Nitric-oxide Dioxygenase Function of Human Cytoglobin with Cellular Reductants and in Rat Hepatocytes*

Cytoglobin (Cygb) was investigated for its capacity to function as a NO dioxygenase (NOD) in vitro and in hepatocytes. Ascorbate and cytochrome b5 were found to support a high NOD activity. Cygb-NOD activity shows respective Km values for ascorbate, cytochrome b5, NO, and O2 of 0.25 mm, 0.3 μm, 40 nm, and ∼20 μm and achieves a kcat of 0.5 s−1. Ascorbate and cytochrome b5 reduce the oxidized Cygb-NOD intermediate with apparent second order rate constants of 1000 m−1 s−1 and 3 × 106 m−1 s−1, respectively. In rat hepatocytes engineered to express human Cygb, Cygb-NOD activity shows a similar kcat of 1.2 s−1, a Km(NO) of 40 nm, and a kcat/Km(NO) (k′NOD) value of 3 × 107 m−1 s−1, demonstrating the efficiency of catalysis. NO inhibits the activity at [NO]/[O2] ratios >1:500 and limits catalytic turnover. The activity is competitively inhibited by CO, is slowly inactivated by cyanide, and is distinct from the microsomal NOD activity. Cygb-NOD provides protection to the NO-sensitive aconitase. The results define the NOD function of Cygb and demonstrate roles for ascorbate and cytochrome b5 as reductants.

Measurements of the NO reactivities of the oxy-complexes of Ngb and Cygb in vitro (12,13) support a NOD function. However, O 2 transport-storage (14,15), peroxidase (16,17), oxidase (18), heterotrimeric G␣ protein guanine nucleotide dissociation inhibitor (19), and other functions (11,20,21) have also been postulated, and it remains unclear which of the proposed functions is physiologically significant and under what circumstances. For example, several NO metabolic activities and sinks have been measured in various cells and organelles (22)(23)(24) that would make a NOD activity appear either redundant or inconsequential. In addition, measurements of NO metabolism by globins with autooxidizable electron donors in the absence of SOD (13) do not allow the discernment of catalytic NO dioxygenation from NO oxidation by O 2 . .
Understanding globin function(s) continues to demand a synthesis of knowledge of structure, ligand affinities, reactivities, autooxidation rates, and electron donors. Moreover, to assess the NOD function of a globin, knowledge of its interactions with NO and O 2 within the cellular milieu is essential. Recently, Cygb-dependent NO dioxygenation and protection of NO-sensitive respiration within mouse NIH3T3 fibroblasts was reported (25), yet the relative capacity of Cygb for a NOD function with NO, O 2 , and cellular reductant(s), as well as its relation to other cellular NO sinks, remains to be defined.
Here we describe interactions of Cygb with NO, O 2 , and reductants that govern its capacity to function as a NOD in vitro and in rat hepatocytes. Investigations of the NOD activities of Cygb, Ngb, and Mb with ascorbate reveal a uniquely efficient redox coupling with Cygb that may be attributed to a unique Cygb structure. In addition, we report that the Cygb-NOD activity is physically and kinetically distinct from the microsomal NOD activity (23,26,27). copper-and zinc-containing SOD (5000 units/mg), and nitrate reductase (10 units/mg) were obtained from Roche Applied Science. DNA restriction and modifying enzymes were obtained from New England Biolabs, Inc. DNA primers, G418, and SeeBlue TM prestained protein molecular weight standards were purchased from Invitrogen. AG 1-X8 ion exchange resin (acetate form) was purchased from Bio-Rad. 99.993% O 2 , 99.999% CO, 99.998% N 2 , and 99.99% argon were from Praxair (Bethlehem, PA). Recombinant rat cytochrome b 5 (soluble form) was expressed in E. coli and purified essentially as described (28). Recombinant human NCB5OR (29) (23).
