Enzymatic and nonenzymatic mechanisms for ferric leghemoglobin reduction in legume root nodules (flavins/ferric leghemoglobin reductase/nitrogen fixadtion/physiological reductants)

Evidence is presented for the operation in nodules of at least four systems for restoring functional ferrous leghemoglobin (Lb2+) from its inactive, ferric form. (i) Reduction offerric leghemoglobin (Lb3+) by a reductase. The enzyme is a flavoprotein of 100 kDa with two equally sized subunits and exhibits a K. of 9 ,uM for soybean Lb3+ component a and a K. of 51 ,IM for NADH. NADPH is only 30% (initial velocities) as effective as NADH. Lb3+ reductase converts 215 nmol of Lb3+to Lb2+.CO (or Lb2+0O2) per mg of protein per min and does not require an exogenous electron carrier. The enzyme shows similar affinity for soybean, bean, and cowpea Lb3+, but different V,. values. The reductase is inactive when Lb3+ is bound to nicotinate or N02-. (ii) Direct reduction of Lb3+ by NAD(P)H, ascorbate, and cysteine. Reduction by NAD(P)H is greatly stimulated by trace amounts of metals such as Mn2+. (iu) Reduction of Lb3+ by the flow of electrons from NAD(P)H to free flavins to Lb3+. The reaction does not occur via 0° or H202, and thus NAD(P)H-reduced flavins can directly reduce Lb3+. The efficiency of the reaction follows the order riboflavin > FMN > FAD. (iv) Reduction of Lb3+ by an unknown compound, B, of nodules. B has a molecular mass < 1 kDa and is heat-stable. The reaction mediated by B differs from those mediated by flavins and metals in several ways, requires NAD(P)H, and generates 02. Only the ferrous forms of hemoglobin (Hb), myoglobin (Mb), and leghemoglobin (Lb) bind 02. Oxidation of these hemoproteins to the ferric form is readily observed in vitro, but the proportions of the ferric forms are remarkably low in vivo. For example, erythrocytes from reptiles and mammals have a steady-state level of 1-3% Hb3+ (1). In human erythrocytes enzymatic and nonenzymatic mechanisms exist for reducing Hb3+ to Hb2+ (2). The contribution of each system to Hb3+ reduction is estimated to be 67% NADH:Hb3+ reductase (also named NADH:cytochrome b5 reductase), 5% NADPH:flavin reductase, 16% ascorbate, and 12% reduced glutathione (2). Analogous systems may exist in skeletal muscles for the reduction of Mb3+ (3, 4). In leguminous nodules a steady-state level of Lb3+ is also believed to result from the autoxidation of Lb2+-O2, which is favored by low pH values (5). Several nodule metabolites, such as O2, NO2-, and H202, may contribute to the oxidation of Lb2+ and Lb2+-O2 (6). Detection of Lb3+ in intact or minimally disturbed nodules is difficult due to the inherent light scattering by nodules, the low extinction coefficient of the diagnostic absorption band of Lb3+ at -625 nm, and the existence in nodules of several ligands, such as nicotinate (7), whose complexes with Lb3` do not exhibit the 625-nm band. The observation that chemically generated Lb3+ is rapidly reduced in soybean nodule slices suggests that nodules are equipped with mechanisms for restoring functional Lb2+ (8). Proteins with Lb3' reductase (FLbR) activity were reported in lupin (9) and soybean (10, 11) nodules. Lupin FLbR is very similar to cytochrome b5 reductase from erythrocytes (9). Puppo et al. (10) partially purified an FLbR-like enzyme from soybean nodules, but their preparation showed very low activity and this was not corrected for nonenzymatic Lb3+ reduction. Saari and Klucas (11) also purified a FLbR from soybean nodules that was shown to be a homodimer of 100 kDa and, therefore, unlike lupin FLbR. They also reported the existence of small, thermostable molecules in nodules that reduced Lb3+ upon addition of NADH and interfered with the purification ofFLbR (11). The identification of these compounds was not attempted and their efficacy for reducing Lb31 was not compared with that of FLbR. In this paper we describe several mechanisms for the reduction of Lb31 to Lb2+ that may be functional in legume nodules: (i) a specific enzyme (FLbR), (ii) endogenous reductants, (iii) NAD(P)H-reduced flavins, and (iv) a nonflavin unknown compound that also requires NAD(P)H for activity. MATERIALS AND METHODS Materials. Equipment for FPLC (fast protein liquid chromatography) and HPLC were purchased from Pharmacia and Waters, respectively. Reagents and chemicals were obtained as follows: hydroxylapatite (Bio-Gel HPT), Bio-Gel P-6DG, and protein assay reagent from Bio-Rad; Sephadex G-25 from Pharmacia; DEAE-cellulose (DE-52) from Whatman; ammonium sulfate and chemicals for HPLC from Baker; and sodium amobarbital from Lilly (Indianapolis, IN). All other