![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Center for Drug Discovery and Design, State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Graduate School of the Chinese Academy of Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China (W.L., H.L., F.G., X.L., J.S., H.J.); Department of Pharmacology and Toxicology, University of Texas Medical Branch, Galveston, Texas (J.R.H.); Department of Medicinal Chemistry, University of Kansas, Lawrence, Kansas (E.E.S.); and School of Pharmacy, East China University of Science and Technology, Shanghai, China (H.J.)
(Received February 11, 2005; Accepted April 6, 2005)
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
). This issue is important because experimental data indicate that the regio- and stereospecificities of P450s may be influenced not only by residues in the active site but also by residues far from the binding site (Domanski and Halpert, 2001). A few structures of bacterial P450s suggest several potential routes for substrate access to and exit from the active site. The structures of substrate-free P450 102 (Ravichandran et al., 1993) and inhibitor-bound P450 101 (Raag et al., 1993) indicate a possible access channel located between the F-G loop and ß1 sheet. This channel has also been suggested as a route for substrate/product egress in several other bacterial P450s (Wade et al., 2004). A very different channel has been described in the structure of P450 51 (Podust et al., 2001). This channel threads through the B-C loop and is almost perpendicular to the channel found in P450 102.
To date, the structures of five mammalian P450s, including rabbit 2C5 (Williams et al., 2000) and 2B4 (Scott et al., 2003), and human 2C8 (Schoch et al., 2004), 2C9 (Williams et al., 2003), and 3A4 (Williams et al., 2004; Yano et al., 2004), have been determined. Most of these structures adopt a closed conformation, in which no obvious channels are present for substrate entry or product egress. Thus, an intriguing problem arises: which parts open up to allow substrate/product passage in the closed form? Specific P450 regions have been shown to alter their conformation in response to ligand. The most dramatic differences in protein conformation are observed for 2B4. The substrate-free structure (Scott et al., 2003) reveals a large open cleft that extends from the protein surface directly to the heme iron, whereas an inhibitor-bound structure (Scott et al., 2004b) adopts a closed conformation similar to that observed in the mammalian 2C enzymes. The differences between the open and closed structures of 2B4 are primarily limited to helices F through G, helices B' through C, the N-terminus of helix I, and the ß4 region. The conformational change upon ligand binding implies that these specific flexible regions may be involved in substrate access or egress channels in the closed P450 form, but the residues involved and the mechanisms of channel opening and closing are unknown. An understanding of these questions is important to explain the broad substrate specificity and regio- and stereospecificities of P450s. In the present study, we use steered molecular dynamics (SMD) simulation to probe the possible substrate exit pathway(s) of P450 2B1. P450 2B1 has been chosen because of the wealth of site-directed mutagenesis information available.
|
). Testosterone, a typical substrate of P450 2B1, was docked into the active site of P450 2B1 in a reactive binding orientation, and then pulled out along three putative exit channels using SMD (Fig. 1), chosen based on available P450 structures. Channel 1 is located between helices I and G and the B'-C loop, and was selected based on the large cleft located in this region in P450 2B4 and an open channel found in P450 2C5 (Williams et al., 2000). Channel 2 is directed through helix B' and the B'-C loop and is consistent with an opening of this region in the unliganded 2B4 structure and the channels found in P450 51 (Podust et al., 2001), P450 152A1 (Lee et al., 2003), and P450 3A4 (Williams et al., 2004). Channel 3 corresponds to a "solvent channel" between the active site and the protein surface found in several P450 crystal structures (Haines et al., 2001; Wester et al., 2003). This channel is lined by helices E, F, and I and ß3 hairpin, and has been suggested to be important for controlling proton access to the active site (Haines et al., 2001). However, it is not yet clear whether this solvent channel can permit the passage of large ligands. The channel between the F-G loop and ß1 sheet in P450 101 and P450 102 was not investigated, because this region is occupied by two short helices, F' and G', in the 2B1 model. Moreover, site-directed mutagenesis data from 2B1 do not support this region as a channel for substrate passage (Scott et al., 2002). |
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
) using the InsightII software package [InsightII; Accelrys Molecular Simulations Inc., San Diego, CA (2000)]. The coordinates of the conserved residues were assigned based on the corresponding residues of P450 2B4 using the Homology module of InsightII. The heme coordinates were copied from P450 2B4 into the corresponding site of the P450 2B1 model.