Expression and Purification of Cygb and Ngb-Human Cygb cDNA Image clone 5193583 (ATCC number 7498923) was obtained from the American Type Culture Collection (Manassas, VA). Cygb cDNA was PCR-amplified with the respective sense and antisense primers GGAGCTGCATATGGAGA-AAGTGCCAGGCGA and CCTCAAGCTTCCTTGGCA-CCCAGAAATGGA engineered with respective NdeI and HindIII sites, cloned into the pET17b vector (Novagen) polylinker, and sequenced. Cygb was expressed and isolated from E. coli BL21(DE3)pLysS. Mouse Ngb cDNA in pET3A was provided by Dr. Thorsten Burmester and Thomas Hankeln (Johannes Gutenberg University of Mainz, Mainz, Germany) and was expressed and isolated from E. coli BL21(DE3)pLysS (31). E. coli expressing isopropylthio-␤-D-galactoside-inducible globin were inoculated in 2 liters of Luria-Bertani medium (32) containing 50 g/ml ampicillin, 1 M hemin, and 5 units/ml catalase in 2.8-liter Fernbach flasks and grown in a 37°C gyrorotatory shaker at 175 rpm. Globin expression was induced with 0.1 mM isopropylthio-␤-D-galactoside when the cell density, as measured by absorbance at 550 nm, was 1.2. The cells were treated with isopropylthio-␤-D-galactoside for 16 h and were harvested by centrifugation. The cell pellets were resuspended and sonicated in a cell:buffer volume ratio of 2:3 in chilled 20 mM potassium phosphate buffer, pH 7.0, containing 1 mM EDTA, 10 g/ml deoxyribonuclease I, and 260 units/ml catalase. The cell lysates were clarified by centrifugation, and the extracts were dialyzed extensively against 1 mM potassium phosphate, pH 7.0. Cell-free extracts were diluted to 20 mg/ml protein, heated for 10 min at 65°C, chilled, and clarified by centrifugation. Cygb and Ngb were isolated by column chromatography on DEAE-Sepharose, Superdex 75, and hydroxyapatite. Protein was measured by the method of Lowry et al. (33) with bovine serum albumin as the standard. Protein purity was assessed by SDS-PAGE using Brilliant Blue G staining. Heme was assayed using the alkaline-pyridine method (34). The spectra were recorded with a Beckman DU 7500 diode array spectrophotometer.
Mammalian Cell Culture-The rat hepatocyte, K9 (CRL-1439), the human lung adenocarcinoma A549 (CCL185), and the human colorectal carcinoma Caco-2 (HTB-37) were obtained from the American Type Culture Collection. The cells were grown, passaged, and harvested as previously described (27). The cells were counted with a hemacytometer, and the weights of the cell pellets were measured with an analytical balance.
Western Blot Analysis-The cells were lysed with an equal volume of detergent buffer containing 20 mM Tris-Cl, pH 7.4, 50 mM NaCl, 1 mM EDTA, and 1% Triton X-100. The extract protein was assayed (33), separated by reducing SDS-PAGE in 1.0-mm Precast 8 -16% gradient gels (Invitrogen), and transferred to nitrocellulose membranes. The membranes were washed, blocked with 5% milk, probed with rabbit anti-Cygb antibodies (15,35), and detected using peroxidase-conjugated goat anti-rabbit IgG (Pierce) and the SuperSignal West Dura extended duration substrate kit (Thermo Scientific) on Kodak X-Omat TM film. Cygb was measured by Western analysis and densitometry with purified human Cygb standards.
Construction of Cygb-expressing Hepatocytes-The 995-bp EcoRI-XbaI fragment of cytoglobin cDNA Image clone 5193583 was isolated and ligated into the EcoRI-XbaI-restricted polylinker region of pcDNA3 v. 1.1 (Invitrogen) generating pcDNA3-hCygb. Rat hepatocytes were transfected with pcDNA3-hCygb or pcDNA3 DNA using FuGENE TM 6 transfection reagent (Roche Applied Science), and stable transfectants were selected with 0.3 mg/ml G418. The clones were isolated and screened for Cygb expression by Western blot analysis.
NO Consumption Assays-NO consumption was measured with ISO-NOP and microchip NO electrodes (WPI Instruments, Inc.) as previously described (34). NO scavenging activities of globins were routinely measured at 37°C in 2 ml of 100 mM sodium phosphate buffer containing 0.3 mM EDTA and 1.0 mg of MnSOD (11 M of the dimer). MnSOD was routinely added to NO consumption assays to competitively scavenge interfering O 2 . generated by reductants and reductases (34).
