Aromatic residues and neighboring Arg414 in the (6R)-5,6,7, 8-tetrahydro-L-biopterin binding site of full-length neuronal nitric-oxide synthase are crucial in catalysis and heme reduction with NADPH.

Nitric-oxide synthase (NOS) requires the cofactor, (6R)-5,6,7, 8-tetrahydrobiopterin (H4B), for catalytic activity. The crystal structures of NOSs indicate that H4B is surrounded by aromatic residues. We have mutated the conserved aromatic acids, Trp(676), Trp(678), Phe(691), His(692), and Tyr(706), together with the neighboring Arg(414) residue within the H4B binding region of full-length neuronal NOS. The W676L, W678L, and F691L mutants had no NO formation activity and had very low heme reduction rates (<0.02 min(-1)) with NADPH. Thus, it appears that Trp(676), Trp(678), and Phe(691) are important to retain the appropriate active site conformation for H4B/l-Arg binding and/or electron transfer to the heme from NADPH. The mutation of Tyr(706) to Leu and Phe decreased the activity down to 13 and 29%, respectively, of that of the wild type together with a dramatically increased EC(50) value for H4B (30-40-fold of wild type). The Tyr(706) phenol group interacts with the heme propionate and Arg(414) amine via hydrogen bonds. The mutation of Arg(414) to Leu and Glu resulted in the total loss of NO formation activity and of the heme reduction with NADPH. Thus, hydrogen bond networks consisting of the heme carboxylate, Tyr(706), and Arg(414) are crucial in stabilizing the appropriate conformation(s) of the heme active site for H4B/l-Arg binding and/or efficient electron transfer to occur.

references therein). In contrast to the well known cytochrome P-450/NADPH-cytochrome P-450-reductase fusion protein, cytochrome P-450BM3, however, all NOS isoforms require the formation of a homodimer and the presence of both Ca 2ϩ / calmodulin (CaM) and (6R)-5,6,7,8-tetrahydrobiopterin (H4B) for catalysis (8 -17). The homodimer formation and the CaM binding seem to be associated with intersubunit/intermolecular electron transfer from the reductase domain to the heme domain, which is an absolute requirement of the activation of molecular oxygen bound to the heme iron.
The intrinsic function of H4B in NOS catalysis is still unclear, although H4B binding affects the conformation of the heme domain (13-15, 17, 18), facilitates substrate binding (16) and also promotes subunit dimerization (9 -12, 19, 20). From previous experimental observations, however, it is likely that the most important function of H4B would be participation in the reaction as a redox-active cofactor (14,17,20,21,25,26). Bec et al. (25) claimed that H4B donates a second electron to the Fe(III)-O-O Ϫ complex generated during catalysis for the monooxidation of L-Arg and to facilitate the O-O bond cleavage. In support of this, Hurshman et al. (26) detected a pterin radical with EPR during catalytic turnover using the oxygenase domain of inducible NOS (iNOS).
The crystal structures of the dimeric oxygenase domains of iNOS and endothelial NOS (eNOS) indicate that H4B is bound to the heme distal site and at such a distance from bound L-Arg that it cannot participate directly in molecular oxygen activation (27)(28)(29)(30). The structures of dimeric eNOS and iNOS oxygenase domains revealed a novel zinc center at the bottom of the dimer interface and the interaction of the heme with the pterin ring located in the dimer interface (28 -30). The structures of the NOS dimers with H4B show that several conserved aromatic residues interact with the pterin ring in the NOS dimer interface (Fig. 1). Fig. 2 shows the structure of the dimer interface of nNOS based on the crystal structure of the eNOS dimer (28). The H4B ring is located perpendicularly to the heme plane and lies between Trp 678 of one subunit and Trp 676 , Phe 691 , and His 692 of the other subunit. Raman et al. (28) claimed that these aromatic residues could stabilize a free radical intermediate with aromatic stacking in the dimer interface, and thus a pterin cation radical may be directly involved in catalysis.
