Sars

Tuesday , May 31, 2005

Dimer of SARS Coronavirus Main Proteinase as Studied by Molecular Dynamics Simulations Vannajan Sanghiran Lee,a,b Tawun Remsungnen,a Siriporn Promsri,a Sornthep

py

Vannarat,d Kitiyaporn Wittayanarakul,a Ornjira Aruksakulwong,a Vudhichai Parasuk,a Pornthep Sompornpisut,a Suwipa Saen-oon,a Wasun Chantratita,c Piyawut Srichaikul,d Supot Hannongbuaa,*

Department of Chemistry, Faculty of Science, Chulalongkorn University, Bangkok

Co

a

10330, Thailand b

Department of Chemistry, Faculty of Science, Chiang Mai University, Chiang Mai

50200, Thailand c

Virology and Molecular Microbiology, Department of Pathology, Faculty of

d

iew

Medicine, Ramathibodi Hospital, Mahidol University, Bangkok 10400, Thailand High Performance Computing Research and Development Division, National

Electronics and Computer Technology Center, Thailand Science Park, 112

Re v

Pahonyothin Road, Klong 1, Klong Luang , Pathumthani 12120, Thailand

TOTAL PAGE 25

TOTAL TABLES 1

TOTAL FIGURES 7

*Corresponding author. Tel.: +66 2218 7602; fax: +66 2218 7603; e-mail address: [email protected]

Elsevier

1 of 25

Tuesday , May 31, 2005

ABSTRACT The severe acute respiratory syndrome - coronavirus (SARS-CoV), a new family of the coronavirus was finally proven to be the cause of SARS. The three-dimensional structure of the enzyme target is required in the development of diagnostic tests,

py

antiviral therapies, and vaccines. Models of SARS-CoV proteinase in monomer and dimer forms were proposed. Although, the monomer form was assumed to be a target for Anti-SARS drug design, its appropriateness as a model has not yet been proven.

Co

The molecular dynamics simulations revealed significant discrepancy between the monomer and dimer models. The deviation of the electrostatic potential confirms the conclusion on the different roles of the two forms. The flexibility of the substratebinding loop regions and the overall residue flexibility in each chain of the dimer form were discussed. Evidence show that the different conformations and interactions

iew

between SARS-CoV proteinase and substrate may be related to the N-terminal interacting with the other chain when the dimer was formed.

Re v

Keywords: SARS; coronavirus; proteinase; MD simulation

Elsevier

2 of 25

Tuesday , May 31, 2005

1. INTRODUCTION The unprecedented technology in the post genomics era, particularly in molecular biology and bioinformatics focused on the severe acute respiratory syndrome (SARS) outbreak is beginning to yield hints about how to fight the disease. Within 3 weeks of

py

searching for the aetiological agent, scientists in several laboratories were able to sequence the novel sequences of the ~30,000-nucleotide RNA genomes of two isolates of the SARS-CoV [1-2]. The genome showed all the feature characteristics of

Co

a coronavirus, but is significantly different from all previously known coronaviruses. Thus, the SARS-CoV represents a new coronavirus group. The genomes of the SARS-CoV Tor2 strain from Toronto [2] and the Urbani strain from Vietnam [1] differ by just eight nucleotides. The amino acid sequences of SARS-CoV isolated from five patients in Singapore and those from Canada, the United States, and China

iew

have been compared. The research teams from the four countries reported that overall, the virus is relatively stable compared to many other RNA viruses [3]. On May 13, 2003 German researchers proposed a probable structure of one of the key proteins involved in the virus's replication, based on the homology modeling technique using the available sequence alignment a development that could provide a possible drug

Re v

target [4]. The predicted amino acid sequences of the viral proteinase enzyme, which is part of the viral replicase polyprotein, and the spike (S) glycoproteins, the envelope (E) protein, the membrane (M) protein, the nucleocapsid (N), and the 3C like (3CL) proteinase of SARS-CoV suggests that they are structurally and functionally homologous to the proteins of known coronaviruses. The pair-wise amino acid sequence identity and similarity with their homologs is about 35 to 65% [5], respectively. The 3CL proteinase is supposed to be one of the potential targets for the

Elsevier

3 of 25

Tuesday , May 31, 2005

development of new drugs. If so, proteinase inhibitors could prevent the process of RNA polymerase or cleavage of the viral S glycoprotein.

