Nucleoside-triphosphatase and hydrolysis of thiamin triphosphate in Escherichia coli

74

Biochimiea et Biophvsica Acta 923 (1987) 74-82

Elsevier BBA 22390

Nucleoside-triphosphatase and hydrolysis of thiamin triphosphate in Escherichia coli Takahiro Nishimune a, Shin-ichi Ito a, Mitsuko Abe a, Mitsuaki Kimoto a and Ryoji Hayashi b " Department of Microbiology, Yamaguehi Universi O, School of Medicine, Ube and h Medical College of Oita, Oita (Japan)

(Received 17 October 1986)

Key words: Thiamin triphosphate; Thiamin-triphosphatase: Thiamin-diphosphate kinase; Nucleoside-triphosphatase

A membrane-bound nonspecific triphosphatase of E. coli was solubilized and purified to a homogeneous SDS-acrylamide gel electrophoresis band. It was found to be a single polypeptide of 16 kDa requiring no Mg 2+, with an optimal pH at 6.5. The substrate specificity was broad and a nonspecific Mg2+-independent ribonucleoside-triphosphatase (NTPase) activity was expressed together with thiamin-triphosphatase activity. The molecular size and characteristics were clearly different from the known NTPases (EC 3.6.1.15). Using the purified thiamin-triphosphatase II, ATP: thiamin-diphosphate phosphoryl transferase (EC 2.7.4.15) activity was demonstrated with an optimal pH of approx. 5.3. Considering its kinetic parameters and other characteristics, however, the thiamin triphosphate synthesizing activity was not thought to take part in cellular thiamin triphosphate synthesis. The possibility that thiamin-triphosphatase lI plays a part in the hydrolysis of thiamin triphosphate to control its cellular level is suggested.

Introduction In a preliminary study, we found thiamin-triphosphatases of E s c h e r i c h i a coli. Mg2+-dependent thiamin-triphosphatase I was mainly in the soluble fraction and t h i a m i n - t r i p h o s p h a t a s e II was solubilized from the membrane. Thiamin-triphosphatase II had a smaller K m for thiamin triphosphate, and it seemed likely that this enzyme had some role in the control of cellular thiamin triphosphate at relatively low concentrations. We have tried to purify thiamin-triphosphatase II and then to examine the relation between its thiamint r i p h o s p h a t a s e and n u c l e o s i d e - t r i p h o s p h a t a s e Correspondence: Dr. T. Nishimune, Osaka Prefectural Institute of Public Health, Nakamichi, Higashinari, Osaka, 537, Japan.

(NTPase) activities. There appears to be no close relationship between the two enzyme activities in the rat [1,2]. It has also been reported that thiamin-diphosphatase and nucleoside-diphosphatase ( N D P a s e ) in bovine liver and rat brain are two activities expressed by one enzyme [3,4]. The results of our studies showed that thiamin-triphosphatase II of E. co# displays a novel nonspecific N T P a s e activity. A n earlier report [5] explored biosynthesis of thiamin triphosphate and suggested the presence of a thiamin triphosphate synthesizing activity in a subcellular fraction of M y c o b a c t e r i u m . We tried to detect it in E. coli and found that, a m o n g several subcellular fractions which showed the activity, the same fraction as thiamin-triphosphatase II was included. However, characterization of apparent thiamin triphosphate synthesizing activity revealed

0304-4165/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)

75 that optimum conditions for this activity differed from the physiological state of bacterial cells both kinetically and in other characteristics. The apparent reversible character of the thiamin-triphosphatase II reaction was shown by detecting both thiamin-triphosphatase and thiamin-diphosphate kinase activities in the purified enzyme preparation. Materials and Methods