Therefore, from spectroscopic, chemical, and structural results, it seems most likely that H4B is involved in the catalysis of NO formation as a redox-active cofactor and/or a modulator of the reactivity of heme iron-oxo complex. Ghosh et al. (31) reported analysis of the H4B binding site by mutating conserved aromatic amino acids in the close proximity of H4B in the oxygenase domain of iNOS. They claimed that aromatic stacking interactions influence the dimeric structure of the oxygenase domain of iNOS, the heme environment, and NO synthesis but are not essential for H4B binding and NO formation activity.
In the present study, we have mutated the conserved aromatic amino acids Trp 676 , Trp 678 , Phe 691 , His 692 , and Tyr 706 of full-length nNOS and studied the effects on the spectroscopic character, dimer formation, catalytic activity, and electron transfer rate from NADPH to the heme. We also generated mutants at Arg 414 , which forms hydrogen bonds between Tyr 706 and Trp 678 . Mutation of these residues greatly reduced both the NO formation activity and the electron transfer rate from the reductase domain to the heme active site of nNOS, even if dimer dissociation caused by the mutations was not marked. We will discuss the role of these aromatic amino acids in catalysis and electron transfer in full-length nNOS.

EXPERIMENTAL PROCEDURES
Materials-A polymerase chain reaction kit for site-directed mutagenesis was obtained from Takara Shuzo (Tokyo, Japan). H4B was purchased from Schircks Laboratories (Jona, Switzerland). 2Ј,5Ј-ADP-Sepharose and CaM-Sepharose were products of Amersham Pharmacia Biotech. Other reagents were obtained from Sigma or Wako Pure Chemicals (Osaka, Japan).
Mutagenesis-The cDNA for rat nNOS was kindly gifted from Dr. S. H. Snyder (Johns Hopkins University School of Medicine). Sitedirected mutagenesis were performed with the polymerase chain reaction-based strategy using a kit from Takara Shuzo. cDNA fragments containing wild type and the desired mutations were cloned into NdeI and XbaI sites of a vector, pCWoriϩ, and transformed into Escherichia coli strain BL21(DE3) (32) in expressing most mutants and the wild type (32,33). However, we used a yeast expression system (34 -36) for obtaining the R414L mutant protein because of the instability of this mutant.
Preparation of Neuronal NOS-Full-length wild type and mutant nNOSs were purified using DEAE-TOYOPEARL, 2Ј,5Ј-ADP-Sepharose, and calmodulin-Sepharose column chromatographies as described previously (34 -36). The purified and concentrated enzyme was dialyzed against 50 mM Tris-HCl (pH 7.5) buffer containing 5 M H4B, 20 M DTT, 0.1 mM EDTA, and 10% glycerol. For the preparation of H4Bdeficient enzyme, E. coli cells expressing the wild type and mutant enzymes were sonicated in the H4B-deficient buffer, and the enzymes were futher purified in the H4B-deficient buffer. Because H4B was deficient in E. coli cells, the enzymes prepared without addition of exogeneous H4B should be H4B-free. All purified full-length nNOSs were more than 95% pure as judged by SDS-polyacrylamide gel electrophoresis stained with Coomassie Blue R-250. The concentration of nNOS was determined optically from the [CO-reduced] Ϫ [reduced] difference spectrum using ⌬⑀ 444 -467 nm ϭ 55 mM Ϫ1 cm Ϫ1 . This ⌬⑀ value was estimated by the pyridine hemochromogen method (34 -36) assuming that one heme is bound to one subunit of this enzyme.
