Jun 9, 2016 - utama yang terhasil daripada amiloid pelopor protein (APP) yang ..... Automated Topology Builder ..... mobile halo around the core structure.
MOLECULAR SIMULATION B ETWEEN AMYLOID B ETA (1-42), A PEPTIDE ASSOCIATED W ITH ALZHEIMER DIS EAS E, AND ZINC(II) ION
By NUR S YAFIQAH ABDUL GHANI
Thesis Submitted to the School of Graduate Studies, Uni versiti Putra Mal aysia, in Fulfillment of the Requirement for the Degree of Master of Science (MSc).
June 2016
All material contained within the thesis, including without limitation text , logos, icons, photographs and all other artwork, is copyright material of Universiti Putra Malaysia unless otherwise stated. Use may be made of any material contained within the thesis for non-commercial purposes from the copyright holder. Co mmercial use of material only be made with the exp ress, prior, written permission of Un iversiti Putra Malaysia.
Copyright © Universiti Putra Malaysia.
Abstract of thesis presented to the Senate of Universiti Putra Malaysia in fu lfillment of the requirement for the Degree of Master of Science
MOLECULAR SIMULATION B ETWEEN AMYLOID B ETA (1-42), A PEPTIDE ASSOCIATED WITH ALZHEIMER DIS EAS E, AND ZINC(II) ION
By NUR S YAFIQAH B INTI ABDUL GHANI June 2016
Chairman
: Roghayeh Abedi Karji ban, PhD
Faculty : Science Alzheimer’s disease (AD) is a brain disorder resulting fro m the accu mulation of amy loid-forming (both as amyloid-β and tau). Aβ peptide is present in everyone’s brain, but the amyloid plaques found in AD’s patients are abnormal, as they can degenerate nerve endings. The number of Alzheimer’s patients is increasing rapidly while there are no specific solutions being reported yet to treat AD effect ively. Amylo id-β(1-42) is a major frag ment fro m amylo id precursor protein (APP) which tends to aggregate into mature amylo id fibrils through a number of intermed iate structural forms, also called the oligo mers or protofibrils. They are to xic to neurons. The mechanis m by which Aβ aggregates in the brain is not fully understood, however there is increasing evidence that metal ions may play an important role in this aggregation process. In a healthy brain, the metal ion content is stringently regulated and the concentration of free metal ions is kept at a very lo w level.
Researchers nowadays are trying to uncover the neurodegenerative role of transition metals and the oxidative stress in AD which has been found to be responsible for major cellu lar p roblems. There are a vast number o f experiment studies trying to shed some light on these processes but the lack of theoretical studies on this matter is quite visib le. Here, we investigated the effect of zinc ion on Aβ (1-42) and its aggregation water and its mixtu re with hexafluoroisopropanol (HFIP) using molecu lar dynamics calc u lations. Fro m our results, the amylo id-β(1-42) fragment and its aggregated structure showed good stability in both conditions which were with and without zinc in water based on the root mean square deviation and radius of gyration calculations over 1 μs a nd 100 ns simu lation time for aggregation process. Besides that, Aβ (1-42) with and without zinc tend to produce more helical structures in solvent mixtu re, but no α -helix was detected in both Aβ-H2 O and Aβ-Zn-H2 O models.
The flexibility of Aβ (1-42) in solvent mixture was lower than Aβ (1-42) in water due to the length of its helical structure. In contrast, the presence of metal ion increased the flexib ility of Aβ (1-42) when the peptide was placed in the solvent mixture, co mpared to its flexib ility in water. Ou r aggregation study showed that 6Aβ-6Zn-HFIP-H2 O model had significant changes in secondary structures, compared to 6Aβ -6Zn-H2 O system. There was also a good correlation with the low flexib ility of peptide in water. In addition, Aβ (1-42) with zinc in water produced less helical structure compared to Aβ (1-42) with zinc in mixed solvent. As shown in secondary structure analysis, the aggregation process occurred rapidly in water after 20 ns compared to solvent mixture where the fully spherical structure was not shown in the mixed solvent.
ii
Abstrak tesis yang dikemu kakan kepada Senat Universit i Putra Malaysia sebagai memenuhi keperluan untuk Ijazah Master Sains
SIMULAS I MOLEKUL DIANTARA AMILOID B ETA (1 -42), PEPTIDA YANG B ERKAITAN DENGAN PEN YAKIT ALZHEIMER DAN ION ZINK(II)
Oleh NUR S YAFIQAH B INTI ABDUL GHANI Jun 2016
Pengerusi
: Roghayeh Abedi Karji ban, PhD
Fakulti : Sains
Penyakit Alzheimer merupakan penyakit gangguan otak yang disebabkan oleh pengumpulan amiloid yang terdiri daripada kedua-dua protein luar sel, amiloid beta (Aβ) dan intrasel (tau). Set iap manusia mempunyai peptida Aβ di dalam otak tetapi kandungan plak-plak amiloid yang ditemui pada pesakit Alzheimer adalah luar b iasa kerana me reka boleh menyebabkan kemerosotan hujung saraf. Kuantiti pesakit Alzheimer semakin men ingkat dengan mendadak sedangkan tiada penyelesaian yang spesifik d ilaporkan untuk merawat AD secara efektif. Aβ (1-42) merupakan pecahan utama yang terhasil daripada amiloid pelopor protein (APP) yang tercenderung untuk berku mpul (agregat) kepada gentian amilo id matang melalu i sebilangan bentuk struktur pertengahan yang juga dikenali sebagai oligo mer atau protofibril. Ia merupakan racun kepada neurons. Mekanisme untuk Aβ bergumpal di dalam otak masih belu m difahami sepenuhnya, walaubagaimanapun, terdapat banyak bukti ion -ion logam yang juga memainkan peranan dalam proses agregat ini. Otak yang sihat mempunyai kandungan ion logam yang spesifik dan kepekatan bagi ion logam yang bebas berada pada tahap yang sangat rendah. Para penyelidik pada masa kini sedang mencuba membongkar peranan neurodegeratif logam peralihan dan tekanan oksidatif yang merupakan punca bagi masalah selular secara keseluruhan. Pelbagai ujikaji telah d ilakukan untuk mencari sinar dalam proses ini tetapi kekurangan ilmu dalam b idang teori dalam kajian ini kelihatan jelas. Di dalam penyelidikan ini, kami mengkaji tentang kesan ion zink terhadap Aβ (1-42) dan penggumpalannya dalam air serta larutan bercampur yang mengandungi air dan larutan hexafluoroisopropanol (HFIP) dengan menggunakan kaedah pengiraan dinamik mo leku l.
iii
Keputusan kami menunjukkan bahawa struktur amyloid -β dan pengumpulan Aβ (1-42) menunjukkan kestabilan yang baik untuk semua keadaan iaitu dengan zink dan tanpa zink di dalam air berdasarkan varians sisihan punca min kuasa dua (RMSD) dan jejari legaran (Rg ) untuk 1 μs and 100 ns untuk proses penggumpalan masa simulasi. Selain itu, Aβ (1-42) dengan zink dan tanpa zink menghasilkan kuantiti alfa -helik yang banyak di dalam larutan bercampur tetapi t idak kelihatan alfa-helik langsung pada model AβH2 O dan Aβ-Zn-H2 O. Fleksibiliti Aβ (1-42) di dalam larutan bercampur adalah rendah berbanding dengan Aβ (1-42) dalam air disebabkan oleh alfa-helik yang panjang. Sebaliknya, kehadiran ion logam meningkat kan fleksibilit i Aβ (1-42) apabila peptida itu diletakkan dalam larutan bercampur berbanding dengan fleksibilit inya di dalam air. Kajian kami dalam proses pengumpulan telah menunjukkan bahawa model 6Aβ-6ZnHFIP-H2 O mempunyai perubahan yang ketara dalam struktur sekunder berbanding dengan sistem 6Aβ-6Zn-H2 O. Terdapat korelasi yang baik dengan peptida yang mempunyai fleksibiliti rendah dalam air. Tambahan pula, Aβ (1-42) dalam larutan bercampur dengan zink ion dalam air menghasilkan struktur helik yang rendah berbanding dengan Aβ(1-42) dan zink ion di dalam larutan bercampur. Proses penggumpalan berlaku dengan pantas di dalam air iaitu selepas 20 ns tetapi pembentukan sfera tidak dapat ditunjukkan sepenuhnya di dalam larutan bercampur.
iv
ACKNOWLEDGEMENTS
Special thanks goes to my supervisor, Dr. Roghayeh Abedi Karjiban for her knowledge, guidance, supervision, encouragement, helpful suggestions and valuable comments throughout the course of this project and supporting me throughout the undertaking of this research wh ich would have been impossible for me to acco mplish. I have gained a lot of knowledge and experience fro m doing this research.
Next, I would like to take this opportunity to say a warm thank you and greatest gratitude to my lab seniors, Alif Latif and Naimah Haron who helped me to prepare the proposal for the first presentation, checking my journal and also my simulations. I also wish to acknowledge my colleagues, Hana Faujan, Ru zana Yahya, Shahidah Shaari and Lim Wui Zhuan for sharing the stress, laughter and also giving advice during this project .
Next, I would like to thank the Graduate Research Fellowship fro m Universit i Putra Malaysia which is providing the funding to take this research.
Last but not least, thank you to my family specially my husband, Nor Irman who helped me financially and gave his mo ral support to me to fin ish this project, and my son, Iqbal who is born during the preparation of this thesis . A lot of love to my parents, Abdul Ghani and Halimah who is always support me with positive words. Thanks also to my siblings Syafirah, Syakir and Anum for listening to my annoying complaint and stressful day. You guys are mean a lot to me. I will remember all the guidance and knowledge that have given and take it as my guidance in further.
v
I certify that a Thesis Examination Co mmittee has met on 09 June 2016 to conduct the final examination of Nur Syafiqah binti Abdul Ghani on her thesis entitled “Molecular Simu lation Between Amy loid Beta (1-42), a Peptide Associated With Alzheimer Disease, and Zinc(Ii) Ion” in accordance with the Universities and University Colleges Act 1971 and the Constitution of the Universiti Putra Malaysia [P.U.(A) 106] 15 March 1998. The Co mmittee reco mmends that the student be awarded the Master of Science. Members of the Thesis Examination Co mmittee were as fo llo ws ;
Mohd B asyaruddi n Bin Abdul Rahman Professor Faculty of Science Universiti Putra Malaysia (Chairman) Haslina Binti Ahmad, PhD Senior Lecturer Faculty of Science Universiti Putra Malaysia (Internal Examiner)
Rozana Binti Othman, PhD Associate Professor Faculty of Medicine Universiti Malaya (External Examiner)
ZULKARNAIN ZAINAL, PhD Professor and Deputy Dean School of Graduate Studies Universiti Putra Malaysia Date: 26 July 2016
vi
This thesis was submitted to the Senate of Universiti Putra Malaysia and has been accepted as fulfillment of the requirement fo r the degree of Master of Science. The members of the Supervisory Co mmittee were as follows:
Roghayeh Abedi Karji ban, PhD Senior Lecturer Faculty of Science Universiti Putra Malaysia (Chairman)
Mahiran Basri, PhD Professor Centre of Foundation Studies for Agriculture Science Universiti Putra Malaysia (Member)
B UJANG KIM HUAT, PhD Professor and Dean School of Graduate Studies Universiti Putra Malaysia
Date:
vii
Declaration by graduate student
I hereby confirm that:
this thesis is my original wo rks; quotations, illustrations and citations have been duly referenced; this thesis has not been submitted previously or concurrently for any other degree at any other institutions; intellectual property fro m the thesis and copyright of thesis are fully -owned by Universiti Putra Malaysia, as according to the Universiti Putra Malay sia (Research) Rules 2012; written permission must be obtained from supervisor and the office of Deputy Vice -Chancellor (Research and Innovation) before thesis is published (in the form of written, printed or in electronic form) including books, journals, modules, proceedings, popular writings, seminar papers, manuscripts, posters, reports, lecture notes, learning modules or any other materials as state in the Universiti Putra Malaysia (Research) Rules 2012; there is no plagiaris m or data falsification/fabrication in the thesis, and scholarly integrity is upheld as according to the Universiti Putra Malaysia (Graduate Studies) Rules 2003 (Rev ision 2012-2013) and the Un iversiti Putra Malaysia (Research) Rules 2012. The thesis has undergone plagiaris m detectio n software.
Signature: ___________________________
Date: __________________
Name and Matric No.: Nur Syafiqah Binti Abdul Ghani (GS33669)
viii
Declaration by Members of Supervisory Committee
This is to confirm that: the research conducted and the writing of this thesis was under our supervision; supervision responsibilities as stated in the Universiti Putra Malaysia (Graduate Studies) Rules 2003 (Revision 2012-2013) are adhered to.
Signature: Name of Chairman of Supervisory Co mmittee:
Dr. Roghayeh Abedi Karjiban
Signature: Name of Member of Supervisory Co mmittee:
Prof. Dr. Mah iran Basri
ix
TABLE OF CONTENTS
Page i
ABSTRACT ABSTRAK
iii
ACKNOWLEDGEMENTS APPROVAL DECLARATION LIST OF TAB LES LIST OF FIGURES LIST OF ABB REVIATIONS LIST OF APPENDICES
v vi viii xii xiii xv i xv iii
CHAPTER 1
2
INTRODUCTION
1
1.1
4
Objectives
LITERATURE REVIEW
5
2.1
Prion Proteins
5
2.1.1 Alzheimer Disease (AD) Amylo id-β Peptide 2.2.1 Theoretical Studies Metal Interaction with A myloid-β Peptide Amylo id-β Peptide Aggregation Molecular Modelling
6 8 9 12 15 17
2.5.1
Molecular Docking Molecular Docking Studies of 2.5.1.1 Aβ-Peptide Molecular Dynamic (MD) Simu lation
17
MET HODOLOGY 3.1 Theoretical Backg round 3.1.1 Molecular Docking 3.1.1.1 Free energy Scoring Function 3.1.1.2 Genetic Algorithms (GAs) 3.1.1.3 Docking Programs 3.1.2 Molecular Dynamics (M D) 3.1.2.1 Force Field 3.1.2.2 Restraints Interaction Period ic Boundary Conditions 3.1.2.3 (PBC) 3.1.2.4 Energy Min imization
20 22 22 23 24 24 26 27 30
2.2 2.3 2.4 2.5
2.5.2 3
18 19
30 31
x
Temperature Coupling Material and Methods 3.1.2.5
3.2
3.2.1 3.2.2
4
Pressure
Materials Methods 3.2.2.1 The Model Structure 3.2.2.2 Molecular Docking Molecular Dynamics 3.2.2.3 Simu lation
4.3
32 34 34 35 35 37
(M D)
RES ULTS AND DISCUSS ION 4.1 Molecular Docking 4.1.1 Binding Site Analysis Ligand Interaction, Ionizability 4.1.2 Hydrophobicity Analysis 4.2 Simu lation of Aβ (1-42) with and without Zinc 4.2.1 Energetics of Model System 4.2.2 Root Mean Square Deviat ion (RMSD) 4.2.3 4.2.4 4.2.5 4.2.6
5
and
39 44 45 45
and
46 49 49 50
Radius of Gy ration (Rg ) Solvent Accessible Surface Area (SASA ) Root Means Square Fluctuations (RMSF) Secondary Structure Analysis
52 55 57 60
Amylo id-β Aggregation Process Simulation 4.3.1 Energetics of Aggregation Process 4.3.2 Root Means Square Deviations (RMSD) 4.3.3 Radius of Gy ration (Rg ) 4.3.4 Solvent Accessible Surface Area (SASA ) 4.3.5 Root Means Square Fluctuations (RMSF) 4.3.6 Secondary Structure Analysis
64 65 66 67 69 71 75
SUMMARY, CONCLUS ION AND RECOMMENDATIONS FOR FUTUR E RES EARCH 5.1 Summary 5.2 Reco mmendations for Future Research
REFERENCES APPENDICES BIODATA OF S TUDENT LIST OF PUB LICATIONS
80 80 81 82 96 135 136
xi
LIST OF TAB LES
Table
Page
1
A summary of docking results for single Amyloid-β peptide bound to zinc ion
45
2
The ligand interaction d istance between Aβ (1-42) and Zn2+ ion
46
3
The potential, kinetic and total energy of Aβ (1-42) in the last 10 ns of all simulat ions
49
4
A summary of So lvent Accessible Surface Area (SASA) average values for Aβ (1-42) in different conditions for 1 μs
56
5
Secondary structure element changes of Aβ (1-42) in different conditions during 1 μs MD simulat ion
61
6
The potential, kinetic and total energy of Aβ(1-42) averaged over the last 10 ns in different solvent
65
7
8
A summary SASA analysis for 6 mo lecules of Aβ (1-42) with zinc during aggregation in different conditions for the last 10 ns of both MD simulat ions Secondary structure element changes of Aβ (1-42) during aggregation involving in zinc ion different conditions during 100 ns MD simu lation
69
77
xii
LIST OF FIGURES
Figure
Page
1
Hu man prion diseases (Selkoe, 2003; Ch iti and Dobson, 2006)
5
2
The comparison between folded and misfolded amyloid (Hutton, 2001)
6
3
NFTs are co mposed of hyperphosphorylated Tau protein (Khatoon et al., 1992; Kopke et al., 1993)
7
4
The outline of M D simu lation steps for studying the interaction of Aβ (1-42) with zinc ion
21
5
The outline of MD simulat ion steps for aggregation study
22
6
Two-dimensional examp le of how a hexagonal box leads to lower volume than a square box with the same separation distance (Lindahl, 2008)
31
7
Amylo id-β(1-42) (PDB code:1IYT); α-helices (red), turns (green) and coils (white)
36
8
Hexfluro isopropanol (HFIP) structure
36
9
10
11
Methodology used for mo lecular docking
Summary of the steps for creating the topology of HFIP molecu le Pictures of simulation bo xes built for (a) single peptide only (b) peptide with zinc and (c) six mo lecules of peptide with zinc ions for aggregation study
38
39
40
12
Pictures of simulat ion box for (a) Aβ (1-42) in water (Aβ-H2 O) and (b) Aβ (1-42) in mixture of water (blue) and HFIP (red) (Aβ-HFIP-H2 O)
41
13
Picture of the simu lations boxes for (a) Aβ (1-42) with zinc (yellow) in water (Aβ-Zn-H2 O) and (b) Aβ (1-42) with zinc in solvent mixture (Aβ-Zn-HFIP-H2 O)
41
xiii
Picture of the simulation bo xes for peptide aggregation study; (a) water (6Aβ-6Zn -H2 O) and (b) solvent mixtu re (6Aβ-6Zn-HFIPH2 O)
42
The ligand interactions of Asp23 and Asn27 for mode 2
47
16
Ionic calcu lation for Aβ(1-42) docked to zinc ion (mode 2)
48
17
Hydrophobic surface calculation for Aβ (1-42) docked Zn 2+ ion
49
18
The RMSD fluctuation over time for Aβ (1-42) in four different conditions [Aβ-H2 O (black), Aβ-HFIP-H2 O (red), Aβ-Zn-H2 O (green) and Aβ-Zn-HFIP-H2 O (b lue)] for 1 μs
52
19
Rg fluctuations over 1 μs simulations for all systems including Aβ H2 O (b lack), Aβ-HFIP-H2 O (red), Aβ-Zn-H2 O (green) and Aβ-ZnHFIP-H2 O (blue)
53
20
The snapshots of conformational changes of amylo id-β(1-42) system in different conditions extracted fro m the Rg results over 1 μs simu lation time for the last 10 ns. (a) Aβ (1-42) in water (b) Aβ (1-42) in mixed solvent (c) Aβ (1-42) with zinc in water (d) Aβ (1-42) with zinc in mixed solvent
54
21
The accessible surface of 3 overlapping spheres is obtained by adding the radius of the solvent sphere to the van der Waals atomic radii. The volu me enclosed by the accessible surface is the excluded volu me with respect to the solvent sphere centre. The overlapping spheres are labeled as i, j, and k (adapted from Timothy, 1984)
55
22
Solvent Accessible Surface Area (SASA) fluctuation hydrophobic content as a function of time for 1 μs
56
23
The fluctuations of Aβ (1-42) residues in water (b lack) and mixed solvents (red) model systems over 1 μs
58
24
The RMSF of Aβ (1-42) residues in different solvents as a function of time over 1 μs
60
25
The visualization of secondary structures of (a) Aβ-H2 O, (b) AβHFIP-H2 O, (c) Aβ-Zn-H2 O and (d) Aβ-Zn-HFIP-H 2 O model systems at the end of 1 μs MD simu lation; α-helix (purple); 310 helix (red); extended β (yellow), bridge-β (p ink); coil (silver) and turn (green)
63
14
15
for
xiv
26
The time expansion of secondary structure elements for respective residues in different conditions; (a) Aβ-H2 O, (b) Aβ-HFIP-H2 O, (c) Aβ-Zn-H 2 O and (d) Aβ-Zn-HFIP-H2 O model systems (1 frame= 1 μs)
63-64
27
The RMSD fluctuations during aggregation over 100 ns simulation time for six molecu les of Aβ (1-42) with zinc ion in different conditions; 6Aβ-6Zn-H2 O (g reen) and 6Aβ-6Zn-HFIP-H2 O (purple)
67
28
The Rg fluctuations during aggregation over 100 ns simu lation time for six mo lecules of Aβ (1-42) with zinc ion in different conditions; 6Aβ-6Zn-H2 O (green) and 6Aβ-6Zn-HFIP-H2 O (purple)
68
29
The picture illustrates the changes of compactness of amy loid -β before and after aggregation in different solvents. The snapshots were taken at (a) 0 ns and (b) 100 ns for amylo id-β in water while (c) 0 ns and (d) 100 ns are amy loid-β in solvent mixture
68-69
30
The solvent accessible surface changes of Aβ (1-42) in different solvents over 100 ns simulat ion time
70
31
The RMSF fluctuations of aggregation of six molecules of Aβ (1-42) in water with zinc ions for 1 μs
71
32
The snapshot of each Aβ (1-42) peptide in water during aggregation for the last 10 ns of simu lation; α-helix (purple), β-sheet (yellow), β-bridge (orange), turn (cyan) and coil (silver)
72
33
The RMSF calcu lation results pf aggregation in mixed solvent for six mo lecules Aβ (1-42) with zinc ions for 100 ns
73
34
The snapshot of each peptide of Aβ (1-42) in solvent mixture during aggregation for the last 10 ns of simulat ion; α-helix (purple), βsheet (yellow), β-bridge (orange), turn (cyan) and coil (silver)
74
35
Snapshot pictures of Aβ (1-42) with zinc during aggregation in water in every 20 ns over 100 ns simu lation time
76
36
The changes of secondary structure for six molecu les of Aβ (1-42) during aggregation with zinc in d ifferent condition; in water (left), in solvent mixture (right) with the 1000 frame over 100 ns simu lation time
78
37
Snapshot pictures of Aβ (1-42) with zinc during aggregation in solvent mixtu re for every 20 ns over 100 ns simu lation time
79
xv
LIST OF ABB REVIATIONS
AD
Alzheimer’s Disease
APP
Amylo id Precursor Protein
APS
Aggregation Prone Structure
ATB
Automated Topology Builder
Aβ
Amylo id Beta
AβPP
Amylo id Beta Precursor Protein
CD
Circular Dichro ism
CHC
Hydrophobic Core
CNS
Central Nervous System
CPU
Central Processing Unit
CQ
Co mpound Dioquinol
CSHA
Canadian Study of Health and Aging
DMD
Discrete Molecular Dynamics
DOPS
Dio leoylphosphatidylserine
DPPC
Dipalmitoylphosphatidycholine
DSSP
Definition of Secondary Structure of Protein
ESR
Electron Spin Resonance
EURODEM
European Co mmunity Concerted Action on Epidemiology and Prevention of Dement ia
HFIP
Hexafluoroisopropanol / 1,1,1,3,3,3-hexafluoropropan-2-ol
IAPP
Islet Amy loid Po lypeptide
LMW
Low Molecular Weight
MD
Molecular Dynamics
xvi
NFTs
Neurofib rillary Tangles
NMR
Nuclear Magnetic Resonance
NOESY
Nuclear Overhauser Effect Spectroscopy
NPT
Nu mber of particle, Pressure, Temperature
NVT
Nu mber of particle, Vo lu me, Temperature
PBN
Phenyl-tert-butyl-nitrone
PBS
Phosphate-Buffered Saline
PCDs
Protein Conformational Disorders
PDB
Protein Data Ban k
POPG
Palmitoyloleoylphosphatidylglycerol
REM D
Replica Exchange Molecular Dynamics
RMSD
Root Means Square Deviation
RMSF
Root Means Square Fluctuation
ROS
Reactive Oxygen Species
SASA
Solvent Accessible Surface Area
SDS
Sodiu m Dodecylsulfate
SPC
Simp le Point Charge
TEM
Transmission Electron Microscopy
TFE
Trifluoroethanol
ThT
Thioflav in-T
VM D
Visual Molecu lar Dynamics
Zn
Zinc
xvii
LIST OF APPENDICES
Appendi x
Page
A
Structure of Aβ (1-42) with pdb code: 1IYT
B
Structure of Molecule HFIP
109
C
Parameter for HFIP
110
D
Script for Packmo l
113
E
Topology for Aβ in water
114
F
Topology for Aβ in Solvent Mixture
115
G
Topology for Aβ with Zinc in Water
117
H
Topology of Aβ with Zinc in So lvent Mixture
119
I
Topology for Aggregation part, 6Aβ with 6 Zinc in Water
121
J
Topology for Aggregation part, 6Aβ with 6 Zinc in Solvent Mixture
124
K
Topology for Zinc
127
L
Example script mdp file for Energy Min imization
128
M
Example script mdp file for Equilibrat ion 1 (NVT)
129
N
Example script mdp file for Equilibrat ion 2 (NPT)
131
O
Example script mdp file for MD production
133
96
xviii
CHAPTER 1
INTRODUCTION
The most regular basis of dementia in aging individuals all over the world is Alzheimer’s disease (AD). As a person gets older, AD beco mes a more crit ical and ever-growing public health problem. A D is known as a type of prion-related illness that shows up in various diseases, for examp le Creut zfeldt-Jakob, bovine spongiform encephalopathy and Mad Cow. Statistically, in 2010, 5.1% o f the US co mmunity was older than 65 years and about 454,000 A D cases were recognized, followed by a 10% boost up to 2,000 nu mbers each year (Abbott, 2011). There are no med icine, efficient approaches or powerful precautionary part for AD but younger generations may take some drugs to imp rove the cholinergic system, such as galantamine (Razadyne), donepezil (Aricept), tacrine (Cognex) and rivastigmine (Exelon) to avoid the risk of developing AD (Mancuso et al., 2011). The AD consequences on the mind are widely man ifested over the failure of cholinergic neurons.
AD causes around 66.67% of extensive crisis of dementia (Pasture and Onkia, 1994). There are two well-known groups which analyse the risk factors of AD; the European Co mmunity Concerted Action on Epidemiology and Prevention of Dementia (EURODEM ), and the Canadian Study of Health and Aging (CSHA) association. The EURODEM stated that smoking might raise the danger of AD (Launer et al., 1999), but there were no specific studies to observe the relationship between smoking and the onset of AD (Hebert et al., 1992, Wang et al., 1999; Do l et al., 2000).
EURODEM also reported that s ex and low educational level were highly associated with the AD cases. However, they were not considered as possible aspects that triggered AD in other researches (Cobb et al., 1995; Yoshitake et al., 1995). On the other hand, the CSHA performed an enormous study of dementia towards aging people by concentrating on its popularity (Posture and Onkia, 1994), incidence (McDo well et al., 1994) and risk factors (McDowell et al., 1994; Lindsay et al., 1997; Hébert et al., 2000). It was found that, either the amyloid-β precursor protein (AβPP) or some enzy mes for examp le, presenilin-1 fro m the metabolism process had contributed about 5% of the AD cases.
Both neural o xidative stress and neuroinflammatory events are crucial factors in the neurodegenerative landscape of AD. AD is regularly defined and distinguished by the existence of both Aβ-rich plaques (neuritic) and neurofibrillary tangles (NFTs) in the range of the entorhinal cortex, hippocampus and isocortex, that combined with synapse loss and clinical dementia (Duyckaerts et al., 2009). Aβ plaques have been effectively investigated for almost three decades and have been carefully analyzed. The 40-42 amino acids of Aβ peptides which are produced from a classic single-pass type 1 transmembrane protein called the amylo id precursor protein (APP) can form neurotoxic oligo mers, to be ultimately restored in the hightly hydrophobic extracellu lar deposits , also known as plaques.
1
These plaques tend to concentrate over time and eventually, they become a dense core or neuritic structures in the advanced stages of disease (D’Alton et al., 2011; Karran et al., 2011). The pathological modification of amy loid precursor protein towards the uncontrolled quantity of Aβ in numerous forms (monomers. oligo mers, protofibrils and fibrils) will lead to precipitation in the downstream processes, including the neuroinfla mmatory activation of microglia and neuritic pathology. Then, it will induce tangles’ formation and cell death. Aβ has also been linked to the initiation of other AD pathophenomena. In addition, some studies propose that the oligomeric Aβ is more toxic than the Aβ fibrils themselves (Glabe et al., 2005). The NFTs are usually found in the neocortical grey matter parenchyma. They are produced from proteins (MAPs) and many other components that have been identified using immunohistochemistry, immunoprecip itation and laser capture microdissection-mass spectrometry methods (Wang et al., 2005; Duyckaerts et al., 2009).
The tau protein functions as a stabilizer by attaching the mict rotubules (MTs) to increase its rigidity along the length of axons (Obulesu et al., 2011). Even though AD diagnosis generally needs a burden of plaques plus tangles, their mere presence do not always coincide with neuron loss or clinical dementia prior to death and autopsy (Green et al., 2000; Price et al., 2009). The hu man brain appears particularly vulnerable to oxidative stress . This necessitates the elaboration of complex antio xidant defenses in order to maintain oxidative balance. Vitamin A, C and E, glutathione and a several number of enzymes are the antioxidants that facilitate electron transfer to a nontoxic species such as catalase, superoxide dismutase and glutathione peroxidase . Each of them has been proven to decrease with age. Hence, the vulnerability of brain to oxidative stress originates from a nu mber o f various mechanisms.
In a healthy brain, the amount of metal ion is strictly standardized and the metal concentrations are kept at a very low level. The metal ions which are necessary for biological function and metal-binding of proteins (i.e. metalloproteins) constitute around one third of the proteome. The transition metals have a growing role of interactions on brain-related diseases because of their participation in biochemical reactions, forming free radicals. It is well known that oxidative stress is responsible for major cellu lar problems . The relationship between the AD disease and metals has been mostly studied by focusing on local accu mulat ions of plaques in brain areas at high risk for AD (Squitti et al., 2013). The hypothesis of Aβ-induced oxidative stress in AD patients (Markesbery et al., 1997; Butterfield et al., 2009) has been supported by Aβinduced elevation of o xidative stress marker in b rain and the subsequent neuronal degeneration (Frautschy et al., 1991).
The research interest on the metal ions’ position, specifically zinc, copper, alu min iu m, and iron in the neurobiological processes is growing rapid ly. There are increasing evidences which demonstrate the interactions of zinc (II) and copper (II) ions with Aβ peptides and their effects towards fibrilization and toxicity. A lot of Zn 2+ and Cu 2+ ions are present in the synaptic area of the brain. It is possible that the age-related dyshomeostasis of these biometals are associated with the AD pathology. The assembly of Aβ into tinctoral aggregates as induced by Zn2+ ion was first reported by Bush and colleagues in 1994 where the aggregation was caused by sub-stoichiometric concentrations of Zn 2+ (Bush et al., 1994).
2
Tougu et al. (2011) also stated that Aβ40 can aggregate in the presence of Zn2+ ions in a millisecond period (Noy et al., 2008). Previous studies reported two possible ways. First, the metal ions will attach to amyloid mono mers and accumulate in the brain to form oligo mers through the aggregation of metallated monomers. Next, they will connect the pre-formed apo-oligo mers. Due to the various arrangements and distributions of the monomer and oligo mer ensembles, these two pathways may eventually show different binding abilities of the metal ions that are connected to oligo mers (M iller et al., 2012). The Alzheimer’s cases are increasing rap idly in nu mber world wide, yet there is no exact medicine for AD to date. Several evidences have shown that the interaction of Aβ with transition metal ions can lead to the aggregation and toxicity (Bo lognin et al., 2011). The mechanis m of its action has been explained by a few experimental works but the data obtained are still limited (Nilsson, 2004; Takano, 2008; Vivekanandan et al., 2011).
The experimental results could not identify the necessary approach to figure out the direct protein-metal interactions at the single-molecular level. The problem faced in most experimental approaches is the direct observation of protein-protein and proteinmetal interactions which require the proteins to be soluble and analyzed using several forqualitatic methods such as electrophoresis, mass spectrometry, and chromatography. Without a doubt, this will pose considerable challenges for those who adopt insoluble amy loid configuration in the direct analysis. In addition, X-ray crystallography and other related procedures depend upon protein crystallizat ion in the first place which is not effective in describing heterogeneously sized oligo mers, poly mers and amorphous aggregates or insoluble amy loid proteins (Pedersen & Heegard, 2013). No wadays, the amount of molecular biological data is increasing rapidly, thus the computer-based analysis of mo lecular interactions has become more and more practical.
Molecular modeling involves all theoretical methods and computational skills to model, predict or even mimic the routine of molecules. In this study, they were applied on Aβ peptide by treating the protein as the monomer to produce mechanistic and structural informat ion of its aggregation processes, because Aβ peptide can aggregate very fast in water. However, using all-ato m force field for full length MD simulations of Aβ 40 and Aβ42 in aqueous solution is very challenging. Therefore, limited findings fro m M D have been reported. For examp le, Santini et al. (2004) and Rod ziewiczMotowidlo et al. (2008) simu lated the hydrophobic core of Aβ 16-22 only, using imp licit solvent model (Santini et al., 2004; Rodziewicz-Motowidło et al. 2008), while others used the coarse-grained MD approaches.
The combination of coarse-grained atomic representations and the enhanced computational power, has allowed us to perform the simulations of biological co mplex systems within microsecond or millisecond time frame (To zzin i, 2005). Moreover, the timescales accessible to simu lation coincide with the particular that is reachable using high advanced spectroscopic techniques . Therefore, it is possible to directly compare MD observations with the experimental results, for examp le, the co mplex aggregation of soluble proteins into fibrillar species (Wu and Shea, 2011). The coupling of powerful co mputers and molecular modeling approaches such as molecular docking and MD makes computational chemistry an exciting area for groundbreaking researches with lots of capabilit ies (Karplus and Kuriyan, 2005).
3
1.1 Objecti ves The main object ives of this study were to identify the amy loid-β(1-42) peptide interaction with zinc ion in water and its mixture with hexafluoroisopropanol (HFIP) fo llowed by exploring the aggregation process of this peptide in the presence of metal by using MD simu lation technique. Therefore, these specific objectives were selected:
To simulate the interaction of Aβ (1-42) peptide with zinc ion in different solvent conditions.
To model the aggregation process of Aβ (1-42) peptide in the presence of zinc ion in different solvent conditions.
To determine the dynamics, flexib ility and structural changes of both model systems after interacting with metal.
4
CHAPTER 2
LITERATURE REVIEW
2.1 Prion Proteins Prions are infectious diseases-causing proteins with a broad spectrum of clinical man ifestations including dementia, ataxia, insomnia, paraplegia, paresthesias and deviant behavior (Prusiner, 2001). Prion disease is a prototypical protein conformat ion disease caused by point mutations in the prion gene that leads to the alterations of prion protein function. The mechanis m of p rion disease has been illu minated by the discovery of prion-like protein conformat ional change in yeast (Ross and Poirier, 2004). They cause a growing problem in aging populations with a great impact on health services and society (Lindsay et al., 2002). The different types of human diseases related to prion disease are Creutzfeldt-Jakob disease (CJD), GerstmannStraüssler-Scheinker syndrome (GSS), Fatal familial insomnia (FFI), Kuru, and Alzheimer Disease (AD) (Selkoe, 2003, Chiti and Dobson, 2006).
Figure 1. Human pri on diseases (Selkoe, 2003; Chiti and Dobson, 2006)
Neurodegenerative disease is often related to unfolded or misfolded structure (Uversky, 2003). Fo lding of secretory proteins provides a number of unique challenges. Fold ing is often accompanied by the format ion of native disulfide bonds and insertion into the lip id bilayers (Bukau et al., 2006). The process of folding often begins co translationally in which the N-terminus begins to fold while the C-terminal portion of protein is still being synthesized by ribosome. The chaperone assisted folding is often necessary in the intracellular environment to prevent misfolding and aggregation that may occur as a consequence of exposure to heat or other changes in the cellular environment (Lee and Tsai, 2005).
