Neurospora crassa NAD(P)H-Nitrite Reductase STUDIES ON ITS COMPOSITION AND STRUCTURE*

Neurospora crassa nitrite reductase (Mr = 290,000) catalyzes the NAD(P)H-dependent 6-electron reduction of nitrite to ammonia via flavin and siroheme prosthetic groups. Homogeneous N. crassa nitrite reductase has been prepared employing conventional purification methods followed by affinity chromatography on blue dextran-Sepharose 4B. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of homogeneous nitrite reductase reveals a single subunit band of Mr = 140,000. Isoelectric focusing of dissociated enzyme followed by sodium dodecyl sulfate-gel electrophoresis in the second dimension yields a single subunit spot with an isoelectric point at pH 6.8-6.9. Two-dimensional thin layer chromatography of acid-hydrolyzed nitrite reductase treated with 5-dimethylaminoaphthalene-1-sulfonyl chloride yields a single reactive NH2-terminal corresponding to glycine. An investigation of the prosthetic groups of nitrite reductase reveals little or no flavin associated with the purified protein, although exogenously added FAD is required for activity in vitro. An iron content of 9-10 Fe eq/mol suggests the presence of nonheme iron in addition to the siroheme moieties. Amino acid analysis yields 43 cysteinyl residues and sulfhydryl reagents react with 50 thiol eq/mol of nitrite reductase. The non-cysteinyl sulfur content, determined as 8.1 acid-labile sulfide eq/mol, is presumably associated with nonheme iron to form iron-sulfur centers. We conclude that N. crassa nitrite reductase is a homodimer of large molecular weight subunits housing an electron transfer complex of FAD, iron-sulfur centers, and siroheme to mediate the reduced pyridine nucleotide-dependent reduction of nitrite to ammonia.