Recently, the X-ray crystal dimer structure of SARS-CoV proteinase has been

py

reported under PDB code 1Q2W.pdb [6]. The ten missing amino acid residues in the binding site vicinity were T45, A46, E47, and D48 for chain A and the residues A46, E47, D48, M49, L50, and N51 for chain B.

Co

Many structural studies have modeled the SARS-CoV proteinase monomer structures based on the crystal structure of an enzyme from the human coronavirus (HcoV) E229 (one cause of the common cold) [4] and an enzyme from a pig coronavirus (transmissible gastroenteritis virus, TGEV) in complex with a peptide inhibitor [4, 7-

iew

9]. In fact, many of these proteins carry out their biological function as multimeric assemblies [10]. Analysis of the dimeric interface suggests that in a solution, the 3CL proteinase may be in dimer form [11]. Dynamic light scattering studies found that both HCoV 299E and TGEV Mpro exist as a mixture of both monomer (~65%) and dimer (~35%) [12] forms. Strong evidence from the X-ray structure (PDB code:

Re v

1Q2W) [6] exhibits the dimerization of the monomer structure.

In the present article, 3D dynamics model of monomer and dimer structures of free SARS coronavirus proteinase and enzyme-substrate analog inhibitor (Thr-Ser-AlaVal-Leu-Gln) in solution were simulated. No attempt was made to understand the mechanism and evolutionary pathway of their formation, but rather an investigation of the different behaviors between the monomer and dimer models.

Elsevier

4 of 25

Tuesday , May 31, 2005

2. METHODS 2.1. Molecular Dynamics Simulations The structure-based sequence alignment of SARS-CoV proteinase which consists of 306 amino acids (GenBank accession number AY278741) was recently obtained by

with mutation data matrix, BLOSUM [1-2, 4].

py

Bioedit software, CLUSTAL X (1.81), and the multiple sequence alignment software

Co

Because the proteinases from SARS-CoV and the pig transmissible gastroenteritis viruses (TGEV) share high homology (35% identity and 65% similarity), the 3D model of the SARS-CoV proteinase in monomer form (Mono-Free) was generated by homology modeling technique [13] using the X-ray structure of the TGEV (PDB data bank: code 1LVO) [12] as the template. The dimer form (Di-Free) was generated by

iew

superposing the Mono-Free with the dimer form (Chain A and B) of the TGEV proteinase. For the structures of the monomer complex (Mono-Cpx) and the dimer complex (Di-Cpx), the SARS-CoV substrate-analog inhibitor (Thr-Ser-Ala-Val-LeuGln) derived from the P6-P1 residues of the NH2-terminal auto-processing site of SARS-CoV Mpro [4] was inserted into the SARS-CoV proteinase in the binding

Re v

pocket close to H41 and C145 of the Mono-Free and Di-Free model. Note that only chain B of the dimer formed a complex with the substrate-analog inhibitor. This configuration is analogous to that found for the porcine coronavirus proteinaseinhibitor complex.

MD simulations were then carried out for the systems consisting of the following 4 forms of the SARS-CoV proteinase: Mono-Free; Di-Free; Mono-Cpx; Di-Cpx. The molecular mechanics potential energy minimizations and MD simulations were

Elsevier

5 of 25

Tuesday , May 31, 2005

carried out using a program package, AMBER7 [14-15]. Calculations were performed using the parm99 force field. All MD runs reported here were done under an isobaricisothermal ensemble (NPT) using a constant pressure of 1 atm and a constant temperature of 298 K. The Mono-Free, Di-Free, Mono-Cpx, and Di-Cpx were

py

solvated by 9572, 10746, 19370 and 19357 TIP3P water molecules in a cubical cell and treated in the simulation under periodic boundary conditions. Energy minimizations were carried out to relax the system prior to MD runs. A cutoff distance of 12 Å was applied for non-bonded pair-wise interactions. The Particle