5
The biological function of a protein is determined by its amino acid sequence. In the last few years, diverse diseases have been found to originate from protein misfold ing which are now grouped together under the name of protein conformational disorders (PCDs) (Tho mas et al., 1995; Kelly, 1996). When the protein is not in its native structure, it will result in too much misfolded protein existence in the cell shown in Figure 2. Fu rthermore, when it fails to fold into native structure, inactive proteins will be produced that are usually to xic. Like familial prion diseases, familial AD has an autosomal do minant pattern of inheritance that can be resulted from a mutation in the gene for APP, presilin 1 or presilin 2 (St George-Hyslop and Peter, 2000).
Figure 2. The comparison between fol ded and misfol ded structure (source from: Hutton, 2001)
2.1.1 Alzheimer Disease AD virtually is the only brain disorder that can be defined by the accumulation of amy loid-forming proteins both ext racellularly amy loid-beta (Aβ) and intracellularly (tau) proteins (Selkoe and Podlisny, 2002). Tau proteins are not involved in AD progression but they cause the less common but equally devastating disorder in which tau-containing neurofibrillary tangles (NFTs) accu mulate in the absence of extracellular amyloid (Hutton, 2001). However, tau protein can still affect AD due to the fact that its post-translational modifications may be toxic (A madoro et al., 2006) and the suppression of tau protein blocks Aβ-induced apoptosis (Rapoport et al., 2002) and reduces memory deficit (Santacruz et al., 2005). Moreover, due to the presence of NFTs in AD brains (Figure 3), tau protein level is 8-fold higher than that of controlled brains (Khatoon et al., 1992; Kopke et al., 1993).
6
Figure 3. NFTs are composed of hyperphos phoryl ated Tau protein (Khatoon et al., 1992; Kopke et al., 1993)
The pathological hallmarks of AD are extracellular senile plaques where the major component is the Aβ peptide. All A D patients have a lot of amy loid p laques with the ability to degenerate nerve endings. Besides that, several neurodegenerative and other diseases are result from the accumulation of amylo id fibrils formed by misfolded proteins (Selkoe, 2003). Aβ is a frag ment formed by APP during proteolysis by enzy mes but the mechanism is still unknown. So me of these fragments can help the fibrils of Aβ clamp and deposit outside neurons in dense format ion known as senile plaques (Ohnishi and Takano, 2004). The cleavage of amylo id beta precursor protein (AβPP) generate s a 40-42 amino acid peptide called Aβ either within the lysomal or endoplasmic reticulu m o r Go lgi compart ments of the cell (Coughlan and Breen, 2000). There are two types of Aβ, one is Aβ40 (DAEFRHDSGYEVHHQ16 KLVFFA 22 EDVGSNKGA IIGLM VGGVV) which is more co mmon co mpared to Aβ 42 that has more fib rillogenic structure (DA EFRHDSGYEVHHQ 16 KLVFFA 22 EDVGSNKGA IIGLM VGGVVIA ). There are some unrecognized mechanis ms that contribute to the preferential deposition of Aβ 42 in AD brains with a higher tendency towards aggregation and toxicity (Gu and Guo, 2013). Mutations in APP which is associated with the early-onset of Alzheimer have been noted to increase the relative production of Aβ42. This may open a great window for Alzheimer therapy through modulating the activity of β and γ secretases to produce mainly Aβ 40 instead of Aβ42 (Yin et al., 2007; Panza et al., 2011; Sathya et al., 2012; Tan and Evin, 2012).
7
2.2 Amyl oi d-β Pepti de The structure of Aβ 40 in cationic, anionic perfluorinated and nonionic surfactants has been revealed and the effect of peptide in their critical micelle concentration has been estimated (Rocha et al. 2012). The charged or nonionic head group and fluorinated chains were used to characterize the forces that might be involved in the stabilization of Aβ 40 secondary structures by amphiphiles. The molecules that could induce a nonaggregated stable state were also screened. Several experimental studies have shown the detection of different Aβ oligomers in the frontal cortex of AD patients. From these experiments, the Aβ oligomers were extracted fro m the autopsy of frontal cortex b rain tissues obtained from A D patients and age-matched control. Their results indicated the presence of different confirmat ions of Aβ oligomers in the soluble, membrane associated and insoluble fraction which could be a potential neuroto xic that affected Aβ plaque deposit ion (Bao et al., 2012). So me studies suggested that amyloid fibrils can not always induce Thioflavin-T (ThT) fluorescence. For examp le, some fragments of TAR DNA -binding protein from amylo id fibrils were investigated, but the binding of these fibrils to ThT did not cause any fluorescence emission (Chen et al., 2010).
Based on the revised criteria for the identificat ion of in vitro amyloid fibrils by Nilsson (2004), it can be concluded that β-amylase forms amylo id fibrils (Luo et al., 2012). A parallel β-sheet structure has been reported for Aβ fibrils by using the solid state NMR. The atomic level structure of Aβ in aqueous environment, however, has not been determined because of its tendency to aggregate. There are several reports that produced soluble forms of Aβ to possess intrinsic neurotoxicity. Recently, the determination of the crystal structure of Aβ frag ments in an aqueous solution without organic solvents and detergents becomes possible using a fusion technique which can also be used for amyloidogenic peptides in aqueous environment (Takano, 2008). The conformational analysis of Aβ fragments in a membrane-mimicking (sodium dodecylsulfate, SDS) environ ment has been done. CD studies were carried out at both micellar and non-micellar concentrations. Additionally, the MD calculations were performed in an exp licit water/SDS environ ment with the spectroscopic studies on frag ment ‘11-28’ of the Aβ and its E22G variant (Rodziewicz-Motowidlo et al., 2008). Several other studies also showed that the fragment ‘11-28’ might be a good model for conformat ional and aggregation studies (Rabana et al., 2002; Juszcyk et al., 2005). It contains the hydrophobic core (CHC) wh ich is responsible of the aggregation process (Hilb ich et al., 1991; Tjernberg et al., 1999) and also the residues which are critical to α-secretase cleavage of APP. In summary, the frag ment was able to form aggregated fibril similar to those found in natural protein (Tjernberg et al., 1999) and displayed cellu lar to xicity in vitro and in vivo (Flood et al., 1994; Rabana et al., 2002).
8
Overall, only the native peptide that was placed in the micellar SDS surface needed to promote the α-helical structure, whereas the Artic variant was found to be in the 3 10 helical conformation. It seems that the specific E22G mutation in the Artic variant entails the weaker interactions of the peptide with the SDS micelle than those characteristic of the native peptide fragment or the Italian or Flemish variants. The glycine residue was also introduced during the mutation of Artic variant which can be responsible for the shortening of the helix in peptide (Rodziewicz-Motowidlo et al., 2008).
2.2.1 Theoretical Studies The interaction of Aβ 1-40 peptide (PDB code: 2LM N; Tycko Model) during Aβ oligo merization was carried out by Rao et al. (2013) with the ability of galantamine to prevent Aβ-oligomerization in consideration (Scott & Orv ig, 2009). The in itial step in Aβ-oligo merization process is the format ion of dimers (Tarus et al., 2005). They used mo lecular docking to determine the potential binding site of galantamine followed by MD simu lation to check the stability of galantamine-Aβ 1-40 peptide dimer binding. They suggested that galantamine can bind at the central region composing of Lys16Ala21 and the C-terminal region consisting of Ile31-Val36. These interactions were primarily van der Waal’s contacts along with another two polar interactions which led to the destabilization of antiparallel β-strand interactions and prevented the format ion of to xic soluble oligomer format ion and assembly where the galantamine binding led to a major structural shift of the dimer consisted of Asp23-Gly29 (Rao et al., 2013). The β-hairpin fo lding with the presence of sodium ion has also been studied. It was found that, the Na+ ions could interact with the N atoms of β -hairp in fro m the backbone while the side chain would be pushed away. This might happen because the backbone N ato m of β -hairp in was all deeply buried and the two side chain N ato ms were in protonated states (Lys50) and in the 1H-indole ring (Trp43), respectively. The conformat ional analysis was further done and the solvent was checked to explore whether they could affect the folding mechanism. They said that there was a strong interaction between the metal ions with the surrounding water molecules but however, it produced a different result with the apo-peptide (Wu et al., 2012). The molecular interaction between Aβ molecular protofilaments and lipid bilayers memb ranes was investigated by Florentina and Nicole (2012) in the presence of explicit water mo lecules by using computational tools and all-atom molecu lar dynamics. Their simulation revealed the relative contributions of different structural elements to the dynamics and stability of Aβ protofilaments segment near membran e and the first step in fibril-memb rane interactions mechanis m. The local symmetryrelated and directional properties of the protofilaments segment models were able to cover a broad range of relative protofilaments -membrane orientations corresponded to systems denoted by S1-S4. The early steps in the interaction mechanism were then detected between Aβ protofilaments and lipid memb ranes. It was then found that the electrostatic interactions between several charged peptide side chain and lipid head groups were highly involved in this process (Tofoleanu & Buchete, 2012).
9
De To ma and co-workers (2012) described two amylo idogenic proteins, Aβ peptides and islet amy loid polypeptide (IAPP) with misfolding propensities which could influence the AD and type II diabetes. Protein misfolding diseases share several similarities fro m the init iation and/or the development of their associated conditions. It was stated that, Aβ and IAPP were representing amyloidoses which highlighted the primary considerations for studying misfolded proteins associated with human diseases. From their results, the mechanism of the diseases generation and aggregation was partially elucidated. However, following studies provided enough evidence that the local environment conditions and interactions with metal ions and other existing proteins can directly affecting the processes.
The unique properties of amyloid proteins pose a challenge for traditional drug design. The development of novel therapeutic strategies is currently limited because of the lack of the unresolved understanding of the etiology of amylo id disease. Hence, the first step is to extend the current understanding of the nature of these diseases and the factors that trigger them. As a result, the full understanding of the patho-physiological relationship between the format ion of misfolded protein, their interactions with other biological co mponents and the onset and progression of these diseases will be vital (De Toma et al., 2012). A various molecular dynamics (MD) simulat ions were performed by Lee and Ham (2010) with SA NDER module of AM BER9 program package (Case et al., 2006) using the ff99 force field on Aβ42 protein in explicit water to characterize the aggregation of prone structure (APS) of Aβ 42 monomer and the early sequential conformat ional transitions for Aβ 42 misfolding in a lag phase. Helical NMR structure was used as initial conformation fo r all simulat ions. The nature of the APS in Aβ 42 misfolding was determined based on the observed correlation between the nonlocal backbone H-bonding and the hydrophobic surface area (Lee & Ham, 2010). They found that the hypothetical aggregation mechanism fro m the APS of Aβ 42 to fibril accounted three mandatory steps in which the strong off registry side chain interactions showed a capacity to impede fibril gro wth. The elimination of the off registry side chain interaction formed by ASP23 in the wild type Aβ sequence could trigger the formation of fibril. The interactions between Aβ peptide and lipid bilayers can promote a peptide distribution on the bilayers surface that is prone to peptide-peptide interactions, which may in fluence the propensity of Aβ to aggregate into higher-order structures (Davis and Berkowit z, 2009).
They used unconstrained and umbrella sampling molecular dynamics simu lation to investigate the interactions between the 42-amino acid of Aβ peptide and model bilayers of zwitterionic dipalmitoylphosphatidycholine (DPPC) and anionic dioleoylphosphatidylserine (DOPS) lip ids. Lip ids with anionic head groups were also used as a model system. Fro m the results, the anionic lipids lowered the pH of solution near the bilayers which would alter the protonation state of proteins near these bilayers (Krishtalik & Cramer, 1995; Terzi et al., 1995; van Klo mpenburg & de Kruijff, 1998). Calculations by Terzi et al. (1995) were consistent with the experimental data in which Aβ was bound to anionic palmitoyloleoylphosphatidylglycerol (POPG) lipids. They showed that the protonation of three histidine residues occurred upon binding . This result is also supported by the use of mu ltiple pH states to study Aβ binding.
10
Thus, they suggested that Aβ might attracted to the surface of DPPC and DOPS bilayers over the small time scales used in these simu lations (Davis & Berkowit z, 2009).The Aβ10-35 peptide folding was probed using replica-exchange molecu lar dynamics (REM D) simulat ions in explicit water (Bau mketner and Shea, 2007). The Aβ 10-35 peptide is a frag ment of the full-length Aβ 40/42 peptide that has many of the amy loidogenic properties of its full-length counterpart. Under physiological temperature and pressure, their simu lations revealed that the Aβ 1035 peptide does not possess a single unique folded state. This peptide exists as a mixture of collapsed globular states that remains in rapid dynamic equilib riu m with each other. This conformat ional ensemble is do minated by random co il and bends structures with the insignificant presence of α-helical or β-sheet structure. Their replica -exchange simu lations also showed that the Aβ10-35 peptide under physiological conditions did not fold to a unique native state but rather exists as an ensemble of inter-converting conformat ions (Baumketner & Shea, 2007). These results which provided the general picture of structural organization was in a good qualitative agreemen t with the earlier experimental results (Zhang et al., 2000) together with the other Aβ peptides’ strudies (Riek et al., 2001; Hou et al., 2004). Fro m NM R studies done by Lee and co-workers (1995), the Aβ 10-35 peptide showed a structured core consisting of the hydrophobic patch L17 VFFA 21 , with the carboxyl and amino terminal parts of the peptide forming a mobile halo around the core structure.
Wei and Shea (2006) carried out another set of MD simulations using the GROMACS software package (Berendsen et al., 1995; Lindahl et al., 2001) and GROM OS 96 force field (van Gunsteren et al., 1996). They demonstrated that Aβ 25-35 adopted mostly the helical structure in HFIP/water co-solvent while in pure water, the peptide wanted to be in mostly collapsed-coil structure as well as a lesser extent of β-hairpins. Thus, different solvents can affect the structures in different ways. Fluorinated alcohols such as hexafluoroisopropanol (HFIP) and trifluoroethanol (TFE) are co mmonly used to stabilize helical structure, however in some instances, these solvents can act as protein denaturants as well (Brooks & Nilsson, 1993).
The use of replica exchange mo lecular dynamic (REM D) simu lations enabled them to obtain a near-comp lete determination of the free energy of the folding of Aβ25-35 peptide for the first time (Wei & Shea, 2006). Daidone (2004) conducted another 50 ns stimulat ion in TFE for the H1 prion peptide and the fragment 12-28 of the Aβ peptide with helical starting structures. Overall, the helical fo rms of these peptides were very stable and retained during the time of simulat ion. The structural properties of Aβ 42 peptide through a combination of ion-mobility mass spectrometry and theoretical modelling were further detected by Baumketner et al. (2006). The modelling was done by using the CHARMM all-ato m force field (MacKerell et al., 1998). They used replica exchange molecular dynamics simulat ions propagated with the time step of 2 fs. Their simulations illustrated that the Aβ42 in aqueous solution could adopt both extended chain as well as collapsed-coil structures.
11
The Aβ 42 peptide did not display a unique fold but it rather comprised a mixture of rapidly inter-converting conformat ions. The results of their simulat ions were consistent with the solution NMR experiment based on NOE constraint collected fro m Nuclear Overhauser Effect Spectroscopy (NOESY) spectra and recorded with different heteronuclear filters, chemical shift indices and J-scalar coupling. It was also consistent with CD spectroscopy results. The experimental results were confirmed by finding the percentage of ordered residues in the soluble Aβ 42 from their simu lations.Aβ peptide can form to xic oligo mers having an important contribution to the onset of Alzheimer disease (Urbanc et al. (2004). The to xicity of Aβ oligomers depends on their structure, which is governed by assembly dynamics.
However, due to the limitations of experimental techniques, a detailed knowledge of oligo mer structure at atomic level can be missed. They used discrete mole cular dynamics simulation (DMD) in wh ich the particles collide elastically and their kinetic energy before and after the collision are conserved. DMD is an event-driven process which requires keeping track of particle positions and velocities only at collision times that have to be sorted and updated. This property makes it different fro m trad itional continuous MD. Using DMD, they predicted the existence of different dimer conformat ions of Aβ oligomers. The all-ato m exp licit solvent (ES) molecu les produce a set of microstates of the macroscopic conformat ion while imp licit solvent (IS) surface energy dielectric continuum model is used to calculate the average salvation free energy as the sum of the free energies by creating the solute size hydrophobic cavity of the van der Waals solute-solvent interaction and the polarization of water solvent by the solute’s charges.
Therefore, a co mbined ES/IS method was applied to study the stability of dimer conformat ions with realistic force fields to measure the free energy of predicted dimer conformat ions (Vo robjev and Hermans, 1999). The co mbined ES/IS method has a particular source of error in the conformational free-energy estimation that comes from applying the implicit solvation method to figure the solvation free energy, which is overestimation. Against these limitations, the combined ES/IS method was successful in proposing some ideas for solving the folding problem. The free energy of misfolded proteins was later shown to be greater than the free energy of naturally folded proteins (Vorobjev and Hermans, 2001). As reported, the planar β -strand Aβ dimers estimated by the coarse-grained MD model could not provide the experimentally observed inequality in Aβ o ligo mer structure between Aβ1–40 and Aβ1–42 (Bitan et al., 2003). 2.3 Metal Interaction wi th Amyloi d-β Pepti de Metals are generally assigned in nature and biological scheme. They can commonly be classified as either toxicologically or biochemically functional (Duce & Bush, 2010). Nowadays, researchers are more focused on transition metals because it is familiar and can participate in biochemical reactions that produced free radicals. As reported experimentally, the concentrations of metal ions in a healthy brain should be at low level.
12
There are several types of protein-bound metal cations like iron (Fe3+/Fe2+), zinc (Zn 2+), copper (Cu 2+/Cu +), calciu m (Ca 2+), magnesium (Mg 2+) or manganese (Mn 2+/ Mn 3+) (Ko zlo wski et al., 2009) and the imbalance of these transition metal cations are assumed to have a contribution into Aβ deposition (Bush, 2003). Copper, zinc and iron are found to be highly concentrated in Aβ plaques inside the brain of AD patients. Some studies recommended that metal binding to Aβ can promote Aβ aggregation and generate reactive oxygen species (ROS) (Pithadia & Lim, 2012). A broad aspect proposing the relationship between AD and systematic changes of metal metabolis m upon genetic variability has also emerged (Squitti et al., 2013). Metal ions are also involved with the ubiquitous protein systems and play an important role during their folding process (Wu et al., 2012). The outcomes of metal ions on the aggregation of human Aβ have been reviewed in vitro (Miura et al., 2000). The links between AD and metals have been mostly investigated with a focus on local accumulations in brain areas critical to AD.Several researchers stated that Zn2+ ions have been shown to trigger Aβ aggregation (Mantyh et al., 1993, Bush et al., 1994, Esler et al., 1996) and an abnormal high level of zinc has been found within amy loid deposits (Lovell et al., 1998, Suh et al., 2000). However, the role of zinc in amyloid deposition is subject ed to debate (Cuajungco & Fagét, 2003; Cuajungo et al., 2005) as Zn 2+ ions induced the deposition of Aβ in the form of nonfibrillar aggregates (Yoshiike et al., 2001; Parbhu et al., 2002).
In addition, zinc has also been shown to function as an antioxidant with the ability to protect brain from extensive redo x chemical wh ich is a reaction that contributes to AD related oxidative stress (Curtain et al., 2001; Valko et al., 2005). The electron spin resonance (ESR) data obtained fro m Brian and co-workers (2002) experiment showed that both Aβ and α-synuclein could generate hydroxyl radicals after pre-incubation in phosphate-buffered saline (PBS) solution followed by the additio n of small amounts of Fe(II). Th is was detected by the metal-dependent formation of hydrogen peroxide by using Fenton’s reaction.
Their results represented that this procedure might be the most obvious chemical route which could be consistent with the inhibitory effects of metals chelators and catalase (Tabner et al., 2002). Ho wever, they could not confirm the existence of peptidyl free radicals or any other free radicals. The very weak four line ESR spectrum gained fro m Aβ and PBN could be due to the oxidation of the PBN by the hydrogen peroxide produced by Aβ or the trapping of hydroxyl rad icals by the PBN which was immed iately degraded to tert-butyl-hydroaminixy l as reported by Bush et al. (1999). Therefore, the degeneration and the loss of nerve cells in brain could be resulted fro m the direct production of hydrogen peroxide during extracellu lar or intracellu lar protein aggregation that would induce oxidative damage, part icularly in the presence of metals where the hydrogen peroxide could be converted via Fenton’s reaction into the highly reactive hydroxy l rad ical.
13
Metals have been observed at the sites of brain lesions in AD d isease (Bush et al., 1999; Campbell et al., 2001; Tabner et al., 2001) and there is a number of evidences for the involvement of reactive o xygen species in the pathogenis of AD and other neurodegenerative diseases (Varadarajan et al., 2000; Adams et al., 2001). Viles (2012) showed how the metal ions could influence the amy loid fo rmation in a variety of neurodegenerative diseases. The various mechanisms by which metal ions might affect the kinetics of amyloid fiber formation were reported. The copper ion, Cu 2+ that bound to Aβ in either monomeric or mature Aβ fibers had identical coordination geometries (Karr & Szalai, 2008; Sarell et al., 2009) and affin ities (Sarell et al., 2009). Solid state NMR of the Cu 2+ complex suggested that the fibrillar structure was not disrupted by Cu 2+. Several studies stated that the Zn2+ and Cu 2+ ions could cause marked aggregation on A peptide aggregation. However, these initial studies did not make the distinction between amorphous aggregates or their non-toxicity to cell and form amy loid fibers. Further investigations conducted using the fiber specific fluorophore thioflav in-T (ThT) suggested that Cu 2+ and Zn2+ promoted only amorphous aggregation of Aβ and actually inhibited fiber formation using primary cell culture and immo rtal cell lines. Their findings indicated that the Cu 2+ induced amorphous aggregates were non-toxic to cells (Viles, 2012).
Other researchers like Crouch and Barnham (2012) planned an alternate strategy to target the interaction between Aβ and metal ions by using the compound that had the potential to redistribute metal ions within the brains. They used the prototype metalchelating compound dioquinol (CQ). The function of this compound is to prevent the Aβ toxicity in vitro, out-competing Aβ for metal ions without affecting the activity of metal-dependent enzymes and also attenuating the rate of cognitive decline in AD in a small phase II clin ical trial. The study was not about the interaction between CQ and Aβ but it prevented the extracellular Aβ-metal interactions by using the ability of the CQ to redistribute Cu 2+ into cells. They also mentioned the formation of sypnaptotoxic Aβ oligomers wh ich might be driven by the Zn release into the synapse, a process that could be prevented by using CQ. Therefore, the use of compounds with the ability to remove metal ions from Aβ and redistribute them within the brain to regions or cellular locations where they can promote synaptic function offers the opportunity to find a more integrated approach to treat AD. Targeting multip le aspects of AD opens the greater therapeutic potential than methods designed to treat just the individual components of this complex d isease (Crouch and Barnham, 2012). In another study by Liu et al. (2011), it was suggested that Fe3+ might influence the morphology of Aβ fibers with shorter curved fibers which could cause toxicity in the fly while calciu m ion, Ca 2+ was found at a very high concentration extracellular typically at 2 mM in the extracellular space. Ca 2+ could also accelerate Aβ fiber formation (Isaacs et al., 2006). Ca2+ is a hard metal ion and coordinates via oxygen ligands in vitro. Although Ca 2+ is an abundant metal ion, it is still not clear that whether the AD pathology will be strongly affected by fluctuations in Ca2+ levels outside the cell.
14
However, the high extracellu lar levels of Ca 2+ may influence the critical concentration for fibril format ion and explain why sub-nanomolar levels of extracellular Aβ can be sufficient to cause fibers to form in vivo (Viles, 2012). A wide range of potential oligo mers of Aβ 42 complexes with Zn 2+ was demonstrated by the covalently linking of the experimental based coordinate sets of Aβ fragments from the N-termini and Ctermin i using molecular dynamics (MD) simulat ions (Miller et al, 2010). Replica exchange molecu lar dynamics (REMD) simu lations were used to further explore the relative stabilities and populations of the ordered aggregates compared to the vast number of conformational ensembles. So far, only several computational studies had investigated the metal ion-Aβ peptide interaction. Their simulat ions produced a stable model in wh ich the Zn 2+ ion was coordinated with the residues of two different peptides; 8Zn 2+-8Aβ 42 and 4Zn 2+-8Aβ 42 (Miller et al., 2010). 2.4 Amyl oi d-β Pepti de Aggregation A general view of the aggregation of soluble mono mers wh ich changes into insoluble fibrils involves; (i) nucleation dominated by a short lag phase, (ii) formation of low mo lecular weight (LMW) oligomers (dimmers to dodecamers), (iii) assembly of LMW oligo mers into layers of β-sheet-rich oligo mers and (iv) the growth of the β-sheet-rich oligo mers into higher order aggregates of heterogeneous size and morphology. Ult imately, the higher order of o ligo meric intermediates disappears with the format ion of fibrils (Hardy and Selkoe, 2002; Walsh and Selkoe, 2004, 2007).
The matured amylo id fibrils constitute plaques that are not thought to be the prime cytotoxic agent (Liu et al., 2011). There are strong evidences from both in vitro and in vivo studies that stated the Aβ derived neurotoxins are soluble and fibril-free (Lambert et al., 1998). Experimental studies have shown that the secondary structure alterations and the aggregation of Aβ peptides are related to AD disease (Hardy and Selkoe, 2002). The peptide is aggregated to amyloid fibrils within the neurotic plaques and vascular deposits that characterize this disease (Kirschner et al., 1986).
The causes and the effects of linking protein aggregation to the degenerative central nervous system (CNS) d iseases are largely unknown but the aggregation of proteins are diagnostically specific (Ross and Poirer, 2004) and its association with the disease has obvious diagnostic and therapeutic implications (Pedersen and Heegaard, 2013). It was proposed that by this an effective strategy to look for the prevention and/or treatment of this disease by inhibiting Aβ42 aggregation (McKoy et al., 2012). The Aβ 42 aggregation is not fully studied and the mechanism of aggregation is still under intense research (Alptuzun et al., 2010). There are growing evidences that the aggregation of Aβ into a variety of supramolecular structures can be critically involved in its insurgence and progression or several numbers of mo lecular aspects of the ‘amylo id cascade hypothesis’, for examp le, the structural characterisation of the amy loid peptide either in mono meric or in aggregate form (Valensin et al., 2012).
15
The results pointed out that the oligomeric Aβ aggregates could be causing the neurotoxicity. Metal-based compounds which are capable of interacting with the Nterminal metal b inding site of amylo id peptides could similarly contrast metal-induced Aβ aggregation and serve as a platform to find potential drugs for AD. Unfo rtunately, there are some crucial aspects needed to be elucidated first before proceeding with an effective pharmaceutical develop ment of metal-based compounds as anti-AD agent such as making sure the metal compounds are able to cross the blood brain barrier. The changes that happen on metal compound before reaching the target organ must also be clarified. The favourable concentration of the metal co mpound should be considered too in order to avoid the negative effect towards brains. A variety of nano scale and toxic aggregate species can produce Aβ forms ranging fro m s mall oligo mers to fibrils. They can build a strong interaction with lipid memb ranes which may represent an important step in several toxic mechanisms . Yates et al. (2013) investigated a variety of Aβ fragments in an effort to understand how specific regions of Aβ could regulate its interaction with lipid membranes. The type of Aβ fragments involved Aβ1-11 , Aβ 1-28 , Aβ10-26 , Aβ12-24 , Aβ16-22 , Aβ22-35 and Aβ1-40 in the presence of the supported model of total brain lipid ext ract (TBLE) bilayers. By using scanning probe techniques, they elucidated aggregate morphologies for these different Aβ frag ments in free solution and in the presence of TBLE b ilayers. These frag ments formed different form of oligo meric and fibrillar aggregates under free solution conditions.
Exposure to TBLE bilayers resulted in distinct aggregate morphologies compa red to free solution and changed the bilayers’ stability based on Aβ sequence by producing the altered mechanical properties of bilayers. Aβ 1-11 , Aβ1-28 and Aβ12-24 produced the minimal interaction with lipid memb ranes and formed only sparse oligomers (Yates et al., 2013). There are several techniques to study amylo id fibril formation in vitro (Nilsson, 2004). A protein aggregate can be classified as an amylo id fibril if some but not all of the criteria are satisfied. For instance, β-sheet rich structures in the β-amylase aggregates were characterized by circular d ichrois m (CD) spectros copy. The results of the Congo red binding assay, Congo red birefringence and transmission electron microscopy (TEM) were used to identify the features of the amyloid fibrils and to prove the format ion of β-amy lase fibrils.
However, the results of the Thioflavin-T binding (ThT) fluorescence assay did not show the formation of amylo id fibrils. The oxidation of Met35 in mono meric Aβ 40 can results in a decrease in C-terminal β-strand format ion as reported by Brown and coworkers (2014). Their results provided an insight onto the conflicting body of literature for this topic (Bro wn et al., 2014). The reduced β-strand content illustrated that the oxidation of Met35 was consistent with a decrease in the aggregation rate that was observed experimentally (Snyder et al., 1994; Hou et al., 2004).
16
Their findings were consistent with the experimental model in wh ich the oxidative stress in the AD brain induced toxicity. However, this was independent of Aβ fibril formation (Bro wn et al., 2014). Bad ja and Filipek (2015) simu lated the early stages of Aβ 13-23 by using MD in imp licit environ ment. They used nine identical α-helices of Aβ 13-23 fragment in one system for 500 ns. Their results showed that the small β-sheets emerged very early with two separated strand structures in the presence of membrane to facilitate this process. Following the hypothesis by Tomaselli et al. (2006), the α-toβ transition preceded Aβ aggregation because the starting structure was helical (Badja and Filipek, 2015).
2.5 Molecul ar Modeling Molecular modeling enco mpasses all theoretical methods and computational simulat ion techniques which are used to model or mimic the behavior of molecules in different environments. These techniques are used in different fields of co mputational che mistry, drug design, computational biology and material science to study the molecular systems ranging from s mall chemical systems to large biological mo lecules and material assemblies.
Classical Mechanics is the most important aspect of molecular simulat ion techniques such as MD in macroscopic world as it refers to the use of mechanical mechanics or Newtonian Mechanics to describe the physical basis behind the real phenomenon. Molecular models are typically described by spring-like interactions representing chemical bonds and van der Waals forces. Another most widely used method is mo lecular docking which is applied to predict the preferred orientation of one molecule to protein or another molecule when bound to each other to form a stable co mplex (Lengauer et al., 1996).
2.5.1 Molecul ar Docking Molecular docking is a co mputational procedure that attempts to predict non -covalent binding of macro molecules or more frequently, of a macro mo lecule (receptor) and a small molecu le (ligand) efficiently starting with their unbound structures, structures obtained from MD simu lations or ho mology modeling, and other built structures . The goal is to predict the bound conformations and the binding affinity (Trott et al., 2010). The prediction of binding of small mo lecules to protein is usually important in drug design and drug development. Docking programs generally use a scoring function to approximate the standard chemical potentials of the system (Gilson et al., 1997; Chang et al., 2007). Docking is also useful for the prediction of the bound conformation of known binders when the experimental studies of hollow structures are unavailable (Sousa et al., 2006; Trott et al., 2010). Successful docking methods are able to effectively search the high dimensional spaces and apply scoring function that correctly rank docking candidates.
17
Docking can also be used to perform virtual screening on large libraries of compounds and then rank the results and propose structural hypotheses on how the ligands inhibit the target which is invaluable in lead optimization during computer-assisted drug design (Morris et al., 2008). The scoring functions take a pose as input and return a number indicating the likelihood that the pose represents a favorable binding interaction. Most scoring functions are physics -based molecular mechanic force fields that estimate the energy of the pose in which a lo w energy indicates a stable system and thus a likely binding interaction. An alternative approach is to derive a statistical potential for interactions from a large database of protein -ligand complexes such as Protein Data Ban k (Halperin et al. 2002).
The most used docking programs are Genetic Optimisation for Ligand Docking (GOLD) (Jones et al. 1997) and AutoDock (Goodsell et al., 1996). Genetic algorith m permits the exp loration of a large conformat ional space which is basically spanned by the protein and ligand to represent each spatial arrangement of the pair as a gene with a particular energy. Even though the genetic algorithm is successful in sampling the large conformat ional space, many docking programs developed require the protein to remain fixed wh ile allowing only the ligand to change the position and adjust to the active sites of the protein. There is also a require ment for mult iple runs to gain reliable results regarding ligands that may bind to the protein. The improvements reported by using grid-based evaluation of energies, limit ing the exploration of the conformational changes at only local areas. These areas can be active sites of interest or improved tabling methods which have significantly enhanced the performance of genetic algorith ms and made them suitable for v irtual screening applications (Jones et al., 1997; Meng et al., 2004; Wei et al., 2004). 2.5.1.1 Molecul ar Docking Studies of Aβ-Pepti de Theoretical methods have been used for several years to expand the experimental work and speed up Alzheimer’s drug design process. Molecular docking is being applied in developing small molecu les as effective therapeutics against monomeric, oligo meric, and fibrillated forms of Aβ. They are usually used to figure out the ability of compounds to bind to proteins with well-defined binding site Aβ aggregation is a challenging area and the related informat ion are difficult to gather using traditional experimental techniques thus theoretical methods such as molecular docking and virtual screening can be very helpful (Teplo w et al., 2006).
For examp le, molecular docking has been used to characterize the interactions of small mo lecules with Aβ monomers. Teper et al. (2005) used the docking study of hydroxycholesterol derivatives to find the different compounds that can bind to large regions of the Aβ surface, encompassing nearly half of the sequence. In another study, Bray mer et al. (2011) investigated the binding of stilbene derivatives to Aβ monomers. They found that these compounds could bind to polar N -terminal residues that are believed to bind to metal ions and contribute to neurotoxicity.
18
A xanthone derivative docked to Aβ could stabilized the α-helical conformation of the peptide during a short (20 ns) MD simu lation (Wang et al, 2012) and by using a similar approach, Yang et al. (2010) docked a pentapeptide (LPFFD) to Aβ and analy zed the conformat ions adopted over the course of a short (30 ns) MD simulat ion, with the initial α-helical character of Aβ being retained as a result of LPFFD binding. Liu et al. (2006) performed MD p rior to molecular docking to generate a conformation to represent the solution conditions better. The compound examined in that work was bound to a large portion of the Aβ surface, but it was not inhibiting β-strand formation. They suggested that its inhibitory mechanism involved interfering with interpeptide hydrogen bonding.
A combined docking and MD approach were run to evaluate the binding energy of Aβ 16-20 with the frag ment Lys, Val, Leu, Phe and Phe (KVLFF) and peptide frag ment Leu-Pro-Phe-Phe-Asp (LPFFD) peptides which is known as a β-sheet breaker on mono meric Aβ 40 (Viet et al., 2011). They found that the LPFFD did not totally loss its total β-strand content but it prevented the formation of β-strands in aggregation-prone regions. Moreover, the LPFFD inhibited the transition of α-helix to rando m coil structures. All of these studies are the early efforts to study and understand the interactions of small molecu les and peptides with monomeric Aβ (Lemkul and Bevan, 2012).
2.5.2 Molecul ar Dynamics (MD) Simulati on Co mputational simulations act as a bridge between microscopic length and time scales and the macroscopic world to guess the interactions between molecules and predict bulk character. MD has been used to imp rove the understanding of phase transitions and the behavior of molecu les at interfaces for several years (Lee et al., 1974; Chapela et al., 1977; Frenkel and McFague, 1980). Quantitative and/or qualitative informat ion about macroscopic properties are usually estimated by the behavior of molecu les at mo lecular level fro m the simulat ion of the model system at atomic level. By this, the hidden detail behind bulk measurement can be revealed. The results produced from simu lation methods can be compared with those of real experiment which are mostly impossible to find during wet lab experiments under extreme pressure or temperature (Allen et al., 1989; A llen, 2004). This procedure will be further exp lained in chapter 3.
19
CHAPTER 3
MET HODOLOGY In this project, the amylo id-β with 42 residues was used as a model system. The interaction of Aβ (1-42) in different solvents involving water and solvent mixtu re was detailed. The solvent mixture was consisted of hexafluoro isopropanol (HFIP) and water. The ro le of HFIP/water mixture in stabilizing the helical structure has been explored by investigating the interactions of the solvent molecules with the peptide backbones experimentally not theoretically (Wei and Shea, 2006). HFIP can dissolve Aβ (1-42) better than all other med ia because of its helix-pro moting ability. The structure of Aβ (1-42) found in aqueous HFIP can be a mediu m that mimics the lip idic environ ment of membranes in boomerang shapes.
