Báo cáo Y học: Mycobacterium tuberculosis FprA, a novel bacterial NADPH-ferredoxin reductase docx

Mycobacterium tuberculosis FprA, a novel bacterialNADPH-ferredoxin reductase Federico Fischer, Debora Raimondi, Alessandro Aliverti and Giuliana Zanetti Dipartimento di Fisiologia e Bioc

Mycobacterium tuberculosis FprA, a novel bacterial

NADPH-ferredoxin reductase

Federico Fischer, Debora Raimondi, Alessandro Aliverti and Giuliana Zanetti

Dipartimento di Fisiologia e Biochimica Generali, Universita` degli Studi di Milano, Milano, Italy

The gene fprA of Mycobacterium tuberculosis, encoding a

putative protein with 40% identity to mammalian

adreno-doxin reductase, was expressed in Escherichia coli and the

protein purified to homogeneity The 50-kDa protein

monomer contained one tightly bound FAD, whose

fluor-escence was fully quenched FprA showed a low ferric

reductase activity, whereas it was very active as a NAD(P)H

diaphorase with dyes Kinetic parameters were determined

and the specificity constant (kcat/Km) for NAD PH was two

orders of magnitude larger than that of NADH Enzyme full

reduction, under anaerobiosis, could be achieved with a

stoichiometric amount of either dithionite or NADH, but

not with even large excess of NADPH In enzyme titration

with substoichiometric amounts of NADPH, only charge

transfer species (FAD-NADPH and FADH2-NADP+)

were formed At NADPH/FAD ratios higher than one, the

neutral FADsemiquinone accumulated, implying that the

semiquinone was stabilized by NADPH binding Stabiliza-tion of the one-electron reduced form of the enzyme may be instrumental for the physiological role of this mycobacterial flavoprotein By several approaches, FprA was shown to be able to interact productively with [2Fe)2S] iron-sulfur pro-teins, either adrenodoxin or plant ferredoxin More inter-estingly, kinetic parameters of the cytochrome c reductase reaction catalyzed by FprA in the presence of a 7Fe ferre-doxin purified from M smegmatis were determined A Km value of 30 nM and a specificity constant of 110 lM )1Æs)1 (10 times greater than that for the 2Fe ferredoxin) were determined for this ferredoxin The systematic name for FprA is therefore NADPH-ferredoxin oxidoreductase Keywords: flavoprotein; ferredoxin reductase; ferredoxin; Mycobacterium tuberculosis

Information available from the complete genome sequence

of Mycobacterium tuberculosis [1] has promoted a wide

investigation of new targets for drugs against tuberculosis

[2] The disease has regained ground in the developed world

due to the increased appearance of resistant strains of the

bacterium and the facile diffusion in the immunodepressed

people M tuberculosis is strongly dependent on iron

availability and on iron-containing cofactors for growth

and survival [3] It is well-known that iron availability in the

host plays a very important role in promoting the infection

by mycobacteria Interestingly, it has been reported that

Nramp1 (natural resistance-associated macrophage protein)

protein of mouse macrophages confers resistance to myco-bacterial infection in mice [4] Recently, a hyphothesis has been proposed based on the homology of Nramp1 to DCT1, a metal-ion transporter [5] Thus, the action of Nramp1 in the phagosomal membrane may be to deplete

Fe2+or other divalent cations from the phagosome, thus hampering the pathogen growth Among possible strategies

to effectively interfere with the pathogen metabolism, the blockage or limitation of Fe2+ availability inside the mycobacterium seems a promising target to pursue Redox systems called ferric reductases use intracellular redox cofactors to reduce the ferric Fe to the ferrous form for biosynthesis of iron-proteins A NAD(P)H:ferrimycobactin oxidoreductase activity was measured in M smegmatis cell extract [6] In Escherichia coli, enzymes of the ferredoxin-NADP+ reductase (FNR) protein family showing iron reductase activity, such as the flavin reductase, sulfite reductase and flavohemoglobin, have been implicated in such metabolism [7] Searches of the M tuberculosis genome for enzymes structurally related to the FNR family was unsuccessful but led to the identification of two genes, fprA and fprB, encoding putative adrenodoxin reductase-like proteins, expected to be functionally related to members

of the FNR family [8], i.e electron transferases that function

as a switch between two-electron and one-electron flow systems This class of enzymes is implicated in a variety of functions such as iron reduction, activation of ribonucleo-tide reductase, response to oxygen stress as well as reduction

of P450 cytochromes [8]

Here, we report on production and biochemical charac-terization of the recombinant FprA The homogeneous protein is shown to be a novel bacterial ferredoxin reductase

Correspondence to G Zanetti, Dipartimento di Fisiologia

e Biochimica Generali, Via Celoria 26, 20133 Milano, Italy.

Fax: + 39 02 50314895 Tel.: + 39 02 50314896,

E-mail: gzanetti@mailserver.unimi.it

Abbreviations: AdR, adrenodoxin reductase; Adx, adrenodoxin;

FNR, ferredoxin-NADP + reductase; Fd I, ferredoxin I; DPIP,

2,6-dichlorophenol-indophenol; SQ, semiquinone; CT,

charge-transfer complex.

Proteins: Bos taurus adrenodoxin, SWISS-PROT entry

ADX1_BOVIN; Spinacia oleracea ferredoxin I, SWISS-PROT entry

FER1_SPIOL; Mycobacterium smegmatis ferredoxin, SWISS-PROT

entry FER_MYCSM.