Gel Filtration-To examine the dimer contents of the mutants, purified full-length NOSs were incubated with 1 mM L-Arg and 0.1 mM H4B overnight at 4°C and analyzed on a Superose 6 HR10/30 size exclusion chromatography column (Amersham Pharmacia Biotech), equilibrated with 50 mM Tris-HCl (pH 7.5) buffer containing 0.2 M NaCl, 0.1 mM EDTA, and 0.1 mM L-Arg, and connected to an fast protein liquid chromatography system (Amersham Pharmacia Biotech). The molecular masses of the protein peaks were estimated relative to the molecular mass of standard proteins: thyroglobulin (669 kDa), ferritin (440 kDa), catalase (232 kDa), aldolase (158 kDa), and albumin (67 kDa).
Enzyme Assay-The rate of NO formation was determined from the NO-mediated conversion of oxyhemoglobin to methemoglobin, monitored at 401 nm using a methemoglobin minus oxyhemoglobin extinction coefficient of 49 mM Ϫ1 cm Ϫ1 (2). The NADPH oxidation rate was determined spectrophotometrically as an absorbance decrease at 340 nm, using an extinction coefficient of 6.22 mM Ϫ1 cm Ϫ1 . Unless otherwise indicated, assays were carried out at 25°C in 50 mM Tris-HCl (pH 7.5) buffer containing 10 M oxyhemoglobin, 0.1 mM NADPH, 5 M each of FAD and FMN, 10 g/ml CaM, 1 mM CaCl 2 , 100 units/ml catalase, 10 units/ml superoxide dismutase, 5 M H4B, 20 M DTT, and 0.05-0.1 M nNOS in the presence or absence of 0.1 mM L-Arg or N G -hydroxy-L-Arg. Cytochrome c reductase activity was determined by monitoring the absorbance at 550 nm using an extinction coefficient of 21 mM Ϫ1 cm Ϫ1 . The H 2 O 2 generation rate was measured by the formation of ferric thiocyanate under similar conditions as described for the NO formation activity and NADPH oxidase activity but without catalase and superoxide dismutase.
Optical Absorption Spectra-Spectral experiments under aerobic conditions were carried out on a Shimadzu UV-2500 spectrophotometer maintained at 25°C by a temperature controller. Anaerobic spectral experiments were conducted on a Shimadzu UV-160A spectrophotometer maintained at 15°C in a glove box under a nitrogen atmosphere with an O 2 concentration of less than 50 ppm. To ensure that the temperature of the solution was appropriate, the cell was incubated for 10 min prior to spectroscopic measurements. Titration experiments were repeated at least three times for each complex. Regression analyses were performed, and lines giving an optimum correlation coefficient were selected. Experimental errors were less than 10%.
Crystal Structure-The crystal structure coordinates of bovine eNOS  heme domain (28) was obtained through the World Wide Web from the Protein Data Bank. RasMac 2.6-ucb1.0 software was used to make a dimer interface structure of nNOS based on the eNOS structure.

Aromatic Mutants
Spectral Properties-The mutants were expressed as fulllength enzymes in 1 liter of culture of E. coli cells yielding 100 -400 nmol. Purification of the mutants in the absence of H4B resulted in unstable enzymes. Therefore, all buffers used for enzyme preparation except for H4B-deficient enzyme preparation contained 5 M H4B. The resting Fe(III) form of the W678L mutant had a Soret peak at 438 nm ascribed to the low spin type heme as shown in Fig. 3A. The position of this peak did not move even after addition of 1 mM L-Arg, in contrast to the wild type enzyme. This suggests that the substrate binding site was changed by the mutation in terms of the absorption spectral character. The Fe(II) heme-CO complex of the W678L mutant had a peak at 444 nm, and no absorption band at around 420 nm ascribable to the denatured form, P420. The pyridine hemochromogen spectrum of the W678L mutant was the same as that of the wild type enzyme, and the heme content was 0.7/subunit. The W676L and F691L mutants showed spectral characteristics similar to those of the W678L mutant (not shown). These results suggest that the heme in W676L, W678L, and F691L has the normal internal thiol-axial ligand from Cys 415 and that the heme environment is well preserved after the mutations. However, the mutations might have led to very low binding affinities for L-Arg in terms of the optical absorption spectral change in the presence of 5 M H4B. In contrast, the absorption spectra of the W678H (Fig. 2B) and H692L, Y706L, and Y706F (not shown) mutants generated in the present study had broad Soret peaks around 400 nm, similar to the wild type (34,35). The addition of L-Arg to these mutants resulted in the shift of the Soret peak from 400 to 396 nm, suggesting that their L-Arg binding sites were not altered by the mutations.