Metal ions are ubiquitous in protein systems and play a significant role during their folding processes (Wu et al., 2012). So me studies have been stated that in the presence of Zn 2+ ion, Aβ (1-42) can form co mplexes with histidine residues (Stellato et al., 2006; Silva and Saxena, 2013). Thus, the zinc ion was docked to peptide by using AutoDock Vina software (Trott & Olson, 2010). The structure with the lowest binding energy was selected as the initial structure for MD simu lations in different conditions . The models were consisted of the peptide in pure water (Aβ-H2 O), peptide in mixed solvents (AβHFIP), peptide with zinc in water (Aβ-Zn-H2 O), peptide with zinc in mixed-solvent (Aβ-Zn-HFIP). A summary of step is shown in Figure 4.
20
Figure 4. The outline of MD simulation steps for studyi ng the i nteraction of Aβ (142) with zinc ion Several researchers have stated that Zn 2+ ions could trigger Aβ aggregation which is very important in AD disease (Mantyh et al., 1993; Bush et al., 1994; Esler et al., 1996). Therefore, this study was continued by explo ring Aβ(1-42) aggregation in the presence of zinc in different conditions. Two models prepared including six mo lecules of peptide with zinc in water (6Aβ-Zn-H2 O) and solvent mixture (6Aβ-Zn-HFIP-H2 O). The outline of MD simulation steps for aggregation study is shown in Figure 5. All six simu lations were carried out using GROMACS software packages (Berendsen, et al., 1995; Lindahl, et al., 2001).
21
System Preparation • 6 molecules of Aβ + 6 Zn + H 2 O • 6 molecules of Aβ + 6 Zn + HFIP + H 2 O
Creating Topology • Define Box
Energy Minimization (EM)
Heating and preEquilibration
MD Production and Data Analysis Figure 5. The outline of MD simulation steps for aggregation study
3.1 Theoretical Background 3.1.1 Molecul ar Docking Molecular docking is used to predict and study on how two or more molecu lar structures like an enzy me and ligand will fit and bind together. Ligand is a small mo lecule that interacts with protein binding sites while binding sites are the areas of protein known to be active in formation of compounds. The several possibilities of mutual conformat ions in which binding may occur is called binding modes. Proteinligand interactions are different fro m generic protein-protein interaction because of the small size of ligand. Proteins are usually treated as rigid body due to their large size. For a successful docking process , the conformational changes of both protein and ligand are necessary.
22
Two main search algorithms are the search algorith m and rigorous searching algorith m. An optimu m nu mber of configurations originated fro m the experimentally determined binding modes is created by search algorithm. By using this algorithm, al l configurations are evaluated to distinguish the experimental binding modes fro m all other modes. The rigorous searching algorithm is used to go through all possible binding modes between two molecules . However, because of the size of the search space this is illogical. Thus, current docking methods have been modified in wh ich a two-stage scoring function is used as an alternative approach to direct the search by a reduced function and then to rank the resulting structures (Kaapro and Ojanen, 2002).
3.1.1.1 Free Energy Scoring Function The force field evaluates binding in two steps. The ligand and protein start in an unbound conformation. Firstly, the intramolecular energies are estimated for the transition from unbound states to the conformat ion of the ligand and protein in the bound states. Then, the intermolecular energetics of co mbined ligand and protein in their bound conformation is determined. The force field includes six pair-wise evaluations (V) and an estimate of the conformational entropy lo st upon binding (∆Sconf);
(1) where L refers to the ligand and P refers to the protein in the ligand-protein docking calculation. An assumption was made in an earlier version of AutoDock the most commonly used software for molecu lar docking in which the unbound form of a ligand (VL-L bound ) is the same as the final docked conformation of the ligand (VL-L unbound) thus a final contribution of VL-L bound - VL-L unbound equal to 0. Each of the pair-wise energetic terms includes the evaluations for dispersion/repulsion, hydrogen bonding, electrostatics and desolvation energies defined by the below equation;
(2)
The weighting constants , W have been optimized to calibrate the empirical free energy based on a set of experimentally determined binding constants. The first term is a typical 6/ 12 potential for dispersion/repulsion interactions. The parameters selected here are based on the AMBER force field. The second term is a directional H-bond term accord ing to a 10/12 potential. The parameter C and D are assigned in a way that they can give a maximal well depth of 5 kcal/ mo l at 1.9 Å for hydrogen bonds with oxygen and nitrogen and a well depth of 1 kcal/ mol at 2.5 Å fo r hydrogen bonds with sulfur.
23
The function E(t) provides directionality based on the angle, t from an ideal H-bonding geometry. The third term is a screened Coulomb potential for electrostatics energy calculations and the final term is a desolvation potential defined by the volume of atoms (V) that surrounds a given atom and shields it fro m solvent by a weighted salvation parameter (S) and an exponential term with distance-weighting factor of σ=3.5 Å (Morris et al., 1998; Huey et al., 2004; Huey et al., 2007; Morris et al., 2009; Cosconati et al., 2010; Forli and Olsom, 2012).
3.1.1.2 Genetic Algorithms (GAs) In the determination of complex optimization during molecular docking, genetic algorith ms and evolutionary programming are very suitable due to their efficiency. The fundamental idea of genetic algorith ms is the growth of a population of possible solutions via genetic operators to a final population by an optimized pre-defined fitness function. The process of applying genetic algorith ms starts with encoding the variables , for examp le b inary strings are ‘genetic code’ that changes by the degrees of freedo m. Later, the random init ial population of solutions is created and genetic operators that are used to this population leading to a new population. Consequently, the new population will score and rank. The probabilities of getting to the next iteration round depend on their score by using ‘the survival of the fittest’. Good solutions will maintain the population if the size of the population is kept constant. The GA is very suitable for parallel co mputing. The programs applying GA are GOLD, AutoDock, DIVA LI and DARWIN (Kaapro and Ojanen, 2002).
3.1.1.3 Docking Programs Docking programs are usually sold in a package with other molecu lar design software. There are no significant differences between different programs and they may all produce false alarms thus combining different searching and scoring functions, can produce more reliable results. Successful docking software not only create the right conformat ion but they are able to recognize it (Kaapro and Ojanen, 2002). AutoDock program has proven to be an effective tool with the ability to quickly and accurately predict bound conformations and binding energies of ligands with proteins (Goodsell et al., 1996; Morris et al., 1996; Morris et al., 1998; Osterberg et al., 2002; Huey et al., 2007). In order to allow the searching of large conformational space available to a ligand around a protein, AutoDock uses a grid-based method to rapidly evaluate the binding energy of trial conformations. The target protein is embedded in a g rid. Then, a probe atom is sequentially placed at each grid point and later, the interaction energy between the probe atom and the target is determined and the obtained value is restored. These grid energies will then be applied as a lookup table during docking (Morris et al., 2009).
24
Another common program for mo lecular docking and virtual screening is AutoDock Vina. Trott and Olson (2010) stated that AutoDock Vina can reach to an approximately two orders of magnitude speed-up compared to other molecular docking softwares. It can also significantly improve the accuracy of the binding mode predictions compared to AutoDock software. The general functional form of the conformation -dependent part of the scoring function of AutoDock Vina is formu lated as below;
(3) where the summation is over all the pairs of atoms that can move relative to each other without considering 1-4 interactions (atoms separated by three consecutive covalent bonds). From equation 3, each atom, i is assigned a type ti , a symmetric set on interaction functions, and the interatomic distance of rij . These values are determined based on a summat ion over all intermo lecular and intramolecu lar interactions;
(4)
The optimization algorith m which will be described further in the follo wing section, tries to find the global min imu m of c and other low-scoring conformat ions which is later to be ranked. The predicted free energy of binding is calculated from the intermolecular part of the lowest-scoring conformat ion by using the following equation;
(5 ) where the function, g can possibly be an arbitrary possibly nonlinear function which increases smoothly. In the output, other low-scoring conformat ions also have s values. Thus, to preserve the ranking, cintra of the best binding mode is usually used according to below equation; (6) For modularity reasons, Vina program does not rely on any functional form of interactions or g. In addition, these functions are considered as a parameter for the rest of the code. However, there are some terms that are different fro m X-score and in tuning the scoring function, therefore, one must note that Vina ranks the conformations according to eq. (4) or equivalent ly eq. (6) while the X-score only counts intermolecular contributions. Vina is compatib le with the file format used for AutoDock version 4 structure files in which PDBQT can be taken as an extension of the PDB file fo rmat (Morris et al., 1998).
25
This possibility makes Vina friendly and fast to be used by the existing au xillary software developed for AutoDock software like AutoDock Tools which is usually used to prepare the files by choosing the search space and checking the results at the same time. In addition, Vina can estimate its own grid map quickly and automatically without saving them on the disk. Hence, identical results will be produced but executed faster (Trott et al., 2010). The size and the location of the binding site can be visualized by PyMOL (Schrodinger, 2010), Discovery Studio version 4.0 ( Studio and Insight, 2009) and also LigPlot program (Wallace et al., 1995; Laskowski and Swindells, 2011) thus it can be adjusted interactively. Several residues within the bindin g site can be optionally defined to be flexible during docking. Subsequently, the necessary files for the receptor definition will be generated automatically. Similarly, file preparations for mu ltip le ligands are also controlled by the plugin.
Then, the docking calculations results are loaded in all programs and visualized. By this, the results of mult iple docking runs can be automatically analyzed (Seeliger et al., 2010). Docking poses generated by Vina program can be direct ly loaded into PyMOL through the plugin. Poses for multip le ligands are handled simu ltaneously using an intuitive notebook layout. For each docking pose, meta-information containing the docking score is displayed in a small text viewer, allo wing the direct analysis of configuration/score relationships. The docking poses are ranked according to their docking scores and both the ranked list of docked ligands and their corresponding binding poses are exported. In fact, the ranked list of docking results are exported in a CSV file format and then directly imported into an Excel file.
AutoDock Tools, created as part of AutoDock software, is a widely accessible tool. AutoDock Tools is able to cluster, display and also analyze the results of docking experiments. AutoDock Tools is imp lemented in the object programming language Python and built fro m reusable software components (Ascher & Lutz, 1999; Sanner et al., 1999). It exists in the context of a rich set of tools for mo lecular modeling, the Python Molecular Viewer (PM V) (Sanner et al., 1999; Sanner et al., 2005).
3.1.2 Molecul ar Dynamics (MD) MD simu lations are considered as one of the principal tools in studying chemical and biological mo lecular systems. They provide a lot of useful and atomic detailed informat ion on the fluctuations and conformational changes of proteins and nucleic acids. These methods are now routinely used to investigate the structure, dynamics and thermodynamic properties of d ifferent model systems .
They are also applied in the determination of structures from X-ray crystallography and NMR (Alder & Wainwright, 1957; Alder & Wainwright, 1959; Rah man, 1964; Stillinger & Rah man, 1974; Mc Cammon et al., 1977). In MD, the motion of the atoms in a molecular assembly is treated by Newton’s dynamics followed by the determination of the net force and acceleration experienced by each atom ( Hansson et al., 2002; Allen, 2004; Binder et al., 2004; Lindahl, 2008; van der Spoel et al., 2010). The Newton second law of mot ion is defined by the following equation;
26
(7 ) Using the positional vector of atom i (ri ) over time (t) for N interacting atoms , equation (7) can also be written as; (8 )
where i=1,2,3… N. The forces (F) calculated fro m MD integration method result in a negative derivative of potential energy function (V) as shown below; (9) These equations are solved simultaneously in a small t ime scale, with temperature and pressure treated imp licit ly or exp licit ly, depending on the type of MD simulations carried out.
3.1.2.1 Force Fiel d There are three main types of force fields available for MD simulat ions. All-atom level force field p rovides parameters for every single ato m within the system wh ile the united-atom force field produces parameters for all ato ms except non-polar hydrogen. In between, coarse-grained force field is an abstract representation of molecules by grouping several atoms into super-atoms.
The numerous sets of parameters can be used within one set of equations. These force field parameters that have been implemented in different softwares are AMBER (Wang et al., 2004), CHARMM (MacKerell et al., 1998), OPLS (Jo rgensen et al., 1996; Kaminski et al., 2001) and GROM OS (Gunsteren et al., 1996; Bonvin et al., 2000; Schuler et al., 2001; Oostenbrink et al., 2004). All force fields are continuously being tested and improved. For examp le, in the GROMOS force field, the A-version is the basic force field designed for molecu les in solution or in crystalline for m while the Bversion which is derived from the A-version is used for simu lating mo lecules in vacuo, where the dielectric screening effect of the environment is being ignored. The atomic charges and van der Waals parameters are changed such that atom charge groups with a non-zero total charge are neutralized while maintaining the hydrogen-bonding capacity of the individual atoms.
In the newer versions of GROM OS force field, build ing blocks are supplied via more than one file. These files have several building blocks that are categorized according to the kind of molecule they show. GROMA CS package which was emp loyed in this study supports the GROMOS-96 force fields (Gunsteren et al., 1996). GROM OS-96 force field is an improved version of the GROMOS-87 (Gunsteren et al., 1987). It has been parameterized with Lennard-Jones cut-off value of 1.4 n m because the LennardJones forces are almost zero beyond 1.4 n m.
27
The GROMOS-96 53A 5 and 53A6 force fields (Oostenbrink et al., 2004) are the result of a co mplete re-parameterization of the non-bonded interaction parameters for condensed phase simu lations of pure small mo lecules in liquid phase (53A 5) and solutions of mo lecular systems in water or apolar solvents (53A6).
All interaction types have been redefined in these force fields by involving another set of parameters for addit ional solvents. A force field is built up fro m a set of equations called potential functions which will then be used to generate the potential energies . The general force field equation was first formulated by Weiner et al. (1984) in which they determined a charge model appropriate for both all -ato m and united-atom representations as follows;
(10)
The potential energy functions are subdivided into two parts consisted of non-bonded and bonded interactions. The non-bonded interactions are calculated based on creating a neighbor list (a list of non-bonded atoms within a certain radius) in which the exclusions have already been removed, while the bonded interactions determine covalent bond-stretching, angle-bending, improper dihedrals and proper dihedrals. The bonded and restraint interactions are determined based on fixed lists which can be seen in the topology file. The bonded interactions are built accord ing to the fixed list of atoms. They also consider 3- and 4-body interactions instead of just pair interactions. A special type of dihedral interaction or improper dihedral is used to force atoms to remain in a p lane or to prevent transition to a configuration of opposite chirality. Improper dihedrals keep planar groups planar or prevent them fro m flipping over their mirror images. The non-bonded interactions which contain a repulsion term, a dispersion term and a Coulo mb term are also known as pair-addictive and centrosymmetric defined by the following general equation;
(11)
(12)
where V is the potential energy between two atoms, i and j within the radius, r. There are two non-bonded terms; the Lennard-Jones potential measures the van der Waals interactions between atomic radii, and the Coulomb force wh ich calculate the electrostatic interactions between charged atoms. The sum of the non-bonded interactions follows the below equation;
28
(13)
The first term, the Lennard-Jones potential, employs unique ε and σ values for each possible pair of atoms, i and j. The ε value is the min imu m potential energy for two interacting species at the ideal radius of separation, and the σ value is the distance at which the potential of interaction is zero. The final variable, rij is the radius of separation between atoms i and j.
In the second term, the Cou lo mb potential or the energy of interaction between two charged atoms (i and j) depends on the charges of each atom (q i and q j ), the permittiv ity of free space, ε0 (a constant), and the radius of separation of the two charged atoms, rij . The repulsion and dispersion terms are co mbined in either the Lennard-Jones (or 6-12 interaction), or the Buckingham (or exp -6 potential). In addit ion, (part ially) charged atoms can also be treated by the Coulomb term. The cut-off range cannot exceed half the box size because of the min imu m-image convention. The interacting pair is considered only one image of each particle in the periodic boundary condition.
To reduce the problem for the systems containing charged particles and also large systems, the long-range electrostatic algorithm such as particle mesh Ewald (PM E) method was proposed (Darden et al., 1993; Essmann et al., 1995). For examp le, Weber et al. (2000) found that a peptide in a smaller periodic bo x tends to be in a stable αhelix form while when the box size became larger, the periodicity artifacts were consequently smaller for the same peptide during unfolding. The conditionally convergent series was first reconstructed by Ewald (1921) to a sum of rapidly convergent series. He considered each point charge surrounded by a charge distribution of equal magnitude and opposite sign.
For many cases, the Ewald summation is more efficient computationally and stable compared to a larger cut-off radius (Essmann et al., 1995; Norberg and Nilsson; 2000). Thus, the sum of the bonded and non-bonded terms fro m each source is the total potential energy of interaction (Vtotal ) of a given atom simp lified as follows;
(14)
The force that is exerted on each atom is equal to the negative derivative of potential energy with respect to the position as explained earlier.
29
3.1.2.2 Restraints Interaction Restraint part is usually used to include experimental data. In fact, they are not part of the force field thus the reliability of the parameters is not important. The position restraint is used to restrain particle to fixed positions. It is usually applied during equilibrat ion to avoid the drastic rearrangement of crit ical parts, for examp le, to restrain motion in a protein that is subjected to large solvent forces when the solvent is not equilibrated yet. The restrain ing energy can maintain the integrity of inner part. It is mostly wise to be used for spherical shells to make the force constant depends more on radius. The restrain is increased from zero at the inner part to a large value at the outer boundary according to the following equation;
(15)
where Ri is the fixed reference positions , ri is the position restraint for particles, i and Vpr is the potential position restraint. The force constant of position restraint is kpr . Positions restraints are applied to a special fixed list of ato ms.
3.1.2.3 Periodic Boundary Conditi on (PB C) The systems under study are usually put in the space-filling bo x surrounded by translated copies of itself called PBC. It is a requirement that the unit cell to remain in a shape that it can tile perfectly into a three-dimensional crystal. Therefore, the artifact fro m unwanted boundaries in an isolated cluster will be replaced by the artifact of periodic conditions. A cube or rectangular prism is the most intuitive and common choice but it can be computationally expensive due to the unnecessary amounts of solvent molecules availab le in the corners, distant from the central macro molecules. Hence, a common way is to introduce a box with less volume wh ich can be more economical such as the truncated octahedron and rhombic dodecahedron. Both shapes (Adam et al., 1979) are closer to a sphere than a cube for simu lating (an approximately spherical) macro molecule in solution, since fewer solvent molecules will be required to fill the box of simulat ion. Thus, a minimu m distance between macro molecu lar images will be supported perfectly (Bekker et al., 1995).
The volume of a rhombic dodecahedron is approximately 71% of a cube with the same spacing pattern and for truncated octahedron is 77%. Even though the difference seems small but it will be 30% more valuable when simulat ions are run for weeks. It is far fro m triv ial to see in three dimensions but Figure 6 shows how a hexagonal cell similarly can be more efficient than a square in two dimensions.
30
Figure 6. Two-di mensional example of how a hexag onal box leads to lower vol ume than a s quare box with the same separati on distance (Lindahl, 2008)
The rhombic dodecahedron is equivalent to a periodic system designed based on a triclinic unit cell. The rhombic dodecahedron is the smallest and most regular spacefilling unit cell wh ich is able to keep each of the 12 image cells in the same distance. As a result, 29% of the CPU time will be saved during the simulat ion of a spherical or flexib le mo lecule in solvent which is shown by the equations below;
(16) (17) (18) where the bo x is defined by 3 bo x vectors which are a, b and c. The equation (16) is suitable for rotation while the equation (17) and (18) is always satisfied for adding and subtracting box vectors.
3.1.2.4 Energ y Mi nimizati on The stable conformer of a molecu le is very important during simu lation because it allo ws understanding its properties and behavior based upon structural considerations. Energy minimization (EM) is usually carried out to produce a stable conformer. EM is a numerical procedure for finding a min imu m on the potential energy surface starting fro m a higher energy initial structure. After performing many steps of EM, a local or global minimu m on the potential energy surface can be reached (Jean, 2015).
31
EM is usually performed by using two different algorithms; steepest descent and the conjugate gradient. Even though the steepest descent technique is certainly not the most efficient algorithm for searching but it is easy to imp lement . First, the forces, F and potential energy are calculated by using the following equation;
(19) where h n is the maximu m d isplacement and Fn is the force or the negative gradient of the potential energy, V. The notation max means the largest absolute values of the force components. The min imization stops when the amount of force is converged. Then, the conjugate gradient (cg) method is applied (van der Spoel et al., 2010). In the cg method, the first portion of search takes place in the direction of the largest gradient determined by the steepest descent method previously.
However, to avoid from the back and forth fluctuation which often plaques the steepest descent algorithm as it move toward the min imu m, the cg method normally uses a part of the previous direction for the following search. By this, this method can move rapidly and then converge to the minimu m value. Th is method is quite common for energy minimizat ion of large mo lecules among researchers (Jean, 2015). Unfortunately, it is sometimes problematic. For examp le, in GROMACS package, cg cannot be used when constraints are applied by the SETTLE algorithm (M iyamoto & Kollman, 1992) fo r water mo lecules but it can be changed in the .mdp file by define = DFLEXIBLE where the water comes fro m a flexible model.
3.1.2.5 Temperature and Pressure Coupling The temperature of system should be controlled during simu lation because the mo lecule might drift during equilibration. This drift can originate from force truncation and integration errors or external or frict ional forces during heating. The weak coupling scheme for controlling temperature was introduced by Berendsen and his colleagues (Berendsen et al., 1984). The Berendsen algorithm mimics weak coupling with firstorder kinetics to an external heat bath according to a given temperature of . Thus, the deviation of system temperature fro m is slowly corrected according to the following equation;
(20) As results, the temperature deviat ion decays exponentially with a t ime constant, . The strength of the coupling can be varied and adapted to the requirement of the simu lation. For equilibrat ion, the coupling time is quite short, for examp le, 0.01 ps but for a reliable equilibriu m it has to be much longer, for example 0.5 ps thus it can hardly influence the conservative dynamics.
32
The Berendsen thermostat also has an additional stochastic term wh ich ensures that a correct kinetic energy distribution can be obtained as follows; (21)
where is the kinetic energy, is the number of degrees of freedom and is a Wiener process. When the NVT ensemb le (constant number of particles, volu me, and temperature) is used for simulat ion, the conserved energy quantity is added to the energy log file. This ensemble is also referred to as canonical ensemble. Eventhough the Berendsen’s thermostat is stable and simp le to implement and physically appealing, it shows some art ifacts. It has no conserved quantity and it is not associated to a well defined ensemble, except in limit ing cases. However, this is not that significant and, this thermostat is widely used. Another method for controlling the temperature is velocity rescaling also known as V-rescale (Bussi et al., 2007). This method is an extension of the Berendsen thermostat to enforce the correct distribution of the kinetic energy by adding a properly constructed random force.
Thus, the trajectories are not affected when a relaxation time is chosen. This method leads to the correct canonical distribution through a unified scheme in wh ich Berendsen thermostat was problemat ic. A remarkable result is defining a quantity which is constant and at the same time plays a role similar to that of the energy in the micro canonical ensemble. Namely, it can be used to verify several procedures that generate configurations belong to the desired NVT ensemble.
The velocity-rescaling method involve the multiplication of the velocities of all the particles by the same factor, α that is calculated by enforcing the total kinetic energy K to be equal to the average kinetic energy at the target temperat ure by using
,
where N f is the number of degrees of freedom and β is the inverse temperature. Thus, the rescaling factor, α for the velocit ies is obtained by the below equation follows;
(22) This will result in no more effect on constrained bond lengths and the motion of the center of mass as the same factor is used for all particles . This operation is generally performed at a predetermined frequency during equilibration or when the kinetic energy exceeds the limits of an interval centered on the target value. The sampled ensemble is not exp licitely known but, since in the thermodynamic limits the average properties do not depend on the ensemble chosen, thus this very simple algorith m will still be very useful. When the simu lation is run in the NPT ensemble where the number of particles, pressure, and temperature are all constant (also known as isothermalisobaric ensemble), controlling of pressure becomes very vital.
33
MD simu lation is usually performed in an isotropic condition where the Berendsen algorith m (Berendsen et al., 1984) rescales the coordinates and box vectors in every step based on a matrix , wh ich affects the first-order kinetic relaxat ion of pressure towards the given reference pressure, as follows; (23)
The scaling matrix,
is calculated as follows;
(24)
where, β is the isothermal co mpressibility of system when the fluctuations in pressure or volume are mo re important. Another option is the Parrinello-Rah man pressure coupling scheme (Parrinello & Rah man, 1981; Nosé & Klein, 1983) that is similar to the Nosé-Hoover coupling in some cases but it may, at least theoretically, be a problem to define the exact and well-defined ensemble co mpared to the weak coupling simu lation. However, this barostat is very helpful. In the Parrinello-Rah man barostat, the box vectors are represented by the matrix, b dictates by the matrix equation of motion as follows;
(25) The volume of the box is denoted by V, and W is a matrix parameter that determines the strength of coupling. The matrices, P and Pref are the current and reference pressure, respectively. It has been suggested that a weak coupling should be done first to get the target pressure, and then switched to Parrinello-Rahman coupling once the system became stable (Parrinello & Rahman, 1981).
3.2 Materials and Methods 3.2.1 Materials Computer Hardware
A Dell Workstation T5500 with Intel® Xeon® with 4 CPUs and 2.5 Gigabyte of RAM. 64-Nodes Co mputer Cluster (Depart ment of Chemistry, Universiti Putra Malaysia).
34
Computer Software
Ubuntu 12.04 LTS Groningen Machine for Chemical Simu lation (Gro macs v4.5.5) Package (van Der Spoel et al., 2013) The PyMOL Molecu lar Graphics System, version 1.4.1 (Schrodinger, 2010) Visualized Mo lecular Dynamic (VMD) (Hu mphrey et al., 1996) Packmo l (Martínez, L., et al., 2009) Grace Plotting Soft ware (Turner et al., 2004) AutoDock Vina (Trott et al., 2010) An Automated Force Field Topology Builder (ATB) and Repository v1.0 (Malde et al., 2011) LigPlot (Wallace et al., 1995; Laskowski et al., 2011) Discovery Studio 4.0 Visualizer (Studio and Insight, 2009)
3.2.2 Methods 3.2.2.1 The Model Structure The model structure used in this study was Aβ (1-42) peptide. That was a α-helix-kink-αhelix obtained by NMR spectroscopy deposited in Protein Data Bank (Entry code: 1IYT) (Creszenci et al., 2002). The PDB structure was selected as the starting structure for both molecular docking and MD simu lation study. There are 42 residues in Aβ (1-42) as shown in Figure 7. It has been reported that the fibrillogenesis which may happen in AD cases involves an oligomeric α-helical intermed iate (Kirkidtadze et al., 2001). The structural characterizat ion of a mono meric and soluble fo rm of Aβ (1-42) in an isotropic med ia is necessary not only to shed some light on the steps involved in the fibro llogenesis but also to evaluate the role of amylo id-β in the interaction with the memb rane. The structure of Aβ (1-42) has been found in aqueous hexafluoroisopropanol (HFIP) which is a mediu m that mimicks the lipidic environ ment of memb ranes as a boomerang-shaped. It has been stated that the second helix (residue 28-38) of Aβ (1-42) structure corresponds to the transmembrane region of APP with the typical amino acid composition of transmembrane helices for example small (Gly and Ala) and hydrophobic (Ile, Leu, Met and Val) residues (Eilers et al., 2000; Creszenci et al., 2002). Its only charged residue along this sequence is Lys28 which is placed at the Nterminal end of the helix (Creszenci et al., 2002).
35
Figure 7. Amyloi d-β(1-42) (PDB code: 1IYT); α-helices (red), turns (green) and coils (whi te)
Hexafluorisopropanol or 1,1,1,3,3,3-hexafluoropropan-2-ol, (HFIP) (Figure 8) is a highly polar water-wh ite solvent which is used for a variety of chemical processes. Rajan et al. (1997) suggested that the peptide fluoroalcohol association could by driven by the hydrophobic effects originated fro m the trifluoro methyl subsituents. Thus, with the larger number of water molecu les involved, the process of dehydration will be entropically favorable. Hexafluoroacetone (HFA) mo lecules effectively provide a ‘teflon face’ which secludes the peptide in a non-interacting environment in which intramolecular hydrogen bond formation becomes energetically favored. Th is tefloncoated peptide can be solubilized by the preponderance of O-H groups on the surface. But, the poor hydrogen bond acceptor property of HFA and other fluoroalcohols limits their ability to insert into C=O-H-N bonds in peptides in contrast to water (Rajan and Balaram, 1996; Rajan et al., 1997).
Figure 8. Hexafluoroisopropanol (HFIP) structure
36
HFIP is an acidic alcohol because of its strong hydrogen bonding properties. It is also very useful as a solution phase for peptide chemistry research. Several NM R measurements on both Aβ (1-40) and Aβ (1-42) which were carried out in different solvents showed that the interface between aqueous and apolar phases similar to SDS micelles (Co les et al., 1998; Shao et al., 1999). Also, it can be found in solvents that reproduce an apolar microenviron ment such as trifuloroethanol/water mixtu res (Sticht et al., 1995). An aqueous mixture of a fluorinated alcohol, HFIP was chosen in this study because it can dissolve Aβ(1-42) better than all other media and at the same time it has a helix-pro moting ability wh ich is very similar to triflouroethanol (Rajan et al., 1997; Bhattacharjya et al., 1999; Crescenzi et al., 2002).
3.2.2.2 Molecul ar Docking Molecular docking experiment for docking Zn 2+ ion to peptide was performed by using AutoDock Vina in advance to the preparation of models involving zinc (Trott et al., 2010). It was done to determine the potential zinc-binding sites of Aβ (1-42). The obtained results were then clustered using a tolerance of 1.0 Å. The single Zn2+ ion was extracted fro m the protein amyloid-β with residue 1 to 16 with the PDB code 1ZE9 and then it docked to the Aβ (1-42) (rigid receptor docking) where the Zn 2+ ion could dock randomly with the bo x centered at x= -0.616 Å, y= -0.065 Å and z= 1.269 Å. The number of points was; x-dimension=104; y-dimension=74 and zdimension=108 with the total grid points per map of 858,375. A total of 10 docking experiments were run for each co mp lex using the Lamarckian genetic algorith m (LGA ) with the default parameters fro m AutoDock 4 (Morris et al., 2009). The main function of LGA resemb les the style of Darwin ian evolution and applies Mendelian genetics (Morris et al., 1998). There are several methods for conformation search in AutoDock Vina but the Lamarckian genetic algorith m g ives the most efficient search for general applications (Morris et al., 2012). A summary of the steps used for molecular docking is presented in Figure 9.
37
Mapping binding site and ligand • coordinate of macromolecule/receptor (1IYT.pdb) • coordinate of ligand (Zinc.pdb) - change from 0 to 2 for Zn 2+
Preparing Grid Parameter • Set spacing to 1 Å and create the box
Evaluating Interaction Energy (Scoring Function) • Preparing Docking parameter (in AutoDock4) • Selecting algorithm >LGA (run GA=10)
Performing Docking • using AutoDock Vina
Results and Analysis
Figure 9. Methodology used for molecular docking
38
3.2.2.3 Molecul ar Dynamics (MD) Simulati on The topology of HFIP was generated by using the steps shown in Figure 10.
A B3LYP/6-31G* geometry optimization
Assigning bonded and van der Waals parameters based on GROMOS 53A6 (Berendsen et al., 1981) force field
Estimation of initial charges using the ESP method of MerzKollman (Singh et al., 1984; Besler et al., 1990)
Generating final charges and charge groups using ATB (Malde et al., 2011)
Figure 10. Summary of the steps for creating the topol ogy of HFIP molecule
39
The water was modeled by the explicit simple point-charge (SPC) model (Berendsen et al., 1981) while HFIP was prepared by Automated force field Topology Builder (ATB) (Malde et al., 2011). All simu lations were run using periodic boundary conditions in the dodecahedron simu lation box with the minimu m distance between the solute and the box wall being 1.0 n m. Figure 11 shows the box created for the simulat ion of single peptide with and without zinc and the aggregation of peptide before it is filled with solvent.
Figure 11. Pictures of simulati on boxes built for (a) single pepti de only (b) pepti de wi th zinc and (c) six molecules of pepti de with six zinc i ons for aggregation study First two models (Aβ-H2 O; Aβ-HFIP-H2 O) were prepared in a solvent mixture composed of hexafluoroisopropanol (HFIP) and water in the ratio of 80:20 v/v. After creating simu lation bo x, the volu me of the bo x was 307.05 n m3 for water but for solvent mixture 272.53 n m3 box created. Both boxes were then filled up with SPC water molecules (Berendsen et al., 1981). The first model filled up with 8,777 water mo lecules. The second model involved 711 mo lecules of HFIP and 4,063 of water mo lecules. The total charge of Aβ (1-42) in both systems was -3 a.u wh ich could represent the biologically relevant state of the system (Ohnishi et al., 2004). The net charge of both model systems were neutralized by replacing three water mo lecules with three sodium ions as shown in Figure 12. For the system of Aβ (1-42) with Zn 2+ ion in different conditions, the charge of both models (Aβ-Zn-H2 O; Aβ-Zn-HFIP-H2 O) was -1 a.u which was neutralized later by adding one sodium ion to replace one water molecu le.
40
Figure 12. Pictures of simulati on box for (a) Aβ (1-42) in pure water (Aβ-H2 O) and (b) Aβ(1-42) in mixture of water (bl ue) and HFIP (red) (Aβ-HFIP-H2 O) Figure 13 shows Aβ (1-42) with zinc (in yellow) including (a) the peptide with the 8,780 mo lecules of water and (b) the peptide in 661 of HFIP and 4,351 of water molecules. For aggregation study, six mo lecules of Aβ (1-42) peptides packed randomly by using packmol software (Mart inez et al., 2009) in different conditions (Figure 13). The first model, 6Aβ-6Zn-HFIP-H2 O was prepared in solvent mixture consisting of 800 HFIP and 6,666 water mo lecules wh ile the system in water (6Aβ-6Zn-H2 O) had 12,142 water mo lecules. The total charge of both prepared models was -6 a.u because the presence of six zinc ions.
Figure 13. Picture of the simulations boxes for (a) Aβ (1-42) with zinc (yellow) in pure water (Aβ-Zn-H2 O) and (b) Aβ(1-42) with zinc in sol vent mi xture (Aβ-ZnHFIP-H2 O)
41
Figure 14. Pictures of the simulation boxes for pepti de aggregati on study; (a) water (6 Aβ-6 Zn-H2 O), (b) sol vent mixture (6Aβ -6Zn-HFIP-H2 O) Four MD simulations including peptide in water (Aβ -H2 O), peptide in the mixed solvent (Aβ-HFIP-H2 O), peptide with zinc in water (Aβ-Zn -H2 O) and peptide with zinc in the mixed solvents (Aβ-Zn-HFIP-H2 O) were performed using GROMACS software packages version 4.5.5 (Berendsen et al.,1995; Lindahl et al., 2001; Van Der Spoel et al., 2005). MD simu lations for amyloid-β with and without zinc ion in different solvents conditions were performed for 1 μs wh ile the simulat ion time period for aggregation study was 100 ns. GROMOS96 53a6 force field was selected to treat the potential energy of the model systems (Van Gunsteren et al., 1996; Bonvin et al., 2000; Oostenbrink et al., 2004).
For each model, the system was first energy minimized with positionally restrained solute followed by a min imization without any restraints to relax the solvent molecules. All systems were minimized by using the same steps involving 250,000 steps for both steepest descent and conjugate gradient methods until the energy was converged with the force less than 12 kJ/ mol for Aβ (1-42) in water, 15 kJ/ mo l for Aβ (1-42) in the mixed solvents and 33 kJ/ mo l for both Aβ (1-42) with and without zinc in water and solvent mixtu re while for the aggregation study, the force was less than 25 kJ/ mol for 6Aβ-6Zn in water and 35 kJ/ mol fo r 6Aβ-6Zn in mixed-solvent when the energies converged. For temperature coupling two separated groups including protein and non-protein were created. In GROMA CS, the command trjconv is used to convert the trajectory to different unit-cell representation while editconf command produces the orientation providing a square intersection with the xy-plane and also creates the simulation bo x.
42
All model systems were heated to 300K in an NVT ensemble with the constant number of particles, volu me, and temperature for 5 ns. The temperature was controlled by temperature coupling using velocity rescaling method (v -rescale) (Bussi et al., 2007) with the coupling constant of T = 0.1 ps. The electrostatic interactions were cut-off at 1.0 n m using the Particle Mesh Ewald (PM E) algorithm (Darden et al., 1993; Essman et al., 1995). The fourth order interpolation was used by setting up the pme-order to four and the fourier spacing was equal to 0.14 n m. The cut-off value of van der Waals interaction was also set to 1.0 n m. The LINCS algorithm (Hess et al., 1997) was used to constrain all bond lengths and also water molecules in the peptides and HFIP and allo w integration time step was 2 fs (see APPENDIX M).