Enzymes: adrenodoxin reductase (EC 1.18.1.2), ferredoxin-NADP+

reductase (EC 1.18.1.2)

Note: a website is available at http://users.unimi.it/phybioch/

Index_htm

(Received 1 February 2002, revised 11 April 2002,

accepted 2 May 2002)

and to possess some properties similar to those of the bovine

adrenodoxin reductase [9]

M A T E R I A L S A N D M E T H O D S

Materials

All chemicals and pyridine nucleotides were purchased from

Sigma–Aldrich Chemical Co Cytochrome c (Sigma C2506)

was further purified by ion-exchange chromatography on

SP-Sepharose (Pharmacia Biotech.) Restriction

endonuc-leases, DNA polymerase and DNA modifying enzymes

were supplied by Amersham Pharmacia Biotech M

tuber-culosis cosmid MTCY164 was kindly provided by S T

Cole, Institut Pasteur, France pGEM-T and pET11a were

from Promega and Novagen, respectively Bovine Adx1was

a generous gift from F Bonomi, University of Milano,

Italy Recombinant spinach ferredoxin I (Fd I) was purified

as described previously [10] M smegmatis ferredoxin has

been purified by a modification of the procedure described

by Imai et al [11] DEAE-cellulose and Sepharose 4B steps

were replaced by chromatoraphy on HiLoad Q-Sepharose

Performance and HiLoad phenyl-Sepharose

High-Performance columns (Pharmacia Biotech) Ferredoxin was

eluted at about 0.7MNaCl from the first column using a

0–1M NaCl gradient in 50 mM Tris/HCl, pH 7.4 The

pooled fractions were brought to 2Mammonium sulphate

and loaded on the second column Elution was performed

with a 2–0 M ammonium sulphate gradient in the same

buffer as above Ferredoxin was desalted by dialysis against

50 mMTris/HCl, pH 7.4

PCR amplification and molecular cloning

The open reading frame of the M tuberculosis gene Rv3106,

named fprA, was amplified from the cosmid MTCY164

(GenBank accession no Z95150) by PCR using the

nucleotides 5¢-GCCATATGATGCGTCCCTATTACA-3¢

and 5¢-GTCATATGTCAGCCGAGCCCAAT-3¢, which

contained the NdeI restriction site (underlined) The

result-ing DNA fragment was cloned into pGEM-T vector and

sequenced The NdeI DNA fragment from the recombinant

plasmid containing fprA was recloned in the NdeI site of the

expression vector pET-11a, yielding pETfprA

Overexpression of fprA

E coli BL21(DE3) cells transformed with pETfprA were

grown in flasks under vigorous shaking at various

temper-atures in 2· YT medium supplemented with 100 mgÆL)1

ampicillin For enzyme purification, E coli cells were grown

in a New Brunswick 12 L fermentor at 25C to midlog

phase (D600¼ 1.2–1.5) The culture, after cooling to 15 C,

was induced with 0.1 mM isopropyl thio-b-D-galactoside

Cells were harvested after 15–17 h

Purification of FprA

All purification steps were performed at 4C except for

FPLC, which was carried out at room temperature E coli

cell paste were resuspended in 2 mLÆg)1 of buffer A

(50 mM Na-phosphate, pH 7.0, containing 1 mM EDTA

and 1 mM 2-mercaptoethanol) supplemented with 1 mM

phenylmethanesulfonyl fluoride and disrupted by sonica-tion After removal of cell debris by centrifugation at

43 000 g for 1 h, the protein concentration of the crude extract was adjusted to 25 mgÆmL)1 The solution was then brought to 40% saturation of ammonium sulphate (1.64M), the precipitate discarded and the soluble fraction loaded on Sepharose 4B column (Pharmacia Biotech) pre-equilibrated with 1.64Mammonium sulphate in buffer A FprA was eluted with the same solution as a single peak well separated from the material eluting in the void volume To the pooled FprA-containing fractions glycerol was added to 10% final concentration and the enzyme was precipitated with 85% saturation ammonium sulphate The pellet was resuspended and dialysed against 25 mM imidazole-HCl, pH 7.0, containing 10% glycerol and

1 mM 2-mercaptoethanol The enzyme was loaded on a HiLoad Q-Sepharose High-Performance column (Pharma-cia Biotech) and eluted with a linear gradient from 150 to

250 mM NaCl The purified FprA was desalted by gel-filtration on PD10 column (Pharmacia Biotech) using

50 mMHepes/KOH, pH 7.0, containing 10% glycerol and

1 mMDTT The enzyme stored at)80 C retained its full activity for more than 1 year

Molecular characterization methods SDS/PAGE was carried out on 10% polyacrylamide gels Microsequencing was performed on an Applied Biosystems 477/A protein sequencer equipped with an on-line HPLC system Analytical gel-filtration analyses were performed on

a HPLC apparatus (Waters) equipped with either Superdex

75 or Superose 12 columns (Pharmacia Biotech) in 50 mM Hepes/KOH, pH 7.0, containing 0.15Mammonium acetate and 2 mM2-mercaptoethanol FprA and ferredoxin (10 and

40 lM, respectively) were cross-linked by treatment with

5 mM N-ethyl-3-(3-dimethylaminopropyl)carbodiimide in

25 mMNa-phosphate, pH 7.0 [12]

Spectral analyses Absorption spectra were recorded with a Hewlett-Packard

8453 diode-array spectrophotometer The extinction coeffi-cient of the protein-bound flavin was determined spectro-photometrically quantitating the FADreleased from the apoprotein following SDS treatment [13] Fluorescence measurements were performed on a Jasco FP-777 spectro-fluorometer at 15C The identity of the enzyme bound flavin was assessed fluorimetrically by treating with phos-phodiesterase the flavin released after thermal denaturation

at 100C of the holoenzyme [13]

Activity assays Enzyme catalyzed reactions were monitored continuously

on a Hewlett-Packard 8453 diode-array spectrophotometer Ferric reductase activity was assayed in both aerobic and anaerobic conditions in 50 mMTris/HCl, pH 7.5 at 25C

as described previously [14] Diaphorase activity was measured in 0.1M Tris/HCl, pH 8.2 at 25C with either

K3Fe(CN)6or DPIP as electron acceptor and NADPH or NADH as reductants Cytochrome c reductase activity was assayed in the same buffer as above with either 5 lM spinach Fd I, bovine Adx or M smegmatis ferredoxin,

using 50 lMcytochrome c as the terminal electron acceptor.

