Drug Metabolism and Disposition Fast Forward
First published on October 8, 2004; DOI: 10.1124/dmd.104.002071
0090-9556/05/3301-10-18$20.00
DMD 33:10-18, 2005
STRUCTURAL AND FUNCTIONAL DIVERSITY IN HEME MONOOXYGENASES
Thomas L. Poulos
Department of Molecular Biology & Biochemistry, Physiology & Biophysics, and Department of Chemistry and the Center in Chemical and Structural Biology, University of California, Irvine, Irvine, California
(Received August 27, 2004;
Accepted October 7, 2004)
 |
Abstract
|
|---|
Recent advances in understanding structure-function relationships in cytochrome P450 (P450), nitric-oxide synthase (NOS), and heme oxygenase are summarized. Of particular importance is the role that dynamics plays in P450 function, where the active site undergoes large open/close motions to enable substrates to bind and products to leave. In sharp contrast, the heme-containing active site of NOS is rigid and remains relatively exposed compared with P450s. This difference in dynamics and active site exposure requires that the O2 activation machinery operate somewhat differently in P450 and NOS. Owing to the open NOS active site, the NOS-oxy complex could be subject to nonspecific protonation that short-circuits the normal reaction path. One working hypothesis holds that NOS recruited the cofactor, tetrahydrobiopterin, to bind near the heme for very rapid coupled electron/proton transfer to the oxy complex, which avoids indiscriminate reaction with bulk solvent. Despite these differences, P450, NOS, and also heme oxygenase use a very similar network of H-bonded water molecules in the active site that are required for oxygen activation. Both P450 and NOS are important drug targets. With NOS, the structural basis for isoform-selective inhibition by a class of dipeptide inhibitors has been worked out, thus providing the basis for structure-based drug design.
I begin this article by thanking those who nominated and selected me for the 2004 Bernard B. Brodie Award in Drug Metabolism from the American Society of Pharmacology and Experimental Therapeutics. In looking at the list of past awardees, I am honored to be included in this group of excellent scientists, many of whom I know and have had the pleasure of working with. The list of awardees also reminded me of one of the main reasons I have remained in the P450 research area for so long. Certainly P450 continues to present surprises and challenges, but just as important are the people in the P450 community. P450 and drug metabolism research has attracted some stellar scientists who also are interesting people, which has continued to make this a fun and productive area of research.
The most generous introduction by Dr. Michael Franklin to my Brodie Award Lecture stressed the importance of the first P450 crystal structure which, of course, was P450cam. Achieving this goal was due to many people, with I. C. Gunsalus (Gunny) heading the list as well as Gerry Wagner, who played a critical role in obtaining the first P450cam crystals. Gunny took a chance with me early on and, after some early successes and setbacks, the structure of P450cam was solved in the early 1980s. I will always be grateful for Gunny's enthusiasm, support, and even occasional chiding. More recently, the structures of the other two redox components of the P450cam system (Fig. 1) have been solved in our lab at UCI (University of California, Irvine): putidaredoxin (Sevrioukova et al., 2003
) and putidaredoxin reductase (Sevrioukova et al., 2004). Given the historical importance of P450cam, it is fitting that the first complete structure determination of all components for a bacterial P450 monooxygenase is the camphor monooxygenase system.
Since the first description of P450cam back in 1968 (Katagiri et al., 1968), P450cam has served as the paradigm in P450 structure-function studies. P450cam, however, is no longer alone. After the initial P450cam structure paper in 1985 (Poulos et al., 1985), there was no new structure until P40BM3 (Boddupalli et al., 1992; Ravichandran et al., 1993), followed by P450terp (Boddupalli et al., 1992) and P450eryF (Cupp-Vickery and Poulos, 1995). In addition to the first structure of a P450 redox complex (Sevrioukova et al., 1999), the next major step was the first structure determination of a membrane-bound P450 (Williams et al., 2000). With increasing advances in cloning, expression, purification, and more user-friendly crystallographic software and hardware, the number of new P450 structures has dramatically increased, and as of this writing there are 20 unique P450 structures on deposit in the Protein Data Base. In addition, the structures of other heme monooxygneases have been solved, which include the heme domain for all three mammalian isoforms of nitric-oxide synthase (NOS) (Crane et al., 1998; Raman et al., 1998; Fischmann et al., 1999; Li et al., 1999, 2002), as well as both mammalian (Schuller et al., 1999; Sugishima et al., 2002) and bacterial heme oxygenases (HOs) (Schuller et al., 2001; Hirotsu et al., 2003; Friedman et al., 2004). The connection between NOS and HO with P450 is clearly evident based on the similar reactions and the requirement for NADPH via a diflavin protein similar to P450 reductase (Fig. 2). This wealth of structural information has made it possible to understand the structural basis for substrate binding and specificity in P450s and has begun to reveal the critical role that dynamics plays in catalysis. A comparison between the P450, NOS, and HO systems also has revealed important similarities and differences in the O2 activation and hydroxylation chemistries. The article that follows focuses on three key aspects of these enzyme systems: 1) the role played by conformational dynamics, 2) similarities and differences in the O2 activation machinery, and 3) the prospects for structure-based inhibitor/drug design.