Co

Mesh Ewald (PME) method was employed for correcting electrostatic interactions. Sodium and Chloride ions were added to neutralize the system. The simulation time step was set at 2 femtosecond (fs). The temperature of the system was gradually raised to 298 K for the first 60 picosecond (ps) and then kept constant according to the

iew

Berendsen algorithm [16] with a coupling time of 0.2 ps. The total simulation time was carried over 1 ns for each system. The quality of the geometry and the

Re v

stereochemistry of the protein structure were validated using PROCHECK [17].

Elsevier

6 of 25

Tuesday , May 31, 2005

3. RESULTS AND DISCUSSIONS

3.1. Comparison between the homology and the MD structures

py

As mentioned earlier the 3D model of the SARS-CoV proteinase in monomer form (Mono-Free) was generated using homology modeling technique [13] starting from the X-ray structure of the porcine coronavirus proteinase. The PROCHECK [17]

Co

program was used to check the geometry and the stereochemistry of the modeled protein structure. The calculation of main chain torsion angles of the protein showed no severe distortion of the backbone geometry. A total of 97% of the backbone dihedrals,

and , fall within the structurally favorable regions in the Ramachandran

plot. The overall average of PROCHECK’s score including covalent geometry,

iew

planarity, main-chain hydrogen bonds, non-bonded interactions, dihedral angle, disulphide bond and chirality checks was above the minimum requirement.

The final structure of the SARS-CoV monomer from the MD simulations and that

Re v

available in the literature [12] determined by the homology modeling, were compared in Figure 1a. The root mean square displacement (RMSD) discrepancy between the homology and the MD structures of 3.1 Å was observed. The different algorithm method used and, in addition, the exclusion of the solvent effects in the homology modeling, leads to the missing of some specific structural details.

The difference between the molecular structures of the Mono-Cpx and chain B (inserted substrate inhibitors chain) of the Di-cpx leads to the RMSD of 3.0 Å (Figure 1b). The reason for this deviation is due to the interaction between the two

Elsevier

7 of 25

Tuesday , May 31, 2005

chains in the complex form of the dimer. Such interaction is not found in the monomer complex. This finding indicates that the use of the monomer form of the SARS proteinase in the investigation of guest-host complexation would lead to wrong results. Therefore, the dimer form of the SARS proteinase is suggested to be an

py

appropriate template for designing and screening of Anti-SARS drug. There are some evidences supporting this conclusion in detail of the next section. 3.2. Electrostatic potentials of monomer and dimer models

Co

In addition to the structural differences between the monomer and dimer in free form, electrostatic potential (ESP), one of the most important factor effecting enzymesubstrate binding, was calculated. The results are given in Figure 2. Visualization of the van der Waals electrostatic potential surface provides considerable insight into the different charge localizations between the binding sites of the monomer and dimer

iew

form in an aqueous solution. Although the global minima of the ESP of the two forms were observed to take place at almost the same region (marked by circle in Figure 2), the deviations of the ESP local minima in terms of both intensity and their distribution (red = -4.0 kT; white = 0 kT; blue = +2 kT). More negative electrostatic potential over the binding pocket surface was found in dimer model. Distinct substrate-binding

Re v

effect should be observed. This confirms the previous conclusion on the different role of the two forms due to the different shape of the ESP plots that can effect directly on the binding affinity between the enzyme and substrates or inhibitors.

