Chimeric Enzymes of Cytochrome P450 Oxidoreductase and Neuronal Nitric-oxide Synthase Reductase Domain Reveal Structural and Functional Differences*

The nitric-oxide synthases (NOSs) are comprised of an oxygenase domain and a reductase domain bisected by a calmodulin (CaM) binding region. The NOS reductase domains share (cid:1) 60% sequence similarity with the cytochrome P450 oxidoreductase (CYPOR), which transfers electrons to microsomal cytochromes P450. The crystal structure of the neuronal NOS (nNOS) connecting/FAD binding subdomains reveals that the structure of the nNOS-connecting subdomain diverges from that of CYPOR, implying different alignments of the flavins in the two enzymes. We created a series of chimeric en- zymes between nNOS and CYPOR in which the FMN binding and the connecting/FAD binding subdomains are swapped. A chimera consisting of the nNOS heme domain and FMN binding subdomain and the CYPOR FAD binding subdomain catalyzed significantly increased rates of cytochrome c reduction in the absence of CaM and of NO synthesis in its presence. Cytochrome c reduction by this chimera was inhibited by CaM. Other chimeras consisting of the nNOS heme domain, the CYPOR FMN binding subdomain, and the nNOS FAD binding subdomain with or without the tail region also catalyzed cytochrome c reduction, were not modulated by CaM, and could not transfer electrons into the heme domain. A chimera consisting of the heme domain of nNOS and the reductase domain of CYPOR reduced cytochrome c and ferricyanide at rates 2-fold higher than that of native CYPOR, suggesting

Nitric-oxide synthases (NOSs) 1 are heme-and flavin-containing enzymes that catalyze the formation of nitric oxide (NO) and citrulline from arginine through two NADPH-requiring monooxygenation steps that are mechanistically similar, but not identical, to those of the cytochrome P450 system. NO is a molecule with diverse physiological effects, including neurotransmission, hemodynamic regulation, and cytotoxicity. There are three isoforms of NOS, neuronal (nNOS), endothelial (eNOS), and macrophage (iNOS), which are differentially expressed, modified, and regulated (for review, see Ref. [1][2][3]. They are arranged in a domain structure, with an N-terminal oxygenase domain, containing the heme active site, the arginine binding site, and the tetrahydrobiopterin cofactor site, and a C-terminal reductase domain, containing FAD and FMN cofactors and the NADPH binding site, connected by a calmodulin (CaM) binding sequence. The transfer of electrons is from NADPH to FAD to FMN to the heme, reminiscent of the cytochrome P450 system. The binding of CaM, however, is required for the NOSs to transfer electrons from the reductase domain to the heme where catalysis occurs.
Despite the fact that all three NOS isoforms share 50 -60% overall sequence similarity, each of the NOS isoforms is intrinsically different from the other with regard to rates of NO synthesis, cytochrome c reduction, electron transfer, and hemenitrosyl complex formation. Thus, structural differences existing among the various isoforms must account for this variability. The structures of the heme domains of all three NOS isoforms have been solved and exhibit apparently minor differences, except in their initial N-terminal sequences (4 -6). Chimeric enzymes in which the heme domain of one isoform is connected to the flavin domain of another (7) indicated that the identity of the flavin domain rather than that of the heme domain determined the rates of NO synthesis and cytochrome c reduction.
The crystal structures of the entire reductase domains have not yet been reported, but several regulatory elements in this domain have been identified and characterized (for review, see Ref. 2). In addition to the CaM binding sequence, an autoregulatory insert was identified in the FMN binding region of the constitutively expressed isoforms, nNOS and eNOS, but not in the inducible isoform, iNOS (8). This autoregulatory insert obstructs transfer of electrons into the heme domain in the absence of CaM. In the presence of CaM, this impediment is at least partially relieved, and electron flux to the heme can occur. If the autoregulatory insert is completely removed, the rate of electron flow increases, indicating that this element intrinsically regulates electron flow to the heme, even in the presence of CaM (9 -11).
