 49 kDa

45

30 B

t h g i e w r a l u c e l o m g o L

6 5.8 5.6 5.4 5.2 5 4.8 4.6 4.4

9 10 11 12 13 14 15 16 Elution volume (mL)

thyroglobulin (669 kDa), (from left to right):

The puriﬁed enzyme exhibits a yellow colour and has a maximum absorption peak at 422 nm (Fig. 2), which are typical characteristics of a PLP-bound enzyme as a result of Schiff’s base formation [10]. When PLP was released from the enzyme by adding 0.4% SDS, the peak at 422 nm disappeared with a concomitant increase of absorbance around 388 and 330 nm as a result of the formation of free PLP (Fig. 2). Under the denatured condition, A280 = 1.29 (Fig. 2) but, after subtracting the absorbance contributed by PLP, A280 derived from protein absorption was 1.03. Based on the molar absorption coefﬁcient of 30 995 m)1Æcm)1 for apoenzyme of PvSHMT (apoPvSHMT) (i.e. calculated on the basis of the amino acid composition using the protparam program; http://www.expasy.ch/tools/ protparam.html) [25], the protein concentration was calculated to be 33.22 lm, which is similar to the con- centration of free PLP released. Thus, the ratio for the binding of PLP to enzyme is 1 : 1 (i.e. referring to enzyme monomer concentration) and the concentration of the holoenzyme of PvSHMT (holoPvSHMT) can be estimated to be equivalent to the concentration of the released PLP (see Experimental procedures). The molar absorption coefﬁcient (e422) of the holoenzyme was cal- culated to be 6370 m)1Æcm)1. This spectral property is similar to the reported value of SHMT from rabbit liver (e430 = 7200 m)1Æcm)1 [26]. PLP-bound SHMTs of various species have their maximum absorption peaks at around 420–430 nm [11,15,26–28].

The dissociation constant (Kd) for the binding of PLP to PvSHMT was measured using the ultraﬁltration

1.5

0.2

Fig. 1. SDS ⁄ PAGE of recombinant PvSHMT at different steps of puriﬁcation and determination of the native molecular weight by gel ﬁltration. (A) Lane M, protein standard markers (sizes are indicated); lane 1, crude extract; lane 2, after 1% poly(ethyleneimine) fraction- ation; lane 3, after DEAE-Sepharose chromatography; lane 4, after SP-Sepharose chromatography. (B) The elution volumes of the stan- dard proteins with known molecular weights are shown as ﬁlled circles ferritin (440 kDa), catalase (232 kDa), aldolase (158 kDa), BSA (66 kDa) and ovalbumin (43 kDa). The elution volume of PvSHMT (empty circle) indicates an apparent molecular weight of 72.4 kDa.

0.15

0.1

0.05

Table 1. Summary for puriﬁcation of recombinant PvSHMT. For enzyme activity, 1 unit is deﬁned as the amount of the enzyme that catalyzes the consumption of 1 lmolÆmin)1 NADH in the MTHFR coupled assay under the buffer system 50 mM Hepes (pH 7.0), 1 mM dithiothreitol and 0.5 mM EDTA at 25 (cid:2)C. % Yield is reported for 4.2 L of cell culture. The data shown are based on one repre- sentative puriﬁcation process. Generally, the values from different puriﬁcation processes can vary by approximately 10–15%.

e c n a b r o s b A

Step of puriﬁcation Total protein (mg) Total activity (unit) Speciﬁc activity (unitÆmg)1 protein) Puriﬁcation fold % Yield

0 300 350 400 450 500 550 Wavelength (nm)

0.5

e c n a b r o s b A

100 3146 1854 520 335 0.16 0.18 1 1.1 64.4

0 250 300 350 400 450 500 550 600 Wavelength (nm)

Crude extract 1% (w ⁄ v) poly (ethyleneimine) DEAE-Sepharose SP-Sepharose 490 44 242 82 0.49 1.86 2.7 3.8 46.5 15.8

studies have shown that most of mammalian enzymes including human SHMT are homotetrameric enzymes with subunit molecular masses of approximately 50 kDa [14,21–23], whereas the enzymes from prokary- otes such as E. coli and Bacillus sp. were found mainly

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4025

Fig. 2. Absorption spectra of the holoenzyme (dashed line) and the released PLP (solid line). The holoenzyme (34 lM) was denatured with 0.4% SDS in 50 mM Hepes (pH 7.0) containing 1 mM dith- iothreitol and 0.5 mM EDTA at 25 (cid:2)C. Inset: spectra of the samples in the range 300–550 nm.

Interaction of PvSHMT with various amino acids

[29],

[31]. All of

(0.23 mm)

method (see Experimental procedures). This procedure separates free PLP from the protein fractions at equilibrium. Using the known molar absorption coefﬁcient of PLP (e388 = 4900 m)1Æcm)1) the concentration of free PLP in the ﬁltrate can be deter- mined. Concentrations of (monomeric) apoPvSHMT and holoPvSHMT at equilibrium were calculated based on concentrations of free PLP and total enzyme used. The analysis (see Experimental procedures) indicates that the Kd value for the binding of PLP and apoenzyme is 0.14 ± 0.01 lm. This value indicates that PLP is tightly bound to the enzyme and that PvSHMT exists mainly as a PLP-bound form at the enzyme concentra- tions used in our experiments. The Kd value of the bind- ing of PLP and PvSHMT is less than the values reported for SHMT from humans (1 lm) [30] and E. coli the biochemical and thermodynamic properties of PvSHMT are summarized in Table 2.

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

Previous studies have shown that SHMTs from various sources can bind to d- and l-amino acids and form an external aldimine, which can be observed by spectral change of the enzyme-bound PLP species [5,32]. In the present study, a similar investigation was carried out to explore the binding of PvSHMT to l- and d-serine, l-alanine, l-threonine and glycine. Incubation of the enzyme with l- and d-serine caused an increase of the maximum absorption peak at 435 nm and, on incuba- tion with l-alanine, an increase at 446 nm (Fig. 3A– C). These results indicate the formation of external aldimine linkages between PLP and the amino acids [7,32]. A plot of the absorbance increase versus the amino acid concentration represents a binding iso- therm of the amino acid to PvSHMT (Fig. 3A–C). The Kd values were calculated from the plots to be 0.11, 17 and 2.3 mm for l-serine, d-serine and l-ala- nine, respectively (Fig. 3A–C and Table 2). It is noted that, although both l- and d-serine can bind PvSHMT, the binding afﬁnity for l-serine is 150-fold higher than that of d-serine.

Table 2. Summary for biochemical, kinetic, and thermodynamic properties of the recombinant PvSHMT.