Catalytic Properties-We next examined the catalytic activities of the mutants by monitoring NO formation, NADPH oxidation, and H 2 O 2 formation rates (Table I). When L-Arg is used as the substrate, the NO formation rates of the W676L, W678L, and F691L mutants were less than 0.1 nmol/min/nmol heme. The rate of the W678H mutant was 17 nmol/min/nmol heme, which is 35% that of the wild type. The H692L mutant had a catalytic activity comparable with that of the wild type enzyme. The mutation of Tyr 706 to Leu and Phe decreased the rate of NO formation down to 13 and 29% that of the wild type, respectively. When N G -hydroxy-L-Arg, the reaction intermediate, was used as the substrate, the W676L, W678L, and F691L mutants again did not show any detectable NO formation activity. The other mutants had slightly higher NO formation rates with N G -hydroxy-L-Arg compared with that with L-Arg, similar to that observed for the wild type enzyme. Thus, the mutations of aromatic residues in the dimer interface had similar effects on the NO formation with either L-Arg or N G -hydroxy-L-Arg.
To elucidate how the mutations in the H4B binding site couple electron transfer from NADPH to NO formation activity, we determined the NADPH oxidation rate during catalysis ( Table I). The NADPH oxidation rate of the wild type corresponds well with the NO formation rate, indicating close coupling of NADPH consumption with NO formation. However, all the mutants had relatively high NADPH oxidation rates compared with NO formation rates. In these mutants, NO synthesis was not coupled well with NADPH consumption, and the electrons from NADPH appear to be used to generate H 2 O 2 or H 2 O. In fact, for most of the mutants except for the Y706L and Y706F mutants, the H 2 O 2 formation rates were larger than that of the wild type (Table I). This uncoupling of the electron transfer was more obvious for the W676L, W678L, and F691L mutants, where NO formation activities were undetectable. It is interesting to note that the NADPH oxidation rates with the F691L and H692L mutants were higher than those observed for the wild type enzyme. The cytochrome c reductase activity of each mutant was essentially the same as that of the wild type enzyme (data not shown).
Heme Reduction with NADPH-To understand the electron transfer from the reductase domain to the heme active site, we examined the heme reduction rate with NADPH under anaerobic conditions. Table II summarizes the rate of heme reduction with NADPH for all the mutants generated in this study. Fig. 4 shows the Soret absorption spectral changes of the W678L and W678H mutants in the presence of CO caused by addition of NADPH, upon which time-dependent Fe(II)-CO complex formation was monitored. The heme of the W678L  mutant was very slowly reduced in the presence of L-Arg and H4B, and only a small fraction of the Fe(II)-CO complex was observed even after 20 min of incubation with NADPH. The W678H mutant was quickly reduced similar to that of the wild type enzyme (1.2 min Ϫ1 ) under the same conditions. The heme reduction rates of the W676L and F691L mutants were less than 0.01 min Ϫ1 . Note that complete removal of L-Arg and H4B from the full-length nNOS does not prevent electron transfer to the heme and the enzyme retains a significant heme reduction rate of 0.08 min Ϫ1 upon addition of excess NADPH. Thus, the very slow heme reduction rates observed for those Trp and Phe mutants are mostly due to the effects of the mutation but not to the low affinity of L-Arg and H4B to the aromatic mutants. The mutation of His 692 to Leu did not markedly change the heme reduction rate, in accordance with the effect on the NO formation activity. The removal of the hydroxyl group from Tyr 706 did not affect the heme reduction rate with NADPH, even though the Y706F mutant had NO formation activity 40% that of the wild type enzyme.