soluble, sulfhydryl-containing metalloflavoprotein which mediates the 2-electron transfer sequence (9, and references therein): NADPH + [-SH + FAD + cytochrome bs;; + Mol -+ NO.%-The second enzyme of the nitrate assimilation pathway, nitrite reductase, mediates the &electron reduction of nitrite to ammonia. Bacterial nitrite reductases are FAD-dependent metalloproteins and t-ypically utilize NADH as electron donor (10,11). The enzyme from photosynthetic organisms is characterized as a low molecular weight (60,000-70,000) FADindependent metalloprotein for which reduced ferredoxin serves as electron donor (12). Nitrite reductases from the fungi A. nidulans (13) and N. crassa (14) can accept electrons from either NADH or NADPH. The nitrite reductase complex from N . crassa is a large molecular weight (290,000) electrontransport system which requires FAD for activity in vitro and which catalyzes the stoichiometric reduction of nitrite to ammonia using 3 eq of reduced pyridine nucleotide to provide the necessary 6 electrons (15).
N . crassa nitrite reductase was fiist described by Nason et al. (14). In 1974, Lafferty and Garrett (15) reported a 90-fold purification of N . crassa nitrite reductase. In addition to the physiological activity of catalyzing the NAD(P)H-dependent reduction of nitrite to ammonia the enzyme also demonstrated FAD-dependent NAD(P)H-hydroxylamine reductase activity as well as FAD-independent dithionite-nitrite reductase activity. In 1975, Vega and Garrett (16,17) reported that nitrite reductase has an associated NAD(P)H-diaphorase activity for which a dye such as cytochrome c, ferricyanide, menadione, or dichloroindophenol may serve as electron acceptor. The absorption spectrum of the partially purified nitrite reductase exhibited maxima at 405, 555, 585, and 600 nm (15) and suggested the presence of a novel heme prosthetic group, siroheme. Vega et al. (18) demonstrated, by spectral analysis of N. crassa nitrite reductase and its extracted heme chromophore, that the heme moiety was indeed siroheme. The oxidized nitrite reductase showed a characteristic heme absorption maximum at 578 nm. A typical "reduced enzyme" spectrum resulted upon the addition of either NADPH and FAD or dithionite. Further spectral changes which represented an interaction of nitrite with the heme moiety resulted when nitrite was added to the reduced enzyme, thereby implying a functional role of the chromophore in nitrite reduction. In addition, incubation of nitrite reductase with CO in the presence of both NADPH and FAD resulted in inhibition of enzymatic activity and the formation of a spectrally distinct complex. Nitrite competed effectively with CO and reversed the inhibition, and the resulting spectrum was identical with the reduced enzyme-nitrite complex spectrum. FAD apparently functions in the transfer of electrons from NAD(P)H to siroheme. Siroheme is presumably the site of nitrite binding. All of the NAD(P)H-dependent activities of nitrite reductase 9711 are sensitive top-hydroxymercuribenzoate (15)(16)(17). It is anticipated that a functional sulfhydryl moiety participates in the initial reactions of the electron transfer sequence (19). These observations are the basis of the following scheme for the nitrite reductase where brackets define the enzyme complex: s2042- The N. crassa nitrite reductase was purified to homogeneity by Greenbaum et al. in 1978 (20). The marked instability of previous preparations was overcome and homogeneous enzyme was obtained following affinity chromatography on blue dextran-Sepharose 4B.
The present paper reports the results of physical and chemical studies on the purified N. crassa nitrite reductase. The amino acid composition and subunit organization of the enzyme have been examined, Determinations of the number of various prosthetic groups essential to nitrite reductase activity are reported and comparisons to other multielectron transferring systems are drawn.
Purification of Nitrite Reductase NAD(P)H-nitrite reductase was purified from N. crassa mutant nit-1 through Fraction 6 of the procedure of Greenbaum et al. (20). In a typical preparation, 600 g of frozen nit-1 mycelia were homogenized at 0 "C with 1500 ml of 0.1 M potassium phosphate buffer, pH 7.3, containing 5 mM EDTA, 5 mM cysteine, and 10% glycerol. The homogenate was centrifuged at 27,000 X g for 20 min. The resulting supernatant, Fraction 1, was treated by conventional methods of ammonium sulfate fractionation and DEAE-cellulose ion exchange chromatography through the step yielding Fraction 6 (37.5-47.5% (NH4),S04 precipitate), which represented a 55% recovery and a 14fold purification. At this stage of purification, Fraction 6 was divided into 0.3-ml aliquots and stored at -70 "C. To obtain homogeneous nitrite reductase, a single aliquot of Fraction 6 was chromatographed on a Sepharose 4B column (1.5 X 30 cm; bed volume, 50 ml) previously equilibrated with 10 mM potassium phosphate buffer, pH 7.0, containing 5 mM EDTA, 5 mM cysteine, 10% glycerol, and 2 mM sodium sulfite. Following elution with this buffer, the active fractions were pooled and designated as Fraction 7.
Fraction 7 was immediately chromatographed on a blue dextran-Sepharose 4B affinity column (1.5 X 2.2 cm; bed volume, 4 ml) equilibrated with the previous buffer. The blue dextran-Sepharose 4B resin was prepared according to Ryan and Vestling (21). The column was washed sequentially with 30 mM potassium phosphate equilibration buffer, buffer containing 0.5 mM ATP and 0.5 m~ NADP, and buffer alone. Nitrite reductase was eluted from the column with 15 ml of equilibration buffer containing 0.5 mM NADPH to yield the final fraction, the pooled blue dextran-Sepharose 4B eluates (Fraction 8). Recoveries of 30-50% of Fraction 7 and a 1400-fold purification were typical.

Assay of NADPH-Nitrite Reductase
Nitrite-dependent pyridine nucleotide oxidation was followed at 340 nm as described previously (15)

Protein Determinations
Protein concentrations were determined by a modified Lowry method using crystalline bovine serum albumin as the standard (22). In the Fraction 8 samples (blue dextran-Sepharose 4B eluates), the protein was first precipitated by addition of 0.025 ml of 2% sodium deoxycholate and 1 ml of 24% trichloroacetic acid (23). This treatment eliminated interferences caused by the nucleotides and other buffer components. The protein concentration was then determined by the modified Lowry method (22).