There are several lists of parameters that can be used to treat the cut-off values which can be found in the .mdp file of GROMA CS program. Both coulomb and van der Waals interactions are denoted by the interaction type selectors termed vdwtype and coulombtype. Adding these two parameters can complete a total of six non -bonded interactions. The neighbor searching (NS) can be performed using a single -range or a twin-range approach which described by two radii rlist and max rcoulomb and rvdw. Usually, one builds the neighbor list every 10 time steps or every 20 fs using parameter nstlist. In the neighbor list, all interaction pairs which fall within rlist are stored. Furthermore, the interactions between pairs that do not fall within rlist but they are within max rcoulomb and rdvw are calculated during NS. The forces and energies are saved separately and added to short-range forces at every time step between successive NS. If rlist = max (rcoulomb,rvdw), no forces are evaluated during neighbor list generation.
After all systems reached the targeted temperature of 300K, the simulation was extended to run a short equilibration for 2 ns in an NPT ensemble (constant number of particles, pressure, and temperature). LINCS algorith m was again used to constrain all bonds and also water (Hess et al., 1997). The Particle Mesh Ewald (Essman et al., 1995) was applied for electrostatic interaction treatment with the cut-off value of 1.4 nm and the van der Waals interactions were also cut at 1.4 n m. The fourth order interpolation was used with the p me-order of four and the spacing of 5 x 10-3 to give electrostatic energies accurately (Lindahl et al., 2001). The pressure was kept constant at 1 bar by using Berendsen thermostat (Berendsen et al., 1994) with p =2.0 ps until all systems were well equilibrated. The pressure coupling type was isotropic (Parrinello et al., 1981) (see APPENDIX N).
Later on, MD production simu lations were carried out in an NPT ensemble. The temperature and pressure were kept constant at 300K and 1.0 bar by using V-rescale algorith m (Bussi et al., 2007) and Berendsen thermostat (Berendsen et al., 1994). PM E algorith m was used for long-range electrostatics interaction treatment (Darden et al., 1993; Essman et al., 1995). The cut-off values were 1.4 nm for both van der Waals and electrostatic forces. All bonds even heavy atom-H bonds were constrained by using LINCS methods (Hess et al., 1997). A ll other parameters were the same as equilibrat ion step (see APPENDIX O).
43
CHAPTER 4
RES ULT AND DISCUSSION A single Amy loid-β (Aβ) chain usually contains fro m 38 to 43 residues and it prones to aggregate in aqueous environment (Dubnovitsky et al., 2013; Cohen et al., 2013). There are two main forms of Aβ including Aβ (1-42) and Aβ (1-40) but the longer one shows much higher tendency towards aggregation and greater toxicity (Gu and Guo, 2013). Pro-o xidant role of transition metal ions such as iron, copper and zinc was experimentally studied by Jamora and Valko (2011). They stated that these ions are essential for neuronal activ ities and the brain regulates the homeostasis of these metal ions as part of its physiology. However, it has been suggested that dyshomeostasis of metal ions in the brain might have contribution to Alzheimer disease (AD) pathogenesis. Aβ demonstrates a high affinity for Cu 2+, Zn2+ and Fe3+ ions producing reactive o xygen species (ROS) which increases the Aβ toxicity (Finefrock et al., 2003). A 1:1 ratio of zinc and Aβ(1-42) at pH 7 was studied. The zinc binding interaction study with Aβ (1-42) followed the works of Zirah et al. (2006) and Curtain et al. (2001) experimental investigations . It stated that zinc binding to apo-Aβ(1-16) could have a stabilization effect instead of a drastic transformation. Miller and co-workers (2010) have also shown that Aβ (1-42) zinc-b inding site may be located in two peptide domains including the disordered N-terminal domain (residues 1-16) and the C-terminal do main (residues 17-42). The His, Glu and Asp residues of Aβ (1-42) have a great potential to coordinate to metal ions (Miller et al., 2010).
Extracellu lar senile plaques and intracellular neurofibrillary tangles are the main hallmarks of AD (Guo and Lee, 2011; Gandy and DeKosky, 2013). The accumulat ion of Aβ (1-42) in the form of insoluble aggregates in the brain is considered as an important step in the pathogenesis of AD in amyloidogenic pathway (Thinakaran and Koo, 2008; Zeynep et al., 2014) while the aggregation mechanism has not been s pecified yet. The abnormal aggregates of Aβ(1-42) are believed to induce hyperphosphorylation among tau proteins, tangle format ion and neuronal loss which eventually results in cognitive impairment. Therefore, the modulation of abnormal Aβ (1-42) aggregation can be a potential therapeutic target for AD (Yanagisawa et al., 2015).
We performed the MD simulat ion of a single amy loid-β peptide with and without zinc ion in different solvent conditions for 1 μs. In addition, two MD simulat ions were carried out to investigate the aggregation process of Aβ (1-42) at different conditions for 100 ns. The water and HFIP were used as solvents in our study. The HFIP acts as hydrogen bond breaker and it is usually applied for eliminating pre-existing structural in homogeneities in Aβ (1-42) (Stine et al., 2003).
44
4.1 Molecul ar Docking 4.1.1 Bindi ng Site Anal ysis A molecular docking study usually starts with definition of binding site which is a restricted region of the protein in general. In our experiment, the receptor was amyloidβ. The zinc ion as ligand mo lecule was docked to amy loid-β peptide using AutoDock Vina (Trott et al., 2010) in order to find the potential interactions between peptide and metal ion.
A plugin was set for PyMOL (Schrodinger, 2010) to carry out the binding site analysis. Fro m the total of nine docked structures generated during molecular docking process, a single docked conformat ion of Aβ (1-42)-Zn with the lowest binding energy was selected as initial structure for MD simulat ion study of Aβ (1-42) with metal. The binding energies of zinc ion which was randomly docked to single Aβ (1-42) is summarized in Table 1. Table 1. A summary of docking results for single Amyl oi d-β pepti de bound to zinc ion Mode
Binding Energy (kcal/ mol)
Distance fro m RMSD (l.b.)
Best Mode RMSD (u.b.)
1
-0.9
0.000
0.000
2 3 4 5 6 7 8 9
-0.9 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8 -0.8
22.684 38.732 41.888 40.395 24.313 32.431 24.338 3.406
22.684 38.732 41.888 40.395 24.313 32.431 24.338 3.406
Fro m the results, the binding affin ity did not show a large difference among different modes and it changed between -0.9 kcal/ mol and -0.8 kcal/ mol. The top two conformat ions with the lowest energy were mode 1 and 2 with the same RMSD values for both lower (l.b) and upper boundaries (u.b). Mode 1 revealed the interaction of zinc ion and Valine-39, Isoleucine-41 and Alanine-42 residues in which RMSD values for distance and best mode were 0.00 Å while in mode 2, zinc ion was connected to the Aspartate-23 and Asparagine-27 with the best mode at RMSD, 22.68 Å.
The other conformations from mode 3 to 9 showed similar binding affin ity of -0.8 kcal/ mo l. Not much difference was found in the position of zinc ion among these conformat ions. Mode 3 with the RMSD values of 38.73 Å showed the interaction between residues Aspartate-7 and Glutamate-11 while in the mode 4 (41.89 Å ), interaction was detected between Glutamate-3 and Histidine-6 and zinc ion.
45
In mode 5, the interaction of residues Arginine-5, Aspartate-7, Serine-8 and Glycine-9 with metal ion was found at 40.40 Å. The residues connected to zinc ion in mode 6 were Glutamic-22 and Serine-26 at 24.31 Å. As seen, RMSD values were decreased fro m mode 7 to 9 means the interaction became mo re and more weak. The residues that bound to zinc ion were Glycine-9 and Tyrosine-10 in mode 7, Aspartate-23 and Asparagine-27 in mode 8 and for Glycine-38, Isoleucine-41 and Alanine-42 were in mode 9.
Thus, we considered the two modes with the lowest binding energy values but we proceed our study with mode 2 as the initial structure for MD simulations. The selected structure was similar to Miller et al. (2010) structure as the metal was attached to the same aspartate as the mode 2 conformat ion. The C-terminal do main of peptide is consisted of Glutamic acid-22 and Aspartic acid-23 whereas another seven potential residues including Aspartic acid-1, Glutamic acid-3, Histidine-6, Aspartic acid-7, Glutamic acid-11, Histidine-13 and Histidine-14 are part of the N-terminal domain. However, some experimental works reported that the Zn 2+ ion can also bind to the Nterminal region of the amyloid-β peptide in residues Histidine-6, Glutamic acid-11, Histidine-13, and Histidine-14 (Zirah et al., 2006; Gaggelli et al., 2008; Minico zzi et al., 2008). Besides that, the obligatory sequence that involves zinc b inding coordination has been mapped to 6-28 region of Aβ (Bush et al., 1994). But overall, all co mplexes demonstrated satisfying binding affinity toward amy lo id-β peptide.
4.1.2 Ligand Interactions, Ionizability and Hydrophobicity Analysis Further analysis was done by using the Discovery Studio 4.0 Visualizer (Studio and Insight, 2009) which had a tool to analyze the protein-ligand interactions automatically. Ligand and the hydrophobic interactions between the receptor, Aβ (1-42) and the ligand, Zn 2+ were analyzed to confirm our selected docked conformation. Table 2. The ligand interacti ons distance between Aβ(1-42) and Zn2+ ion Mode 1
2 3 4 5
Residue Name 2+
Distance in Å
Val 39 – Zn Ile 41 – Zn 2+ Ala 42 – Zn 2+ Asp 23 – Zn 2+ Asn 27 – Zn 2+ Asp 7 – Zn 2+ Glu 11 – Zn 2+
2.46 2.50 2.57 2.48, 2.48, 4.43 2.12 2.40, 2.51, 4.46 4.08
Glu 3 – Zn 2+ His 6 – Zn 2+ Arg 5 – Zn 2+ Asp 7 – Zn 2+ Ser 8 – Zn 2+
2.51 2.50 2.47 2.30 2.42, 2.43
46
6 7
8 9
Gly 9 – Zn 2+ Glu 22 – Zn 2+ Ser 26 – Zn 2+ Gly 9 – Zn 2+ Tyr 10 – Zn 2+ His 14 – Zn 2+ Asp 23 – Zn 2+ Asn 27 – Zn 2+ Gly 38 – Zn 2+ Ile 41 – Zn 2+ Ala 42 – Zn 2+
2.19 2.52, 3.12 2.35, 2.38 2.45 2.47 5.31 2.39, 2.52, 3.50 2.12 2.43 2.31 3.12
Fro m results summarized in Table 2, three ligand interactions were found between zinc ion and Aβ (1-42) in mode 1. In addition, two ligand interactions were formed in mode 2 where the interaction between Hydrogen atom (H) that connected to Nitrogen atom (N) in Asparagine-27 was strong due to its shorts distance with the length of 2.12 Å. The second interactions are between the oxygen atoms of Aspartate-23 with zinc ion with the length of 2.48 Å and 4.43Å as illustrated by Figure 15.
Figure 15. The ligand interactions of As p23 and Asn27 for mode 2
47
Figure 16 shows the ionizability of Aβ (1-42) docked to Zn2+ ion in the selected conformat ion. The procedure to arrive at the color of the surface vertices is analogous to the procedure described for the hydrogen bond donor or acceptor surface. Invented charges are assigned to heavy atoms covered by the surface which are also part of an amino acid residue. The charge value depends on the residue type. The positive values (Blue) correspond to basic residues for example the Arginine residues (NE, NH1 and NH2) are assigned a charge of +1/ 3 while Histidine residues (ND1 and NE2) are assigned a charge of +1/2. In addition, the Lysine residues, atom NZ is assigned a charge of +1. Other residues like Asparagine residues, the atoms OD1 and OD2 are assigned a charge of -1/2 and Glutamate residues, OE1 and OE2 are assigned a charge of -1/2 which the negative values (Red) correspond to acidic residues. The color of the surface vertices was derived from the classification of the atoms covered by the surface as either hydrogen bond donors or acceptors. The picture is a result of numerically mapped color spectrum.
Figure 16. Ionic calculation for Aβ(1-42) docked to zinc ion (mode 2)
For each atom covered by the calculated surface, the hydrophobicity of the residue according to the Kyte-Doolittle scale was used as the fictitious charge (Figure 17). The default color spectrum was Blue-White-Brown. The full color map range of values is 4.5 to +4.5 but, it was narrowed to -3.0 to +3.0 by default for imp roving the color contrast. The negative values (Blue) correspond to hydrophilic residues, whereas positive values (Brown) are denoted to hydrophobic ones. From Figure 17, the hydrophobicity value of the carboxy l groups was in between -1 to -2.
48
Figure 17. Hydrophobic surface calculati on for Aβ(1-42) docked Zn2+ ion for mode 2
4.2 Simulati on of Aβ (1-42) wi th and wi thout Zinc 4.2.1 Energetics of Model S ystem Table 3 reports the calculated results of average kinetic, potential and total energies of four Aβ (1-42) MD simulat ion in d ifferent conditions. Table 3. The potenti al, kinetic and total energy of Aβ (1-42) in the last 10 ns of all simulati ons Energy (kJ/ mo l)
Simu lations (1 μs) Potential
Kinetic
Total
Aβ in water
-371,608.00±0.00
66,701.10±0.00
-304,907.00±0.00
Aβ in HFIP
-127,468.00±0.00
55,005.90±0.00
-72,462.20±0.00
Aβ with zinc in water
-372,738.00±0.00
66,719.70±0.00
-306,019.00±0.00
Aβ with zinc in HFIP
-105,803.00±0.00
54,179.70±0.00
-51,623.70±0.00
49
As shown in Table 3, the average total energies for Aβ (1-42) in water and solvent mixture were -304,907.00±0.00 kJ/ mol and -72,462.20±0.00 kJ/ mol, respectively while for Aβ (1-42) with zinc in water and mixed solvent, the average total energies of 306,019.00±0.00 kJ/ mol and -51,623.70±0.00 kJ/ mo l were obtained. Both simulations in water produced the lowest average value of total energy where the average potential energy showed the main contribution in which the average potential values for Aβ (1-42) without and with zinc ion in water were -371,608.00±0.00 kJ/ mol and 372,738.00±0.00 kJ/ mo l, respectively. Additionally, the average potential energy in the solvent mixture for Aβ (1-42) simulat ions without and with zinc ion were 127,468.00±0.00 kJ/ mol and -105,803.00±0.00 kJ/ mol, respectively. As seen, the stability of Aβ(1-42) decreased in aqueous solution because water molecules could interact with the peptide polar groups and form hydrogen bonds (Levy et. al, 2001) which increased the motion of mo lecules as shown by its increased kinetic energy as well. The simulat ions with zinc ion also showed the same trends in which the Aβ (1-42) with zinc in water produced higher kinetic energy. Thus, the α-helix hydrogen bond formation might result in destabilizat ion and even unwinding of the entire Aβ (1-42).
The average volume of the model systems were also calculated for the last 10 ns of each simu lation. The average calculated volume of Aβ (1-42) in water was 275.58±0.00 nm3 but in the presence of zinc ion, the average volume Aβ (1-42) slightly increased to 275.69±0.00 n m3 which was almost similar. In addition, for the Aβ (1-42) in mixed solvent, the average volume was 302.20±0.00 n m3 but after inserting zinc ion, the average volume became 305.76±0.00 n m3 . It has been said that the coil structures in which the polar group are exposed to the solvent are likely to be mo re accessible in water than in organic mediu m (Levy et al., 2001). Therefore, when the number of conformat ion that peptide can adopt increasing, the entropy and the flexibility of the peptide is also become larger. The calculated flexib ility can be represented by the volume of the conformational space that a peptide can occupy in specific environment. As seen, a larger conformational space is available in aqueous solution compare in the mixtu re of water and HFIP. In water, the peptide can be found in many conformations as compared to mixed solvents environment.
4.2.2 Root Mean S quare Deviation (RMS D) The stability of obtained trajectories was checked by using the root mean square deviation (RMSD) o f the modeled structure relative to the reference structure as a function of time. The RMSD is calculated by the g_rms command in GROMACS which applies the least-square fitting of the structure to the reference structure ( =0). All gained conformations were compared with the reference structure which was the minimized structure in all model systems , according to the following equation;
(26)
50
where is the total mass and is the position of atom i at t ime t. The RMSD was calculated by the least-square fitting of the Cα-atoms in all structures. Figure 18 represents the RMSD fluctuation over time for Aβ (1-42) in four different conditions for 1 μs. Fro m the results, the most stable structure throughout the simu lation was found in the Aβ-H2 O model (Black) with an average RMSD value of 2.35±0.17 n m. At the beginning of simu lation, the average RMSD value of Aβ-HFIPH2 O model (Red) increased fro m 0.60±0.00 n m to 1.05±0.00 ns fro m 0 ns to 27 ns. After that, the deviation was slightly decreased from 28 ns to 199 ns. It was remained more or less stable between 200-286 ns with an average value of 1.07±0.14 n m. The deviation was then sharply increased from 287 ns to 432 ns. However, towards 734 ns, a slight decrease was observed but then was kept constant in the remaining time of simu lation with an average value of 1.38±0.02 n m. A steep increase in RMSD was found in the Aβ-H2 O-Zn model (Green) in the first 50 ns of simulation. It was slightly decreased from 51 ns to 164 ns. Then, the deviation continued to increase fro m 165 ns to 190 ns. The RM SD was then remained constant between 191-636 ns with an average value of 1.46±0.04 n m with a sudden increase fro m 637 ns to 673 ns. Later on, the fluctuation remained stable with an average value of 1.64±0.03 nm until the end of simu lation. The RMSD of Aβ -Zn-HFIP-H2 O model (Blue) slo wly increased fro m 0 ns to 145 ns. It was then fluctuating back and forth showing an increasing pattern while fro m 174 ns to 331 ns, the fluctuation decreased and then it changed significantly until 438 ns with no specific pattern. After an almost sharp increase towards 883 ns, the system became stable until the end of simulat ion with an average value of 0.90±0.04 n m. Our findings indicated the Aβ (1-42) produced more various conformations in the systems involving HFIP. According to Zimm-Bragg theory of the helix-coil equilibriu m (Zimm and Bragg, 1959) short polypeptides might not form helices in water. This phenomenon has also been proven by numerous studies in which the tendency of helix formation in water is low (So man et al., 1991; Buuren and Berendsen, 1993). Thus, fewer conformat ions will be produced in water than in organic solvent as shown by our results. Also, it can be suggested that the addition of metal could highly affect the peptide structure.
51
Figure 18. The RMS D fl uctuati on over ti me for Aβ(1-42) in four di fferent condi tions [Aβ -H2 O (black), Aβ-HFIP-H2 O (red), Aβ-Zn-H2 O (green), and Aβ-ZnHFIP-H2 O (bl ue)] for 1 μs
4.2.3 Radi us of Gyration (Rg ) The radius of gyration, Rg is a measure of protein co mpactness which is calculated by using the g_gyrate command in GROMACS. If the protein remains folded and stable, Rg will likely maintain a relatively steady value. But, if the protein loses its’ secondary structure and unfolds, the Rg value will change over time. The rad ius of gyration is calculated based on a group of atoms at the center of mass and the radii about x-, y- and z-axes. The atoms are explicitly mass weighted. To check the co mpactness of the peptide in different solutions and conditions, the radius of gyration (R g ) is calculated as follows; (27) where is the mass of atom, i and is the position of atom, i with respect to the center of mass of the molecule. In this study, gyration calculations were performed to investigate the compactness of amyloid-β peptide in d ifferent conditions. The fluctuation of Rg values over 1 μs simulation time for different model systems is shown in Figure 19. The Rg value for Aβ-H2 O model showed a slight increase in the first 100 ns. It then decreased from 101 ns to 200 ns and later it became constant until the end of simu lation with an average value of 2.51±0.25 n m. A d ifferent Rg variation of Aβ (1-42) was found when it placed in solvent mixture (Aβ-HFIP-H2 O). The radius decreased and then remained constant towards a minimu m of 1.11 n m at 157 ns.
52
Figure 19. Rg fluctuati ons over 1 μs simul ati on for all systems includi ng Aβ -H2 O (black), Aβ-Zn-H2 O (green) Aβ-HFIP (red), and Aβ-Zn-HFIP-H2 O (bl ue)
After a sudden increase of Rg value from 158 ns to 196 ns, it continued to decrease and remained more or less stable until 693 ns and kept constant until the end of simulation, with an average value 1.06±0.01 n m. In the presence of zinc ion, the Aβ-H2 O-Zn system overally showed a lo wer radius of gyration. The Rg value decreased fro m 0 ns to 42 ns, and later it increased and decreased sharply between 43 ns and 180 ns with a stable region in between with an average value of 1.03±0.07 n m. It was fluctuated back and forth between the average values of 0.93±0.01 n m to 0.92±0.02 n m with some specific stable regions from 738 ns to 745 ns. The radius of gyration slowly increased after that but it became stable again for the last 50 ns of simu lation with an average value of 0.92±0.02 n m. Overall, no significant increased or decreased pattern was found for Aβ-Zn-HFIP-H2 O model except for the first 200 ns of simulat ion with a sharp increase and decrease followed by a stable conformation for a while. Then, it decreased for a few seconds and had an obvious stable fluctuation with an average value of 1.19±0.04 n m for the remain ing time of simu lation. As seen, our RMSD and Rg results were consistent. The shorter distance between the atomic contacts can maintain the folding conformat ion and its stability (Huang et al., 2015). Figure 20 illustrates the most obvious changes in amy loid-β conformations highlighted by Rg analysis for different systems at different conditions.
53
(a)
(c)
(b)
(d)
Figure 20. The snapshots of conformational changes of amyloi d-β(1-42) system in di fferent condi tions obtained by Rg results over 1 μs simulation time for the l ast 10 ns. (a) Aβ(1-42) in water (b) Aβ(1-42) in mixed sol vents (c) Aβ(1-42) with zinc in water (d) Aβ(1-42) with zinc in mixed sol vents
54
4.2.4 Solvent Accessible Surface Area (SAS A) The min imization of solvent accessible surface area which is usually applied to evaluate the interaction of protein surface with solvent produces a powerful constraint for the prediction of the ultimate conformat ion of these macromo lecular assemblies (Chotia & Janin, 1975; Rich mond & Richard, 1978; Wodak & Janin, 1978; Chen et al., 1979, 1982). The SASA value of a solute is the surface of closest distance of the centre of solvent molecu les where both solute and solvent are rep resented by hard sphere as shown by Figure 21.
The SASA analysis was done by using g_sas in GROMACS to calcu late the hydrophobic, the hydrophilic and the total solvent accessible surface area. A group was defined for the surface calculation and another group for the output. The calculation group is always consisted of all non-solvent atoms in the system while the output group can be the whole or part of the calculation group. By default, periodic boundary conditions are taken in a way that the radius is set for the solvent probe. SASA analysis was carried out to explore the changes of peptide surface area in this study when it inserted into the solvent mixture models and in water in all M D simu lations.
Figure 21. The accessible surface of 3 overlappi ng spheres is obtained by adding the radius of the sol vent s phere to the van der Waals atomic radii. The vol ume enclosed by the accessible surface is the excluded volume with respect to the solvent sphere centre. The overlapping spheres are labeled as i, j and k (adapted from Ti mothy, 1984)
55
Table 4. A summary of Sol vent Accessible Surface Area (SASA) average values for Aβ in different conditi ons for 1 μs Simulati on Times (1 μs)
Hydrophobic
SASA (nm2 ) Hydrophilic
Total
Aβ in water
110.22±1.96
0.00
110.22±1.96
Aβ in HFIP
37.81±2.79
0.00
37.81±2.79
Aβ with zinc in water
32.97±1.81
0.00
32.97±1.81
Aβ wi th zinc in HFIP/ H2 O
36.84±2.07
0.00
36.84±2.07
The main purpose of SASA calculation was to determine the amount of hydrophilic, hydrophobic and total areas available to the solvent. It was found that 179 out of 409 atoms could be classified as hydrophobic and the presence of metal ion did not affect the SASA values significantly. Statistically, at 1 μs (Table 4) the amount of hydrophobic contents of Aβ(1-42) in water produced the highest value of both hydrophobic and total SASA with an average value of 110.22±1.96 n m2 followed by 37.81±2.79 n m2 in solvents’ mixture co mpared to the SASA average values of Aβ (1-42) models with and without zinc ion in different conditions with the average values of 32.97±1.81 n m2 and 36.84±2.07 n m2 , respectively. No SASA contents were found as hydrophilic content.
Figure 22. Sol vent Accessible Surface Area (SASA) fluctuation for hydroph obic content as a functi on of ti me for 1 μs
56
Our results might suggest that the existence of peptide in solvent mixture can help retain the helical structure and make the peptide to accumulate mo re rather than interact with the water mo lecules. This is in accordance with the RMSD and Rg results in which folded and stable conformat ions can be detected during the course of simu lation. The significant change of the SASA value in the first two model systems can originate fro m polar solvation effect which results fro m the native state in aqueous solution that will be characterized by larger entropy and consequently, more secondary structure changes. The dependence of Aβ (1-42) stability to the solvent characteristics, for examp le, the polarity and the dielectric constant of solvent is supported by several experimental evidence. The helical propensity of Aβ (1-42) in water shows a dramatic increase by the addition of certain alcohols like trifluoroethanol (Cammers -Goodwin et al., 1996; Rohl et al., 1996). It has been proposed that the alcohol can act on the exposed CO and NH groups by diminishing their exposure to the solvent, for examp le by shifting the conformat ional equilibriu m toward more co mpact structures, such as α -helical conformat ion (Vila et al., 2000). Another reason for observed results can be the helix– coil equilib riu m of Aβ (1-42) toward maintaining α-helical conformat ion resulting fro m the charged residues which has no effect in here as hydrophilic SASA values are zero (Marqusee et al., 1989; W illiam et al., 1998).
4.2.5 Root Mean S quare Fluctuation (RMS F) Flexib ility is an integral property to determine the substrate specificity of a particular biomo lecules (Koshland, 1959; Koshland, 1963).
To estimate the flexib ility of structure, root mean square fluctuation (RMSF) is usually used which is defined by the following equation; (28)
where T is the time taken for averaging and is the reference position of atom, i. The reference position is taken as the time-averaged position of the same atom. The averaging of fluctuation for calculating flexib ility is done over atoms by producing the time-specific difference to RM SD in which the average was taken over atoms to give the time -specific values. For further estimat ion of the flexib ility of protein , the RMSF values can be converted into theoretical B-factors, Bi (Hu nenberger et al., 1995) according to equation 28 and the region where the flexibility of protein affected by observing the fluctuation of each residue. Bi is fo rmulated as follows;
(29)
57
This definition is usually used to estimate the flexibility of residues among protein during folding or unfolding events where the crystal structure is available. As there is no report of Aβ (1-42) x-ray crystallized in water available yet, we skipped this analysis and just focused on RMSF calculations to gain a view of the flexibility of peptide upon zinc binding. Figure 23 represents the fluctuation of the Aβ (1-42) residues in water (Black) and in mixed solvent (Red). Fro m the results, Aβ-H2 O system showed several significant peaks compared to Aβ-HFIP-H2 O model. The fluctuation among peptide was decreased from 0.46 n m to 0.41 n m fro m residues 16 to 21. Then, a sudden increase fro m 0.92 n m to 1.98 n m at residues 27 to 28 was observed. Therefore, the β-sheet among peptide structure for Aβ-H2 O system appeared in an extended-β region with the most flexible residues including Lys16-A la21, Gly29-Val36 and Val39-Ile41. The bridge-β region (residues Tyr10 and Gly37) were also showed high flexib ility during simulation. In summary, residues 28-42 in water model system demonstrated a high flexibility (3.07±0.06 n m) fo r the last 10 ns of simulat ion producing an extended-β sheet. Additionally, a α-helix was found at His13-Lys16 with a low flexibility compared to other residues . The system showed an almost stable fluctuation during simu lation with no significant changes . This system overall produced a low flexib ility (0.17±0.08 n m) most probably due to the presence of a long α-helix with most obvious fluctuations and flexibility at residues His13-Lys16.
Figure 23. The fluctuations of Aβ (1-42) residues in water (bl ack) and mi xed sol vents (red) model systems over 1 μs
58
Figure 24 shows the RMSF results of the model systems in different conditions with zinc ion. The fluctuation of Aβ-Zn-H2 O system followed an almost similar pattern to Aβ-Zn-HFIP-H2 O model. The difference of Aβ-Zn-H2 O and Aβ-Zn-HFIP-H2 O model systems were at residues 3 to 5 which in water was slightly lower co mpared to mixed solvents. In addition, at residue 7 to 10 for Aβ-Zn-HFIP-H2 O model, flexibility slightly decreased compared to the same residues in water system. Another obvious change was at residues 13 to 15 in water with a higher flexib ility compared to solvent mixture. The flexib ility of residues 20 and 21 in mixed solvent was higher with the values of 0.26 nm and 0.22 n m while in water it was 0.09 n m for residue 20 and 0.12 n m for residue 21. The residues 22-27 in solvent mixture in total showed a higher flexibility compared to water system.
Also, the residues 28 to 30 in mixed solvents had higher fluctuations than in water. But, fro m residues 36 to 42 the difference in flexib ility was significantly higher in the mixtu re of solvents compared to water. Overall the flexib ility of Aβ (1-42) in solvent mixtu re was higher than the flexibility of Aβ (1-42) in water in the presence of metal ion. No α-helix was present in the Aβ-Zn-H2 O model system except β-sheet at residues Phe4-His16, Leu17-Ala21, Lys28-Gly 33 and Val39-Ile41 co mpared to Aβ-Zn-HFIPH2 O model which produced two α-helices at residues Val12-A la21 and Ser26-Met35. Hence, it can be seen that the zinc ion affects the Aβ (1-42) flexib ility in both models significantly co mpared to systems without Zn 2+ ion. Some experimental studies have stated that amyloid-β cannot dissolve in water and it is tended to aggregate. These NMR experiments were done on small amy loid-β frag ments in aqueous solution (Zhang et al., 2000; Yang et al., 2009).
The specific effect of the mixture of water and organic solvents particularly fluorinated alcohols, such as TFE and HFIP are still not completely explained (Crescenzi et al., 2002; Sticht et al., 1995). Water is a good proton acceptor and proton donor and it can easily form hydrogen bonds with Aβ (1-42) peptide. In addition, water has a s mall size, so it can easily break the intramolecular hydrogen bonds between Aβ(1-42) molecules. Therefore, the Aβ (1-42) molecules become more flexib le. On the other hand, HFIP has a poor ability to donate and accept proton, and it is considerably larger than water. As a result, the intramolecu lar hydrogen bond cannot be easily broken by adding HFIP to the system. This can explain the higher strictness of the Aβ (1-42) molecule in HFIP in comparison with water (Huang et al,, 2015).
59
Figure 24. The RMS F of Aβ(1-42) residues in different sol vents as a function of ti me over 1 μs
Our results are also consistent with the study performed by Yang et al. (2009) in which they could detect the higher flexib ility of the residues of Aβ (1-42) in water wh ile the peptide was less flexib le in organic solvent because of its long α-helix. The higher flexib ility of C-terminus implies that this region is very sensitive to the environment and it may play a vital ro le in the unfolding process of Aβ (1-42). MD simulat ion calculations and NOESY spectrum determination also showed that Aβ(1-42) peptide was more flexib le in water than in HFIP solution (Huang et al., 2015). The peptide become less flexible in the organic solvent environment and the extended conformat ions are more dominant. In contrast, the metal ion in Aβ-H2 O-Zn was still flexib le but its structure was mo re co mpact than the peptide without zinc ion.
4.2.6 Secondary Structure Anal ysis All structures of Aβ (1-42) at the end of each simu lation were characterized by applying the secondary structure analysis using the definition of secondary structure according to protein (DSSP) protocol (Kabsch & Sander, 1983) and the pictorial database PDBsum which provides summarized informat ion about each experimentally determined structural model in the Protein Data Bank (PDB) (Lasko wski, 2009). The VM D suite of program (Hu mphrey et al., 1996) was used for visualizations . Additionally, the interaction between single peptide with zinc was further visualized by using PyMOL software (Schrodinger, 2010) and Discovery Studio 4.0 software (Studio and Insight, 2009).
60
Table 5. Secondary Structure element changes of Aβ(1-42) in different condi tions during 1 μs MD simulation Simu lations Aβ in water Aβ in mixed solvent Aβ with Zinc in water Aβ with Zinc in mixed solvent
Element
Time (ns)
Alpha Helix
200 0
400 0
Beta Sheet Turns/Coils Alpha Helix Beta Sheet Turns/Coils Alpha Helix
14 28 4 6 32 0
13 29 0 12 30 0
Beta Sheet Turns/Coils Alpha Helix
11 31 17
12 30 6
Beta Sheet Turns/Coils
0 25
6 30
600 0 14 28 5 10 27 0 12 30 10 6 26
800 0 19 23 4 16 22 0 12 30 9 4 29
1000 0 17 25 4 22 16 0 17 25 9 6 27
Table 5 reports the amount of helical structure, β-sheet, turns and coils of Aβ (1-42) in different conditions over 1 μs. All obtained structures in different conditions are illustrated in Figure 25 and Figure 26. For the first model, fro m 200 ns until the end of simu lation, no α-helix structure was obtained in water only 17 residues changed into βsheet and 25 residues turn into coils and turns. Further analysis showed that, the structure of Aβ-H2 O model system contained three β-sheet and two bridged-β, with altogether formed 33.3% of the total structure excluding the α-helix. It is proven that in water, the Aβ (1-42) tends to produce more β-sheet structures and no helical structure. In an aqueous solution, the water molecu les with strong hydrogen bonding capabilit ies can be inserted between carbonyl, CO and amidic, NH groups of the peptide backbone and disrupt intra-peptide hydrogen bonds, thus the helical structures are being destabilized (Jalili and Akhavan, 2009). It has been reported that, Aβ (1-42) can adopt a turn-helical structure in solutions containing at least 50% HFIP in volume (To maselli et al., 2006; Jalili and Akhavan, 2009). We found a stable helix reg ion in Aβ-HFIP-H2 O system. Additionally, there was one α-helix (4 residues), three β-sheet (22 residues), four turns and three coils (16 residues) at the end of simulat ion. As seen, HFIP can locally enhance the backbone hydrogen bonding by removing the water mo lecules fro m the pro ximity of peptide through the formation of hydrogen bonds with backbone carbonyl groups even it is a poor proton acceptor and proton donor. However, it can not affect the hydrogen bond between CO and NH groups. As a result, the carbonyl groups are engaged in bifurcated hydrogen bonds with both HFIP and NH thus the helical conformation can still be stabilized (Buck, 1998).
61
The disruption of water mo lecular structure around the peptide can be as a result of hydrophobic association of TFE and also HFIP mo lecules with apolar side chains of peptide, which is an entropically favorable process (Walgers et al., 1998). The structure of Aβ (1-42) in organic co-solvent may reveal the underlying conformational changes tendencies of the peptide. The structures with organic solvent (i.e. HFIP or TFE), the central hydrophobic core (Leu17-A la21, Leu17-Phe20 and Val18-Phe20) can form a helix similar to what we observed in our model Aβ-HFIP (To maselli et al., 2006; Vivekanandan et al., 2011) suggesting that this region is the most favorable region for helix formation in the absence of organic co-solvents. In addition, the long N-terminal helix can be retained by changing fro m apolar to polar solutions but the shorter Cterminal helix may be lost.
Crescenzi and co-workers (2002) reported that their obtained second helix (residues 28 38) could be corresponding to the trans-membrane region of APP protein. The second helix fro m our simulat ion was detected at residues 26 -35 which is consistent with the experimental data reported. The addition of zinc ion in both conditions showed that the Aβ-Zn-H2 O system produced four β-sheet (17 residues), four turns and coils (25 residues) with no helical structure. The amount of α-helix in Aβ-Zn-HFIP-H2 O was about 21.4% including two helices from Val12 to Ala21 and from Ser26 to Met35. The N-terminal region alternated to turn and the helical structure while the C-terminus adopted a mostly turn like structure. However, a β -sheet within 6 residues appeared at the end of simu lation. The α-helix structure is usually replaced by π-helix, turn, and coil conformat ions which are intermediates for peptide unfolding (Luttmann and Fels, 2006).