Elsevier

8 of 25

Tuesday , May 31, 2005

3.3. Flexibility of the substrate-binding loop regions

In the substrate catalytic reaction by the SARS 3CL proteinase, H41 is supposed to act as an acid-base catalyst, and C145 as the nucleophilic attacking agent, identical to

py

the H41 and C145 of TGEV Mpro [4]. The conformational structure and the movement of the substrate-binding loop regions, which is known to relate to the enzymatic reaction, has been evaluated focusing on the amino acid sequence around the active

Co

sites (H41 and C145) of the coronavirus proteinase. The results are investigated in terms of the cavity size in which the distances d1 from the center of mass of H41 to C145 and d2 from the center of mass of D48 and Q189 were defined (Figure 3a). Changes of d1 (Figure 3b) and d2 (Figure 3c) as a function of time were calculated. The average RMSDs for the free and complexed forms of the monomer (Mono-Free)

iew

and each chain of the dimer (Di-Free (A) and Di-Free (B)) were averaged from 8001200 ps and summarized in Table 1. Note that the inhibitor was only inserted into chain B.

Unexpected information is a decrease of d1 and an increase of d2 in changing from

Re v

monomer to dimer. This fact is observed for both free and complex forms. Although this finding is not yet understood, the opening of the outer region (d2) and the contracting of the inner one (d1) suggests that the two chains of the dimer play

different roles in the catalytic process. An increase of the outer regions of both chains (d2 distance) would facilitate the complex formation by opening the entrance of the active site, and hence, reducing the energy barrier when the substrate enters the active site. The effect of the loop flexibility via substrate binding were also detected and shown in Figure 3c. The d2 distance in the Di-Cpx chain B (dark blue line) is shorter

Elsevier

9 of 25

Tuesday , May 31, 2005

than that of chain A (no substrate, green line). This is in consistent with the data shown in Table 1 in which the d2 distances of the two chains of the dimer in free form were almost equivalent while d2 of 12.93 Å for chain A in the complex form is

3.4. Residue flexibility in each chain of dimer

py

significantly longer than that of 9.66 Å for chain B.

To seek for more structural details of the dimer, RMSDs of all residues from each

Co

chain of the Di-free compared to their average structure after 800 ps simulations were evaluated and plotted in Figure 4a with their subtraction (RMSD of chain A minus that of chain B). The flexibility of each residue in the two chains is exhibited by the RMSD peaks (solid and dash lines). Comparing the total positive and negative filled

iew

areas in Figure 4a, different flexibilities of the two chains due to inter-chain contacts were found. Higher flexibility of chain B than that of chain A is clearly exhibited by the several negative subtracted peaks (RMSD of each residue of chain A minus that of chain B). The major changes (RMSD > ±0.5 Å) in residue flexibility are highlighted on the resudues F3, D48, N72, G109, A129, D153, P168, N228, and G275 (Figures

Re v

4a and 4b). The residues D48 and P168 are in the binding pocket vicinity which can cause the different binding. The superposition of the average monomer structure for all atoms with dimer chain A and chain B shows the difference in the structures with RMSDs of 2.81 Å and 2.74 Å, respectively. This indicates nonequivalent role of the two chains of the dimer of the SARS proteinase and confirms our previous conclusion that the dimer form should be considered in the investigation of the guest-host complexation.

Elsevier

10 of 25

Tuesday , May 31, 2005

To understand and clarify the above conclusions, detailed analysis was performed focusing on the N-terminal of both chains (Figure 5) in which its folding approaches the catalytic area of the other chain. This is fully consistent with the report for the dimer structure of the porcine coronavirus where the N-terminal takes part in the

py

catalytic process, known as auto-processing activity [12].

3.5. N-terminal induced conformational changes in the binding pocket

Co

In the Mpro dimer, the NH2-terminal amino acid residues from the other chain seems to have a structural role for the mature Mpro. The deletion of residues 1 to 5 leads to a decrease in activity of 0.3% in the standard peptide-substrate assay [12]. Although the binding orientation of the short peptide inhibitor to the SARS-CoV binding site (H41

iew

and C145) is well conserved, some changes occurred in the orientation between the inhibitor and the N-terminal domain of SARS-CoV involving movement of the Nterminal of the SARS-CoV with respect to the binding site. A closer look at the two N-terminal regions of the dimer structure in Figure 5, the amino acid residues shown in black were observed to form hydrogen bonds with the N-terminus of chain A