Electrophoretic Analyses
Polyacrylamide gel electrophoresis was performed according to Clark (24). A discontinuous Tris/glycine buffer system (pH 8.6) was used. Fraction 8 eluates (10-100 pg of protein/O.l-to 0.2-ml samples) were electrophoresed on 5% cross-linked gels for 4 h at 4 "C and 2 mA/tube. SDS'-polyacrylamide gel electrophoresis was performed according to the method of Maize1 (25). Fraction 8 eluates were dialyzed overnight against 10 mM sodium phosphate buffer, pH 7.2. Samples containing 10-50 pg of protein/O.l ml were treated with 1-2% SDS and 1-2% 8-mercaptoethanol and incubated for 5 min in a boiling water bath. Standard proteins of known subunit molecular weight were similarly prepared. Approximately 10 pg each of myosin, RNA polymerase p and p, ,&galactosidase, and phosphorylase a were treated alone or with the sample nitrite reductase. Electrophoresis proceeded at room temperature as described above.
The dissociation of Fraction 8 eluates in guanidine HCl was performed according to Weber et al. (26). The treated protein samples were dialyzed overnight against 10 mM sodium phosphate buffer, pH 7.2, containing 0.1% SDS and then resolved on 5% polyacrylamide gels as described above.
Electrophoresis in the nonequilibrium system (32) proceeds toward the cathode with 0.01 M &PO4 in the upper reservoir and 0.02 M NaOH in the lower reservoir. Since the polarity is reversed, basic proteins enter the gel first and are better resolved in this system than on isoelectric focusing gels (pH 4-7). Fraction 8 samples were dissociated and prepared in sample dilution buffer as described above. Electrophoresis proceeded for 5-6 h at 300 V. Gels were equilibrated for 30-60 min prior to electrophoresis in the second dimension.
' The abbreviations used are: SDS, sodium dodecyl sulfate; dansyl, Amino Acid Analysis Amino acid analysis was performed on a Beckman automatic amino acid analyzer Model 120 C. Aliquots of Fraction 8 were hydrolyzed under vacuum in 6 N HCI for 24,48, and 72 h at 105 "C. Cysteine was determined as carboxymethyl cysteine following reduction in the presence of P-mercaptoethanol, alkylation of the reduced protein with iodoacetic acid (33), and hydrolysis at 105 "C for 18 h. Tryptophan was estimated spectrophotometrically by the method of Edelhoch (34).

NH2-terminal Analysis
The NH2 termini of nitrite reductase were labeled with 5-dimethylaminonaphthalene-1-sulfonyl chloride by a modification of the method of Gray (35,36). Aliquots of Fraction 8 eluates in 0.5 M sodium bicarbonate, pH 9.5, were treated with an equal volume of dansyl chloride (20 mg/ml of acetone) for 16-20 h. The labeled protein was hydrolyzed under vacuum in 6 N HCI at 105 "C. After 6-18 h the sample was dried and the amino acids were resuspended in deionized H20. A by-product of the dansylation reaction, dansyl sulfonic acid, was extracted with toluene. The dansyl amino acids were extracted with ethyl acetate, dried under N2, and resuspended in methanol.
The dansylated amino preparation (approximately 0.3 nmol of nitrite reductase) was applied to a polyamide thin layer chromatography sheet (20 X 20 cm). Two-dimensional chromatography was performed according to Woods and Wang (37). The dansyl amino acids were separated in a fmt dimension solvent composed of water and 90% formic acid (1W1.5). After 45 min the sheet was dried and then developed in n-heptane/l-butanol/glacial acetic acid (331) for 60 min. The fluorescent spots corresponding to the dansyl amino acids were visualized under UV light. The two-dimensional migration of the spots was compared to a reference map constructed with standard dansylated amino acids. In addition, bovine pancreatic insulin served as a standard protein. Labeling, hydrolysis and chromatography of the standard preparations proceeded as described above for nitrite reductase.
Prosthetic Group Analysis Flavin-The flavin content of homogeneous preparations of nitrite reductase was determined fluorometrically by the method of Faeder and Siegel (38). Samples contained 5-20 pg of protein.
Iron-Following electrophoresis of the native protein the iron content of nitrite reductase was demonstrated in situ on polyacrylamide gels by two staining methods. Gels were incubated in 0.2% dimethoxybenzidine, 5% acetic acid, 0.03% Hz02 in the dark for 30-45 min. Heme iron appeared as a dark brown band following extensive rinsing of the gel in 15% acetic acid (39). Nonheme iron was stained upon incubation of the gel in 0.7% a,a'-dipyridyl and 8% thioglycollic acid for 10-30 min (40).
The total iron content of nitrite reductase was measured on a Perkin-Elmer atomic absorption spectrophotometer (Model 370) equipped with a graphite furnace HGA 2100 attachment. Aliquots of Fraction 8 (60-80 pg/ml) were dialyzed extensively against distilled water. Samples of 20 pl of standard iron and protein preparations were dried at 110 "C for 20 s, charred at 1100 "C for 60 s, and atomized at 2600 "C for 15 s. Absorbance was measured at 248.3 nm with a slit width of 0.2 nm.
Total iron content was also measured colorimetrically by a modification of the method of Fortune and Mellon (41). Fraction 8 eluates (0.14-0.24 nmol of nitrite reductase/0.4 ml of elution buffer prepared without EDTA) were treated with 0.05 ml of 10% HCI and then incubated at 80 "C for 10 min. In rapid succession, 0.25 ml of 10% hydroxylamine and 0.3 ml of 0.5% o-phenanthroline were added. Color development was complete in 15 min and absorbance was measured at 512 nm.
Labile Sulfide Content-The acid-labile sulfide content of nitrite reductase was determined by measuring methylene blue formation at 670 nm using the procedure described by King and Moms (42) and Siegel et al. (43). Aliquots of Fraction 8 contained 75-200 pg of protein and were eluted from the blue dextran-Sepharose 4B affinity column in elution buffer prepared without cysteine or sulfite.
Sulfiydryl Content-The total sulfhydryl content of nitrite reductase was examined with the thiol reagent 4,4'-dithiodipyridine (44). Aliquots of Fraction 8 (15-37 pg of nitrite reductase in blue dextran-Sepharose 4B elution buffer prepared without cysteine or sulfite) were incubated with the reagent in the presence and absence of 8 M urea. The absorbance of 4-thiopyridone, the species formed upon reaction of thiols with 4,4'-dithiodipyridine, was measured at 324 nm.

