Key interactions of the mutant HIV-1 reverse transcriptase/ efavirenz : an evidence obtained from ONIOM method

Pornthip Boonsri; Mayuso Kuno; Supa Hannongbua

doi:10.1039/C1MD00162K

View PDF VersionPrevious ArticleNext Article

DOI: 10.1039/C1MD00162K (Concise Article) Med. Chem. Commun., 2011, 2, 1181-1187

Show CompoundsShow Chemical Terms

Key interactions of the mutant HIV-1 reverse transcriptase/efavirenz: an evidence obtained from ONIOM method†

Pornthip Boonsri ^ab, Mayuso Kuno ^c and Supa Hannongbua *^ab
^aDepartment of Chemistry, Faculty of Science, Kasetsart University, Bangkok, 10900, Thailand. E-mail: fscisph@ku.ac.th; Fax: +66-2-5793955; Tel: +66-2-5625555 ext 2111
^bCenter of Nanotechnology KU, Kasetsart University and NANOTEC Center of Excellence at Kasetsart University, Bangkok, 10900, Thailand
^cDepartment of Chemistry, Faculty of Science, Srinakharinwirot University, Bangkok, 10110, Thailand

Received 23rd June 2011 , Accepted 10th September 2011

First published on 27th September 2011

Abstract

Two-layered ONIOM calculations were performed in order to compare the binding of efavirenz (EFV) to the HIV-1 RT binding pocket of both wild type (WT) and K103N enzymes. The K103N mutation reduces the binding affinity of the inhibitor by 5.81 kcal mol⁻¹ as obtained from the ONIOM2 (B3LYP/6-31G(d,p):PM3) method. These indicate that the loss of binding energy to K103N mutation can attribute to a weakened attractive interaction between the drug and residues surrounding in the binding pocket. The deformation of the K103N binding pocket requires more energy for structural rearrangement than that of the WT by approximately 4.0 kcal mol^¬1. Moreover, the pairwise energies perfectly demonstrate that the K103N mutation affects on the loss of the interaction energy. In addition, the main influences are due to residues surrounding in the binding pocket; K101, K102, S105, V179, W229, P236 and E138. In particular, two residues; K101 and S105, established hydrogen bondings with the inhibitor. ONIOM calculations, resulting in the details of binding energy, interaction energy and deformation energy can be used to identify the key interaction and structural requirements of more potent HIV-1 RT inhibitor.

Introduction

The human immunodeficiency virus type 1 (HIV-1) has been identified as the causative agent of acquired immunodeficiency syndrome (AIDS).¹ At present, there are many efforts being made to stop reverse transcriptase (RT) replication. Therefore, this enzyme has been one of the important and excellent targets for drug discovery. The enzyme inhibitors are currently among the most widely used in treating HIV infection. In spite of the search of new potent and specific HIV inhibitors, understanding the details of molecular mechanisms has been still unclear and important for further drug development.

The most common HIV-1 RT mutations associated with resistance to RT inhibitors are found to be substitution of the Lys103Asn (K103N) and Tyr181Cys (Y181C) within the non-nucleoside inhibitor binding pocket (NNIBP).^2–4 Ribone et al. reported that the presence of the K103N mutation causes to a more voluminous NNIBP and modifies its electrostatic properties.⁵ The K103N was the primarily associated with the treatment failure of efavirenz.^6–8 Therefore, it is important to understand the effect of this mutation to drug-enzyme interaction. However, increasing the size of binding pocket resulted in more tight fitting of the large inhibitor in the enzyme.⁹ Thus, the development of new NNRTIs to improve resistance profiles continues to be an active area of research. According to the experimental data, the binding energy of EFV and navirapine to the K103N HIV-1 RT decreased by 6-fold and 40-fold compared to the wild type (WT), respectively.¹⁰ In the present work, we focus on EFV, which have been found to show greater resilience and smaller reductions in activity to mutations.^11–13

