Indian Journal of Biochemistry & Biophysics Vol. 36, June 1999, pp. 137-142

Purification and characterization of flavokinase from Neurospora crassa

Sivalenka Raja Rajeswari, Vidya S lonnalagaddat and Sobhanaditya lonnalagadda*

Department of Biochemistry, University College of Science, Osmania University, Hyderabad SOO 007

Received 4 January 1999; revised 19 April 1999

The ATP-dependent phosphorylation of riboflavin to FMN by flavokinase is the key step in flavin biosynthesis. Flavokinase has been purified from a fungal source for the first time. The enzyme purified from a cell wall lacking mutant of Neurospora .crassa, slime, is a monomer of Mr 3S.S kDa with maximal activity at alkaline pH and high temperature (SS°C). The Km for both substrates is the lowest repprted for flavokinase from any source so far (120 nM for riboflavin and 210 nM for MgATp2.). The enzyme exhibits preference for Mg2+ over Zn2

+ as the essential activator and is also significantly activated by several cations. Activation by orthophosphate may be physiologically relevant for the intracellular regulation of flavokin ase.

Vitamin B2 (riboflavin) exists in cells mainly as coenzyme forms flavin mononucleotide, FMN and flavin adenine dinucleotide, FAD. The first step in biosynthesis of these flavocoenzymes is the phosphorylation of riboflavin to form FMN by flavokinase [ATP:riboflavin 5' phosphoryl transferase (EC 2.7.1.26)]. Flavokinase activity has been characterized from animals1.2, bacteria3

,4 and plants5.6

.

However no information is available regarding this important enzyme in lower eukaryotes such as fungi.

No isozymes of flavokinase have been reported from any source so far, leading one to believe it may be a conserved, single copy gene. However, several functional differences have been observed In

flavokinase from various organisms, for example, it occurs as a bifunctional flavokinase : FAD synthetase complex in Corynebacterium ammoniagenes7

• In rat liver, flavokinase is a monomeric protein of molecular weight of 28 kDa(ref. I), with Zn2+ as the preferred essential activator8 and follows an ordered sequential Bi-Bi reaction mechanism9

.lo. Flavokinase from mung

bean is a monomer of Mr 35 kDa(ref. 6) with a preference for Mg2

+ as the essential activator ll . Both the rat liver and mung bean enzymes are inactivated by modification of Arg residue(s) by phenyl glyoxaI9

•12

• In addition, mung bean flavokinase shows conformational flexibility in presence of orthophosphate (Pi), which is required for binding to affinity matrices viz. Blue Sepharose, riboflavin

• To whom all correspondence shou ld be addressed t present address: Bharatiya Vidya Bhavan Degree College, Sainikpuri, Secunderabad. Telephone No. 91-040-7017661 , FAX No. 91-040-7019020.

Sepharose and ATP Agarosell .13. Pi also acts as a non­

essential activator of this enzyme· and may function as a physiological regulator of flavin metabolism II .

The aim of this investigation was to purify and characterize flavokinase from a lower eukaryote­fungal source so that the conserved features of this enzyme can be deduced by comparison of its properties with flavokinase from ot.her organisms. The source organism chosen was slime, a cell wall-less mutant of Neurospora crassa.

Material and Methods

Materials Slime (FGSC 4761) was from Fungal Genetics

Stock Center, Kansas, USA. 10 w-amino-octyl Flavin was a generous gift from Prof D B McCormick, Emory University, Atlanta, Georgia, USA . Riboflavin, A TP (equine muscle, disodium salt) , Sephacryl S300, DEAE Sepharose, Sepharose CL-6B, C8-A TP Agarose, bovine serum albumin (BSA) , SDS, ammonium persulphate, I)-mercaptoethanol , standard molecular weight markers, Cibacron Brilliant Blue G250 and silver nitrate were pl;rchased from Sigma Chemical Co, USA. All other reagents used were of analytical grade, purchased from local compames.