Unless otherwise stated, the NADPH concentration

was kept constant by regeneration with 2.5 mM glucose

6-phosphate and 2 lgÆmL)1glucose 6-phosphate

dehydro-genase Steady-state kinetic parameters for the diaphorase

activities and for the cytochrome c reductase activity with

mycobacterial ferredoxin were determined by varying the

concentrations of the substrates Double-reciprocal plots of

the data yielded parallel lines Initial rate data (v) were fitted

by nonlinear regression using GRAFIT 4.0 (Erythacus

Software Ltd, Staines, UK) to a ping-pong Bi-Bi

mechan-ism equation (Eqn 1):

v¼ V  A  B=ðKa B þ Kb A þ A  BÞ ð1Þ

where A and B, and Kaand Kbare the molar concentrations

and the Michaelis constants for the two substrates,

respect-ively

Enzyme titrations and photoreductions

Titrations of oxidized FprA with NADP+, NAD+, or

spinach Fd I were performed spectrophotometrically at

15C The enzyme was diluted to a final concentration of

12–15 lMin 10 mM Tris/HCl, pH 7.7 NADP+titrations

were carried out at different ionic strength by varying the

NaCl concentration between 0 and 150 mM D ifference

spectra were computed by subtracting from each spectrum

that obtained in the absence of ligand, after correction for

dilution Kd values were obtained by fitting data sets by

nonlinear regression to the theoretical Eqn (2) for a 1 : 1

binding, using the softwareGRAFIT4.0 (Erythacus Software

Ltd, Staines, UK)

DA

¼ De Lþ P þ Kd

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Lþ P þ Kd

ð Þ2 4  L  P q

2

ð2Þ

DA is the value of the difference spectrum at a selected

wavelength; De is the difference extinction coefficient at that

wavelength of the protein-ligand complex; L is the total

molar concentration of added ligand; P is the total molar

concentration of FprA

All reduction experiments were carried out in anaerobic

cuvettes at 15C Solutions were made anaerobic by

successive cycles of equilibration with O2-free nitrogen

and evacuation Reductive titrations with Na-dithionite,

NADPH, or NADH were carried out using 15–50 lMFprA

solutions in 10 mMTris/HCl, at pH 7.4 Photoreductions of

FprA using EDTA/light [15] were performed in 10 mM

Hepes/KOH, at pH 7.0, containing 15 mM EDTA and

1.8 lM 5-deazariboflavin NADP+ titration of reduced

enzyme was carried out by additions of an anaerobic

solution of NADP+to FprA previously photoreduced as

described above The amount of FADSQ was calculated by

subtraction of the contribution of the CT species [16,17]

from the absorbance at 625 nm according to Eqn (3):

Asq¼ A625 ð2:79  A750Þ ð3Þ

The contribution of CT species to A625can be estimated

taking into account that SQ does not absorb at 750 nm [18]

and that a A625/A750value of 2.79 for CT species could be determined from experiments in which no SQ was formed

R E S U L T S

Identification of fprA and fprB The search of M tuberculosis genome [1] for enzymes potentially involved in iron metabolism led to the identifi-cation of two genes, fprA and fprB, whose predicted protein products are related to each other They share a domain with significant similarity ( 40% identity) with mamma-lian AdR (Table 1) FprB contains a C-terminal domain homologous to FprA (42% identity) plus an N-terminal moiety comprising an iron-sulfur binding region signature typical of bacterial 7Fe ferredoxins It is remarkable that AdR homologs are present in very few bacteria, whereas two such proteins are found in mycobacteria (Table 1) To our knowledge, the fusion protein does not have a counter-part in other organisms, except for other mycobacteria Production of FprA

We tried to heterologously express both cloned genes, yet we were only successful in obtaining FprA in a soluble active form In a preliminary series of experiments, E coli BL21(DE3) strain harboring pETfprA was grown at

37C Upon induction, a novel protein band of 50-kDa was clearly visible in SDS/PAGE, but most of the protein was present in insoluble form Growth and induction conditions were varied to optimize the production of the recombinant protein in a soluble form (not shown) The amount of the soluble recombinant protein increased greatly

by lowering the growth temperature Concomitantly, the NADPH-ferricyanide reductase specific activity of the soluble cell extracts also increased, being highest in cells grown at 15C and harvested about 16 h after induction The purification of FprA was achieved by a three-step procedure as described in Materials and methods An ammonium sulphate fractionation coupled to a salt-pro-moted adsorption chromatography on Sepharose 4B, and followed by an anion-exchange chromatography on Table 1 Sequence comparison matrix stating percentage of identical residues of bacterial and selected eukaryotic adrenodoxin reductase-like proteins Proteins with sequence identity lower than 20% were omitted Individual proteins are: A, FprA (M tuberculosis); B, FprA (M lep-rae); C, FprB C-terminal domain (M tuberculosis); D, FprB C-term-inal domain (M leprae); E, probable ferredoxin reductase (Deinococcus radiodurans); F, putative ferredoxin reductase (Strepto-myces coelicolor); G, Arh1p (Saccharo(Strepto-myces cerevisiae); H, bovine adrenodoxin reductase.

Q-Sepharose, yielded about 2 mg of FprA per gram of cells,

with an overall yield of 25% and a purification factor of 18

SDS/PAGE of the various fractions of the purification is

shown in Fig 1

FprA is a flavoprotein

The visible absorption spectrum of the purified protein is

presented in Fig 2 The absorbance in the visible region is

that typical of a flavoprotein with bands centered at 381 and

452 nm and shoulders at 422 and 473 nm Maximal

absorbance in the ultraviolet region was at 272 nm A value

of 7.0 for the A272/A452 ratio was calculated from the

spectrum Flavin fluorescence was almost completely

quenched The non covalently bound flavin in FprA was

shown to be FAD The flavin fluorescence of the released

cofactor increased about 10-fold after phosphodiesterase

treatment, as expected for the conversion from FADto

FMN The extinction coefficient of the enzyme at 452 nm

was calculated to be 10 600M )1Æcm)1from the amount of

FADreleased after protein denaturation by SDS A

stoichiometry of 0.98 mol FADper mol of 50 kDa

monomer was established The flavin was reducible by

dithionite (Fig 2) and an anaerobic titration of FprA with

this reductant showed that 0.92 molÆmol FAD)1or about

two electrons per flavin were required for full reduction (see

inset of Fig 2) This excludes the presence of additional

redox cofactors in the enzyme No changes in absorbance

beyond 550 nm were observed, indicating that the flavin

semiquinone (SQ) did not accumulate [18] Thus, only two

forms of the FADprosthetic group were present during titration, the oxidized form and the fully reduced one, as can

be deduced from the presence of an isosbestic point at

340 nm The same pattern of reduction was obtained by photoreduction [15] FprA was rapidly reduced by succes-sive periods of irradiation in the presence of 5-deazaribo-flavin and EDTA, yielding the hydroquinone spectrum after