These results are also consistent with the results of RM SD, Rg and RM SF analysis of our MD simulat ions in which the Aβ is more flexib le in water than in solvent mixture but in the presence of zinc ion, the structure of Aβ shows the reversible trend. No αhelix was detected in both Aβ-H2 O and Aβ-Zn-H2 O models. Both structures in water contained more β-sheet compared to Aβ (1-42) in solvent mixture. The insertion of the strong electronegative trifluoro methyl group can significantly increase the acidic nature of the carboxylic mostly in peptide. The fluorinated particles or micelles have a higher charge surface density than their hydrogenated analogs thus they can bind more strongly to peptide molecule. If the α-helix is being stable enough, the amyloid fibrils will not be formed.
This means that the fibrils are formed only in the conditions with the lower energy barrier for helix formation without significant increase in the energy for the conversion to β-sheet (Rocha et al., 2012). The creation of a deep energy well for the α-helix can prevent the fibril fo rmation (Kirkitadze et al., 2001).
62
Figure 25. The visualization of secondary structures of (a) Aβ -H2 O, (b) Aβ-HFIPH2 O, (c) Aβ-Zn-H 2 O and (d) Aβ -Zn-HFIP-H2 O model systems at the end of 1 μs MD Si mulation; α-helix (purple); 3 10 helix (red); extended-β (yellow), bri dge-β (pink); coil (sil ver) and turn (green)
(a)
(b)
63
(c)
(d)
Figure 26. The time expansion of secondary structure elements for respecti ve residues in different conditions; a) Aβ-H2 O, (b) Aβ-HFIP-H2 O, (c) Aβ-Zn-H2 O and (d) Aβ-Zn-HFIP-H2 O model systems (1 frame = 1 ns) of 1 μs 4.3 Amyl oi d-β Aggregati on Process Simulati on The pathology of the protein and peptide aggregation are considered as key events in a number of chronic and devastating neurodegenerative conditions including dementias such as AD. More than half of the 25 recognized neurodegenerative diseases are associated with the specific protein aggregates.
The causes and the effects of relating the protein aggregation and the degenerative central nervous system (CNS) d iseases to amyloid-β aggregation are still unknown but the aggregated proteins are diagnostically specific (Ross and Poirier, 2004) and the association of disease with the aggregation has obvious diagnostic and therapeutic implications (Pedersen and Heegaard, 2013). Therefore, shading some light on the aggregation process of this peptide can be helpful to prevent and/or the treatment of these diseases, for examp le, by inhibit ing Aβ (1-42) aggregation (McKoy et al., 2012). The Aβ (1-42) aggregation has not been fully understood. However, several co mpeting pathways suggested which could happen simu ltaneously. Several studies imp ly that the simu lation of not necessarily full-length amy loid proteins can be very useful for understanding the amylo id format ion (Osborne et al., 2014).
The presence of amy loid aggregates can be used to define the entire class of amylo id related diseases. Because the production of amyloid is so often observed during disease progression and any suspected toxic species could be present under the physiologica l conditions that produce amyloid. To accu mulate under the disease-relevant amy loidogenic conditions, the aggregates must be cytotoxic and stable enough.
64
Experimental tests of the amyloid cascade and toxic o ligo mer hypotheses mainly evaluate the first requirement of toxicity. There are two hypotheses for this idea that require d ifferent free -energy landscape (He et al., 2010). The amy loid cascade hypothesis states that the accumulation of autocatalytic amyloid fibrils and plaques are causing the deficits in amy loid-related disease (Hardy and Higgins, 1992).
First, the reservoir must fill up until the nucleating oligomers are formed and allow the creation of seeding oligo mers. Next, the reservoir o ligo mers must accumu late in sufficient numbers to support elongation of protofibrils. Therefore, two classes of oligo mers are relevant to disease progression under the amyloid cascade hypothesis which is the seeding oligo mers and reservoir oligo mers. The rate of amylo id format ion often increases the rate of disease progression by increasing the rate of amyloid formation because of the single amino acid mutations (Conway et al., 1998; Narh i et al., 1999; Greenbaum et al., 2005).
4.3.1 Energetics of Aggregati on Process An increase in cross-β-propensity would stabilize the amylo id product but it would not increase the rate of amylo id format ion unless the transition state at the nucleation step resembles cross-β. This idea can explain the inconsistency in the secondary structure prediction of amy loid-prone sequences. A net kinetic effect will appear only if the nucleation barrier is reduced or if the population of the reservoir oligomers is increased. Table 6. The potenti al, kinetic and total energy of Aβ (1-42) averaged over the last 10 ns in different sol vent Simu lati ons
Energy (kJ/ mo l) Potential
Kinetic
Total
Water
552,011.00±64.0
96,977.50± 1.1
455,034.00±6.4
Mixed solvent
224,651.00±84.0
80,934.80± 1.0
143,716.00±85.0
Table 6 summarizes the gained energies for the aggregation of six molecules of Aβ (1-42) with zinc in different conditions for the last 10 ns of simulat ion. The average kinetic energy of peptide 6Aβ-6Zn-H2 O system was higher (96,977.50±1.10 kJ/ mol) co mpared to six 6Aβ-6Zn-HFIP-H2 O model with an average value of 80,934.80±1.00 kJ/ mol. The peptide molecules in water can produce strong interactions with water mo lecules compared to the solvent mixtu re wh ich highly affects the mobility of peptide in the system. This can also affect the structural changes from α-helix to β-sheet during aggregation.
65
Additionally, the large positive entropy change that can drive protein aggregation reactions is usually ascribed to the release of bound water molecules when the protein mo lecules aggregate (Blank, 1994). Besides that, the average potential energ ies showed that in water was -552,011.00±64.0 kJ/ mo l lo west values compared to in solvent mixtu re (-224,651.00±84.0 kJ/ mol) and make the total energy in water became 455,034.00±6.4 while in mixed solvent was -143,716.00±85.0 kJ/ mol.
4.3.2 Root Mean S quare Deviation (RMS D) We only used different solvents (water and HFIP) as our manipulat ive parameters because solvent can largely affect the secondary structure of Aβ (1-42). Several studies reported that in the presence of zinc ion, the peptide tends to aggregate (Miller et al., 2010; Bo lognin et al., 2011; Liu et al., 2011; Rezaei-Ghaleh et al., 2011). In order to determine the stability of each simulat ion, we utilized root mean square deviations of the Cα carbon at each time frame. RMSD calculations are a preliminary step to determine whether any significant conformat ional changes occur in the model system. Figure 27 presents the RMSD calcu lations results of our aggregation study involving zinc ions. The average RMSD value fo r 6Aβ-6Zn-H2O system was 1.02±0.01 n m which is almost similar to another simulation in solvent mixture (6Aβ-6Zn-HFIP-H2 O) with an average value of 1.05±0.01 n m for the last 10 ns of simulat ion. At the beginning of the simulat ion (10 ns), the RMSD of 6Aβ-6Zn-H2 O model was increased sharply from 0 n m to 0.77 n m. Then, it fluctuated almost stable from 11 ns to 57 ns with an average value of 0.94±0.03 n m and after that, it decreased. Then, it remained constant fro m 60 ns onwards until the end of the simulat ion time. A sharp increase was detected for 6Aβ-6Zn-HFIP-H2 O until 11 ns. The fluctuations followed a p lateau between 11 ns and 30 ns with an average value of 0.81±0.03 n m and then slightly increased and remained stable fro m 31 ns to 45 ns (0.90±0.04 n m). There was a significant increase and decrease for another 8 ns but then it remained stable until the end of simulat ion.
66
Figure 27. The RMSD fluctuations during aggregation over 100 ns simulation time for six molecules of Aβ (1-42) with zinc ion in different conditi ons; 6Aβ-6 ZnH2 O (green) and 6 Aβ-6 Zn-HFIP-H2 O (purple)
4.3.3 Radi us of Gyration (Rg ) As discussed in previous section, increasing of gyration radii indicates an expanded conformat ion and decreasing of gyration radii may show some shrinking conformat ions. The average value of Rg in the 6Aβ-6Zn-HFIP-H2 O model was higher than 6Aβ-6Zn-H2 O for the last 10 ns of simulat ion with the values of 2.03±0.03 n m and 1.81±0.04 n m, respectively.
Thus, the less compactness of peptide in water during aggregation helps explain its aggregation process. From Figure 28, the Rg of 6Aβ-6Zn-HFIP-H2 O was increased at first 5 ns with detectable changes from 6-20 ns. The Rg fluctuations were then more or less constant (with moderate changes between 63-80 ns) until the end of simulation. The 6Aβ-6Zn-H2 O system showed the same pattern but lower Rg fluctuations. Thus, the system overall kept its folded structure even after aggregation.
67
Figure 28. The Rg fluctuations during aggregation over 100 ns simulati on ti me for six molecules of Aβ (1-42) with zinc ion in di fferent condi tions; 6Aβ-6 Zn-H2 O (green) and 6Aβ -6Zn-HFIP-H2 O (purple)
Figure 29 illustrates the snapshots of both models at the beginning and at the end of both simulat ions. The Rg fluctuations were not consistent with the RMSD deviations results, mainly as a result of secondary structure changes which later will be discussed .
(a)
(b)
68
(c)
(d)
Figure 29. The picture illustrates the changes of compactness of amyloi d-β before and after aggregation in different condi tions. The snapshots were taken at (a) 0 ns and (b) 100 ns for amyl oi d-β in water, while (c) 0 ns and (d) 100 ns are amyloi d-β in sol vent mixture.
4.3.4 Solvent Accessible Surface Area (SAS A) The hydrophobic effect is responsible for many observed properties such as the stability of cell memb ranes during protein folding as well as the insertion of membrane proteins into the nonpolar lipid environ ment and stabilizing small scale interactions among protein mo lecules (Kauzmann, 1959; Charton and Charton, 1982; Lockett et al., 2013; Breiten et al., 2013). The significant changes in conformat ion usually originate fro m intermolecu lar and intramolecular interactions. Table 7: A summary SASA analysis for 6 molecules of Aβ (1-42) with zinc during aggregation in di fferent condi tions for the last 10 ns of both MD simul ations
Simu lations (100 ns)
6 mo lecules of Aβ (1-42) with zinc in water
6 mo lecules of Aβ (1-42) with zinc in mixed solvent
Hydrophobic (nm2 )
79.95±0.69
83.84±0.66
60.80±0.83
60.55±0.76
140.75±1.14
144.39±1.08
2
Hydrophilic (n m ) 2
Total (n m )
The average solvent accessible surface area can be used to determine the magnitude of binding-induced conformat ional changes from the structures of either monomeric proteins or bound sub-units. Applying this calculation to a large set of protein complexes has suggested that large conformational changes upon binding are common. In addition, a considerable enrich ment of intrinsically disordered sequences in proteins can be related to large conformat ional changes (Marsh and Teichmann, 2011).
69
Table 7 reports the SASA results for the last 10 ns of Aβ aggregation simulations in different solutions involving zinc ion. The hydrophobicity of 6Aβ-6Zn-HFIP-H2 O system was much higher with an average SASA value of 83.84±0.66 n m2 compared to 6Aβ-6Zn-H2 O model (79.95±0.69 n m2 ). This trend was the same fo r the average total SASA values in solvent mixture (144.39±1.08 n m2 ) than in water (140.75±1.14 n m2 ) while the hydrophilic SASA were similar. Lo wer value of SASA can be as a result of increasing the ionic strength of solution indicating Aβ (1-42) can interact more with the solvent. From Figure 30, the total surface area in water (yello w) was decreased sharply in the first 3 ns of the simulat ion. Then, it remained more or less stable except a few spikes between 12-23 ns and 24-84 ns before it became stable until the end of simu lation.
The hydrophobic and hydrophilic and the total SASA fluctuations of model aggregation in water showed almost similar patterns with its total SASA. In contrast, the total SASA fluctuation in solvent mixture (brown) produced higher average values. There was a slightly sharp decrease at the beginning of simu lation but it then increased and remained constant with an obvious spike around 50 ns . Later on, it remained constant until the end of simu lation. The similar patterns were observed for the hydrophobic and hydrophilic SASA fluctuations of model aggregate in the mixture of solvents.
Figure 30. The sol vent accessible surface changes of 6 aggregated Aβ(1-42) pepti des in different sol vents over 100 ns simul ation time.
70
The larger SASA represents a greater solvent exposure of residues which indicates fewer intra-peptide interactions and greater affin ity of the peptide for solvent mo lecules. More solvent exposure towards the peptide can disrupt the β -strand formation by changing the backbone of hydrogen bonding if not favors the β -strand formation and side interactions that increase to a hydrophobic nucleus that is shielded more fro m solvent (Brown et al., 2014). The energetics of protein reactions can be evaluated in terms of the changes in the surface free energy of the protein-water interface that can be estimated fro m the changes in the area and the charge density of the protein-water interface. Aggregation is energetically equivalent to the loss of protein-water interface (Blank, 1994).
Furthermore, there are some evidences that state more flexib le proteins tend to undergo larger conformational changes upon binding (Dobbins et al., 2008). Ho wever, there is a slight difference between the predicted flexibility of protein binding sites which undergo large conformat ional changes upon ligand binding and some proteins may not show any changes (Gunasekaran and Nussinov, 2007).
4.3.5 Root Mean S quare Fluctuation (RMS F) Figure 31 shows RMSF calcu lations results of Aβ (1-42) involving zinc ion in water during aggregation for different conformat ions. The similar trend with the single peptide with zinc in different solvents was observed here where the system in water produced low flexib ility than in solvent mixture. In addition, Aβ (1-42) with zinc in water produced less helical structure compared to Aβ (1-42) with zinc in mixed solvent.
Figure 31. The RMS fluctuations of aggregation of six molecules of Aβ (1-42) in water wi th zi nc ions for 1 μs
71
All obtained conformat ions produced β-sheet, β-bridge, turn and coil structures except structure C, D and F wh ich only showed α-helix, coils and turns structures as seen in Figure 32. Transitions from α-helical to β-sheet structures have been experimentally observed after oligo mer format ion (Kirkitadze et al., 2001) suggesting that a critical mass of chains is needed before ordered aggregates form.
Figure 32. The snapshot of each Aβ(1-42) pepti de in water during aggregati on for the last 10 ns of simulati on; α-helix (purple), β-sheet (yellow), β-bri dge (orange), turn (cyan) and coil (silver)
72
The flexib ility of model system during aggregation in the mixture of solvents was similar to the model system in water (Figure 33). In mixed solvent, C, D, E and F conformat ion produced two helical structures with one β -bridge, turns and coils structure except structure D, where the β-bridge was absent. Structure A and B only had one α-helix each in addition to turns and coils structure. In structure A, one π-helix (Leu34 to Gly 39) was found and there were two β-bridges (Leu 34 and Val 39) in structure B at the end of simu lation as shown in Figure 34. In fact, the helical structures were supposed to retain at residues 10-24 but due to the HFIP effect, the helical structures were retained significantly in C, D, E and F conformations with more helical elements than in water. These four structures also contained another helical structure at the end of C-terminal fro m residue 32 until 40.
Figure 33. The RMS F calcul ati ons results of aggregation in mi xed-sol vent for six molecules of Aβ(1-42) with zinc ions for 100 ns.
73
Figure 34. The snapshot of each pepti de of Aβ (1-42) in solvent mixture during aggregation for the last 10 ns of simulation; α-helix (purple), β-sheet (yellow), βbri dge (orange), turn (cyan) and coil (sil ver)
74
4.3.6 Secondary Structure Anal ysis An important indicator of the potential of amy lo id-β aggregation is secondary structure. The initial α-helical conformat ion of the peptides can change to several secondary structures like β-sheet, turn and coil during MD simulat ion. The obtained results indicated that solvent could affect the secondary structure of Aβ (1-42). The secondary structure in water slowly changed from α-helix to π-helix and β-bridge around 20 ns. One of six peptide structures started to alter to β-sheet around 60 ns while other structures were changed to β-bridge (Figure 35). At 80 ns, more β-sheet was formed and the π-helix started to disappear. Therefore the peptides in water tend to aggregate faster due to the residues were more flexib le.
75
Figure 35. Snapshot pictures of Aβ(1-42) with zinc during aggregation in water in every 20 ns over 100 ns simulation ti me
76
Table 8. Secondary Structure element changes of Aβ (1-42) during aggregation invol ving zinc i on i n different conditi ons during 100 ns MD simulation in Water (ns) Element
in Mixed-So lvent (ns)
20
40
60
80
100
20
40
60
80
100
Alpha Helix
90
82
87
77
80
115
98
91
90
96
Beta Sheet
0
0
12
12
6
6
4
4
4
4
Turns/Coils
162
170
153
163
166
127
147
157
158
152
At the beginning of simu lation, the packed system of peptides was co mpact involving a hydrophobic core but later the percentage of β-sheets was also changed. From Table 8, 6Aβ-6Zn-HFIP-H2 O model produced more α-helices (96 residues) while in water, only 80 residues remained as α-helix at the end of simu lation. No β-sheet was observed at the beginning of simulation until 60 ns (12 residues) and then it turned to 6 residues at the end of the simulation for aggregation in water. In addition, the peptides that dissolved in mixed solvent produced less amount of β-sheet with only 4 residues that almost preserved during whole simu lation time.
Other remaining residues, 166 residues in water and 152 residues in solvent mixture were changed to coils and turns by the end of aggregation simulation. Generally, the 6Aβ-6Zn-HFIP-H2 O model showed the secondary structure changes more significantly compared to 6Aβ-6Zn-H2 O system. Figure 36 demonstrates the time evolution of secondary structure for six mo lecules of Aβ (1-42) in the presence of zinc ion during aggregation in different conditions into more or less twisted helices, 310 and π-helix, respectively. The secondary structure codes are represent α-helix, H (purple), π-helix, I (red), 3-10 helix, G (blue), extended conformation, E (yellow), Isolated bridge, B (dark yellow), turn, T (green) and coil (white).
77
Figure 36. The changes of secondary structure for six molecules of Aβ (1-42) during aggregation wi th zinc in di fferent condi tions; i n water (left), in sol vent mixture (right) with the 1000 frames over 100 ns simulation ti me
According to Figure 37 two chains among all peptides produced 3 10 -helix (blue) while other structure maintained their α-helix. The same peptides were changed back into initial structure while the other three structures altered to π -helix (red) around 20 ns . Most peptides maintained their helical structure in the solvent mixture because β -sheet can hardly produced in HFIP. Several studies have stated that amyloid -β proceeds with many intermed iates as it transfers from helix to β-strand in the aggregation of mono meric Aβ to fibrillar Aβ. It has also been proposed that transient intermediates during the aggregation process contain a considerable amount of α-helix structure (up to 32%) wh ich may convert into β-sheet depending on solution conditions (Kirkitadze et al., 2001; Fezoui and Teplow, 2002; To masselli et al., 2006). Monomeric Aβ (1-40) and Aβ (1-42) are intrinsically unstructured, implying that in solution they do not assume to have any compact tertiary fold but rather they populate to a large number of different conformations. As such, they cannot be crystallized . Most of structural knowledge on Aβ peptides has been derived from NM R experiments and mo lecular dynamics (MD) studies (Valensin et al., 2012). Additionally, some reports have shown that the smaller frag ments or derivatives of Aβ (with oxidized Met35) may be realistically described as random coils with only a small percentage of local nonrandom structures in aqueous solution (Jarvet et al., 2000; Riek et al., 2001; Hou et al., 2004; Lazo et al., 2005). Our results are mostly consistent with studies in which the peptide could alter the structure from α-helix to β-sheet, turn and coil in water. In solvent mixture, most researchers showed the helical structures with no obvious formation of β -sheet.
78
Figure 37. Snapshot pictures of Aβ(1-42) with zinc during aggregation in solvent mixture for every 20 ns over 100 ns simul ation time
79
CHAPTER 5
CONCLUS ION
5.1 Summary
In this study, we investigated the behavior of Aβ (1-42) single peptide in different solvents; in water and solvent mixture containing hexafluoroisopropanol, using MD simu lations. Successfully simu lated the dynamics, flexibility as well as the structural changes of both model systems interacting with zinc ions in different solvents , to understand the mechanism of Aβ aggregation. The simulat ion time for each model system was 1 μs, respectively. The aggregation process took 100 ns only because the presence of zinc ion had induced the early fo rmation of aggregates. The initial structure for MD simulat ion study was prepared using molecular docking process, which predicted a single docked conformation of Aβ (1-42)-Zn with the binding energy of -0.9 kcal/ mo l.
The docking conformation consisted of two ligand interactions between zinc ion and the oxygen atom of Aspartic acid-23, and the nitrogen atom of Asparagine-27, with the length of 2.48 Å, 4.43 Å and 2.12 Å, respectively. Our Rg results were consistent with the RMSD analysis, wh ich reported a shorter distance between the atomic contacts of peptide therefore the folding conformat ion of peptide and its stability were maintained. Statistically, the amount of hydrophobic contents in water produced the highest SASA value (110.22±1.96 n m2 ), fo llo wed by SASA in solvent mixture is 37.81±2.79 n m2 . In contrast, the SASA of Aβ (1-42) with zinc was slightly higher in mixed solvent with a value of 36.84±2.07 n m2 compared Aβ-Zn-H 2 O, wh ich was 32.97±1.81 n m2 .
The MD results revealed that the system in water was more flexib le than the model in the solvent mixture because of the length of helical structure. The Aβ-Zn-H2 O system was also considered flexib le but its structure was compact compared to the peptide without zinc. Fro m our secondary structure analysis, a collapsed-coil structure of Aβ(142) was produced in aqueous solution. In addition, h igher amount of β-sheet structures were found. The peptide in the mixed solvent system was more stable. The presence of helical structure was evident based on the format ion of central hydrophobic core and the close contacts of N- and C- termini with the central helix. In the aggregation studies, the average RMSD value for 6Aβ-6Zn-H2O system was 1.02±0.01 n m in the last 10 ns of simulation, similar to the system in solvent mixture (6Aβ-6Zn-HFIP-H2 O) wh ich reported an average value of 1.05±0.01 n m. Ho wever, the average value of Rg in the 6Aβ-6Zn -HFIP-H2 O model was higher than 6Aβ-6Zn-H2 O in the last 10 ns of simulat ion, with the values of 2.03±0.03 n m and 1.81±0.04 n m, respectively.
80
Thus, the greater compactness of peptide in water helps exp lain its aggregation process. The hydrophobicity of 6Aβ-6Zn-HFIP-H2 O system was higher with an average SASA value of 83.84±0.66 n m2 compared to 6Aβ-6Zn-H2 O model (79.95±0.69 n m2 ). In contrast, the total SASA fluctuation in mixed solvent gives higher average results. The lower value of SASA could be the result of increasing ionic strength in solution, which indicated the favorable interaction between Aβ(1-42) and the solvent. In the RMSF calculat ion, the Aβ (1-42) peptide with zinc produced lower flexib ility in water than in solvent mixture. There is also a similar trend in the secondary structure analysis. In the presence of zinc ion, the peptide in water produced less helical structures. The 6Aβ-6Zn-HFIP-H2 O model produced 96 α-helices. Ho wever, only 80 residues were remained in the 6Aβ-6Zn -H2 O system at the end of simulat ion, because β-sheets were barely produced in HFIP.
5.2 Recommendati on for Future Studies
1. The effect of different metal concentrations on Aβ (1-42) and its aggregation process at different pH should be highlighted. 2. M D simu lations at different temperatures and pressures should also be investigated due to the difference in climate and at mosphere around the world. The vary ing conditions could affect the human brain and subsequently the progress of Alzheimer disease. 3. The simulation time should be extended. Other simulat ion methods such as CG-M D can also be used, especially for aggregation study with more than six mo lecules, to gain more details of the mechanis m. 4. The use of other metals such as Fe2+ or Cu 2+ is highly suggested, to see the comparison each metal towards the effect of AD.
81
REFERENCES Adams, D. J., Adams, E. M ., & Hills, G. J. (1979). The computer simu lation of po lar liquids. Molecular Physics, 38(2), 387-400. Adams, J. D., Jr., Chang, M. L., & Klaid man, L. (2001). Parkinson's Disease-Redox Mechanisms. Current Medicinal Chemistry Journal, 8(7), 809-814. Alder, B. J., & Wainwright, T. E. (1957). Phase Transition for A Hard Sphere System. The Journal of Chemical Physics, 27(5), 1208-1209. Alder, B. J., & Wainwright, T. E. (1959). Studies in Molecular Dynamics. I. General Method. The Journal of Chemical Physics, 31(2), 459-466. Allen, M. P. (2004). Introduction to Molecular Dynamics Simu lation. Computational Soft Matter: From Synthetic Polymers to Proteins, 23, 1-28. Amadoro, G., Ciotti, M. T., Costanzi, M., Cestari, V., Calissano, P., & Canu, N. (2006). NM DA Receptor Mediates Tau-induced Neurotoxicity by Calpain and ERK/MAPK Activation. Proceeding of the National Academy of Science of the U S A., 103(8), 2892-2897. Epub 2006 Feb 2813. Ascher, D., & Lutz, M. (1999). Learning Python: O'Reilly. Bajda M, Filipek S (2015) Study of early stages of amy loid Aβ13-23 fo rmation using mo lecular dynamics simu lation in implicit environments. Computational Biology and Chemistry 56 (0):13-18. Bao, F., Wicklund, L., Lacor, P. N., Klein, W. L., No rdberg, A., & Marutle, A. (2012). Different β-amyloid o ligo mer assemblies in Alzheimer brains correlate with age of disease onset and impaired cholinergic activ ity. Neurobiology of Aging, 33(4), 825.e821-825.e813. Bau mketner, A., & Shea, J. E. (2007). The Structure of the Alzheimer A mylo id β 10 -35 Peptide Probed through Replica-Exchange Molecular Dynamics Simu lations in Exp licit Solvent. Journal of Molecular Biology, 366(1), 275-285. Bau mketner, A., Bernstein, S. L, Wyttenbach, T., Bitan, G., Teplow, D. B., Bowers, M. T., & Shea, J. E. (2006). A myloid β-Protein Monomer Structure: A Co mputational and Experimental Study. Protein Science, 15(3), 420-428. Beck, D. A., & Daggett, V. (2004). Methods for Molecular Dynamics Simulat ions of Protein Fo lding/Unfold ing in So lution. Methods, 34(1), 112-120. Bekker, H., Dijkstra, E. J., Renardus, M. K. R., & Berendsen, H. J. C. (1995). An Efficient, Bo x Shape Independent Non-Bonded Force and Virial Algorith m for Molecular Dynamics. Molecular Simulation, 14(3), 137-151.
82
Berendsen, H. J. C., van der Spoel, D., & van Drunen, R. (1995). GROMACS: A Message-Passing Parallel Molecular Dynamics Implementation. Computer Physics Communications, 91(1), 43-56. Bhattacharjya, S., Venkatraman, J., Balaram, P., & Ku mar, A. (1999). Fluoroalcohols as Structure Modifiers in Peptides and Proteins: Hexafluoroacetone Hydrate Stabilizes a Helical Conformat ion of Melittin at Lo w p H. The Journal of Peptide Research, 54(2), 100-111. Bitan, G., Kirkitad ze, M. D., Lo makin, A., Vo llers, S. S., Benedek, G. B., & Teplo w, D. B. (2003). Amy loid Beta-Protein (Abeta) Assembly: Abeta 40 and Abeta 42 Oligomerize Through Distinct Pathways. Proceedings of the National Academy of Science of the United States of America, 100(1), 330-335. Epub 2002 Dec 2027. Bolognin, S., Messori, L., Drago, De., Gabbiani, C., Cendron, L., & Zatta, P. (2011). Aluminu m, Copper, Iron and Zinc Differentially Alter A mylo id -Aβ1–42 Aggregation and Toxicity. The International Journal of Biochemistry & Cell Biology, 43(6), 877-885. Bray mer J. J.; Choi J.-S.; DeTo ma A. S.; Wang C.; Nam K.; Kamp f J. W .; Ramamoorthy A.; Lim M. H. (2011) Develop ment of Bifunctional Stilbene Derivatives for Targeting and Modulating Metal-Amy loid-β Species. Inorganic Chemistry. 50, 10724– 10734. Brooks, C. L., & Nilsson, L. (1993). Pro motion of Helix Formation in Peptides Dissolved in Alcohol and Water-Alcohol Mixtures. Journal of the American Chemical Society, 115(23), 11034-11035. Bro wn, A. M., Lemku l, J. A., Schau m, N., & Bevan, D. R. (2014). Simulat ions of Monomeric A my loid β-Peptide (1–40) with Vary ing Solution Conditions and Oxidation State of Met35: Imp lications for Aggregation. Archives of Biochemistry and Biophysics, 545(0), 44-52. Bukau, B., Weissman, J., & Horwich, A. (2006). Molecular Chaperones and Protein Quality Control. Cell, 125(3), 443-451. Bush, A. I. (2003). The Metallobio logy of Alzheimer's Disease. Trends in Neurosciences, 26(4), 207-214. Bush, A. I., Huang, X., & Fairlie, D. P. (1999). The Possible Origin of Free Radicals fro m A my loid ß Peptides in A lzheimer’s Disease. Neurobiol Aging, 20, 335-7. Bush, A. I., Pettingell, W. H., Multhaup, G. D. P. M., d Paradis, M., Vonsattel, J. P., Gusella, J. F., … & Tanzi, R. E. (1994). Rapid Induction of Alzheimer A Beta Amylo id Formation by Zinc. Science, 265(5177), 1464-1467. Campbell, A., Smith, M. A., Sayre, L. M., Bondy, S. C., & Perry, G. (2001). Mechanisms by which Metals Pro mote Events Connected to Neurodegenerative Diseases. Brain Research Bulletin, 55(2), 125-132.
83
Case, D. A., Darden, T., Cheatham III, T. E., Simmerling, C., Wang, J., Duke, R. E., . . . Crowley, M. (2006). AMBER 9. University of California, San Francisco, 45. Chang, C.-e. A., Chen, W., & Gilson, M. K. (2007). Ligand Configurational Entropy and Protein Binding. Proceedings of the National Academy of Sciences, 104(5), 1534-1539. Chen, A. K. H., Lin, R. Y. Y., Hsieh, E. Z. J., Tu, P. H., Chen, R. P. Y., Liao, T. Y., Chen, W. L., Wang, C. H., & Huang, J. J. T. (2010). Induction of Amyloid Fibrils by the C-Terminal Frag ments of TDP-43 in A myotrophic Lateral Sclerosis. Journal of the American Chemical Society, 132(4), 1186-1187 Chit i, F., & Dobson, C.M. (2006). Protein Misfold ing, Functional A my loid, and Hu man Disease. Annual Review of Biochemistry, 75, 333-366. Cobb, J. L., Wolf, P. A., Au, R., White, R., & D'agostino, R. B. (1995). The Effect of Education on the Incidence of Dementia and Alzheimer's disease in the Framingham Study. Neurology, 45(9), 1707-1712. Coles, M., Bicknell, W., Watson, A. A., Fairlie, D. P., & Craik, D. J. (1998). Solut ion Structure of Amyloid β-Peptide(1−40) in a Water−Micelle Environ ment. Is the Membrane-Spanning Do main Where We Think It Is? Biochemistry, 37(31), 11064-11077. Conway, K. A., Harper, J. D. & Lansbury, P. T. (1998). Accelerated in vitro fibril formation by a mutant alpha-synuclein lin ked to early-onset Parkinson disease.Nat. Med. 4, 1318–1320. Coughlan, C. M., & Breen, K. C. (2000). Factors Influencing the Processing and Function of the Amyloid β Precursor Protein-A Potential Therapeutic Target in Alzheimer's Disease? Pharmacology & Therapeutics, 86(2), 111-144. Crescenzi, O., To maselli, S., Guerrini, R., Salvadori, S., D'Ursi, A. M., Temussi, P. A., & Picone, D. (2002). So lution Structure of the Alzheimer A mylo id Beta Peptide (1-42) in an Apolar Microenviron ment. Similarity with a Virus Fusion Do main. European Journal of Biochemistry, 269(22), 5642-5648. Crouch, P. J., & Barnham, K. J. (2012). Therapeutic Redistribution of Metal Ions to Treat Alzheimer's Disease. Accounts of Chemical Research, 45(9), 1604-1611. Epub 2012 Jun 1629. Cuajungco, M. P, Frederickson, C. J., & Bush, A. I. (2005). A myloid-β Metal Interaction and Metal Chelation. In Alzheimer’s Disease, (pp. 235-254). Springer US. Cuajungco, M. P., & Fagét, K. Y. (2003). Zinc takes the center stage: its paradoxical role in Alzheimer’s disease. Brain Research Reviews, 41(1), 44-56. Curtain, C. C., Ali, F., Vo litakis, I., Cherny, R. A., Norton, R. S., Beyreuther, K., . . . Barnham, K. J. (2001). Alzheimer's Disease Amyloid-Beta Binds Copper and
84
Zinc to Generate an Allosterically Ordered Membrane-Penetrating St ructure Containing Superoxide Dismutase-Like Subunits. Journal of Biological Chemistry, 276(23), 20466-20473. Daidone, I., Simona, F., Roccatano, D., Broglia, R. A., Tiana, G., Colo mbo, G., & Di Nola, A. (2004). Beta-Hairp in Conformation of Fibrillogenic Peptides: Structure and Alpha-Beta Transition Mechanism Revealed by Molecular Dynamics Simulat ions. Proteins, 57(1), 198-204. Davis, Charles H., & Berkowit z, Max L. (2009). Interaction Between A myloid -β (1– 42) Peptide and Phospholipid Bilayers: A Molecular Dynamics Study. Biophysical Journal, 96(3), 785-797. DeTo ma, Alaina S., Salamekh, Samer, Ramamoorthy, Ayyalusamy, & Lim, Mi Hee. (2012). M isfolded proteins in Alzheimer's disease and type II diabetes. Chemical Society Reviews, 41(2), 608-621. Doll, R., Peto, R., Boreham, J., & Sutherland, I. (2000). Smo king and Dementia in Male Brit ish Doctors: Prospective Study. British Medical Journal, 320(7242), 1097-1102. Duce, J. A., & Bush, A. I. (2010). Biological Metals and Alzheimer's Disease: Implications for Therapeutics and Diagnostics. Progress in Neurobiology, 92(1), 1-18. Esler, W. P., St imson, E. R., Jennings, J. M., Gh ilardi, J. R., Mantyh, P. W ., & Maggio, J. E. (1996). Zinc-Induced Aggregation of Human and Rat Beta-Amy loid Peptides in Vitro. Journal of Neurochemistry, 66(2), 723-732. George-Hyslop, S., & Peter H. (2000). Molecu lar Genetics of Alzheimer’s Disease. Biological Psychiatry, 47(3), 183-199. Gilson, M. K., Given, J. A., Bush, B. L., & McCammon, J. A. (1997). The Statistical Thermodynamic Basis for Computation of Binding Affinities: A Critical Review. Biophysical Journal, 72(3), 1047-1069. Goodsell, D. S., Morris, G. M., & Olson, A. J. (1996). Automated Docking of Flexib le Ligands: Applications of Autodock. Journal of Molecular Recognition, 9(1), 1-5. Greenbaum, E. A., Graves, C. L., M ishizen-Eberz, A. J.,Lupoli, M. A., Lynch, D. R., Englander, S. W. et al.(2005). The e46k mutation in alpha-synuclein increasesamyloid fibril format ion. Journal of Biological Chemistry, 280, 7800–7807. Gu L, Guo Z (2013). Alzheimer's Aβ42 and Aβ40 peptides form interlaced amy loid fibrils. Journal of Neurochemistry, 126 (3):305-311.