Strain, medium, and culture conditions. Thiamintriphosphatase II was prepared from cells of Escherichia coli W70-23 (thiamin-) grown in a synthetic minimal medium [6] at double strength supplemented with 1 • 10-8 M thiamin. Cells were grown for approx. 12 h at 37°C under aerobic conditions to early stationary phase. Assay of thiamin-triphosphatase H and NTPase. The reaction mixture contained enzyme, 0.1 M Tris-maleate buffer (pH 6.5), 40 mM KC1, 0.2 mM 2-mercaptoethanol, 0.05% Lubrol PX and 0.5 mM thiamin triphosphate or nucleoside triphosphate in 50/tl. The mixture was incubated for 2 h at 37°C and the Pi liberated was determined by the method of Concustell et al. [7], or by the method of Baginski et al. [8] for pre-gel filtration samples. Since thiamin-triphosphatase II expressed acidic NTPase activity, both thiamin triphosphate and ATP were used as substrate where indicated. The inhibitory effect of thiamin triphosphate against NTPase was studied using [32p]ATP as substrate, and approx. 0.26 /~mol K H 2 P O 4 was added as carrier before developing a 10 /~1 aliquot of the reaction mixture on a cellulose MN300G thin layer plate (Macherey-Nagel), using tert-pentan o l / f o r m i c a c i d / w a t e r ( 3 : 2 : 1 , v / v ) [9]. The Pi spot was detected by spraying with Concustell's reagents (urea-albumin-phosphate solution followed by malachite green solution) and developing at 37°C for 15 min. The green spot was removed, added to a scintillation vial with a toluene based scintillator and counted in a liquid scintillation counter.

Assay of thiamin-diphosphate kinase and NDP kinase. The assay mixture contained enzyme, 0.1 M potassium acetate buffer (pH 5.5), 40 mM KC1, 2 mM 2-mercaptoethanol, 0.06% Lubrol PX, 0.27 M thiamin diphosphate or CDP and 0.5 M [,/-

32p]ATP (2.5 m C i / m m o l ) in 25 #1. The reaction was carried out for 1 or 2 h at 37°C, and 5 ktl 25 mM thiamin triphosphate or CTP was added as carrier. The reaction proceeded linearly for at least 3 h, and the rate was also linearly dependent on the amount of enzyme. For the Mg2+-NDP kinase assay, the buffer was Tris-HC1 (pH 8.0), and 10 mM MgC12 was added with 4 mM CDP and 6 mM ATP [10]. The reaction was carried out for 10 min and carrier was added as above. 10 /~1 was analyzed via high voltage paper electrophoresis at p H 6.0 (0.1 M sodium phosphate buffer), and at 1. kV and 100-300 mA for approx. 3 h in n-hexane maintained at below 10°C using Toyo No. 50 paper (60 cm length). After drying, thiamin triphosphate and other components were detected under ultraviolet light. The areas were cut out of the paper, added to a vial containing toluene based scintillator and counted in a liquid scintillation counter (efficiency, approx. 70%). A control, which had not received the substrate, was also assayed and the count level was subtracted from that of test assays. For assay of [32p]pi liberated during thiamin [32p]triphosphate synthesis (Fig. 3b), the assay mixture was developed on a cellulose MN300G thin-laver plate as above.

Solubilization of thiamin-triphosphatase 11. E. coli W70-23 cells were grown as above, pelleted, washed with cold 0.01 M Tris-HC1 (pH 7.4) containing 0.03 M KC1 at 4°C and shocked osmotically. The shocked cell was dispersed in 10 mM Tris-HC1 (pH 7.4) containing 0.5 mM 2-mercaptoethanol (40 m l / l i t e r culture) and disrupted in a milder manner by sonication. The undisrupted cells were spun down and washed with 30 m l / l i t e r culture of the same buffer. From the supernatant and the washing combined, membrane fraction was collected by 150 000 x g centrifugation at 4 ° C for 120 min and washed with the same buffer. To the membrane fraction, 5 M NaC104 (in 0.1 M Tris-HC1 (pH 7.4), 0.5 mM 2-mercaptoethanol and 1 mM EDTA) was added to give 0.5 M NaCIO 4 and shaken gently for 20 min at 30°C. The solubilized thiamin-triphosphatase was obtained in the supernatant from the 150000 x g centrifugation at 4°C and dialyzed immediately as described below.