Dimer Formation-It has been suggested that only the dimeric form of NOS is catalytically active (9). The dimer formation ability of each mutant enzyme was analyzed in the absence of L-Arg and H4B, or after overnight incubation with L-Arg and H4B, using gel filtration column chromatography, to evaluate whether the mutations influence the monomer-dimer equilibrium. The gel filtration profiles shown in Fig. 5 (A and B) indicate that the full-length wild type and W678H mutant were predominantly oligomers in the absence of L-Arg and H4B. The overnight incubation with L-Arg and H4B resulted in an increase in the fraction of the dimeric form. In contrast, the W678L mutant was mainly oligomeric even after the incubation (Fig. 5C). Interestingly, all mutants studied here always had a proportion of the dimeric form present, with the different percentages summarized in Table II. Thus, it appears that the aromatic mutations of the H4B binding site never completely abolish the dimer formation ability of full-length nNOS. This finding is in contrast with that observed for the oxygenase domain of iNOS, where some aromatic mutations at the H4B binding site dissociate the dimer to the monomer (31). We also estimated the relative NO formation activity for each mutant per dimeric form as shown in Table II. Here again, it appears that the aromatic amino acids such as Trp 676 , Trp 678 , and Phe 691 are important for NO formation activity and are involved in the electron transfer from NADPH to the heme active site.
Effect of H4B on the Activity-Next we prepared H4B-free enzymes to examine the H4B dependence of the NO formation activity. The resting form of H4B-free wild type enzyme had absorption spectra with a maximum at 418 nm ascribed to the low spin complex (not shown). The H4B-free wild type fulllength nNOS did not show any obvious high spin shift on addition of H4B even after 4 h of incubation at 15°C. However, the addition of L-Arg resulted in a high spin shift within 10 min a Estimated by gel filtration chromatography after overnight incubation with 0.1 mM L-Arg and 5 M H4B. b NO formation activity from L-Arg relative to that of the wild type per dimer. Values were estimated from the data in Table I and the dimer formation fractions described in this Table. c (not shown). This is in contrast to iNOS, where addition of H4B to H4B-free enzyme caused a clear Soret spectral shift from the low spin (420 nm) to the high spin (396 nm) (13). H4B-free W678H, H692L, Y706L, and Y706F mutants of full-length nNOS showed similar spectral shift to the wild type enzyme. All of these H4B-free enzymes did not show any NO formation activity in the absence of H4B as with the wild type enzyme. However, NO formation was observed immediately after addition of the enzyme to the reaction mixture containing H4B with L-Arg. The NO formation rates of the mutants increased with increasing the H4B concentration as with the wild type enzyme (Fig. 6). The EC 50 values for H4B on NO formation are summarized in Table II. All the mutations studied here resulted in a dramatic increase in the EC 50 values, indicating a substantial decrease in H4B binding affinity. Note that especially, the mutation of Tyr 706 affected H4B binding by increasing the EC 50 value by 30 -40-fold compared with the wild type enzyme.