View larger version (37K):
[in this window]
[in a new window]
|
FIG. 2. The structures and reactions catalyzed by the three monooxygenases under consideration: P450, NOS, and HO.
|
|
|
Conformational Dynamics
|
|---|
Open and Closed P450 Structures. One puzzling problem that was apparent from the P450cam structure is how the substrate gets into the active site and how product leaves. A simple examination of a space-filled model of P450cam (Fig. 3) shows that the substrate is buried, so there must be an open/close motion to enable the substrate to penetrate into the active site and bind near the heme iron. Unfortunately, the crystal structure of substrate-free P450cam (Poulos et al., 1986) did not reveal any major differences except for thermal parameters of residues forming a small opening above the substrate. It was hypothesized that these high thermal factor regions must open up to allow substrate to enter. The types of motions required were not evident until the P450BM3 structure. This P450 was initially crystallized in the substrate-free form (Ravichandran et al., 1993), which showed an open active site that could easily accommodate a fatty acid substrate (Fig. 4). Subsequent computational studies illustrated that the active site can close and that crystal packing contacts had trapped P450BM3 in the open conformation (Li and Poulos, 1995). The structure of P450BM3 complexed with a fatty acid (Li and Poulos, 1997; Haines et al., 2001) showed that the F and G helices move as a unit to provide the opening and closing of the active site entryway. Quite interestingly, the computational results closely matched the experimental results. These results also revealed a sobering experimental limitation. P450BM3 had been locked in the open conformation owing to crystal contacts. The cynic might call this an artifact of crystallization, but the fact is that the forces that hold crystals together are quite weak, and crystallization cannot force a protein to adopt a conformation that is not accessible in solution. This is why deoxy hemoglobin crystals crack when oxygenated, as do P450BM3 substrate-free crystals when fatty acids are added. The conformational changes required to bind substrate are not compatible with the crystal lattice. This means that we are somewhat at the whims of Nature on precisely what form of substrate-free P450 crystals we grow. In many ways, we were lucky with P450BM3. Perhaps the most spectacular example of how conformational state and crystallization are linked is P4502B4, where the ligand-free and -bound structures illustrate a huge range of motions, with some residues being repositioned by as much as 18 Å (Scott et al., 2003, 2004).
View larger version (117K):
[in this window]
[in a new window]
|
FIG. 3. A space-filled model of P450cam viewed along the distal surface. The F and G helices are in magenta and the substrate, camphor, is red. Note that the camphor is barely visible through a small opening in the presumed substrate access channel.
|
|
View larger version (43K):
[in this window]
[in a new window]
|
FIG. 4. Space-filled models of P450BM3 in the open and closed conformation. The heme is red and the fatty acid substrate is blue. For the energy-minimized structure, the substrate-free structure was first energy-minimized only, with no molecular dynamics, followed by docking of the fatty acid substrate (Li and Poulos, 1995). Note that the minimized structure closely resembles the crystal structure of substrate-bound P450BM3.
|
|
Induced Fit. The P450BM3 and P4502B4 structures illustrate that the basic P450 fold is very flexible. One possible reason why such flexibility is required is the necessity to sequester the O2 activation machinery from bulk solvent. Without protection from bulk solvent, the P450 catalytic cycle would be subject to indiscriminate protonation of the iron-linked O2 molecule, leading to the nonproductive release of H2O2 rather than substrate hydroxylation. A second reason is the well known lack of specificity among many P450s. Many drug-metabolizing P450s are able to hydroxylate a wide range of substrates. These P450s may be able to adapt to substrates of different sizes not by simply opening and closing but by adapting the shape of the active site to the bound substrate. If so, then we might expect the same P450 to exhibit different active site structures, depending on the ligand bound.
One of the best illustrations of such adaption derives from a most unlikely source, CYP119, from the thermophilic archaeon Sulfolobus solfataricus. The primary interest in this P450 is its remarkable thermal stability and the potential for engineering such stability into P450s for industrial applications. Exploiting the thermal stability properties of CYP119 has yet to be realized, nor is the function of CYP119 known. However, CYP119 does provide an interesting example of induced fit. The initial structure was solved with two different ligands coordinated to the heme iron: 4-phenylimidazole and imidazole (Yano et al., 2000). When the smaller imidazole is bound, the C-terminal end of the F helix "unfolds," resulting in a longer F/G loop that then can dip into the active site and interact with ligand (Fig. 5). This enables Leu155 to directly contact the imidazole. The energetic cost of losing helical H-bonds in the imidazole complex is partially compensated for by a new ion pair between Arg154 and Glu212 (Fig. 5). Thus, the CYP119 active site adapts itself to the ligand bound, and new intramolecular protein contacts balance the energetic cost of those contacts lost during the conformational transition. This could well foreshadow the sorts of rearrangements that occur in drug-metabolizing P450s that are able to oxidize a wide variety of compounds. The implications of such change are well outside the typical view of an enzyme adopting only two states: substrate -free and -bound or even the traditional induced fit hypothesis. Instead, some P450s may be far more chameleon-like in their flexibility and adapt a variety of active site shapes, depending on the ligand bound. The challenge is to be able to understand what features of the P450 fold and the various interactions between structural elements enable such transitions to occur.
View larger version (29K):
[in this window]
[in a new window]
|
FIG. 5. Comparison of CYP119 with imidazole (red) and 4-phenylimidazole (blue) bound to the heme iron. In the imidazole complex the C-terminal end of the F helix unwraps, which allows Arg154 to swing down and form an ion pair with Glu212. Leu155 also is free to move close to and interact with the imidazole ligand.
|
|
The Rigidity of NOS. NOS is a typical P450 monooxygenase in the first step of the reaction, which is hydroxylation of L-Arg (Fig. 2). The architecture of NOS is similar to that of P450BM3 in having the FAD/FMN P450 reductase-like domain attached to the C-terminal domain, giving the following modular enzyme: heme-FMN-FAD. As in P450s, the electron flow is NADPH-FAD-FMN-heme. NOS forms a homodimer through the heme domains, and a majority of the structural work to date has been on the dimeric heme domains. As shown in Fig. 6 the substrate is positioned over the heme near the iron, similar to how P450s position substrates for regioselective hydroxylation. Also, like P450, Cys serves as the proximal heme ligand. This, however, is where the similarity ends. The overall architecture of NOS is totally different from that of P450 (Fig. 1). Most notable is that the active site is open in NOS, and there is little indication that the NOS heme domain undergoes the types of structural changes observed in P450s. It has been possible to cocrystallize NOS with a large number of inhibitors and ligands, and the only changes in structure observed are in the repositioning of side chains (Raman et al., 2001). Part of the rigidity in NOS is due to beta structure serving as the primary active site architectural motif, rather than helices as in P450s. In addition, the heme domain forms a very tight dimer near the active site. The ability of helices to slide over one another is a key feature in P450 flexibility, whereas sheet structures are more rigid. Possible reasons for these striking differences between NOS and P450 are linked to the O2 activation chemistry and electron transfer, to be considered in the next section.