8 min of light exposure (data not shown) Reoxidation of fully reduced enzyme by molecular oxygen occurred without any detectable SQ formation Thus, the one-electron reduced form of FADis not stabilized in the enzyme Molecular properties

FprA showed a Mrof about 50 000 in denaturing PAGE (Fig 1) This value is in good agreement with that of 49 341 calculated from the sequence The identity of the overpro-duced protein was assessed by N-terminal analysis The first

21 amino-acid residues of the purified protein were identical

to those deduced from the gene sequence: MRPYYIAIVG SGPSAFFAAAS The Mr of the recombinant FprA in solution was determined in several conditions Gel filtration experiments in FPLC, either on Superose 12 or Superdex

75, allowed the determination of a value of 53 ± 5 kDa, when the protein was maintained in 10% glycerol and 1 mM dithiothreitol, indicating that under these conditions the protein is a monomer The addition of glycerol and 2-mercaptoethanol were required to avoid formation of aggregates

Catalytic properties The ferric reductase activity of the purified protein was investigated by using Fe3+-EDTA in the presence of the

Fe2+-chelator ferrozine [14] The activity was very low both

in the presence and absence of oxygen and/or FAD: 0.5–1 (mol NADPH)Æmin)1Æ(mol FAD))1 Furthermore, addition of 1 l 7Fe ferredoxin from M smegmatis (see

Fig 1 Purification of recombinant FprA as analysed by SDS/PAGE.

Lanes 1 and 5, molecular mass markers (values in kDa are indicated);

lane 2, crude extract; lane 3, after Sepharose 4B; lane 4, after

Q-Sepharose.

Fig 2 Electronic absorption spectrum of purified FprA and dithionite titration The enzyme was 26 l M in 10 m M Tris/HCl, pH 7.4, con-taining 10% glycerol and 1 m M dithiothreitol FprA was stepwise reduced with dithionite under anaerobiosis The spectra recorded at 0, 0.2, 0.4, 0.5, 0.7, 0.9, and 1 reductant/FADmolar ratios are reported The inset shows the plot of the fractional absorbance change at 452 nm

as a function of dithionite/FADmolar ratio A i and A f are the initial and final values of absorbance at 452 nm, respectively.

below) to the assay did not increase the iron reduction rate.

On the other hand, the protein was found able to catalyze

electron transfer from NADPH as well as NADH to

artificial electron acceptors like ferricyanide and DPIP The

steady-state kinetic parameters for the ferricyanide and

DPIP activities were determined at pH 8.2 (Table 2) The

double-reciprocal plots of initial velocities obtained by

varying the reduced pyridine nucleotide at various fixed

levels of the artificial dye showed a pattern of parallel lines

Data were fitted to Eqn (1) For the K3Fe(CN)6reductase

activity, the experiments revealed that ferricyanide

concen-trations above 1 mM were inhibitory A 100- to 150-fold

lower Kmvalues for NADPH with respect to NADH were

observed in the diaphorase reactions, whereas similar values

of kcat were obtained with both coenzymes, thus the

specificity constant ratio NADPH/NADH was 225 in the

ferricyanide reaction and 116 in the DPIP one The catalytic

efficiencies of FprA with respect to the acceptors differed by

10-fold with preference for the one-electron reducible

substrate, i.e ferricyanide To study the interaction with

pyridine nucleotides in details, FprA was titrated with both

NADP+and NAD+ In both cases, the visible spectrum of

the enzyme was perturbed The difference spectra elicited by

the ligand binding are shown in Fig 3A The features of the

difference spectra produced by NADP+or NAD+are very

similar, but the intensity of the 500 nm peak was fourfold

higher in the case of NADP+(Fig 3A) Titrations with this

coenzyme were performed at increasing ionic strength to

obtain an accurate estimate of the Kdby extrapolation of the

linear part in the graph of log Kdvs. I Thus, Kdvalues of

FprA for NADP+ of 6 nM at I¼ 0, and 0.4 lM at

I¼ 50 mMwere calculated (data not shown) In contrast,

the Kdvalue for NAD+was in the millimolar range

Identification of a physiological electron acceptor

The physiological activity of the mammalian homolog of

FprA is to reduce the [2Fe)2S] iron–sulfur protein Adx

[9,19,20] Nevertheless, there are no genes coding for

[2Fe)2S] ferredoxins in the M tuberculosis genome [1] At

first, we studied the interaction of the recombinant enzyme

with the bovine Adx and with another [2Fe)2S] protein, the

spinach leaf Fd I Cytochrome c was used as final electron

acceptor in these reactions Its reduction was observed only

when either Adx or Fd I was added in the assay, indicating

that FprA was able to interact productively with both these

electron carrier proteins FprA was 10-fold more active with

the plant type Fd I than with Adx under the same

conditions In the mean time, we cloned M tuberculosis

genes coding for 7Fe and 3Fe ferredoxins, but failed in

obtaining the overexpression in E coli Several years ago, a

7Fe ferredoxin was purified from M smegmatis [11] By using a similar procedure, we obtained a reasonable amount

of the M smegmatis 7Fe ferredoxin in homogeneous form

as judged by several criteria (native and denaturing PAGE, protein determination/molarity determined by using the reported extinction coefficient at 406 nm) N-Terminal analysis of the purified protein confirmed its identity with

Fig 3 Spectral perturbations elicited by ligand binding to FprA All measurements were performed in 10 m M Tris/HCl, pH 7.7 with 15 l M

enzyme Difference spectra were computed by subtracting from spectra recorded at titration end-points those of unbound FprA and ligand (A) difference spectra of the complexes between FprA and NADP+ (solid line) or NAD+(dashed line) (B) difference spectrum of the complex between FprA and Fd I.