Subunit Composition of Nitrite Reductase
The homogeneity of nitrite reductase preparations was routinely confirmed by electrophoresis of blue dextran-Sepharose 4B eluates (Fraction 8) on polyacrylamide gels. Fig. 1A  tase on a 5% gel stained with Coomassie brilliant blue R250. Samples of homogeneous nitrite reductase were dissociated with SDS and P-mercaptoethanol and electrophoresis in a discontinuous Tris/glycine/O.l% SDS buffer system followed. Fig. 1B shows the single protein band observed following such treatment on a 5% polyacrylamide gel. Similarly, samples of nitrite reductase denatured in the presence of guanidine hydrochloride revealed one sharp protein band following electrophoresis (Fig. 1 0 . The molecular weight of the nitrite reductase subunit corresponding to the single protein band was determined by coelectrophoresis of SDS-dissociated nitrite reductase and standard proteins of known subunit molecular weight. A plot of mobility factor versus log of the subunit molecular weight revealed a subunit molecular weight of 140,000 for nitrite reductase (Fig. 2). N . crassa nitrite reductase (M, = 290,000) is apparently a dimer of identical molecular weight subunits.

Two-dimensional Electrophoresis
Isoelectric focusing of dissociated nitrite reductase samples in a pH gradient of pH 4-7 or electrophoresis in a nonequilibrium pH gradient followed by a size separation of subunits on a 5% slab gel in a Tris/glycine/O.l% SDS system in the second dimension revealed a single protein spot corresponding to the M , = 140,000 subunit of nitrite reductase. The subunit has an isoelectric point of pH 6.8-6.9. The homodimeric nature of the enzyme is further supported since the subunits demonstrated the same isoelectric point as well as identical molecular weight.

Amino Acid Composition
The amino acid composition of nitrite reductase is shown in Table I. A partial specific volume of 0.730 cm:' g" was calculated by the method of Cohn and Edsall (46). Assuming integral values for the amino acid composition, a molecular weight of 289,353 was calculated.

NH2-terminal Analysis
The NH2 termini of nitrite reductase were labeled with 5dimethylaminonaphthalene-1-sulfonyl chloride. Following hydrolysis of the protein and partial removal of the by-products of the dansylation procedure, the highly fluorescent, ethyl acetate-extracted dansyl amino acids were applied to a polyamide thin layer chromatography sheet. Fig. 3 shows a UV-  " Measured as carboxymethylcysteine by the method of Anfinsen et al. (33).
Estimated by extrapolation to zero time of hydrolysis. Determined spectrophotometrically by the method of Edelhoch Dansyl sulfonic acid demonstrated a blue fluorescence, dansyl sulfonamide appeared orange and the dansyl amino acid was yellow when visualized under UV light.
The dansylated nitrite reductase preparation was also chromatographed in a mixture of several standard dansyl amino acids. When compared to a two-dimensional map constructed with 18 dansyl amino acids, the single fluorescent product of the reaction of nitrite reductase with dansyl chloride corresponded to dansyl glycine.