After the coordinate assignment, the initial model of P450 2B1 was refined using the GROMACS 3.1 software package (Lindahl et al., 2001) with an extended version of the GROMOS87 force field (van Gunsteren et al., 1996). Energy minimization was performed with 100 steps of steepest descent followed by 300 steps of conjugate gradient to release conflicting contacts among residues. The protein was then solvated with water in a rectangular periodic box. The simple point charge model (Berendsen et al., 1981) was used to describe water molecules. Since the protein-water system has a total charge of –5e, the system was neutralized by Na+ ions, which replaced water molecules at the positions of lowest Coulomb potential in the system. The solvent was relaxed by energy minimization while restraining the protein atomic positions with a harmonic potential. The system was then energy-minimized without restraints for 2000 steps using a combination of steepest descent and conjugate gradient. Afterwards, molecular dynamics (MD) equilibration at constant temperature was performed to provide further structural relaxation. Equilibration was conducted at 300 K with decreasing harmonic restraints over a 15-ps time interval followed by 1 ns of equilibration without restraints. The model obtained was energy-minimized again for 1000 steps using the conjugate gradient method. The optimized structure was used as the model for subsequent substrate docking and SMD simulations. During the MD simulation, the LINCS algorithm (Hess et al., 1997) was used to constrain all bonds. A dielectric constant of 1 and a time step of 2 fs were used. The electrostatic interactions were calculated using the particle-mesh Ewald method (Essmann et al., 1995) with a 0.9 nm cutoff. The temperature was kept constant by coupling solute and solvent separately to a thermal bath at 300 K with a coupling constant 
T = 0.1 ps. Pressure was kept constant by coupling to a pressure bath at 1.0 bar, using a coupling constant P = 0.5 ps.
The overall quality of the final 3D model of P450 2B1 was assessed with respect to its geometry and energy. The Procheck (Laskowski et al., 1993) and Prostat modules of InsightII [InsightII; Accelrys (2000)] were used for geometric evaluation. Prosa 2003 (Sippl, 1993) was used to evaluate the quality of consistency between the native fold and the sequence and to examine the energy of residue-residue interactions. The energy is transformed to a z-score by
 | (1) |
S is the associated standard deviation.
Docking Simulation. The structure of testosterone was constructed using the Builder module of the Insight II package [InsightII; Accelrys (2000)]. Placement of testosterone into the active site of the 3D model of P450 2B1 was accomplished by using the Affinity program encode in the Insight II package [InsightII; Accelrys (2000)] as previously described (Scott et al., 2004a). Testosterone was automatically docked in a reactive orientation leading to 16
-hydroxylation.
SMD Simulation. Prior to the SMD simulations of the P450 2B1-testosterone complex, a 200-ps MD simulation was performed to evaluate the stability of testosterone in the reactive binding site of the 3D model, yielding an equilibrated starting structure for SMD simulation. The force field parameters for testosterone were generated using the Dundee Prodrug Server (van Aalten et al., 1996). Atomic charges of testosterone were calculated using the ChelpG method, an electrostatic potential-fitting approach (Breneman and Wiberg, 1990), at the HF/6–31G* level. This ab initio calculation was carried out using the Gaussian98 program (Frisch et al., 1998).
SMD is an extended molecular dynamics simulation method mimicking the principle of atomic force microscopy (Grubmuller et al., 1996). SMD has been widely used to explore the binding and unbinding properties of biomolecules and their responses to external mechanical manipulations at the atomic level (Isralewitz et al., 2001). SMD has also been successfully applied to identify ligand pathways and to explore the elastic properties of several proteins (Isralewitz et al., 2001; Shen et al., 2003; Xu et al., 2003). In SMD simulation, time-dependent external forces are applied to a ligand to accelerate binding or unbinding processes. From the accelerated dissociation process of the ligand, the SMD simulation can reveal information about the protein's flexibility and its response to the dissociation of ligand. Analyses of interactions between the ligand and the protein and the relationship between the applied forces and the ligand pathway can yield important information about the structure-function relationships of the protein-ligand complex, the binding and unbinding pathway(s), and possible mechanisms of ligand recognition and inhibition.