HIV-1 RT inhibitors can be divided into two classes: nucleoside inhibitors (NRTIs) and latter the newer, non-nucleoside inhibitors (NNRTIs). The latter inhibitors bind to an allosteric binding pocket and inhibit RT by the structural distortion of the catalytic residues in the hydrophobic pocket leading to a large conformational shift into an inactive form of the protein.¹⁴ There are currently five approved NNRTIs drugs that target NNIBP including nevirapine,¹⁵delavirdine,¹⁶ efavirenz,¹⁰etravirine,¹⁷ and recently rilpivirine.¹⁸Delavirdine and nevirapine are considered as the first generation NNRTIs and show dramatic loss of activity with single point mutations in the NNIBP. EFV, a second generation NNRTIs, is one of the most potent and selective inhibitors with high inhibitory activity to HIV-1 RT. It is effective against many point mutations; however, efficacy is strongly compromised by the K103N mutation. TMC125 (etravirine) and TMC278 (rilpivirine) are diarylpyrimidine (DAPY) NNRTIs which have high inhibitory affinity for WT and mutant virus strains including K103N.¹⁸ However, the binding mode and protein conformation of RT/DAPY complexes have been investigated and shown different results, as compared to the first and second generation drugs, due to the flexibility of these drugs when bound to RT enzyme. As a result, computational calculations has opened up the possibilities to analyse in more details of protein–ligand interactions, therefore, this method has been widely applied on HIV-1 RT.¹⁹ Mei et al. reported a quantum chemical calculations (QC) study on HIV-1 RT/EFV binding provided useful information in designing more potent inhibitors.²⁰ The free energy of binding that affect the K103N mutation of this enzyme was investigated, based on Monte Carlo/free energy perturbation (MC/FEP) and molecular dynamics (MD) with Molecular Mechanics Poisson Boltzmann/surface area (MM-PBSA) calculations.^21–23 To this end, we are presenting insights into the drug-enzyme interactions using quantum mechanical (QM) method which is the more general combined methods and more easier performance than using on MM sampling. We employ the ONIOM (Our own N-layered Integrated molecular Orbital and molecular Mechanics) method, developed by Morokuma et al., applied on the protein–ligand interactions of the HIV-1 RT complex systems.^10,24–28 The obtained results from ONIOM methods agreed well with the experimental investigation.^29,30 It has been proven to be an appropriate method for the theoretical treatment on large molecular systems.

This method is based on the approach of partitioning a large molecular system into onion-skin-like layers, subsequently followed by the use of a variety of quantum chemical and molecular mechanics methods. In the ONIOM approach, a small part of the system such as the inhibitor and a reacting amino acid in the binding pocket, is treated at a high level of theory, whereas the larger surrounding region is modeled using a lower level of theory.^30–32 This technique balances the computational expense and the accuracy of calculations. The basis of ONIOM method is shown in eqn (1), the total energy of the entire system can be obtained from two-layer ONIOM (ONIOM2). Details of the ONIOM2 approach are given by Kuno et al.²⁸ Using ONIOM, the total energy system E^ONIOM can be expressed as:


E^ONIOM = E_{(high, model)} + E_(low, real) − E_{(low, model)},	(1)

where E_{(high, model)} is the energy of the inner layer at high level of theory, E_(low, real) is the energy of the entire system at low level of theory, and E_{(low, model)} is the energy of the inner layer at the low level of theory.

Quantum chemical calculations and ONIOM method can bring to insight into the interaction mechanisms of these inhibitors to the mutant HIV-1 RT and can also accurately predictive binding energies of the complexed system. Therefore, we firstly compared an overall structural conformation of WT and K103N mutation that taken from X-ray crystal structures. Then, the pairwise analysis between EFV and individual residue in the binding pocket of both complexed systems were investigated by using quantum chemical calculations. Binding energy (BE) analysis of these two complexes allowed the fundamental features of the drug-enzyme interactions to be assessed based on ONIOM method.

Method of calculations

System setup

Models of HIV-1 RT were created from the Protein Data Bank (PDB) structures with the PDB entry codes 1FK9 ²⁵ and 1IKV ³³ for WT and K103N mutant type, respectively. Residues with at least one atom within a 7 Å diameter were centered on the EFV structure. The model system consists of the EFV bound into NNIBP of 22 residues: Pro95, Leu100-Lys101-Lys102-Lys/Asn103-Lys104-Ser105-Val106, Val179-Ile180-Tyr181, Tyr188-Val189-Gly190, Phe227-Leu228-Trp229, Leu234-His235-Pro236, and Tyr318 of the p66 palm domain, and Glu138 of the p51 palm domain of HIV-1 RT. Hydrogen atoms were added to complete the structure of the model system. The geometry and the position of EFV and also the position of all hydrogen atoms were optimized to give the two starting geometries for all subsequent calculations, defined this optimization as heavy atom fixing (HAF). The set up system is considered as the “real model” (RM) in the ONIOM calculation. Fig. 1 shows the schematic representation of the WT and K103N HIV-1 RT binding pockets.