Column chromatographic purification of thiamin-triphosphatase II. Procedures were carried

76 out at 4°C. Sample for Sephacryl S-300 gel filtration was obtained after NaC104 solubilization, as described above, of the membrane fraction of E. coli W70-23, which had been cultured in three batches of 15 1 medium. The solubilized preparation was dialyzed immediately against 50 mM Tris-HC1 (pH 7.4) containing 1 mM 2-mercaptoethanol, 1 mM EDTA and 0.05% Lubrol PX (buffer 1) added with 0.1 M KC1, and concentrated by ultrafiltration before applying to the Sephacryl column. The sample for phosphocellulose chromatography was dialyzed against buffer 2 (10 mM Tris-maleate (pH 6.5), 1 mM EDTA, 1 mM 2-mercaptoethanol, 0.05% Lubrol PX, and 0.17 M KC1). The column was pre-equilibrated and washed after sample application with the same buffer, and developed with a linear KC1 gradient of 0.17 1.2 M (buffer 2). The sample for phenylSepharose chromatography was passed, just before application, through a Sephadex G-25 column to equilibrate with 10 mM Tris-maleate (pH 6.5) containing 1 mM EDTA and 1 mM 2-mercaptoethanol (buffer 3) supplemented with 4 M NaCI. Elution from a phenyl-Sepharose column (approx. 10 ml gel) was carried out by a linear gradient of NaC1 and nonionic detergent starting from buffer 3 containing 4 M NaC1 to the final buffer 3 containing 0.4% Lubrol PX. Each fraction from phenyl-Sepharose was processed immediately through columns of Sephadex G-25 (2 × 8 cm) to re-equilibrate the buffer into buffer 1 containing 0.1 M KC1. The sample of Blue Sepharose was dialyzed against 10 mM Tris-HC1, 0.5 mM 2-mercaptoethanol, 0.5 mM EDTA, 0.04 M KCI and 0.05% Lubrol PX (pH 6.5) (buffer 4). Thiamin-triphosphatase II was eluted by a linear gradient of ATP and pH from 2 mM ATP in buffer 4 (pH 6.5) to 18 mM ATP in buffer 4 (pH 7.4). Analytical methods and reagents. Molecular weight estimation by Sephadex G-200 chromatography (2.5 × 90 cm) was carried out in buffer 1 containing 0.1 M KC1. Standard proteins used were ribonuclease A 13.7 kDa, chymotrypsinogen A 25 kDa, ovalbumin 43 kDa, and bovine serum albumin 67 kDa. Polyacrylamide gel electrophoresis was carried out at pH 4.5 [11] with 7% acrylamide. The sample for the detection of activity was overlaid on the stacking gel with 15% sucrose and approx. 50 mM 2-mercaptoethylamine [12]. After

running the gel at 4°C, the gel was cut into 2 mm lengths, the pieces were ground and extracted with 0.2 M Tris-maleate buffer (pH 6.5) containing 0.5 mM 2-mercaptoethanol and 0.1 m g / m l bovine serum albumin. The recovery was around 10% for thiamin-triphosphatase II and even lower for thiamin-diphosphate kinase activity. In polyacrylamide gel at pH 8.9 (7.5% gel), the activity did not penetrate into the gel toward the anode. SDS-acrylamide gel was run following the method of Laemmli [13] with 15% acrylamide. Protein was measured, depending on the stage of purification, by the Biuret method [14], or the method of Lowry et al. [15], or Bradford [16]. The N-terminal amino acid was determined by a gas phase amino acid sequencer (Applied Biosystems, type 470A). Isoelectric focusing was carried out using Ampholines (pH 7-9) in a flat bed equipment (LKB) for 18 h at 4°C. The sample (450/~mol/g per h) was from the Sephacryl S-300 eluate. After electrofocusing, each fraction was eluted with 0.4 M Tris-maleate buffer (pH 6.5), dialyzed against the same buffer of 0.1 M, and a sample of it was developed by polyacrylamide gel electrophoresis (pH 4.5) at 4°C. The gel was fractionated and extracted as above to measure thiamin-triphosphatase II. Thiamin triphosphate was a generous gift of Dr. M. Yamazaki of Sankyo Co., Tokyo. /3,y-Methylene-ATP (miles Laboratories), phosphocellulose P l l (Whatman), phenyl-Sepharose CL-4B and Blue Sepharose CL-6B (Pharmacia Fine Chemicals), and Lubrol PX (Sigma) were commercially obtained. Results

Co-purification of thiamin-triphosphatase 11 and NTPase Thiamin-triphosphatase II was solubilized from the membrane fraction of E. coli W70-23 as described in Materials and Methods and purified further by gel filtration, ion exchange and group specific affinity chromatography (Table I and see below). Throughout these steps, thiamin-triphosphatase II co-purified with MgZ+-independent, acid optimum ATPase activity (activity ratio of NTPase/thiamin-triphosphatase was approx. 1). Substrate specificity among nucleoside phosphates and thiamin phosphates was measured (Table II).