Properties Values

Biochemical properties Molecular weight (kDa)

)1Æcm)1)

subunit molecular weight native molecular weight (dimer) The molar extinction coefﬁcient (M 49 72.4 6370 ± 48

(50 mM Hepes, pH 7.0, 1 mM dithiothreitol and 0.5 mM EDTA at 422 nm) 8.0 19 Optimum pH (50 mM Hepes) Transition temperature ((cid:2)C)

0.18 ± 0.03 0.14 ± 0.02 0.98 ± 0.06

cat)

47 ± 6.0 0.26 ± 0.01

7.0 ± 0.7 0.015 ± 0.0004

0.32 ± 0.04 0.044 ± 0.002

8.6 ± 1.2 0.12 ± 0.01

Kinetic properties (25 (cid:2)C) Km for L-serine (mM) Km for THF (mM) Turnover number (kcat) for L-serine and THF (s)1) app for D-serine (mM) Km Turnover number (kapp for D-serine (s)1) Km for L-threonine (mM) Turnover number (kcat) for L-threonine (s)1) Km for L-allo-threonine (mM) Turnover number (kcat) for L-allo-threonine (s)1) Km for D,L-b-phenylserine (mM) Turnover number (kcat) for D,L-b-phenylserine (s)1) Thermodynamic properties Kd for bindings:

0.14 ± 0.01 0.11 ± 0.001

Unlike the binding of PvSHMT to the above men- tioned amino acids, distinct absorption spectra for the binding of PvSHMT to l-threonine and glycine were observed. Binding of PvSHMT to l-threonine resulted in absorbance changes at 440 and 498 nm, whereas absorbance changes at 440 and 496 nm were noted for binding to glycine (Fig. 3D,E). An increase of absor- bance at approximately 500 nm for PLP-dependent enzymes is typically a result of the formation of an enzyme–quinonoid species (E–Q) [5,10]. Our result is in agreement with all previous studies demonstrating that an E–Q intermediate was found when SHMT bound to glycine [7]. When THF was added to the mixture of PvSHMT and glycine, the equilibrium shifted towards the greater formation of the E–Q species (Fig. 3F). Based on the absorbance increase at 496 nm and a molar absorption coefﬁcient of E–Q of 40 000 m)1Æcm)1 [27], it was calculated that THF shifted almost all of the enzyme concentration to the E–Q form (Fig. 3F). In the absence of THF (i.e. the lowest line in Fig. 3F), only 0.073 lm of E–Q was detected, whereas, in the presence of 4.8 mm THF (i.e. the highest line in Fig. 3F), the con- centration of E–Q increased to 6.53 lm. The total enzyme concentration used was 7.85 lm. The inﬂuence of THF or THF analogues in increasing the formation of E–Q in the reaction with glycine was previously observed for SHMTs from rabbit liver [33], B. stearo- thermophilus [19] and E. coli [11]. The binding of THF to the enzyme–glycine complex is considered to affect interactions of active site residues with the substrate

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4026

PLP and apoPvSHMT (lM) PvSHMT and L-serine (mM) PvSHMT and D-serine (mM) PvSHMT and L-alanine (mM) 17.0 ± 4.7 2.3 ± 0.3

0.06

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

B

A

0.04

0.03

0.05

0.03 5 3 4

5 3 4 A Δ Δ

0.02 A Δ

0.04

0.01

0.02

0.06 0.05 0.04 0.03 0.02 0.01 0

0.03

0 0 0.5 1 1.5 2 [L-serine], mM

0 10 20 30 40 [D-serine], mM

0.02

0.01

e c n a b r o s b A Δ

–0.01

–0.02

300

500

400

700

300

400

500

700

600 Wavelength (nm)

D

C

0.04

0.03

0.02 6 4 4 A Δ Δ

0.01

0.02

0

0.02

0 2 4 6 8 10 [L-alanine], mM

0.01

e c n a b r o s b A Δ

–0.01

300

400

500

700

300

400

500

700

600 Wavelength (nm)

F

E

0.05

0.3

0.2 6 9 4

0.04

0.25

A Δ Δ

0.1

0.03

0.2

0 0 1 2 3 4 5 6 [THF], mM

0.02

0.15

0.01

0.1

e c n a b r o s b A Δ

e c n a b r o s b A

0.05

–0.01

300

500

400

700

600 Wavelength (nm)

0 350 400 450 500 550 600 650 Wavelength (nm)

carboxyl group, thus lowering the pKa of the 2S proton of glycine and promoting the enzyme to adopt the E–Q form [7,34].

is used instead of

[37]

Figure 3 illustrates the similarity in catalytic proper- ties of PvSHMT and SHMTs from other organisms. PvSHMT resembles mammalian and E. coli enzymes for the ability to form external aldimine with various l- and d-amino acids [5,34]. Additionally, PvSHMT forms an E–Q intermediate with glycine, and the presence of THF promotes a greater extent of E–Q complex formation. However, the Kd for the binding of l-serine to PvSHMT is tighter (0.11 mm) compared to the value of 1 mm for SHMTs from sheep liver and E. coli [35,36].

considered as an improvement of

THF-dependent activity assay and steady-state kinetics of PvSHMT

Several methods for assessing SHMT activity have been reported [5,8]. In general, these methods mea-

sure the activity of the enzyme based on the use of an amino acid substrate alone, such as those mea- suring the formation of benzaldehyde from d,l-ery- thro-b-phenylserine and d,l-threo-b-phenylserine, or acetaldehyde from l-threonine and l-allo-threonine. In the present study, we have developed an alterna- tive assay for measuring THF-dependent SHMT activity by coupling with the reaction of 5,10-methyl- enetetrahydrofolate reductase (MTHFR) from E. coli. This method is based on a previously used assay [20], although the NADH-speciﬁc recombinant MTHFR from E. coli the NADPH- speciﬁc MTHFR from pig liver [20]. Therefore, the activity of PvSHMT can be measured continuously by monitoring the NADH consumption. Because this assay NADH is less expensive than NADPH, may be the MTHFR coupled assay and be more cost effective. In addition, all assays in the present study were carried out under anaerobic conditions using a stopped-ﬂow

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4027

Fig. 3. Spectral changes observed upon titration of PvSHMT with amino acids (A) L-serine, (B) D-serine, (C) L-alanine, (D) L-thre- onine, (E) glycine and (F) 25 mM glycine with various concentrations of THF (0.15– 4.8 mM). (A–E) PvSHMT (cid:2)35 lM in 50 mM Hepes buffer (pH 7.0), 1 mM dithiothreitol and 0.5 mM EDTA was placed in both sample and reference cuvettes and used as a baseline. Each spectrum was recorded after each addition of the amino acid into the sample cuvette and the same volume of buffer into the reference cuvette. Insets: plots of absorbance changes at the wave- length where the signal was maximum against concentrations of the amino acid: 435 nm for L- and D-serine, 446 nm for L-ala- nine. The Kd was calculated using nonlinear least square algorithms in KALEIDAGRAPH. (F) PvSHMT (7.85 lM) was added with 25 mM glycine (the lowest line) and then THF from 0.15–4.8 mM (upper lines). Using the molar absorption coefﬁcient of E–Q of 40 000 M )1Æcm)1 for calculation, at 4.8 mM THF (the highest line), the concentration of E–Q was 6.53 lM.

THF, Km

spectrophotometer to avoid interference from the oxi- dation of THF and MTHF [29] and NADH oxidase activity of MTHFR [37].

Initial rates of PvSHMT reaction at various concen- trations of THF and l-serine were analysed according to Dalziel’s equation (see Experimental procedures) [38]. The double reciprocal plot in Fig. 4 shows a pattern of converging lines, indicating that the reaction of PvSHMT uses a ternary-complex mechanism. The

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

A

5

4 e/v (× 103 s)