Re v

(N_A) and chain B (N_B). Four amino acid residues, F140_A, E166_A, H172_A and N214_B were detected to locate around the N-terminal of chain B whereas K137_B, E166_B, G170_B, and N214_A were interacted with the N-terminal of chain A. In addition, E166 was found to interact with the N-terminal of both chains. These results are corresponding well to the recent report of the crystal structure by Yang and coworkers [18] mentioned the important role of N-terminal in the dimerization and formation of the active site of Mpro. Furthermore, the comparison between the

orientation of E166_A and E166_B in their substrate-binding loops has indicated that

Elsevier

11 of 25

Tuesday , May 31, 2005

E166_B locates closer to its substrate-binding loop region more than that of E166_A. This may lead consequently to more conformational change in the binding pocket of chain B than the other one. This explains the dissimilarity of the binding pocket of

3.6. Conformations of the catalytic residues

py

each chain in the dimer.

More details have been investigated on the conformation of the residues in the

Co

binding site and their interactions with the substrate. The orientation of the H41 and C145 residues of the two chains of the dimer were numerically averaged. The torsional angle, T1 (CD-CG-CB-CA) of the H41 and T2 (N-CA-CB-SG) of C145, in the binding site were defined (Figure 6a). Changes of the two angles for the monomer

iew

and each chain of the dimer in the free form as a function of time steps were schematically compared in Figures 6b-6c.

In Figure 6b, the flexibility of the H41 residue has been clearly visualized. Surprisingly, the T1 angles for both Mono-Free and Di-Free fluctuates in the range of

Re v

approximately 100o. For example, T1 angles of Mono-Free are in the range between 80o and 180o with the average value of 133o. The plot shows the two possible states of the selected torsion of H41 in which the optimal orientation takes place at the average T1 value of 26o and 133o. Among the hydrogen bonds between the enzyme and substrate, only one of them binds with the catalytic residues of the SARS proteinase, O atom of Q312 and H atom of C145 in the Mono-Cpx. In addition, only the first configuration (T1 = 26o) was detected for Mono-Free and chain A of Di-Free and 2 configurations (T1 = 26o and 133o) were observed for chain B of Di-free. For the

Elsevier

12 of 25

Tuesday , May 31, 2005

C145 binding site, the T2 torsion angles of the three plots fluctuates within 50o, half that of H41, indicating a less flexibility of the C145 than H41 binding site. The T2 average angles for chain A and B of the dimer are -66o and -150o, respectively. Here, the specified group of C145 catalytic residue of the monomer in free form is stable in

py

both configurations.

Taken into account all the data in Figures 6b-6c, the high flexibility of the catalytic residues, especially the H41, is supposed to facilitate the enzyme-substrate (or

Co

inhibitor) binding in the catalytic process. In addition, different orientations of the two catalytic residues, H41 and C145, in the binding pocket of both chains of dimer as well as of the monomer support our previous conclusion on the different roles of the three catalytic pockets. This leads directly to selectivity of the binding site. Thus, this

iew

would yield different results when using various models, monomer or dimer, in drug designing and screening.

3.7. Interaction between SAR-CoV proteinase and substrate

Re v

Interest is centered on the complex conformation. Snapshots of the enzyme-substrate equilibrium geometry were taken from the trajectory yielded from the molecular dynamics simulation. Possible hydrogen bonding between substrate and enzyme of Mono-Cpx and Di-Cpx were shown in Figures 7a and 7b, respectively. Based on the donor-acceptor distance criteria (distance < 3 Å and angle < 60 degree), six hydrogen bonds were detected between substrate and enzyme in the Mono-Cpx , five hydrogen bonds were found in the Di-Cpx. Besides the different in hydrogen bonding, hydrophobic residues within 5 Å from the substrate in the monomer and dimer models

Elsevier

13 of 25

Tuesday , May 31, 2005

are also different. L50, F140, L141, G143, C145, M165, L167, P168, and A191 were found in the monomer model whereas L141, G143, M165, L167, P168, F185, V186, A191 and, A193 were detected in the dimer model.