77 TABLE I P U R I F I C A T I O N OF NTPase ( T H I A M I N - T R I P H O S P H A T A S E ) A N D T H I A M I N - D I P H O S P H A T E K I N A S E n.d., not determined. Protein (/~g) a

Sephacryl eluate Phosphocellulose eluate Sephadex G-100 eluate Phenyl-Sepharose eluate Blue Sepharose eluate

364000 18 800 280 25 2

NTPase ~/ ThTPase

~ m o l / g per h

Thiamindiphosphate kinase/ ThTPase b

4.34 57.4 3.4.103 35 • 103 0.31-106

0.040 0.067 0.081 0.097 0.83

1.09 1.00

NTPase ( thiamin- triphosphat ase)

Thiamin-diphosphate kinase

/Lmol/h a

/~mol/g per h

/,mol/h a

39.9 16.2 11.7 9.34 7.35

0.11.103 0.86.103 41.8 - 103 359 - 103 3.7 - 106

1.58 1.08 0.95 0.91 0.61

n.d. n.d. 1.06

a Total protein and total activity (measured as thiamin-triphosphatase) for 45 liter culture. b Thiamin-triphosphatase II. Measured as ATPase at pH 6.5 without Mg 2+.

The results indicate that this phosphatase was a nonspecific NTPase exhibiting more activity towards thiamin triphosphate and ribose-NTPs than the deoxy-form and the diphosphates. These two T A B L E II SUBSTRATE P H A T A S E II

SPECIFICITY

OF

THIAMIN-TRIPHOS-

Substrate

Activity (%)

T h i a m i n triphosphate Thiamin diphosphate Thiamin monophosphate Glucose-6-P p-Nitrophenyl phosphate /3-3,-Methylene-ATP ATP GTP CTP UTP dATP dGTP dCTP dUTP dTI?P ADP GDP CDP UDP AMP dAMP dGMP dCMP dTMP

1(30 a 13 0 0 8 0 103 102 105 80 89 83 79 85 82 35 36 17 11 0 3 2 0 0

100% corresponded to 440 /~mol/g per h.

activities of thiamin-triphosphatase II and riboseNTPase were found to be catalyzed by a single protein since they were mutually inhibited, as shown in Fig. la and b. To determine whether the amount of thiamin diphosphate decreased by the addition of ATP through the mechanism of NDPase activation [3], thiamin monophosphate and free thiamin were measured in all samples of Fig. la, but the results were nearly equal to zero. Thus, participation of NDPase was excluded. 150

b

11 1012 .c o~

~lcx ~5

i5 E :a.

E :a,

gs0

O O.

i-ei--

0

3 5 ATP (raM)

0

1

3 5 ThTP (mIM)

Fig. 1. Apparent inhibition of thiamin-triphosphatase II by the two substrates ATP and thiamin triphosphate. (a) Thiamin-triphosphatase II (ThTPase) assay was carried out as in Materials and Methods but at 5-times the scale (250/zl) for approx. 15 h and the thiamin phosphates were analyzed by a gel filtration method [24]. Activity was calculated from the a m o u n t of thiamin diphosphate formed. (b) NTPase was assayed as in Materials and Methods by developing on a thin layer of M N 3 0 0 G cellulose and the activity was expressed as the a m o u n t of [32pIP i formed.

78

Q

enzyme. Further, both activities were recovered at exactly the same position on polyacrylamide gel electrophoresis (data not shown). Hence, we concluded that thiamin-triphosphatase II was a kind of nonspecific ribose-NTPase. Considering the high activity shown on thiamin triphosphate and the rareness of an enzyme active on both nucleoside phosphates and thiamin phosphates, we called this enzyme thiamin-triphosphatase II. This enzyme showed differences from the k n o w n NTPases (EC 3.6.1.15) as described below.

b

~°cI

100

:a,~Pase

o2 E z~

..~

0

ThTPose~ 50 100 MgCl2 ( rnM )

0

I

2

3

4

p.-f -CH2-ATP (ram)

Fig. 2. Parallel inhibition by Mg 2+ and fl,),-methylene-ATP of thiamin-triphosphatase II and acid NTPase. Thiamin-triphosphatase II (ThTPase) (Sephacryl S-300 eluate) was assayed according to Materials and Methods except for the absence of detergent from the reaction mixture and the addition of Mg 2 + (a) or fl,y-methylene-ATP (b).