85
Hammarström, P., W iseman, R.L., Powers, E.T., & Kelly, J.W. (2003). Prevention of transthyretin amyloid d isease by changing protein misfolding energetics. Science, 299(5607), 713-716. Hardy, J. A. & Higgins, G. A. (1992). Alzheimer's disease: the amyloid cascade hypothesis. Science, 256, 184–185. Hardy, J., & Selkoe, D. J. (2002). The A myloid Hypothesis of Alzheimer's Disease: Progress and Problems on the Road to Therapeutics. Science, 297(5580), 353356. Hebert, L. E., Scherr, P. A., Beckett, L. A., Funkenstein, H. H., A lbert, M. S., Chown, M. J., & Evans, D. A. (1992). Relat ion of Smo king and Alcohol Consumption to Incident Alzheimer's Disease. American Journal of Epidemiology, 135(4), 347-355. Hébert, R., Lindsay, J., Verreau lt, R., Rockwood, K., Hill, G., & Dubois, M. F. (2000). Vascular Dementia Incidence and Risk Factors in the Canadian Study of Health and Aging. Stroke, 31(7), 1487-1493. Hilbich, C., Kisters-Woike, B., Reed, J., Masters, C. L., & Beyreuther, K. (1991). Aggregation and secondary structure of synthetic amyloid βA4 peptides of Alzheimer's disease. Journal of Molecular Biology, 218(1), 149-163. Hou, L., Shao, H., Zhang, Y., Li, H., Menon, N. K., Neuhaus, E. B., Brewer, J. M., & Zagorski, M. G. (2004). Solution NM R Studies of the Aβ(1−40) and Aβ(1−42) Peptides Establish that the Met35 Oxidation State Affects the Mechanism of Amylo id Formation. Journal of the American Chemical Society, 126(7), 1992-2005. Huey, R., Morris, G. M., Olson, A. J., & Goodsell, D. S. (2007). A Semiemp irical Free Energy Force Field with Charge-Based Desolvation. Journal of Computational Chemistry, 28(6), 1145-1152. Hutton, M. (2001). M issense and Splice Site Mutations in Tau Associated with FTDP 17: Mu ltiple Pathogenic Mechanisms. Neurology, 56(11 Suppl 4), S21-25. Isaacs, A. M., Senn, D. B., Yuan, M., Shine, J. P., & Yankner, B. A. (2006). Acceleration of amyloid β-peptide aggregation by physiological concentrations of calcium. Journal of Biological Chemistry, 281(38), 2791627923. Jalili, S, & Akhavan, M. (2009). A Molecular Dynamics Simu lations Study of Conformat ional Changes and Solvation of Aβ Peptide in Trifluoroethanol and Water. Journal of Theoretical and Computational Chemistry, 8(02), 215-231. Juszczyk, P., Ko lodziejczyk, A. S., & Grzonka, Z. (2005). Circular Dichrois m and Aggregation Studies of Amyloid Beta (11-8) Frag ment and Its Variants. Acta Biochimica Polonica, 52(2), 425-431. Epub 2005 Jun 25.
86
Karr, J. W., & Szalai, V. A. (2008). Cu (II) Binding to Monomeric, Oligo meric, and Fibrillar Forms of the Alzheimer’s Disease Amyloid-β Peptide. Biochemistry, 47(17), 5006-5016. Kelly, J. W. (1996). Alternative Conformations of Amylo idogenic Proteins Govern Their Behavior. Current Opinion in Structural Biology, 6(1), 11-17. Khatoon, Sabiha, Grundke‐Iqbal, I, & Iqbal, K. (1992). Brain Levels of Microtubule‐Associated Protein τ Are Elevated in Alzheimer's Disease: A Radio immuno‐Slot‐Blot Assay for Nanograms of the Protein. Journal of Neurochemistry, 59(2), 750-753. Kirschner, D. A., Abraham, C., & Selkoe, D. J. (1986). X-ray diffraction fro m intraneuronal paired helical filaments and extraneuronal amy loid fibers in Alzheimer d isease indicates cross -beta conformation. Proceedings of the National Academy of Sciences, 83(2), 503. Kopke, E., Tung, Y. C., Shaikh, S., Alonso, A. C., Iqbal, K., & Grundke -Iqbal, I. (1993). M icrotubule-associated Protein Tau. Abnormal phosphorylation of a Non-Paired Helical Filament Pool in Alzheimer Disease. Journal of Biological Chemistry, 268(32), 24374-24384. Kozlowski, H., Janicka-Klos, A., Brasun, J., Gaggelli, E., Valensin, D., & Valensin, G. (2009). Copper, Iron, and Zinc Ions Homeostasis and Their Role in Neurodegenerative Disorders (Metal Uptake, Transport, Distribution and Regulation). Coordination Chemistry Reviews, 253(21– 22), 2665-2685. Krishtalik, L. I., & Cramer, W. A. (1995). On the physical basis for the cis -positive rule describing protein orientation in biological memb ranes. FEBS Letters, 369(2– 3), 140-143. Lambert , M. P., Barlow, A. K., Chro my, B. A., Ed wards, C., Freed, R., Liosatos, M., . . . Klein, W. L. (1998). Diffusible, Nonfib rillar Ligands Derived fro m Aβ1–42 are Potent Central Nervous System Neurotoxins. Proceedings of the National Academy of Sciences, 95(11), 6448-6453. Launer, L. J., Andersen, K., Dewey, M., Letenneur, L., Ott, A., A maducci, L. A., ... & Hofman, A. (1999). Rates and Risk Factors for Dementia and Alzheimer’s disease Results from EURODEM Pooled Analyses. Neurology, 52(1), 78-78. Lee, Chewook, & Ham, Sihyun. (2010). Charactering Amylo id-Beta Protein Misfolding fro m Molecu lar Dynamics Simu lation with Explicit Water. Biophysical Journal, 98(3, Supplement 1), 649a. Lee, J. P., Stimson, E. R., Gh ilardi, J. R., Mantyh, P. W., Lu, Y. A., Felix, A. M ., Llanos, W., Behbin, A., & Cu mmings, M. (1995). 1H NMR of A.beta. Amylo id Peptide Congeners in Water Solution. Conformational Changes Correlate with Plaque Co mpetence. Biochemistry, 34(15), 5191-5200.
87
Lee, S., & Tsai, F. T. (2005). Molecular Chaperones in Protein Quality Control. Journal of Biochemistry and Molecular Biology, 38(3), 259-265. Lindahl, E., Hess, B., & Van Der Spoel, D. (2001). GROMACS 3.0: A Package for Molecular Simulat ion and Trajectory Analysis. Journal of Molecular Modeling, 7(8), 306-317. Lindsay J, Laurin D, Verreault R, Hebert R, Helliwell B, Hill GB, McDo well I (2002) Risk factors for Alzheimer's disease: a prospective analysis from the Canadian Study of Health and Aging. American Journal of Epidemiology 156 (5):445453. Lindsay, J., Hébert, R., & Rockwood, K. (1997). The Canadian Study of Health and Aging Risk Factors for Vascular Dementia. Stroke, 28(3), 526-530. Liu D.; Xu Y.; Feng Y.; Liu H.; Shen X.; Chen K.; Ma J.; Jiang H. (2006) Inhibitor Discovery Targeting the Intermediate Structure of the β-Amyloid Peptide on the Conformat ional Transition Pathway: Imp lications in the Aggregation Mechanism of β-A my loid Peptide. Biochemistry 45, 10963–10972. Liu, B., Moloney, A., Meehan, S., Morris, K., Thomas, S. E., Serpell, L. C., Hider, R ., Marciniak, S. J., Lo mas, D. A., & Crowther, D. C. (2011). Iron Pro motes the Toxicity of Amylo id β Peptide by Impeding Its Ordered Aggregation. Journal of Biological Chemistry, 286(6), 4248-4256. Lovell, M. A., Robertson, J. D., Teesdale, W. J., Campbell, J. L., & Markesbery, W. R. (1998). Copper, Iron and Zinc in Alzheimer's Disease Senile Plaques. Journal of the Neurological Sciences, 158(1), 47-52. Luo, J. C., Wang, S. C., Jian, W. B., Chen, C. H., Tang, J. L., & Lee, C. I. (2012). Formation of amy loid fibrils fro m β-amy lase. FEBS Letters, 586(6), 680-685. MacKerell, A. D., Bashford, D., Bellott, Dunbrack, R. L., Evanseck, J. D., Field, M. J., Fischer, S., Gao, J., Guo, H., Ha, S., Joseph-McCarthy, D., Kuchnir, L., Kuczera, K., Lau, F. T. K., Mattos , C., M ichnick, S., Ngo, T., Nguyen, D. T., Prodhom, B., Reiher, W. E., Rou x, B., … & Karp lus, M. (1998). All-Ato m Emp irical Potential for Molecular Modeling and Dynamics Studies of Proteins†. The Journal of Physical Chemistry B, 102(18), 3586-3616. Malde, A. K., Zuo, L., Breeze, M., Stroet, M., Poger, D., Nair, P. C., . . . Mark, A. E. (2011). An Automated Force Field Topology Bu ilder (ATB) and Repository: Version 1.0. Journal of Chemical Theory and Computation, 7(12), 4026-4037. Mantyh, P. W., Gh ilard i, J. R., Rogers, S., DeMaster, E., Allen, C. J., Stimson, E. R., & Maggio, J. E. (1993). Alu minum, Iron, and Zinc Ions Promote Aggregation of Physiological Concentrations of β-Amylo id Peptide. Journal of Neurochemistry, 61(3), 1171-1174.
88
Martínez, L., Andrade, R., Birg in, E., & Mart ínez, J. (2009). Packmo l: A Package for Building In itial Configurations for Molecular Dynamics Simu lations. Journal of Computational Chemistry, 30(13), 2157-2164. McCammon, J. A., Gelin, B. R., & Karplus, M. (1977). Dynamics of Folded Proteins. Nature, 267(5612), 585-590. McDowell, I., Hill, G., Lindsay, J., Helliwell, B., Costa, L., Beattie, L., ... & Buehler, S. (1994). The Canadian study of health and aging: risk -factors for A lzheimers disease in Canada. Neurology, 44(11), 2073-80. McKoy, A. F., Chen, J., Schupbach, T., & Hecht, M. H. (2012). A Novel Inhib itor of Amylo id β (Aβ) Peptide Aggregation: fro m High Throughput Screening to Efficacy in an Animal Model of Alzheimer Disease. Journal of Biological Chemistry, 287(46), 38992-39000. Miller, Y., Ma, B., & Nussinov, R. (2010). Zinc ions promote Alzheimer Aβ aggregation via population shift of polymorphic states. Proceedings of the National Academy of Sciences, 107(21), 9490-9495. Miura, T., Su zuki, K., Kohata, N., & Takeuchi, H. (2000). Metal Binding Modes of Alzheimer's Amylo id β-Peptide in Insoluble Aggregates and Soluble Co mplexes. Biochemistry, 39(23), 7024-7031. Morris, G. M., Goodsell, D. S., Halliday, R. S., Huey, R., Hart, W. E., Belew, R. K., & Olson, A. J. (1998). Automated Docking Using a Lamarckian Genetic Algorith m and an Emp irical Binding Free Energy Function. Journal of Computational Chemistry, 19(14), 1639-1662. Morris, G. M., Huey, R., Lindstrom, W., Sanner, M. F., Belew, R. K., Goodsell, D. S., & Olson, A. J. (2009). AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility. Journal of Computational Chemistry, 30(16), 2785-2791. Morris, G., Goodsell, D., Huey, R., & Olson, A. (1996). Distributed Automated Docking of Flexib le Ligands to Proteins: Parallel Applications of AutoDock 2.4. Journal of Computer-Aided Molecular Design, 10(4), 293-304. Narhi, L., Wood, S. J., Steavenson, S., Jiang, Y., Wu,G. M., Anafi, D. et al. (1999). Both familial Parkinson'sdisease mutations accelerate alpha-synuclein aggregation. Journal of Biological Chemistry. 274, 9843–9846 Nilsson, M. R. (2004). Techniques to Study Amyloid Fibril Formation in Vit ro. Methods, 34(1), 151-160. Ohnishi, S., & Takano, K. (2004). Amy loid fib rils fro m the viewpoint of protein folding. Cellular and Molecular Life Sciences, 61(5), 511-524. Österberg, F., Morris, G. M., Sanner, M. F., Olson, A. J., & Goodsell, D. S. (2002). Automated Docking to Multiple Target Structures: Incorporation of Protein
89
Mobility and Structural Water Heterogeneity in AutoDock. Proteins: Structure, Function, and Bioinformatics, 46(1), 34-40. Panza F, Frisardi V, Solfrizzi V, Imb imbo BP, Logroscino G, Santamato A, Greco A, Seripa D, Pilotto A (2011). Interacting with gamma-secretase for treating Alzheimer's disease: fro m inhib ition to modulation. Current Medicinal Chemistry, 18 (35):5430-5447. Parbhu, A., Lin, H., Th imm, J., & Lal, R. (2002). Imag ing Real-Time Aggregation of Amylo id Beta Protein (1–42) by Atomic Force Microscopy. Peptides, 23(7), 1265-1270. Pasture, O., & Onkia, O. (1994). Canadian study of Health and Aging: Study Methods and Prevalence of Dementia. Canadian Medical Assocociation Journal, 150(6), 899-913. Pedersen, J. T., & Heegaard, N. H. H. (2013). Analysis of Protein Aggregation in Neurodegenerative Disease. Analytical Chemistry, 85(9), 4215-4227. Pithadia, A. S., & Lim, M. H. (2012). Metal-Associated Amyloid-β Species in Alzheimer's Disease. Current Opinion in Chemical Biology, 16(1), 67-73. Prusiner, S. B. (2001). Neurodegenerative Diseases and Prions. New England Journal of Medicine, 344(20), 1516-1526. Rabanal, F., Tusell, J. M., Sastre, L., Qu intero, M. R., Cruz, M., Grillo, D., Pons, M., Albericio, F., Serratosa, J., & Giralt, E. (2002). Structural, Kinetic and Cytotoxicity Aspects of 12-28 Beta-A myloid Protein Frag ment: A Reappraisal. Journal of Peptide Science, 8(10), 578-588. Rah man, A. (1964). Correlations in the Motion of Atoms in Liquid Argon. Physical Review, 136(2A), A405-A411. Rajan, R., Awasthi, S. K., Bhattacharjya, S., & Balaram, P. (1997). “Teflon -Coated Peptides”: Hexafluoroacetone Trihydrate as a Structure Stabilizer for Peptides. Biopolymers, 42(2), 125-128. Rao, Praveen P. N., Mohamed, Tarek, & Os man, Wesseem. (2013). Investigating the Binding Interactions of Galantamine with β-A mylo id Peptide. Bioorganic & Medicinal Chemistry Letters, 23(1), 239-243. Rapoport, M., Dawson, H. N., Binder, L. I., Vitek, M. P., & Ferreira, A. (2002). Tau is Essential to Beta-Amylo id-Induced Neurotoxicity. Proceeding of the National Academy of Science of the U S A., 99(9), 6364-6369. Epub 2002 Apr 6316. Riek, R., Güntert, P., Döbeli, H., W ipf, B., & Wüthrich, K. (2001). NMR studies in Aqueous Solution Fail to Identify Significant Conformat ional Differences between the Monomeric Forms of Two Alzheimer Peptides with Widely Different Plaque-Co mpetence, Aβ(1–40)o x and Aβ(1–42)o x. European Journal of Biochemistry, 268(22), 5930-5936.
90
Rocha, S., Loureiro, J. A., Brezesinski, G., & Pereira, M . D. C. (2012). Peptide– surfactant interactions: Consequences for the Amyloid -Beta Structure. Biochemical and Biophysical Research Communications, 420(1), 136-140. Rodziewicz-Motowidło, S., Czap lewska, P., Siko rska, E., Spodzieja, M., & Kołodziejczy k, A. S. (2008). The Arctic Mutation Alters Helix Length and Type in the 11–28 β-A my loid Peptide Monomer—CD, NM R and MD Studies in an SDS M icelle. Journal of Structural Biology, 164(2), 199-209. Ross, C. A., & Po irier, M. A. (2004). Protein A ggregation and Neurodegenerative Disease. Nature Medicine, 10(7), S10-17. Sanner, M. F. (2005). A Co mponent-Based Software Environment for Visualizing Large Macro molecular Assemblies. Structure, 13(3), 447-462. Santacruz, K, Lewis, J, Spires, T, Paulson, J, Kotilinek, L, Ingelsson, M., … & Ashe, K. H. (2005). Tau Suppression in a Neurodegenerative Mouse Model Improves Memory Function. Science, 309(5733), 476-481. Sarell, C.J., Sy me, C.D., Rigby, S.E.J., & Viles, J.H. (2009). Copper (II) Binding to Amylo id-β Fibrils of Alzheimer’s Disease Reveals a Pico molar Affin ity: Stoichio metry and Coordination Geo metry are Independent of Aβ Oligo meric Form. Biochemistry, 48(20), 4388-4402. Sathya, M., Premku mar, P., Karthick, C., Moorthi, P., Jayachandran, K. S., & Anusuyadevi, M. (2012). BACE1 in Alzheimer's disease. Clinica Chimica Acta, 414, 171-178. Scott, L. E., & Orvig, C. (2009). Medicinal Inorganic Chemistry Approaches to Passivation and Removal of Aberrant Metal Ions in Disease. Chemical Reviews, 109(10), 4885-4910. Selkoe, D. J. (2003). Fold ing Proteins in Fatal Ways. Nature, 426(6968), 900-904. Selkoe, D. J., & Podlisny, M. B. (2002). Deciphering the Genetic Basis of Alzheimer's Disease. Annual Review of Genomics and Human Genetics, 3, 67-99. Shao, H., Jao, S.-c., Ma, K., & Zagorski, M. G. (1999). So lution Structures of MicelleBound Amyloid β-(1-40) and β-(1-42) Peptides of Alzheimer’s Disease. Journal of Molecular Biology, 285(2), 755-773. Snyder, S. W., Ladror, U. S., Wade, W. S., Wang, G. T., Barrett, L. W., Matayoshi, E. D., Huffaker, H. J., Krafft , G. A., & Holzman, T. F. (1994). A myloid -Beta Aggregation: Selective Inh ibition of Aggregation in Mixtures of A my loid with Different Chain Lengths. Biophysical Journal, 67(3), 1216-1228. Sousa, S. F., Fernandos, P. A., & Ramos, M. J. (2006). Protein–ligand docking: Current status and future challenges. Proteins: Structure, Function, and Bioinformatics, 65(1), 15-26.
91
Squitti, R., Siotto, M., Salustri, C., & Polimanti, R. (2013). Metal Dysfunction in Alzheimer’s Disease. In Studies on Alzheimer's Disease (pp. 73-97). Hu mana Press. Sticht, H., Bayer, P., Willbold, D., Dames, S., Hilbich, C., Beyreuther, K., . . . Rösch, P. (1995). Structure of A my loid A4-(1–40)-Peptide of Alzheimer's Disease. European Journal of Biochemistry, 233(1), 293-298. Stillinger, F. H., & Rah man, A. (1974). Improved Simulat ion of Liquid Water by Molecular Dynamics. The Journal of Chemical Physics, 60, 1545. Suh, S. W., Jensen, K. B., Jensen, M. S., Silva, D. S., Kesslak, P. J., Danscher, G., & Frederickson, C. J. (2000). Histochemically-reactive zinc in amy loid plaques, angiopathy, and degenerating neurons of Alzheimer's diseased brains. Brain Research, 852(2), 274-278. Tabner, B. J., Turnbull, S., El-Agnaf, O. M . A., & A llsop, D. (2002). Formation of Hydrogen Pero xide and Hydro xyl Radicals fro m Aβ and α-Synuclein as a Possible Mechanism of Cell Death in Alzheimer’s Disease and Parkinson’s Disease. Free Radical Biology and Medicine, 32(11), 1076-1083. Takano, K. (2008). A my loid Beta Conformat ion in Aqueous Environment. Current Alzheimer Research, 5(6), 540-547. Takano, K., Endo, S., Mukaiyama, A., Chon, H., Matsumura, H., Koga, Y., & Kanaya, S. (2006). St ructure of Amy loid β Frag ments in Aqueous Environments. Federation of European Biochemical Societies Journal, 273(1), 150-158. Tan, J., & Evin, G. (2012). β‐Site APP‐cleav ing enzyme 1 trafficking and Alzheimer’s disease pathogenesis. Journal of Neurochemistry, 120(6), 869-880. Tarus, Bogdan, Straub, John E., & Th iru malai, D. (2005). Probing the Init ial Stage of Aggregation of the Aβ10-35-protein : Assessing the Propensity for Peptide Dimerization. Journal of Molecular Biology, 345(5), 1141-1156. Teper G. L.; Lecanu L.; Greeson J.; Papadopoulos V. (2005) Methodology for MultiSite Ligand-Protein Docking Identification Developed for the Optimizat ion fo Spirostenol Inhibition of β -A mlo id-Induced Neurotoxicity. Chemistry and Biodiversity 2, 1571–1579. Teplow D. B.; Lazio N. D.; Bitan G.; Bernstein S.; Wyttenbach T.; Bowers M. T.; Bau mketner A.; Shea J.-E.; Urbanc B.; Cru z L.; Borreguero J.; Stanley H. E. (2006) Elucidating Amy loid β-Protein Fold ing and Assembly: A Multidisciplinary Approach. Accounts of Chemical Research. 39, 635–645. Terzi, E., Ho lzemann, G., & Seelig, J. (1995). Self -association of beta-amyloid peptide (1-40) in solution and binding to lipid membranes. Journal of Molecular Biology. 252(5), 633-642.
92
Thomas, P. J, Qu, B. H., & Pedersen, P. L. (1995). Defective Protein Folding as a Basis of Hu man Disease. Trends in Biochemical Sciences, 20(11), 456-459. Tjernberg, L. O., Callaway, D. J., Tjernberg, A., Hahne, S., Lilliehook, C., Teren ius, L., Thyberg, J., & Nordstedt, C. (1999). A Molecular Model of Alzheimer Amylo id Beta-Peptide Fibril Format ion. Journal of Biological Chemistry, 274(18), 12619-12625. Tofoleanu, Florentina, & Buchete, Nicolae -Viorel. (2012). Molecular Interactions of Alzheimer's Aβ Protofilaments with Lip id Membranes. Journal of Molecular Biology, 421(4– 5), 572-586. Tomaselli S, Esposito V, Vangone P, van Nuland NAJ, Bonvin AMJJ, Guerrini R, Tancredi T, Temussi PA, Picone D (2006) The α-to-β Conformational Transition of Alzheimer's Aβ-(1–42) Peptide in Aqueous Media is Reversible: A Step by Step Conformational Analysis Suggests the Location of β Conformat ion Seeding. ChemBioChem 7 (2):257-267. Trott, O., & Olson, A. J. (2010). AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and mu ltithreading. Journal of Computational Chemistry, 31(2), 455-461 Turner, P., Team, G. D., & Stambulchik, E. (2004). xmGrace Modeling Software.
Urbanc, B., Cru z, L., Ding, F., Sammond, D., Khare, S., Buldyrev, S. V., Stanley, H. E., & Do kholyan, N. V. (2004). Molecular Dynamics Simulat ion of Amy loid Beta Dimer Formation. Biophysical Journal, 87(4), 2310-2321. Uversky, V. N. (2003). Protein fold ing revisited. A polypeptide chain at the fold ing – misfolding – nonfolding cross-roads: which way to go? Cellular and Molecular Life Sciences, 60(9), 1852-1871. Valensin, D., Gabbiani, C., & Messori, L. (2012). Metal Co mpounds as Inhibitors of β Amylo id Aggregation. Perspectives for an Innovative Metallotherapeutics on Alzheimer's Disease. Coordination Chemistry Reviews, 256(19–20), 23572366. Valko, M., Morris, H., & Cronin , M. T. (2005). Metals, To xicity and Oxidative Stress. Current Medicinal Chemistry, 12(10), 1161-1208. van der Spoel, D., Lindhal, E., Hess, B., van Buuren, A. R., Apol, E., Meulenhoff, P. J., ... & Berendsen, H. J. C. (2013). Gro macs User Manual version 4.5. 5, 2011. van Gunsteren, W. F., Billeter, S. R., Eising, A. A., Hünenberger, P. H., Krüger, P., Mark, A. E., Scott, W. R. P., & Tironi, I. G. (1996). Bio molecu lar Simulat ion: The {GROMOS96} manual and user guide.
93
Van Klo mpenburg, W, & De Kru ijff, B. (1998). The Role of Anionic Lip ids in Protein Insertion and Translocation in Bacterial Membranes. Journal of Membrane Biology, 162(1), 1-7. Varadarajan, S., Yatin, S., Aksenova, M., & Butterfield, D.A. (2000). Rev iew: Alzheimer's Amylo id β-Peptide-Associated Free Radical Oxidative Stress and Neurotoxicity. Journal of Structural Biology, 130(2-3), 184-208. Viet M. H.; Ngo S. T.; Lam N. S.; Li M. S. (2011) Inhibit ion of Aggregation of Amylo id Peptides by Beta-Sheet Breaker Peptides and Their Bind ing Affin ity. The Journal of Physical Chemistry B, 115, 7433–7446. Viles, J. H. (2012). Metal ions and amyloid fiber fo rmation in neurodegenerative diseases. Copper, zinc and iron in Alzheimer's, Parkinson's and prion diseases. Coordination Chemistry Reviews, 256(19–20), 2271-2284. Vivekanandan, S., Brender, J. R., Lee, S. Y., & Ramamoorthy, A. (2011). A Part ially Folded Structure of Amyloid-Beta(1–40) in an Aqueous Environment. Biochemical and Biophysical Research Communications, 411(2), 312-316. Vo robjev, Y. N., & Hermans, J. (1999). ES/IS: Estimat ion of Conformational Free Energy by Co mb ining Dynamics Simulat ions with Exp licit Solvent with an Implicit Solvent Continuum Model. Biophysical Chemistry, 78(1-2), 195-205. Vo robjev, Yu ry N, & Hermans, Jan. (2008). Free Energ ies of Protein Decoys Provide Insight into Determinants of Protein Stability. Protein Science, 10(12), 24982506. Walsh, D. M., & Selkoe, D. J. (2004). Oligomers on the Brain: The Emerging Ro le of Soluble Protein Aggregates in Neurodegeneration. Protein and Peptide Letters, 11(3), 213-228. Walsh, D. M., & Selkoe, D. J. (2007). Aβ Oligo mers – A Decade of Discovery. Journal of Neurochemistry, 101(5), 1172-1184. Wang Y.; Xia Z.; Xu J.-R.; Wang Y.-X.; Hou L.-N.; Qiu Y.; Chen H.-Z. (2012) αMangostin, a polyphenolic xanthone derivative fro m mangosteen, attenuates β-amyloid oligo mers-induced neurotoxicity by inhibit ing amy loid aggregation. Neuropharmacology, 62, 871–881. Wang, H. X., Frat iglioni, L., Frisoni, G. B., Viitanen, M., & Winblad, B. (1999). Smoking and the Occurence of Alzheimer's disease: Cross -sectional and Longitudinal Data in a Population-Based Study. American Journal of Epidemiology, 149(7), 640-644. Wang, R., Lai, L., & Wang, S. (2002). Further Develop ment and Validation of Emp irical Scoring Functions for Structure-Based Binding Affinity Prediction. Journal of Computer-Aided Molecular Design, 16(1), 11-26.
94
Wei, G., & Shea, J. E. (2006). Effects of Solvent on the Structure of the Alzheimer Amylo id-β(25– 35) Peptide. Biophysical Journal, 91(5), 1638-1647. Weiner, S. J., Kollman, P. A., Nguyen, D. T., & Case, D. A. (1986). An All Atom Force Field for Simu lations of Proteins and Nucleic Acids. Journal of Computational Chemistry, 7(2), 230-252. Wu, X., Yang, G., Zu, Y., & Zhou, L. (2012). Mo lecular Dynamics Studies of β hairpin Fo lding with the Presence of the Sodium Ion. Computational Biology and Chemistry, 38, 1-9. Wu, Xiao min, Yang, Gang, Zu, Yuangang, & Zhou, Lijun. (2012). Molecu lar dynamics studies of β-hairpin folding with the presence of the sodium ion. Computational Biology and Chemistry, 38(0), 1-9. Yang C.; Zhu X.; Li J.; Shi R. (2010) Explorat ion of the mechanism for LPFFD inhibit ing the format ion of β-sheet conformat ion of Aβ(1–42) in water. Journal of Molecular Modeling. 16, 813–821. Yates, E. A., Owens, S. L., Lynch, M. F., Cucco, E. M., Umbaugh, C. S., & Leg leiter, J. (2013). Specific Do mains of Aβ Facilitate Aggregation on and Association with Lipid Bilayers. Journal of Molecular Biology, 425(11), 1915-1933. Yin, Y. I., Bassit, B., Zhu, L., Yang, X., Wang, C., & Li, Y. M. (2007). γ Secretase Substrate Concentration Modulates the Abeta42/Abeta40 Ratio: Implications for Alzheimer Disease. Journal of Biological Chemistry, 282(32), 2363923644. Epub 22007 Jun 23637. Yoshiike, Y., Tanemura, K., Murayama, O., Akagi, T., Murayama, M., Sato, S., Sun, X., Tanaka, N., & Takashima, A. (2001). New Insights on How Metals Disrupt Amylo id Beta-Aggregation and Their Effects on Amyloid -Beta Cytotoxicity. Journal of Biological Chemistry, 276(34), 32293-32299. Epub 2001 Jun 22. Yoshitake, T., Kiyohara, Y., Kato, I., Oh mura, T., Iwamoto, H., Nakayama, K., ... & Fujishima, M. (1995). Incidence and risk factors of vascular dementia and Alzheimer's disease in a defined elderly Japanese population The Hisayama Study. Neurology, 45(6), 1161-1168. Zhang, S., Iwata, K., Lachen mann, M. J., Peng, J. W., Li, S., Stimson, E. R., Lu, Y. A., Felix, A. M., Maggio, J. E., & Lee, J. P. (2000). The Alzheimer's Peptide Aβ Adopts a Collapsed Coil Structure in Water. Journal of Structural Biology, 130(2–3), 130-141.