	Fig. 1 Schematic representation of the adopted real system (RM) of EFV bound to the HIV-1 RT binding pocket. Layer of calculations for ONIOM2 method is shown which A is the inner layer and B is the outer layer of 1) WT HIV-1 RT and 2) K103N HIV-1 RT.

Quantum chemical calculations

To investigate the interaction energy (INT) between efavirenz (EFV) and the individual residues, defined as X_i, QC calculations at B3LYP/6-31G(d,p) level were performed. Using the starting geometries derived from the X-ray crystal structures, HAF optimization by PM3 semi-empirical method was used. The interaction energy of each efavirenz-residue pair, INT_[EFV +_Xi_], is defined as:


INT_[EFV + Xi] = E_[EFV + Xi] − E_[EFV] − E_[Xi],	(2)

where E_[EFV +_Xi_] is the uncorrected energy of EFV and each residue X_i pair, and E_[EFV] and E_[Xi] are the energies of EFV and individual residue, respectively.

ONIOM calculations

The ONIOM method is a hybrid computational method that allows different levels of theory to be applied to different parts of the molecular system. Kuno et al. investigated the applicability of the ONIOM2 and ONIOM3 methods for studying nevirapine and HIV-1 RT binding. The results indicated that despite using different ONIOM methods, the intermolecular distances and the binding energies were quite similar.²⁸ The previous published work applied the ONIOM method on the 8-ClTIBO and HIV-1 RT binding pocket complexes.³⁴ Two optimization procedures, HAF which saves more computational time than backbone atom fixing (BBF), were used and the results clearly indicate that there are insignificantly different of structural and binding affinities as using a different ONIOM approach.³⁴ Therefore, in order to reduce computational expenses, the two-layered ONIOM (ONIOM2) method with HAF optimization procedure was selected to investigate the interaction between EFV and its residues in both WT and K103N HIV-1 RT binding pockets. The important interactions between K103 or N103 and EFV are needed to be included into the model layer which high-level of calculations were applied. In the ONIOM2 method, the system was divided into two layers. The inner layer included the interaction region (Fig. 1, A region) was treated at a higher level of calculation, using the HF/6-31G(d,p) and B3LYP/6-31G(d,p) methods. The outer layer (Fig. 1, B region) was calculated by the PM3 semi-empirical method. All the calculations were performed by Gaussian03.³⁵

Binding energy

The binding energy (BE) of EFV bound to the binding pocket was calculated using ONIOM2 method and is defined as:


BE^ONIOM = E^ONIOM_complex − E^ONIOM_pocket − E^QM_EFV,	(3)

where E^ONIOM_complex is the energy of the whole model structure, calculated by using ONIOM2 method, and E^ONIOM_pocket and E^QM_EFV are the energies of the binding pocket and efavirenz, respectively. The BE can be decomposed into the combination of interaction energy (INT) and deformation energy (DEF) as shown in eqn (4),


BE = INT + DEF,	(4)

To analyze the energy contribution in more details, the BE components of EFV bound to its binding pocket was calculated from eqn(5) and eqn (6):


BE = E[Cpx]^ONIOM2_Opt − E[BP]^ONIOM2_Opt − E[EFV]^QM_Opt,	(5)


BE = INT^Cpx_Opt + DEF_BP + DEF_EFV,	(6)

and


INT^Cpx_Opt = E[Cpx]^ONIOM2_Opt − E[BP]^Cpx_Sp − E[EFV]^Cpx_Sp,	(7)


DEF_BP = E[BP]^Cpx_Sp − E[BP]^ONIOM2_Opt,	(8)


DEF_EFV = E[EFV]^Cpx_Sp − E[EFV]^QM_Opt,	(9)

where Cpx denotes the complex, BP denotes the binding pocket, Opt denotes optimized structures, Sp is single point calculations

Results and discussion

Structural analysis of WT and K103N HIV-1 RT

In order to compare the different binding energies of both WT and K103N HIV-1 RT/EFV systems, the starting geometries from X-ray crystallographic structures were analysed. The added hydrogen atoms and inhibitor were allowed to be optimized by PM3 semiempirical calculations. The root-mean-square distances (RMSD) of all atoms in the bound enzyme-drug complex structures are found to be 0.15 and 0.18 Å for the WT and K103N mutant types, respectively. The obtained result indicated that both bound systems can be used as good starting geometries to represent the whole complex structures of WT and mutant HIV-1 RT/EFV systems.