py

In conclusion, all amino acid residues associated in hydrogen bonding plays a role in holding the substrate in the configuration suitable for the catalytic process. It is reconfirmed by the enzyme-substrate binding that the monomer and the dimer forms play different roles in the catalytic process. As it is known that the proteinase is

Co

usually active in the dimer form [10-12] the proposed 3D structure in the dimer form of the SARS proteinase is an appropriate template for designing and screening of Anti-SARS drugs.

iew

4. CONCLUSIONS

MD simulations and analysis of Mono-Free, Di-Free, Mono-Cpx, and Di-Cpx systems give an insight into the dynamic characteristics of monomer and dimer models in term of the flexibility, electrostatic potential, N-terminal interaction, conformation and

Re v

hydrogen bonding with the substrate. These results suggest that the dimer form is an active form for the enzyme-substrate binding and should be used to determine the binding direction of inhibitors for SARS-CoV proteinase. Also, the dimer form is an appropriate target for the design of Anti-SARS drugs.

Elsevier

14 of 25

Tuesday , May 31, 2005

ACKNOWLEDGEMENTS

This work was financially supported by the Thailand Research Fund (RTA4680008). The authors would like to thank C. Sangma, S. Hannongbua, P. Saparpakorn, W.

py

Treesuwan, E. Pasomsub and D. Chuakheaw for their comments and suggestions. The structure-based sequence alignment of coronavirus proteinase was done at the Department of Pathology, Faculty of Medicine, Ramathibodi Hospital, Mahidol All simulations and calculations were performed on multiple Linux

Co

University.

cluster systems where computing facilities were provided by the National Electronics and Computer Technology Center (NECTEC), the Computing Center for advance research at the Faculty of Science, Chulalongkorn University, the Computational Chemistry Unit Cell at Chulalongkorn University, and the Chemistry Department,

Re v

iew

Faculty of Science, Kasetsart University.

Elsevier

15 of 25

Tuesday , May 31, 2005

REFERENCES

[1] P.A. Rota, M.S. Oberste, S.S. Monroe, W.A. Nix, R. Campagnoli, J.P. Icenogle, S. Penaranda, B. Bankamp, K. Maher, M.H. Chen, S. Tong, A. Tamin, L. Lowe, M.

py

Frace, J.L. DeRisi, Q. Chen, D. Wang, D.D. Erdman, T.C. Peret, C. Burns, T.G. Ksiazek, P.E. Rollin, A. Sanchez, S. Liffick, B. Holloway, J. Limor, K. McCaustland, M. Olsen-Rasmussen, R. Fouchier, S. Gunther, A.D. Osterhaus, C.

Co

Drosten, M.A. Pallansch, L.J. Anderson, W.J. Bellini, Science 300 (2003) 1394.

[2] M.A. Marra, S.J. Jones, C.R. Astell, R.A. Holt, A. Brooks-Wilson, Y.S. Butterfield, J. Khattra, J.K. Asano, S.A. Barber, S.Y. Chan, A. Cloutier, S.M. Coughlin, D. Freeman, N. Girn, O.L. Griffith, S.R. Leach, M. Mayo, H.

iew

McDonald, S.B. Montgomery, P.K. Pandoh, A.S. Petrescu, A.G. Robertson, J.E. Schein, A. Siddiqui, D.E. Smailus, J.M. Stott, G.S. Yang, F. Plummer, A. Andonov, H. Artsob, N. Bastien, K. Bernard, T.F. Booth, D. Bowness, M. Czub, M. Drebot, L. Fernando, R. Flick, M. Garbutt, M. Gray, A. Grolla, S. Jones, H. Feldmann, A. Meyers, A. Kabani, Y. Li, S. Normand, U. Stroher, G.A. Tipples, S.

Re v

Tyler, R. Vogrig, D. Ward, B. Watson, R.C. Brunham, M. Krajden, M. Petric, D.M. Skowronski, C. Upton, R.L. Roper, Science 300 (2003) 1399.

