Characterization of the flavoprotein moieties of NADPH-sulfite

0 downloads 0 Views 2MB Size Report
Sep 25, 2018 - safranine T (-289 mV), and neutral red (-325 mV). All potentials in this paper are referred to the standard hydrogen electrode. After equilibration ...
THEJOURNAL OF BIOLOGICAL CHEMISTRY

Val. 264, No. 27, Issue of September 25, pp. 15796-15808,1989 Printed in U.S.A.

Characterization of the Flavoprotein Moietiesof NADPH-Sulfite Reductase from Salmonella typhimurium and Escherichia coli PHYSICOCHEMICAL AND CATALYTIC PROPERTIES, AMINO ACID SEQUENCE DEDUCED FROM DNA SEQUENCE OF CYSJ,AND COMPARISON WITH NADPH-CYTOCHROME P-450 REDUCTASE* (Received for publication, January 9, 1989)

Jacek Ostrowski$, Michael J. BarbertllII, David C. RuegerQll**,Barney E.Miller$ $$, Lewis M. Siegel$ll$$,and Nicholas M. Kredich$$ $$VI From the $Howard Hughes Medical Institute Laboratories and the Departments of §Biochemistry and WMedicine, Duke University Medical Center, Durham, North Carolina 27710 and the VBasic Science Division, Veterans AdministrationMedical Center, Durham, NorthCarolina 27705

NADPH-sulfite reductase flavoprotein (SiR-FP) was R. (1984) Biochemistry 23,6576-6583). Comparison purified from a Salmonella typhimurium cysG strain of the deduced amino acid sequences of SiR-FP and that does not synthesize the hemoprotein component of NADPH-cytochrome P-450 oxidoreductase (Porter, T. the sulfite reductase holoenzyme. cysJ, which codes D., and Kasper, C. B. (1985)Proc. Natl. Acad. Sei.U. for SiR-FP, wascloned from S. typhimurium LT7 and S. A. 82,973-977) also showed identitiesthat suggest Escherichia coli B, and both genes were sequenced. these two proteinsare descended from a common prePhysicochemical analyses and deduced amino acid se- cursor, which contained binding regions for both FMN quences indicatethat SiR-FPis an octamer of identical and FAD. 66-kDa peptides and contains 4 FAD and 4 FMN per octamer. Potentiometric titrationsof SiR holoenzyme, SiR-FP, andFMN-depleted SiR-FP yielded the following redox potentials for the prosthetic groups at pH Sulfite reductase (NADPH) (EC 1.8.1.2) from Salmonella 7.7: E ; (FMNH’FMN) = -152 mV; E ; (FMNH2/ typhimurium and Escherichia coli catalyzes the 6-electron FMNH’) = -327 mV; EA (FADH’FAD) = -382 mV; reduction of sulfite to sulfide (1)and is one of several activities E ; (FADHdFADH’) = -322 mV. Microcoulometric titration of SiR-FP at 25“C yielded data which were in required for the biosynthesis of L-cysteine from sulfate (2). full agreement with these potentials. Spectroscopic The and native enzyme has a subunit structure a& (3), where catalytic studies of native SiR-FP and of SiR-FP de- the octamer constituent a is a flavoprotein (SiR-FP)’ containing FAD and FMN andis coded for by the cysJ gene (4), pleted of FMN support the following electron flow and p is a hemoprotein (SiR-HP) containing siroheme and a 4 FMN. FMN canthen sequence: NADPHFAD Fe4S4 cluster and is coded for by cysl. The holoenzyme concontribute electrons to the hemoproteincomponent of sulfite reductase, as well as to cytochrome c and var- tains four FMN, four FAD, four FerS4 clusters, and four ious diaphorase acceptors. The FMN is postulated to sirohemes (5,6). Dissociation of holoenzyme with urea allows cycle between the FMNHz and FMNH’ oxidation states isolation of free SiR-FP as anoctamer and free SiR-HP asa during catalysis; inthis sense SiR-FP shares a catalytic monomer (3). SiR is unusual in that it contains both FAD mechanism with NADPH-cytochrome P-450 oxidore- and FMN, with each flavin playing a distinctive role in the ductase. electron transfer sequence (7). The FAD serves as an “input” SiR-FP domains involvedin bindingFMN, FAD, and center for receipt of an electron pair from NADPH, while the NADPH are proposed from amino acid sequence ho- FMN serves as anobligatory agent for rapid transfer of these mologies with Desulfovibrio vulgarisflavodoxin (Du- electrons to the heme prosthetic group of the hemoprotein bourdieu, M., and Fox, J. L. (1977) J. Biol. Chem. 252, component of SIR holoenzyme as well as toartificial acceptors 1453-1463) and spinach ferredoxin-NADP+oxidore- such as cytochrome c. The FAD is believed to use all three of ductase (Karplus,P. A., Walsh, K. A., and Herriott, J. its possible oxidation states during catalysis, whereas the * This work was supported by National Institutes of Health Grants FMN, afterits initial reduction, appears to cycle only between the hydroquinone (2-electron reduced) and the semiquinone DK 12828 (to N. M. K.), GM 32210 and GM 21226 (to L. M. S.), and by Veterans Administration Project Grant 215406554-01 (to L. M. (l-electron reduced) states (4). In this sense the FAD funcS.). The costs of publication of this article were defrayed in part by tions much like the FAD moiety of ferredoxin-NADP+ oxithe payment of page charges. This article must therefore be hereby doreductase (EC 1.18.1.2), while the FMNfunctions as itdoes marked “advertisement” in accordance with 18 U.S.C. Section 1734 in flavodoxin. solely to indicate this fact. The FAD and FMN prosthetic groups of SiR-FP share This paper is dedicated to thememory of Henry Kamin, a pioneer similarities in functionwith those of the well studied enzyme in research on the biochemistry of flavins in both sulfite reductase NADPH-cytochrome P-450 oxidoreductase (NADPH-ferriand cytochrome P-450 reductase. hemoprotein oxidoreductase, EC 1.6.2.4) from mammalian The nucleotide sequence(s) reported in this paper hus been submitted to the GenBankTM/EMBLData Bankwith accession number(s) 505025. I( Present address: Dept. of Biochemistry, University of South Florida Medical School, Tampa, FL 33612. ** Present address: Creative Biomolecules, Boston, MA 01748. $$ Present address: Abbott Laboratories, Abbott Park, IL 60064. To whom correspondence should be addressed.