Prosthetic Group Analysis
Flavin-Quantitative fluorescent measurements of the flavin content of homogeneous nitrite reductase were made by a procedure based on the pH-dependent fluorescent behavior of FAD and FMN. The fluorescence of supernatants from boiled Fraction 8 eluates was measured at pH 7.7 and pH 2.6. Table  I1 shows that very little flavin is associated with the purified protein. The amount of FAD ranged from 0.13-0.2 mol/mol of nitrite reductase.
Iron Analysis-Atomic absorption spectroscopy measurements of the iron content of homogeneous nitrite reductase preparations dialyzed extensively against glass-distilled water gave an average value of 10.2 mol of Fe/mol of nitrite reductase (Table II). An alternate analytical procedure, based on the reaction of reduced iron with o-phenanthroline, permitted the direct assay of iron in Fraction 8 eluates. An average value of 9 mol of iron/mol of nitrite reductase was obtained by this method (Table 11). The total iron content of nitrite reductase presumably includes heme iron associated with siroheme moieties and nonheme iron involved in iron-sulfur centers.
Sulfiydryl Analysis-The available versus total sulfhydryl content of nitrite reductase was analyzed with the thiol reagent 4,4'-dithiodipyridine. The reaction of standard cysteine solutions with 4,4'-dithiodipyridine was complete in 5 min and absorbance at 324 nm was linear from 2-20 nmoI of sulfhydryl equivalents. The standard proteins, Klebsiella nitrogenase and rabbit muscle aldolase, required a 90 min incubation period in the presence of 8 M urea to completely expose and react all sulfhydryl equivalents. Table I11 shows that under these reaction conditions, 50 sulfhydryl eq/mol of nitrite reductase reacted with 0.08 mM 4,4'-dithiodipyridine. In the absence of urea, only 37.4 sulfhydryl groups were reactive or accessible.
The sulfhydryl content of homogeneous nitrite reductase was also examined with 5,5'-dithiobis(2-nitrobenzoic acid). A high concentration of reagent and the addition of 8 M urea to   Table III shows that, after 30 min, 50 sulfhydryl eq/mol of nitrite reductase reacted with 2 mM 5,5'dithiobis(2-nitrobenzoic acid) in the presence of 8 M urea. Only 34.6 sulfhydryl eq/mol of nitrite reductase were accessible in the native enzyme. No additional sulfhydryl groups were demonstrated with 4 mM 5,5'-dithiobis(2-nitrobenzoic acid).
Substrate Affinities The Michaelis constants of NADPH-nitrite reductase for the substrates NADPH and nitrite and the cofactor FAD were determined with homogeneous enzyme preparations. Nitrite-dependent NADPH oxidation was followed at 340 nm and initial velocities were determined from absorbance values recorded every 3 s over a 20-s period. A least squares regression analysis of all kinetic data was performed. The apparent K , values are 15 PM for NADPH, 7.5 PM for nitrite, and 0.02 p~ for FAD.

DISCUSSION
The Neurospora crassa nitrite reductase (Mr = 290,000) has been purified to homogeneity by conventional methods of ammonium sulfate fractionation, DEAE-cellulose ion exchange chromotography, and Sepharose 4B gel fitration chromatography, followed by affinity chromatography on blue dextran-Sepharose 4B. The purified enzyme exhibited one band upon polyacrylamide gel electrophoresis under nondissociating conditions. Various staining techniques were employed to demonstrate that heme iron, nonheme iron, and two partial activities of N. crassa nitrite reductase, namely the NADPH-diaphorase and the dithionite-nitrite reductase activities, were associated with this nitrite reductase protein band.
Electrophoresis of homogeneous nitrite reductase under dissociating conditions revealed a single protein band of M, = 140,000. Denaturation of the enz-yme in the presence of 8 M guanidine hydrochloride followed by alkylation and SDSpolyacrylamide gel electrophoresis also resulted in a single protein band, indicating that the smallest dissociable unit of nitrite reductase has a molecular weight of 140,000. Results of two-dimensional polyacrylamide gel electrophoresis indicate that the subunits are identical in isoelectric point as well as molecular weight. Separation of subunits in either the isoelectric focusing or nonequilibrium pH gradient system yielded a single protein spot on the second dimension SDS-polyacrylamide gel. The M, = 140,000 subunit has an isoelectric point at pH 6.8-6.9.
Two-dimensional thin layer chromatography of dansylated nitrite reductase revealed a single fluorescent-labeled NH2terminal amino acid corresponding to glycine. Nitrite reductase apparently possesses only one type of reactive NH2 terminus. Since the nitrite reductase subunits also display identical isoelectric points and molecular weights, it is concluded that N. crassa nitrite reductase is a homodimeric protein. In addition, a single N . crassa locus, the nit-6 gene, has been designated as the structural gene encoding the nitrite reductase apoprotein (47).
Atomic absorption spectroscopy of nitrite reductase purified to homogeneity yielded an average of 10.2 Fe eq/mol of nitrite reductase. Colorimetric determinations of total iron with ophenanthroline gave an average value of 9 Fe/mol of nitrite reductase. Since it is assumed that each subunit of nitrite reductase binds a single siroheme, presumably 2 iron eq correspond to the heme iron of the siroheme moiety. This hypothesis is supported by the evidence that the sulfite reductases of E. coli and spinach (48) bind one siroheme/hemoprotein subunit.