Thiamin-triphosphatase II shows Mg 2+ inhibition (data not shown). The effect of Mg 2+ on ribose-NTPase was similar (Fig. 2a). B,7-Methylene-ATP, an inhibitor of NTPase, inhibited thiamin-triphosphatase II as if it was an analogue of thiamin triphosphate (Fig. 2b). Those results indicated that the activities of thiamin-triphosphatase II and ribose-NTPase were expressed by the same Q

1.C

~ 0.~

~

:k

."

~.az

A~ATP

b

~1C E

/,"

?

~'0.; 2

4 6 Time ( h )

0

2

4 6 Time ( h )

Fig. 3. Thiamin-diphosphate kinase activity detected in a thiamin-triphosphatase II preparation. A thiamin-triphosphatase II preparation (1.9 mmol/g per h) was used for the detection of thiamin-diphosphate kinase activity. The assay mixture contained 2.5 mM MgC12 in the case of hatched circle data. (a) Time-dependent accumulation of thiamin [32p}triphosphate. (b) The same sample as in (a) was analyzed for the remaining [32p]ATP and [32p]p i released. ThTP, thiamin triphosphate.

Thiamin diphosphate kinase activity of thiamin-triphosphatase H W h e n E. coli was loaded with thiamin diphosphate, cellular thiamin triphosphate increased (data not shown). One of these thiamin triphosphate synthesizing activities ( A T P : thiamin-diphosphate phosphotransferase, EC 2.7.4.15) was assigned to the same fraction as thiamin-triphosphatase II when surveyed in a preliminary experiment. This reaction proceeded linearly for more than 3 h (Fig. 3a), with the simultaneous hydrolysis of approx. 20-times the n u m b e r of moles of A T P (Fig. 3b) (see Table I for purified preparations; the ratio thiamin-diphosphate k i n a s e / thiamin-triphosphatase II). The enzyme used (Fig. 3) was free from k n o w n Mg2+-dependent, alkaline o p t i m u m N D P kinase (EC 2.7.4.6) (see Fig. 5). The substrate specificity of thiamin-diphosphate kinase activity was relatively broad, exhibiting comparable activities on some other nucleoside diphosphates (Table III). A parallel effect on p H on thiamin-diphosphate kinase and on the Pi liberation from [y-32p]ATP (observed with the kinase reaction) was seen (data not shown). Further, the two activities were recovered from exa c t l y the same position on polyacrylamide gel electrophoresis (data not shown).

Co-purification of thiamin triphosphatase II and thiamin-diphosphate kinase To confirm, that thiamin-triphosphatase II expresses a thiamin-diphosphate kinase activity, we purified thiamin-triphosphatase II from E. coli W70-23. The solubilized thiamin-triphosphatase II was developed on a Sephacryl S-300 colunm (Fig. 4). The profile showed that the major fraction of MgZ+-ATPase (EC 3.6.1.3) eluted differently from

79 T A B L E III SOME C H A R A C T E R I S T I C S O F T H E T W O ACTIVITIES OF T H I A M I N - T R I P H O S P H A T A S E II Substrate

Thiamin-diphosphate kinase activity (%)

Thiamin diphosphate Thiamin monophosphate dTDP CDP UDP

expt. 1

expt. 2

100 a 17 107 49 38

100 0 87 68 60

T h i a m i n diphosphate + 10 m M E D T A Thiamin diphosphate + 10 m M dithiothreitol

a

83

Additives

Thiamin- t riphosphat ase II activity (%)

control 1 m M thioglycollic acid 1 m M dithiothreitol 1 mM EDTA 1 m M metal mixture (Cu, Zn, Mn, Ca)

100 b 99 103 101 72

5 m M pyrithiamin

84

85

a 100% corresponded to 6 0 / ~ m o l / g per h. b 100% corresponded to 2.5 m m o l / g per h.