chemicals were from Sigma. Mega-Pure (Coming) water was used throughout the study. Bacterial and Plant Culture. Rhizobium leguminosarum bv. phaseoli 3622, Bradyrhizobium japonicum 3I1b110, and Bradyrhizobium spp. (Vigna) BR7301 were used to elicit root nodules on seedlings of bean (Phaseolus vulgaris L. cv. Canadian Wonder), soybean [Glycine max (L.) Merr. cv. Hobbit], and cowpea [Vigna unguiculata (L.) Walp. cv. California Blackeye]. Bacteria and plants were grown as indicated previously (12) except that the nutrient solution for plants had 15 mg of Sequestrene 330Fe (10o Fe; Ciba-Geigy) per liter instead of ferric citrate. Nodules were harvested from plants at the late vegetative growth stage: bean, 35-40 days; soybean, 30-35 days; and cowpea, 40 days. Purification of Lbs. All operations were conducted at 0-40C. Nodules (50 g or 100 g) were homogenized with an ice-cold Sorvall Omni-mixer (3 x 1 min; maximum setting) in 3 ml of 50 mM KP, (pH 7.0) per g and 0.3 g of polyvinylpolypyrrolidone per g. The homogenate was filtered through eight layers of cheesecloth and centrifuged at 30,000 x g for Abbreviations: DCIP, 2,6-dichloroindophenol; Lb, leghemoglobin; Lb21/3+, ferrous/ferc Lb; Lba, -b, -c, and -d, different Lb isoprteins or components from the same legume species; FLbR, Lb3+ reductase; SOD, superoxide dismutase. 7295 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 7296 Botany: Becana and Klucas 60 min. The clear supernatant was applied to a hydroxylapatite column (15 x 2.5 cm) previously equilibrated with 10 mM KP; (pH 7.0). Three protein fractions were eluted by sequentially washing the column with 2-3 column volumes of 50 mM, 200 mM, and 700mM KPj (pH 7.0), respectively. The fraction that was eluted with 50 mM KP1 buffer was free of FLbR and was used for further purification of Lbs by DE-52 chromatography with sodium acetate at pH 5.2 (soybean; ref. 13) or with Tris-HCl at pH 8.0 (bean and cowpea; G. Sarath, H. K. Jun, and F. Wagner, personal communication). Concentrations of Lb3+ (purified as above), whale skeletal muscle Mb3+ (Sigma), and bovine Hb3+ (Sigma) were determined by the pyridine-hemochrome method (13). Purification of Soybean FLbR. The enzyme was purified according to the procedure of Saari and Klucas (11), as revised by Ji (14). The entire purification process was carried out at 0-40C and required less than 4 days to complete. FLbR was purified from the 700 mM KP, fraction mentioned above by several steps of FPLC, involving anion-exchange, gelfiltration, and hydrophobic columns (14). Protein concentration was determined by the Bio-Rad microassay (Bio-Rad Bulletin), with bovine serum albumin as a standard. Assay of Diaphorase Activity. FLbR was routinely assayed during the purification by using its diaphorase activity as a convenient assay (15). One unit of activity was defined as the amount of enzyme that reduced 1 nmol of 2,6-dichloroindophenol (DCIP) per min (15). Assay of FLbR Activity. FLbR activity was measured by following the conversion of Lb3" to Lb2`CO at 562 nm. The reaction was linear for at least 15 min. The reaction mixture (1 ml) contained 60-85 mM KPj (CO-saturated) at pH 7.0, 50 AM Lb3+ (from soybean, bean, or cowpea), 3 ,ug of enzyme, and 700,M NADH to initiate the reaction. One unit ofFLbR activity was defined as the amount of enzyme that produced 1 nmol of Lb2+CO per min. All FLbR activities were corrected for nonenzymic reduction of Lb3" by NADH and were determined at 23 + 2°C with a Cary 219 (Varian) spectrophotometer. Kinetic Parameters of FLbR. A linear relationship between initial velocities (V0) and enzyme concentration was observed at least in the range of 1-8 ,g of protein. Km values of soybean FLbR for several Lbs and NADH were determined essentially as indicated above during the first 5-10 min. Values of Km and Vma, were calculated from EadieHofstee plots. Effect of Metabolites, Inhibitors, and Other Modulators on FLbR Activity. FLbR activity was assayed as described above except 50 ,uM soybean Lb isoform c (Lbc) was used, and the relevant compound at the concentrations indicated in Table 1 was added to the reaction solutions. All tested compounds were also assayed in the absence of enzyme to correct for nonenzymatic reduction of Lb3+ by the compound(s) alone. Reactions were followed by sequential scanning (450-650 nm) at 0 (100% Lb3+), 0.5, 1, and 4 hr. After the last scan, a few crystals of dithionite were added to the cuvette and CO was gently bubbled through the cuvette for a few seconds. A new scan was run after 1 min, which corresponded to 100% Lb2+.CO. Percentages of Lb2+.CO formed were then calculated