[3] Y.J. Ruan, C.L. Wei, A.L. Ee, V.B. Vega, H. Thoreau, S.T. Su, J.M. Chia, P. Ng, K.P. Chiu, L. Lim, T. Zhang, C.K. Peng, E.O. Lin, N.M. Lee, S.L. Yee, L.F. Ng, R.E. Chee, L.W. Stanton, P.M. Long, E.T. Liu, Lancet 361 (2003) 1779.

Elsevier

16 of 25

Tuesday , May 31, 2005

[4] K. Anand, J. Ziebuhr, P. Wadhwani, J.R. Mesters, R. Hilgenfeld, Science 300 (2003) 1763.

[5] V.S. Lee, K. Wittayanarakul, T. Remsungnen, V. Parasuk, P. Sompornpisut, W.

py

Chantratita, C. Sangma, S. Vannarat, P. Srichaikul, S. Hannongbua, P. Saparpakorn, W. Treesuwan, O. Aruksakulwong, E. Pasomsub, S. Promsri, D. Chuakheaw, S. Hannongbua, ScienceAsia 29 (2003) 181.

Co

[6] J.B. Bonanno, R. Fowler, S. Gupta, J. Hendle, D. Lorimer, R. Romero, M. Sauder, C.L. Wei, E.T. Liu, S.K. Burley, T. Harris, To be published.

[7] B. Xiong, C.S. Gui, X.Y. Xu, C. Luo, J. Chen, H.B. Luo, L.L. Chen, G.W. Li, T.

iew

Sun, C.Y. Yu, L.D. Yue, W.H. Duan, J.K. Shen, L. Qin, T.L. Shi, Y.X. Li, K.X. Chen, X.M. Luo, X. Shen, J.H. Shen, H.L. Jiang, Acta Pharmacol Sin. 24 (2003) 497.

[8] K.C. Chou, D.Q. Wei, W.Z. Zhong, Biochem. Biophys. Res. Commun. 308

Re v

(2003) 148.

[9] H.L. Liu, J.C. Lin, Y. Ho, C.W. Chen, Chem. Phys. Lett. 401 (2005) 24.

[10] D.S. Goodsell, A.J. Olson, Annu. Rev. Biophys. Biomol. Struct. 29 (2000) 105.

[11] S. Liu, J. Pei, H. Chen, X. Zhu, Z. Liu, W. Ma, F. He, L. Lai, Beijing Da Xue Xue Bao 35 Suppl (2003) 62.

Elsevier

17 of 25

Tuesday , May 31, 2005

[12] K. Anand, G.J. Palm, J.R. Mesters, S.G. Siddell, J. Ziebuhr, R. Hilgenfeld, Embo J. 21 (2002) 3213.

py

[13] C. Combet, M. Jambon, G. Deleage, C. Geourjon, Bioinformatics 18 (2002) 213.

[14] D.A. Case, D.A. Pearlman, J.W. Caldwell, T.E. Cheatham III, J. Wang, W.S. Ross, C.L. Simmerling, T.A. Darden, K.M. Merz, R.V. Stanton, A.L. Cheng, J.J. Vincent, M. Crow-ley, V. Tsui, H. Gohlke, R.H. Radmer, Y. Duan, J. Pitera,

Co

I. Massova, G.L. Seibel, U.C. Singh, P.K. Weiner, P.A. Kollman, AMBER 7: University of California, San Francisco, 2002.

[15] D.A. Pearlman, D.A. Case, J.W. Caldwell, W. S. Ross, T.E. Cheatham III, S.

(1995) 1.

iew

DeBolt, D. Ferguson, G. Seibel, P.A. Kollman, Comp. Phys. Commun. 91

[16] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, A. DiNola, J.R. Haak, J.

Re v

Comp. Phys. 81 (1984) 3684.

[17] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J. Appl. Cryst. 26 (1993) 283.

[18] H. Yang, M. Yang, Y. Ding, Y. Liu, Z. Lou, Z. Zhou, L. Sun, L. Mo, S. Ye, H. Pang, G.F. Gao, K. Anand, M. Bartlam, R. Hilgenfeld, Z. Rao, Proc. Natl. Acad. Sci. U.S.A. 100 (2003) 13190.