thiamin-triphosphatase II, and that thiamin-diphosphate kinase paralleled thiamin-triphosphatase II. Active fractions were pooled and applied on a phosphocellulose column. Thiamin-diphosphate kinase activity showed affinity for the phosphocellulose, whereas MgZ+-dependent, alkaline optimum, NDP kinase (EC 2.7.4.6) did not bind at pH 6.5 and eluted in the flow-through fractions (Fig. 5). This activity (flow-through fractions) was not active on thiamin diphosphate at

pH 6.5 (Fig. 5) or at pH 8.0 with Mg 2+ added (data not shown). Thiamin-triphosphatase II was eluted in two peaks by KC1 concentrations higher than 0.5 M, accompanied by thiamin-diphosphate kinase activity. The amount of former peak varied from experiment to experiment and was usually minor. The latter peak was pooled and applied on Sephadex G-100 column. Thiamin-triphosphatase II was eluted at the fractions which corresponded 45.

A1

-g 1,0 t-

/I

o

E

&,0-

t~

I'

i a.

=.35.

J'o

"

i~

A28Orv~I 'I

1.0

t~ 3.0-

~o!

''

h,

I

/i~

1

I

E

C

E

I-.

~

__-v/_°_ _L_v¢

x n

100

200 300 &O0 Elution volume ( m l )

06"

=k

0.4"

500

Fig. 4. Gel filtration chromatography of thiamin-triphosphatase II (ThTPase II). A membrane fraction (60.8 g protein) of E. coli W70-23 (45 1 culture) was solubilized giving 1.61 g of solubilized protein and applied on a Sephacryl S-300 superfine (2.5 × 95 cm) column and developed with 'buffer 1' containing 0.1 M KC1. Thiamin-triphosphatase II (ThTPase II) activity was assayed using A T P as substrate at p H 6.5. Mg2+-ATPase was assayed at p H approx. 8.0 using [32p]ATP as the substrate and the activity was expressed by the a m o u n t of [32p]ATP degraded. Thiamin-diphosphate (ThDP) kinase was assayed as in Materials and Methods. ThTP, thiamin triphosphate.

o

"~ 0.2'

Q.

1

0

-"~:

•

10

J

210

-J

i

30

/~0

Fraction number

Fig. 5. Separation of thiamin-diphosphate (ThDP) kinase from MgZ+-NDP kinase by phosphocellulose chromatography. The column (22 ml wet volume) was eluated with a linear gradient of KC1. Eluate was fractionated into 10 g in fractions Nos. 1-15, or into 4.5 g in subsequent fractions. ThTP, thiamin triphosphate.

80

ct

b

ThTPose A/

[

ThTPose

o o.6r2o~

0

50 1(30 150 Eiutior~ v o l u m e (ml)

20 40 60 80 ~00 E l u t i o n volume ( m[ )

Fig. 6. Co-purification of thiamin-triphosphatase II with thiamin-diphosphate (ThDP) kinase and acidic NTPase. (a) Phenyl-Sepharose CL-4B column chromatography. Elution was accomplished by a linear gradient of NaC1 and Lubrol PX, but only the NaC1 is plotted. Each eluate fraction was applied on Sephadex G-25 column to re-equilibrate the buffer into 'buffer 1' containing 0.1 M KC1. Enzyme activities were assayed for this eluate. Thiamin-triphosphatase (ThTPase) II activity was assayed with ATP as the substrate at pH 6.5. (b) Blue Sepharose CL-6B column chromatography. To 2.5 ml wet volume of Blue Sepharose CL-6B, the sample eluted from phenyl-Sepharose was applied after dialysis against 'buffer 4'. The column was washed with 'buffer 4 containing 2 mM ATP and eluted with a gradient of ATP and pH (only ATP concentration is shown) with a flow rate of 2 ml/h. ThTP, thiamin triphosphate; ThDP, thiamin diphosphate.