R414 Mutants
Tyr 706 probably interacts with both Arg 414 and the heme carboxylate via hydrogen bonds and is located in close proximity to Trp 678 in full-length nNOS, based on homology-based comparison with the crystal structures of iNOS and eNOS oxygenase-domain dimers (Fig. 7). Arg 414 is well conserved among NOSs. To evaluate the role of this residue in the catalytic activity of and electron transfer within full-length nNOS, Arg 414 mutants were prepared and characterized. The resting state of R414Q had a Soret peak at 400 nm, and addition of L-Arg resulted in high spin shift similar to the wild type enzyme. In contrast, R414E and R414L mutants had their Soret peaks at 420 and 441 nm, respectively (not shown). Addition of L-Arg to the resting R414E and R414L mutants did not change the Soret band (not shown), similar to that observed for the W676L, W678L, and W692L mutants. The Fe(II)-CO complexes of all Arg 414 mutants had their Soret absorption bands at around 445 nm. In the case of R414E mutant, the absorption peak of the Fe(II)-CO complex quickly moved from 445 to 420 nm, suggesting that the heme active site of the enzyme was very unstable after the mutation. Table III summarizes various kinetic parameters for the Arg 414 mutants. The NO formation activities of R414E and R414L mutants were less than 0.1 nmol/min/nmol heme, whereas the R414Q mutant had activity comparable with that of the wild type enzyme. The heme re-duction rates of the R414E and R414L mutants with excess NADPH were less than 0.01 min Ϫ1 , whereas the reduction rate of R414Q mutant was similar to that of the wild type enzyme. The EC 50 values for H4B on NO formation activity for the R414Q mutant was 0.4 M, which is similar to that of the wild type. Thus, it appears that Arg 414 is important in the catalytic activity associated with electron transfer from the reductase domain to the heme active site. DISCUSSION In the present study, mutational analysis was used to address the specific roles of Arg 414 and the aromatic residues of the H4B binding site in catalysis and electron transfer. Hydrogen bond networks between H4B and the backbone carbonyls of eNOS residues Trp 449 (which corresponds with Trp 678 of nNOS) and Phs 462 (Phe 691 of nNOS), and a water-bridged hydrogen bond to His 463 (His 692 of nNOS) are to be considered (28). In the dimer interface of nNOS, the H4B ring must be located between Trp 678 of one subunit and Trp 676 , Phe 691 , and His 692 of the other subunit, based on the crystal structure of eNOS (Fig. 2). The mutation of the aromatic residues Trp 676 , Trp 678 , and Phe 691 to Leu and the mutation of Arg 414 to Glu or Leu resulted in a clear loss of NO formation activity and a loss of heme reduction ability in the presence of NADPH under anaerobic conditions (Tables I-III).
Spectral Properties-The Soret absorption bands of the W678L (Fig. 3B), W676L (Fig. 3A), F691L, and R414L mutants were located at around 438 -441 nm. Low spin Fe(III) cytochrome P-450 complexes with oxygen or nitrogen sixth axial ligands have Soret bands at around 416 -425 nm (44,45). Low spin Fe(III) complexes with sulfur as the sixth axial ligand have Soret band positions at around 461-465 nm for cytochrome P-450s (46) or 455-460 nm for NOSs (47,48), respectively. Subunit dissociation or denaturation with urea resulted in a shift in the Soret absorption band to 460 nm (48). The cytochrome P-450-cyanide complex has a Soret absorption band at around 440 nm (46). Therefore, at this point, it is difficult to assign the 6th axial ligand of these mutants.
The absorption band of the W678L, W676L, F691L, and R414L mutants did not move on addition of excess L-Arg. This suggests that the substrate binding site of the full-length nNOS was greatly affected and that the affinity of L-Arg may be greatly decreased. It is difficult to examine how the H4B binding site in full-length nNOS was changed by the aromatic and Arg 414 mutations. However, the presence of 5 M H4B markedly stabilized the mutant structures, suggesting that H4B binding is still effective in structuring the active site cavity appropriately.
Catalytic Properties-NO formation activities for wild type nNOS correlate with NADPH oxidation rates, because electron transfer is well coupled with the catalytic activity. Accordingly, for the aromatic mutants with low NO formation activities, NADPH oxidation rates are correspondingly low (Table I). However, the F691L, H692L, R414Q, and R414E mutants had relatively high NADPH oxidation rates compared with the wild type, considering their NO formation activities (Tables I and  III). It is suggested, therefore, that those amino acid residues are important in the interdomain or intersubunit electron transfer of the nNOS enzyme. When NADPH oxidation and/or electron transfer is not well coupled with the molecular oxygen activation, H 2 O 2 and/or H 2 O is formed (34, 35, 49 -51). Thus, mutants with low NO formation activities produce more H 2 O 2 (Table I). Also, it is highly possible that H4B is closely associated with the electron transfer process in this enzyme, because H4B binding in these mutants may be markedly different to that of the wild type enzyme, and appropriate electron transfer involving H4B is no longer feasible in the mutant enzymes.