Superimposition the X-ray crystal structure of the EFV-K103N RT on the WT complex was investigated and shows a RMSD of the backbone atoms of about 0.5 Å. The obtained result shows that structural rearrangement is not significantly different. However, the subtle changes in the overall positioning and orientation of residues lining in the NNIBP are observed, such as L100, K101, K102, S105 V179, L228, H235, and P236. These residues are mostly located around the mutation residue and the wing shape of EFV inhibitor (see ESI†). The interatomic distances between EFV and those residues in the binding pocket are observed.

We additionally determined the hydrogen bonding interactions between the nitrogen atom of benzoxazin-2-one and the backbone carbonyl oxygen of K101 (–C [double bond, length as m-dash] O_K⋯N_B_⁻), and a second between a carbonyl group of the benzoxazin-2-one and the backbone nitrogen atom of K101 (–N_K⋯CO_B⁻) which are shown in Fig. 2.^11,27 The first type of hydrogen bond distance was calculated to be 2.75 and 3.17 Å from the WT and the K103N, respectively, and the second, to be 3.17 and 3.26 Å, respectively.


	Fig. 2 Hydrogen bonds representation of (A) –CO_B⋯H–N_K_⁻, between benzoxazine-2-one and the amine of K101; and (B) –N–H_B⋯OC_K⁻, between benzoxazine-2-one–NH and the carbonyl of K101.

We also found that the interactions between EFV and K101 residue are slightly weakened for the K103N than that of the WT. Indeed, the mutation at 103 position results in destabilizing the formation of a hydrogen bond between the K101 residue and EFV. It is found to be the main contribution of the EFV/RT interactions which will be undoubtedly discussed in the following part.

The WT RT only allows small contaction between the K103 side chain and the EFV inhibitor. The interacting distances from the side chain of K103, C_β and C_γ to the benzoxazin-2-one ring of EFV are all optimal van der Waals distances. For the K103N RT, the methylene group of K103 is replaced by a bulky amide group of N103 that disturbs the interaction as shown in Fig. 3. Thus the substitution of the linear side chain of lysine by the branched one of asparagines at position 103 appears to greatly affect the chemical environment in the K103N mutation, leading to the different interactions.^33,34


	Fig. 3 Electron density map of EFV and residue K103 in the WT NNIBP (A) and residue N103 in the mutant NNIBP (B). The dashed lines show the van der Waals interaction between backbone N of EFV and the C_γ atom of the residue.

Fundamental knowledge of these nonbonded interactions is essentials in the design of new potent inhibitors towards specific target of HIV-1 RT. However, the reason to explain why the mutation encodes resistance to a wide range of NNRTIs and it is rather difficult to test experimentally. Therefore, quantum chemical calculations are needed to investigate the interaction energy between EFV and each residue in the binding pocket.

Interaction energy of both WT and K103N RT

The calculated interaction energies between EFV and each residue surrounding the binding pocket are shown in Table 1 for the WT and K103N RT. The obtained total interaction energy for the WT/EFV is −22.75 kcal mol⁻¹ and reduces to −15.00 kcal mol⁻¹ for K103N. It is obviously seen that there are more repulsive interactions between EFV and the surrounding residues of the K103N than that of the WT.

Table 1 Interaction energies (kcal mol⁻¹) of EFV with individual residues (X_i), calculated by the B3LYP/6-31G(d,p) method

Residues	Interaction energy (kcal mol⁻¹)
Residues	WT^a	K103N mutant	ΔE^b
a The interaction energy taken from Ref. 11. b ΔE = INT_wild type − INT_K103N RT.
Pro(P)095	−0.35	−0.33	−0.02
Leu(L)100	−3.98	−2.66	−1.32
Lys(K)101	−11.29	−9.47	−1.82
Lys(K)102	1.46	−0.96	2.42
Lys(K)103 Asn(N)	−1.53	−0.70	−0.83
Lys(K)104	−0.09	−0.16	0.07
Ser(S)105	−2.33	0.35	−2.68
Val(V)106	1.01	0.70	0.31
Val(V)179	1.34	−1.89	3.23
Ile(I)180	−0.39	−0.34	−0.05
Tyr(Y)181	−0.38	1.06	−1.44
Tyr(Y)188	−1.09	−0.29	−0.80
Val(V)189	−1.90	−0.49	−1.41
Gly(G)190	−0.49	−0.45	−0.04
Phe(F)227	−0.06	0.27	−0.33
Leu(L)228	0.13	0.01	0.12
Typ(W)229	−0.23	1.4	−1.63
Leu(L)234	0.19	1.58	−1.39
His(H)235	−2.24	−1.47	−0.77
Pro(P)236	−1.72	−0.14	−1.58
Tyr(Y)318	−0.27	−0.81	0.54
Glu(E)138(B)	1.46	−0.21	1.67
Total	−22.75	−15.00	−7.75