Another regulatory element exists at the C terminus of all known NOSs. An extension of between 21-42 amino acids, beyond the residues homologous to the terminus of CYPOR, impedes electron flow through the flavin domain in the absence of CaM (12,13). The crystal structure of the FAD and NADPH binding regions suggests that this tail region bisects the flavin domain between the FMN and FAD and emerges on the other side of the molecule (14), thus impeding the flow of electrons between the flavins. In the presence of CaM, this blockade is partially removed, presumably by repositioning of the tail, facilitating flux between the flavins (12,13). The tail regions of nNOS and eNOS contain a serine residue that has been demonstrated in eNOS to be phosphorylated as a result of physiological stimuli, resulting in higher rates of both NO production and cytochrome c reduction (15)(16)(17)(18). Mutation of this residue in either eNOS or nNOS from serine to aspartate mimics the negative charge of phosphorylation and results in an enzyme with increased NO synthesis and cytochrome c reduction (13,19,20). Partial removal of the tail region of eNOS, from serine 1179 to the end, resulted in an enzyme with similar characteristics (21). These results are consistent with a model in which the negative charge of phosphorylation or aspartate mutation repositions the tail such that electron flow is potentiated (13).
Although the amino acid sequence of the nNOS reductase domain is 58% similar to that of CYPOR (22), CYPOR completely lacks the two aforementioned regulatory elements. The crystal structure of CYPOR, unlike that of the NOS reductase domains, has been solved (23), and the structure of the FAD and NADPH binding subdomain is similar to the only portion of the nNOS reductase domain structure thus far crystallized and solved (14). Given the similarity in sequence, the presumed similarity in structure, and the contrast in strict regulatory elements present in the NOS reductase domains but absent in CYPOR, chimeric enzymes were created in which the FMN and/or FAD binding subdomains of CYPOR are incorporated into nNOS. These chimeric molecules were used to investigate the interplay between these elements and other domain sequences and their effects on electron flow by measuring their capacities to catalyze both flavoprotein-mediated electron transport and heme-mediated NO production.

Chemicals
(6R)-5,6,7,8-Tetrahydrobiopterin was from Research Biochemicals International (Natick, MA). All other chemicals were from Sigma and were of the highest grade available.

Enzymes
Pfu Turbo polymerase was from Stratagene (La Jolla, CA), shrimp alkaline phosphatase was from U. S. Biochemical Corp., and restriction enzymes were from Promega (Madison, WI) or New England Biolabs (Beverly, MA).

Recombinant DNA Manipulations
All of the following constructs are summarized in Table I and illustrated in Fig. 1.
Soluble CYPOR-The codons encoding residues 79 -678 of cytochrome P450 reductase were amplified by PCR, using the original CYPOR plasmid kindly provided by Dr. Charles Kasper as template, with the following primers: Upstream, atg cag cat atg gct aac att atc gta ttc ta; Downstream, aag ctt atg cat tct aga cta gct cca cac atc tag tga. The forward primer encodes an NdeI site, and the reverse primer encodes an XbaI site for insertion into the vector. The PCR product was restriction-digested with NdeI/XbaI and ligated with NdeI/XbaI-digested pCW vector. This ligation was used to transform XL10-gold cells (Stratagene), and colonies were screened by restriction digest. The correct construct was then used to transform Escherichia coli BL21 cells.
nNOS ox,FMN /C FAD,NADPH -The codons encoding the initial 963 amino acids of nNOS, encompassing the heme domain and FMN binding subdomain, were amplified by PCR using nNOS pCW (24) as template, with the following primers: Upstream, tag cta cat atg gct gaa gag aac acg; Downstream, ctt cca gga tcg atc att gct aat gag gga gtt. The forward primer encodes a NdeI site for insertion into the vector. The return primer incorporates a silent mutation generating a ClaI site. The codons encoding amino acids 241-678 of CYPOR, encompassing the connecting and FAD/NADPH binding subdomains, were amplified by PCR with the following primers: Upstream, ggg gag gat cga tcc att cgc cag tat gag ctc; Downstream, aag ctt atg cat tct aga cta gct cca cac atc tag tga. The forward primer incorporates a silent mutation generating a ClaI site, and the return primer incorporates an XbaI sequence for insertion into the vector. The two PCR products were restrictiondigested with NdeI/ClaI (nNOS) or ClaI/XbaI (CYPOR) and ligated with NdeI/XbaI-digested pCW vector. This ligation was used to transform XL10-gold cells (Stratagene), and colonies were screened by restriction digest. The correct construct was then used to cotransform E. coli BL21 cells along with groELS plasmid.