The abbreviations used are: SiR-FP, sulfite reductase flavoprotein; SiR, NADPH-sulfite reductase; SiR-HP, sulfite reductase hemoprotein; des-FMN SiR-FP, sulfite reductase flavoprotein depleted of FMN; DCIP, 2,6-dichloroindophenokAcPyADP+,3-acetylpyridine adenine dinucleotide phosphate; HPLC, high pressure liquid chromatography; kb, kilobase(s); SDS, sodium dodecyl sulfate.

15796

Sulfite Reductase Flavoprotein liver microsomes (8). The two enzymes differ significantly, however, both in physiological function (9) and in the fact that SiR-FPbinds one FMN and one FAD to two apparently identical peptide chains (3), while NADPH-cytochrome P450 oxidoreductase binds one FMN and one FAD to a single peptide chain (10). Studies on SiR-FP derived from urea-dissociated E. coli SiR holoenzyme have been reported, but utilized protein that was substantially depleted of FMN and with low catalytic activity (3,7). We report here (a) a study of the physicochemical, enzymatic, and electrochemical properties of FMN-replete SiR-FP purified directly from a mutant strain of S. typhimurium that lacks SiR-HP; ( b ) the cloning and DNA sequence of the cysJ genes of S. typhimurium and E. coli B, which codefor SiR-FP; ( c ) a comparison of the deduced amino acid sequence of SiR-FP with that reported for rat liver NADPH-cytochrome P-450 oxidoreductase (11). EXPERIMENTALPROCEDURES