Heme Reduction-The heme reduction rates for mutants with very low NO formation activities (less than 0.1 nmol/min/ nmol heme) were also very low (less than 0.01 min Ϫ1 ) (Table  II). We also found that H4B-free wild type enzyme showed heme reduction at a significant rate (0.08 min Ϫ1 ) on addition of NADPH even in the absence of L-Arg (not shown). Therefore, the very slow heme reduction rates for the mutants are ascribed to the mutation effect rather than the effect of the substrate or the cofactor. This suggests that the first electron to the heme might come from the reductase domain in an H4Bindependent fashion but that the pterin radical, if this exists, may still be involved in the second electron transfer to the oxy-ferrous complex, as proposed by Bec et al. (25). Taken together, these results suggest that the aromatic residues Trp 676 , Trp 678 , and Phe 691 are important for proper H4B binding and crucial for electron transfer from the reductase domain and for NO formation. If a pterin radical is truly involved in the electron transfer, these aromatic amino acids should help to stabilize the radical during catalysis (25,26,28).
Dimeric Formation Ability-H4B binding has been suggested to promote dimerization of the enzyme that is thought to be essential for formation of an active enzyme (9 -17). A recent report suggested that electrons supplied to the heme of the oxygenase domain originate from the reductase domain of the alternate subunit by intersubunit electron transfer within the dimer (52). Therefore, it is possible that the mutations caused the loss of H4B binding and also affected dimer formation. However, gel filtration analysis of the mutants (Fig. 5) indicates that they did not totally abolish dimer formation, suggesting that H4B binding is not absolutely essential for the dimer formation in full-length nNOS, as has already been suggested by other workers (9,10,13,23,49). They have reported that dimer formation of nNOS and eNOS is H4Bindependent and regulated solely by heme availability. The crystal structural study on iNOS agrees with this suggestion (3), although another group has claimed that H4B binding appears to play a critical role in iNOS dimer formation (11,12). The present results are consistent with previous reports that full-length nNOS makes dimers in an L-Arg-and H4B-independent fashion (9,10,23,49).
H4B and Aromatic Character-Several reports have suggested that H4B functions commonly in all NOS isomers as an electron supplying cofactor and an allosteric effector (9 -17). A comparison of the iNOS and eNOS crystal structure reveals the striking conservation of the H4B binding site (27)(28)(29)(30). As mentioned above, from the eNOS crystal structure, a catalytic model was proposed involving a pterin cation radical based on the affinity of L-Arg to the H4B binding site (28). Also, analysis of the nNOS reaction at Ϫ30°C suggested that a pterin radical may deliver an electron to the oxy-ferrous intermediate complex generated during catalysis (25). The pterin radical was recently detected with EPR under specific turnover conditions, although the specific nature of the radical could not be determined (26). If that is the case, aromatic residues in the dimer interface could stabilize the radical intermediate with aromatic stacking interactions. In fact, the W676L, W678L, and F691L mutants, which have both lost the aromatic character, did not have any NO formation activity, and heme reduction was extremely slow for these mutants in the presence of NADPH. On the other hand, the W678H mutant that still conserves the aromatic stacking showed normal spectral properties, NO formation activity, and heme reduction rate with NADPH in the presence of excess L-Arg and H4B, although its H4B affinity was greatly reduced (EC 50 value of H4B increased 18-fold of wild type). These results suggest that the aromatic character of Trp 676 , Trp 678 , and Phe 691 are essential for efficient electron transfer to the heme, probably by stabilizing the pterin radical during catalysis (26,28).