View larger version (28K):
[in this window]
[in a new window]
|
FIG. 6. Active site structure of eNOS. The conserved Glu363 H-bonds with the substrate, which helps to position and hold the L-Arg (yellow) N atom to be hydroxylated (indicated by arrow) close to the iron for hydroxylation by the Fe(IV)=O intermediate. The cofactor (BH4 in yellow) H-bonds with the same propionate that H-bonds with the substrate  -amino group.
|
|
|
Oxygen Activation
|
|---|
Introduction. The activation of O2 must, in principle, progress through several intermediates (Scheme 1). Much of the effort in understanding this process has been devoted to how protons are delivered to the oxy complex. This is required to give heterolytic cleavage of the O-O bond from the dihydroperoxy intermediate to leave behind Fe(III)-O or its electronic equivalent, which is an O atom with only six valence electrons, a potent oxidant. The proper delivery of protons is important to avoid uncoupling, which normally involves the release of peroxide. This could occur if the Fe(III)-O-OH- were improperly protonated to give
Peroxidases carry out a similar reaction and, in this case, an active site His residue shuttles protons to the leaving oxygen atom to facilitate O-O bond cleavage (Poulos and Kraut, 1980). It was a surprise to find no similar residue in P450cam. Instead, it appears that a necklace of ordered solvent molecules provides an H-bonded network that delivers protons to the oxy complex (Fig. 7). NOS and HO are very similar since here, too, there is no amino acid side chain suitably positioned to deliver protons to the oxy complex. The nitric oxide complexes of both NOS (Li et al., 2001) and HO (Lad et al., 2003) show a network of H-bonded solvent very similar to P450, which clearly implicates a very similar solvent proton network operating in all three enzyme systems.
View larger version (23K):
[in this window]
[in a new window]
|
FIG. 7. -The P450cam-oxy complex. This structure is based on the 1.8Å structure determined in our laboratory (S. Nagano and T. C. Poulos, unpublished) and is very similar to the structure reported by Schlichting et al. (2000) with the exception that occupancy for O2 is higher in our more recent structure.
|
|
O2 Activation Intermediates. Despite such similarities, there are important differences. For example, the various intermediates involved in O2 activation that can be trapped for spectroscopic studies differ. Cryogenic reduction methods have enabled spectroscopic identification of intermediates in both NOS and P450 between the initial oxy, Fe(II)-O-O, complex and product (Davydov et al., 2001, 2002). In P450, it is possible to observe the peroxy, Fe(III)-O-O-2, and hydroperoxy, Fe(III)-O-OH-, intermediates. With NOS, however, only the Fe(III)-O-O-2 intermediate can be trapped. For both proteins it is assumed that a second proton must be added to Fe(III)-O-OH-, giving Fe(III)-O-OH2, leading to cleavage of the O-O bond. This leaves Fe(III)-O or it electronic equivalent, Fe(IV)=O, as the active hydroxylating species. Since neither the Fe(III)-O-OH2 nor Fe(III)-O is observed, the second protonation and O-O bond cleavage steps followed by substrate hydroxylation must be very rapid with, perhaps, protonation and O-O bond cleavage being concerted.
A key difference between NOS and P450 in the O2 activation step is that NOS requires the cofactor tetrahydrobiopterin or BH4 (Tayeh and Marletta, 1989). It now appears that the function of BH4 is to donate an electron to the oxy complex and, hence, forms a short-lived pterin radical (Hurshman et al., 1999). To carry out the cryo-trapping experiments, it was necessary to replace BH4 with an inactive pterin. Davydov et al. (2002) argued that without the coupled proton-electron transfer capability of BH4, only the electron is delivered to the oxy complex and, thus, NOS becomes "stuck" in the Fe(III)-O-O-2 state. A similar role for BH4 serving as an electron and proton donor has been proposed by Sørlie et al. (2003). With P450s, the electron is delivered from the FMN of P450 reductase and the current structural information indicates that the FMN-to-heme distance is on the order of 18 Å (Fig. 8) (Sevrioukova et al., 1999), whereas the proton is delivered from a nearby active site group, most likely water. As shown in Figs. 6 and 8, the BH4 directly H-bonds to the heme propionate so that electron transfer to the heme iron requires a short through-bond distance. Figure 9 presents a possible mechanism for O2 activation in NOS. Whether or not the protons derive from water and/or BH4, it is clear that electron and proton transfer must be close together in space and, hence, time in NOS, whereas in P450s, the electron donor is far from the proton donor. Note, too, that the L-Arg guanidinium N atom that is hydroxylated could also serve as a source of one proton.
View larger version (39K):
[in this window]
[in a new window]
|
FIG. 8. The NOS active site and the structure of the electron transfer complex formed by the FMN and heme domains of P450BM3 (Sevrioukova et al., 1999). The distance between the FMN and heme is 17 to 18 Å compared with the direct H-bonded link between the electron donating BH4 cofactor in NOS and the heme.
|
|
View larger version (16K):
[in this window]
[in a new window]
|
FIG. 9. Possible NOS O2 activation mechanism. A critical difference between NOS and P450 is that in NOS, BH4 delivers the electron required to activate the oxy complex. The precise proton donors shown are speculative and may or may not involve proton donation by the substrate, L-Arg, or coupled proton-electron transfer from BH4. A tightly coupled proton/electron donation ensures that there is no buildup of the Fe(III)-OOH- hydroperoxide intermediate that could potentially be released as H2O2 by solvent donating a proton to the iron-linked O atom resulting in Fe-O bond cleavage.