Table 2 Kinetic parameters for the ferricyanide and DPIP reductase reactions of FprA.

Electron

acceptor k cat (e – Æs)1)

KNADðPÞHm (l M )

k cat /K m

(e – Æs)1Æl M )1 )

K acceptor m

(lM)

k cat /K m

(e – Æs)1Æl M )1 ) NADPH

NADH

the ferredoxin isolated by Imai et al [11] This ferredoxin

has 88% identity with FdxC of M tuberculosis [1] The

steady-state kinetic parameters for both the 2Fe and 7Fe

ferredoxin reductase activities are reported in Table 3 The

kinetic data obtained with the protein substrates yielded

parallel lines in double-reciprocal plots and were fitted to

Eqn (1) The kcat measured with the 7Fe ferredoxin was

30% of that with the spinach protein, whereas the Kmfor

the homologous protein substrate was about 30-fold lower,

suggesting a much higher affinity of FprA for the

mycobacterial ferredoxin Due to the ability to reduce

iron-sulfur proteins using preferentially the pyridine

nuc-leotide phosphate, the systematic name for FprA is thus

NADPH-ferredoxin oxidoreductase or NFR Although

Fd I is not the physiological substrate of the bacterial

reductase, a titration of FprA with Fd I was attempted to

demonstrate that an interaction between the two proteins

was indeed occurring thus supporting the activity data

Figure 3B shows the difference spectrum obtained at

saturating concentration of Fd I Two positive peaks

appeared centered around 450 and 380 nm, where FprA

has absorption maxima An approximate Kdvalue of 2 lM

was obtained by titration The interaction between the two

proteins was further investigated by using cross-linking

agents Following incubation of the two proteins with

N-ethyl-3-(3-dimethylaminopropyl)carbodiimide, FprA was

fully converted to protein adducts of about 66 kDa as

determined by SDS/PAGE This is the expected value

for a 1 : 1 cross-linked complex between the

flavopro-tein and Fd I [12] The cross-linked species acquired the

capacity to reduce directly cytochrome c as judged by

measuring the cytochrome c reductase activity in the

absence of added Fd I The same type of experiments were

repeated replacing the spinach protein with the 7Fe

ferredoxin A cross-linked protein of about 66 kDa was

also obtained, although at a lower rate of formation with

respect to the plant ferredoxin

Anaerobic reduction of FprA with NAD(P)H

Bovine AdR shows peculiar behavior when anaerobically

reduced by NADPH [21] We therefore tried to verify

whether FprA presented the same reduction pattern when

treated with physiological reductants Identification of

reduced intermediates could help in elucidating the

mech-anism of action of FprA The titration of FprA with the less

efficient substrate NADH practically superimposed to that

with dithionite (Fig 2) About 1 (mol NADH)Æ(mol

FAD))1 was sufficient to fully reduce the enzyme, again

without significant changes at wavelengths longer than

550 nm (data not shown) The full reduction of the enzyme

FADby just an equimolar amount of NADH implies that

the enzyme redox potential is far more positive than that of

the pyridine nucleotide couple A completely different

pattern was observed when FprA was titrated with NADPH (Fig 4A) Clearly, upon reduction with substoi-chiometric amounts of the reduced coenzyme, absorption in the 500–800 nm region built up with a broad peak at

550 nm These spectral changes are usually ascribed to formation of charge-transfer (CT) species (FAD-NADPH and/or FADH2-NADP+) [16,17] Only two species are present during titration, as indicated by the presence of two isosbestic points (373 and 490 nm, respectively) After addition of more than 1 mol NADPH per mol FAD (Fig 4B), the spectra in the long wavelength region changed A peak at 580 nm with a shoulder at 625 nm developed This type of spectrum (peaks in the 600 nm region with no absorption beyond 700 nm) can be attrib-uted to the flavin neutral SQ [18] In an attempt to further characterize the various species formed during NADPH titration of FprA, a titration with NADP+ of the fully

Fig 4 NADPH reduction of FprA The titration was performed in

10 m M Tris/HCl, pH 7.4 under anaerobiosis 47 l M FprA was titrated with NADPH The spectra recorded at 0, 0.3, 0.45, 0.6, 0.7, 0.9, 1 (A) and at 1.3, 1.6, 1.9, 3, 6 (B) NADPH/FAD molar ratios are reported The inset shows the plot of the absorbance at 625 nm due to SQ, obtained by subtracting the contribution of charge-transfer species as detailed in Materials and methods, as a function of NADPH/FAD molar ratio.

Table 3 Kinetic parameters for the 2Fe and 7Fe ferredoxin reductase reactions of FprA.

Electron acceptor

k cat

(e – Æs)1)

K NADPH m

(l M )

k cat /K m

(e – Æs)1Æl M )1 )

K acceptor m

(l M )

k cat /K m

(e – Æs)1Æl M )1 )

