Luận án tiến sĩ: Nghiên cứu tính toán TRPC6, purine biosynthesis và azobenzene oligomers

COMPUTATIONAL STUDIES TO UNDERSTAND MOLECULAR REGULATION OF THE TRPC6 CALCIUM CHANNEL, THE MECHANISM OF PURINE BIOSYNTHESIS, AND THE FOLDING OF AZOBENZENE OLIGOMERS DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Peng Tao, M. ***** The Ohio State University 2007 Dissertation Committee: Approved by Professor Christopher M. Hadad, Advisor Professor Russell M. Pitzer _________________________________ Professor James V.

Coe Advisor Graduate Program in Chemistry UMI Number: 3241692 UMI Microform 3241692 Copyright 2007 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.

Box 1346 Ann Arbor, MI 48106-1346 ABSTRACT Different computational chemistry methods were applied to study a variety of problems at the molecular level. These problems concern protein-protein interactions, the mechanism of reaction for enzymatic purine biosynthesis, structural interconversion in non-natural oligomeric folding, and carbohydrate synthesis. Transient receptor potential-canonical 6 (TRPC6) calcium channels are currently the subject of intense investigation for their role in modulating smooth muscle tone in blood vessels and lung tissue. Binding of a protein, FKBP12, is a prerequisite for the formation of a multiprotein complex involved in channel regulation.

To study the elements of molecular recognition in FKBP12 for binding to the TRPC6 intracellular domain, 20 nanosecond molecular dynamic simulations were performed on the complex of FKBP12 and a peptide model of the wild-type TRPC6 intracellular domain, a phosphorylated Ser768 analog of the wild-type peptide as well as Ser768Asp and Ser768Glu mutants. The phosphorylated peptide demonstrated the greatest binding affinity by the MM-GB/SA method, due to the strong interaction between the phosphate group and two lysine (Lys44 and Lys47) residues of FKBP12 at the binding site. These trajectories also revealed transient, non-simultaneous interactions with the ε-NH3⊕ group of these lysine residues. This feature was not observed in simulations containing the other peptides.

Decomposition of the binding free energies into each amino acid residue ii identified important additional structural elements necessary for this protein-protein interaction. Potential catalytic reaction mechanisms of the enzyme PurE Class I, which catalyzes the transformation from N5-carboxyaminoimidazole ribonucleotide (N5-CAIR) to 4-carboxyaminoimidazole ribonucleotide (CAIR) in the purine biosynthetic pathway, were investigated by density functional theory (DFT) methods. The potential energy surfaces (PES) of model processes for these enzymatic reactions have been explored, and have aided in identifying the most energetically feasible pathway. Calculations using a simplified model system, containing only the essential atoms involved in the chemical process, revealed seven potential reaction pathways for transformation of N5-CAIR to CAIR.

The experimental results exclude four of these pathways. Two of the remaining three pathways involve deprotonation of one carbon atom (C4) of the imidazole ring. This process makes the relative energy of transition states of these two pathways higher than the third pathway. The full N5-CAIR structure was studied via PES calculations, including consideration of the ribose-5-phosphate unit and its different charge states.

There are 48 structures of N5-CAIR and CAIR regarding the different protonation states of the substrate. Four reaction pathways were identified based on the available structures after optimization. One pathway involved a cationic substrate. Two involved anionic iii substrates.

A fourth one involved a neutral substrate. These four pathways have similar PES to their counterparts in the simplified model. The cationic pathway is the most favorable pathway in both the simplified and full reaction models. In this pathway, an intramolecular proton donor/acceptor is required before and after migration of the carboxyl group (-CO2H).

According to the spatial arrangement of catalytic amino acids at the active site of the PurE Class I crystal structure (PDB ID: 1D7A), a conserved histidine 45 (His45) residue could be such a proton donor/acceptor. Based on these results, a stepwise enzymatic reaction mechanism for PurE Class I is proposed. First, His45 at the PurE Class I active site protonates the amide nitrogen of N5-CAIR. Second, the carboxyl group migrates to carbon 4 (C4) with concomitant C-C bond formation.