|
|
The Role of BH4 in NOS Catalysis. Left open is the question of why NOS uses BH4 at all. Why not simply use the P450-like electron transfer process? The answer could relate to another stark difference between P450 and NOS noted earlier, which is that NOS has an exposed active site and does not undergo the sorts of larger scale conformational changes observed in P450s. To avoid uncoupling and H2O2 release, it is necessary to protect the P450 O2 activation machinery from bulk solvent. This is one reason the active site in P450s must undergo an open/close motion, which effectively buries the iron-linked O2. NOS, however, must have an open active site for two reasons. First, the substrate, L-Arg, is a polar molecule. Sequestering such a polar molecule within a buried active site would be energetically costly. Second, and more important, is that the product, NO, must be released to serve its function as a biological signaling molecule. An open active site ensures release of NO, whereas burial would trap the NO within the active site pocket. Such exposure increases the chances of uncoupling by bulk solvent. To circumvent this problem NOS recruited BH4 as a nearby electron transfer cofactor to give a rapid and coupled proton-electron electron transfer system. This prevents the buildup of Fe(III)-OOH-, which is subject to uncoupling. Thus, NOS proceeds from Fe(III)-O-O-2 to Fe(III)-OOH2 by rapid coupled proton-electron transfer followed by rapid cleavage of the O-O bond.
Heme Oxygenase. HO presents an interesting twist on P450 O2 activation chemistry. Rather than the iron-linked oxygen serving as the source of the OH group in the product, it is the distal O atom that ends up in the product, -hydroxy-heme (Fig. 10) (Ortiz de Montellano, 1998). Why, then, do P450, NOS, and peroxidases cleave the O-O heterolytically while HO does not? One important difference is that in HOs, diatomic ligands like NO and CO are forced to bend toward the -meso heme carbon owing to steric crowding by the distal helix. The most relevant, of course, is the O2 complex, which has a remarkable Fe-O-O bend angle of 110°, which places the distal O atom about 3.3Å from -meso heme carbon (Unno et al., 2004). Elegant kinetic isotope experiments indicate the mechanism outlined in Fig. 10 (Davydov et al., 2003). Unlike P450s, the proximal O atom linked to the iron is protonated. In P450s this would lead to uncoupling and H2O2 release, but because the distal O atom is so close to the -meso heme carbon, the electrophilically activated peroxide reacts with the -meso heme carbon. Mutations within the HO active site pocket can lead to peroxidase-like activities owing to heterolytic cleavage of the O-O bond, thus giving the Fe(IV) = O intermediate (Liu et al., 2000; Lightning et al., 2001). The structural effects of one of these mutations in human heme HO1, Asp140Ala, are small, leading to subtle changes in the intricate H-bonding network involving active site solvent (Lad et al., 2003). It thus appears that steric factors and H-bonding networks are responsible for tipping the balance in favor of electrophilic attack on the -meso heme carbon by the iron-linked peroxide rather than O-O bond cleavage. In summary, the common theme that links the O2 activation process in P450, NOS, and HO is water. None of these enzymes use an amino acid side chain as the direct proton donor, as do peroxidases and many other enzymes. Instead, an intricate network of H-bonded solvent appears to be crucial in delivering solvent protons to the iron-linked oxygen molecule.
Structure-Based Inhibitor Design. All three enzymes under consideration are to varying degrees potential drug targets. Knowing the crystal structure of a druggable target should, in principle, greatly aid and speed the search for candidate therapeutic agents. Indeed, in those drug discovery projects where the emphasis has been placed, early on, in structure determination, structure-based approaches have proven quite successful. The classic example is the currently used HIV protease inhibitors. HIV protease was identified as a viable drug target in 1988, and by 1996, three drugs were on the market (http://www.nigms.nih.gov/news/science_ed/structlife/chapter4.html). The crystal structure was solved in 1989 (Miller et al., 1989) and played an early and important role in the rapid development of antiprotease drugs (Wlodawer and Erickson, 2003).
P450s also are the target for various therapeutic agents, with the most well known being antifungal agents. With an increasing number of groups involved in successful structure determination of membrane-bound P450s, there is renewed optimism that the structure of most of the important P450 drug targets will be solved in the not-too-distant future. Despite such optimism, one of the more fascinating aspects of P450 function, its flexibility, may make structure-based inhibitor design more difficult. The simplest drug design projects are those in which the enzyme is relatively inflexible and there is little difference in the inhibitor-free and -bound structures. It then becomes possible to solve the structure of many complexes in the same crystal form and, hence, develop a large structural database critical in inhibitor design. Such rapid structure determinations may prove more difficult with drug-metabolizing P450s, since flexibility and adaption may require cocrystallization in different crystal forms. In contrast, specific P450s, such as those involved in steroid metabolism that operate on only one or a limited number of similar substrates, may not be as adaptable. The main limitation here will be solubility of steroids at the concentrations required for crystallography.
NOS, however, has none of these limitations. As we have seen, NOS is rigid compared with P450s and does not undergo large open/close motions. The NOS active site is open and remains open when inhibitors and substrates bind, with major changes occurring only in the position of some side chains (Raman et al., 2001). Thus, it has been possible to solve the structure of dozens of complexes in the same crystal form, owing to the basic lock-and-key mode of binding. Moreover, most NOS inhibitors that mimic the properties of the NOS substrate, L-Arg, are quite soluble.
Unlike P450s, where there are many isoforms, there are only three mammalian NOS isoforms with nearly identical active site structures. In one respect, this makes the problem of selective inhibitor design easier than with P450s, since there are far fewer isoforms to compare. However, because the active sites are so similar, there are few obvious structural differences that can be exploited for inhibitor design. This makes the development of isoform-selective NOS inhibitors a challenging problem. The three main NOS isoforms participate in distinct physiological processes: endothelial NOS (eNOS; blood pressure regulation), neuronal NOS (nNOS; neural transmission), and inducible NOS (iNOS; immune system). The reason for targeting NOS is that the over- and underproduction of NO is associated with various pathological states including hypertension, septic shock, stroke, and degenerative neural diseases (Bredt and Snyder, 1994). Diminished levels of NO contribute to chronic hypertensive diseases. In contrast, NO overproduction contributes to pathological conditions related to primary neurodegenerative, inflammatory, and vascular disorders (Gross and Wolin, 1995). In septic shock, NOS inhibitors can restore vascular tone and blood pressure (Kilbourn and Griffith, 1992). Blocking NO production also limits ischemia-elicited infarct size in animal models (Patel et al., 1993), and NO can act as a neurotoxic agent in Alzheimer's disease (Law et al., 2001). NO has been found to stimulate breast cancer tumor growth (Alalami and Martin, 1998), whereas NOS inhibitors have been shown to block tumor growth (Iwasaki et al., 1997). The culprit in many of these cases appears to be iNOS-generated NO. Thus, it is desirable to block the cytotoxic effects of iNOS-generated NO but not the beneficial NO production of eNOS-generated NO involved in blood pressure regulation.