C FMN /nNOS FAD,NADPH -The codons encoding residues 79 -238 of CY-POR, encompassing the FMN binding subdomain, were amplified by PCR with the following primers: Upstream, atg cag cat atg gct aac att atc gta ttc tat; Downstream, gcg aat gga tcg atc ctc ccc agt ggc ttc tac. The forward primer encodes a NdeI site for insertion into the vector. The return primer incorporates a silent mutation generating a ClaI site. The codons encoding amino acids 962-1430 of nNOS, encompassing the connecting and FAD/NADPH binding subdomains, were amplified by PCR with the following primers: Upstream, agc aat gat cga tcc tgg aag agg aac aag ttc; Downstream, ctg cag gtc gac tct aga. The forward primer incorporates a silent mutation, generating a ClaI site, and the return primer incorporates an XbaI sequence for insertion into the vector. The two PCR products were restriction-digested with NdeI/ ClaI (CYPOR) or ClaI/XbaI (nNOS) and ligated with NdeI/XbaI-digested pCW vector. This ligation was used to transform XL10-gold cells (Stratagene), and colonies were screened by restriction digest. The correct construct was then used to cotransform E. coli BL21 cells along with groELS plasmid.
C FMN /nNOS FAD,NADPH (tr1)-The codons encoding residues 79 -238 of CYPOR, encompassing the FMN binding subdomain, were amplified by PCR with the following primers: Upstream, atg cag cat atg gct aac att atc gta ttc tat; Downstream, gcg aat gga tcg atc ctc ccc agt ggc ttc tac. The forward primer encodes an NdeI site for insertion into the vector. The return primer incorporates a silent mutation generating a ClaI site. The codons encoding amino acids 962-1397 of nNOS, encompassing the connecting and FAD/NADPH binding subdomains, but not the C-terminal residues, were amplified by PCR with the following primers: Upstream, agc aat gat cga tcc tgg aag agg aac aag ttc; Downstream, c tag tct aga tta tcc aaa gat gtc ctc gtg. The forward primer incorporates a silent mutation generating a ClaI site, and the return primer incor- porates an XbaI sequence for insertion into the vector. The two PCR products were restriction-digested with NdeI/ClaI (CYPOR) or ClaI/ XbaI (nNOS) and ligated with NdeI/XbaI-digested pCW vector. This ligation was used to transform XL10-gold cells (Stratagene), and colonies were screened by restriction digest. The correct construct was then used to cotransform E. coli BL21 cells along with groELS plasmid. nNOS ox /C FMN /nNOS FAD,NADPH -The codons encoding residues 1-754 of nNOS, encompassing the heme domain, as well as vector sequence from nucleotide 9101 to the beginning of the nNOS cDNA at 212 were amplified by PCR with the following primers: Upstream, cac tga cgc gtt gcg cga; Downstream, cgc ctt gac gcg ttt ggc cat ggc ctg ccc cat tag. The forward primer spans a vector-encoded MluI site at 9106. The return primer incorporates a silent mutation generating an MluI site in the nNOS cDNA. The codons encoding amino acids 1-596 of the abovedescribed C FMN /nNOS FAD,NADPH plasmid, encompassing the FMN binding subdomain of CYPOR and the connecting and FAD/NADPH binding subdomains of nNOS, along with vector sequence from the end of the chimera coding sequence to nucleotide 9118 were amplified by PCR with the following primers: Upstream, ggg gag gat cga tcc att cgc cag tat gag ctc; Downstream, aag ctt atg cat tct aga cta gct cca cac atc tag tga. The forward primer incorporates a silent mutation, generating an MluI site, and the return primer spans the vector-encoded MluI site at 9106. The two PCR products were restriction-digested with MluI and ligated together. This ligation was used to transform XL10-gold cells (Stratagene), and colonies were screened by restriction digest. The correct construct was then used to cotransform E. coli BL21 cells along with groELS plasmid.
nNOS ox /C FMN /nNOS FAD,NADPH (tr1)-This chimera was made exactly as nNOS ox /C FMN /nNOS FAD,NADPH described above except that the template DNA for the second PCR was from C FMN /nNOS FAD,NADPH (tr1) rather than C FMN /nNOS FAD,NADPH .