Materials-Calcium phosphate gel and L-djenkolic acid were from Sigma. Sephadex G-10 was purchased from Pharmacia, and DE52 microgranular DEAE-cellulosewas from Whatman. Deazaflavin was a generous gift from Dr.David Seibert. SiR holoenzymes werepurified from S. typhimurium LT2 and E. coli B by the procedure of Siegel et ul. (5). The optical and EPR spectroscopic properties of the two enzymes were indistinguishable. Bucterial Strains, Media, and Plasmids-S. typhimurium strains were from the Salmonella Genetic Stock Centre, University of Calgary, Alberta, Canada and included the LT2 derivatives cysG439 and AcysJZH383, and the LT7 strain hisG70. The E. coli B strain used wasATCC11303.EC1124was isolated by A. Wiater as a cysZ derivative of the E. coli K12 strain JA199 (from J. Carbon) and is AtrpES leu-6 thi hsdR hsdM+ cysZ. Minimal medium was medium E (12) supplemented with 0.5% glucose. L-Djenkolic acid at 0.15 mM was substituted for sulfate as a limiting sulfur source to derepress cysJ and maximize SiR-FP levels in cysG439. Rich media consisted of LB for plasmid transformations and YTfor M13 phage production (13). Solid media contained 1.5% agar, and ampicillin at 25 mg/liter was added when appropriate for selection of AmpRstrains. Plasmids pGBK5 and pJYW2 (Fig. 1) were isolated by selecting for Cys+transformants of the E. coli cysZstrain EC1124. pGBK5 was isolated by G. Jagura-Burdzy and carried cysZ from the S. typhimurium LT7 strain hisG70 on a 13.0-kb SaZI fragment inserted in

I

PStI

PStI

o

I

I

I

a

4

e

I

I

I

I

I

I I

e

10

la

14

16

18

kb

FIG. 1. Plasmids carrying the cysJIHregion from S. typhimurium LT7 and E. coli B. Both are derivatives of pBR322 and are shown in linear form beginning with the unique EcoRI site of this vector. The insert (bold line) in pGBK5 is a 13.0-kb Sal1 fragment from the S. typhimurium LT7 strain hisG70. Plasmid pJYW2 contains a 9.5-kb fragment (bold line) obtained from a partial Sau3A digest of E. coli B chromosomal DNA and inserted into the BamHI site of pBR322. The approximate positions of cysJ, cysZ, and cysH are shown abovethe scale.

RH

15797

-" -

"

t " c "