) s

3 t p e c r e t n

[L-serine], mM

2 3 0 1 × (

0.05

1 i y r a m

i r P

0

5 10 15 20 25

–1

0 1/[L-serine], mM

B

0.10

0.5

0.4

0.3 ) s M m

0.2 0.20 0.40 0.80 1.6

e p o l s y r a m

0.1 3 0 1 × (

i r P

(5.0 s)1),

0

5 10 15 20 25

1/[L-serine], mM

–1

–50

–5

1/[THF], mM–1

–10

l-serine and kcat were calculated from values of Km the secondary plots (Fig. 4) to be 0.14 ± 0.02 mm, 0.18 ± 0.03 mm and 0.98 ± 0.06 s)1, respectively (Table 2). The kinetic parameters of PvSHMT were compared with the data for other organisms and are summarized in Table 3. The parameters shown in Table 3 were obtained either from the assays using a radioactive method, in which the product [14C] formal- dehyde from 3-[14C] l-serine reacts with dimedon to form a 14C-formaldemethone complex that is trapped in toluene solution [39], or from the coupled enzymatic reaction in which MTHF was oxidized by NADP+ to form 5,10-methenyl-THF and NADPH by using the reaction of MTHF dehydrogenase [5]. PvSHMT has similar Km values for l-serine and THF to those of Trypanosoma cruzi, a kinetoplastid parasite. Nonethe- less, the kcat data for these parasite enzymes cannot be compared because the appropriate information for T. cruzi are unavailable. When the kcat of PvSHMT (0.98 s)1) is compared with that of other organisms, it is signiﬁcantly lower than that reported for E. coli (10.67 s)1), B. stearothermophilus human (9.58 s)1) and rabbit liver (14.17 s)1) [11,14,15,24]. The discrepancy is partly a result of the temperature used in monitoring the kinetic properties of these enzymes (25 (cid:2)C for PvSHMT and 30 or 37 (cid:2)C for other SHMTs). However, the intrinsic values of kcat also lead to such a distinction because the kcat values at 30 and 37 (cid:2)C were calculated to be 1.3 and 2.6 s)1, respec- tively (based on the results shown in Fig. 7, below). These numbers are still two- to ten-fold lower than those of the other SHMTs.

THF-independent activity of PvSHMT

SHMT has been known to catalyze a broad range of reactions and various amino acids can be used as

Fig. 4. Steady-state kinetics of the reaction of PvSHMT. The dou- ble reciprocal plot of a bisubstrate reaction catalyzed by PvSHMT reveals an intersecting-line pattern, indicating that the kinetic mech- anism is a ternary-complex type. Initial rates of PvSHMT were mea- sured by coupling with the reaction of MTHFR at 25 (cid:2)C. Concentrations of serine and THF used were 0.05–1.6 mM for ser- ine and 0.025–0.4 mM for THF. The reactions were carried out in 50 mM Hepes (pH 7.0). Insets: secondary plots of the ordinate intercepts (A) and slopes (B) of the primary plot as a function of 1 ⁄ [serine].

Table 3. Summary for kinetic parameters of the THF-dependent SHMT reactions of PvSHMT and SHMTs from various organisms. ), value THF not reported. Kinetic parameters of T. cruzi, B. stearothermophilus and sheep liver cytosolic SHMTs were determined by for kcat and Km the L-[3-14C-serine] method [16,19,20]. Kinetic parameters of E. coli, human and rabbit cytosolic SHMTs were determined by coupling with MTHF dehydrogenase using NADPH as reducing equivalent [14,15,50].

L-serine (mM)

THF(mM)

Kinetic parameters

Organism References Temperature ((cid:2)C) (assay) Km Km kcat (s)1)

0.14 ± 0.02 Present study 25 0.98 ± 0.06 (0.26 ± 0. 01)

– 10.67

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4028

Plasmodium vivax (D-serine) Trypanosoma cruzi Escherichia coli Bacillus stearothermophilus Sheep (liver cytosolic) Rabbit (liver cytosolic) Human (cytosolic) 0.18 ± 0.03 (47 ± 6) 0.15 0.30 0.90 ± 0.03 1.0 0.30 0.30 0.11 0.025 – 0.82 0.02 0.04 5.0 ± 0.6 – 14.17 9.58 37 37 37 37 30 30 [16] [11] [24] [18] [14] [15]

indicate

results

Nevertheless, the enzyme cannot catalyze the conver- sion of d-serine and THF to yield glycine and MTHF [5]. This characteristic feature is contradictory to PvSHMT. Figure 5 shows a plot of initial rates of the PvSHMT reaction versus concentrations of d-serine at a saturating concentration of THF. Because the assays measured the formation of MTHF using the coupled reaction of MTHFR, that the PvSHMT can catalyze the transfer of the hydroxym- ethyl moiety from d-serine to THF. The Km app and app values for the d-serine reaction were calculated kcat to be 47 ± 6 mm and 0.26 ± 0.01 s)1, respectively (Fig. 5). These values indicate that l-serine is a better substrate than d-serine because its Km is lower and its kcat is higher (Table 2).

substrates [5,6,8]. Reactions of PvSHMT with d,l-threo- b-phenylserine, l-allo-threonine and l-threonine were explored to determine whether PvSHMT shares any common catalytic properties with other SHMTs. When d,l-threo-b-phenylserine was incubated with PvSHMT, the reaction formed glycine and benzaldehyde, as noted by a large increase in absorbance at 248 and 279 nm as a result of the absorption of benzaldehyde (e248 = 12 000 and e279 = 1400 m)1Æcm)1) [40]. This indicates that PvSHMT can catalyze a THF-indepen- dent retro-aldol cleavage of d,l-threo-b-phenylserine, as is typically observed for the reaction of SHMTs from other species. The assays of l-threonine and l-allo-thre- onine were monitored by coupling with the reaction of yeast alcohol dehydrogenase [5]. Steady-state kinetic parameters of these reactions are summarized in Table 2, and also compared with those reported for other SHMTs (Table 4). Based on the Km and kcat values, l-allo-threonine serves as a better substrate than l-threonine for PvSHMT. In general, PvSHMT cata- lyzes these substrates less efﬁciently than SHMTs from other species because the kcat value of the reaction of PvSHMT and l-allo-threonine is noticeably lower than those of other SHMTs (Table 4).

Reaction of PvSHMT with D-serine

It is widely known that SHMT has a broad range of reactions catalyzed as well as substrate speciﬁcity.

PvSHMT also differs from mammalian SHMT in that E–Q formation was not observed upon incubation the enzyme with l-serine at pH 8 or above of (Fig. 6A). This indicates that PvSHMT strictly requires THF for the removal of the a-H from the external aldimine compound of l-serine. This property is unlike the rabbit SHMT, where E–Q was detected when l- serine alone was incubated with the enzyme [41]. By contrast, incubation of PvSHMT with d-serine at pH 8 resulted in the formation of E–Q, as noted by an in absorbance at approximately 500 nm increase rabbit and E. coli SHMTs were (Fig. 6B). The reported to bind to glycine and many l- and d-amino acids (e.g. l-serine, d-alanine and l-alanine), which

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

Table 4. Summary for kinetic parameters of the THF-independent reactions (retrol-aldol cleavage of 3-hydroxy amino acid) of SHMTs from various organisms.

Kinetic parameters

Organisms References Km (mM) kcat (s)1)

Plasmodium vivax

D,L-threo-b-phenylserine L-allo-threonine L-threonine Sheep (liver)

Present study 8.6 ± 1.2 0.32 ± 0.04 7.0 ± 0.7 0.12 ± 0.01 0.044 ± 0.002 0.015 ± 0.0004

D,L-threo-b-phenylserine D,L-erythro-b-phenylserine L-threonine

L-allo-threonine

D,L-allo-threonine Bacillus stearothermophilus

L-allo-threonine

D,L-allo-threonine

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4029

Native cytosolic enzyme [27] 84 9.5 7 21 32 0.09 [24] Recombinant enzyme 0.7 ± 0.01 3.7 ± 0.4 Rabbit (liver) Recombinant form of cytosolic enzyme [14] 1.5 1.34 [24] 0.4 ± 0.009 0.9 ± 0.002 Escherichia coli [11] 1.5 0.5

0.15

0.1

0.05

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

) 1 – s M µ ( e t a r l a i t i n

I

100

150

200 250

[D-serine], mM

resulted in the formation of E–Q [5,35]. However, incubation of d-serine with these enzymes did not result in the formation of E–Q [5,35]. Therefore, it appears that, with the d-serine isomer, only PvSHMT can catalyze the breakage of the a-H and stabilize the E–Q form. This difference in substrate speciﬁcity between PvSHMT and other SHMTs implies that the structural details of PvSHMT active site differ from those of mammalian enzymes, which may allow the development of speciﬁc inhibitors against PvSHMT for anti-malarial purpose.