Copper-and zinc-containing SOD reacted with NO and was thus unsuitable. Cell NO metabolism was measured at 37°C in Dulbecco's phosphate-buffered saline containing 5 mM glucose and 100 g/ml cycloheximide (27,36). NO consumption measurements were corrected for background rates. NO-saturated water (1.94 mM) was prepared over AG 1-X8 resin as previously described (34). CO-saturated water (1.0 mM) was prepared under 99.5% CO (34). O 2 -saturated buffer (1.14 mM) was prepared, and O 2 concentrations were varied in reactions as previously described (27,34,36 Nitrate and Nitrite Assays-Nitrate was reduced to nitrite with NADPH and nitrate reductase. Nitrite was assayed using the Griess reagent (34).
Globin Reduction Assays-Reduction of ferric Cygb, Ngb, and Mb was measured by following the formation of the ferrous-CO complex at 422, 416, and 422 nm, respectively, in a 1-cm thermostatted quartz cuvette at 37°C. The reactions were initiated by adding 10 M of globin (heme) to an anaerobic 1-ml reaction mix containing buffer, 20 M CO, and the indicated concentrations of ascorbate. O 2 was depleted from reactions by scrubbing the reaction mix with N 2 and reacting 5 mM glucose with 2 units of glucose oxidase and 260 units of catalase for 10 min prior to adding 4 -8 l of the concentrated globin.
Cell NO Exposures and Aconitase Assays-Freshly harvested K9neo and K9Cygb hepatocytes were resuspended at a density of 3.2 ϫ 10 6 cells/ml in 3-ml of serum-free F12K medium buffered to pH 7.4 with 50 mM sodium HEPES and containing 100 g/ml cycloheximide. O 2 , N 2 , and NO gas mixtures were delivered at 30 ml/min using three-way gas proportioners (27). The cells were harvested, and the extracts were assayed for aconitase activity and protein as previously described (27).
Data Analysis-The Tukey-Kramer (Honestly Significant Difference) statistical analysis method in the program JMP (SAS Institutes Inc.) was used for the analysis of significance (p Ͻ 0.05). The data presented are representative of the results of two or more trials.

Ascorbate-driven NO Metabolic Activity of Cygb, Ngb, and
Mb-Ascorbate, a potential electron donor for plant hemoglobins (37,38) and Mb (39), was investigated for its ability to support enzymic NO scavenging by Cygb, Ngb, and Mb. With 10 mM ascorbate, the enzymic NO scavenging activity is linearly dependent upon globin concentration. Human Cygb shows a turnover rate of 0.25 NO/heme/s with 100 nM NO at 37°C (Fig.  1A, F). An ϳ4-fold slower turnover is seen at 20°C (Fig. 1A, f). Mouse Ngb (Fig. 1B, F) and sperm whale Mb (Fig. 1C, F) show respectively 150-and 70-fold lower activities than Cygb.
The Cygb activity shows an apparent K m (ascorbate) value of 2 mM ( Fig. 2A), an apparent K m for NO of 40 nM, and k cat of 0.5 s Ϫ1 with saturating ascorbate and O 2 in neutral phosphate buffer (Fig. 2B). With 100 nM NO, half-maximal activity is seen at ϳ20 M O 2 (Fig. 2C). Inhibition of the activity occurs at [NO]/[O 2 ] ratios Ͼ1:500 (see below), thus complicating the determination of a K m (O 2 ) value. Similar to other NOD activities (3,23,27,40), the activity is inhibited by CO and inactivated cyanide. With 20 M O 2 , 10 M CO inhibits the Cygb activity by ϳ50% (Fig. 2D). NO progressively inhibits the activity during turnover with 400 nM NO and 200 M O 2 (Fig. 2E, F), and the activity is slowly inactivated by 250 M cyanide (Fig. 2E, E). Preincubation of Cygb with cyanide for 7 min does not cause a greater loss of activity (Fig. 2E, compare E and ‚), thus demonstrating the importance of turnover for cyanide-mediated inactivation.