95
APPENDICES APPENDIX A Structure of Aβ(1-42) with pdb code: 1 IYT REMARK 4 XXXX COM PLIES WITH FORMAT V. 2.0 ATOM 1 N ASP A 1 -18.493 4.557 22.977 1.00 0.00 ATOM 2 CA ASP A 1 -17.779 4.595 21.686 1.00 0.00 ATOM 3 C ASP A 1 -17.049 5.927 21.534 1.00 0.00 ATOM 4 O ASP A 1 -17.666 6.978 21.679 1.00 0.00 ATOM 5 CB ASP A 1 -18.738 4.399 20.499 1.00 0.00 ATOM 6 CG ASP A 1 -19.302 2.986 20.392 1.00 0.00 ATOM 7 OD1 ASP A 1 -18.830 2.093 21.089 1.00 0.00 ATOM 8 OD2 ASP A 1 -20.201 2.781 19.390 1.00 0.00 ATOM 9 H1 ASP A 1 -17.830 4.666 23.731 1.00 0.00 ATOM 10 H2 ASP A 1 -19.152 5.323 23.014 1.00 0.00 ATOM 11 H3 ASP A 1 -18.976 3.675 23.080 1.00 0.00 ATOM 12 HA ASP A 1 -17.042 3.790 21.685 1.00 0.00 ATOM 13 HB2 A SP A 1 -19.554 5.116 20.586 1.00 0.00 ATOM 14 HB3 A SP A 1 -18.193 4.603 19.577 1.00 0.00 ATOM 15 HD2 ASP A 1 -20.482 3.586 18.952 1.00 0.00 ATOM 16 N ALA A 2 -15.750 5.876 21.228 1.00 0.00 ATOM 17 CA A LA A 2 -14.939 7.059 20.973 1.00 0.00 ATOM 18 C A LA A 2 -13.825 6.704 19.989 1.00 0.00 ATOM 19 O ALA A 2 -13.882 7.091 18.827 1.00 0.00 ATOM 20 CB A LA A 2 -14.398 7.624 22.293 1.00 0.00 ATOM 21 H ALA A 2 -15.322 4.974 21.098 1.00 0.00 ATOM 22 HA A LA A 2 -15.555 7.830 20.504 1.00 0.00 ATOM 23 HB1 A LA A 2 -15.230 7.958 22.914 1.00 0.00 ATOM 24 HB2 A LA A 2 -13.834 6.869 22.841 1.00 0.00 ATOM 25 HB3 A LA A 2 -13.752 8.479 22.087 1.00 0.00 ATOM 26 N GLU A 3 -12.827 5.941 20.450 1.00 0.00 ATOM 27 CA GLU A 3 -11.659 5.574 19.654 1.00 0.00 ATOM 28 C GLU A 3 -11.221 4.123 19.876 1.00 0.00 ATOM 29 O GLU A 3 -10.225 3.675 19.318 1.00 0.00 ATOM 30 CB GLU A 3 -10.527 6.587 19.895 1.00 0.00 ATOM 31 CG GLU A 3 -10.279 6.898 21.383 1.00 0.00 ATOM 32 CD GLU A 3 -9.124 7.880 21.569 1.00 0.00 ATOM 33 OE1 GLU A 3 -8.146 7.798 20.834 1.00 0.00 ATOM 34 OE2 GLU A 3 -9.295 8.850 22.511 1.00 0.00 ATOM 35 H GLU A 3 -12.840 5.661 21.418 1.00 0.00 ATOM 36 HA GLU A 3 -11.930 5.598 18.602 1.00 0.00 ATOM 37 HB2 GLU A 3 -9.609 6.205 19.441 1.00 0.00 ATOM 38 HB3 GLU A 3 -10.784 7.519 19.389 1.00 0.00 ATOM 39 HG2 GLU A 3 -11.176 7.334 21.818 1.00 0.00 ATOM 40 HG3 GLU A 3 -10.026 5.981 21.918 1.00 0.00 ATOM 41 HE2 GLU A 3 -10.115 8.786 23.001 1.00 0.00 ATOM 42 N PHE A 4 -11.999 3.375 20.657 1.00 0.00 ATOM 43 CA PHE A 4 -11.743 1.968 20.932 1.00 0.00
N C C O C C O O H H H H H H H N C C O C H H H H H N C C O C C C O O H H H H H H H N C
96
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
44 C PHE A 4 -12.131 1.108 19.732 1.00 0.00 C 45 O PHE A 4 -11.503 0.092 19.449 1.00 0.00 O 46 CB PHE A 4 -12.483 1.558 22.212 1.00 0.00 C 47 CG PHE A 4 -12.105 2.413 23.409 1.00 0.00 C 48 CD1 PHE A 4 -10.854 2.223 24.026 1.00 0.00 C 49 CD2 PHE A 4 -12.938 3.469 23.831 1.00 0.00 C 50 CE1 PHE A 4 -10.434 3.085 25.053 1.00 0.00 C 51 CE2 PHE A 4 -12.510 4.339 24.851 1.00 0.00 C 52 CZ PHE A 4 -11.257 4.149 25.460 1.00 0.00 C 53 H PHE A 4 -12.813 3.804 21.055 1.00 0.00 H 54 HA PHE A 4 -10.678 1.835 21.090 1.00 0.00 H 55 HB2 PHE A 4 -13.560 1.607 22.048 1.00 0.00 H 56 HB3 PHE A 4 -12.235 0.518 22.434 1.00 0.00 H 57 HD1 PHE A 4 -10.207 1.418 23.710 1.00 0.00 H 58 HD2 PHE A 4 -13.915 3.609 23.394 1.00 0.00 H 59 HE1 PHE A 4 -9.476 2.929 25.530 1.00 0.00 H 60 HE2 PHE A 4 -13.150 5.144 25.183 1.00 0.00 H 61 HZ PHE A 4 -10.932 4.810 26.251 1.00 0.00 H 62 N ARG A 5 -13.177 1.538 19.027 1.00 0.00 N 63 CA A RG A 5 -13.685 0.870 17.839 1.00 0.00 C 64 C ARG A 5 -12.779 1.131 16.636 1.00 0.00 C 65 O ARG A 5 -12.306 0.195 15.994 1.00 0.00 O 66 CB A RG A 5 -15.117 1.339 17.543 1.00 0.00 C 67 CG A RG A 5 -16.163 0.663 18.439 1.00 0.00 C 68 CD A RG A 5 -16.283 -0.837 18.126 1.00 0.00 C 69 NE A RG A 5 -17.648 -1.332 18.353 1.00 0.00 N 70 CZ ARG A 5 -18.682 -1.123 17.523 1.00 0.00 C 71 NH1 A RG A 5 -18.528 -0.375 16.425 1.00 0.00 N 72 NH2 A RG A 5 -19.872 -1.666 17.798 1.00 0.00 N 73 H ARG A 5 -13.591 2.401 19.334 1.00 0.00 H 74 HA ARG A 5 -13.690 -0.203 18.019 1.00 0.00 H 75 HB2 A RG A 5 -15.182 2.420 17.673 1.00 0.00 H 76 HB3 A RG A 5 -15.362 1.114 16.504 1.00 0.00 H 77 HG2 A RG A 5 -15.914 0.800 19.492 1.00 0.00 H 78 HG3 A RG A 5 -17.120 1.153 18.255 1.00 0.00 H 79 HD2 A RG A 5 -15.999 -1.052 17.095 1.00 0.00 H 80 HD3 A RG A 5 -15.604 -1.394 18.774 1.00 0.00 H 81 HE A RG A 5 -17.795 -1.893 19.180 1.00 0.00 H 82 1HH1 ARG A 5 -17.629 0.043 16.238 1.00 0.00 H 83 2HH1 ARG A 5 -19.297 -0.210 15.793 1.00 0.00 H 84 1HH2 ARG A 5 -19.988 -2.229 18.628 1.00 0.00 H 85 2HH2 ARG A 5 -20.663 -1.514 17.191 1.00 0.00 H 86 N HIS A 6 -12.568 2.407 16.301 1.00 0.00 N 87 CA HIS A 6 -11.902 2.823 15.070 1.00 0.00 C 88 C HIS A 6 -10.382 2.666 15.169 1.00 0.00 C 89 O HIS A 6 -9.644 3.631 14.989 1.00 0.00 O 90 CB HIS A 6 -12.292 4.270 14.739 1.00 0.00 C 91 CG HIS A 6 -13.749 4.441 14.389 1.00 0.00 C 92 ND1 HIS A 6 -14.260 4.531 13.108 1.00 0.00 N 93 CD2 HIS A 6 -14.788 4.555 15.272 1.00 0.00 C 94 CE1 HIS A 6 -15.591 4.694 13.211 1.00 0.00 C
97
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
95 NE2 HIS A 6 -15.932 4.709 14.514 1.00 0.00 96 H HIS A 6 -12.961 3.126 16.886 1.00 0.00 97 HA HIS A 6 -12.245 2.193 14.248 1.00 0.00 98 HB2 HIS A 6 -12.036 4.914 15.582 1.00 0.00 99 HB3 HIS A 6 -11.704 4.595 13.879 1.00 0.00 100 HD1 HIS A 6 -13.730 4.500 12.249 1.00 0.00 101 HD2 HIS A 6 -14.728 4.548 16.350 1.00 0.00 102 HE1 HIS A 6 -16.274 4.809 12.381 1.00 0.00 103 HE2 HIS A 6 -16.868 4.832 14.874 1.00 0.00 104 N ASP A 7 -9.921 1.441 15.430 1.00 0.00 105 CA A SP A 7 -8.515 1.094 15.512 1.00 0.00 106 C ASP A 7 -8.074 0.412 14.222 1.00 0.00 107 O ASP A 7 -7.204 0.916 13.521 1.00 0.00 108 CB ASP A 7 -8.262 0.208 16.739 1.00 0.00 109 CG A SP A 7 -6.831 -0.322 16.737 1.00 0.00 110 OD1 ASP A 7 -5.916 0.459 16.499 1.00 0.00 111 OD2 ASP A 7 -6.698 -1.677 16.696 1.00 0.00 112 H ASP A 7 -10.605 0.706 15.550 1.00 0.00 113 HA ASP A 7 -7.929 2.005 15.616 1.00 0.00 114 HB2 ASP A 7 -8.406 0.803 17.642 1.00 0.00 115 HB3 ASP A 7 -8.971 -0.618 16.754 1.00 0.00 116 HD2 ASP A 7 -7.476 -2.155 16.982 1.00 0.00 117 N SER A 8 -8.682 -0.725 13.882 1.00 0.00 118 CA SER A 8 -8.334 -1.479 12.686 1.00 0.00 119 C SER A 8 -8.387 -0.583 11.446 1.00 0.00 120 O SER A 8 -7.426 -0.509 10.688 1.00 0.00 121 CB SER A 8 -9.294 -2.663 12.568 1.00 0.00 122 OG SER A 8 -10.609 -2.211 12.845 1.00 0.00 123 H SER A 8 -9.478 -1.059 14.406 1.00 0.00 124 HA SER A 8 -7.317 -1.862 12.789 1.00 0.00 125 HB2 SER A 8 -9.230 -3.096 11.567 1.00 0.00 126 HB3 SER A 8 -9.014 -3.425 13.298 1.00 0.00 127 HG SER A 8 -11.218 -2.952 12.767 1.00 0.00 128 N GLY A 9 -9.500 0.129 11.260 1.00 0.00 129 CA GLY A 9 -9.682 1.039 10.133 1.00 0.00 130 C GLY A 9 -8.715 2.229 10.135 1.00 0.00 131 O GLY A 9 -8.589 2.897 9.112 1.00 0.00 132 H GLY A 9 -10.265 -0.035 11.901 1.00 0.00 133 HA2 GLY A 9 -9.534 0.481 9.206 1.00 0.00 134 HA3 GLY A 9 -10.700 1.422 10.121 1.00 0.00 135 N TYR A 10 -8.031 2.508 11.251 1.00 0.00 136 CA TYR A 10 -6.962 3.493 11.301 1.00 0.00 137 C TYR A 10 -5.652 2.812 10.894 1.00 0.00 138 O TYR A 10 -4.966 3.236 9.965 1.00 0.00 139 CB TYR A 10 -6.897 4.090 12.715 1.00 0.00 140 CG TYR A 10 -5.820 5.135 12.925 1.00 0.00 141 CD1 TYR A 10 -5.920 6.386 12.290 1.00 0.00 142 CD2 TYR A 10 -4.768 4.892 13.829 1.00 0.00 143 CE1 TYR A 10 -5.006 7.408 12.598 1.00 0.00 144 CE2 TYR A 10 -3.856 5.915 14.137 1.00 0.00 145 CZ TYR A 10 -4.013 7.189 13.568 1.00 0.00
N H H H H H H H H N C C O C C O O H H H H H N C C O C O H H H H H N C C O H H H N C C O C C C C C C C
98
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
146 OH TYR A 10 -3.152 8.190 13.909 1.00 0.00 147 H TYR A 10 -8.133 1.924 12.070 1.00 0.00 148 HA TYR A 10 -7.179 4.298 10.603 1.00 0.00 149 HB2 TYR A 10 -7.861 4.542 12.950 1.00 0.00 150 HB3 TYR A 10 -6.740 3.296 13.442 1.00 0.00 151 HD1 TYR A 10 -6.711 6.572 11.578 1.00 0.00 152 HD2 TYR A 10 -4.668 3.930 14.310 1.00 0.00 153 HE1 TYR A 10 -5.073 8.354 12.082 1.00 0.00 154 HE2 TYR A 10 -3.045 5.725 14.824 1.00 0.00 155 HH TYR A 10 -3.399 9.041 13.542 1.00 0.00 156 N GLU A 11 -5.328 1.723 11.593 1.00 0.00 157 CA GLU A 11 -4.104 0.961 11.432 1.00 0.00 158 C GLU A 11 -3.960 0.352 10.036 1.00 0.00 159 O GLU A 11 -2.830 0.161 9.593 1.00 0.00 160 CB GLU A 11 -4.051 -0.135 12.505 1.00 0.00 161 CG GLU A 11 -3.883 0.434 13.923 1.00 0.00 162 CD GLU A 11 -2.445 0.843 14.215 1.00 0.00 163 OE1 GLU A 11 -1.687 0.036 14.741 1.00 0.00 164 OE2 GLU A 11 -2.038 2.043 13.723 1.00 0.00 165 H GLU A 11 -5.965 1.424 12.319 1.00 0.00 166 HA GLU A 11 -3.270 1.644 11.575 1.00 0.00 167 HB2 GLU A 11 -4.973 -0.716 12.457 1.00 0.00 168 HB3 GLU A 11 -3.216 -0.806 12.295 1.00 0.00 169 HG2 GLU A 11 -4.541 1.285 14.096 1.00 0.00 170 HG3 GLU A 11 -4.146 -0.342 14.642 1.00 0.00 171 HE2 GLU A 11 -1.100 2.142 13.896 1.00 0.00 172 N VA L A 12 -5.075 0.019 9.369 1.00 0.00 173 CA VA L A 12 -5.085 -0.666 8.081 1.00 0.00 174 C VA L A 12 -4.060 -0.066 7.143 1.00 0.00 175 O VA L A 12 -2.973 -0.601 6.991 1.00 0.00 176 CB VA L A 12 -6.496 -0.736 7.456 1.00 0.00 177 CG1 VA L A 12 -7.260 0.587 7.320 1.00 0.00 178 CG2 VA L A 12 -6.468 -1.438 6.090 1.00 0.00 179 H VA L A 12 -5.963 0.145 9.841 1.00 0.00 180 HA VA L A 12 -4.769 -1.689 8.263 1.00 0.00 181 HB VA L A 12 -7.079 -1.342 8.134 1.00 0.00 182 1HG1 VA L A 12 -8.328 0.384 7.361 1.00 0.00 183 2HG1 VA L A 12 -6.986 1.270 8.121 1.00 0.00 184 3HG1 VA L A 12 -7.074 1.056 6.358 1.00 0.00 185 1HG2 VA L A 12 -5.975 -0.817 5.340 1.00 0.00 186 2HG2 VA L A 12 -5.934 -2.385 6.166 1.00 0.00 187 3HG2 VA L A 12 -7.488 -1.630 5.757 1.00 0.00 188 N HIS A 13 -4.409 1.048 6.521 1.00 0.00 189 CA HIS A 13 -3.605 1.689 5.504 1.00 0.00 190 C HIS A 13 -2.238 2.057 6.050 1.00 0.00 191 O HIS A 13 -1.253 1.890 5.340 1.00 0.00 192 CB HIS A 13 -4.292 2.935 4.986 1.00 0.00 193 CG HIS A 13 -5.525 2.680 4.156 1.00 0.00 194 ND1 HIS A 13 -6.837 2.899 4.539 1.00 0.00 195 CD2 HIS A 13 -5.531 2.221 2.868 1.00 0.00 196 CE1 HIS A 13 -7.627 2.547 3.507 1.00 0.00
O H H H H H H H H H N C C O C C C O O H H H H H H H N C C O C C C H H H H H H H H H N C C O C C N C C
99
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
197 NE2 HIS A 13 -6.854 2.139 2.482 1.00 0.00 198 H HIS A 13 -5.308 1.411 6.762 1.00 0.00 199 HA HIS A 13 -3.466 1.009 4.668 1.00 0.00 200 HB2 HIS A 13 -4.518 3.547 5.855 1.00 0.00 201 HB3 HIS A 13 -3.548 3.420 4.357 1.00 0.00 202 HD1 HIS A 13 -7.150 3.259 5.430 1.00 0.00 203 HD2 HIS A 13 -4.669 1.983 2.263 1.00 0.00 204 HE1 HIS A 13 -8.708 2.582 3.492 1.00 0.00 205 HE2 HIS A 13 -7.188 1.832 1.579 1.00 0.00 206 N HIS A 14 -2.174 2.535 7.300 1.00 0.00 207 CA HIS A 14 -0.902 2.796 7.959 1.00 0.00 208 C HIS A 14 0.069 1.633 7.721 1.00 0.00 209 O HIS A 14 1.256 1.873 7.508 1.00 0.00 210 CB HIS A 14 -1.088 3.044 9.465 1.00 0.00 211 CG HIS A 14 -1.621 4.398 9.869 1.00 0.00 212 ND1 HIS A 14 -1.437 4.970 11.114 1.00 0.00 213 CD2 HIS A 14 -2.276 5.304 9.080 1.00 0.00 214 CE1 HIS A 14 -1.945 6.212 11.074 1.00 0.00 215 NE2 HIS A 14 -2.471 6.432 9.855 1.00 0.00 216 H HIS A 14 -3.023 2.647 7.836 1.00 0.00 217 HA HIS A 14 -0.468 3.683 7.499 1.00 0.00 218 HB2 HIS A 14 -1.733 2.283 9.890 1.00 0.00 219 HB3 HIS A 14 -0.110 2.943 9.937 1.00 0.00 220 HD1 HIS A 14 -0.983 4.541 11.909 1.00 0.00 221 HD2 HIS A 14 -2.557 5.187 8.044 1.00 0.00 222 HE1 HIS A 14 -1.900 6.929 11.879 1.00 0.00 223 HE2 HIS A 14 -2.907 7.291 9.550 1.00 0.00 224 N GLN A 15 -0.438 0.391 7.719 1.00 0.00 225 CA GLN A 15 0.342 -0.796 7.432 1.00 0.00 226 C GLN A 15 0.215 -1.296 5.994 1.00 0.00 227 O GLN A 15 1.246 -1.481 5.357 1.00 0.00 228 CB GLN A 15 0.038 -1.869 8.477 1.00 0.00 229 CG GLN A 15 1.093 -2.981 8.452 1.00 0.00 230 CD GLN A 15 0.699 -4.132 9.367 1.00 0.00 231 OE1 GLN A 15 1.396 -4.445 10.325 1.00 0.00 232 NE2 GLN A 15 -0.428 -4.774 9.075 1.00 0.00 233 H GLN A 15 -1.425 0.252 7.911 1.00 0.00 234 HA GLN A 15 1.376 -0.492 7.493 1.00 0.00 235 HB2 GLN A 15 0.045 -1.417 9.471 1.00 0.00 236 HB3 GLN A 15 -0.958 -2.270 8.287 1.00 0.00 237 HG2 GLN A 15 1.215 -3.373 7.443 1.00 0.00 238 HG3 GLN A 15 2.049 -2.571 8.783 1.00 0.00 239 1HE2 GLN A 15 -0.991 -4.489 8.288 1.00 0.00 240 2HE2 GLN A 15 -0.715 -5.536 9.670 1.00 0.00 241 N LYS A 16 -0.989 -1.520 5.455 1.00 0.00 242 CA LYS A 16 -1.248 -1.784 4.046 1.00 0.00 243 C LYS A 16 -0.330 -0.961 3.137 1.00 0.00 244 O LYS A 16 0.125 -1.479 2.123 1.00 0.00 245 CB LYS A 16 -2.758 -1.618 3.738 1.00 0.00 246 CG LYS A 16 -3.145 -0.697 2.567 1.00 0.00 247 CD LYS A 16 -2.933 -1.357 1.199 1.00 0.00
N H H H H H H H H N C C O C C N C C N H H H H H H H H N C C O C C C O N H H H H H H H H N C C O C C C
100
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
248 CE LYS A 16 249 NZ LYS A 16 250 H LYS A 16 251 HA LYS A 16 252 HB2 LYS A 16 253 HB3 LYS A 16 254 HG2 LYS A 16 255 HG3 LYS A 16 256 HD2 LYS A 16 257 HD3 LYS A 16 258 HE2 LYS A 16 259 HE3 LYS A 16 260 HZ1 LYS A 16 261 HZ2 LYS A 16 262 HZ3 LYS A 16 263 N LEU A 17 264 CA LEU A 17 265 C LEU A 17 266 O LEU A 17 267 CB LEU A 17 268 CG LEU A 17 269 CD1 LEU A 17 270 CD2 LEU A 17 271 H LEU A 17 272 HA LEU A 17 273 HB2 LEU A 17 274 HB3 LEU A 17 275 HG LEU A 17 276 1HD1 LEU A 17 277 2HD1 LEU A 17 278 3HD1 LEU A 17 279 1HD2 LEU A 17 280 2HD2 LEU A 17 281 3HD2 LEU A 17 282 N VA L A 18 283 CA VA L A 18 284 C VA L A 18 285 O VA L A 18 286 CB VA L A 18 287 CG1 VA L A 18 288 CG2 VA L A 18 289 H VA L A 18 290 HA VA L A 18 291 HB VA L A 18 292 1HG1 VA L A 18 293 2HG1 VA L A 18 294 3HG1 VA L A 18 295 1HG2 VA L A 18 296 2HG2 VA L A 18 297 3HG2 VA L A 18 298 N PHE A 19
-4.168 -2.153 0.765 1.00 0.00 -3.922 -2.865 -0.501 1.00 0.00 -1.809 -1.483 6.041 1.00 0.00 -1.017 -2.833 3.873 1.00 0.00 -3.183 -2.606 3.561 1.00 0.00 -3.263 -1.227 4.616 1.00 0.00 -4.201 -0.441 2.667 1.00 0.00 -2.586 0.239 2.617 1.00 0.00 -2.732 -0.579 0.458 1.00 0.00 -2.080 -2.029 1.239 1.00 0.00 -4.430 -2.884 1.531 1.00 0.00 -5.009 -1.472 0.625 1.00 0.00 -3.664 -2.206 -1.222 1.00 0.00 -3.173 -3.531 -0.377 1.00 0.00 -4.760 -3.355 -0.782 1.00 0.00 -0.024 0.293 3.486 1.00 0.00 0.800 1.123 2.625 1.00 0.00 2.198 0.517 2.565 1.00 0.00 2.727 0.244 1.487 1.00 0.00 0.796 2.584 3.108 1.00 0.00 -0.465 3.341 2.648 1.00 0.00 -0.655 4.621 3.472 1.00 0.00 -0.376 3.739 1.166 1.00 0.00 -0.228 0.624 4.429 1.00 0.00 0.390 1.075 1.624 1.00 0.00 0.866 2.595 4.197 1.00 0.00 1.671 3.100 2.716 1.00 0.00 -1.342 2.708 2.790 1.00 0.00 -0.756 4.378 4.530 1.00 0.00 0.201 5.283 3.340 1.00 0.00 -1.558 5.138 3.146 1.00 0.00 0.480 4.396 1.006 1.00 0.00 -0.274 2.864 0.527 1.00 0.00 -1.284 4.270 0.875 1.00 0.00 2.757 0.245 3.742 1.00 0.00 4.057 -0.380 3.870 1.00 0.00 4.039 -1.768 3.225 1.00 0.00 4.854 -2.060 2.357 1.00 0.00 4.511 -0.402 5.337 1.00 0.00 5.792 -1.226 5.529 1.00 0.00 4.755 1.025 5.848 1.00 0.00 2.190 0.359 4.571 1.00 0.0 0 4.739 0.256 3.329 1.00 0.00 3.723 -0.851 5.931 1.00 0.00 5.604 -2.279 5.317 1.00 0.00 6.577 -0.860 4.867 1.00 0.00 6.130 -1.145 6.563 1.00 0.00 5.555 1.495 5.274 1.00 0.00 3.850 1.625 5.757 1.00 0.00 5.046 0.995 6.898 1.00 0.00 3.109 -2.622 3.650 1.00 0.00
C N H H H H H H H H H H H H H N C C O C C C C H H H H H H H H H H H N C C O C C C H H H H H H H H H N
101
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349
CA PHE A 19 C PHE A 19 O PHE A 19 CB PHE A 19 CG PHE A 19 CD1 PHE A 19 CD2 PHE A 19 CE1 PHE A 19 CE2 PHE A 19 CZ PHE A 19 H PHE A 19 HA PHE A 19 HB2 PHE A 19 HB3 PHE A 19 HD1 PHE A 19 HD2 PHE A 19 HE1 PHE A 19 HE2 PHE A 19 HZ PHE A 19 N PHE A 20 CA PHE A 20 C PHE A 20 O PHE A 20 CB PHE A 20 CG PHE A 20 CD1 PHE A 20 CD2 PHE A 20 CE1 PHE A 20 CE2 PHE A 20 CZ PHE A 20 H PHE A 20 HA PHE A 20 HB2 PHE A 20 HB3 PHE A 20 HD1 PHE A 20 HD2 PHE A 20 HE1 PHE A 20 HE2 PHE A 20 HZ PHE A 20 N A LA A 21 CA A LA A 21 C A LA A 21 O A LA A 21 CB A LA A 21 H A LA A 21 HA A LA A 21 HB1 A LA A 21 HB2 A LA A 21 HB3 A LA A 21 N GLU A 22 CA GLU A 22
2.950 -3.993 3.182 1.00 0.00 2.859 -4.060 1.655 1.00 0.00 3.601 -4.803 1.016 1.00 0.00 1.708 -4.600 3.845 1.00 0.00 1.234 -5.899 3.225 1.00 0.00 1.938 -7.093 3.469 1.00 0.00 0.141 -5.901 2.337 1.00 0.00 1.536 -8.287 2.846 1.00 0.00 -0.257 -7.094 1.712 1.00 0.00 0.439 -8.288 1.968 1.00 0.00 2.468 -2.280 4.353 1.00 0.00 3.820 -4.571 3.496 1.00 0.00 1.915 -4.766 4.903 1.00 0.00 0.899 -3.878 3.780 1.00 0.00 2.790 -7.097 4.134 1.00 0.00 -0.395 -4.987 2.124 1.00 0.00 2.070 -9.205 3.046 1.00 0.00 -1.102 -7.097 1.038 1.00 0.00 0.125 -9.208 1.495 1.00 0.00 1.946 -3.292 1.059 1.00 0.00 1.748 -3.307 -0.379 1.00 0.00 3.014 -2.830 -1.074 1.00 0.00 3.489 -3.461 -2.016 1.00 0.00 0.557 -2.431 -0.776 1.00 0.00 -0.069 -2.813 -2.100 1.00 0.00 -0.914 -3.938 -2.169 1.00 0.00 0.182 -2.055 -3.258 1.00 0.00 -1.525 -4.288 -3.384 1.00 0.00 -0.429 -2.407 -4.474 1.00 0.00 -1.286 -3.520 -4.537 1.00 0.00 1.371 -2.675 1.615 1.00 0.00 1.542 -4.338 -0.659 1.00 0.00 -0.219 -2.545 -0.029 1.00 0.00 0.866 -1.383 -0.788 1.00 0.00 -1.087 -4.544 -1.291 1.00 0.00 0.844 -1.201 -3.222 1.00 0.00 -2.171 -5.153 -3.436 1.00 0.00 -0.240 -1.822 -5.363 1.00 0.00 -1.755 -3.790 -5.472 1.00 0.00 3.575 -1.717 -0.592 1.00 0.00 4.801 -1.192 -1.160 1.00 0.00 5.929 -2.214 -1.064 1.00 0.00 6.729 -2.318 -1.983 1.00 0.00 5.209 0.085 -0.424 1.00 0.00 3.165 -1.228 0.200 1.00 0.00 4.592 -1.001 -2.221 1.00 0.00 4.424 0.833 -0.509 1.00 0.00 5.385 -0.138 0.630 1.00 0.00 6.133 0.470 -0.855 1.00 0.00 5.999 -2.972 0.032 1.00 0.00 6.920 -4.085 0.187 1.00 0.00
C C O C C C C C C C H H H H H H H H H N C C O C C C C C C C H H H H H H H H H N C C O C H H H H H N C
102
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
350 C GLU A 22 351 O GLU A 22 352 CB GLU A 22 353 CG GLU A 22 354 CD GLU A 22 355 OE1 GLU A 22 356 OE2 GLU A 22 357 H GLU A 22 358 HA GLU A 22 359 HB2 GLU A 22 360 HB3 GLU A 22 361 HG2 GLU A 22 362 HG3 GLU A 22 363 HE2 GLU A 22 364 N ASP A 23 365 CA A SP A 23 366 C ASP A 23 367 O ASP A 23 368 CB ASP A 23 369 CG A SP A 23 370 OD1 ASP A 23 371 OD2 ASP A 23 372 H ASP A 23 373 HA ASP A 23 374 HB2 ASP A 23 375 HB3 ASP A 23 376 HD2 ASP A 23 377 N VA L A 24 378 CA VA L A 24 379 C VA L A 24 380 O VA L A 24 381 CB VA L A 24 382 CG1 VA L A 24 383 CG2 VA L A 24 384 H VA L A 24 385 HA VA L A 24 386 HB VA L A 24 387 1HG1 VA L A 24 388 2HG1 VA L A 24 389 3HG1 VA L A 24 390 1HG2 VA L A 24 391 2HG2 VA L A 24 392 3HG2 VA L A 24 393 N GLY A 25 394 CA GLY A 25 395 C GLY A 25 396 O GLY A 25 397 H GLY A 25 398 HA2 GLY A 25 399 HA3 GLY A 25 400 N SER A 26
6.676 -5.118 -0.899 1.00 0.00 7.594 -5.523 -1.605 1.00 0.00 6.736 -4.719 1.580 1.00 0.00 7.876 -4.423 2.555 1.00 0.00 9.104 -5.281 2.287 1.00 0.00 8.980 -6.501 2.261 1.00 0.00 10.300 -4.652 2.407 1.00 0.00 5.344 -2.792 0.781 1.00 0.00 7.925 -3.717 0.023 1.00 0.00 5.820 -4.357 2.039 1.00 0.00 6.628 -5.800 1.487 1.00 0.00 8.137 -3.364 2.530 1.00 0.00 7.519 -4.683 3.550 1.00 0.00 11.004 -5.300 2.333 1.00 0.00 5.424 -5.549 -1.007 1.00 0.00 5.016 -6.579 -1.946 1.00 0.00 5.446 -6.206 -3.363 1.00 0.00 6.046 -7.014 -4.065 1.00 0.00 3.500 -6.791 -1.857 1.00 0.00 3.074 -8.008 -2.662 1.00 0.00 2.842 -7.877 -3.859 1.00 0.00 2.830 -9.142 -1.948 1.00 0.00 4.752 -5.132 -0.373 1.00 0.00 5.546 -7.492 -1.669 1.00 0.00 3.193 -6.899 -0.817 1.00 0.00 2.984 -5.927 -2.274 1.00 0.00 3.031 -9.055 -1.015 1.00 0.00 5.159 -4.968 -3.762 1.00 0.00 5.574 -4.430 -5.047 1.00 0.00 7.094 -4.290 -5.119 1.00 0.00 7.730 -4.705 -6.087 1.00 0.00 4.844 -3.099 -5.300 1.00 0.00 5.397 -2.397 -6.545 1.00 0.00 3.336 -3.338 -5.468 1.00 0.00 4.642 -4.365 -3.129 1.00 0.00 5.300 -5.133 -5.818 1.00 0.00 4.995 -2.441 -4.445 1.00 0.00 6.388 -1.990 -6.342 1.00 0.00 5.473 -3.115 -7.359 1.00 0.00 4.742 -1.577 -6.838 1.00 0.00 3.152 -4.004 -6.310 1.00 0.00 2.913 -3.786 -4.570 1.00 0.00 2.829 -2.390 -5.648 1.00 0.00 7.667 -3.672 -4.093 1.00 0.00 9.062 -3.277 -4.031 1.00 0.00 9.941 -4.497 -4.241 1.00 0.00 10.753 -4.540 -5.164 1.00 0.00 7.079 -3.466 -3.296 1.00 0.00 9.265 -2.532 -4.801 1.00 0.00 9.272 -2.844 -3.052 1.00 0.00 9.713 -5.536 -3.439 1.00 0.00
C O C C C O O H H H H H H H N C C O C C O O H H H H H N C C O C C C H H H H H H H H H N C C O H H H N
103
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
401 CA SER A 26 402 C SER A 26 403 O SER A 26 404 CB SER A 26 405 OG SER A 26 406 H SER A 26 407 HA SER A 26 408 HB2 SER A 26 409 HB3 SER A 26 410 HG SER A 26 411 N ASN A 27 412 CA A SN A 27 413 C ASN A 27 414 O ASN A 27 415 CB ASN A 27 416 CG A SN A 27 417 OD1 ASN A 27 418 ND2 ASN A 27 419 H ASN A 27 420 HA ASN A 27 421 HB2 ASN A 27 422 HB3 ASN A 27 423 1HD2 ASN A 27 424 2HD2 ASN A 27 425 N LYS A 28 426 CA LYS A 28 427 C LYS A 28 428 O LYS A 28 429 CB LYS A 28 430 CG LYS A 28 431 CD LYS A 28 432 CE LYS A 28 433 NZ LYS A 28 434 H LYS A 28 435 HA LYS A 28 436 HB2 LYS A 28 437 HB3 LYS A 28 438 HG2 LYS A 28 439 HG3 LYS A 28 440 HD2 LYS A 28 441 HD3 LYS A 28 442 HE2 LYS A 28 443 HE3 LYS A 28 444 HZ1 LYS A 28 445 HZ2 LYS A 28 446 HZ3 LYS A 28 447 N GLY A 29 448 CA GLY A 29 449 C GLY A 29 450 O GLY A 29 451 H GLY A 29
10.423 -6.794 -3.545 1.00 0.00 9.845 -7.688 -4.643 1.00 0.00 9.930 -8.910 -4.546 1.00 0.00 10.410 -7.484 -2.178 1.00 0.00 10.872 -6.591 -1.179 1.00 0.00 8.980 -5.481 -2.738 1.00 0.00 11.446 -6.574 -3.835 1.00 0.00 9.390 -7.793 -1.938 1.00 0.00 11.045 -8.371 -2.212 1.00 0.00 11.706 -6.209 -1.464 1.00 0.00 9.286 -7.087 -5.696 1.00 0.00 8.762 -7.790 -6.853 1.00 0.00 8.736 -6.864 -8.074 1.00 0.00 7.900 -6.999 -8.968 1.00 0.00 7.380 -8.361 -6.513 1.00 0.00 6.942 -9.450 -7.481 1.00 0.00 7.749 -10.014 -8.213 1.00 0.00 5.652 -9.774 -7.480 1.00 0.00 9.205 -6.080 -5.695 1.00 0.00 9.453 -8.596 -7.084 1.00 0.00 7.403 -8.812 -5.520 1.00 0.00 6.651 -7.552 -6.511 1.00 0.00 5.015 -9.309 -6.851 1.00 0.00 5.337 -10.506 -8.097 1.00 0.00 9.702 -5.941 -8.144 1.00 0.00 9.855 -5.015 -9.254 1.00 0.00 10.034 -5.788 -10.549 1.00 0.00 9.467 -5.425 -11.570 1.00 0.00 11.029 -4.044 -9.053 1.00 0.00 10.597 -2.732 -8.383 1.00 0.00 11.669 -1.642 -8.571 1.00 0.00 11.037 -0.337 -9.076 1.00 0.00 12.058 0.673 -9.405 1.00 0.00 10.368 -5.904 -7.399 1.00 0.00 8.930 -4.463 -9.335 1.00 0.00 11.836 -4.515 -8.491 1.00 0.00 11.406 -3.795 -10.047 1.00 0.00 9.664 -2.398 -8.839 1.00 0.00 10.410 -2.906 -7.321 1.00 0.00 12.180 -1.467 -7.623 1.00 0.00 12.413 -1.964 -9.301 1.00 0.00 10.465 -0.536 -9.984 1.00 0.00 10.366 0.067 -8.317 1.00 0.00 12.606 0.890 -8.585 1.00 0.00 12.665 0.319 -10.131 1.00 0.00 11.604 1.513 -9.737 1.00 0.00 10.790 -6.881 -10.480 1.00 0.00 10.958 -7.825 -11.574 1.00 0.00 9.625 -8.200 -12.230 1.00 0.00 9.589 -8.458 -13.430 1.00 0.00 11.206 -7.076 -9.585 1.00 0.00
C C O C O H H H H H N C C O C C O N H H H H H H N C C O C C C C N H H H H H H H H H H H H H N C C O H
104
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
452 HA2 GLY A 29 453 HA3 GLY A 29 454 N A LA A 30 455 CA A LA A 30 456 C A LA A 30 457 O A LA A 30 458 CB A LA A 30 459 H A LA A 30 460 HA A LA A 30 461 HB1 A LA A 30 462 HB2 A LA A 30 463 HB3 A LA A 30 464 N ILE A 31 465 CA ILE A 31 466 C ILE A 31 467 O ILE A 31 468 CB ILE A 31 469 CG1 ILE A 31 470 CG2 ILE A 31 471 CD1 ILE A 31 472 H ILE A 31 473 HA ILE A 31 474 HB ILE A 31 475 2HG1 ILE A 31 476 3HG1 ILE A 31 477 1HG2 ILE A 31 478 2HG2 ILE A 31 479 3HG2 ILE A 31 480 1HD1 ILE A 31 481 2HD1 ILE A 31 482 3HD1 ILE A 31 483 N ILE A 32 484 CA ILE A 32 485 C ILE A 32 486 O ILE A 32 487 CB ILE A 32 488 CG1 ILE A 32 489 CG2 ILE A 32 490 CD1 ILE A 32 491 H ILE A 32 492 HA ILE A 32 493 HB ILE A 32 494 2HG1 ILE A 32 495 3HG1 ILE A 32 496 1HG2 ILE A 32 497 2HG2 ILE A 32 498 3HG2 ILE A 32 499 1HD1 ILE A 32 500 2HD1 ILE A 32 501 3HD1 ILE A 32 502 N GLY A 33
11.617 -7.387 -12.323 1.00 0.00 11.423 -8.732 -11.188 1.00 0.00 8.536 -8.220 -11.451 1.00 0.00 7.200 -8.500 -11.952 1.00 0.00 6.466 -7.198 -12.276 1.00 0.00 5.980 -7.012 -13.392 1.00 0.00 6.432 -9.342 -10.935 1.00 0.00 8.613 -7.912 -10.488 1.00 0.00 7.281 -9.091 -12.859 1.00 0.00 6.987 -10.256 -10.720 1.00 0.00 6.298 -8.777 -10.016 1.00 0.00 5.453 -9.604 -11.334 1.00 0.00 6.384 -6.280 -11.304 1.00 0.00 5.631 -5.035 -11.458 1.00 0.00 6.218 -4.116 -12.544 1.00 0.00 5.575 -3.164 -12.990 1.00 0.00 5.351 -4.379 -10.087 1.00 0.00 4.197 -5.098 -9.356 1.00 0.00 4.918 -2.916 -10.246 1.00 0.00 4.550 -6.479 -8.803 1.00 0.00 6.815 -6.479 -10.407 1.00 0.00 4.681 -5.332 -11.875 1.00 0.00 6.251 -4.398 -9.471 1.00 0.00 3.897 -4.493 -8.504 1.00 0.00 3.333 -5.188 -10.015 1.00 0.00 5.754 -2.305 -10.586 1.00 0.00 4.105 -2.874 -10.967 1.00 0.00 4.561 -2.499 -9.307 1.00 0.00 4.663 -7.197 -9.611 1.00 0.00 5.468 -6.418 -8.221 1.00 0.00 3.745 -6.821 -8.153 1.00 0.00 7.410 -4.458 -13.024 1.00 0.00 8.081 -3.937 -14.206 1.00 0.00 7.100 -3.523 -15.307 1.00 0.00 7.286 -2.474 -15.912 1.00 0.00 9.163 -4.936 -14.686 1.00 0.00 10.542 -4.293 -14.462 1.00 0.00 9.002 -5.407 -16.142 1.00 0.00 11.716 -5.181 -14.884 1.00 0.00 7.908 -5.145 -12.474 1.00 0.00 8.585 -3.034 -13.875 1.00 0.00 9.110 -5.828 -14.063 1.00 0.00 10.595 -3.355 -15.013 1.00 0.00 10.647 -4.078 -13.398 1.00 0.00 8.028 -5.875 -16.285 1.00 0.00 9.115 -4.570 -16.832 1.00 0.00 9.751 -6.162 -16.377 1.00 0.00 11.600 -6.180 -14.465 1.00 0.00 11.774 -5.244 -15.971 1.00 0.00 12.645 -4.744 -14.516 1.00 0.00 6.048 -4.311 -15.556 1.00 0.00
H H N C C O C H H H H H N C C O C C C C H H H H H H H H H H H N C C O C C C C H H H H H H H H H H H N
105
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
503 CA GLY A 33 504 C GLY A 33 505 O GLY A 33 506 H GLY A 33 507 HA2 GLY A 33 508 HA3 GLY A 33 509 N LEU A 34 510 CA LEU A 34 511 C LEU A 34 512 O LEU A 34 513 CB LEU A 34 514 CG LEU A 34 515 CD1 LEU A 34 516 CD2 LEU A 34 517 H LEU A 34 518 HA LEU A 34 519 HB2 LEU A 34 520 HB3 LEU A 34 521 HG LEU A 34 522 1HD1 LEU A 34 523 2HD1 LEU A 34 524 3HD1 LEU A 34 525 1HD2 LEU A 34 526 2HD2 LEU A 34 527 3HD2 LEU A 34 528 N M ET A 35 529 CA M ET A 35 530 C M ET A 35 531 O M ET A 35 532 CB M ET A 35 533 CG M ET A 35 534 SD M ET A 35 535 CE M ET A 35 536 H M ET A 35 537 HA M ET A 35 538 HB2 M ET A 35 539 HB3 M ET A 35 540 HG2 M ET A 35 541 HG3 M ET A 35 542 HE1 M ET A 35 543 HE2 M ET A 35 544 HE3 M ET A 35 545 N VA L A 36 546 CA VA L A 36 547 C VA L A 36 548 O VA L A 36 549 CB VA L A 36 550 CG1 VA L A 36 551 CG2 VA L A 36 552 H VA L A 36 553 HA VA L A 36
5.001 -3.945 -16.503 1.00 0.00 4.509 -2.508 -16.291 1.00 0.00 4.512 -1.697 -17.216 1.00 0.00 5.950 -5.167 -15.023 1.00 0.00 5.387 -4.049 -17.518 1.00 0.00 4.159 -4.626 -16.377 1.00 0.00 4.104 -2.173 -15.064 1.00 0.00 3.613 -0.871 -14.697 1.00 0.00 4.734 0.139 -14.768 1.00 0.00 4.540 1.250 -15.247 1.00 0.00 2.907 -0.967 -13.332 1.00 0.00 3.481 -0.201 -12.121 1.00 0.00 3.563 1.329 -12.269 1.00 0.00 2.573 -0.464 -10.910 1.00 0.00 4.260 -2.775 -14.273 1.00 0.00 2.902 -0.592 -15.458 1.00 0.00 1.885 -0.674 -13.509 1.00 0.00 2.856 -2.016 -13.037 1.00 0.00 4.473 -0.605 -11.916 1.00 0.00 2.722 1.699 -12.856 1.00 0.00 3.536 1.800 -11.285 1.00 0.00 4.498 1.640 -12.728 1.00 0.00 1.631 0.071 -11.033 1.00 0.00 2.352 -1.524 -10.803 1.00 0.00 3.063 -0.115 -10.000 1.00 0.00 5.912 -0.262 -14.307 1.00 0.00 7.083 0.596 -14.307 1.00 0.00 7.327 1.111 -15.730 1.00 0.00 7.407 2.315 -15.961 1.00 0.00 8.287 -0.161 -13.724 1.00 0.00 9.223 0.790 -12.973 1.00 0.00 8.557 1.382 -11.381 1.00 0.00 8.687 3.171 -11.636 1.00 0.00 5.957 -1.218 -13.986 1.00 0.00 6.850 1.449 -13.670 1.00 0.00 7.945 -0.910 -13.009 1.00 0.00 8.839 -0.662 -14.521 1.00 0.00 10.145 0.255 -12.749 1.00 0.00 9.464 1.636 -13.614 1.00 0.00 9.729 3.446 -11.790 1.00 0.00 8.093 3.460 -12.503 1.00 0.00 8.303 3.675 -10.750 1.00 0.00 7.364 0.196 -16.698 1.00 0.00 7.510 0.528 -18.108 1.00 0.00 6.291 1.332 -18.576 1.00 0.00 6.427 2.363 -19.240 1.00 0.00 7.749 -0.745 -18.929 1.00 0.00 7.783 -0.459 -20.437 1.00 0.00 9.076 -1.412 -18.536 1.00 0.00 7.199 -0.771 -16.438 1.00 0.00 8.401 1.126 -18.228 1.00 0.00