Moreover, some residues show different interaction energy approximately 1.5 kcal mol⁻¹ as compared with the binding of EFV between the WT and K103N RT. These residues are K101, K102, S105, V179, W229, P236 and E138. The results demonstrate that the loss of binding interaction of the mutation not only comes from the residue K103N, but also some residues located around the binding pocket which corresponding to the interatomic distances (see ESI†). Interestingly, the interaction energy of the inhibitor with residue S105 is much different between the WT and mutation binding to be 2.68 kcal mol^¬1. On the other hand, the occurrence of K103N mutation causes the favorable binding interaction of EFV to residue V179, shown tightly binding of about 3.23 kcal mol⁻¹ as compared to that of the WT structure. Thus, the loss of these interactions can stabilize the structural conformations and help to maintain the bound structure for drug resistance, as anticipated.³³

Furthermore, the interaction between EFV and K101 residue shows the strongest attractive energy to be −11.29 and −9.47 kcal mol⁻¹ for WT and K103N enzymes, respectively. The result corresponds to the interatomic distances that obtained from two hydrogen bondings between EFV and K101 (see Fig. 2). The results show that the loss of attractive interaction of EFV with K101 of the mutant enzyme is about 2.0 kcal mol⁻¹ as compared to the WT binding. The reason to explain dues to the fact that there is an effect of the amide group at N103 which caused the strength of hydrogen bonding between benzoxazin-2-one and K101 residue. Thus, strong binding of EFV to K101 and S105 residues is a key for drug binding to the mutant enzyme.

In general, the binding of NNRTIs in the binding pocket is driven by the stabilizing nonbonded interactions with the amino acid residues, which was termed as hydrophobic and electrostatic interactions to override RT-resistance.^5,36 The cyclopropyl-propynyl group is positioned in the top sub-pocket surrounded by the aromatic side-chains of Y181, Y188, W229, and F227. The benzoxazine-2-one ring is sandwiched between the side-chains of L100 and V106, while also making edge-on contacts with Y318 and V179. The nonbonded interactions play an important role to the binding energy of EFV upon the HIV-1 RT of both types. The more details of the interaction analysis are further investigated.

Binding energy calculations for WT and K103N RT

The ONIOM method was applied to investigate the binding energies of the WT and K103N RT. In this study, the inner layer consists of [EFV + K103] and [EFV + N103] for WT and mutant type systems, respectively (Fig. 1). Based on the results in Table 1, the interaction between EFV and K101 residue shows the highest interaction energy when compared to other residues due to hydrogen bondings found in the binding (see Fig. 2). We take into account the effect of K101 residue into the model system of EFV/HIV-1 RT binding pocket, consequently, [EFV + K101 + N103] was used to represent another model layer for the calculations. Therefore, the generated ONIOM2 models are:

ONIOM2A: HF/6-31G(d,p)[EFV + K103N]:PM3[RM]

ONIOM2B: B3LYP/6-31G(d,p)[EFV + K103N]:PM3[RM]

ONIOM2C:HF/6-31G(d,p)[EFV + K101 + K103N]:PM3[RM]

[RM] stands for real model in ONIOM calculations. The results of BE, INT and DEF energy calculations from ONIOM2A and ONIOM2B are listed in Table 2. It was found that the binding energy of EFV to the WT is about twice as compared with the mutant binding pocket based on ONIOM2 methods. The BE difference between WT and K103N RT are 5.01 for ONIOM2A and 5.81 kcal mol⁻¹ for ONIOM2B calculations. Based on different method of calculations, the BE analyzes obtained from ONIOM2B calculations give stronger binding of about 2.19 and 3.31 kcal mol⁻¹ as compared with ONIOM2A for the WT and mutant type, respectively. Considering more into details, it was found that the different INT of both WT and mutant systems are similar of about 3.00 kcal mol^¬1. The results indicate that the K103N significantly influences the enzyme-inhibitor interactions.