nNOS ox /C FMN,FAD,NADPH -A segment of the nNOSpCW plasmid from the vector-encoded MluI site at nucleotide 9101 through the start of the nNOS cDNA at nucleotide 212 to the codon for residue 754 of nNOS, encompassing the heme domain, was amplified by PCR. The following primers were used: Upstream, cac tga cgc gtt gcg cga; Downstream, cgc ctt gac gcg ttt ggc cat ggc ctg ccc cat tag. The return primer incorporates a silent mutation generating an MluI site. A segment of the CYPOR plasmid from the start of the soluble CYPOR cDNA at residue 79 (nucleotide 212 in pCW) to the codon for final residue of CYPOR (678) through the vector-encoded MluI site was amplified by PCR. The following primers were used: Upstream, gac gta aaa cgc gtc aac att atc gta ttc tat; Downstream, tcg cgc aac gcg tca gtg. The forward primer incorporates a silent mutation generating an MluI site. The two PCR products were restriction-digested with MluI and ligated together. This ligation was used to transform XL10-gold cells (Stratagene), and colonies were screened by restriction digest. The correct construct was then used to cotransform E. coli BL21 cells along with groELS plasmid.

Protein Expression and Purification
nNOS, CYPOR, and chimeric proteins were expressed and purified as previously described (24) except that cultures were grown in 500 ml of medium in Fernbach flasks, all of the column equilibration and wash buffers contained 100 mM NaCl, and the protein was eluted with buffer containing 500 mM NaCl and 15 mM 2Ј,3Ј-AMP. The reductase proteins (CYPOR, C FMN /nNOS FAD,NADPH , and C FMN /nNOS FAD,NADPH (tr1)) were fully oxidized by ferricyanide titration before gel filtration chromatography. Calmodulin was prepared by the method of Zhang and Vogel (25). Based on the gel filtration chromatography step, all of the heme domain-containing chimeras were primarily dimeric, including NCC, which contains no nNOS-derived reductase domain modules, indicating that all the determinants necessary for dimerization are contained within the heme domain of nNOS.

Spectrophotometric Methods
CO difference spectra were performed as described (24). The molar protein concentrations for heme-containing chimeras were determined based on heme content via reduced CO difference spectra, where ⑀ ϭ 100 mM Ϫ1 cm Ϫ1 for ⌬A 445-470 . The molar protein concentrations for non-heme-containing chimeras were determined based on flavin content via absolute absorbance, where ⑀ ϭ 21 mM Ϫ1 cm Ϫ1 (10.5 mM Ϫ1 cm Ϫ1 / flavin) at 454 nm. All spectral analyses were performed using a Shimadzu model 2401PC UV-visible dual-beam spectrophotometer.

Stopped-flow Spectroscopy
Stopped-flow reactions were performed aerobically under turnover conditions at 23°C, as described in Miller et al. (26) and Roman et al. (12,13), using an Applied Photophysics SX.18MV diode array stoppedflow spectrophotometer, with a dead time of 2 ms. Reactions contained 3 M enzyme and 100 M NADPH in 50 mM Tris-HCl, pH 7.4, and 100 mM NaCl. Where indicated, 10 M CaM was also added. Flavin reduction was monitored at 485 nm.

Measurement of Activity
Nitric oxide formation (hemoglobin capture assay) and cytochrome c reduction were measured at 23°C as described (27,28), with the exception that both assays were performed in a buffer containing 50 mM Tris-HCl, pH 7.4, and 100 mM NaCl. Rates of NO synthesis and cytochrome c reduction were determined using extinction coefficients of 60 mM Ϫ1 cm Ϫ1 at 401 nm and 21 mM Ϫ1 cm Ϫ1 at 550 nm, respectively. Ferricyanide reduction was performed in the same manner as cytochrome c reduction, except that the ferricyanide concentration was varied, and the extinction coefficient was 1.02 mM Ϫ1 cm Ϫ1 at 420 nm. The reoxidation of reduced flavins was monitored at 485 nm for all enzymes in the presence of 50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 20 M NADPH, and 2 M enzyme at 23°.