Elsevier

18 of 25

Tuesday , May 31, 2005

FIGURES

Fig. 1. The superimpositions of SARS-coronarvirus proteinase (a) Mono-Free yielded from our MD simulation (grey) and from homology modeling (black) which available

py

in the literature [13]. The binding site was marked by arrow. (b) Mono-Cpx (grey) and Di-Cpx chain B (black) where both of them obtain from our MD simulations. The

Co

inhibitor inserted region was indicated by arrow.

Fig. 2. Electrostatic potential of the Mono-Free (a) and Di-Free (b) forms of the

iew

SARS coronarvirus proteinase (red = -4 kT; white = 0 kT;blue = +2 kT) in which the

Re v

most negative regions are marked by circles.

Elsevier

19 of 25

Tuesday , May 31, 2005

Fig. 3. Changes of d1 (a) and d2 (b) distances as a function of time step where d1 and

Re v

iew

Co

py

d2 are defined as those from H41-C145 and D48-Q189 shown in (a), respectively.

Elsevier

20 of 25

Tuesday , May 31, 2005

Fig. 4. a) Comparison of RMSD (dashed and dotted lines) with respect to their average structure for non-hydrogen atoms of the structures taken from the MD trajectory between 800 - 1200 ps and their residue-wise subtraction (filled bars) for the indicated pairs of the SARS-CoV proteinase structures. b) The flat ribbon

py

representation of the high flexibility residues (RMSD > 0.5 as marked in Figure

Re v

iew

Co

4a) were shown in ball and stick.

Elsevier

21 of 25

Tuesday , May 31, 2005

Fig. 5. Ribbon diagram shown the intermolecular contacts of SARS-CoV of Di-Free model (chain A in light grey and chain B in dark grey). N-terminus of chain A (N_A), N-terminus of chain B (N_B) and the associated amino acids residues were labeled. The upper case letters after the residue number indicate the corresponding chain (_A

Re v

iew

Co

py

for chain A and _B for chain B) marked as black.

Elsevier

22 of 25

Tuesday , May 31, 2005

Fig. 6. a) Definitions of the T1(CD-CG-CB-CA) and T2(N-CA-CB-SG) torsional angles at the H41 and C145 residues, respectively. (b) and (c) Changes of T1 and T2

Re v

iew

Co

py

angles for the indicated system, respectively.

Elsevier

23 of 25

Tuesday , May 31, 2005

Fig. 7. Possible hydrogen binding between substrate and the amino acid residues in the binding pocket of a) Mono-Cpx model b) Di-Cpx model.

Enzyme

R1) Q312(O)

S144(H)

R2) Q312(O)

C145(H)

R3) V310(O)

E166(H)

R4) V310(H)

E166(O)

R5) Q312(HE22) H163(NE2)

Fig. 7b. Di-Cpx

Enzyme

Re v

Substrate

T190 (O)

iew

R6) A309 (H)

Co

Substrate

py

Fig. 7a. Mono-Cpx

Elsevier

S1) Q312(H)

H41(NE2)

S2) V310(H)

E166(O)

S3) V310(O)

E166(H)

S4) Q312(OE1)

E166(H)

S5) S308(HG)

T190 (O)

24 of 25

Tuesday , May 31, 2005

TABLES Table 1 Changes of d1 and d2 distances, defined in Figure 3, the MD trajectory between 800-1200 ps where Mono, Di, Free , and Cpx stand for monomer, dimer, free

respectively.

Mono-

Mono-

Di-Free

Di-Free

Free

Cpx

(A)

(B)

d1 /Å

8.69

6.72

7.78

d2 /Å

6.22

6.53

8.38

py

form and complex form, which (A) and (B) denote chain A and chain B of the dimer,

Di-Cpx

Di-Cpx

(A)

(B)

7.64

6.56

6.55

8.17

12.93

9.66

Re v

iew

Co

Distance

Elsevier

25 of 25