~

~i

~

~

!!ii

i i¸

ilii~

t

IF

Fig. 7. SDS-polyacrylamide gel eleetrophoresis analysis of thiamin-triphosphatase II in each step of purification. Samples were analyzed with 15% gel and detected by silver staining. 1, NaC1Oa-solubilized sample; 2, Sephacryl S-300 eluate; 3, phosphocellulose eluate; 4, Sephadex G-100 eluate, 5, phenyl-Sepharose eluate; 6 and 8, Blue Sepharose eluate; 7, molecular weight standards (67 kDa, 43 kDa, 25 kDa and 13.7 kDa; 100 ng each was applied).

to a molecular weight of approx. 16 000 (data not shown, see below for the molecular weight). The fractions containing activity were concentrated with a membrane filter (Diafilter G05T), and the buffer was changed to a detergent-free high-salt one and applied on a column of phenyl-Sepharose. Thiamin-diphosphate kinase activity and thiamintriphosphatase II were eluted at approx. 1.5 M NaC1 plus 0.25% Lubrol PX (Fig. 6a). The pooled fractions containing the activity exhibited contaminating bands on the SDS-polyacrylamide gel electrophoresis (Fig. 7), and were purified further by Blue Sepharose CL-6B (Fig. 6b). In contrast to the known examples, thiamin-triphosphatase II was adsorbed firmly at acidic pH, and eluted at alkaline pH or when ATP was added. A combined linear gradient of 5.0-6.0 mM ATP and pH of 5.3-7.6 gave a better purification. Each step was analyzed with SDS-polyacrylamide gel electrophoresis (Fig. 7), as summarized in Table I.

Characteristics of thiamin-triphosphatase II In the assay of thiamin-diphosphate kinase, approx. 10-times the number of moles of ATP was hydrolyzed in the purified enzyme (Table I, thiamin-diphosphate kinase/thiamin-triphosphatase II). A saturating concentration of thiamin diphosphate was used in these assays ( K m for thiamin diphosphate was 56 mM and for ATP was 10 mM, data not shown). The optimum pH of thiamin-diphosphate kinase activity was approx. 5.3 (data not shown). The effect of temperature was studied and the optimum was approx. 50°C (data not shown). The effect of Mg 2+ was inhibitory and a curve very similar to Fig. 2a was obtained (data not shown). The molecular weight of thiamin-triphosphatase II was estimated by SDS-acrylamide gel electrophoresis (Fig. 7), giving a value of 15900, and also by gel filtration chromatography with Sephadex G-200, with the value of 15 100. Thus, thiamin-triphosphatase II was shown to be a single polypeptide of molecular weight approx. 16000. Using an eluate of Blue Sepharose, the N-terminal amino acid was determined to be glutamic acid. In experiments to determine the isoelectric point of thiamin-triphosphatase II, activity peaks were found at pH 8.54 (main peak), 8.30, 8.0, 7.6, and 7.00 (graph not shown).

81

Discussion A bacterial thiamin-triphosphatase of E. coli was evidenced for the first time in a purified form. Purified enzyme was found to be accompanied by acid NTPase activity. Another example of an enzyme which shows reactivity on both thiamin phosphate and nucleoside phosphate is thiamindiphosphatase of bovine liver and rat tissue [3,4]. In rat NDPase, two types of enzymes with clear differences in the pattern of isoelectric focusing have been observed [4]. Thiamin-triphosphatase II in this paper showed the typical pattern of a multi-isoelectric point enzyme and resembled more the brain-type thiamin diphosphatase of rat. These enzymes also resembled each other in that both thiamin phosphate and nucleoside phosphate could act as substrate at neutral pH. We concluded that thiamin-triphosphatase II showed the activity of a nonspecific NTPase on the basis of the following observations. (i) The two activities were co-purified throughout the steps. (ii) Both activities were detected at the same position on acrylamide gel electrophoresis. (iii) The two activities were inhibited similarly by Mg 2÷ and /3,7-methylene-ATP. (iv) Thiamin-triphosphatase II was inhibited by ATP, and the Mg 2+independent ATPase activity was inhibited by thiamin triphosphate. (v) Ribonucleoside triphosphates other than ATP were hydrolyzed at nearly the same rate. As the substrate, deoxyNTPs were 80-90% active, nucleoside diphosphates were approx. 30% active, and monophosphates were almost inactive. However, NTPase (EC 3.6.1.15) already reported [17-21] are all clearly different from thiamin-triphosphatase II. Their molecular sizes are between 100 kDa [20] and 260 kDa [17], and some of them [19] require dithiol for activation. Thus, thiamin-triphosphatase II is a novel NTPase of E. coli. On the other hand, the well known Mg2+-de pendent, alkaline optimum N D P kinase (EC 2.7.4.6) showed no activity on thiamin diphosphate (Fig. 5). Enzymes for nucleoside metabolism seem to be strictly differentiated from thiamin metabolism in vivo, and there should be some particular function(s) for those enzymes which can act on both types of substrates. Besides the activity of thiamin-triphosphatase II, the purified en-