Growth of slime Slime was grown in Vogel 's supplemented medium

at 2rC with mild agitation (100 rpm) in an environmental incubator. The culture medium was prepared by mixing Vogel 's 50X stock (2%), mannitol (2%), sucrose (2%), yeast extract (0.75%),

138 INDIAN J BIOCHEM. BIOPHYS, VOL. 36, JUNE 1999

peptone (0.5%), beef extract (0.3%) and trace elements solution stock (0.2%). Vogel's 50X stock contains trisodium citrate (12.5%), potassium dihydrogen phosphate anhydrous (25%), ammonium nitrate (10%), magnesium sulphate heptahydrate (1%) and calcium chloride dihydrate (0.5%). Trace elements solution stock contains citric acid (5%), zinc sulphate heptahydrate (1 %), cupric sulphate pentahydrate (0.05%), manganous sulphate monohydrate (0.05%), boric acid (0.05%) and sodium molybdate dihydrate (0.05%)

Preparation of jIavin Sepharose affinity matrix

Sepharose CL6B, activated by incubation with epichlorohydrin (98%) at 37°C overnight, was coupled to 10 co-amino-octyl flavin in coupling buffer (carbonate-bicarbonate, pH 9.0, overnight incubation at 37°C). The unbound ligand was filtered out and unreacted groups blocked with 0.1 M ethanolamine. The gel was washed three times with 0.1 M acetate buffer (PH 4.0) containing 0.5 M NaCl, water, followed by 0.1 M borate buffer (PH 8.0) containing 0.5 MNaCI and water.

Purification of jIavokinase from slime

Cells from log phase cultures (750 ml of (:ulture yielded around 12 g wet weight of cells) of slime were collected by centrifugation. All subsequent steps were carried out at 4°C. The cells were suspended in 5 volumes (w/v, i.e. 60 ml) of buffer A (50 mMTris.CI, pH. 8.0, 2 mM EDTA, 5 mM j3-mercaptoethanol, 10% (v/v, glycerol) and lysed by sonication. The extract was clarified by centrifugation at 8000xg for 10 min. The supernatant was loaded onto a DEAE Sepharose column (25 ml bed volume) pre-equilibrated with buffer A. The column was washed with buffer A and activity eluted with buffer A contain 100 mM KCl. The active fractions were pooled, mixed with an equal volume of 1 M phosphate buffer, pH 7 .4,and loaded onto the flavin Sepharose column (25 ml bed volume) pre-equilibrated with buffer B (0.5 M phosphate buffer, pH 7.4). The column was washed with buffer B and activity eluted With the buffer contain 100 J.lM riboflavin. The active fractions were pooled and loaded in presence of 1 mM EDT A onto an A TP agarose column (2.5 ml bed volume) pre-equilibrated with buffer B containing 1 mM EDTA. The column was washed sequentially with (i) buffer B, (ii) buffer B containing 0.5 M KC1, and (iii) buffer B, plior to .elution of flavokinase with distilled water.

Protein estimation This was done by micro dye-binding assay of

Bradfordl4 with BSA as standard.

Native and SDS-PAGE Polyacrylamide gel electrophoresis under non­

denaturing conditions was done as described by Garfin l5 and in the presence of SDS was done as described by Thomas et a116

•

Gel filtration Flavokinase was loaded onto a calibrated Sephacryl

S-300 column (1.8 X 80 cm) and individual fractions assayed for activity.

Assays for jlavokinase, FMN hydrolase and FAD hydrolase

Enzyme was incubated at 37°C for 5-10 min in reaction mixture containing 0.2 M Tris.Cl, pH 8.5, 1 mM MgS04 and (a), 10 ~ riboflavin and 40 J.lM A TP for flavokinase or (b), 100 J.lM FMN for FMN hydrolase or (c), 100 J.lM FAD for FAD hydrolase/ pyrophosphatase. Reactions were stopped with 0.2 M formic acid and stored at -20°C until analysis by reverse phase: HPLC.

Reverse phase HPLC analysis of assay samples Riboflavin and FMN were separated as described

earlier I 3. One unit I)f enzyme activity is defined as that producing 1 nmole of FMN per min at 37°C.

Optimum pH and temperature for jIavokinase Flavokinase was assayed in buffers of different pH

ranging from 4 to 10, or under standard assay conditions at different temperatures and FMN produced detl~rmined as described above.

Effect of cations and anions

Flavokinase was assayed in the presence of (a), mM sulphate salts of mono- or di- valent cations or (b), 50 mM sodium or potassium salts ofmono-di-and trivalent anions. FMN produced was determined as described above.

Determination of kinetic constants

Saturation with riboflavin-Flavokinase was assayed in 0.2 M Tris-CI (PH 8.5) in the presence of 40 J.lM ATP, 1 mMMgS04 and 0.1 to 10 J.lM riboflavin for 5 min. Saturation with MgATp2--flavokinase was assayed in 0.2 M Tris-Cl (PH 8.5) in the presence of

RAJESWARI et al.: PURIFICATION AND CHARACTERIZATION OF FLA VOKINASE FROM NEUROSPORA CRASSA 139

Table I - Purification of slime flavokinase

(from 750 ml culture yielding 12 g wet weight of cells)

Stage Volume Protein (ml) (mg)

Crude 60 494 DEAE Sepharose 20 62 Flavin Sepharose 50 0.34 ATP Agarose 3 0.034

(Figures are rounded off to the nearest integer).