This step generates the protonated CAIR intermediate. In the final step, His45 deprotonates the protonated CAIR intermediate to produce the final product, CAIR, and regenerates its initial state. For the first time, an atomistic description of the PurE Class I enzyme catalytic mechanism is provided. This information can be applied to the development of transition state analogs and mechanism- based inhibitors of PurE Class I.

These newly designed molecules could be used as potential new drug leads for optimization by synthetic and medicinal chemistry. Foldamers are defined as unnatural polymers/oligomers with a well-defined, compact, three-dimensional folding capability. Azobenzene units are common linkages in foldamer designs. Four alternating pyridinedicarboxamide/m-(phenylazo)azobenzene iv oligomers which could fold into both right- and left-handed helices were studied computationally for their dynamical properties.

Two helices were shown as the global minimum among the conformations generated by Monte Carlo simulation. Extended conformations have higher potential energies than compact ones. Molecular dynamics simulations (100 ns) at a single temperature were also performed to study the interconversion process between two helices of these oligomers. However, the molecules were trapped at the local minimum, and no interconversion was observed.

To overcome this difficulty, replica-exchange molecular dynamic (REMD) simulations which apply a parallel tempering algorithm were performed on the azobenzene oligomers. Both right- and left-handed helices were successfully sampled in the simulation for all four oligomers. Careful investigation of REMD trajectories revealed twisted conformations as intermediate structures in the interconversion pathway between two helices. The temperature weighted histogram analysis method (T-WHAM) was applied on the REMD simulation results to generate contour maps of the potential of mean force (PMF).

Atomic pair distances and dihedral angles were used as reaction coordinates for PMF contour plots. Analysis showed that right- and left-handed helices are equally sampled in REMD simulations. In large oligomers, both right- and left-handed helices could be adopted by different parts of the molecule simultaneously. The interconversion between two helices could occur in the middle of the helical structure, which is not necessarily at the end of the molecule.

v Oligosaccharides play a variety of important roles in a number of biological events, and the stereoselective synthesis of carbohydrates is critical. Computational studies have been done to understand the mechanism of the regioselective epoxide ring- opening process catalyzed by the presence of (-)-sparteine in the synthesis of β- arabinofuranosides. Using a simple diamine, N,N,N’,N’-tetramethylpropanediamine, to mimic (-)-sparteine, DFT was applied to search for transition states connected to two unique epoxide ring-opening events for attack at two different carbons of the epoxide ring. However, the reaction barriers calculated by DFT in this simplified model could not explain the observed experimental regioselectivity.

DFT single-point energy calculations using (-)-sparteine lead to similar results to those of the simplified model. Molecular dynamics (MD) methods were applied to simulate five systems containing the (-)- sparteine-Li+ complex and different substrates. MD simulations revealed interesting conformational changes of the (-)-sparteine unit. By switching between two conformers, (-)-sparteine could push Li+ toward the epoxide oxygen in a catalytically useful fashion.

Since the relative orientation of (-)-sparteine with respect to the furanose ring is largely influenced by steric effects due to the substituent group at the C5 position of substrate, (-)-sparteine helps Li+ selectively approach the epoxide oxygen on one face, which leads to specific ring-opening at the C3 position. In addition to traditional MD, ab initio molecular dynamics methods were applied to simulate four systems for the epoxide ring- opening process. The conformations of (-)-sparteine during the ring-opening process vi agree with the observation in our earlier classical MD simulations, and this conformational control provides the origin of the experimental selectivity. vii Dedicated to my beautiful wife Jin viii ACKNOWLEDGMENTS In the past five years, my advisor Dr.

Hadad gave me tremendous support for my study and research at the Ohio State University. Without his help, none of this would have been possible. His enthusiasm about science and high productivity in research set up a lifetime model for me to follow. I also would like to thank Dr.

Pitzer for allowing me to attend his group meetings in the past year. This enjoyable experience gave me lots of education in quantum chemistry. Many other faculty members in the physical division also educated me about physical chemistry. I would give my special thanks to Dr.