zyme catalyzed the reverse directed reaction. However, the kinetic parameters and characteristics of the reverse reaction ( K m for thiamin diphosphate of 56 mM, for ATP of 10 mM, optimum pH of 5.3, hydrolysis of approx. 10-times the number of moles of ATP) indicated that the thiamin-triphosphatase II activity was relatively favoured in the physiological state of the bacterial cell. Incidentally, the alkaline optimum N D P kinase of E. coli detected in Fig. 5 were recovered in the fractions which corresponded to an approximate molecular size of 16 kDa (Fig. 4). Thus E. coli Mg2+-NDP kinase did not agree in size with the reported N D P kinase of Salmonella typhimurium (85 kDa) [22] or of Bacillus subtilis (100 kDa) [23]. On the basis of the data obtained, we deduce that some other enzyme than thiamin-triphosphatase II is responsible for the cellular thiamin triphosphate synthesis, and thiamin-triphosphatase I is now under investigation.

Acknowledgements The authors express their gratitude to Dr. M. Yamazaki of Sankyo Co., Tokyo for his generous gift of thiamin triphosphate.

References 1 Barchi, R.L. and Braun, P.E. (1972) J. Biol. Chem. 247, 7668-7673 2 Barchi, R.L. (1976) in Thiamine (Gubler, C.J., Fujiwara, M. and Dreyfus, P.M., eds.), pp. 195-225, John Wiley & Sons, New York 3 Yamazaki, M. and Hayaishi, O. (1968) J. Biol. Chem. 243, 2934-2942 4 Sano, S., Matsuda, Y., Miyamoto, S. and Nakagawa, H. (1984) Biochem. Biophys. Res. Commun. 118, 292-298 5 FossiFanelli, A., Ipata, P.L. and Fasella, P. (1961) Biochem. Biophys. Res. Commun. 4, 23-27 6 Davis, B.D. and Mingioli, E.S. (1950) J. Bacteriol. 60, 17-23 7 Concustell, E., Cortes, M., Ferragut, A. and Gener, J. (1977) Clin. Chim. Acta 81,267-272 8 Baginski, E.S., Foa, P.P. and Zak, B. (1967) Clin. Chim. Acta 15, 155-158 9 Randerath, K. (i962) Biochem. Biophys. Res. Cornmun. 6, 452-457 10 Ingraham, J. and Ginther, C.L. (1978) Methods Enzymol. 51,371-375 11 Reisfeld, R.A., Lewis, U.J. and Williams, D.E. (1962) Nature 195, 281-283

82 12 Brewer, J.M. (1967) Science 156, 256-257 13 Laemmli, U.K. (1970) Nature 277, 680-685 14 Gornall, A.G., Bordawill, C.J. and Maxima, M.D. (1949) J. Biol. Chem. 177, 751-766 15 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 16 Bradford, M.M. (1976) Anal. Biochem. 72, 248-254 17 Lanka, E., Mikolajczyk, M., Schlicht, M. and Schuster, H. (1978) J. Biol. Chem. 253, 4746-4753 18 Raacke, I.D. and Fiala, J. (1964) Proc. Natl. Acad. Sci. USA 51, 323-329 19 Asai, T., O'Sullivan, W.J. and Tatibana, M. (1983) J. Biol. Chem. 258, 6816-6822

20 Brightwell, R. and Tappel, A.L. (1968) Arch. Biochem. Biophys. 124, 333-343 21 Lewis, M. and Weissman, S. (1965) Arch. Biochem. Biophys. 109, 490-498 22 Ginther, C.L. and Ingraham, J.L. (1974) J. Biol. Chem. 249, 3406-3411 23 Sedmak, J. and Ramaley, R. (1971) J. Biol. Chem. 246, 5365-5372 24 Nishimune, T. and Hayashi, R. (1980) Experientia 36, 916-918