Hydrogen Bonding Network Involving the Heme Carboxylate-The mutation of Tyr 706 dramatically increased the EC 50 value for H4B (40-fold that of wild type) on NO formation activity, although the same mutations did not significantly affect the heme reduction rate with NADPH (Table II). The replacement of Arg 414 with Glu (acidic residue) or Leu (neutral residue) resulted in a clear loss of NO formation activity and an extreme decrease in the rate of heme reduction with NADPH (Table III). Interestingly, the mutation of Arg 414 to Gln had little effect on NO formation activity, rate of heme reduction, and H4B binding, despite resulting in a loss of positive charge. It appears, therefore, that the hydrogen bonding role of Arg 414 is more important than its charge. Tyr 706 and Arg 414 are well conserved among all NOSs. In the eNOS dimer structure (28), Tyr 706 is located at the side of the heme plane and interacts with the carboxyl moiety of heme propionate as shown in Fig. 7. Tyr 706 may also interact via a hydrogen bond with Arg 414 , which is the adjacent residue to the axial ligand, Cys 415 (with RasMac 2.6-ucb1.0 software) (Fig. 7). Hydrogen bonds between the heme and Tyr 706 and between Tyr 706 and Arg 414 could stabilize the active conformations of the heme environment. The other propionate group of heme is hydrogen bonded to the pterin and simultaneously interacts with the amino group of the substrate L-Arg. Thus, several hydrogen bond networks appear to keep the structure of the heme active site in an optimal conformation for maximal electron transfer and NO formation to occur. The charge balance between H4B, Trp 678 , Arg 414 , Tyr 706 , and the heme propionates appear to be a critical aspect of the active site. The mutation of Tyr 706 markedly reduced the H4B binding affinity by disrupting the important hydrogen bond network, despite the fact that it is located at a significant distance from H4B. The charge and volume of Arg 414 residue is likely important to hold the heme plane in a stable fashion and/or to control the redox potential of the heme iron for the optimal catalytic reaction.
Comparison with Aromatic Mutants of iNOS Oxygenase Domain-A similar study involving the mutational analysis of the H4B binding site in iNOS oxygenase domain was reported (31). Ala mutations of Trp 455 and Phe 470 of the iNOS oxygenase domain, which correspond to Trp 676 and Phe 691 in nNOS, respectively, dissociated the dimeric form of the oxygenase domain of iNOS. The dissociated monomer did form heterodimers with a full-length iNOS subunit. Interestingly, these heterodimers retained partial NO formation activities. Ala mutation of Trp 457 of iNOS, which corresponds to Trp 678 in nNOS, partially retained dimer formation and NO formation activity. It was suggested that the aromatic character of these residues is not essential for H4B binding or NO formation activity within the iNOS oxygenase domain. These results are not in accordance with our results obtained using full-length nNOS mutants. It is not certain whether these differences are due to differences between the two NOS isomers or/and between the heterodimer (consisting of the oxygenase domain subunit and the full-length subunit) and the full-length homodimer.
Addition of L-Arg alone, H4B alone, or both combined to iNOS oxygenase domain increased the heme iron reduction potential (54). In contrast, addition of L-Arg alone, H4B alone, or both to nNOS oxygenase domain do not markedly change the heme iron potential (54). These results suggest that there are some differences in the heme environment and H4B binding aspects between the isomers. Further experiments should be done to understand the precise role of H4B in the reaction cycle of each NOS isomer.
Summary-In summary, it was found that aromatic residues Trp 676 , Trp 678 , and Phe 691 in the H4B binding site of the full-length nNOS enzyme are crucial in electron transfer for heme reduction with NADPH as well as the binding of the substrate and/or H4B. Although Tyr 706 is located distant from H4B, the mutation of Tyr 706 dramatically increased the EC 50 value of H4B (30 -40-fold that of wild type) for NO formation activity, indicating that the hydrogen bond formed with one of the propionates influences the H4B binding affinity, probably via its influence on the other propionate which interacts directly with H4B. Arg 414 is also crucial for NO formation activity and electron transfer through its involvement in this key hydrogen bonding network in the heme active site.