The results demonstrate K m values for NO and O 2 within physiologically relevant concentration ranges. However, the K m (ascorbate) value of 2 mM is greater than the K m values of 0.4 -0.9 mM reported for ascorbate-utilizing enzymes (41,42), and the ascorbate concentrations measured in some tissues (43).
Anions Competitively Inhibit Activity-Salts were tested for effects on the K m (ascorbate) value and steady-state behavior of Cygb. Lowering the buffer sodium phosphate concentration from 100 to 10 mM decreases the K m (ascorbate) value to 0.25 mM (Fig.  3A, line 1). Salts, including potassium chloride (lines 2-4) or sodium chloride (line 5), competitively inhibit the activity with respect to [ascorbate]. Increasing the sodium phosphate concentration or including sodium phosphate or NaCl in a 25 mM sodium citrate buffer also competitively inhibits the activity and increases the K m (ascorbate) value (Fig. 3B). Furthermore, the competitive effect of buffer salt depends more on the anion than the ionic strength. Salts do not significantly affect the k cat achieved with saturating ascorbate (Fig. 3A) or the K m (NO) (data not shown). The results suggest electrostatic interactions in the binding of the negatively charged ascorbate anion.
Cygb Reduction by Ascorbate-The rates of ferric globin reduction were measured under conditions similar to those for ascorbate-driven NO metabolism ( Fig. 2A). Globin reduction was measured in the presence of excess CO and the absence of  O 2 to stabilize the reduced heme from reoxidation. The rate of Fe 2ϩ (CO) complex formation was measured by following the absorbance increase at 422 nm as described under "Materials and Methods." Under these conditions, the Cygb reduction rate shows a linear dependence upon [ascorbate] up to 8 mM (data not shown). The apparent second order rate constant for ascorbate-mediated reduction of ferric Cygb is 1.
The corresponding rate constants for Ngb and Mb are 0.10 and 0.13 M Ϫ1 s Ϫ1 , respectively. The results demonstrate an ϳ10fold faster reduction rate for Cygb than for Ngb and Mb. However, these differences do not account for the larger differences in turnover rates measured in Fig. 1.
Cygb NO Metabolic Activity Produces Nitrate-Slow and gradual addition of 32 nmol of NO-saturated water (16-l) with a gas tight syringe to a 2-ml reaction mix containing 50 mM Tris-Cl, pH 7.5, 2 mM ascorbate, 200 M O 2 , 11 M MnSOD, and 1.0 M Cygb at 37°C yields 28.7 Ϯ 1.0 nmol of NO 3 Ϫ and 2.8 Ϯ 0.5 nmol of NO 2 Ϫ or 91 Ϯ 3% NO 3 Ϫ and 9 Ϯ 2% NO 2 Ϫ . The results demonstrate a nitrate-generating NOD mechanism for Cygb (4,5). However, the data also indicate secondary NO oxidation reactions, possibly representing residual reactions of O 2 . with NO to form peroxynitrite and NO 2 Ϫ (44). Cygb-NOD Activities with NADH, NADPH, cytochrome b 5 , and Cellular Reductases-Several cellular reductants and reductases with globin reducing capacity (5,13,45,46) were tested for their ability to support the NOD activity of Cygb in the presence of a high [SOD].
NADH and NADPH support a low Cygb-NOD turnover that is Ͼ15-fold slower than with ascorbate under otherwise comparable conditions (Table 1), demonstrating a preference for ascorbate. Reduced cytochrome b 5 also supports the NOD activity of Cygb and shows saturation with half-maximal activity at 0.3 M (Fig. 4A). Ferredoxin reductase and NADPH produce a low background activity in the cytochrome b 5 reducing system. CYPOR can also support a low activity of Cygb-NOD (Fig. 4B). NCB5OR, an enzyme with a flavin-containing reductase domain and a cytochrome b 5 domain that is found in the lumen of the endoplasmic reticulum (29), also supports the NO metabolic activities of Cygb (Fig. 4C), Ngb (Fig. 4D, f), and Mb (‚). Cygb shows the highest activity of the three (Fig. 4D, compare F with f and ‚). Microsomes containing membranebound CYPOR (14 milliunits) and incubated with 100 M NADPH can supply electrons for the NO metabolic activity of Cygb. With 300 nM Cygb, microsomes show greater activity than that observed with NADPH alone and catalyzed by the microsomal NOD activity (Fig. 4E) (23).