RESULTS
Expression and Spectral Characterization of Chimeric Enzymes-All of the chimeric enzymes expressed very well in E. coli, with yields comparable with those of wild-type nNOS and CYPOR (i.e. 100 -150 nmol/liter cells), and all were of the correct mass as visualized by SDS-PAGE. All of the chimeric enzymes looked spectrally healthy. All chimeric holoenzymes contained both heme and flavins, as determined by absolute spectra, and were isolated in a mixed low spin/high spin form that shifted completely to low spin with the addition of 1 mM imidazole and completely to high spin upon the addition of 100 M arginine (Fig. 2A). The absorbance peak shifted an additional 4 nm lower upon the addition of H 4 B (not shown), suggesting that the binding site for this cofactor is intact. All formed the characteristic peak at ϳ445 nm in the reduced, CO-bound form, indicating proper insertion of the heme moiety in the chimeras. Spectral measurement of the binding constant for arginine (K s ) yielded values comparable with that of wildtype NOS (717 nM; Ref. 24). Shown in Fig. 2B are the substrate difference spectra for titration of nNOS ox /C FMN,FAD,NADPH by arginine, which yielded K s ϭ ϳ300 nM. The similarity of these values to the wild type suggests that the binding site for arginine is unimpaired in the chimeras. All of these characteristics are similar in both the wild-type and chimeric proteins, suggesting an intact and potentially functional heme domain.
All of the non-heme domain-containing flavoprotein constructs were isolated as greenish in color, indicating the forma-tion of an air-stable flavin semiquinone, which was confirmed spectrophotometrically (not shown), as is typically seen with CYPOR. This semiquinone form persisted until oxidized by the addition of ferricyanide. None of the chimeras were as stable as nNOS or CYPOR, as judged by loss of activity over a period of a week on ice, but all were stable (less than 10% drop in activity) for at least 2 days. Therefore, all activity measurements were performed on freshly prepared enzyme within 2 days of isolation.
Cytochrome c Reduction-To determine the functionality of the reductase domains, cytochrome c reduction was assayed. Reduction of cytochrome c involves passage of electrons from NADPH to FAD to FMN to cytochrome c and, thus, requires an intact reductase domain. Fig. 3 shows that all of the chimeric holoenzymes were able to reduce cytochrome c. Wild-type nNOS, nNOS (tr1), and CYPOR are shown for comparison. As reported previously, the rates of reduction by nNOS and nNOS (tr1) in the presence of CaM are very similar to that of CYPOR, ϳ3000 -4500 min Ϫ1 . The K m values for cytochrome c were determined for all of the enzymes shown, and all are between 3-6 M (not shown), so any differences observed among the chimeras are not due to altered interaction with cytochrome c.
The profile for NNC resembles that of nNOS (tr1) in that in the absence of CaM, cytochrome c reduction is very high, 2-2.5fold higher than either CYPOR or nNOS in the presence of CaM. In the presence of CaM, cytochrome c activity decreases by 75%, to a level half that of CYPOR. NNC retains the auto- regulatory insert in the FMN subdomain but has lost the regulatory tail region.
NCN and NCN (tr1) reduce cytochrome c at rates significantly less than that of CYPOR (ϳ700 -1000 versus ϳ3000 min Ϫ1 ) but higher than nNOS in the absence of CaM (ϳ350 min Ϫ1 ). These rates are unaffected by CaM. These chimeras have lost the autoregulatory insert, since their FMN domains are from CYPOR, and NCN (tr1) has also lost the C-terminal tail regulatory region. The two comparable reductase domain constructs, CN and CNtr1 (not shown), also reduced cytochrome c, but at very low rates, 7 Ϯ 1.2 and 15.4 Ϯ 2.5, respectively.
Although the entire reductase domain of NCC has been donated by CYPOR, the rate of cytochrome c reduction reaches the level achieved by NNC in the absence of CaM, ϳ7500 min Ϫ1 . Interestingly, this is more than twice the rate of CYPOR alone (ϳ3000 min Ϫ1 ) even though there is no contribution by nNOS to the reductase domain. Unlike NNC, however, this rate is unaffected by CaM. It is notable that neither regulatory region is present in this chimera.