" " " " E. coll

FIG. 2. Sequencing strategies for regions of pGBK6 and pJYW2 corresponding tocysJ. Overlapping single-stranded templates were generated 'for dideoxy sequencing by the method of Dale et al. (19), and both strands were completely sequenced for the S. typhimurium and E. coli genes. The KpnI and EcoRI sites shown are present only in the pGBK5 sequence (uboue); the PstI and Hind111 sites are present only in the pJYW2 sequence (below). cysJ was identified as a 1.8-kb open reading frame in both sequences. pBR322. pJYW2 was isolated by J.-Y. Wu (in this laboratory) and carries cysZ from E. coli B on a 9.5-kb fragment obtained from a partial Sau3A digest, which was inserted in the BamHIsite of pBR322. When transferred to S. typhimurium AcysJZH383 by phage PBBHT-rn&liated transduction (14), pGBK5 gave a Cys+ phenotype with each, indicating this plasmid carries the entire cysJZH region. DNA sequence analyses reported here and elsewhere (66)' establish the relative positions of cysJ, cysZ, and cysH as shown in Fig. 1 and indicate that pJYW2 also carries the entire cysJZH region. DNA Methods-Recombinant DNA methods were those of Maniatis etal. (15) using enzymes and reagents purchased from Bethesda Research Labs, International Biotechnologies, Inc.,New England Biolabs, and Pharmacia-LKB Biotechnology Inc. DNA sequencing was performed by the method of Sanger et al. (16) using deoxyadenosine 5'-(a-[36S]thio)triphosphate(17) from Du Pont-New England Nuclear and field-strength gradient electrophoresis gels (18) maintained at 55 "C with thermostatic plates. Overlapping fragments of single-stranded DNA templates were generated from M13 phage derivatives by the method of Dale et al. (19). Sequencing strategies are shown in Fig. 2. Enzyme Assays-Assays of enzymatic activity were performed in 1.0-ml reaction volumes containing 0.1 M potassium phosphate, pH 7.7,0.1 mM Na2EDTA, 0.2 mM NADPH, 1 p M FMN, an electron acceptor, and enzyme. Electron acceptors were present at thefollowing concentrations: 0.5 mM sulfite, 10 mM hydroxylamine, 0.1 mMof mM either cytochrome c, DCIP, ferricyanide, or menadione, 0.2 AcPyADP+. Rates were measured at 25 "C in a Cary model 14spectrophotometer against a reference solution containing all reaction components except enzyme. Absorbance changes werefollowed at 340 nm for sulfite, hydroxylamine, ferricyanide, or menadione as acceptors; at 550 nm for cytochrome c; at 600 nm for DCIP and at 363 nm for AcPyADP, using extinction coefficients reported by Siegel et al. (1). Oneunit of SiR-FP activity is defined as theamount that catalyzes oxidation of 1pmol of NADPH/min with cytochrome c as acceptor in the reaction mixture described above. In assays carried out without added FMN, enzyme at a concentration appropriate for each assay was prepared by rapid dilution of a concentrated stock solution ( X . 1 pM) at 0 "C just prior to assay, in order to minimize dissociation of FMN (7). Purification of SiR-FP-The S. typhimurium strain cysG439 lacks the ability to synthesize siroheme, a cofactor of the hemoprotein component of SIR holoenzyme (20), and is a convenient source of free SiR-FP, which can be assayed by its NADPH-cytochrome c reductase activity (3). cysG439 was grownwith vigorous aeration to a density of 5-10 X 10" ml" in minimal medium containing the limiting sulfur source L-djenkolic acid at 0.15 mM. Cells were harvested by centrifugation and stored at -70 "C until further use. For the preparation summarized in Table I, 230 g of cells were thawed overnight at 4 "C, suspended in 460 ml of standard buffer (0.05 M potassium phosphate, 0.1 mM NaZEDTA, pH 7.7), and disrupted by sonic oscillation at temperatures notexceeding 15 "C. Allsubsequent steps were performed at 0-4 "C. Particulate matter was removed by centrifugation at 44,000 X g for 30 min and resuspended in another 230 ml of standard buffer. Following further treatment with sonic oscillation, 'Ostrowski, J., Wu, J-Y., Rueger, D. C., Miller, B. E., Siegel, L. M., and Kredich, N. M. (1989) J. Bioi. Chem. 264, in press.

15798

Flavoprotein Reductase

Purification step

Sulfite

TABLEI Purification of SiR-FP from S. tvphimurium cysG439 Total Total Volume enzymatic protein activitP ml

0.45

Crude 8,100 extract' 17,900 Streptomycin supernatant First ammonium sulfate 97 0.64 Calcium phosphate 4,580 gel 208 Second ammonium sulfate Bio-Gel A-1.5 D O O ~(concentrated)

795 1,135 2707,840 1,475 10.6 3,380 4.0

mg

units

Specific activity

Yield

unitslmg

%

8,000

99

12,200 47 13

1.050

22 71 81

57 42 13

"Assayed as NADPH-cytochrome c reductase. One unit of enzyme catalyzes the reduction of 1 pmol of cytochrome c/min. * From 230 g of cell paste.