10 ~ riboflavin, 1 mM MgS04, and 0.1 to 10 ~ A TP for 8 min. FMN produced was determined as described above.

Analyses of data A commercially available computer program,

Enzfit (Biosoft, Elsevier), was. used for the fit of observed kinetic data to determine values for Km and keat by nonlinear regression to Michaelis Menten equation. Least squares linear regression was used to determine the slopes of Lineweaver Burk plots and Hill plots.

Results and Discussion Purification of jlavokinase

The slime extract was subjected to anion exchange chromatography on DEAE as the first step in the purification. This led to an apparent loss of almost 60% of the total activity. However, an apparent gain of activity was observed upon subjecting the active fractions to affinity chromatography on flavin Sepharose. One reason for this observation could be the co-elution or activation of FMN hydrolase in the DEAE eluate. As the assay procedure estimates the amount of FMN formed in a mixture of riboflavin and A TP, presence of FMN hydrolase, which converts FMN to riboflavin and ri, will lead to an underestimation of the flavokinase activity. Regaining of flavokinase activity following chromatography on flavin Sepharose is consistent with the removal of the hydrolase. Another reason for enhanced activity in the riboflavin Sepharose eluate could be the activating effect of Pi (described later) as the enzyme in eluted in buffer containing 0 .5 M Pi . The final step of affinity chromatography on A TP Agarose yielded a preparation purified 8,500 fold with a specific activity of 3,000 U/mg (Table I). This enzyme showed no contaminating activities of FMN hydrolase or FAD synthetase (data not shown).

The enzyme recovered from A TP Agarose column

Activity Sp. Act. Recovery Fold (U) (U/mg) (%) purity

174 0.35 100 I 66 I 38 3 151 445 87 1270 100 2950 58 8427 .

Fig. I - Purification of flavokinase [Aliquots from various stages of purification using the standard protocol were electrophoresed on 12% SDS-PAGE and stained with Coomassie Blue. Lane I , crude extract; lane 2, DEAE eluate; lane 3, flavin Sepharose eluate; lanes 4 and 6, standard molecular weight markers: phosphorylase b (97.4 kDa), BSA (66.2 kDa), ovalbumin (45 kDa), carbonic anhydrase (31 kDa), soybean trypsin inhibitor (21.5 kDa), lysozyme (14.4 kDa) and aprotinin (6.5 kDa); lanes 5 and 7, A TP agarose eI uate ]

migrated as a single band on SDS-PAGE (Fig. 1) as well as native PAGE (not shown). The molecular mass of the enzyme was determined to be 35.5 kDa by SDS-PAGE and gel filtration (Fig. 2). This is the first time flavokinase has been assayed and purified to homogeneity from a lower eukaryote - a fungus. It resembles the plant enzyme in being a monomer (identical molecular mass under native and denaturing conditions). It also resembles the plant enzyme in its requirement for Pi to induce binding to A TP Agarose (Table 1) and blue Sepharose (data not shown). The final preparation (3 ,000 U /mg) however shows much greater activity than the rat liver (80 U /mg) and mung bean enzyme (100 U/mg).

Enzymatic activity Flavokinase from most sources has a pH optimum

of around 8, but this enzyme showed maximal activity

140 INDIAN J BIOCHEM. BIOPHYS, VOL. 36, JUNE 1999

at pH range 8 to 10 (Fig. 3a). Activity was maximal at temperatures between 37°C and 57°C (Fig. 3b). The high temperature optimum was rather unexpected since the source organism, slime is thermo-sensitive and is cultured at 27°C rather than at 37°C used for the wild type N.crassa.

Mg2+ was found to be the most effective essential activator (Fig. 4a), the order of effectiveness of divalent cations being Mg2+ > C02+ > Zn2+ > Cu2+ > Cd2+ > Fe2+. In this aspect also this enzyme resembles the plant and other microbial enzymes rather than the rat liver flavokinase (which prefers Zn2+ over Mg2+ ).

5.2e (a)'

4.8

4.4

1:

• ADH ... BSA

• Flavokinase

• Cytc ..... .. Regression line

•• ".

0> '<V , • ~ i 4.~wO~~~8eO~~~9~0~~~1~0~0~~~11~0 "* Elution volume (ml) o Q.J '0 ~ , 0>14.8 .31

4.4

4.0

3,.61

• .. . ..