Ferricyanide Reduction-The reduction of ferricyanide can be accomplished by passage of electrons from NADPH to FAD to ferricyanide. The transfer of electrons from FAD to FMN is not required, so ferricyanide reduction is a good indicator of the integrity of the FAD/NADPH binding regions. Table II lists the turnover numbers and K m values of ferricyanide in ferricyanide reduction for CYPOR, nNOS, and each of the chimeras. Note that the K m values for CYPOR and wild-type nNOS (32 M and 1.8 mM, respectively) as well as the turnover numbers (6,160 and 34,630 min Ϫ1 , respectively) vary greatly, and thus, the conditions established for measuring the rate of ferricyanide reduction of CYPOR (i.e. 1 mM ferricyanide) cannot be used for the NOSs. The kinetic constants of the chimeric proteins fall into two categories, one with K m values similar to nNOS and one with K m values similar to CYPOR. NNC and NCC have K m values of 34 -35 M, whereas NCN and NCN (tr1) have K m values of 1.4 -1.5 mM. These K m values correspond to that of the parent of the FAD subdomain (i.e. NNC and NCC have the CYPOR FAD subdomain, whereas NCN and NCN (tr1) have the nNOS FAD subdomain), lending support to the idea that ferricyanide is reduced through FAD and not FMN. Furthermore, although NCN and NCN (tr1) have turnover numbers comparable with nNOS, the parent of their FAD subdomains, NNC and NCC, have turnover numbers twice that of CYPOR, indicating that the presence of the heme domain has a direct effect on activity by the reductase domain, particularly in the case of NCC, where the entire reductase domain was donated by CYPOR.
In the presence of CaM, the turnover number for wild-type nNOS ferricyanide reduction does not change, but the K m value decreases by 50% (to 780 M). The activity of the chimeras is essentially unchanged by CaM (not shown).
NO Synthesis-Only one of the chimeras was competent to synthesize NO. NNC catalyzed NO production in the presence of CaM at a rate of 88 Ϯ 1 min Ϫ1 , as compared with wild-type nNOS, which has a turnover number of 71 Ϯ 6 min Ϫ1 under the same conditions (Fig. 4). Thus, despite the chimeric nature of its reductase domain, NNC is at least as good as, and perhaps better than, the wild-type enzyme at synthesizing NO. NNC also synthesizes NO at almost 3ϫ the rate of nNOS (tr1) in the presence of CaM but, unlike nNOS (tr1), forms no NO in the absence of CaM. NNC and nNOS (tr1) have identical heme domains and FMN subdomains. They differ only in the connecting/FAD subdomain, and that of nNOS (tr1) was initially designed to mimic that of CYPOR, which is the parent of the NNC connecting/FAD subdomain.
To determine whether the other chimeric holoproteins were able to transfer electrons to their heme domains even though NO is not formed, heme reduction by electrons donated by NADPH was measured in the presence of CO, superoxide dismutase, and catalase. Heme reduction was monitored by an increase in absorbance at 445 nm, the characteristic peak observed for the reduced, CO-bound form of the enzyme. None of the chimeric enzymes NCN, NCN (tr1), or NCC formed the reduced, CO-bound heme, indicating that electrons are not passed from the reductase domain to the heme in these proteins.
Flavin Reduction-Initial flavin reduction in the presence of oxygen and NADPH was measured via stopped-flow spectroscopy, and the results are shown in Table III. All enzymes showed the biphasic flavin reduction characteristic of the wildtype nNOS (26). There is clearly no effect of CaM on the rate of either the fast or slow phase of flavin reduction in any of the chimeric holoproteins. Furthermore, the rates of flavin reduction in the chimeras are similar to that of wild-type nNOS in the presence of CaM, indicating that whatever elements or conformation that the binding of CaM alters to relieve inhibition of electron flow through the flavin domain are no longer present or effective in the chimeras.