In the current SMD simulations, testosterone was pulled out from the binding pocket through the putative exit channels (Fig. 1) by an external force. The movement of testosterone along the channels was determined using the criterion of minimal collision with amino acid residues. The center of mass of testosterone was harmonically restrained to a point moving with a constant velocity along the desired directions. Steered molecular dynamics were performed using the GROMACS 3.1 software package (Lindahl et al., 2001) and an extended version of GROMOS87 force field (van Gunsteren et al., 1996). During the SMD simulations, the pulling velocity was set to 0.005 nm · ps–1 and the spring force constant was assigned as 500 kJ · mol–1 · nm–2. The forces of testosterone with P450 2B1 were monitored throughout the SMD simulations, and the interactions between the protein residues and testosterone were analyzed along the SMD trajectory. At each time point, the presence of direct hydrogen bonds, water bridges, and hydrophobic interactions between testosterone and 2B1 was analyzed using the LIGPLOT program (Wallace et al., 1995).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
) confirmed that the sequence identity (79%) is the highest between 2B1 and 2B4 among all the known P450 structures in the Protein Data Bank. The high sequence identity between 2B1 and 2B4 ensures the accuracy of the modeled structure of 2B1 based on the crystal structure of 2B4. A comparison between the ligand-free and ligand-bound structures of 2B4 reveals that the latter adopts a closed conformation that is similar to 2C5 (Wester et al., 2003; Schoch et al., 2004). The substrate-bound structure of 2B1 is likely similar to the ligand-bound structure of 2B4. Accordingly, the crystal structure of a 4-(4-chlorophenyl)imidazole-2B4 complex (PDB entry 1SUO) (Scott et al., 2004b) was selected as the template to construct the 3D model for 2B1. In the crystal structure of the P450 2B4 complex, the coordinates for the N-terminal residues 1–27 were missing. Since the N-terminus does not affect substrate binding, the 3D model of P450 2B1 was constructed only for residues 28–491. Figure 2 displays the sequence alignment between P450 2B1 and P450 2B4. Because of the lack of gaps in the alignment, the entire structures were considered as conserved regions.
|
|
-
angles in core Ramchandran regions was 84.9% in the P450 2B1 model. The geometric assessment of P450 2B1 was also performed using Prostat/InsightII [InsightII; Accelrys (2000)]. The cutoff used, which represents the significant difference for bond length, bond angle, and torsion from the reference value obtained from known protein crystal structures, is 5 S.D. For the P450 2B1 model, none of the bond distances, none of the bond angles, and only three dihedral angles were found to have more than 5 S.D. Thus, the P450 2B1 model is of reasonable quality compared with the crystal structure of the P450 2B4 complex.
|
-hydroxylation orientation leading to the formation of the major product. Contacts between protein and substrate are predominately hydrophobic. The side chains of Arg98, Gly99, Thr100, Ile101, Ile104, Ile114, Phe115, Phe206, Phe297, Ala298, Glu301, Thr302, Val363, Ile365, Val367, Pro368, and Ile477 lie within 5 Å of testosterone. Most of these residues, including Ile114, Phe115, Phe206, Phe297, Ala298, Thr302, Val363, Ile365, Val367, and Ile477, correspond to key residues responsible for regio- and stereoselectivities revealed by previous mutagenesis studies on family 2 enzymes (Domanski and Halpert, 2001). Additional residues, including Thr100, Ile101, and Ile104, are located in the first of six substrate recognition sites (SRS-1), as suggested by Gotoh (1992), based on a comparison between family 2 enzymes and P450 101, and have been shown to influence the kinetic constants of some substrates in very recent site-directed mutagenesis experiments (Honma et al., 2005). To evaluate the stability of testosterone in the active site of the P450 2B1 model, a 200-ps MD simulation was performed. Figure 5 shows the RMSD of testosterone and the distance between the iron atom and oxidation site C16 of testosterone as a function of simulation time. The RMSD fluctuates around 0.1 nm during the simulation, which indicates that the testosterone structure does not deviate significantly from the initial docked pose. The iron-C16 distance in the complex fluctuates around 0.45 nm, the typical length found in the crystal structures of other P450s for carbon oxidation. Thus, during the free MD simulation, the testosterone remained in the active site and maintained an orientation allowing C16 hydroxylation.