Table 2 Binding energy (BE), interaction energy (INT) and deformation energy (DEF) of EFV inhibitor complex with the WT and mutant type bound into the RT binding pocket (BP), obtained from ONIOM2A and ONIOM2B calculations

Energy components (kcal mol⁻¹)	ONIOM2A		ONIOM2B
Energy components (kcal mol⁻¹)	WT	K103N mutant	WT	K103N mutant
BE	−10.66	−5.65	−12.71	−6.90
INT	−12.97	−10.30	−15.16	−13.61
DEF	2.31	4.64	2.45	6.70
DEF_EFV	1.02	1.51	1.17	1.44
DEF_BP	1.29	3.14	1.28	5.27

The deformation energy represents the energy for the conformational change of the ligand to bind into the binding pocket. The results in Table 2 clearly show that the DEF of the K103N binding pocket requires more energy for the structural rearrangements than that of the WT by approximately 3.99 kcal mol⁻¹ as obtained from the ONIOM2B calculations. The reason to explain are due to the fact that the mutant binding pocket needs more energy than that of the wild type and these caused the loss of inhibitory affinity of the drug.

We then compare the different ONIOM methods that used to treat the model layer between EFV to both WT and K103N systems (see ESI†). The BE of the enzyme-inhibitor complexes that obtained from ONIOM2A is more slightly higher than ONIOM2B approximately 2 and 1 kcal mol⁻¹ for WT and K103N, respectively. The effect of hydrogen bonding of K101 residue on the binding energy of the first generation NNRTIs drug was studied.³³ It was found that there is not much different calculated results between B3LYP/6-31G(d,p) and HF/6-31G(d,p). Therefore, HF/6-31G(d,p) method was used to treat the inner layer for ONIOM2C calculations in order to reduce the computational time. This model included the effect of the hydrogen bonding between EFV and the backbone of K101.

To this respect, the result shows the BE, INT and DEF of EFV bound to the K103N RT with and without the K101 residue in the inner layer (see ESI†). The obtained result from ONIOM2C are stronger binding energy than ONIOM2A of about 3.22 kcal mol^¬1. Considering the decomposition of BE components, it can be seen that the INT of ONIOM2A and ONIOM2C are significantly different. Thus inclusion of the K101 residue into the inner layer (ONIOM2C) gives the strong interaction energy of about 5.11 kcal mol⁻¹ as compared to the ONIOM2A calculations.

Based on these knowledge, inhibitor rearrangement could be adapted as the result from mutated NNIBP which appears much significantly important for the next generation NNRTIs. From these observations, it can be concluded that the hydrogen bonding between EFV inhibitor and the backbone of K101 residue has significant effect on the binding energy calculations which is the primary mechanism of drug binding to both WT and K103N RT. Also, the K103N causes significantly large conformational rearrangement in term of deformation energy of the binding pocket as compared to the WT RT. The results agree with the previous work that calculated from molecular mechanics for free energy of binding.³⁷

Conclusions

Quantum chemical calculations was applied to investigate the pairwise analysis between EFV and individual residue in the binding pocket of both WT and K103N RT. The results show that the K103N RT is more repulsive interactions between EFV and residues surrounding the binding pocket than of the WT structure. The results from pairwise energies perfectly demonstrate that the K103N RT slightly affect on the loss of interaction energy. The main influences are due to residues around the binding pocket (Lys101, Lys102, Ser105, Val179, Trp229, Pro236 and Glu138). These serve as the main contribution for drug binding which support the results obtained from ONIOM calculations.

Based on ONIOM2 calculations, these evidently show that the K103N decreases the stabilization energy of EFV bound to its binding pocket. Furthermore, the analyses of the energy components in terms of interaction energy and deformation energy also shows a significant structural rearrangement upon EFV binding, and more energy is needed for conformational adaptation in the binding pocket. The potent inhibitor accommodates the K103N RT by the formation of an additional interaction to the asparagine side chain and minor rearrangement of the inhibitor position in the binding pocket. It is worth to note that the K101 and S105 residues are important as strong interaction for further novel inhibitor development. Taken into account, the K101 residue shows the key main interaction of the binding due to the presence of hydrogen bonding between EFV and the backbone of K101.

Acknowledgements

This work was supported by grants from the Thailand Research Fund (TRF) and the Commission on Higher Education (RTA5380010). P.B. is grateful to the Royal Golden Jubilee Ph.D. Scholarship (PHD02622549). The National Center of Excellence in Petroleum, Petrochemical Technology and Advanced Materials are gratefully acknowledged for research facilities. Partial supporting by NECTEC and NANOTEC, Ministry of Science and Technology, Thailand, through its program of Center of Excellence Network is grateful. Thanks are due to Dr Matthew Paul Gleeson for reading of the manuscript.