Neurospora crassa NAD(P) H-Nitrite Reductase
Estimation of the acid-labile sulfide content of N. crassa nitrite reductase by the methylene blue formation assay shows 8.1 S"/mol of enzyme. Thus the remainder of the iron as determined above is apparently associated with sulfur as nonheme iron-sulfur centers which may be present as four Fe2S2* or two Fe4S4* centers. It is likely that the iron is organized as tetranuclear centers in nitrite reductase since other 6-electron transferring enzymes such as E. coli sulfite reductase (48), spinach nitrite reductase (49), and nitrogenase (50) contain Fe4S4* centers. The sulfhydryl reagents, 5,5"dithiobis(2-nitrobenzoic acid) and 4,4'-dithiodlpyridine, each reacted with 50 sulfhydryl eq/ mol of nitrite reductase. Amino acid analysis of carboxymethylated nitrite reductase shows 43 carboxymethylcysteinyl residues. The difference of 7 sulfhydryl eq confirms the presence of noncysteine sulfur associated with the nonheme iron-sulfur centers. In addition, titration of total uersus available sulfhydryl groups indicated that a class of relatively inaccessible or slow reacting sulfhydryl equivalents exists. An interesting speculation is that the 14-16 groups which consistently required the addition of 8 M urea and high concentrations of reagents for complete reaction may be the sulfide equivalents of two Fe4S4* centers and 8 cysteinyl sulfurs which are necessary to bind the centers to the protein.
In spite of the absence of FAD in highly purified prepara- and siroheme (18). In addition, the K,,, for FAD is 0.02 p~; the constants for the substrates of the reaction, NADPH and nitrite, are 400-to 800-fold greater.
Comparison of assimilatory nitrite reductase of heterotrophic organisms, where this nitrite reductase from N . crassa serves as the prototype, with the nitrite reductase of photosynthetic organisms, the spinach enzyme being the best characterized example (49, and references therein), suggests a similar functional organization to achieve their common physiological role of nitrite reduction. Both contain nonheme ironsulfur centers and siroheme as prosthetic groups. Both are composed of a single polypeptide type. However, the spinach nitrite reductase is a M, = 61,000 monomeric protein. Thus the enzymes differ markedly in their size. They also differ dramatically in their electron donor specificity. Spinach nitrite reductase employs photosynthetically reduced ferredoxin to reduce nitrite in a flavin-independent process whereas N . crassa nitrite reductase utilizes reduced pyridine nucleotides in an FAD-dependent reaction.
Like the spinach nitrite reductase, nitrite reductases of algae and nitrate reductases from blue-green algae and photosynthetic bacteria are relatively small, monomeric proteins which utilize reduced ferredoxin as electron donor (12). The multielectron transferring proteins which utilize reduced pyridine nucleotides as elect,ron donors are generally of high molecular weight and require the presence of a flavoprotein diaphorase moiety for electron transfer. For example, the flavoprotein subunit (Mp = 54,000) of E. coli sulfite reductase (M, = 700,000) binds FAD and FMN prosthetic groups and is the initial portion of the electron transfer apparatus required for the NADPH-dependent, 6-electron reduction of sulfite to sulfide (51). The complexity and increased molecular weight of the pyridine nucleotide-dependent proteins is perhaps a reflection of this requirement for a flavoprotein to accept electrons from NAD(P)H. Multielectron transfers are cer-tainly possible in the absence of flavin, i.e. the ferredoxindependent 6-electron reduction of nitrite catalyzed by spinach nitrite reductase and the dithionite-dependent 6-electron reduction of nitrite and sulfite mediated by N. crassa nitrite reductase and E. coli sulfite reductase, respectively. However, successful utilization of reducing power of the physiological electron donor NAD(P)H by N . crassa nitrite reductase and E. coli sulfite reductase requires the participation of a flavincontaining diaphorase function. In light of the absolute requirement of N. crassa nitrite reductase for FAD in its NAD(P)H-dependent activities, FAD is assigned as a cofactor of the enzyme.