|
Egress of Testosterone from P450 2B1. The force profile of testosterone egress along channel 1 is shown in Fig. 6A. During the first 230 ps, a steady increase of the applied force was observed. At the beginning of the simulation, the substrate-protein complex is stabilized by a water bridge to Glu301 and hydrophobic contacts with Ile104, Ile114, Phe297, Thr302, Val363, Ile365, Pro368, and Ile477, as shown in Fig. 7A. At 230 ps, testosterone forms a new hydrogen bond to Gly99 and hydrophobic interactions with Arg98, Ile101, Ile104, Ile114, Ala298, Val367, Pro368, and Ile477. These residues are located in the active site, demonstrating that the substrate has not left the vicinity of the heme. Breaking these hydrogen bond and hydrophobic interactions produces the highest peak in the force profile, corresponding to a rupture force of 860 pN. At 260 ps, testosterone has moved out of the active site and reached the entrance of channel 1, and a local minimum appears in the force profile. At this time, the direct hydrogen bond between substrate and Gly99 is broken, but testosterone forms a water bridge with Ser294 and several hydrophobic contacts with Ile104, Phe108, and Tyr111 in helix B' and the B'-C loop, residue Leu293 in helix I, and Leu238 and Leu242 in helix G, as shown in Fig. 7B.
|
|
Figure 6B shows the force profile of testosterone exiting the P450 2B1 active site via channel 2. Initially, the exit of testosterone proceeded slowly, since the substrate remained tightly bound while the applied force increased steadily. At 240 ps of the simulation time, testosterone forms water bridges with Arg98, Gly99, Phe115, Asn117, Ser294, Glu301, and Arg370 and hydrophobic interactions with Ile114, Ile362, Val367, Pro368, and Ile477. Breaking these interactions produces the highest force peak, corresponding to a rupture force of 830 pN. At 275 ps, the substrate forms two water bridges with Asn117 and Arg370 and six hydrophobic contacts with residues Gly99, Thr100, Ile101, Ala102, Glu105, and Phe115, as shown in Fig. 7E. All of these residues except Arg370 are located in helix B' or the B'-C loop. These interactions demonstrate that the substrate has moved out of the active site pocket and reaches the entrance of channel 2. At 310 ps, the force decreases down to a local minimum, there is no water bridge between testosterone and 2B1, and the substrate only forms a hydrophobic interaction with Gly99. At 395 ps, a moiety of the substrate has penetrated the B'-C loop and testosterone adjusts its orientation to interact with other residues, which results in a small increase in the rupture force. A force of 400 pN is necessary to overcome the hydrophobic interactions of the substrate with residues Ile104, Glu105, Phe108, and Phe115 (Fig. 7F). After 500 ps, the force has reached a global valley and the substrate has been completely pulled out of the protein.
Figure 6C shows the force profile of testosterone leaving the P450 2B1 active site via channel 3. The rupture force for the substrate leaving this channel is up to 1400 pN, which is more than 1.5 times higher than those of channels 1 and 2. The highest force peak is produced at 400 ps. At this point, testosterone forms two hydrogen bonds to Lys479 and Ile480 and hydrophobic interactions with Phe206, Leu209, Phe297, Glu301, Pro364, Ile477, and Gly478. Several residues, including Leu209, Gly478, Lys479, and Ile480, are located at the entrance of channel 3, indicating that at this time testosterone has moved out of the active site and is penetrating through the channel. During the exit of testosterone via this channel, a hydrogen bond between O
2 in Glu301 and N in Lys479 is noted, as shown in Fig. 8. During the first 250 ps, the hydrogen bond between residue Glu301 and residues Lys479 is relatively stable. From 250 ps to 400 ps, the hydrogen bond is formed intermittently. After 400 ps, the hydrogen bonding is entirely broken, which enlarges the space enough to allow testosterone to pass through. Therefore, the hydrogen bonding between Glu301 and Lys479 may play an important role in preventing substrates from passing through this channel.
|
To compare the protein motions for testosterone exit via these three putative channels, the nonhydrogen atom RMSDs and C RMS fluctuations (RMSFs) from the initial structure were monitored during SMD simulations. The global RMSD after testosterone expulsion from channels 1, 2, and 3 fluctuated around 0.21 nm, 0.19 nm, and 0.18 nm, respectively, as shown in Fig. 9. Figure 10 shows the C RMSF from the initial structure during SMD simulations along these three putative channels. The maximal C RMSF for testosterone exit via channel 3 is 0.27 nm, which is contributed by the flexible ß3 hairpin region (the third shaded region in Fig. 10C). The maximal C RMSFs for testosterone exit via channel 1 and channel 2 are 0.22 nm and 0.21 nm, respectively, which are contributed by the H-I loop region. Since this loop is completely exposed to the surface of protein and thus freely mobile, it is not surprising that the region has a large RMSF for all the putative channels. In addition, the B' helix/B'-C loop region displays a relatively large RMSF in testosterone exit via channel 1 and channel 2 (the first shaded regions in Fig. 10, A and B), which indicates that testosterone exit via both channel 1 and channel 2 involves the flexibility of this region.