this suspension was again centrifuged, and the combined supernatants were mixed with 0.5 volume of 10% (g/volume) streptomycin sulfate that had been dissolved in standard buffer and neutralized with 5 M KOH. After stirring for 1 h the mixture was centrifuged at 16,000 X g for 40 min, and the supernatant was brought to 27% saturation with ammonium sulfate by the addition of157 mg/ml. After 1 h of stirring, precipitated protein was removed by centrifugation at 16,000 X g for 30 min, and the supernatant was brought to 45% saturation by the addition of another 113 mg/ml of ammonium sulfate. After 30 min the precipitate was collected by centrifugation at 12,000 X g for 40 min and dissolved in buffer to a concentration of approximately 5 units of enzyme/ml. Calcium phosphate gel was suspended in water at 10 mg dry weight/ ml and added to theenzyme solution a t a ratioof approximately 0.25 mgof calcium phosphate/unit of enzyme. The amount of gel used varied from one preparation to another and was determined each time in a small pilot run as the amount required to adsorb 80-90% of enzyme activity. The suspension was stirred for 1 h and centrifuged at 12,000 X g for 40 min. The gel was resuspended in one-third the starting volume of standard buffer, stirred for 1h, andagain collected by centrifugation. Enzyme activity was eluted by suspending the gel in one starting volume of0.15 M potassium phosphate, pH 6.75, containing 0.1 mM Na2EDTA. After 2 h of stirring the gel was removed by centrifugation, and thesupernatant was brought to 42% saturation by the addition of 244 mg/ml ammonium sulfate. Thirty minutes later precipitated protein was collected by centrifugation, dissolved in a small amount of standard buffer, and centrifuged for 2 h a t 44,000 X g to remove any residual particulate matter. Size exclusion chromatography was carried out with 10-mgportions of protein using a 90 X 2.5-cm column of Bio-Gel A-1.5m agarose (Bio-Rad) equilibrated with standard buffer and run at a flow rate of 5 ml/h. Fractions of 2.5 ml were collected, and those with a specific activity of 75 units/mg or greater were pooled and concentrated by ultrafiltration with a collodion bag apparatus (Schleicher & Schuell, Inc.). This final product contained 10-20% of the enzyme activity in the crude extract and represented a purification of approximately 200-fold (Table I). Polyacrylamide Gel Electrophoresis-SDS-polyacrylamide gel electrophoresis was performed according to the procedure of Laemmli (21), and protein bands were stained with Coomassie Blue.Polyacrylamide gel electrophoresis in 8 M urea and 5% polyacrylamide (22) was performed with protein samples that had been incubated overnight at 23 "C in standard buffer containing 8 M urea and 0.3 M 2mercaptoethanol. Sedimentation Measurements-These were conducted with a Beckman Model E ultracentrifuge equipped with electronic speed and temperature control units and a photoelectric scanner with multiplexer accessories. Protein samples were dialyzed extensively against reference solutions. Sedimentation velocity experiments were carried out at 25 "C in an An-F rotor with Schlieren optics. Sedimentation equilibrium experiments were performed at 20 "C in An-J and An-F rotors. Solution absorbance was measured at 278 nm and also at 456 nm for native enzyme. M, values were calculated from the equation M, = 2RT [(dlnA)/(dr2)]/[w2(l- Bp)]. A value for B was determined from the amino acid composition (23), and p was determined pycnometrically. Measurements at 456 and 278 nm for native enzyme gave results that differed by no more than &5%, andM, values are given as the average for both wavelength scans. Spectroscopy-UV and visible spectra were measured in 1-cm path-