The main problem with isoform-selective NOS inhibitors is that a majority are similar in structure to L-Arg, and although many bind tightly, they exhibit little discrimination. This is gradually changing, and our own efforts have been directed toward dipeptide inhibitors developed in the laboratory of Rick Silverman at Northwestern University (Huang et al., 1999) (Fig. 11). These inhibitors are based on L-nitro-Arg, which binds the same to all NOS isoforms (Fig. 12). Whereas L-nitro-Arg is not specific, dipeptide inhibitors based on the L-nitro-Arg skeleton exhibit up to 2600-fold selectivity (Fig. 11). To understand the structural basis for such selectivity, the structures of eNOS and nNOS complexed with these inhibitors have been determined (Flinspach et al., 2004). The discussion will be limited to DIP1 (Fig. 11) since the results are very much the same for the other inhibitors.
View larger version (23K):
[in this window]
[in a new window]
|
FIG. 12 Superimposition of the L-nitro-Arg (magenta) structure on the L-Arg (yellow) structure. They are nearly identical with the exception of the nitro group in L-nitroArg, which makes favorable electrostatic contacts with the peptide backbone. This helps to explain the higher affinity of L-nitro-Arg for NOS over L-Arg.
|
|
The structures of eNOS and nNOS complexed with DIP1 show clear differences in the conformation of the inhibitor (Fig. 13). In nNOS, the inhibitor is "curled," whereas in eNOS it adopts an extended conformation. A detailed comparison of the structures (Fig. 14) suggests a structural basis for this difference. In nNOS, the curled conformation allows the -amino group of the inhibitor to directly H-bond with Glu592. In eNOS, the extended conformation places the -amino group too far for direct H-bonding to the homologous Glu363 and, instead, a water molecule forms an H-bonding bridge. The main reason for this difference is due to one amino acid difference: where eNOS has Asn368, nNOS has Asp597. This means that nNOS has two negatively charged side chains near the inhibitor -amino group, Asp597 and Glu592, whereas eNOS has only one, Glu363. As a result, the inhibitor adopts the curled conformation in nNOS, which optimizes interactions between the inhibitor -amino group and the two active site carboxylate groups. An energetic analysis of the crystal structures illustrates that the inhibitor in nNOS is, indeed, electrostatically more stable. The nonbonded energy of interaction between the inhibitor and protein is -30.4 kcal/mol in eNOS and -28.2 kcal/mol in nNOS, whereas the electrostatic stabilization is -34.6 kcal/mol in eNOS and -70.9 kcal/mol in nNOS. A dissection of the electrostatic interactions shows that the main contribution is between the inhibitor -amino group and the nearby Asp597 and Glu592.
View larger version (34K):
[in this window]
[in a new window]
|
FIG. 13 Electron density maps of the DIP1 inhibitor bound to eNOS and nNOS. Note that DIP1 adopts a curled conformation in nNOS but an extended conformation in eNOS.
|
|
To further test the hypothesis that the Asp/Asn difference is the main factor controlling isoform selectivity, the reciprocal mutants were made, crystal structures solved, and Ki values measured (Fig. 15). The inhibitor adopts the eNOS-like extended conformation in the nNOS Asp-to-Asn mutant, whereas it adopts the curled conformation in the eNOS Asn-to-Asp mutant. Moreover, the Ki drops in the eNOS mutant and increases in the nNOS mutant (Fig. 15). Using this information for the design of new inhibitors assumes that the energetic cost of forcing the inhibitor into the "curled" conformation is unfavorable but is compensated for by the favorable electrostatic interactions between the inhibitor -amino group and the Asp and Glu in the nNOS active site. Removing the need to curl the inhibitor by placing the amino group on a rigid scaffold should give an even more selective inhibitor. This design principle now is being used to develop a new generation of NOS inhibitors.
View larger version (46K):
[in this window]
[in a new window]
|
FIG. 15 Electron density maps of DIP1 bound to wild-type and mutant eNOS and nNOS. Note that DIP1 adopts the curled conformation with a low Ki in the eNOS mutant, whereas DIP1 adopts the extended conformation with a higher Ki in the nNOS mutant.
|
|
|
Conclusions
|
|---|
The ever-increasing structural database for P450s has enabled a reasonably good understanding on P450 substrate specificity and has revealed the remarkable flexibility of the P450 fold. It remains to be seen whether a single P450 can acquire a diverse range of active site conformations to accommodate substrates of quite different sizes and shapes. This could be called the "multi-induced-fit" hypothesis. With respect to O2 activation, P450, together with heme oxygenase and nitric-oxide synthase, illustrates a surprisingly consistent pattern of solvent H-bonding networks as being critical to the O2 protonation and activation. Finally, nitric-oxide synthase has proven to be an excellent system for working out the structural basis for inhibitor selectivity. It remains to be seen whether or not this information can be exploited in the design of isoform-selective inhibitors and drugs, but the level of optimism is high.
|
Acknowledgments
|
|---|
I want to thank the many current and former members of the laboratory who have contributed to work summarized in this article, including Huiying Li, Irina Sevrioukova, Shingo Nagano, David Schuller, Latesh Lad, and C. S. Raman. I also am indebted to some outstanding collaborators and friends including Paul Ortiz de Montellano, Bettie Sue Masters, and Richard Silverman.
|
Footnotes
|
|---|
Work in the Poulos laboratory was supported by National Institutes of Health Grants GM33688 and GM57353.
doi:10.1124/dmd.104.002071.