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
RMSF via channel 3 is higher than those via channel 1 and channel 2; 3) the hydrogen bond formed between Glu301 and Lys479, which stabilizes the protein structure, has to be broken in order for testosterone to exit through channel 3. Since our simulations yield a similar force profile and rupture force for exit along channel 1 and channel 2, our results do not allow discrimination between these two channels for substrate egress. However, opening mechanisms of these two channels differ, as discussed below.
Although there is no obvious opening between helices I and G and loop B'-C in the closed, inhibitor-bound 2B4 structure, the ligand-free 2B4 structure (Scott et al., 2003) reveals a large cleft in this region that extends directly from protein surface to the heme iron. Site-directed mutagenesis revealed that substitution of residues in the N-terminus of helix I can alter the activity and stereoselectivity of P450 2B1 (Scott et al., 2004a). Moreover, random repulsion molecular dynamics simulations have indicated a similar substrate egress pathway in P450CAM (Ludemann et al., 2000). In the current SMD simulation, the secondary structures of helices I and G are well maintained, and the rupture force is relatively small during testosterone exit along channel 1. All these findings indicate that a channel may exist between helices I and G and the B'-C loop. The simulation results suggest some residues that may play important roles in the egress of testosterone along channel 1. After entering channel 1, testosterone forms a water bridge to Ser294 which, in our simulation, persisted until 450 ps. During this period, the hydrogen-bonding network formed between Ser294 and testosterone discourages the egress of testosterone. It can be speculated that abolishing the hydrogen-bonding network will alter the enzyme activity. This is consistent with the recent mutagenesis data indicating that S294A mutant exhibited a 3-fold decrease in Kcat/Km for the 16 product (Scott et al., 2004a). A similar function is attributed to Glu286, which also formed a water bridge to testosterone. The replacement of Glu286 by Ala led to a 2-fold decrease in Kcat/Km for the 16 product (Scott et al., 2004a).
An analysis of the trajectory of SMD along channel 1 reveals that two residues located at the entrance of the putative channel 1, Phe108 and Phe297, appear to act as two flexible clamps during substrate egress. The hydrophobic interactions of Phe108 and Phe297 with testosterone not only stabilize substrate binding and guide the orientation of the substrates, but also prevent the substrate from leaving the active site. The B'-C loop must rearrange to open wide enough for testosterone passage, and at the same time, the side chain torsion of Phe297 rotates to enlarge the space, as shown in Fig. 11. To quantify this opening and rotation, the distance between Phe297 and Phe108 and the side chain torsions 
of Phe297 have been monitored, as shown in Fig. 12. In the first 230 ps, the distance remained around 0.9 nm. In the following 50 ps, displacement of the B'-C loop occurred, and the distance increased significantly to 1.4 nm in order for testosterone to pass through. At the same time, the torsion of Phe297 increased from 22° to a maximal 120°. After testosterone crossed over the bottleneck, the torsion returned to its initial value.
|
|
). Mutagenesis data on Phe108 show that substitution of this residue can also alter the kinetic constants of 2B1 (Honma et al., 2005). The SMD simulation result together with the site-directed mutagenesis results support the inference that channel 1 is a possible route for substrate egress in 2B1.
The structure of P450 51 (Podust et al., 2001) reveals an access channel from protein surface to the heme, which penetrates through the open B-C loop. In P450 152A1 (Lee et al., 2003), a similar channel is proposed to permit water to escape from the active site and to allow access of hydrogen peroxide to substrate-bound enzyme. An opening between helices B' and C is also apparent in the ligand-free P450 2B4 structure (Scott et al., 2003) and 3A4 structures (Williams et al., 2004; Yano et al., 2004). Based on these observations, this region was selected as a possible exit channel in the current study. The SMD simulation results show that a relatively small rupture force and only slight backbone motions are required for testosterone exit along channel 2, as shown in Fig. 6B and Fig. 11B. Our results, in conjunction with findings from crystal structures of several mammalian P450s, suggest that channel 2 may serve as a common channel for ligand passage in some bacterial and mammalian P450s.