... .... ... ..... • PhosPhorYlase b .. BSA

• Ovalbumin

• Flavokinase ... Carbo Anhy. • Soya. Trp. Inh.

• Lysozyme .. Aprotinin · . Reg'easion line

(b)

... ... ..

• ' . ' .

•

...

~~------------~~~~----~--,~~~ 0,1 0.2 0 .. 3 0.4 0.5 0.6 0.7

Relative MobiJjty

Fig, 2 - Determination of molecular weight [(a): Plot of Log molecular weight versus elution volume. Flavokinase was subjected to gel filtration on a calibrated column of Sephacryl S300. Standards used: alcohol dehydrogenase (150 kDa), BSA (66 .2 kDa) and cytochrome c (12 kDa). (b): Flavokinase was subjected to electrophoresis on 12% SDS-PAGE and mobility compared to that of standard molecular weight markers: phosphorylase b (97.4 kDa), BSA (66 ,2 kDa), ovalbumin (45 kDa), carbonic an hydrase (31 kDa), soybean trypsin inhibitor (2 1,5 kDa), lysozyme (14,4 kDa) and aprotinin (6,5 kDa)]

The slime: enzyme was. activated by anions Pi, nitrate, carbonate and sulphate (Fig. 4b). Thus, these ions may have a potential role as physiological activators of phosphorylation of riboflavin. Pi acts as a mixed-type, non-essential activator for mung bean flavokinase 11

, where the Km for both substrates and the Vmax is enhanced in the presence of Pi . The activation of the slime enzyme also by Pi supports a role for Pi as a general physi910gical "feedback activator", where Pi, the product of hydrolysis of FMN, may stimulate the phosphOlylation of the other product, riboflavin, to its biologically utilizable form.

Kinetic constants The Km of the enzyme for riboflavin (Fig. 5) was

... e '·

4 , . '" ." ..

~

c 2, ·E .... . . Q.J Q. . .. .... .

"U • Q.J 1 ,

E 4, i5 6 .... 9: 8 10 .E z ~ :25

(b) •

, . 0

K:20 .. 15 ... . .' 10

5 .. ' ,

', . .., 01L-~"""""~--'-_--'-~---L~_"--~..J... -10 20 30 40 50 60 70 80

Assay Temperature (0C)

Fig. 3 - Effect of pH and temperature on flavokinase act ivity [(a): Effect of pH . Flavokinase (5 mU) was incubated in acetate buffer (PH 4 to 6) or Tris,H CI buffer (PH 7-1 0) for 10 min at 30·C and FMN formed quantitated by reverse phase HPLC. (b): Effect of temperature. Flavokinase ( 15 mU) was incubated in standard reaction mixture at th e given temperature for 10 min and FMN formed quantitated by reverse phase HPLC]

RAJESWARI el af.: PURIFICATION AND CHARACTERIZATION OF FLA VOKINASE FROM NEUROSPORA CRASSA 141

120 nM (std. elIor 0.02) and that for MgA TP2- (Fig. 6)

was 210 nM (std. error 0.05). This represents a very high affinity for the substrates. Flavokinase from other sources show a Km of around 4 J.1M for A TP and. 10 J.1M for riboflavin (with the exception of flavokinase from Bacillus subtilis4 and mung bean l 2

•13

which have a Km of 100 nM for riboflavin). The Hill plots showed a slope of 1.00 for riboflavin saturation and 0.96 for MgATp2

. saturation (data not shown).

100 (a)

80

60

40

20

Z' .s: O· '---L..L.L.L.J..---L..LL..LJ'---"'-L..£..."----'lL"'::.<-"--~.L..L-~~~CL...<1..--' 15 Co Zn Na Cu Cd « Cation added (1 mM)

Fig. 4 - Effect of cations and anions [(a): Effect of cations. Flavokinase (20 mU) was incubated in standard reaction mi xture containing I mM of sulfate salt of the cations at 30·C for 10 min and FM N formed quantitated by reverse phase HPLC. Activity in presence of Mg2+ is taken as I 00%. (b): Effect Qf anions. Flavokinase (20 mU) was incubated in standard reaction m ix ture contai ning 50 mM of sod ium salt of the anions at 30·C for 10 m in and FMN formed quantitated by reverse phase HPLC. Acti vity in absence of added anions is taken as 100%. Abbrev iations: Nit , nitrate; Crb, carbonate; Su i, sul fa te; Phs, orthophosphate; Ars, arsenate; Fi r, fl uoride; C it, citrate; Chi , ch loride; Ace, acetate]

The kcat for this enzyme is 20 pmqle FMN/min/llg enzyme (equivalent to approximately 3 pmole of this 35 .5 kDa protein), giving a turnover number of around 6 catalytic cycles per minute at room temperature (30°C).