Additional experiments to observe the formation of E–Q that may originate from a retro-aldol cleavage of the PvSHMT-d-serine complex, with the consequence of free formaldehyde release, were carried out similar to those described previously [42]. Formaldehyde is known to spontaneously react with THF in solution to

1.4

Fig. 5. Initial rates of the reaction of PvSHMT as a function of D- serine concentration. The reactions contained 6S-THF at a ﬁxed concentration of 400 lM and D-serine concentrations in the range 5–200 mM. The initial rates were plotted against various D-serine concentrations and the kinetic parameters were calculated using app and nonlinear least square algorithms in KALEIDAGRAPH. The Km app were determined to be 47 ± 6 mM and 0.26 ± 0.01 s)1, kcat respectively.

A

1.2

1

0.3

A

0 t a c

–1

'

0.25

–2

k n

l

0.8

–3

) 1 – s (

0.2

t a c

–4 0.0032

0.0036

'

0.6

k

0.0034 1/T(K–1)

0.15

0.4

0.1

e c n a b r o s b A

0.2

0.05

0 300

350

400

500

550

600

25 10 Temperature (ºC)

450 Wavelength (nm)

B

0.3

B

0.8

0.05

0.04

0.25

0.03

0.02

0.6

e c n a b r o s b A

0.01

0.2

0 300 350 400

500 550 600

450 Wavelength (nm)

0.15

) 1 – s M µ (

0.4

0.1

e c n a b r o s b A

'

x a m V

0.05

0.2

0 300

350

400

500

550

600

450 Wavelength (nm)

0 5

8 pH

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4030

Fig. 7. Effects of temperature and pH on PvSHMT activity. (A) Turnover numbers of PvSHMT (k¢cat) at 5–30 (cid:2)C were measured and plotted as a function of the temperature. Inset: Arrhenius plot based on the equation ln k¢cat = ln (A))Ea ⁄ RT. The arrow indicates the transition temperature at (cid:2)19 (cid:2)C. For the pH dependence study, (B) initial velocities of the reaction were measured at various pH (6.5–10.0) at 25 (cid:2)C. The curve indicates ﬁtting to the plot according to Eqn (4). Both studies were performed at the saturating concentrations of THF (0.8 mM) and serine (2 mM). Fig. 6. Spectral change of PvSHMT upon titration with L- and D-ser- ine. (A) PvSHMT spectrum ((cid:2)35 lM) in 50 mM Hepes buffer (pH 8), 1 mM dithiothreitol and 0.5 mM EDTA is shown as the solid line, and those with 2.8 mM L-serine at pH 8, 9 and 10 are shown as the dashed line, the line with rectangles, and the line with circles, respectively. (B) PvSHMT ((cid:2)35 lM) in 50 mM Hepes buffer (pH 8), 1 mM dithiothreitol and 0.5 mM EDTA is shown as the solid line. After adding 355 mM of D-serine, the dashed line spectrum was obtained, Inset: titration of indicating the formation of the E–Q. PvSHMT with D-serine (0.04–355 mM) at pH 8.0.

failed to detect

yield MTHF, a substrate for MTHF dehydrogenase [42] or MTHFR. We employed the reaction of MTHFR that can measure the amount of free formal- dehyde (down to the level of 0.2 lm) by monitoring the oxidation of NADH by MTHF (see Experimental procedures). free formaldehyde under the conditions studied: A422 = 0.4 or equivalent to 63 lm of PvSHMT and 500 mm of d- or l-serine in 50 mm Hepes (pH 8.0). However, this cannot rule out the possibility of PvSHMT undergoing a retro-aldol cleavage and forming formaldehyde entrapped inside the enzyme active site.

Effects of temperature and pH on the catalysis of PvSHMT

in the

range 5–30 (cid:2)C. The

(k¢cat)

that will be used for performing PvSHMT assays in the future. The ﬁtting curve yielded pK1 and pK2 of 6.5 ± 0.2 and 8.9 ± 0.2, respectively. These pKa val- ues indicate that, in the enzyme–substrate complex, a functional group with a pKa of (cid:2)6.5 needs to be in the deprotonated form, whereas the other, with a pKa of (cid:2)8.9, has to be in the protonated form for the reaction to proceed with a maximum rate [43]. However, these data have not been used for insightful interpretation of the reaction mechanism of PvSHMT because further such as measuring kcat and studies are required, kcat ⁄ Km at various pHs. The results obtained in the present study cannot be compared directly with those of other SHMTs because previous investigations of the effect of pH on SHMT activity mostly measured the retro-aldol cleavage activity using l-threonine [44] or b-phenylserine as substrates [45]. The optimum pH for the reaction of SHMT from Hyphomicrobium methy- lovorum with b-phenylserine is in the range 7.5–8.0 [45], whereas the optimum pH for SHMT from Strep- tococcus thermophilus is 6.5 when using l-threonine as a substrate [44].

The effect of temperature on the THF-dependent activ- ity of PvSHMT was examined by measuring turnover number at the saturating concentrations of both sub- results strates obtained indicated that the catalytic activity increased upon temperature increment (Fig. 7A). A plot of ln (k¢cat) versus reciprocal values of absolute temperature (Arrhenius plot) reveals a pattern consisting of two linear relationships with a transition temperature at 19 (cid:2)C (Fig. 7A). When each line was ﬁtted to the Arrhenius equation, the activation energies (Ea) below and above the transition temperature were calculated to be 132.24 and 79.05 kJÆmol)1, respectively. This demonstrates that, at temperatures below and above 19 (cid:2)C, the rate determining steps of the reaction are different. A previous study investigating a temperature effect on the reaction of Methanococcus jannaschii SHMT using l-allo-threonine as a substrate revealed a linear relationship of the Arrhenius plot in the temper- ature range 25–65 (cid:2)C [12]. The Arrhenius plot shown in Fig. 7 is also useful for the approximation of k¢cat the reactions catalyzed by PvSHMT at various of temperatures to allow comparison with the values of other enzymes summarized in Table 3. Under the con- ditions used, the rate was 0.8 s)1 at 25 (cid:2)C and 1.3 s)1 at 30 (cid:2)C. Using the Arrhenius plot shown in Fig. 7 (Ea = 79.05 kJÆmol)1, R = 8.314 JÆmol)1ÆK)1, T = 310 K, ln A = 31.625 s)1), as well as the assumption that Ea and A are constant in this temperature range, the rate of the PvSHMT reaction at 37 (cid:2)C was calcu- lated to be 2.6 s)1, which is three times the value at 25 (cid:2)C.

The effect of pH on the THF-dependent SHMT activity was investigated for pH 6.5–10. The plot of apparent maximum velocities for the THF-dependent reaction catalyzed by PvSHMT as a function of pH revealed an optimum pH at 8.0, which is a condition

In conclusion, the recombinant PvSHMT can be overexpressed and puriﬁed. In the present study, an alternative assay for THF-dependent SHMT activity was developed and used to explore the kinetic proper- ties of the enzyme. PvSHMT is largely similar to SHMTs from other species with respect to its ability to form external aldimines with various amino acids and catalyze the cleavage of 3-hydroxy-amino acids. How- ever, the present study shows that features unique to PvSHMT can be drawn. First, PvSHMT reacts with d-serine, but not with l-serine, to form E–Q complex independent of THF under high pH condition. This property is the reverse of that for the mammalian l-serine. Second, enzyme, which can only use PvSHMT is capable of utilizing d-serine, in addition to l-serine, as a substrate to generate glycine in the THF-dependent reaction. These results demonstrate that the environment of the active sites between the parasite and mammalian enzymes are not identical. The stereoisomer for serine is restricted to the l-isomer for mammalian enzymes, whereas the parasite enzyme can accommodate both stereoisomers. This informa- tion is important because the difference in substrate speciﬁcity between the enzymes from the parasite and mammalian host can be useful for designing selec- tive inhibitors against the malarial enzyme. Future studies of PvSHMT X-ray structures will be help- ful for elucidating the difference in the microenviron- ment of the active sites of SHMTs at the atomic level, which will facilitate the structure-based design of inhibitors.