with the 0%o and 100% values of A562. Although FLbR activity is similar with airand COsaturated buffers (11), CO was preferred for long incubations to avoid problems of Lb2` autoxidation. Lb3+-nicotinate and Lb3+-nitrite were produced by adding a few crystals of nicotinic acid and KNO2 just prior to the addition of the enzyme and NADH. Inhibitors were preincubated with the enzyme in buffer at 23°C for 1 hr. Flavin Content ofNodules. Free flavins were extracted from nodules essentially as described by Cerletti and Giordano (16), at 0C in the dark. Nodules (0.3 g) were extracted twice with ice-cold trichloroacetic acid, and to the pooled supernatants 2 M KPj (pH 7.0) was added to give a final pH of 6.1. Aliquots were stored at -700C until further analysis (2-10 days later) of flavins according to Light et al. (17). Extraction of Low Molecular Mass Compounds from Nodules. These were prepared either from the supernatant after the 55-85% ammonium sulfate fractionation (13) used for Lb purification (for experiment in Table 3) or directly from the nodule cytosol (for experiment in Table 4). Both the supernatant and the cytosol were filtered sequentially through YM10 (10-kDa nominal cutoff) and YM2 (1-kDa nominal cutoff) membranes (Amicon). Extinction Coefficients. The following E or As values (mM-1cm-1) were used for calculations. For diaphorase activity: DCIP (600 nm), 21 (15). For FLbR activity: soybean Lba2+.CO minus Lba3+ (562 nm), 8.26; bean Lba2+CO minus Lba3+ (562 nm), 5.87; cowpea Lbb2+-CO minus Lbb3+ (562 nm), 6.44. For ferric hemoprotein-reducing activity of small molecules: Lb2+'O2 minus Lb3+ (574 nm), 10.2 (11); Mb2+*O2 minus Mb3+ (581 nm), 11.9 (18); Hb2+-O2 minus Hb3+ (576 nm), 11.7 (19). For concentrations of pyridine nucleotides and flavins (20): NADH (340 nm), 6.22; NADPH (340 nm), 6.20; riboflavin (450 nm), 12.20; FMN (450 nm), 12.20; FAD (450 nm), 11.30. RESULTS AND DISCUSSION Purification of Lb Components. Soybean, bean, and cowpea Lbs were separated by anion-exchange chromatography (DE-52) into four, two, and three components, respectively. These components were named according to their order of elution. Bean and cowpea Lbs were extracted from fresh nodules and the relative abundances of the components were calculated from the peak areas. The proportions were as follows: bean Lba (87%) and Lbb (13%); cowpea Lba (2%), Lbb (83%), and Lbc (15%). FLbR: Purification, Molecular Mass, and Kinetic Characteristics. A protein that catalyzes the reduction of Lb3+ to Lb2+ using NADH has been purified to homogeneity from soybean nodules, asjudged by a single protein band of 55 kDa on silver-stained gels after SDS/PAGE. Likewise, the molecular mass of the native enzyme, determined by gel filtration on a Superose-12 column, was 100 kDa, in agreement with an earlier report (11). Small molecules that are present in the nodule cytosol and facilitate reduction of Lb3+ and DCIP precluded a reliable determination of specific activities at the initial stages of purification. The maximal specific activity achieved was 2000 nmol of DCIP reduced per mg of protein per min or 215 nmol of Lb3+ reduced per mg ofprotein per min. The kinetic parameters of soybean FLbR were calculated using soybean Lba as well as closely related Lbs from bean and cowpea to determine whether the enzyme can distinguish natural and extraneous substrates in terms of affinity and catalytic activity. The Km values of FLbR for Lb3+ from soybean, bean, and cowpea nodules were very similar, ranging from 8.8 to 13.3 ,M, but significant differences were found in the Vm. values for Lbs from different species. Relative to the Vmax with soybean Lba, V,n values were 42% and 45% for bean Lba and cowpea Lbb, respectively. The Km value forNADH was 51 ,uM with soybean Lba as the other substrate. If a steady-state proportion of 41% Lb3+ exists in soybean nodules, as occurs for Hb3+ in human erythrocytes (2), Lb3+ concentrations in nodules would be 10-30 ,uM. This range of values would be consistent with the postulated function of FLbR in vivo. FLbR: Effect of Physiological Reductants, Enzyme Inhibitors, and Lb Ligands. The effect of NAD(P)H and various inhibitors on the reaction catalyzed by FLbR was studied over 4-hr incubation periods (Table 1). FLbR can use NADPH instead of NADH as a reductant, but the activity was only 54% and 80% that using NADH, after 0.5 hr and 4 hr, respectively (Table 1). This is consistent with a previous Proc. Natl. Acad. Sci. USA 87 (1990) Proc. Natl. Acad. Sci. USA 87 (1990) 7297 Table 1. Effect of physiological reductants, inhibitors, and Lb ligands on FLbR activity of soybean nodules