After the flavins have been reduced and catalysis ends, the nNOS reductase domain assumes a partially reduced, air stable semiquinone form (26,32), a state also observed with CYPOR. Eventually, over a period of hours for nNOS and days for CYPOR, the flavins will become fully oxidized. When the tail region of any of the NOS isoforms was removed, the en- zymes no longer formed the semiquinone but quickly became fully reoxidized (12,13), suggesting significant differences in the exposure of the flavins to solvent between CYPOR, which does not have the tail region, and nNOS (tr1), which had the tail region removed to resemble CYPOR. To determine whether the air-stable semiquinone form was generated in the chimeras, flavin reoxidation was followed. In Fig. 5, the enzyme is initially in its resting state, with flavins fully oxidized. NADPH is added, the flavins reduced, and catalysis is begun. When the NADPH is exhausted, the flavin absorbance returns to a plateau whose absorbance represents the air-stable form. If the absorbance plateaus at the same absorbance as it started, the flavins have been fully reoxidized. If it regains only half the absorbance change, the semiquinone form is present. Fig. 5 shows that NCN (tr1) forms a stable semiquinone, and these data are representative of all the chimeras; they all reoxidize to the stable semiquinone state. DISCUSSION We have produced a series of chimeric enzymes by swapping the FMN binding and FAD/NADPH binding subdomains be-tween nNOS and CYPOR. The successful production of enzymes able to reduce cytochrome c indicates that the FMN binding and FAD/NADPH binding subdomains have discrete domain structures that are appropriately folded and similar enough in structure that swapping domains still yields proteins that in each case support electron transfer through the two flavins.
The heme domain-containing constructs are more active than their corresponding reductase domain constructs (NCN versus CN; NCN (tr1) versus CN (tr1)), indicating an influence of the heme domain on folding/stability of the flavin domain. In fact, NCC, in which the entire reductase domain is donated by CYPOR, exhibits a 2-fold increase in both cytochrome c and ferricyanide reduction over that of CYPOR. Such behavior is not observed with nNOS holoenzyme versus the nNOS reductase domain. Thus, it is clear that the presence of the nNOS heme domain has a very dramatic effect on activity in the CYPOR reductase domain, perhaps by altering the angle between the two flavin moieties or by an effect of the heme on the redox potentials of the two flavins.  Interestingly, CaM regulation is completely gone in all chimeras except for NNC, which contains the nNOS FMN binding subdomain and, thus, the autoregulatory insert. A nNOS construct in which the autoregulatory insert has been removed still responds to CaM (9,33), indicating that something more that the autoregulatory insert was involved. The chimera NCN could, thus, potentially still retain modulation by CaM, but it clearly does not. The cytochrome c reduction activities of these chimeras, however, are closer to those observed with wild-type enzyme in the presence of CaM, so the chimeras must be missing those inhibitory elements that CaM relieves to allow electron transfer through the flavin domain. Experiments by Matsuda and Iyanagi (34) demonstrate that CaM activates the transfer of electrons between the two flavin subdomains. This is consistent with our data showing that CaM has no effect on the K cat value of ferricyanide reduction, but a significant effect on the K cat value of cytochrome C reduction by nNOS. Because CaM appears to have no effect on either the fast or slow phases of flavin reduction or on cytochrome c activity in any of the chimeric constructs, it is obvious that the block CaM relieves in the wild-type enzyme reductase domain is not present in the chimeras. Craig et al. (35) propose that this block, which is relieved by CaM binding, is a conformational block imposed by NADPH binding, which restricts the access of external electron acceptors to the FMN. In NNC, which retains the nNOS FMN subdomain, or in NCN and NCN (tr1), which retain the nNOS FAD subdomain and NADPH binding site, or even in nNOS (tr1), which retains all FMN, FAD, and NADPH binding sites from nNOS but lacks the C-terminal tail, no such block appears to exist, indicating that both nNOS subdomains must be intact and present for the NADPH/CaM mediation to occur.
Electron transfer to the heme domain, however, does not occur in NCN, NCN (tr1), or NCC. This may be because of the inability of the FMN subdomain of CYPOR to interact properly with the heme domain of nNOS, as discussed below, but may also be related to the lack of CaM modulation. Sagami et al. (36) propose that the mechanism of activation of nNOS by CaM was not merely dependent on the activation of electron transfer to the heme domain. Because nNOS mutants in which the calmodulin site had been deleted still show heme reduction but not NO production, they suggest that additional structural features in the heme domain must be affected by CaM binding. These structural features are still present, because the heme domains of the chimeras are intact and respond appropriately to substrate. One possibility suggested by these non-functional chimeric proteins is an interaction between the autoinhibitory region, which is not present in the chimeras, and sites on the heme domain.