length silica cells at 23 "C with an Aminco DW-2 spectrophotometer operating with a 2 nm bandwidth. EPR spectra were recorded a t X band on a Varian E9 spectrometer equipped with a Varian variable temperature liquid N2 cryostat and 100 kHz modulation. EPR spectra were recorded at 173 K, 2.5 G modulation amplitude, and 10 microwatt power, which wasshown to be nonsaturating for SiR-FP flavin radical. Spin concentrations were determined by double integration with CuEDTA as a standard. Protein Methods-Unless otherwise noted, protein concentrations were measured by the Zamenhof (24) microadaptation of the microbiuret method described previously (5). Bovine serum albumin assayed spectrophotometrically with an E;% value of 6.7 cm" (25) was used as a protein standard. Dry weight measurements of protein were performed on samples of SIR-FP that had been dialyzed exhaustively against standard buffer (26). Duplicate 2.0-ml samples of SiR-FP with A456 = 0.52 were found to contain 6.0 and 6.1 mg of protein. Amino acid analyses were performed according to Spackman et al. (27) with a Beckman Model 120B automatic amino acid analyzer. A known amount of L-norleucine was added as a standard to each protein sample, which was then hydrolyzed in 6 N HCl in uucuo for 28, 48, and 72 h at 110 "C. Half-cystine was determined as cysteic acid after performic acid oxidation (28), and tryptophan was estimated spectrophotometrically (29) on protein samples that had been dialyzed extensively against 6 M guanidine HC1 to remove prosthetic groups. Protein S-carboxymethylation was carried out with samples dissolved in 6M guanidine HC1,0.14 M 2-mercaptoethanol, pH 8.6. After incubation at 23 "C for 4 h, iodoacetamide in 6 M guanidine HCl, pH 8.6, wasadded in slight excess over the 2-mercaptoethanol. Details of our methods for cyanogen bromide digestion of proteins, purification of peptides by high performance liquid chromatography, gas phase peptide sequencing (30), and carboxyl-terminal analyses of peptides with carboxypeptidases have been described (31). Cofactor Analyses-FMN and FAD were determined fluorometrically on boiled enzyme samples by the method of Faeder and Siegel (32). Total iron content was measured as described by van de Bogart and Beinert (33) using iron wire dissolvedin 20% HN03 asa standard. Acid-labile sulfide was determined by the procedure of Siegel et al. (5), and siroheme was determined spectrophotometrically in acetoneHCl extracts of enzyme, to which pyridine had been added (34). Preparation of FMN-depleted SiR-FP-FMN was removed byphotolysis as it dissociated from dilute solutions of SiR-FP (A156 = 0.0200.025) in 0.1 M potassium phosphate, pH 7.7, containing 0.1 mM Na2EDTA and 1.2 M ammonium sulfate (7). Irradiation for 7h resulted in removal of 90-95% of the FMN and less than 10% of the FAD. Longer treatment periods were not used in this work because they resulted in the progressive and irreversible loss of both FAD and FMN. This preparation is termed "des-FMN" SiR-FP in this work. Photochemical Reduction of SiR-FP-Solutions of SIR-FP or desFMN SiR-FP at A456 = approximately 0.3 were prepared in 0.1 M potassium phosphate, pH 7.7, containing 10 mM Na2EDTA and 2.5 p~ deazaflavin, placed in sealed 1-cm silica cuvettes in the dark (351, and made anaerobic by repeated evacuation and flushing with argon (Matheson, high purity). Residual O2was removedfrom the argon by passing it successively, in an all glass system, through a BASF copper catalyst column at 190°C and then through a solution of reduced methyl viologen. The enzyme solutions were then subjected to short (1-2-min) periods of illumination from a General Electric 200-watt sealed beam spot lamp placed 25 cm from the cuvette. The cuvettes

Fluvoprotein Sulfite Reductase were placed in a water-ice mixture to prevent overheating of the enzyme solution during the illumination procedure. After each illumination period, absorption spectra were measured with an Aminco DW-2 spectrophotometer. Aqueous solutions of deazaflavin were prepared and analyzed spectrophotometrically as described (36). Ekctrochemical Experiments-Potentiometric titrations were carried out at 25 "C under argon in an anaerobic glass cell of the type described by Dutton (3711, modified by addition of an extramicroelectrode to allow continuous monitoring of pH. The cell was fitted with a gold foil indicator electrode and a calomel electrode (Radiometer K401) calibrated against a standard solution of quinhydrone. Enzyme solutions, 62-86 p~ in total flavin in 0.1 M potassium phosphate, pH 7.7, containing 0.1 mM NazEDTA, were first reduced by either photoreduction or by addition of NADPH to a final concentration of 1.0 mM (sum of NADPH an,d NADP+). They were then adjusted to more positive potentials by addition of small volumes of 0.2 M potassium ferricyanide in 0.1 M potassium phosphate, pH 7.7, containing 0.1 mM Na,EDTA. Photoreduction was achieved by irradiation of the enzyme solution, containing 20 p~ deazaflavin and 10 mM Na2EDTA with a 200-watt sealed beam lamp until a stable potential was achieved. Equilibration between enzyme and theindicating electrode was facilitated by use of a mixture of the following dye mediators (present at 20 p M each) of appropriate midpoint potential: 2,5dihydroxybenzoquinone (-60 mV), indigodisulfonate (-125 mV), anthroquinone-2,7-disulfonate(-182 mV), phenosafranine (-255 mV), safranine T (-289 mV), and neutral red (-325 mV). All potentials in thispaper are referred to the standardhydrogen electrode. After equilibration periods of a t least 5 min a t each potential(until the potential stopped drifting with time), a sample of enzyme was removedby means of a gas-tight Hamilton syringe fitted with a 30-cm stainless steel needle, transferred to an argon-flushed EPR tube, frozen in liquid NZ,and stored at 77 K for subsequent recording of EPR spectra. Equilibrium was achieved after 5-20 min in all redox titrations reported in this work. Theoretical curves were calculated as described by Iyanagi et al. (38) for flavin semiquinone species formed as intermediates in two consecutive single-electron processes. Precision errors in redox potentials derived from the type of analysis used in the pyesent work are expected to be f 1 5 mV (39). Microcoulometry was performed a t 25 "C as described by Spence et al. (40) with the exception that thefollowing mediators were used indigodisulfonate (-125 mV), anthroquinone-2,7-disulfonate(-182 mV), phenosafranine (-255 mV), benzyl viologen (-311 mV), and methyl viologen (-440 mlV).