M smegmatis

ferredoxin

reduced enzyme obtained by photoreduction was performed

(Fig 5A) The spectra resemble those already observed

during the early steps of NADPH titration of oxidized

enzyme (Fig 4A) It can be noted that both the absorbance

at 450 and 340 nm of the solution increased at each addition

of NADP+up to 1 NADP+per FAD(see inset) and no SQ

was formed Thus, the spectrum of the CT formed in this

experiment, which is superimposable to that formed in the

titration of the oxidized enzyme with a molar amount of

NADPH, is mostly due to FAD-NADPH charge transfer,

as can be judged from the high absorbance at 340 and

450 nm, and low absorbance at 750 nm The SQ amount

present during NADPH titration could then be calculated

by subtracting from the spectra the contribution of the CT

species as obtained from the experiment shown in Fig 5A

In the inset of Fig 4B, the absorption changes due to SQ

accumulation are plotted against the NADPH/flavin molar

ratio It can be observed that the SQ built up only after one

NADPH/flavin was added, reached its maximum after

addition of slightly more than two NADPH/flavin, and

then remained at this level notwithstanding the high amount

of NADPH added Indeed, full reduction of the bound

flavin to FADdihydroquinone was not achieved even by prolonged incubation or by using NADPH in the presence of

a NADPH regenerating system This suggests that the SQ is stabilized by complexation with NADPH This is further confirmed by photoreduction experiments carried out in the presence of 1.5 (mol NADP+)Æ(mol FAD))1 (Fig 5B) Whereas photoreduction of uncomplexed FprA did not elicite accumulation of reduced intermediates, in the presence

of NADP+the formation of a long wavelength band with a peak at 550 nm (ascribable to CT species) in the early steps of reduction was observed With further irradiation, the spectral features typical of the SQ appeared This indicates that the SQ accumulated only after NADPH was formed, thus suggesting that this intermediate is a complex between flavin SQ and NADPH These data can be rationalized according to the scheme presented below:

Eoxþ NADPH $ CT

CTþ NADPH $ Ered-NADPHþ NADPþ

CTþ Ered-NADPH$ 2Esq-NADPH where CT indicate an equilibrium mixture of the two charge-transfer species FAD-NADPH and FADH2 -NADP+

D I S C U S S I O N

The functional annotation of proteins identified in genome sequencing projects is based on protein sequence similarities

to homologs in the databases However, due to the possibility of divergent evolution, homologous enzymes may not catalyze the same reaction Thus, a biochemical characterization of the gene product is required to establish the protein’s real function in that organism This was particularly necessary in the case of the fprA gene product of

M tuberculosis, because of the absence in the bacterial genome of genes coding for [2Fe)2S] ferredoxins, the expected protein substrate for an adrenodoxin reductase-like enzyme To our knowledge, this is the first adrenodoxin reductase-like protein from a bacterium to be characterized The recombinant enzyme was shown to be a flavoprotein containing noncovalently bound FAD, whose fluorescence was nearly fully quenched This is a remarkable difference from the mammalian enzyme, the flavin of which is fluorescent [22,23] The fprA gene product did not show significant activity as ferric reductase as was at first hypothesized Instead, it possesses the activities typical of the mammalian AdR [9,24], including the capacity to reduce the mammalian Adx However, FprA was more efficient with plant Fd I and more interestingly, with a 7Fe ferredoxin of M smegmatis This ferredoxin is a homolog

of M tuberculosis FdxC (88% identity between the sequences) The higher affinity of the reductase for the 7Fe ferredoxin is in keeping with the absence of 2Fe ferredoxins

in mycobacteria The elucidation of the three-dimensional structure of the enzyme will provide more information on the structural basis for the specificity in protein–protein recognition The enzyme can use both NADPH and NADH

as a reductant; however, the specificity constant (kcat/Km) of NADPH is two orders of magnitude larger than that of NADH Furthermore, binding of NADP+ to FprA is extremely tight with Kdvalues in the nanomolar region The affinity of FprA for NADP+is at least 10 times higher than

Fig 5 Effect of NADP+addition after or before FprA photoreduction.

Photoreduction of FprA was performed in 10 m M Hepes-KOH,

pH 7.0, in the presence of 15 m M EDTA and 1.8 l M

5-deazaribofla-vin (A) NADP + titration of 26.5 l M photoreduced FprA The spectra

recorded at 0, 0.1, 0.3, 0.4, 0.5, 0.7, 0.8, 1 NADP + /FADmolar ratios

are reported The inset shows the absorbance changes at 452 (s), 550

(d) and 750 nm (h) as a function of NADP + /FADmolar ratio The

absorbance change at 750 nm has been multiplied by four for clarity.

(B) photoreduction of 20 l M FprA in the presence of 30 l M NADP+.

The spectra recorded before and after 1.5, 2.5, 3.5 min irradiation

(dashed line) and after 6, 10, 13, 17 min irradiation (solid lines) are

shown The inset shows an enlargement of the spectral data in the 500–

750 nm region.

that of bovine AdR [9] The tight binding of NADP(H) may

have physiological implications Anaerobic titrations with

NADPH of FprA revealed a completely different pattern

from that obtained with dithionite, NADH or

photoreduc-tion In the latter cases, only two forms of the enzyme, the

oxidized and the fully reduced ones, were observed With

NADPH or NADP+present during reduction of FprA,

two additional forms were identified: CT species

(FAD-NADPH and FADH2-NADP+) and FADsemiquinone

Unlike bovine AdR, FprA highly favored the CT species

FAD-NADPH, as judged by comparison of the spectra [21]

By analysis of the conditions in which the SQ accumulated,

it can be inferred that this intermediate results from

NADPH binding to the flavin SQ, as observed in the case

of bovine AdR [21] This complex is assumed to be a

compulsory intermediate in the catalytic cycle of these

enzymes, whose functional role is to mediate electron

transfer between two-electron donors (NADPH) and

one-electron acceptors (iron-sulfur protein substrates) [9]