The results demonstrate that, in addition to ascorbate, cytochrome b 5 and NCB5OR effectively support Cygb-NOD activity. We can estimate apparent second order rate constants for cytochrome b 5 and NCB5OR-mediated reduction of Cygb of 3 ϫ 10 6 and 6 ϫ 10 6 M Ϫ1 s Ϫ1 , respectively. None of the electron donors is as effective in supporting the NO scavenging activity of either Ngb or Mb, thus suggesting electron donor specificity for Cygb.    (Fig. 5A). Human Cygb migrates as a monomer (20.9 kDa) on reducing SDS-PAGE gels. A weak signal from the disulfide-linked dimeric Cygb is detected with Ͼ25 ng of purified Cygb/lane after long film exposures. The results are consistent with immunohistochemical assays of Cygb in rat hepatocytes (15,16).
Rat hepatocytes offered a cell model for measuring the capacity of Cygb to function as a NO scavenger. Rat hepatocytes express negligible Cygb and relatively low levels of NO metabolic activity, synthesize ascorbate (47,48), and contain microsomal detoxification enzymes including CYPOR and cytochrome b 5 . Transfection of hepatocytes with pcDNA3-hCygb expressing the human Cygb gene under control of the cytomegalovirus promoter produced several stable cell lines with elevated Cygb expression (data not shown). A representative clone, K9Cygb, was selected for further investigation. K9Cygb expressed 50 Ϯ 10 ng of Cygb/100 g of soluble protein, whereas the control cell line, K9neo, which was stably transfected with the pcDNA3 vector, showed no detectable Cygb protein (Fig. 5B). If we assume that ϳ90% of the cell weight is water, we can estimate that K9Cygb hepatocytes express 1.  (Fig. 6A). With 10 nM NO, the activity difference is larger. The Cygb-NOD activity is ϳ13fold greater for 200 M O 2 and ϳ9-fold higher at a more physiological O 2 concentration of 10 M (Fig. 6B).
The NO metabolic activity shows an apparent K m (NO) value of 40 nM with 200 M O 2 (Fig. 7A). Half-maximal activity is observed with Յ5 M O 2 with 20 nM NO (Fig. 7B, F). NO inhibits the activity at higher NO:O 2 concentration ratios precluding the determination of a true K m (O 2 ) value. CO also inhibits the activity, and inhibition is also less pronounced with higher O 2 concentrations (Fig. 7B, E). Similar to in vitro reactions (Fig. 2E), the NO metabolic activity is weakly yet progressively inactivated by 250 M cyanide during turnover (Fig. 7C, compare E with the control F). In contrast, the basal NO metabolic activity in K9neo cells is rapidly inactivated by 250 M   cyanide (Fig. 7C, compare Ⅺ with the control f). NO metabolism by K9Cygb cells is also insensitive to the CYPOR and flavoenzyme inhibitor, DPI. DPI (50 M) inactivates Ͻ10% of the total activity (Fig. 7D, compare E with the control F). The results demonstrate the capacity of Cygb to function as a NOD. Further, the data show that the activity is distinct from the cyanide and DPI-sensitive microsomal NOD (23,27). The results also suggest that the cyanide-sensitive NO metabolic activity previously attributed to Cygb in NIH3T3 fibroblasts (25) is due to the ubiquitous microsomal NOD.