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4031

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

Experimental procedures

Chemicals and reagents

proofreading polymerase (Pfu). NdeI and SpeI restriction sites (underlined) were incorporated into the 5¢ and 3¢ ends of the resulting PCR product, which covers a complete sequence of pvshmt. Cycling parameters used for the PCR reaction were: a ﬁrst cycle at 95 (cid:2)C for 5 min; 30 cycles of 95 (cid:2)C for 30 s, 52 (cid:2)C for 30 s, 68 (cid:2)C for 150 s, and a ﬁnal cycle at 95 (cid:2)C for 30 s, 52 (cid:2)C for 30 s, 68 (cid:2)C for 7 min. The 1.3 kb PCR product obtained was ligated into an expression vector pET17b at the NdeI and SpeI restriction sites. The resulting plasmid was named pET17b-pvshmt. The sequence was veriﬁed by DNA sequencing at the Bioservice Unit, National Science and Technology Develop- ment Agency (Bangkok, Thailand).

Expression and puriﬁcation of recombinant PvSHMT

The expression plasmid pET17b-pvshmt was transformed into E. coli BL21 (DE3) and expression of PvSHMT was induced by addition of 1 mm IPTG into the culture when D600 (cid:2)1.0 was reached at 16 (cid:2)C. Cells were allowed to grow at this temperature overnight (until D600 (cid:2)4.0 was reached), which typically yielded 3.6 gÆL)1 cell culture. The cell pellet was stored at )80 (cid:2)C until use.

Restriction endonucleases, DNA ligase and other modifying enzymes were obtained from New England Biolabs (Ipswich, MA, USA) and Promega (Madison, WI, USA). l-serine, NADH, PLP, d,l-threo-b-phenylserine, poly(ethyleneimine) solution (stock 50% w ⁄ v) and 37% formaldehyde solution were purchased from Sigma-Aldrich (St Louis, MO, USA). THF was purchased from Merck Eprova AG (Schaffhausen Switzerland). Phenylmethanesulfonyl ﬂuoride and dithiothre- itol were from Bio-Science Inc. (Allentown, PA, USA). IPTG was purchased from Fermentas Life Sciences (Glen Burnie, MD, USA). EDTA and ammonium sulfate were purchased from Merck (Gibbstown, NJ, USA). Chromatographic media were purchased from GE Healthcare Biosciences (Uppsala, Sweden). Hepes was purchased from Research Organics (Cleveland, OH, USA). All other chemicals used were of analytical grade. pET23b-metF [46], a plasmid for expression of MTHFR, was a generous gift from E. Trimmer (Department of Chemistry, Grinnell College, Grinnell, IA, USA). THF stock solution was prepared under anaerobic condition in an anaerobic glove box (Belle Technology, Portesham, UK) by dissolving solid THF in an anaerobic buffer of 50 mm Hepes buffer (pH 7.0) containing 1 mm dith- iothreitol and 0.5 mm EDTA. The concentration of the stock solution of THF was determined as described previously [47] with some slight modiﬁcations. A solution of THF prepared was mixed with 99% formic acid containing 1 mm dithiothre- itol and 0.5 mm EDTA, and was incubated in a boiling water bath for 5 min. Under these conditions, THF is converted to 5,10-methenyl-THF, which has an absorption peak at 350 nm. The amount of 5,10-methenyl-THF formed was calculated using the known molar extinction coefﬁcient; e350, pH 3 = 26 000 m)1Æcm)1 [29].

Bacterial strains and plasmids

Escherichia coli DH5a (Stratagene, La Jolla, CA, USA) was used as a common host strain for plasmid-mediated transfor- mation or manipulation of recombinant plasmids. E. coli BL21 (DE3) (Novagen, Madison, WI, USA) was served as a host strain for expression of the PvSHMT. The expression vector pET17b (3306 bp; Novagen) was used for construct- ing an expression vector for expressing the pvshmt gene.

Plasmid construction

Frozen cell paste (30.4 g) was thawed, resuspended in a lysis buffer (50 mm Hepes buffer, pH 8.3, containing 100 lm phenylmethanesulfonyl ﬂuoride, 1 mm dithiothreitol, 1 mm EDTA and 10 lm PLP) and disrupted by sonication. The suspension was centrifuged (Avanti J-E centrifuge; Beckman Coulter, Fullerton, CA, USA) at 39 200 g for 1 h at 4 (cid:2)C. Polyethyleneimine was added to the supernatant obtained to make up a ﬁnal concentration of 1% w ⁄ v and this suspension was centrifuged at 39 200 g for 30 min at 4 (cid:2)C. The resulting supernatant was collected and applied onto a DEAE-Sepha- rose column (2.5 · 25 cm) pre-equilibrated with Buffer 1 (50 mm Hepes, pH 8.3, containing 1 mm dithiothreitol and 0.5 mm EDTA). The column was then washed with the same buffer. PvSHMT was eluted as a ﬂow through fraction that exhibited yellow colour (absorption peak at 422 nm). PLP was added to the enzyme solution to make up a ﬁnal concen- tration of 10 lm. The enzyme solution was then exchanged into Buffer 2 (50 mm Hepes, pH 7.0, containing 1 mm dith- iothreitol and 0.5 mm EDTA) by passing through a Sepha- dex G-25 gel ﬁltration column. The PvSHMT solution was loaded onto an SP-Sepharose column (2.5 · 17 cm) pre- equilibrated with Buffer 2. The column was eluted with 600 mL of 0–300 mm NaCl gradient in Buffer 2. PvSHMT was eluted around the mid of the gradient. Fractions exhibit- ing the 280 ⁄ 422 ratio at (cid:2)8.2 were pooled and concentrated by a Centriprep YM-30 unit or a stirred-cell with 30 kDa molecular weight cut-off membrane (Millipore Corporation, Billerica, MA, USA). Puriﬁed PvSHMT was stored at )80 (cid:2)C after exchanging into 50 mm Hepes buffer (pH 7.0) containing 1 mm dithiothreitol, 0.5 mm EDTA and 10 lm

The full-length pvshmt gene was ampliﬁed from pGemT- 4RT(ms2002) [4], a plasmid containing pvshmt, using pri- mers 5¢-CTAATCATATGTTTAACAACGAGCCGCTGGA (sense strand) and 5¢-ACTGTACTAGTCACAC AC-3¢ (antisense strand) with a CGTTGGATGCCCTCAGA-3¢

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4032

PLP. PLP (10 lm) was added to the enzyme solution before exchanging the buffer to ensure that PLP was fully bound. The protein content was determined by Bradford method [48] using BSA as a protein standard.

was prepared in an anaerobic glove box. One unit of PvSHMT was deﬁned as the amount of enzyme that cata- lyzes the consumption of 1 lmolÆmin)1 NADH in the MTHFR coupled assay at 25 (cid:2)C, and the speciﬁc activity was expressed as a ratio of the enzyme unit per mg protein.