This enzyme must be of relevance to mycobacteria

because a homolog is present in M leprae, whose genome is

greatly downsized and degraded [25] On the basis of the

high similarity of FprA with mammalian AdR (Table 1), its

enzymatic function may be inferred In mitochondria, AdR,

with a [2Fe)2S] ferredoxin, is part of an electron chain

which delivers electrons from NADPH to cytochrome P450

enzymes, mainly involved in hydroxylation reactions

[9,19,20] The M tuberculosis genome is rich in genes

encoding P450 cytochromes (22 genes, see [1]), whereas it

lacks genes coding for Adx-type ferredoxins and it contains

only genes encoding 7Fe and 3Fe ferredoxins [1] In

bacteria, different systems for P450 cytochrome reduction

are employed Well known is the system comprising

putidaredoxin reductase, a NADH-dependent flavoprotein,

and putidaredoxin (2Fe ferredoxin), which transfers

elec-trons to P450cam [26] This system is similar to the

mammalian AdR-Adx A microsomal-type P450 reductase

instead is present in Bacillus megaterium [27] Apparently,

purification of the reductase from other bacteria was

unsuccessful due to protein instability and low expression

level A microbial cytochrome P450 reduction system was

purified from Streptomyces griseus grown in a soybean

flour-enriched medium [28] The ferredoxin reductase was

a NADH-dependent flavoprotein of 60 kDa with a

N-terminal sequence comprising a FADbinding consensus

sequence (GXGXXG), which is typical of the glutathione

reductase large family [29], to which AdR also belongs

They showed that this enzyme can couple electron transfer

from NADH to cytochrome P450soy in the presence of

S griseus7Fe ferredoxin The activity value measured in the

cytochrome c assay is in agreement to that obtained with

FprA and M smegmatis ferredoxin The low Kmvalue of

FprA for this iron-sulfur protein strengthens the hypothesis

that a 7Fe ferredoxin could be the physiological partner of

the enzyme Nevertheless, in herbicide-induced S griseolus

cells [30], two small 3Fe ferredoxins were found highly

expressed, which could reconstitute an in vitro electron chain

to P450 cytochromes using spinach FNR Recently, a

cytochrome P450 and a 3Fe ferredoxin were purified from

Mycobacteriumsp strain HE5, grown on morpholine [31]

In both these cases, it was hypothesized that the reductase is

constitutively formed and it has a broad specificity with

respect to the ferredoxin substrate

Further roles for AdR have been discovered The AdR-Adx system of the lower eukaryote Saccharomyces cerevisiaewas shown to be essential for yeast viability by gene knockout [32–34] and to be involved in the biosynthe-sis of the cell iron-sulfur clusters [35–37] Furthermore, mammalian AdR has been recently identified to play a role

in the p53-dependent apoptosis, due to its potential to produce reactive oxygen species (ROS) [38] Accordingly, it can be assumed that the mycobacterial FprA may have similar functions in iron-sulfur cluster synthesis or oxidative stress response It is likely that FprA is primarily involved in the reduction of P450 enzymes as is the case of the other bacerial reductases cited above Recently, the P450 14a-demethylase of M tuberculosis has been characterized and suggested to be involved in the cholesterol biosynthetic pathway [39] Cholesterol has been shown to be essential to

M tuberculosis infection [40] Furthermore, some of the cytochrome P450 enzymes could be involved in the synthesis

of the complex cell wall components Thus, if FprA provides electrons to several pathways through the inter-action with several ferredoxins, it represents a potential target for antimycobacterial drugs Crystals of FprA have been obtained and the three-dimensional structure is being currently determined

A C K N O W L E D G E M E N T S

This work was carried out with funds from the Ministero dell’Univer-sita` e della Ricerca Scientifica e Tecnologica (Prin 1999) and European Union (EU Cluster QLK2-2000–01761) We thank Dr G Riccardi (University of Genova), Dr R Cantoni and Dr M Branzoni (University of Pavia) for help in cloning and DNA sequencing,

Dr A Negri and Dr G Tedeschi for protein microsequencing, and

Dr M A Vanoni and Dr B Curti for helpful discussions.

R E F E R E N C E S

1 Cole, S.T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris, D., Gordon, S.V., Eiglmeier, K., Gas, S., Barry III, C.E.

et al (1998) Deciphering the biology of Mycobacterium tuber-culosis from the complete genome sequence Nature 393, 537–544.

2 McKinney, J.D (2000) In vivo veritas: the search for TB drug targets goes live Nat Med 6, 1330–1333.

3 De Voss, J.J., Rutter, K., Schroeder, B.G & Barry, C.E III (1999) Iron acquisition and metabolism by mycobacteria J Bacteriol.

181, 4443–4451.

4 Vidal, S.M., Malo, D., Vogan, K., Skamene, E & Gros, P (1993) Natural resistance to infection with intracellular parasites: isola-tion of a candidate for Bcg Cell 73, 469–485.

5 Gunshin, H., Mackenzie, B., Berger, U.V., Gunshin, Y., Romero, M.F., Boron, W.F., Nussberger, S., Gollan, J.L & Hediger, M.A (1997) Cloning and characterization of a mammalian proton-coupled metal-ion transporter Nature 388, 482–488.

6 Brown, K.A & Ratledge C (1975) Iron transport in Myco-bacterium smegmatis:ferrimycobactin reductase (NAD(P)H:ferri-mycobactin oxidoreductase), the enzyme releasing iron from its carrier FEBS Lett 53, 262–266.

7 Fontecave, M., Coves, J & Pierre, J.L (1994) Ferric reductases or flavin reductases? Biometals 7, 3–8.

8 Karplus, P.A & Bruns, C.M (1994) Structure–function relations for ferredoxin reductase J Bioenerg Biomembr 26, 89–99.

9 Nonaka, Y., Miura, R & Yamano, T (1991) NADPH-adreno-doxin oxidoreductase In Chemistry and Biochemistry of Flavo-enzymes, Vol 2 (Mu¨ller, F., ed.), pp 329–341 CRC Press, Boca Raton, FL.

10 Piubelli, L., Aliverti, A., Bellintani, F & Zanetti, G (1995)

Spi-nach ferredoxin I: overproduction in Escherichia coli and

puri-fication Prot E xp Purif 6, 298–304.

11 Imai, T., Matsumoto, T., Ohta, S., Ohmori, D , Suzuki, K.,

Tanaka, J., Tsukioka, M & Tobari, J (1983) Isolation and

characterization of a ferredoxin from Mycobacterium smegmatis

Takeo Biochim Biophys Acta 743, 91–97.

12 Zanetti, G., Curti, B & Aliverti, A (1984) A cross-linked complex

between ferredoxin and ferredoxin-NADP+reductase J Biol.

Chem 259, 6153–6157.

13 Aliverti, A., Curti, B & Vanoni, M.A (1999) Identifying and

quantitating FADand FMN in simple and iron-sulfur-containing

flavoproteins In Methods in Molecular Biology, Vol 131

Flavo-protein Protocols (Chapman, S.K & Reid, G.A., eds), pp 9–23.