Sensitivity of Aconitase to NO-mediated Inactivation in Cygbexpressing Hepatocytes-Exposure of K9neo and K9Cygb hepatocytes to an atmosphere containing 480 ppm of NO balanced with 21% O 2 and N 2 for 60 min inactivates the NO-sensitive aconitase (27,32) to 61 Ϯ 9% (Ϯ S.D., n ϭ 4) and 75 Ϯ 5% (Ϯ S.D., n ϭ 4) of its control activity levels, respectively. The 14% difference between the values is significant (p Ͻ 0.05). In the absence of cells and under similar exposure conditions, NO rapidly achieves a concentration of 400 Ϯ 50 nM. Exposure of either cell line to 240 ppm NO did not result in significant inactivation of aconitase, suggesting effective protection by the basal NO metabolic activity under these conditions. Aconitase activities in control K9neo and K9Cygb cell extracts were 11.5 Ϯ 1.0 and 11.3 Ϯ 1.5 milliunits/mg (Ϯ S.D.; n ϭ 5), respectively. The results demonstrate Cygb-dependent protection of aconitase in rat hepatocytes.

DISCUSSION
Like other globins (4,5), Cygb, Ngb, and Mb can function as enzymic NODs when coupled to suitable electron donors. With ascorbate and reduced cytochrome b 5 , the human Cygb-NOD shows a maximal turnover rate of ϳ0.5 s Ϫ1 . With a net NOD activity of 2 nmol NO/min/10 7 cells at 200 nM NO and 200 M O 2 (Fig. 6A) and expression of 550 ng of Cygb/1.10 Ϯ 0.21 mg of soluble protein (Ϯ S.D., n ϭ 3) or 10 7 cells (Fig. 5B), we can estimate a similar turnover rate of 1.2 s Ϫ1 within engineered K9Cygb hepatocytes. We can also calculate a k cat /K m (NO) (kЈ NOD ) value of 1.3-3 ϫ 10 7 M Ϫ1 s Ϫ1 for the Cygb-catalyzed NO dioxygenation reaction, which is in range of the 3.4 to ϳ7 ϫ 10 7 M Ϫ1 s Ϫ1 measured for the structurally similar Mb and Ngb using stopped flow analysis (12,30) and the ϳ2.2 ϫ 10 7 M Ϫ1 s Ϫ1 estimated from NO electrode measurements of the reaction of NO with oxy-Cygb (13). Hence, electron transfer to Cygb does not significantly limit enzymatic activity in vitro or in rat hepatocytes.
With saturating ascorbate or reduced cytochrome b 5 , the Cygb-NOD activity is limited by its kЈ NOD and susceptibility to NO inhibition. These are determined, respectively, by NO diffusion and reactivity with the active site oxy-complex (kЈ NOD ) and the competition between O 2 and NO for binding to the ferrous heme (K d (O 2 )/K d (NO)) (5). NOD activity is ultimately determined by globin structure. By comparison, flavohemoglobin-NOD is Ͼ5-fold less sensitive to NO inhibition with a relatively small K d (O 2 )/K d (NO), reacts with NO up to 100-fold faster with a relatively large kЈ NOD Յ 2.9 ϫ 10 9 M Ϫ1 s Ϫ1 , and shows Ͼ100-fold faster turnover rates (5,49,50).
By analogy to the peroxidases and peroxiredoxins and the catalases that scavenge H 2 O 2 in cells (51), individual NODs may be suited for scavenging NO under different conditions. Cygb-NOD is clearly distinct from the microsomal CYPOR-dependent NOD activity. Caco-2, A549 cells, and hepatocytes express negligible Cygb (Fig. 5A) but express a DPI and cyanidesensitive microsomal activity that shows only weak NO inhibition at an [NO]/[O 2 ] ratio of 1:100 (23,27). In contrast, the NO metabolic activity of Cygb in hepatocytes is DPI-resistant, slowly inactivated by cyanide (Fig. 7C), and inhibited by NO at [NO]/[O 2 ] ratios of Ͼ1:500 (Fig. 2E).
Our results demonstrate that ascorbate and cytochrome b 5 both act as efficient electron donors for the Cygb-NOD activity and, moreover, suggest that the preference for electron donors in cells will be dependent upon their relative abundance. The low K m (ascorbate) value of 0.25-0.67 mM measured at physiological salt concentrations (Fig. 3B) argues for a role for ascorbate in ascorbate-utilizing collagen-synthesizing hepatic stellate cells and fibroblasts and ascorbate-rich neurons (15,16). We have not measured ascorbate or cytochrome b 5 levels in cultured rat (K9) hepatocytes. However, hepatocytes synthesize ascorbate and are a major source of ascorbate in rats (43,48). Ascorbate concentrations of 6 nmol/mg of protein (ϳ1.2 mM) and 2 mM have been measured in cultured primary rat hepatocytes (47,48).