Thus flavokinase from N crassa resembles that from Vigna radiata in its monomeric nature, molecular mass, cation preference, activation by Pi

c 4 ~ 'E .... a> • a. 3 • •

"0 • • a>

15f 41

E .... 2 .E • • 'Ii

Z • Ao "' 6" ~!; ~

;'0 .~ . u.. 1 •

0.5 0.0 / , , , ,

0 • '-;--10 ·0 10 20 ·30 40 E 11 [S) a. OL-~~--~--~--~--~--~~~~--~~

0.0 0.10.2 0.3 0.4 ~0.5 0.6 2.0 4.0 [Riboflavin] IlM

.Fig. 5 - Riboflavin saturation. [Flavokinase (5 mU) was incubated in reaction mixture containing 200 mM Tris. He l, p H 8.5, I mM MgS04, 40 11M ATP and varying concentrations of riboflavin (0.01-5 11M) at 30·C for 5 min and FMN fo rmed quantitated by reverse phase HPLC. v VS IS] plot. Inset Lineweaver Burk plot]

.... a> Q.

"0

~4 .... .E • z • ~ 2 • "0 • E • a.

•

0 0 .2

• • • •

0.8 6

" > ;::0.4 " .. '

~ .. ,

." 0.0

0 4 8 12 11 [S)

4 6 ·8 10

[MgA Tp21llM

Fig. 6 - ATP saturation. [Flavok inase (5 mU) was incubated in react ion mixture contai ning 200 mM T ris. HC I, pH 8.5 , I mM MgS04' 10 11M riboflav in and varying concent rations of A TP (0.0 I- I 0 11M) at 30·C for 8 min and FMN formed quantitated by reverse phase HPLC. y Vs [S] plot. Inset Lineweaver Burk plot)

142 INDIAN J BIOCHEM. BIOPHlYS, VOL. 36, JUNE 1999

and low Km for riboflavin. However, the slime flavokinase has a very low Km for MgATp2

- as well (no other flavokinase has been reported with a Km for MgA TP2

- below 4 J,LM), and has the highest sJX~cific activity reported so far. The relative abundance of this enzyme in slime and the simple, three step purification procedure developed for its isolation will permit molecular characterization of this enzyme.

Acknowledgement We thank Prof D B McCormick of Emory

University for the gift of a flavin analog used to prepare flavin Sepharose. This study was supported by a "Start up Grant" from Smith Kline & French Laboratory, PA, USA, and a grant (37/920/96/EMR­II) from CSIR, India, to SJ. A SRF award from UGC, India, to SRR and an award of RA to VSJ from the CSIR grant (37/920/96/EMR-II) is gratefully acknowledged.

References 1 McCormick D 8 (1962) J Bioi Chem 237, 959-962

2 Zak Z & McCormick D B (1982) Comp Biochem Physiol

738, 341-345

3 Kearney E B, Goldenberg J, Lipsick J & Perl M (1979) J BioI Chem 254, 9551-9557

4 Manstein D .r & Pai E F (1986) J Bioi Chem 261, 16169-16173

5 Giri K V, Krishnaswamy P R & Rao N A (1958) Bioohem J 70, 66-71

6 Sobhanaditya J & Rao N A (1981) Biochem J 197, 227-232

7 Nakagawa S, Igarashi A, Ohta T, Hagih'ara T, Fujio T & Aisaka K (1995) Biosci Biotech Biochem 59, 694-702

8 Yamada Y, Merrill A H JI' & McCormick D B (1990) Arch

Biochem Biophys 278, 125-130

9 Nakano H & McCormick D B (1992) Biochem Int 28, 441 -450

10 Sobhanaditya J & Rao N A (1981) Biochem fnt, 5, 567-573

II Das-Panja K, Jonnalagadda V S & Jonnalagadda S (1998) Biochem Mol Biollnt (In Press)

12 Das-Panja K (1997) Study of Vigna radiata Flavokinase: Substrate, saturation, effect of orthophosphate alld chemical modification , Ph.D thesis, Osmania University, Hyderabad

13 Das-Panja K, Jonnalagadda V S & Jonnalagadda S (1997)

Anal Biochem 252, 271-276

14 Bradford M B (1976) Anal Biochem 72, 248-252

15 Gartin D E (1990) Methods Enzymol182 , 425

16 Thomas E (1984) Protein Biochemistry Cold Spring Harbour

pp 231-283