Humana Press Inc., Totowa, NJ.

14 Coves, J & Fontecave, M (1993) Reduction and mobilization of

iron by a NAD(P)H: flavin oxidoreductase from Escherichia coli.

E ur J Biochem 211, 635–641.

15 Massey, V & Hemmerich, P (1977) A photochemical procedure

for reduction of oxidation-reduction proteins employing

deazari-boflavin as catalyst J Biol Chem 252, 5612–5614.

16 Massey, V & Palmer, G (1962) Charge-transfer complexes

of lipoyl dehydrogenase and free flavins J Biol Chem 237,

2347–2358.

17 Sakuray, T & Hosoya, H (1966) Charge-transfer complexes of

nicotinamide-adenine dinucleotide analogues and flavin

mono-nucleotide Biochim Biophys Acta 112, 459–468.

18 Massey, V & Hemmerich, P (1980) Active-site probes of

flavo-proteins Biochem Soc Trans 8, 246–257.

19 Hanukoglu, I (1992) Steroidogenic enzymes: structure, function,

and role in regulation of steroid hormone biosynthesis J Steroid

Biochem Mol Biol 43, 779–804.

20 Bernhardt, R (1996) Cytochrome P450: structure, function, and

generation of reactive oxygen species Rev Physiol Biochem.

Pharmacol 127, 137–221.

21 Nonaka, Y., Fujii, S & Yamano, T (1986) The semiquinone state

of NADPH-adrenodoxin oxidoreductase in the course of

anae-robic reduction with NADPH J Biochem 99, 803–814.

22 Chu, J & Kimura, T (1973) Studies on adrenal steroid

hydro-xylases Complex formation of the hydroxylase components.

J Biol Chem 248, 2089–2094.

23 Foster, R.P & Wilson, L.D (1975) Purification and

character-ization of adrenodoxin reductase from bovine adrenal cortex.

Biochemistry 14, 1477–1484.

24 Lambeth, J.D & Kamin, H (1976) Adrenodoxin reductase.

Properties of the complexes of reduced enzyme with NADP + and

NADPH J Biol Chem 251, 4299–4306.

25 Cole, S.T., Eiglmeier, K., Parkhill, J., James, K.D., Thomson,

N.R., Wheeler, P.R., Honore´, N., Garnier, T., Churcher, C.,

Harris, D et al (2001) Massive gene decay in the leprosy bacillus.

Nature 409, 1007–1011.

26 Peterson, J.A., Lorence, M.C & Amarneh, B (1990)

Putidar-edoxin reductase and putidarPutidar-edoxin Cloning, sequence

determi-nation, and heterologous expression of the proteins J Biol Chem.

265, 6066–6073.

27 Narhi, L.O & Fulco, A.J (1986) Characterization of a catalyti-cally self-sufficient 119,000-dalton cytochrome P-450 mono-oxygenase induced by barbiturates in Bacillus megaterium J Biol Chem 261, 7160–7169.

28 Ramachandra, M., Seetharam, R., Emptage, M.H & Sariaslani, F.S (1991) Purification and characterization of a soybean flour-inducible ferredoxin reductase of Streptomyces griseus J Bacteriol.

173, 7106–7112.

29 Dym, O & Eisenberg, D (2001) Sequence-structure analysis of FAD-containing proteins Protein Sci 10, 1712–1728.

30 O’Keefe, D.P., Gibson, K.J., Emptage, M.H., Lenstra, R., Romesser, J.A., Litle, P.J & Omer, C.A (1991) Ferredoxins from two sulfonylurea herbicide monooxygenase systems in Strepto-myces griseolus Biochemistry 30, 447–455.

31 Sielaff, B., Andreesen, J.R & Schra¨der, T (2001) A cytochrome P450 and a ferredoxin isolated from Mycobacterium sp strain HE5 after growth on morpholine Appl Microbiol Biotechnol 56, 458–564.

32 Manzella, L., Barros, M.H & Nobrega, F.G (1998) ARH1 of Saccharomyces cerevisiae: a new essential gene that codes for a protein homologous to the human adrenodoxin reductase Yeast

14, 839–846.

33 Barros, M.H & Nobrega, F.G (1999) YAH1 of Saccharomyces cerevisiae: a new essential gene that codes for a protein homo-logous to human adrenodoxin Gene 233, 197–203.

34 Lacour, T., Achstetter, T & Dumas, B (1998) Characterization of recombinant adrenodoxin reductase homologue (Arh1p) from yeast Implications in in vitro cytochrome P45011b monooxyge-nase system J Biol Chem 273, 23984–23992.

35 Lange, H., Kaut, A., Kispal, G & and Lill, R (2000) A mitochondrial ferredoxin is essential for biogenesis of cellular iron-sulfur proteins Proc Natl Acad Sci USA 97, 1050–1055.

36 Lill, R & Kispal, G (2000) Maturation of cellular Fe-S proteins:

an essential function of mitochondria Trends Biochem Sci 25, 352–356.

37 Mu¨hlenhoff, U & Lill, R (2000) Biogenesis of iron-sulfur proteins

in eukaryotes: a novel task of mitochondria that is inherited from bacteria Biochim Biophys Acta 1459, 370–382.

38 Hwang, P.M & Bunz, F., YuJ., Rago, C., Chan, T.A., Murphy, M.P., Kelso, G.F., Smith, R.A., Kinzler, K.W & Vogelstein, B (2001) Ferredoxin reductase affects p53-dependent, 5-fluoro-uracil-induced apoptosis in colorectal cancer cells Nat Med 7, 1111–1117.

39 Bellamine, A., Mangla, A.T., Dennis, A.L., Nes, W.D & Waterman, M.R (2001) Structural requirements for sub-strate recognition of Mycobacterium tuberculosis 14a-demethy-lase: implications for sterol biosynthesis J Lipid Res 42, 128–136.

40 Gatfield, J & Pieters, J (2000) Essential role for cholesterol

in entry of mycobacteria into macrophages Science 288, 1647–1650.