The participation of sulfhydryl groups in the catalysis mediated by nitrite reductase is still undetermined. In contrast to the sulfhydryls of N . crassa nitrate reductase which participate in the proposed sulfhydryl-disulfide transition in the flow of electrons from NADPH to FAD (52), functional sulfhydryls of nitrite reductase are implicated by inhibitor studies, e.g., p-hydroxymercuribenzoate, but their role has not been definitively established. The sequence of electron transfer mediated by nitrite reductase is best described presently as: Although such electron transfer sequences are illustrative, they cannot account for the mechanistic complexities implicit in the transfer of 6 electrons. Siegel (48) states that the minimum catalytic unit for sulfite and nitrite reduction is very simple: one Fe4S4* center and one siroheme moiety. Such is the case for E. coli sulfite reductase hemoprotein and spinach nitrite reductase and the role of each member of the catalytic unit has been examined by spectrophotometric and EPR studies. Rueger and Siegel (53) showed that spectral alterations at siroheme absorption maxima are induced upon reaction of sulfite reductase with CO or CN-or with the substrate sulfite. 35SOa2was shown to bind and remain closely associated with the heme moiety during the course of its reduction. In addition, Siegel (48) demonstrated that the Fe4S4* center undergoes a modification in EPR signal and is rendered more readily reducible upon the formation of CO-or CN-siroheme complexes which suggests that there is an interaction between the heme and Fe. S center of sulfite reductase during sulfite turnover. Siroheme has been designated as the nitrite binding site for spinach nitrite reductase and a role for the Fe4S4* center in nitrite reduction has been proposed by Lancaster et al. (49). Recently, Christner et al. (54) demonstrated by EPR and Mossbauer spectroscopic analyses that the siroheme and Fe. S center of the E. coli sulfite reductase hemoprotein subunit are closely linked as a functional unit. A common bridging ligand may provide the linkage since the two chromophores appeared to share a single electronic spin and the electronic environments of 5 Fe atoms were altered upon the addition of a single electron to the oxidized enzyme.
Since the reduction of sulfite or nitrite involves the transfer of 6 electrons and the maximum electron storage capacity of either E. coli sulfite reductase hemoprotein subunit or spinach nitrite reductase is apparently only 2 electrons (1 in the Fe S center and 1 in siroheme), Siegel (48) concludes that intermediates of the reaction remain bound throughout the catalytic cycle. Indeed, like E. coli sulfite reductase, N. crassa nitrite reductase can catalyze the 6-electron reduction of nitrite utilizing dithionite as reductant. The iron-sulfur centers and siroheme are probably involved in this reaction in a manner similar to that described for E. coli sulfite reductase and spinach nitrite reductase. However, the presence of FAD prosthetic groups as well as sulfhydryl groups which may be catalytically active raises the electron storage capacity of N . crassa nitrite reductase.
The ability of nitrite reductase to catalyze a 6-electron reaction is one of the most interesting properties of the enzyme. Other proteins catalyzing multielectron transfers are typically heteromultimeric, including cytochrome oxidase, which catalyzes the 4-electron reduction of 0 2 to HzO, and the &electron transferring enzymes sulfite reductase and nitrogenase. Cytochrome oxidase (Mr = 171,300) is composed of Seven different subunits (55), E. coli sulfite reductase (Mr = 700,000) has the subunit structure asp4 (51), and nitrogenase (MoFe protein; M, = 220,000-245,000) has an a&-type subunit organization (50). The fact that N . crassa nitrite reductase is constituted as a homodimer of high molecular weight subunits which house a complex electron transfer apparatus places this enzyme in a unique position among the proteins which catalyze multielectron transfers.