C C O H H H N C C O C C C C H H H H H H H H H H H N C C O C C S C H H H H H H H H H N C C O C C C H H
106
ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM ATOM
554 HB VA L A 36 555 1HG1 VA L A 36 556 2HG1 VA L A 36 557 3HG1 VA L A 36 558 1HG2 VA L A 36 559 2HG2 VA L A 36 560 3HG2 VA L A 36 561 N GLY A 37 562 CA GLY A 37 563 C GLY A 37 564 O GLY A 37 565 H GLY A 37 566 HA2 GLY A 37 567 HA3 GLY A 37 568 N GLY A 38 569 CA GLY A 38 570 C GLY A 38 571 O GLY A 38 572 H GLY A 38 573 HA2 GLY A 38 574 HA3 GLY A 38 575 N VA L A 39 576 CA VA L A 39 577 C VA L A 39 578 O VA L A 39 579 CB VA L A 39 580 CG1 VA L A 39 581 CG2 VA L A 39 582 H VA L A 39 583 HA VA L A 39 584 HB VA L A 39 585 1HG1 VA L A 39 586 2HG1 VA L A 39 587 3HG1 VA L A 39 588 1HG2 VA L A 39 589 2HG2 VA L A 39 590 3HG2 VA L A 39 591 N VA L A 40 592 CA VA L A 40 593 C VA L A 40 594 O VA L A 40 595 CB VA L A 40 596 CG1 VA L A 40 597 CG2 VA L A 40 598 H VA L A 40 599 HA VA L A 40 600 HB VA L A 40 601 1HG1 VA L A 40 602 2HG1 VA L A 40 603 3HG1 VA L A 40 604 1HG2 VA L A 40
6.933 -1.423 -18.716 1.00 0.00 6.818 -0.088 -20.781 1.00 0.00 8.553 0.281 -20.660 1.00 0.00 8.007 -1.378 -20.978 1.00 0.00 9.911 -0.756 -18.784 1.00 0.00 9.108 -1.628 -17.470 1.00 0.00 9.189 -2.350 -19.080 1.00 0.00 5.095 0.883 -18.184 1.00 0.00 3.828 1.505 -18.534 1.00 0.00 3.766 2.957 -18.067 1.00 0.00 3.077 3.772 -18.671 1.00 0.00 5.058 0.095 -17.548 1.00 0.00 3.690 1.462 -19.615 1.00 0.00 3.023 0.948 -18.054 1.00 0.00 4.491 3.281 -16.998 1.00 0.00 4.601 4.622 -16.455 1.00 0.00 5.796 5.386 -17.021 1.00 0.00 5.837 6.604 -16.890 1.00 0.00 4.949 2.523 -16.500 1.00 0.00 3.694 5.190 -16.666 1.00 0.00 4.722 4.544 -15.377 1.00 0.00 6.764 4.699 -17.639 1.00 0.00 7.913 5.344 -18.273 1.00 0.00 7.521 5.891 -19.647 1.00 0.00 8.031 6.929 -20.063 1.00 0.00 9.105 4.368 -18.332 1.00 0.00 10.159 4.751 -19.380 1.00 0.00 9.794 4.315 -16.962 1.00 0.00 6.673 3.689 -17.720 1.00 0.00 8.221 6.204 -17.674 1.00 0.00 8.742 3.374 -18.586 1.00 0.00 9.751 4.654 -20.387 1.00 0.00 10.494 5.777 -19.223 1.00 0.00 11.016 4.082 -19.298 1.00 0.00 10.312 5.255 -16.772 1.00 0.00 9.064 4.157 -16.171 1.00 0.00 10.518 3.500 -16.945 1.00 0.00 6.654 5.184 -20.380 1.00 0.00 6.247 5.614 -21.707 1.00 0.00 5.586 7.005 -21.701 1.00 0.00 5.836 7.799 -22.606 1.00 0.00 5.432 4.524 -22.428 1.00 0.00 4.187 4.087 -21.655 1.00 0.00 5.026 4.974 -23.835 1.00 0.00 6.386 4.262 -20.072 1.00 0.00 7.181 5.691 -22.251 1.00 0.00 6.075 3.648 -22.534 1.00 0.00 4.486 3.681 -20.693 1.00 0.00 3.516 4.930 -21.508 1.00 0.00 3.661 3.308 -22.207 1.00 0.00 4.303 5.789 -23.782 1.00 0.00
H H H H H H H N C C O H H H N C C O H H H N C C O C C C H H H H H H H H H N C C O C C C H H H H H H H
107
ATOM 605 2HG2 VA L A 40 5.906 5.310 -24.383 1.00 0.00 ATOM 606 3HG2 VA L A 40 4.573 4.139 -24.369 1.00 0.00 ATOM 607 N ILE A 41 4.748 7.307 -20.701 1.00 0.00 ATOM 608 CA ILE A 41 4.202 8.648 -20.486 1.00 0.00 ATOM 609 C ILE A 41 5.081 9.342 -19.435 1.00 0.00 ATOM 610 O ILE A 41 5.998 8.736 -18.891 1.00 0.00 ATOM 611 CB ILE A 41 2.715 8.572 -20.076 1.00 0.00 ATOM 612 CG1 ILE A 41 1.923 7.544 -20.913 1.00 0.00 ATOM 613 CG2 ILE A 41 1.960 9.913 -20.147 1.00 0.00 ATOM 614 CD1 ILE A 41 1.614 6.311 -20.066 1.00 0.00 ATOM 615 H ILE A 41 4.635 6.642 -19.951 1.00 0.00 ATOM 616 HA ILE A 41 4.239 9.223 -21.405 1.00 0.00 ATOM 617 HB ILE A 41 2.723 8.270 -19.035 1.00 0.00 ATOM 618 2HG1 ILE A 41 0.969 7.952 -21.248 1.00 0.00 ATOM 619 3HG1 ILE A 41 2.476 7.256 -21.808 1.00 0.00 ATOM 620 1HG2 ILE A 41 2.269 10.588 -19.353 1.00 0.00 ATOM 621 2HG2 ILE A 41 2.112 10.383 -21.119 1.00 0.00 ATOM 622 3HG2 ILE A 41 0.892 9.747 -19.998 1.00 0.00 ATOM 623 1HD1 ILE A 41 2.538 5.937 -19.632 1.00 0.00 ATOM 624 2HD1 ILE A 41 0.928 6.575 -19.261 1.00 0.00 ATOM 625 3HD1 ILE A 41 1.160 5.536 -20.684 1.00 0.00 ATOM 626 N A LA A 42 4.799 10.609 -19.131 1.00 0.00 ATOM 627 CA A LA A 42 5.507 11.384 -18.116 1.00 0.00 ATOM 628 C A LA A 42 7.006 11.477 -18.430 1.00 0.00 ATOM 629 O A LA A 42 7.318 11.704 -19.620 1.00 0.00 ATOM 630 CB A LA A 42 5.219 10.817 -16.716 1.00 0.00 ATOM 631 OXT A LA A 42 7.809 11.369 -17.478 1.00 0.00 ATOM 632 H A LA A 42 4.062 11.060 -19.645 1.00 0.00 ATOM 633 HA A LA A 42 5.108 12.398 -18.145 1.00 0.00 ATOM 634 HB1 A LA A 42 5.625 11.486 -15.957 1.00 0.00 ATOM 635 HB2 A LA A 42 4.143 10.730 -16.566 1.00 0.00 ATOM 636 HB3 A LA A 42 5.675 9.834 -16.595 1.00 0.00 TER 637 A LA A 42
H H N C C O C C C C H H H H H H H H H H H N C C O C O H H H H H
108
APPENDIX B Structure of Molecule HFIP ALL ATOM STRUCTURE FOR M OLECULE LIG 12 0hfi C1 1 0.127 0.012 -0.009 0hfi F1 2 0.233 -0.046 -0.067 0hfi F2 3 0.144 0.007 0.124 0hfi F3 4 0.125 0.142 -0.046 0hfi C2 5 -0.002 -0.058 -0.053 0hfi H2 6 -0.005 -0.054 -0.163 0hfi O1 7 0.005 -0.189 -0.003 0hfi H1 8 -0.057 -0.245 -0.053 0hfi C3 9 -0.129 0.012 -0.002 0hfi F4 10 -0.129 0.027 0.132 0hfi F5 11 -0.145 0.133 -0.058 0hfi F6 12 -0.237 -0.063 -0.034 0.00000 0.00000 0.00000
109
APPENDIX C Parameter for HFIP [ mo leculetype ] ; Name nrexcl hfi 3 [ atoms ] ; nr type resnr resid atom cgnr charge mass 1 CH0 1 hfi C1 1 0.384 12.0110 2 F 1 h fi F1 1 -0.128 18.9984 3 F 1 h fi F2 1 -0.128 18.9984 4 F 1 h fi F3 1 -0.128 18.9984 ; 5 C 1 hfi C2 2 -0.036 12.0110 6 HC 1 hfi H2 2 0.155 1.0080 7 OA 1 hfi O1 2 -0.566 15.9994 8 H 1 hfi H1 2 0.447 1.0080 ; 9 CH0 1 hfi C3 3 0.384 12.0110 10 F 1 hfi F4 3 -0.128 18.9984 11 F 1 hfi F5 3 -0.128 18.9984 12 F 1 hfi F6 3 -0.128 18.9984 ; ; total charge of the mo lecule: 0.000 [ bonds ] ; ai aj funct c0 1 2 2 0.1360 1 3 2 0.1360 1 4 2 0.1360 1 5 2 0.1530 5 6 2 0.1100 5 7 2 0.1430 5 9 2 0.1530 7 8 2 0.1000 9 10 2 0.1360 9 11 2 0.1360 9 12 2 0.1360
total_charge
0.000
0.000
0.000
c1 7.2300e+06 4.7700e+06 7.2300e+06 7.1500e+06 1.2100e+07 8.1800e+06 7.1500e+06 1.8700e+07 4.7700e+06 4.7700e+06 7.2300e+06
[ pairs ] ; ai aj funct ; all 1-4 pairs but the ones excluded in GROMOS itp 1 8 1 1 10 1 1 11 1 1 12 1
110
2 2 2 3 3 3 4 4 4 6 6 6 6 7 7 7 8
6 7 9 6 7 9 6 7 9 8 10 11 12 10 11 12 9
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
[ angles ] ; ai aj ak funct angle fc 2 1 3 2 107.60 507.00 2 1 4 2 107.60 507.00 2 1 5 2 111.40 532.00 3 1 4 2 107.60 507.00 3 1 5 2 111.40 532.00 4 1 5 2 111.40 532.00 1 5 6 2 109.50 285.00 1 5 7 2 109.50 320.00 1 5 9 2 111.30 632.00 6 5 7 2 115.00 460.00 6 5 9 2 106.75 503.00 7 5 9 2 110.30 524.00 5 7 8 2 108.53 443.00 5 9 10 2 111.40 532.00 5 9 11 2 111.40 532.00 5 9 12 2 109.50 618.00 10 9 11 2 107.60 507.00 10 9 12 2 107.60 507.00 11 9 12 2 107.60 507.00 [ dihedrals ] ; GROM OS imp roper dihedrals ; ai aj ak al funct angle fc
111
[ dihedrals ] ; ai aj ak 2 1 5 1 5 7 1 5 9
al funct ph0 cp mult 9 1 0.00 3.77 3 8 1 0.00 1.26 3 10 1 0.00 3.77 3
[ exclusions ] ; ai aj funct ; GROMOS 1-4 exclusions
112
APPENDIX D Script for Pack mol
# A combination of 6 p rotein # tolerance 2.0 filetype pdb output aggregatepack.pdb structure aggregatemode2.pdb number 6 inside bo x 0. 0. 0. 40. 40. 40. end structure
113
APPENDIX E Topology for Aβ in water ; File 'water.top' was generated ; By user: onbekend (0) ; On host: onbekend ; At date: Thu Apr 4 09:10:40 2013 ; ; This is a standalone topology file ; ; It was generated using program: ; pdb2gmx - VERSION 4.5.5 ; ; Co mmand line was: ; pdb2gmx -f 1IYT_edit.pdb -o protein.gro -p water.top -ignh ; ; Force field was read fro m the standard Gro macs share directory. ; Include forcefield parameters #include "gromos53a6.ff/forcefield.itp" ; Include topology for protein #include "protein.itp" ; Include Position restraint file #ifdef POSRES #include "posre.itp" #endif ; Include water topology #include "gromos53a6.ff/spc.itp" #ifdef POSRES_WATER ; Position restraint for each water o xygen [ position_restraints ] ; i funct fcx fcy 1 1 1000 1000 #endif
fcz 1000
; Include topology for ions #include "gromos53a6.ff/ions.itp" [ system ] ; Name ALZHEIM ER'S DISEASE AM YLOID in water [ mo lecules ] ; Co mpound # mols Protein_chain_A 1 SOL 8777 NA 3
114
APPENDIX F Topology for Aβ in Sol vent Mi xture ; File 'control.top' was generated ; By user: onbekend (0) ; On host: onbekend ; At date: Mon Jun 24 12:55:18 2013 ; ; This is a standalone topology file ; ; It was generated using program: ; pdb2gmx - VERSION 4.5.5 ; ; Co mmand line was: ; pdb2gmx -f 1IYT_edit.pdb -o protein.pdb -p control.top -ignh ; ; Force field was read fro m the standard Gro macs share directory. ; ; Include forcefield parameters #include "gromos53a6.ff/forcefield.itp" ; Include topology for protein #include "protein.itp" ; Include Position restraint file #ifdef POSRES #include "posre.itp" #endif ; Include HFIP topology #include "hfi.itp" ; Include Position restraint file #ifdef POSRES_hfi #include "posre_hfi.itp" #endif ; Include water topology #include "gromos53a6.ff/spc.itp" #ifdef POSRES_WATER ; Position restraint for each water o xygen [ position_restraints ]
115
; i funct 1 1 #endif
fcx 1000
fcy 1000
fcz 1000
; Include topology for ions #include "gromos53a6.ff/ions.itp" [ system ] ; Name ALZHEIM ER'S DISEASE AM YLOID in water [ mo lecules ] ; Co mpound # mols Protein_chain_A 1 hfi 711 SOL 4063 NA 3
116
APPENDIX G Topology for Aβ wi th Zinc in Water ; File 'metal.top' was generated ; ; By user: onbekend (0) ; On host: onbekend ; ; At date: Wed May 8 09:52:29 2013 ; ; This is a standalone topology file ; ; It was generated using program: ; pdb2gmx - VERSION 4.5.5 ; ; Co mmand line was: ; pdb2gmx -f 1IYT_edit_zn.docked_2.pdb -o protein.gro -p metal.top -ignh ; ; Force field was read fro m the standard Gro macs share directory. ; ; Include forcefield parameters #include "gromos53a6.ff/forcefield.itp" ; Include chain topologies #include "metal_Protein_chain_A.itp" ; Include position restraint for protein #ifdef POSRES #include "posre_Protein_chain_A.itp" #endif ; Include ion chain topologies #include "metal_Ion_chain_A2.itp" ; Include position restraint for metal #ifdef POSRES_ION #include "posre_Ion_chain_A2.itp" #endif ; Include water topology #include "gromos53a6.ff/spc.itp" #ifdef POSRES_WATER ; Position restraint for each water o xygen [ position_restraints ] ; i funct fcx fcy fcz 1 1 1000 1000 1000 #endif
117
; Include topology for ions #include "gromos53a6.ff/ions.itp" [ system ] ; Name Protein in water [ mo lecules ] ; Co mpound # mols Protein_chain_A 1 Ion_chain_A2 1 SOL 8780 NA
1
118
APPENDIX H Topology of Aβ wi th Zinc in S ol vent Mi xture ; File 'mixmetal.top' was generated ; ; By user: onbekend (0) ; On host: onbekend ; ; ; At date: Thu Jul 18 12:42:02 2013 ; ; This is a standalone topology file ; ; It was generated using program: ; pdb2gmx - VERSION 4.5.5 ; ; Co mmand line was: ; pdb2gmx -f 1IYT_edit_zn.docked_2.pdb -o protein.gro -p mixmetal.top -ignh ; ; Force field was read fro m the standard Gro macs share directory. ; ; Include forcefield parameters #include "gromos53a6.ff/forcefield.itp" ; Include chain topologies #include "mixmetal_Protein_chain_A.itp" ; Include position restraint for protein #ifdef POSRES #include "posre_Protein_chain_A.itp" #endif ; Include position restraint for metal #ifdef POSRES_ION #include "posre_Ion_chain_A2.itp" #endif ; Include ion chain topologies #include "mixmetal_Ion_chain_A2.itp" ; Include hfi topologies #include "hfi.itp" ; Include water topology #include "gromos53a6.ff/spc.itp" #ifdef POSRES_WATER ; Position restraint for each water o xygen [ position_restraints ]
119
; i funct 1 1 #endif
fcx 1000
fcy 1000
fcz 1000
; Include topology for ions #include "gromos53a6.ff/ions.itp" [ system ] ; Name Protein in water [ mo lecules ] ; Co mpound # mols Protein_chain_A 1 Ion_chain_A2 1 hfi 661 SOL 4351 NA 1
120
APPENDIX I Topology for Aggregation part, 6Aβ with 6 Zinc i n water ; File 'aggregate.top' was generated ; By user: onbekend (0) ; On host: onbekend ; At date: Wed Dec 3 08:20:43 2014 ; This is a standalone topology file ; It was generated using program: ; pdb2gmx - VERSION 4.5.5 ; Co mmand line was: ; pdb2gmx -f aggregatepack_new.pdb -o aggregate.gro -p aggregate.top -ignh ; Force field was read fro m the standard Gro macs share directory. ; ; Include forcefield parameters #include "gromos53a6.ff/forcefield.itp" ; Include chain topologies #include "aggregate_Protein_chain_A.itp" #ifdef POSRES #include "posre_Protein_chain_A.itp" #endif ; Include chain topologies #include "aggregate_Protein_chain_B.itp" #ifdef POSRES #include "posre_Protein_chain_B.itp" #endif ; Include chain topologies #include "aggregate_Protein_chain_C.itp" #ifdef POSRES #include "posre_Protein_chain_C.itp" #endif ; Include chain topologies #include "aggregate_Protein_chain_D.itp" #ifdef POSRES #include "posre_Protein_chain_D.itp" #endif ; Include chain topologies #include "aggregate_Protein_chain_E.itp" #ifdef POSRES #include "posre_Protein_chain_E.itp" #endif ; Include chain topologies #include "aggregate_Protein_chain_F.itp" #ifdef POSRES #include "posre_Protein_chain_F.itp"
121
#endif ; Include ligand topologies #include "aggregate_Ion_chain_A2.itp" #ifdef POSRES #include "posre_Ion_chain_A2.itp" #endif
; Include ligand topologies #include "aggregate_Ion_chain_B2.itp" #ifdef POSRES #include "posre_Ion_chain_B2.itp" #endif ; Include ligand topologies #include "aggregate_Ion_chain_C2.itp" #ifdef POSRES #include "posre_Ion_chain_C2.itp" #endif ; Include ligand topologies #include "aggregate_Ion_chain_D2.itp" #ifdef POSRES #include "posre_Ion_chain_D2.itp" #endif ; Include ligand topologies #include "aggregate_Ion_chain_E2.itp" #ifdef POSRES #include "posre_Ion_chain_E2.itp" #endif ; Include ligand topologies #include "aggregate_Ion_chain_F2.itp" #ifdef POSRES #include "posre_Ion_chain_F2.itp" #endif ; Include water topology #include "gromos53a6.ff/spc.itp" #ifdef POSRES_WATER ; Position restraint for each water o xygen [ position_restraints ] ; i funct fcx fcy fcz 1 1 1000 1000 1000 #endif ; Include topology for ions #include "gromos53a6.ff/ions.itp"
122
[ system ] ; Name Built with Packmo l in water [ mo lecules ] ; Co mpound # mols Protein_chain_A 1 Ion_chain_A2 1 Protein_chain_B 1 Ion_chain_B2 1 Protein_chain_C 1 Ion_chain_C2 1 Protein_chain_D 1 Ion_chain_D2 1 Protein_chain_E 1 Ion_chain_E2 1 Protein_chain_F 1 Ion_chain_F2 1 SOL 12142 NA 6
123
APPENDIX J Topology for Aggregation part, 6Aβ with 6 Zinc i n Sol vent Mixture ; ; ; ; ; ; ; ; ; -ignh ;
File 'mixaggregate.top' was generated By user: onbekend (0) On host: onbekend At date: Wed Dec 3 08:42:37 2014 This is a standalone topology file It was generated using program: pdb2gmx - VERSION 4.5.5 Co mmand line was: pdb2gmx -f aggregatepack_new.pdb -o mixaggregate.gro -p mixaggregate.top Force field was read fro m the standard Gro macs share directory.
; Include forcefield parameters #include "gromos53a6.ff/forcefield.itp" ; Include chain topologies #include "mixaggregate_Protein_chain_A.itp" #ifdef POSRES #include "posre_Protein_chain_A.itp" #endif ; Include chain topologies #include "mixaggregate_Protein_chain_B.itp" #ifdef POSRES #include "posre_Protein_chain_B.itp" #endif ; Include chain topologies #include "mixaggregate_Protein_chain_C.itp" #ifdef POSRES #include "posre_Protein_chain_C.itp" #endif ; Include chain topologies #include "mixaggregate_Protein_chain_D.itp" #ifdef POSRES #include "posre_Protein_chain_D.itp" #endif ; Include chain topologies #include "mixaggregate_Protein_chain_E.itp" #ifdef POSRES #include "posre_Protein_chain_E.itp" #endif ; Include chain topologies #include "mixaggregate_Protein_chain_F.itp" #ifdef POSRES #include "posre_Protein_chain_F.itp"
124
#endif ; Include ligand topologies #include "mixaggregate_Ion_chain_A2.itp" #ifdef POSRES #include "posre_Ion_chain_A2.itp" #endif ; Include ligand topologies #include "mixaggregate_Ion_chain_B2.itp" #ifdef POSRES #include "posre_Ion_chain_B2.itp" #endif ; Include ligand topologies #include "mixaggregate_Ion_chain_C2.itp" #ifdef POSRES #include "posre_Ion_chain_C2.itp" #endif ; Include ligand topologies #include "mixaggregate_Ion_chain_D2.itp" #ifdef POSRES #include "posre_Ion_chain_D2.itp" #endif ; Include ligand topologies #include "mixaggregate_Ion_chain_E2.itp" #ifdef POSRES #include "posre_Ion_chain_E2.itp" #endif ; Include ligand topologies #include "mixaggregate_Ion_chain_F2.itp" #ifdef POSRES #include "posre_Ion_chain_F2.itp" #endif ; Include Topology for hfip #include "hfi.itp" #ifdef POSRES_HFI #include "posre_hfi.itp" #endif ; Include water topology #include "gromos53a6.ff/spc.itp" #ifdef POSRES_WATER ; Position restraint for each water o xygen [ position_restraints ] ; i funct fcx fcy fcz
125
1 1 #endif
1000
1000
1000
; Include topology for ions #include "gromos53a6.ff/ions.itp" [ system ] ; Name Built with Packmo l in water [ mo lecules ] ; Co mpound # mols Protein_chain_A 1 Ion_chain_A2 1 Protein_chain_B 1 Ion_chain_B2 1 Protein_chain_C 1 Ion_chain_C2 1 Protein_chain_D 1 Ion_chain_D2 1 Protein_chain_E 1 Ion_chain_E2 1 Protein_chain_F 1 Ion_chain_F2 1 hfi 800 SOL 6666 NA 6
126
APPENDIX K Topology for Zinc HETATM
1 ZN ZN A 18
-7.502 1.199 1.466 1.00 1.00
Zn
END
127
APPENDIX L Example script mdp file for Energy Mini mization: ; Preprocessing title = steep (steep; cg) cpp = /usr/bin/cpp include = -I/usr/local/gro macs/share/gromacs/top define = -DFLEXIBLE ; ; Run Control ; integrator = steep nsteps = 250000 ; ; Energy Min imization emstep = 0.01 emtol =5 nstcgsteep = 10 ; ; Neighbor Search ing nstlist = 10 ns_type = grid pbc = xyz rlist = 2.0 ;nm ; ; Electrostatics coulombtype = p me rcoulomb = 2.0 ;nm vdw-type = cut-off rvdw = 2.0 fourierspacing = 1.0 ; grid spacing for FFT ; ; Output Control nstenergy = 10
128
APPENDIX M Example script mdp file for Equili bration 1 (NVT): ; Preprocessing title = heat ; init ial equilibrate in canonical ensemble (NVT) cpp = /usr/bin/cpp include = -I/usr/share/gromacs/top ; ; Run parameters integrator = md ; leap-frog integrator nsteps = 2500000 ; 0.2 * 2500000 = 5000 ps = 5 ns dt = 0.002 ; 2.0 fs ; ; Output control nstxout = 100 ; save coordinates every 0.2 ps nstvout = 100 ; save velocities every 0.2 ps nstenergy = 100 ; save energies every 0.2 ps nstlog = 100 ; update log file every 0.2 ps xtc-precision = 500 nstxtcout = 100 ; xtc co mpressed trajectory output every 2 ps ; ; Bond parameters continuation = no ; first dynamics run constraint_algorithm = lincs ; holonomic constraints constraints = all-bonds ; all bonds (even heavy atom-H bonds) constrained lincs_iter =1 ; accuracy of LINCS lincs_order =4 ; also related to accuracy ; ; Neighborsearching ns_type = grid ; search neighboring grid cells nstlist = 10 ; 10 fs rlist = 1.0 ; short-range neighborlist cutoff (in n m) rcoulomb = 1.0 ; short-range electrostatic cutoff (in n m) rvdw = 1.0 ; short-range van der Waals cutoff (in n m) ; ; Electrostatics coulombtype = PM E ; Particle Mesh Ewa ld for long-range electrostatics pme_order =4 ; cubic interpolation fourierspacing = 0.16 ; grid spacing for FFT ; ; Temperature coupling is on tcoupl = V-rescale ; modified Berendsen thermostat tc-grps = Protein non-Protein ; tau_t = 0.1 0.1 ; time constant, in ps ref_t = 300 300 ; reference temperature, one for each group, in K ; ; Pressure coupling is off pcoupl = no ; no pressure coupling in NVT
129
; Energy monitoring energygrps = Protein_non-Protein ; ; Periodic boundary conditions pbc = xy z ; 3-D PBC ; ; Dispersion correction DispCorr = EnerPres ; account for cut-off vdW scheme ; ; Velocity generation gen_vel = yes ; assign velocities fro m Maxwell d istribution gen_temp = 300 ; temperature for Maxwell distribution gen_seed = 173529 ; generate a random seed
130
APPENDIX N Example script mdp file for Equili bration 2 (NPT): title = Turning on the P coupling cpp = /usr/bin/cpp include = -I/usr/local/gromacs/share/gromacs/top ;define = -DPOSRES (-DPOSRES_hfi) ; ; RUN CONTROL PA RAMETERS integrator = md dt = 0.002 ; 2.0 fs nsteps = 1000000 ; 0.002 * 1000000 = 2000 ps = 2 ns ; ; OUTPUT CONTROL OPTIONS nstxout = 500 ; save coordinates every 0.2 ps nstvout = 500 ; save velocities every 0.2 ps nstenergy = 500 ; save energies every 0.2 ps nstlog = 500 ; update log file every 0.2 ps xtc_precision = 1000 nstxtcout = 500 ; xtc co mpressed trajectory output every 2 ps ; ; BOND PARAM ETER continuation = yes ; Restarting after NVT constraint_algorithm = lincs ; holonomic constraints constraints = all-bonds ; all bonds (even heavy atom-H bonds) constrained lincs_iter =1 ; accuracy of LINCS lincs_order =4 ; also related to accuracy ; ; NEIGHBORSEA RCHING ns_type = grid ; search neighboring grid cells nstlist = 10 ; 10 fs rlist = 1.4 ; short-range neighborlist cutoff (in n m) rcoulomb = 1.4 ; short-range electrostatic cutoff (in n m) rvdw = 1.4 ; short-range van der Waals cutoff (in n m) ; ; OPTIONS FOR ELECTROSTATICS A ND VDW coulombtype = p me ; Particle Mesh Ewald for long-range electrostatics fourierspacing = 0.5 ; grid spacing for FFT pme_order =4 ; cubic interpolation ; ; TEM PERATURE COUPLING IS ON tcoupl = V-rescale ; modified Berendsen thermostat tc-grps = Protein non-Protein tau_t = 0.1 0.1 ; time constant, in ps ref_t = 300 300 ; reference temperature, one for each group, in K ; ; Energy monitoring energygrps = Protein_non-Protein
131
; ; Periodic boundary conditions pbc = xy z ; 3-D PBC ; ; PRESSURE COUPLING IS ON pcoupl = Berendsen ; Pressure coupling on in NPT (cant used Parrinello-Rah man it will gve the presseure -ve value) pcoupltype = isotropic ; uniform scaling of bo x vectors tau_p = 2.0 ; time constant, in ps ref_p = 1.0 ; reference pressure, in bar compressibility = 4.5e-5 ; isothermal co mpressibility of water, bar^ -1 refcoord_scaling = co m ; ; GENERATE VELOCITIES FOR STA RTUP RUN gen_vel = no gen_temp = 300.0 gen_seed = 173529
132
APPENDIX O Example script mdp file for MD producti on: title = MD simulat ion cpp = /usr/bin/cpp include = -I/usr/local/gromacs/share/gromacs/top ; ; RUN CONTROL PA RAMETERS integrator = md dt = 0.002 ; 2 fs nsteps = 100000000 ; 100000000 * 0.002 = 200000 ps = 200 ns ; ; OUTPUT CONTROL OPTIONS nstxout = 500 nstvout = 1000 nstlog = 500 nstenergy = 500 nstxtcout = 500 xtc_precision = 2000 ; ; BOND PARAM ETER continuation = yes ; Restarting after NVT constraint_algorithm = lincs ; holonomic constraints constraints = all-bonds ; all bonds (even heavy atom-H bonds) constrained lincs_iter =1 ; accuracy of LINCS lincs_order =4 ; also related to accuracy ; ; NEIGHBORSEA RCHING PA RAMETERS nstlist = 10 ns-type = Grid pbc = xy z rlist = 1.4 ; ; OPTIONS FOR ELECTROSTATICS A ND VDW coulombtype = p me rcoulomb = 1.4 vdw-type = Cut-off rvdw = 1.4 fourierspacing = 0.16 pme_order =4 ewald_rtol = 1e-5 table-extension= 1.0 optimize_fft = yes ; ; TEM PERATURE COUPLING IS ON tcoupl = V-rescale ; modified Berendsen thermostat tc-grps = Protein Non-Protein ; mo re accurate tau_t = 0.1 0.1; time constant, in ps ref_t = 300 300 ; reference temperature, one for each group, in K
133
; PRESSURE COUPLING IS ON pcoupl = Berendsen ; Pressure coupling on in NPT pcoupltype = isotropic ; uniform scaling of bo x vectors tau_p = 2.0 ; time constant, in ps ref_p = 1.0 ; reference pressure, in bar compressibility = 4.5e-5 ; isothermal co mpressibility of water, bar^ -1 refcoord_scaling = co m ; ; Energy monitoring xtc-grps = Protein_non-Protein energygrps = Protein_non-Protein ; ; GENERATE VELOCITIES FOR STA RTUP RUN gen_vel = yes gen_temp = 300.0 gen_seed = 173529
134
BIODATA OF S TUDENT Nur Syafiqah Binti Abdul Ghan i was born on 31st August 1989 and received her first educational at Sek. Ren. Rantau Panjang at Melaka and continued for secondary level at Sek. Men. Sultan Alauddin also in Melaka. After SPM she got offered to continue studied in Johor Matriculat ion College, Tangkak, Johor for one year and further her high level education in Universiti Putra Malaysia on 2008. She graduated with a major of Bachelor of Science (Hons.) Petroleu m Chemistry in 2011 and her final year project was in organic field which entitled “Chemical Consistuent in Cratoxylum Arborescens” under supervision Prof Gwendoline Ee Cheng Liang. Before she started her studies as a Master student, she was working as a Material Engineer at Shah Alam, Selangor. Almost one year working, she got an offered fro m her supervisor, Dr. Roghayeh Abedi Karjiban who is looking a master student in computational of theoretical and chemistry on March 2012. She thought this was the best opportunity to further her studies in computational field and she also wants to learn something new because she preferred chemistry in her life more. She worked as research assistant for six months before continued as official master student under supervision Dr Roghayeh and co -supervisor, Prof Mahiran Basri. On the way to complete her worked, in the middle of the end of semester she was married with her husband, Nor Irman Yajis on 14 th of June 2014. Right now in the middle of writ ing her thesis, she was pregnant and got delivered on 27th August 2015 her son was coming out and the name was Muhammad Iqbal.
135
LIST OF PUB LICATIONS
Published Nur Syafiqah Abdul Ghani, Roghayeh Abedi Karjiban, Mahiran Basri, NurHana Faujan and Lim Wui Zhuan. Unveiling A myloid-β(1-42) interaction with Zinc in water and mixed hexafluoroisopropanol solution in Alzheimer’s disease . International Journal of Peptide Research and Therapeutics. DOI: 10.1007/s10989-016-9570-4 (Accepted) Proceedings Zarina bt Ahmad, Roghayeh Abedi Karjiban, Nur Syafiqah Abdul Ghani (2015). Co mparison of Amylo id-β (1-42) in Different So lvent and Temperature. 18th Industrial Chemistry Seminar, 9 June, Cyberview Resort & Spa, Cyberjaya, Malaysia. Nur Syafiqah Abdul Ghani, Roghayeh Abedi Karjiban, Mahiran Basri (2014). Structural Co mparison Between Aβ (1-42) and Aβ (1-42)-Zn in Water. 18th Malaysian International Chemical Congress (18MICC) 2014, PWTC, 3 -5th November, Kuala Lampur, Malaysia. Nur Syafiqah Abdul Ghani, Roghayeh Abedi Karjiban, Mahiran Basri (2014). A Co mparison between the Structure of Amyloid-β (1-42) without and with Zinc in Mixed So lvent. The Fundamental Science Congress (FSC 2014), 19-20th August, Auditorium Jurutera, Universit i Putra Malaysia, Malaysia. Nur Syafiqah Abdul Ghani, Roghayeh Abedi Karjiban, Mahiran Basri (2014). Molecular Simu lation of Amylo id-β(1-42) with Zinc in Alzheimer’s Disease. Postgraduate Seminar 2013/14, 25-27th June, Bilik Saintis Gemilang, Depart ment of Chemistry, Facu lty of Science, UPM, Malaysia. Nur Syafiqah Abdul Ghani, Roghayeh Abedi Karjiban, and Mahiran Basri (2013). Co mputational Simu lation of Amy loid-β(1-42) in the Mixed Solvent. 38th Annual Conference of the Malaysian society for Biochemistry and molecu lar Biology, Putrajaya Marriot hotel & spa, Putrajaya, 28-29th August, Malaysia.
136