Coe as my committee member. Hadad’s research group members are always friends and colleagues for me. John C Hackett provided much help in my research and in critical reading of my manuscripts. Hayes and Mr.

DeMatteo also helped me with their collaboration and critical reading. I also want to thank all other group members for their friendship and help. Financial support from the National Science Foundation and the National Institute of Health is gratefully acknowledged. ix My mother, Mrs.

Xiuying Liang, my father, Mr. Junhe Tao, my mother-in-law, Mrs. Qiaoyun Liu, and my father-in-law, Mr. Binyan Liu all gave me strong support when I was far away from my home country, People’s Republic of China.

I want to dedicate this dissertation to the most important person in my whole life, my beautiful wife Dr. Your support was always with me during the completion of this dissertation. Your intelligence sheds lights on many of my scientific problems. Your enthusiasm about life always encourages me to move forward.

Any achievement would be meaningless for me without your sharing. You make me the luckiest person in the whole world with your endless love, support, and encouragement. No words are adequate enough to express my gratitude. A journey with you is the most beautiful thing I can ask for in my life.Born – Xinxiang, Henan Province, P.

Chemistry Peking University, Beijing, China 1994 – 1998. Chemistry Peking University, Beijing, China 2001 – 2005. Graduate Teaching and Research Associate, The Ohio State University. PUBLICATIONS Research Publications 1.

“Synthesis of L-Daunosamine and L-Ristosamine Glycosides via Photoinduced Aziridination. Conversion to Thioglycosides for Use in Glycosylation Reactions” J. “Protein ligand docking based on empirical method for binding affinity estimation. “Calculating partition coefficients of peptides by the addition method.

“Calculation of peptide's partition coefficients by amino acid addition model.” Wuli Huaxue Xuebao 1999, 15, 449-453. xi FIELDS OF STUDY Major Field: Chemistry xii TABLE OF CONTENTS Page Abstract. xi List of Tables. xviii List of Figures.

xxiii Chapters: 1 Introduction .1 TRPC ion channels .2 N5-CAIR mutase mechanism in purine biosynthesis pathway .4 Regioselective ring-opening of epoxides in 2,3-anhydrosugars .5 References for Chapter 1 .1 Density functional theory .2 Monte Carlo simulations .3 Molecular dynamics simulations .4 Quantum molecular dynamics .5 References for Chapter 2. 31 xiii 3 Molecular determinants of TRPC6 channel recognition by FKBP12 .1 TRPC6 peptide docking .2 Molecular dynamics simulation of FKBP12-TRPC6 peptide complexes and separated FKBP12 and TRPC6 peptides .3 MM/GB-SA free energy of binding calculations .3 Results and discussion .1 TRPC6 peptide docking .2 RMSD fluctuation of complexes .3 Residue fluctuation in each simulation .4 Binding free energy calculations by MM-GBSA based on Single-Trajectory .5 Free energy of binding calculations of FKBP12 with unphosphorylated and phosphorylated wild-type peptide complexes based on Single-Trajectory .6 Free energy of binding calculations of Ser768Asp and Ser768Glu mutants based on Single-Trajectory .7 Entropic contribution of each complex based on Single- Trajectory .8 Free energy of binding calculations of WT complex based on Three-Trajectory.9 Free energy of binding calculations of mutant complexes based on Three-Trajectory .10 Decomposition analysis of the free energy of binding based on Single-Trajectory .11 Decomposition analysis of the free energy of binding based on Three-Trajectory .12 Spatial distribution of residues with major contributions to the free energy of binding .13 Interactions among Lys44, Lys47 and Residue768 75 3.14 Comparison of RMSD fluctuation between bound and unbound peptides .15 Comparison of RMSD fluctuation between bound and unbound FKBP12 .5 References for Chapter 3. 91 xiv 4 Computational studies of the N5-CAIR mutase mechanism in the purine biosynthesis pathway .3 Results and Discussions .2 Thermochemistry for transformation of AMI to AMIC .