ABBREVIATIONS: P450, cytochrome P450; NOS, nitric-oxide synthase; HO, heme oxygenase; BH4, tetrahydrobiopterin; eNOS, endothelial NOS; nNOS, neuronal NOS; iNOS, inducible NOS; NO, nitric oxide.
Thomas L. Poulos received a Ph.D. in Biology from the University of California, San Diego in 1972. During his postdoctoral work in the Chemistry Department at UCSD, he solved the first crystal structure of a heme enzyme, cytochrome c peroxidase. After his postdoctoral work, Dr. Poulos held the position of Director of Protein Engineering at Genex Corp. in Gaithersburg, Maryland and, later, Director of the University of Maryland's Center for Advanced Research in Biotechnology and Professor of Chemistry at the College Park Campus of the University of Maryland, where he was awarded the Presidential Meritorious Service Award. While at Genex, Dr. Poulos completed work on the first structure determination of a cytochrome P450, P450cam. In 1992 he was recruited by the University of California, Irvine, where he now holds the position of Chancellor's Professor of Biochemistry, Biophysics, and Chemistry. His current work centers on structure-function relationships in heme enzymes and related proteins including nitric-oxide synthase, P450s, and heme oxygenase.
Address correspondence to: Thomas L Poulos, Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA 92697-3900. E-mail: poulos{at}uci.edu
|
References
|
|---|
Alalami O and Martin JHJ (1998) ZR-75-1 human breast cancer cells: expression of inducible nitric oxide synthase and effect of tamoxifeb and phorbol ester on nitric oxide production. Cancer Lett 123: 99-105.[CrossRef][Medline]
Boddupalli SS, Hasemann CA, Ravichandran KG, Lu JY, Goldsmith EJ, Deisenhofer J, and Peterson JA (1992) Crystallization and preliminary x-ray diffraction analysis of P450terp and the hemoprotein domain of P450BM-3, enzymes belonging to two distinct classes of the cytochrome P450 superfamily. Proc Natl Acad Sci USA 89: 5567-5571.[Abstract/Free Full Text]
Bredt DS and Snyder SH (1994) Nitric oxide: a physiological messenger molecule. Annu Rev Biochem 63: 175-195.[CrossRef][Medline]
Crane BR, Arvai AS, Ghosh DK, Wu C, Getzoff ED, Stuehr DJ, and Tainer JA (1998) Structure of nitric oxide synthase oxygenase dimer with pterin and substrate. Science (Wash DC) 279: 2121-2126.[Abstract/Free Full Text]
Cupp-Vickery JR and Poulos TL (1995) Structure of cytochrome P450eryF involved in erythromycin biosynthesis. Nat Struct Biol 2: 144-153.[CrossRef][Medline]
Davydov R, Ledbetter-Rogers A, Martasek P, Larukhin M, Sono M, Dawson JH, Masters BS, and Hoffman BM (2002) EPR and ENDOR characterization of intermediates in the cryoreduced oxy-nitric oxide synthase heme domain with bound L-arginine or N(G)hydroxyarginine. Biochemistry 41: 10375-10381.[CrossRef][Medline]
Davydov R, Makris TM, Kofman V, Werst DE, Sligar SG, and Hoffman BM (2001) Hydroxylation of camphor by reduced oxy-cytochrome P450cam: mechanistic implications of EPR and ENDOR studies of catalytic intermediates in native and mutant enzymes. J Am Chem Soc 123: 1403-1415.[CrossRef][Medline]
Davydov R, Matsui T, Fujii H, Ikeda-Saito M, and Hoffman BM (2003) Kinetic isotope effects on the rate-limiting step of heme oxygenase catalysis indicate concerted proton transfer/heme hydroxylation. J Am Chem Soc 125: 16208-16209.[CrossRef][Medline]
Fischmann TO, Hruza A, Niu XD, Fossetta JD, Lunn CA, Dolphin E, Prongay AJ, Reichert P, Lundell DJ, Narula SK, and Weber PC (1999) Structural characterization of nitric oxide synthase isoforms reveals striking active-site conservation. Nat Struct Biol 6: 233-242.[CrossRef][Medline]
Flinspach M, Li H, Jamal J, Yang W, Huang H, Hah J-M, Gomez-Vidal JA, Litzinger EA, Silverman RB, and Poulos TL (2004) Structural basis for dipeptide amide isoform-selective inhibition of neuronal nitric oxide synthase. Nat Struct Mol Biol 11: 54-59.[CrossRef][Medline]
Friedman J, Lad L, Li H, Wilks A, and Poulos TL (2004) Structural basis for novel deltaregioselective heme oxygenation in the opportunistic pathogen Pseudomonas aeruginosa. Biochemistry 43: 5239-5245.[CrossRef][Medline]
Gross SG and Wolin MS (1995) Nitric oxide: pathophysiological mechanisms. Annu Rev Physiol 57: 737-769.[CrossRef][Medline]
Haines DC, Tomchick DR, Machius M, and Peterson JA (2001) Pivotal role of water in the mechanism of P450BM-3. Biochemistry 40: 13456-13465.[CrossRef][Medline]
Hirotsu S, Chu GC, Unno M, Lee D-S, Yoshida T, Park S-Y, Shiro Y, and Ikeda-Saito M (2003) The crystal structures of the ferric and ferrous forms of the heme complex of HmuO, a heme oxygenase of Corynebacterium diphtheriae. J Biol Chem 279: 11937-11947.