The K m (ascorbate) and k cat values for the Cygb-NOD activity (Figs. 2 and 3) allow us to estimate apparent second order rate constants for Cygb reduction ranging from 125 to 1000 M Ϫ1 s Ϫ1 and dependent upon anion concentrations, whereas the ferric Cygb was reduced by ascorbate with an ϳ100-fold smaller rate constant under similar buffer conditions. The disparate rate constants indicate that ferric Cygb, per se, is not an obligate intermediate in the catalytic cycle. The rate of Cygb reduction may be greater for the high potential Fe 3ϩ ( Ϫ OONO) intermediate (4, 5) like the fast ascorbate-mediated reduction of the ferryl form of leghemoglobin (52) and Compound I or II in ascorbate peroxidase (53)(54)(55). The high reduction rate constants estimated for cytochrome b 5 and NCB5OR may be similarly explained. The steady-state kinetic results suggest caution when evaluating roles of potential electron donors based solely on transient measurements of ferric globin reduction (12,13,45,56,57).
The Michaelis-Menten behavior of the Cygb-NOD activity with ascorbate and the competitive inhibition by anions ( Figs.  2A and 3) strongly suggest a positively charged binding site for ascorbate. Cygb has unique invariant ArgE10 -84 and LysFG2-116 residues near a solvent-exposed heme proprionate carboxylate (58, 59) that may form a binding site similar to that in ascorbate peroxidase (60). In the peroxidase-ascorbate complex (Protein Data Bank code 1OAF) (Fig. 8, top panel), Arg-172 facilitates ferric heme reduction by hydrogen bonding and stabilizing ascorbate ene-diol(ate) (53,55,60), and a Lys-30 amine hydrogen bonds the ascorbate 6-hydroxyl. In the dimeric Cygb structure (Protein Data Bank code 1UMO) (58), the heme proprionate, ArgE10 and LysFG2 residues are located in different positions in the A and B subunits (Fig. 8, compare bottom left  and bottom right panels), indicating a large conformational shift, and a possible inducible ascorbate-binding site. Moreover, the A and B subunit conformations are associated with changes between hexa-and pentacoordinate iron (58), imply-ing communication between the iron and the putative ascorbate-binding site during catalysis. It is noteworthy that the B subunit carboxylate O-atom and LysFG2 N-atom are separated by 10.7 Å (Fig. 8, bottom right panel) as seen in the ascorbate peroxidase-ascorbate complex (Fig. 8, top panel). In the monomeric Cygb structure (Protein Data Bank code 1V5H) (59), the ArgE10 guanidinium nitrogen is 6.6 Å from the carboxylate O-atom and 8.5 Å from the LysFG2 N-atom, the latter two being only 2.9 Å from each other, further illustrating the large structural flexibility of the site.
Elevated Cygb expression in activated hepatic stellate cells, chondroblasts, osteoblasts, and hypoxic neurons (15,16,21,61) suggests an important role for Cygb in the protection of these cells against NO toxicity; however, quantitative knowledge of Cygb expression levels and the capacity for NO detoxification has been lacking. We have determined that ϳ1.3 M Cygb confers a NOD activity equal to ϳ1. 6 (44) and protect respiration (25) and enzymes important to fibrogenesis such as the prolyl 4-hydroxylase (61,62). Cygb could also regulate the nanomolar [NO] found in tissues (63) and feedback regulate O 2 delivery to tissues (8,64). The large effects of [O 2 ] on activity (Fig. 7B) may also explain the benefit of hypoxic induction of Cygb (15,65). Experiments can now be directed toward determining the roles of an ascorbate and cytochrome b 5 -driven Cygb-NOD activity during fibrosis (16) and other conditions and elucidating the electron transfer mechanisms for ascorbate and cytochrome b 5 .