Determination of native and subunit molecular masses

Steady-state kinetic parameters (Km of l-serine and THF, kcat) were analysed according to Dalziel’s equation for a bisubstrate reaction as shown in Eqn (1) [38]. Initial rate data were from the reactions varying concentrations of l-serine in the range 0.05–1.6 mm and THF in the range 0.025–0.4 mm.

(cid:3)

ð1Þ

(cid:3) þ /B= B½

(cid:3) þ /AB= AB½

e=v ¼ /0 þ /A= A½

serine ¼ /B=/o:

where: kcat ¼ 1=/0; Km Kinetic parameters

THF ¼ /A=/0; Km for

A molecular mass of native PvSHMT was determined using Superdex 200 HR 10 ⁄ 30 gel ﬁltration chromatography operated by A¨ KTA puriﬁer (GE Healthcare Bioscienc- es). Protein elution was carried out at a ﬂow rate of 0.5 mLÆmin)1 with 50 mm sodium phosphate buffer (pH 7). Protein standards used were ovalbumin (43 kDa), BSA (66 kDa), aldolase (158 kDa), catalase (232 kDa), ferritin (440 kDa) and thyroglobulin (669 kDa).

app, d-serine and kcat

reaction using d-serine as a substrate were determined using a ﬁxed concentration of 6S-THF at 0.4 mm and varying concentrations of d-serine in the range 5–200 mm. The Km app values were calculated using nonlinear least square algorithms in kaleidagraph (Synergy Software, Reading, PA, USA).

A subunit molecular mass was determined by 12% SDS ⁄ PAGE with the protein markers: phosphorylase b (97 kDa), BSA (66 kDa), ovalbumin (45 kDa), carbonic anhydrase (30 kDa), soybean trypsin inhibitor (20 kDa) and lysozyme (14.4 kDa). Protein bands were visualized by staining with Coomassie brilliant blue dye.

Absorption measurements

PvSHMT activity assay and kinetic analysis

UV-visible absorption spectra were recorded with double- beam spectrophotometers (Cary300, Varian Inc., Palo Alto, CA, USA; Shimazu 2501PC, Shimazu Corp., Kyoto, Japan) or an Agilent 8453 diode-array spectrophotometer (Hewlett-Packard, Palo Alto, CA, USA). All spectropho- tometers were equipped with thermostat cell compartments. All spectra were recorded at 25 (cid:2)C using buffer as blank.

absorbance

Determination of the molar absorption coefﬁcient of PvSHMT

An absorption spectrum of PvSHMT (A422 (cid:2)0.21 AU) in 50 mm Hepes buffer (pH 7.0), 1 mm dithiothreitol and 0.5 mm EDTA was recorded. Then, SDS was added to a ﬁnal concentration of 0.4% to denature the protein. The concentration of the released PLP, equivalent to the con- centration of holoPvSHMT, was calculated based on the extinction coefﬁcient of free-PLP (e388 = 4970 m)1Æcm)1) in the same buffer condition. The molar extinction coefﬁcient of PvSHMT at the wavelength 422 nm was then calculated using Beer–Lambert equation. The stoichiometry between apoenzyme and free-PLP was 1 : 1 because both of their concentrations were comparable (see Results). All enzyme concentrations reported in the present study are in accor- dance with subunit concentrations because they indicate active site concentrations.

The dissociation constant for the binding of apoenzyme and PLP

The dissociation constant for the binding of apoenzyme and PLP was calculated based on concentrations of the free and

PvSHMT activity was measured by coupling its reaction with the reaction of MTHFR [37]. SHMT catalyzes the conver- sion of serine and THF to glycine and MTHF, for which the latter and NADH are substrates of MTHFR. The reaction of MTHFR yields 5-methyltetrahydrofolate and NAD+ as products and can be measured by monitoring the decrease of NADH absorbance. We measured the consumption of NADH by monitoring 375 nm the (e375 = 1.92 mm)1Æcm)1) instead of at 340 nm in order to avoid interference from the THF absorbance. The assays were performed using a Hi-Tech Scientiﬁc Model SF-61SX stopped-ﬂow spectrophotometer (TgK Scientiﬁc Limited, Bradford-on-Avon, UK) in single mixing mode under anaer- obic conditions. The stopped-ﬂow apparatus was made anaerobic by protocatechuate acid–protocatechuate 3,4-di- oxygenase oxygen-scrubbing solution as described previously [49]. The oxygen-scrubbing solution contained 50 mm sodium phosphate buffer (pH 7.0), 400 lm protocatechuate acid or 3,4-dihydroxybenzoate and (cid:2)0.1 unitÆmL)1 protoca- techuate 3,4-dioxygenase. The solution was allowed to stand in the ﬂow system overnight and was then thoroughly rinsed with an anaerobic buffer [de-aerated buffer of 50 mm Hepes buffer (pH 7.0) containing 1 mm dithiothreitol and 0.5 mm EDTA] before starting the experiments. The assay measuring SHMT speciﬁc activity typically contains 150 lm NADH, 2 mm l-serine, 400 lm THF, 3 lm MTHFR, 1 lm PvSHMT, 1 mm dithiothreitol and 0.5 mm EDTA in 50 mm Hepes buffer (pH 7.0). All concentrations mentioned in these exper- iments refer to those after mixing. The reaction assays were performed at 25 (cid:2)C in three replicates. A solution of THF

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4033

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

curve, syringes containing formaldehyde solutions (0.2, 0.4, 0.8, 1.6 and 3.2 lm) were bubbled with N2 gas and added to an anaerobic solution of 150 lm NADH and 400 lm THF to preform MTHF, and then loaded onto the stopped-ﬂow machine. The solutions were mixed with an anaerobic solu- tion of 3 lm MTHFR and the reactions were monitored for decrease of A340. All concentrations indicated are after the ﬁnal stopped-ﬂow mixing. This method can detect formalde- hyde concentration down to the level of 0.2 lm.

bound species measured at equilibrium (Eqn 2) by ultraﬁltra- tion method using a Centriprep unit with 30 kDa molecular weight cut-off (YM-30). A protein solution (10 mL) contain- ing 24 lm holoenzyme (A422 (cid:2)0.15 AU) in 50 mm Hepes buffer (pH 7.0), 1 mm dithiothreitol and 0.5 mm EDTA was placed in a Centriprep unit. Then, the unit was centrifuged at 780 g for 3 min at 25 (cid:2)C (Avanti J-E centrifuge). Approxi- mately 1 mL of the ﬁltrate was obtained and A388 to was measured determine the concentration of the released free PLP. The Kd value under the equilibrium condition was cal- culated according to Eqn (3) where [Enz], [PLP] and [Enz- PLP] are the concentrations of free enzyme, free PLP and enzyme-bound PLP complex, respectively.

A solution of PvSHMT (63 lm) with d- or l-serine (500 mm) in Hepes buffer (pH 8.0) was ultraﬁltrated using a microcon concentrator. The ﬁltrate was ﬁlled into a syr- inge, bubbled with N2 gas, mixed with NADH and THF, and the reaction was monitored as described above.

Enz (cid:4) PLP Ð Enz þ PLP

ð2Þ

Effect of temperature on PvSHMT activity

ð3Þ

Kd ¼

½Enz(cid:3)½PLP(cid:3) ½Enz (cid:4) PLP(cid:3)

temperatures

the reaction at various

Equilibrium binding of L-serine and other amino acids to PvSHMT

using

spectroscopically

The effect of temperature on the reaction catalyzed by PvSHMT was investigated by measuring the initial velocity of in the range 5–30 (cid:2)C. The catalytic rate constant (k¢cat) measured at the saturating concentrations of both substrates was plotted as a function of absolute temperature (K) using the Arrhenius equation, ln (k¢cat) = ln (A))Ea ⁄ RT, where A, Ea, R and T represent the Arrhenius constant, activation energy, gas constant (8.314 JÆmol)1ÆK)1) and absolute temperature (K), respectively. The activation energy of the reaction catalyzed by PvSHMT was calculated from the slope ()Ea ⁄ R) of the Arrhenius plot [ln (k¢cat) versus 1 ⁄ T].