In addition, little difference exists between the chimeras whether the tail region is present or not (NCN versus NCN (tr1)), even in the flavin reoxidation state. All of the chimeras are able to form a stable semiquinone, unlike nNOS (tr1), whether the FAD binding subdomain is donated by CYPOR or by nNOS, indicating that either the nNOS or the CYPOR FMN domains has the capacity to stabilize the semiquinone form in these chimeras. Thus, in the chimeras, the tail does not obstruct or limit electron flow through the reductase domain even in the absence of CaM, due to either loss of a direct interaction between the tail region and the autoregulatory insert, as proposed by Lane and Gross (21), or an alternate alignment of FMN binding and FAD binding subdomains. Because nNOS still responds to CaM in the absence of its autoregulatory insert, it is most likely that the spatial apposition of the flavins in the chimeric proteins is such that the tail does not obstruct electron flow.
The reductase domain of nNOS shares 58% amino acid sequence similarity with CYPOR, and both CYPOR and the NOS reductase domains can be divided into four subdomains: FMN binding, connecting, FAD binding, and NADPH binding subdomains. In addition to these similarities, significant differences exist between the two. The two most obvious regions of dissimilarity are the autoregulatory insert and the tail region of nNOS, both of which are completely absent in CYPOR. The complete structure of CYPOR has been published (23) but, because the complete structure of the nNOS reductase domain has not yet been reported, no direct comparison can be made. However, Zhang et al. (14) published the structure of a partial nNOS reductase protein consisting of the connecting, FAD binding, and NADPH binding subdomains. They found that the overall polypeptide fold of this protein closely resembled the corresponding regions in CYPOR and that the FAD binding site, the NADPH binding site, and the residues implicated in hydride transfer were strictly conserved between CYPOR and all the NOSs. Given that electron transfer between the FMN and FAD/NADPH binding subdomains occurs readily in the chimeras, particularly when the FAD/NADPH binding subdomain is from CYPOR and the FMN binding subdomain is from nNOS (NNC), this suggests that the electron transfer mechanism of the NOSs is very similar to that of CYPOR.
In the chimeric enzymes, the connecting region between the FMN and FAD binding subdomains is donated by the parent of the FAD binding subdomain. This connecting region shows the least sequence and structural similarity between CYPOR and nNOS (14). The N-terminal 27 residues of the nNOS-connecting region align with 39 residues in CYPOR, making this region 12 residues shorter in nNOS than in CYPOR, and form an ␣-helical structure; in CYPOR, these residues form a loop. Also, amino acids 1077-1091 in nNOS form a ␤-strand finger that may interact with the FMN binding subdomain. They align with a region in CYPOR that is eight residues shorter, and in CYPOR, this region is a short loop.
Zhang et al. (14) also identify a putative FMN domain binding site on the connecting region. In CYPOR, the interaction between the FMN and connecting/FAD binding subdomains is highly electrostatic, with the FMN binding subdomain negatively charged and the connecting/FAD binding subdomain positively charged. In nNOS, the corresponding residues of the connecting region are more neutral, suggesting that the FMN binding subdomain interaction site must also be less charged than in CYPOR (14). The chimeras indicate that the negatively charged FMN binding subdomain is tolerated for interaction with the connecting/FAD binding subdomain but is not tolerated for interaction with the heme domain. Also, the positively charged connecting/FAD binding subdomain of CYPOR is tolerated and in fact performs better than the more neutral nNOS connecting/FAD binding subdomain.
In summary, we have constructed functional chimeric enzymes between nNOS and CYPOR, suggesting that 1) the FMN binding and FAD/NADPH binding subdomains of nNOS and CYPOR are discrete structures that are similar enough in structure to reconstitute electron flow; 2) the electron transfer mechanism of the NOSs is very similar to that of CYPOR; 3) the presence of the nNOS heme domain significantly affects the CYPOR reductase domain; 4) the chimeras are missing the inhibitory elements that CaM relieves; and 5) the tail region of the chimeras does not obstruct electron flow as in wild-type nNOS.
These experiments suggest that regions or subdomains of the nNOS flavoprotein could be uniquely targeted to control electron flux through this NOS isoform. Such experiments also provide insight into the functional relationships among the various interactive subdomains that lead to inhibition or activation of these signaling molecule-producing enzymes.