Notes and references

R. C. Gallo, S. Z. Salahuddin, M. Popovic, G. M. Shearer, M. Kaplan, B. S. Haynes, T. J. Palker, R. Redfield, G. Oleske, B. Safai, G. White, P. Foster and P. D. Markham, Science, 1984, 224, 500–503 CAS.
E. Muraglia, O. D. Kinzel, R. Laufer, M. D. Miller, G. Moyer, V. Munshi, F. Orvieto, M. C. Palumbi, G. Pescatore, M. Rowley, P. D. Williams and V. Summa, Bioorg. Med. Chem. Lett., 2006, 16, 2748–2752 CrossRef CAS.
M. Udier-Blagovi, E. K. Watkins, J. Tirado-Rives and W. L. Jorgensen, Bioorg. Med. Chem. Lett., 2003, 13, 3337–3340 CrossRef.
K. Das, J. Ding, Y. Hsiou, A. D. ClarkJr, H. Moereels, L. Koymans, K. Andries, R. Pauwels, P. A. J. Janssen, P. L. Boyer, P. Clark, R. H. Smith Jr, M. B. K. Smith, C. J. Michejda, S. H. Hughes and E. Arnold, J. Mol. Biol., 1996, 264, 1085–1100 CrossRef CAS.
R. R. Sergio, A. Q. Mario, M. Marcela and C. B. Margarita, J. Chem. Inf. Model., 2011, 51, 130–138 CrossRef.
L. T. Bacheler, E. D. Anton, P. Kudish, D. Baker, J. Bunville, K. Krakowski, L. Bolling, M. Aujay, X. V. Wang, D. Ellis, M. F. Becker, A. L. Lasut, H. J. George, D. R. Spalding, G. Hollis and K. Abremski, Antimicrob. Agents Chemother., 2000, 44, 2475–2484 CrossRef CAS.
L. Bacheler, S. Jeffrey, G. Hanna, R. D'Aquilla, L. Wallace and K. Logue, J. Virol., 2001, 75, 4999 CrossRef CAS.
B. Clotet, Int. J. Clin. Pract. Suppl., 1999, 103, 21–25 CAS.
D. Kalyan, D. B. Joseph, S. R. Angela, D. Chhaya Jr, D. C. Arthur, C. María-Jos, B. Jan and A. Eddy, J. Med. Chem., 2011, 54, 2727–2737 CrossRef.
S. D. Young, S. F. Britcher, L. O. Tran, L. S. Payne, W. C. Lumma, T. A. Lyle, J. R. Huff, P. S. Anderson, D. B. Olsen, S. S. Carroll, D. J. Pettibone, J. A. O'Brien, R. G. Ball, S. K. Balani, J. H. Lin, I. W. Chen, W. A. Schleif, V. V. Sardana, W. J. Long, V. W. Byrnes and E. A. Emini, Antimic. Agents. Chemo., 1995, 39, 2602–2605 CAS.
P. Nunrium, M. Kuno, S. Saen-oon and S. Hannongbua, Chem. Phys. Lett., 2005, 405, 198–202 CrossRef CAS.
P. Srivab and S. Hannongbua, Chem. Med. Chem., 2008, 3, 803–811 CrossRef CAS.
C. F. Venezia, K. J. Howard, M. E. Ignatov, L. A. Holladay and M. D. Barkley, Biochemistry, 2006, 45, 2779–2789 CrossRef CAS.
M. Udier-Blagovic, E. K. Watkins, J. Tirado-Rives and W. L. Jorgensen, Bioorg. Med. Chem. Lett., 2003, 13, 3337–3340 CrossRef CAS.
D. Richman, A. S. Rosenthal, M. Skoog, R. J. Eckner, T. C. Chou, J. P. Sabo and V. J. Merluzzi, Antimicrob. Agents. Chemother., 1991, 35, 305–308 CAS.
T. J. Dueweke, S. M. Poppe, D. L. Romero, S. M. Swaney, A. G. So, K. M. Downey, I. W. Althaus, F. Reusser, M. Busso, L. Resnick, D. L. Mayers, J. Lane, P. A. Aristoff, R. C. Thomas and W. G. Tarpley, Antimic. Agents. Chemo., 1993, 37, 1127–1131 CAS.
K. Das, A. D. ClarkJr, P. J. Lewi, J. Heeres, M. R. de Jonge, L. M. H. Koymans, H. M. Vinkers, F. Daeyaert, D. W. Ludovici, M. J. Kukla, B. De Corte, R. W. Kavash, C. Y. Ho, H. Ye, M. A. Lichtenstein, K. Andries, R. Pauwels, M. P. de Bethune, P. L. Boyer, P. Clark, S. H. Hughes, P. A. J. Janssen and E. Arnold, J. Med. Chem., 2004, 47, 2550–2560 CrossRef CAS.