Optimum pH for the reaction of PvSHMT

the

the optimum pH for

reaction of To determine PvSHMT, initial rates of the reaction were measured at various pH conditions (6.5–10). The assays were carried out in 100 mm buffer containing 0.5 mm EDTA and 1 mm dith- iothreitol. The buffers used in this study were Hepes for pH 6.5–8.5 and carbonate for pH 9–10. A plot of apparent V¢max as a function of pH was analysed according to Eqn (4), where Y is apparent maximum velocity (V¢max), C is the pH-independent value of V¢max, and K1 and K2 are the dissociation constants for the ionizable groups.

Y ¼

ð4Þ

1 þ

C 10(cid:4)pH 10(cid:4)pK1

10(cid:4)pK2 10(cid:4)pH

The dissociation constant for the binding of l-serine to PvSHMT was determined a Cary300 double-beam spectrophotometer. The difference in spectral property as a result of binding was recorded directly after adding a series of small volumes of a stock serine solution into a 1 mL solution of PvSHMT [35 lm enzyme with A422 (cid:2)0.23 AU in 50 mm Hepes buffer (pH 7.0), 1 mm dithiothreitol and 0.5 mm EDTA] placed in the sample cuvette, whereas an equal volume of the buffer was added into the enzyme solution placed in the reference cuv- ette. The ﬁnal concentrations of l-serine used were in the range 0.05–1.8 mm. The change in absorbance at the wave- length at which the signal was most pronounced (y-axis) was plotted against serine concentration (x-axis) after cor- rection for dilution effect. The Kd value for the PvSHMT– serine least complex was determined using nonlinear squares ﬁtting in kaleidagraph. Similar experiments were performed for glycine (0.2–20 mm), d-serine (1–20 mm), l-alanine (0.1–8 mm) and l-threonine (0.1–8 mm). Only the data for l-serine, d-serine and l-alanine were to calculate Kd values for the formation of the external aldimine com- plex. Because the binding of glycine and l-threonine resulted in the formation of E–Q species, their Kd values cannot simply be calculated from the absorbance change.

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

Acknowledgements

Determination of formaldehyde formation

This work was supported by a grant from National Center for Genetic Engineering and Biotechnology (BT-B-02-MG-BC-5006) to U.L. and P.C., and par- tially supported by grants from The Thailand Research Fund (BRG5180002) and the Faculty of Science, Mahidol University (to P.C). K.S. is a recipient of a

Experiments similar to those described previously [42] were carried out to determine whether free formaldehyde forma- tion occurred during the incubation of PvSHMT with d- or l-serine. MTHFR was used instead of MTHF dehydro- genase because of its availability. To establish a standard

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4034

12 Angelaccio S, Chiaraluce R, Consalvi V, Buchenau B,

Giangiacomo L, Bossa F & Contestabile R (2003) Cata- lytic and thermodynamic properties of tetrahydrometha- nopterin-dependent serine hydroxymethyltransferase from Methanococcus jannaschii. J Biol Chem 278, 41789–41797.

scholarship from Thailand Graduate Instituted of Sci- ence and Technology (TGIST). S.M. received a fellow- ship from the Commission on Higher Education Staff Development Project, Thailand. We thank Dr Eliza- for providing the plasmid for beth E. Trimmer MTHFR expression, and Merck Epova Eprova AG (Schaffhausen Switzerland) for providing THF and THF analogues. We thank Drs Jeerus Sucharitakul and Jack F. Kirsch for their valuable suggestions.

13 Delle Fratte S, White RH, Maras B, Bossa F & Schirch V (1997) Puriﬁcation and properties of serine hydrox- ymethyltransferase from Sulfolobus solfataricus. J Bacte- riol 179, 7456–7461.

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax

References

14 Di Salvo ML, Delle Fratte S, De Biase D, Bossa F & Schirch V (1998) Puriﬁcation and characterization of recombinant rabbit cytosolic serine hydroxymethyltrans- ferase. Protein Expres Purif 13, 177–183.

1 Snow RW, Guerra CA, Noor AM, Myint HY & Ha SI (2005) The global distribution of clinical episodes of Plasmodium falciparum malaria. Nature 434, 214–217. 2 Yuthavong Y, Kamchonwongpaisan S, Leartsakulpa-

15 Kruschwitz H, Ren S, Di Salvo ML & Schirch V (1995) Expression, puriﬁcation, and characterization of human cytosolic serine hydroxymethyltransferase. Protein Expres Purif 6, 411–416.

nich U & Chitnumsub P (2006) Folate metabolism as a source of molecular targets for antimalarials. Future Microbiol 1, 113–125.

16 Capelluto DG, Hellman U, Cazzulo JJ & Cannata JJ (2000) Puriﬁcation and some properties of serine hy- droxymethyltransferase from Trypanosoma cruzi. Eur J Biochem 267, 712–719.

3 Alfadhli S & Rathod PK (2000) Gene organization of a Plasmodium falciparum serine hydroxymethyltransferase and its functional expression in Escherichia coli. Mol Biochem Parasitol 110, 283–291.

17 Vatsyayan R & Roy U (2007) Molecular cloning and biochemical characterization of Leishmania donovani serine hydroxymethyltransferase. Protein Expres Purif 52, 433–440.

4 Leartsakulpanich U, Kongkasuriyachai D, Imwong M, Chotivanich K & Yuthavong Y (2008) Cloning and characterization of Plasmodium vivax serine hydroxym- ethyltransferase. Parasitol Int 57, 223–238.

5 Schirch V (1982) Serine hydroxymethyltransferase. In

Adv Enzymol Relat Areas Mol Biol 53 (Meister A, ed.), pp 83–112. John Wiley & Sons, Inc., New York, NY.

18 Jagath-Reddy J, Ganesan K, Savithri HS, Datta A & Rao NA (1995) cDNA cloning, overexpression in Esc- herichia coli, puriﬁcation and characterization of sheep liver cytosolic serine hydroxymethyltransferase. Eur J Biochem 230, 533–537.

6 Schirch V (1998) Mechanism of folate requiring

enzymes in one carbon metabolism. In Comprehensive Biological Catalysis: A mechanistic Reference (Sinnott M, ed.), pp 211–252. Academic press limited, Academic press, San Diego, CA.

19 Bhavani S, Trivedi V, Jala VR, Subramanya HS, Kaul P, Prakash V, Rao NA & Savithri HS (2005) Role of Lys- 226 in the catalytic mechanism of Bacillus stearothermo- philus serine hydroxymethyltransferase-crystal structure and kinetic studies. Biochemistry 44, 6929–6937.

7 Schirch V & Szebenyi DME (2005) Serine hydroxym-

ethyltransferase revisited. Curr Opin Chem Biol 9, 482– 487.

20 Matthews RG, Ross J, Baugh CM, Cook JD & Davis L (1982) Interaction of pig liver serine hydroxymethyl- transferase with methylenetetrahydropteroylpolygluta- mate inhibitors and with tetrahydropteroylpolyglutamate substrates. Biochemistry 21, 1230–1238.