Huang H, Martasek P, Roman LJ, Masters BS, and Silverman RB (1999) N(omega) Nitroarginine-containing dipeptide amides. Potent and highly selective inhibitors of neuronal nitric oxide synthase. J Med Chem 42: 3147-3153.[CrossRef][Medline]
Hurshman AR, Krebs C, Edmondson DE, Huynh BH, and Marletta MA (1999) Formation of a pterin radical in the reaction of the heme domain of inducible nitric oxide synthase with oxygen. Biochemistry 38: 15689-15696.[CrossRef][Medline]
Iwasaki T, Higashiyama M, Kuriyama K, Sasaki A, Mukai M, Shinkai K, Horai T, Matsuda T, Matsuda H, and Akedo H (1997) NG-nitro-L-arginine methyl ester inhibits bone metastasis after modified intracardiac injection of human breast cancer cells in a nude mouse model. Jpn J Cancer Res 88: 861-866.[Medline]
Katagiri M, Ganguli BN, and Gunsalus IC (1968) A soluble cytochrome P-450 functional in methylene hydroxylation. J Biol Chem 243: 3543-3546.[Abstract/Free Full Text]
Kilbourn RG and Griffith OW (1992) Overproduction of nitric oxide in cytokine-mediated and septic shock. J Natl Cancer Inst 84: 1671-1672.[Free Full Text]
Lad L, Wang J, Li H, Friedman J, Bhaskar B, Ortiz de Montellano PR, and Poulos TL (2003) Crystal structures of the ferric, ferrous and ferrous-NO forms of the Asp140Ala mutant of human heme oxygenase-1: catalytic implications. J Mol Biol 330: 527-538.[CrossRef][Medline]
Law A, Gauthier S, and Quirion R (2001) Say NO to Alzheimer's disease: the putative links between nitric oxide and dementia of the Alzheimer's type. Brain Res Rev 35: 73-96.[CrossRef][Medline]
Li H and Poulos TL (1995) Modeling protein substrate interactions in the heme domain of cytochrome P450BM-3. Acta Crystallogr D 51: 21-32.[CrossRef][Medline]
Li H and Poulos TL (1997) The structure of the cytochrome P450BM-3 haem domain complexed with the fatty acid substrate, palmitoleic acid. Nat Struct Biol 4: 140-146.[CrossRef][Medline]
Li H, Raman CS, Glaser CB, Blasko E, Young TA, Parkinson JF, Whitlow M, and Poulos TL (1999) Crystal structures of zinc-free and -bound heme domain of human inducible nitric oxide synthase and a comparison with endothelial nitric oxide synthase. J Biol Chem 274: 21276-21284.[Abstract/Free Full Text]
Li H, Raman CS, Martasek P, Masters BS, and Poulos TL (2001) Crystallographic studies on endothelial nitric oxide synthase complexed with nitric oxide and mechanism-based inhibitors. Biochemistry 40: 5399-5406.[CrossRef][Medline]
Li H, Shimizu H, Flinspach M, Jamal J, Yang W, Xian M, Cai T, Wen EZ, Jia Q, Wang PG, and Poulos TL (2002) The novel binding mode of N-Alkyl-N'-hydroxyguanidine to neuronal nitric oxide synthase provides mechanistic Insights into NO biosynthesis. Biochemistry 41: 13868-13875.[CrossRef][Medline]
Lightning LK, Huang H, Moenne-Loccoz P, Loehr TM, Schuller DJ, Poulos TL, and de Montellano PR (2001) Disruption of an active site hydrogen bond converts human heme oxygenase-1 into a peroxidase. J Biol Chem 276: 10612-10619.[Abstract/Free Full Text]
Liu Y, Lightning LK, Huang HW, Moenne-Loccoz P, Schuller DJ, Poulos TL, Loehr TM, and Ortiz de Montellano PR (2000) Replacement of the distal glycine 139 transforms human heme oxygenase-1 into a peroxidase. J Biol Chem 275: 34501-34507.[Abstract/Free Full Text]
Miller M, Schneider J, Sathyanarayana BK, Toth MV, Marshall GR, Clawson L, Selk L, Kent SB, and Wlodawer A (1989) Structure of complex of synthetic HIV-1 protease with a substrate-based inhibitor at 2.3 A resolution. Science (Wash DC) 246: 1149-1152.[Abstract/Free Full Text]
Ortiz de Montellano PR (1998) Heme oxygenase mechanism: evidence for an electrophilic ferric peroxide species. Acc Chem Res 31: 543-549.[CrossRef]
Patel VC, Yellon DM, Singh KJ, Neild GH, and Woolfson RG (1993) Inhibition of NO limits infarct size in the in situ rabbit heart. Biochem Biophys Res Commun 194: 234-238.[CrossRef][Medline]
Poulos TL, Finzel BC, Gunsalus IC, Wagner GC, and Kraut J (1985) The 2.6-Å crystal structure of Pseudomonas putida cytochrome P-450. J Biol Chem 260: 16122-16130.[Abstract/Free Full Text]
Poulos TL, Finzel BC, and Howard AJ (1986) Crystal structure of substrate-free Pseudomonas putida cytochrome P450. Biochemistry 25: 5314-5322.[CrossRef][Medline]
Poulos TL and Kraut J (1980) The stereochemistry of peroxidase catalysis. J Biol Chem 255: 8199-8205.[Free Full Text]
Raman CS, Li H, Martasek P, Kral V, Masters BSS, and Poulos TL (1998) Crystal structure of constitutive endothelial nitric oxide synthase: a paradigm for pterin function involving a novel metal center. Cell 95: 939-950.[CrossRef][Medline]
Raman CS, Li H, Martasek P, Southan G, Masters BSS, and Poulos TL (2001) Crystal structure of nitric oxide synthase bound to nitro indazole reveals a novel inactivation mechanism. Biochemistry 40: 13448-13455.[CrossRef][Medline]
Ravichandran KG, Boddupalli SS, Hasermann CA, Peterson JA, and Deisenhofer J (1993) Crystal structure of hemoprotein domain of P450BM-3, a prototype for microsomal P450's. Science (Wash DC) 261: 731-736.[Abstract/Free Full Text]
Schlichting I, Berendzen J, Chu K, Stock AM, Maves SA, Benson DE, Sweet RM, Ringe D, Petsko GA, and Sligar SG (2000) The catalytic pathway of cytochrome P450cam at atomic resolution. Science (Wash DC) 287: 1615-1622.[Abstract/Free Full Text]
Schuller DJ, Wilks A, Ortiz de Montellano PR, and Poulos TL (1999) Crystal structure of human heme oxygenase-1. Nat Struct Biol 6: 860-867.[CrossRef][Medline]
Schuller DJ, Zhu W, Stojiljkovic I, Wilks A, and Poulos TL (2001) Crystal structure of heme oxygenase from the Gram-negative pathogen Neisseria meningitidis and a comparison with mammalian heme oxygenase-1. Biochemistry 40: 11552-11558.[CrossRef][Medline]
Scott EE, He YA, Wester MR, White MA, Chin CC, Halpert JR, Johnson EF, and Stout CD (2003) An open conformation of mammalian cytochrome P450 2B4 at 1.6-Å resolution. Proc Natl Acad Sci USA 100: 13196-13201.[Abstract/Free Full Text]
Scott EE, White MA, He YA, Johnson EF, Stout CD, and Halpert JR (2004) Structure of mammalian cytochrome P450 2B4 complexed with 4-(4-chlorophenyl)imidazole at 1.9-Å resolution: insight into the range of P450 conformations and the coordination of redox partner binding. J Biol Chem 279: 27294-27301.[Abstract/Free Full Text]
Sevrioukova I, Gracia C, Li H, Bhaskar B, and Poulos TL (2003) Crystal structure of putidaredoxin, the [2Fe-2S] component of the P450cam monooxygenase system from Pseudomonas putida. J Mol Biol 333: 377-392.[CrossRef][Medline]
Sevrioukova I, Li H, and Poulos TL (2004) Crystal structure of putidaredoxin reductase from Pseudomonas putida, the final structural component of the P450cam monooxygenase system. J Mol Biol 236: 889-902.