E. B. Lansdon, K. M. Brendza, M. Hung, R. Wang, S. Mukund, D. Jin, G. Birkus, N. Kutty and X. Liu, J. Med. Chem., 2010, 53, 4295–4299 CrossRef CAS.
X. He, Y. Mei, Y. Xiang, D. W. Zhang and J. Z. H. Zhang, Proteins: Struct., Funct., Bioinf., 2005, 61, 423–432 CrossRef CAS.
Y. Mei, X. He, Y. Xiang, D. W. Zhang and J. Z. H. Zhang, Proteins: Struct., Funct., Bioinf., 2005, 59, 489–495 CrossRef CAS.
M. Udier-Blagovic, J. Tirado-Rives and W. L. Jorgensen, J. Med. Chem., 2004, 47, 2389–2392 CrossRef CAS.
M. B. Kroeger Smith, L. H. Rader, A. M. Franklin, E. V. Taylor, K. D. Smith, R. H. Smith Jr., J. Tirado-Rives and W. L. Jorgensen, Bioorg. Med. Chem. Lett., 2008, 18, 969–972 CrossRef CAS.
P. Weinzinger, S. Hannongbua and P. Wolschann, J. Enzyme Inhib. Med. Chem., 2005, 20, 129–134 CrossRef CAS.
N. Kitisripanya, P. Saparpakorn, P. Wolschann and S. Hannongbua, Nanomed.: Nanotechnol., Biol. Med., 2011, 7, 60–68 CrossRef CAS.
J. Ren, J. Milton, K. L. Weaver, S. A. Short, D. I. Stuart and D. K. Stammers, Structure, 2000, 8, 1089–1094 CrossRef CAS.
M. Kuno, R. Palangsuntikul and S. Hannongbua, J. Chem. Inf. Model., 2003, 43, 1584–1590 CrossRef CAS.
S. Promsri, P. Chuichay, V. Sanghiran, V. Parasuk and S. Hannongbua, THEOCHEM, 2005, 715, 47–53 CrossRef CAS.
M. Kuno, S. Hannongbua and K. Morokuma, Chem. Phys. Lett., 2003, 380, 456–463 CrossRef CAS.
S. Saen-oon, S. Hannongbua and P. Wolschann, J. Chem. Inf. Model., 2003, 43, 1412–1422 CrossRef CAS.
M. Svensson, S. Humbel, R. D. J. Froese, T. Matsubara, S. Sieber and K. Morokuma, J. Phys. Chem., 1996, 100, 19357–19363 CrossRef CAS.
S. Dapprich, I. Komaromi, K. S. Byun, K. Morokuma and M. J. Frisch, THEOCHEM, 1999, 461–462, 1–21 CrossRef CAS.
G. S. Tschumper and K. Morokuma, THEOCHEM, 2002, 592, 137–147 CrossRef CAS.
J. Lindberg, S. Sigurdsson, S. Löwgren, H. O. Andersson, C. Sahlberg, R. Noréen, K. Fridborg, H. Zhang and T. Unge, Eur. J. Biochem., 2002, 269, 1670–1677 CrossRef CAS.
S. Saen-oon, M. Kuno and S. Hannongbua, Proteins: Struct., Funct., Bioinf., 2005, 61, 859–869 CrossRef CAS.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, et al., Gaussian03, Pittsburgh, Gaussian, Inc, 2003 Search PubMed.
R. K. Raju, N. A. Burton and I. H. Hillier, Phys. Chem. Chem. Phys., 2010, 12, 7117–7125 RSC.
Z. Zhang, M. Zheng, L. Du, J. Shen, X. Luo, W. Zhu and H. Jiang, J. Comput.-Aided Mol. Des., 2006, 20, 281–293 CrossRef CAS.

Footnote

† Electronic supplementary information (ESI) available: Figure of interatomic distances between EFV and the residues in the binding pocket for WT and K103N mutation and also the results of included the effect of the hydrogen bonding between EFV and the backbone of K101 by using ONIOM calculations. See DOI: 10.1039/c1md00162k

Click here to see how this site uses Cookies. View our privacy policy here.

Key interactions of the mutant HIV-1 reverse transcriptase/COMPOUND LINKSRead more about this on ChemSpiderDownload mol file of compoundefavirenz: an evidence obtained from ONIOM method†