A detailed analysis of the interactions between testosterone and the protein revealed that Phe115 acts as a gatekeeper, not only playing an important role in stabilizing the orientation of testosterone in the active site, but also preventing ligands from escaping from the active site. The corresponding Phe114 in 2C9 was found to have a strong hydrophobic contact with the phenyl group of S-warfarin and, thereby, to stabilize its orientation in the 2C9 enzyme (Williams et al., 2003). An analysis of snapshots of the exit process through channel 2 shows that after testosterone leaves the active site, the phenyl ring of Phe115 flips to occupy the space left by testosterone and consequently closes the channel. This implies that a mutation of Phe115 to other residues might change the orientation of substrate in the active site and influence substrate egress from the active site. This suggestion is in agreement with the site-directed mutagenesis data, which indicate that F115A exhibits a 3-fold decrease in the testosterone hydroxylase activity (Domanski et al., 2001).
Although the opening of both channels is due to the flexibility of helix region B'-C loop/B', the opening mechanism of channel 2 is different from that of channel 1. The opening of channel 1 is characterized by a rotation of Phe297 in conjunction with the B'-C loop rearrangement, involving a relatively large displacement of the B'-C loop (the first shadow region in Fig. 10A and Fig. 11A). In contrast, testosterone exit through channel 2 is achieved by the expansion of B'-C loop (Fig. 11B), together with a relatively small displacement of the backbone (Fig. 10B). It has been noted that in family 2 P450s, two highly conserved GlyXGly motifs flank the region between helix B' and the B'-C loop. Glycines are known for their torsional flexibility due to the lack of a side chain, and the peptide backbone can readily adopt a wide range of conformations. The GlyXGly motif thus can be assumed to constitute the hinge for the B'-C loop conformational changes.
The solvent channel, our channel 3, has been suggested to play an important role in controlling proton access to active site (Haines et al., 2000; Wester et al., 2003). However, no evidence at present directly indicates a ligand pathway. Our simulation shows that substrate egress via this route requires large rupture forces, large backbone motion, and the rupture of a hydrogen bonding network involving Glu301. This indicates a low probability of channel 3 to serve as an egress channel for testosterone in P450 2B1. In light of apparent opening of this channel in the structures of 2C5, 102, and inhibitor-bound 2B4, and the smaller ligand size in the 2B4 complex, access/egress through channel 3 cannot be ruled out. This remains to be tested by additional experimental and computational studies.
In conclusion, we used SMD simulations to investigate three possible ligand channels of P450 2B1 and to clarify the role of some residues previously identified by site-directed mutagenesis experiments. Our results demonstrate that of the three possible channels, the "solvent channel" is unlikely to serve as exit channel, whereas both other channels are equally good candidates for the egress channels. However, opening mechanisms of these two channels differ. The opening of channel 1 is characterized by a rotation of Phe297 in conjunction with a relatively large displacement of the B'-C loop. In contrast, testosterone exit through channel 2 is achieved by the expansion of B'-C loop together with a relatively small displacement of the backbone.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
ABBREVIATIONS: P450, cytochrome P450; SMD, steered molecular dynamics; 3D, three-dimensional; PDB, Protein Data Bank; MD, molecular dynamics; RMS, root-mean-square; RMSD, RMS deviation; RMSF, RMS fluctuation.
Address correspondence to: Dr. James R. Halpert, Department of Pharmacology and Toxicology, The University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555-1031. E-mail: jhalpert{at}utmb.edu or Dr. Hualiang Jiang, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555 Zu Chong Zhi Road, Zhangjiang Hi-Tech Park, Shanghai 201203. E-mail: hljiang{at}mail.shcnc.ac.cn
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Zhang, C. Kenaan, D. Hamdane, G. H. B. Hoa, and P. F. Hollenberg Effect of Conformational Dynamics on Substrate Recognition and Specificity as Probed by the Introduction of a de Novo Disulfide Bond into Cytochrome P450 2B1 J. Biol. Chem., September 18, 2009; 284(38): 25678 - 25686. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Li, H. Liu, X. Luo, W. Zhu, Y. Tang, J. R. Halpert, and H. Jiang Possible Pathway(s) of Metyrapone Egress from the Active Site of Cytochrome P450 3A4: A Molecular Dynamics Simulation Drug Metab. Dispos., April 1, 2007; 35(4): 689 - 696. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. W. Locuson and T. S. Tracy Identification of Binding Sites of Non-I-Helix Water Molecules in Mammalian Cytochromes P450 Drug Metab. Dispos., December 1, 2006; 34(12): 1954 - 1957. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