Sevrioukova IF, Li H, Zhang H, Peterson JA, and Poulos TL (1999) Structure of a cytochrome P450-redox partner electron transfer complex. Proc Natl Acad Sci USA 96: 1863-1868.[Abstract/Free Full Text]
Sørlie M, Gorren ACF, Marchal S, Shimizu T, Lange R, Andersson KK, and Mayer B (2003) Single-turnover of nitric-oxide synthase in the presence of 4-amino-tetrahydrobiopterin: proposed role for tetrahydrobiopterin as a proton donor. J Biol Chem 278: 48602-48610.[Abstract/Free Full Text]
Sugishima M, Sakamoto H, Higashimoto Y, Omata Y, Hayashi S, Noguchi M, and Fukuyama K (2002) Crystal structure of rat heme oxygenase-1 in complex with heme bound to azide: implication for regiospecific hydroxylation of heme at the alpha-meso-carbon. J Biol Chem 277: 45086-45090.[Abstract/Free Full Text]
Tayeh MA and Marletta MA (1989) Macrophage oxidation of L-arginine to nitric oxide, nitrite and nitrate. Tetrahydrobiopterin is required as a cofactor. J Biol Chem 264: 19654-19658.[Abstract/Free Full Text]
Unno M, Matsui T, Chu GC, Couture M, Yoshida T, Rousseau DL, Olson JS, and Ikeda-Saito M (2004) Crystal structure of the dioxygen-bound heme oxygenase from Corynebacterium diphtheriae: implications for heme oxygenase function. J Biol Chem 279: 21055-21061.[Abstract/Free Full Text]
Williams PA, Cosme J, Sridhar V, Johnson EE, and McRee DE (2000) Mammalian microsomal cytochrome P450 monooxygenase: structural adapations for membrane binding and functional diversity. Mol Cell 5: 121-131.[CrossRef][Medline]
Wlodawer A and Erickson JW (2003) Structure-based inhibitors of HIV-1 protease. Annu Rev Biochem 62: 543-585.[CrossRef]
Yano J, Koo L, Schuller D, Li H, Ortiz de Montellano PR, and Poulos TL (2000) Crystal structure of a thermophilic cytochrome P450 from the archaeon Sulfolobus solfataricus. J Biol Chem 275: 31086-31092.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
Y. Tu, R. Deshmukh, M. Sivaneri, and G. D. Szklarz
Application of Molecular Modeling for Prediction of Substrate Specificity in Cytochrome P450 1A2 Mutants
Drug Metab. Dispos.,
November 1, 2008;
36(11):
2371 - 2380.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P. J. Murphy
The Development of Drug Metabolism Research as Expressed in the Publications of ASPET: Part 3, 1984-2008
Drug Metab. Dispos.,
October 1, 2008;
36(10):
1977 - 1982.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H.-P. Hersleth, T. Uchida, A. K. Rohr, T. Teschner, V. Schunemann, T. Kitagawa, A. X. Trautwein, C. H. Gorbitz, and K. K. Andersson
Crystallographic and Spectroscopic Studies of Peroxide-derived Myoglobin Compound II and Occurrence of Protonated FeIV O
J. Biol. Chem.,
August 10, 2007;
282(32):
23372 - 23386.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Kumar, L. Sun, H. Liu, B.K. Muralidhara, and J. R. Halpert
Engineering mammalian cytochrome P450 2B1 by directed evolution for enhanced catalytic tolerance to temperature and dimethyl sulfoxide
Protein Eng. Des. Sel.,
December 1, 2006;
19(12):
547 - 554.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. Zhao, M. A. White, B. K. Muralidhara, L. Sun, J. R. Halpert, and C. D. Stout
Structure of Microsomal Cytochrome P450 2B4 Complexed with the Antifungal Drug Bifonazole: INSIGHT INTO P450 CONFORMATIONAL PLASTICITY AND MEMBRANE INTERACTION
J. Biol. Chem.,
March 3, 2006;
281(9):
5973 - 5981.
[Abstract]
[Full Text]
[PDF]
|
|
|
![[Feedback]](/icons/banner/feedbackACT.gif)
![[For Subscribers]](/icons/banner/subscriberACT.gif)
![[Archive]](/icons/banner/archiveACT.gif)
![[Search]](/icons/banner/searchACT.gif)
![[Contents]](/icons/banner/tocACT.gif)
Top
Abstract
Conformational Dynamics