15799 TABLE I1

Sedimentation equilibrium of purified SiR-FP from S. typhimurium Protein solutions were dialyzed extensively against the solvent indicated and were centrifuged to equilibrium a t 25 "C in a Beckman Model E ultracentrifuge. Solvent

Protein concentration wlml

6 M guanidine HCI" 0.38

8 M ureab

Mean value (+S.D.) Standard buffef

0.28 0.19 0.10 0.38 0.19 0.39 66,600 29,000 0.10

Rotor speed

Calculated molecular weight

rpm

36,000 61,900 36,000 55,400 36,000 58,400 36,000 65,800 29,000 57,800 29,000 58,600 29,000 59,900 60,600 (+ 4,000)

0.40 0.30 0.20

488,000 12,000 12,000 12,000 12,000 8,000 8,000 8,000 8,000

458,000 459,000 0.10 478,000 0.40 452,000 0.30 455,000 0.20 477,000 0.10 548,000 Mean value (+S.D.) 488,000 (+- 37,000) S-Carboxymethylated protein was used for equilibrium sedimentation in standard buffer containing 6 M guanidine HCl. 2-Mercaptoethanol was included a t 0.14 M in standard buffer containing 8 M urea. Standard buffer was 0.05 M potassium phosphate, 0.1 mM Na2EDTA,pH 7.7.

ratios of absorbances at these respective maxima, when corrected for elevated base-line absorbance at 740 nm, are 8.4:0.87:1.0. Based on dry weight protein measurements the E:z of SiR-FP was calculated to be 1.72 cm". Analyses of three different preparations of SiR-FP showed 15.8 f 0.8 nmol of flavin/mg protein, with a molar ratio of FMN toFAD of approximately 1.0 (Table 111). These data yield an extinction coefficient of 10.9 (mM flavin)" cm" a t 456 nm for SiRFP. The minimum molecular weight of 63,300 k 3,000 per RESULTS flavin is in good agreement withthe subunitmolecular weights Purity, Molecular Weight, and SubunitComposition of SiR- obtained from SDS-polyacrylamide gel electrophoresis, ultraFP from S. typhimurium-Purified preparations of S. typhi- centrifugation, and from the deduced amino acid sequence murium SiR-FP yielded a single band after polyacrylamide (see below). From these data we conclude that each SiR-FP gel electrophoresis in ISDS or 8 M urea. A subunit mass of 66 subunit contains either 1 mol of FMN or 1mol of FAD and kDa was estimated by SDS-polyacrylamide gel electrophore- that theoctameric protein contains 4 FAD and 4 FMN. sis. The electrophoretic migration of the SiR-FPsubunit in 8 Analysis of two SiR-FP preparations for iron yielded only M urea and in SDS was identical to thatof the SiR holoenzyme 1.2 f 0.1 nanoatoms of iron/mg dry weight, i.e.