8 Matthews RG & Drummond JT (1990) Providing one carbon units for biological methylations: mechanistic studies on serine hydroxymethyltransferase, methylene- tetrahydrofolate reductase, and methyltetrahydrofolate- homocysteine methyltransferase. Chem Rev 90, 1275–1290.

21 Schirch L & Peterson D (1980) Puriﬁcation and proper- ties of mitochondrial serine hydroxymethyltransferase. J Biol Chem 25, 7801–7806.

9 Agrawal S, Kumar A, Srivastava V & Mishra BN

22 Manohar R, Ramesh KS & Rao NA (1982) Puriﬁca-

(2003) Cloning, expression, activity and folding studies of serine hydroxymethyltransferase: a target enzyme for cancer chemotherapy. J Microbiol Biotechnol 6, 67–75.

tion, physiochemical and regulatory properties of serine hydroxymethyltransferase from sheep liver. J Biosci 4, 31–50.

10 Eliot AC & Kirsch JF (2004) Pyridoxal phosphate

23 Garrow TA, Brenner AA, Whitehead VM, Chen XN,

enzymes: mechanistic, structural and evolutionary con- sideration. Annu Rev Biochem 73, 383–415.

Duncan RG, Korenberg JR & Shane B (1993) Cloning of human cDNAs encoding mitochondrial and cytosolic serine hydroxymethyltransferases and chromosomal localization. J Biol Chem 268, 11910–11916.

11 Schirch V, Hopkins S, Villar E & Angelaccio S (1985) Serine hydroxymethyltransferase from Escherichia coli: puriﬁcation and properties. J Bacteriol 163, 1–7.

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4035

24 Jala VR, Prakash V, Rao NA & Savithri HS (2002)

37 Trimmer EE, Ballou DP & Matthews RG (2001) Meth- ylenetetrahydrofolate reductase from Escherichia coli: elucidation of the kinetic mechanism by steady-state and rapid-reaction studies. Biochemistry 40, 6205–6215.

Overexpression and characterization of dimeric and tet- rameric forms of recombinant serine hydroxymethyl- transferase from Bacillus stearothermophilus. J Biosci 27, 233–242.

25 Gasteiger E, Hoogland C, Gattiker A, Duvaud S,

38 Dalziel K (1957) Initial steady-state velocities in the evaluation of enzyme-coenzyme-substrate reaction mechanism. Acta Chem Scand 11, 1706–1723.

39 Taylor RT & Weissbach H (1965) Radioactive assay for serine transhydroxymethylase. Anal Biochem 13, 80–84.

Wilkins MR, Appel RD & Bairoch A (2005) Protein identiﬁcation and analysis tools on ExPASy server. In Proteomics Protocol Handbook (Walker JM, ed.), pp 571–607. Humana Press, Totowa, NJ.

40 Forbes WF, Ralph AS & Gosine R (1958) Light-

26 Schirch V & Mason M (1963) Serine transhydroxymeth-

absorption studies: part X. Ultraviolet spectra in acid and basic media-some further observations on the ortho effect. Can J Chem 36, 869–878.

ylase: a study of the properties of a homogeneous enzyme preparation and of the nature of its interaction with substrates and pyridoxal-5 phosphate. J Biol Chem 238, 1032–1037.

41 Chen MS & Schirch V (1973) Serine hydroxymethyl- transferase: studies on the role of tetrahydrofolate. J Biol Chem 248, 7979–7984.

42 Contestabile R, Paiardini A, Pascarella S, di Salvo ML, D’Aguanno S & Bossa F (2001) l-Threonine aldolase, serine hydroxymethyltransferase and fungal alanine racemase: a subgroup of strictly related enzymes special- ized for different functions. Eur J Biochem 268, 6508– 6525.

43 Leskovac V (2003) Comprehensive Enzyme Kinetics.

27 Ulevitch RJ & Kallen RG (1977) Puriﬁcation and char- acterization of pyridoxal 5¢-phosphate dependent ser- ine hydroxymethylase from lamb liver and its action upon b-phenylserines. Biochemistry 16, 5342–5350. 28 Bhatt AN & Bhakhuni V (2008) Characterization of pyridoxal-5-phosphate binding domain and folding intermediate of Bacillus subtilis serine hydroxymethyl- transferase: an autonomous folding domain. J Biochem 144, 295–303.

Springer, New York, NY.

29 Dawson RMC, Elliott DC, Elliott WH & Jones KM

(1986) Data for Biochemical Research, 3rd edn. Oxford University Press, New York, NY.

30 Fu TF, Hunt S, Schirch V, Safo MK & Chen BH

44 Vidal L, Calveras J, Clapes P, Ferrer P & Caminal G (2005) Recombinant production of serine hydrox- ymethyltransefrase from Steptococus thermophilus and its preliminary evaluation as a biocatalyst. Appl Micro- biol Biotechnol 68, 489–497.

45 Miyazaki SS, Toki S, Izumi Y & Yamada H (1987)

(2005) Properties of human and rabbit cytosolic serine hydroxymethyltransferase are changed by single nucleo- tide polymorphic mutation. Arch Biochem Biophys 442, 92–101.

Puriﬁcation and characterization of a serine hydroxym- ethyltransferase from an obligate methylotroph, Hypho- microbium methylovorum GM2. Eur J Biochem 162, 533–540.

31 Malerba F, Belleli A, Giorgi A, Bossa F & Contestabile R (2007) The mechanism of addition of pyridoxal-5- phosphate to Escherichia coli apo-serine hydroxymethyl- transferase. Biochem J 404, 447–485.

32 Schirch V (1984) Folates in serine and glycine metabo- lism. In Folate and Pterins (Blankley RL & Benkovic SJ, eds), pp 399–431. Wiley, New York, NY.

46 Sheppard CA, Trimmer EE & Matthews RG (1999) Puriﬁcation and properties of NADH-dependent 5,10-methylenetetrahydrofolate reductase (MetF) from Escherichia coli. J Bacteriol 181, 718–725. 47 Rosenthal S, Smith LC & Buchanan JM (1965)

33 Schirch V & Ropp M (1967) Serine transhydroxymethy- lase. Afﬁnity of tetrahydrofolate compounds for the enzyme and enzyme-glycine complex. Biochemistry 6, 253–257.

34 Stover P & Schirch V (1991) 5-Formyltetrahydrofolate

Enzymatic synthesis of the methyl group of methionine: IX Transmethylation from S-adenosylmethionine and 5-methyltetrahydrofolate to 2-mercaptoethanol and homocysteine. J Biol Chem 240, 836–843.

48 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilization the principle of protein-dye binding. Anal Biochem 72, 248–254.

polyglutamates are slow tight binding inhibitors of serine hydroxymethyltransferase. J Biol Chem 266, 1543–1550. 35 Shostak K & Schirch V (1988) Serine hydroxymethyl- transferase: mechanism of the racemization and trans- amination of d- and L-alanine. Biochemistry 27, 8007– 8014.

36 Jagath JR, Sharma B, Rao NA & Savithri HS (1997)

49 Patil PV & Ballou DP (2000) The use of protocatech- uate dioxygenase for maintaining anaerobic conditions in biochemical experiments. Anal Biochem 286, 187–192.

50 Schirch V & Gross T (1968) Serine hydroxymethyltrans- ferase: identiﬁcation as the threonine and allo-threonine aldolase. J Biol Chem 243, 5651–5655.

The role of His-134, -147, and -150 residues in subunit assembly, cofactor binding, and catalysis of sheep liver cytosolic serine hydroxymethyltransferase. J Biol Chem 272, 24355–24362.

FEBS Journal 276 (2009) 4023–4036 ª 2009 The Authors Journal compilation ª 2009 FEBS

4036

K. Sopitthummakhun et al. Biochemical characterization of SHMT from P. vivax