Upload
lyphuc
View
233
Download
0
Embed Size (px)
Citation preview
i
Nutraceuticals Based Computational Medicinal Chemistry
KAYATHRI RAJARATHINAM
Licentiate Thesis - 2013
School of Biotechnology
Division of Theoretical Chemistry and Biology
Royal Institute of technology
Stockholm, Sweden
i
© Kayathri Rajarathinam, 2013 ISBN 978-91-7501-806-5 ISSN 1654-2312 TRITA-BIO Report 2013:11 Printed by Universitets service US-AB Stockholm, Sweden 2013
ii
To my lovable family
iii
Abstract
In recent years, the edible biomedicinal products called nutraceuticals have been becoming more
popular among the pharmaceutical industries and the consumers. In the process of developing
nutraceuticals, in silico approaches play an important role in structural elucidation, receptor-
ligand interactions, drug designing etc., that critically help the laboratory experiments to avoid
biological and financial risk. In this thesis, three nutraceuticals possessing antimicrobial and
anticancer activities have been studied. Firstly, a tertiary structure was elucidated for a coagulant
protein (MO2.1) of Moringa oleifera based on homology modeling and also studied its
oligomerization that is believed to interfere with its medicinal properties. Secondly, the
antimicrobial efficiency of a limonoid from neem tree called ‘azadirachtin’ was studied with a
bacterial (Proteus mirabilis) detoxification agent, glutathione S-transferase, to propose it as a
potent drug candidate for urinary tract infections. Thirdly, sequence specific binding activity was
analyzed for a plant alkaloid called ‘palmatine’ for the purpose of developing intercalators in
cancer therapy. Cumulatively, we have used in silico methods to propose the structure of an
antimicrobial peptide and also to understand the interactions between protein and nucleic acids
with these nutraceuticals.
iv
List of papers
1. Dimerization of a flocculent protein from Moringa oleifera: experimental evidence and
in silico interpretation
Asalapuram R. Pavankumar, Rajarathinam Kayathri, Natarajan A. Murugan, Qiong Zhang,
Vaibhav Srivastava, Chuka Okoli, Vincent Bulone and Hans Ågren;
Journal of Biomolecular Structure and Dynamics, 2013 (In press)
2. A plant alkaloid, azadirachtin as an inhibitor for glutathione-S-transferase from
Proteus mirabilis
Rajarathinam Kayathri, Natarajan A. Murugan, Premkumar Albert, Hans Ågren;
(Journal of Chemical Information and Modeling - communicated)
3. Palmatine, a nucleotide sequence specific inhibitor for cancer therapy.
Rajarathinam Kayathri, Natarajan A. Murugan, Hans Ågren (in manuscript)
Paper 1: It is teamwork, where I am responsible for the calculations and participated actively in
manuscript preparation.
Paper 2 and Paper 3: In which, I have done the calculations. The manuscript was solely
prepared by me and actively participated in discussions with the co-authors.
v
Acknowledgements
First of all, I would like to express my heartfelt and sincere gratitude to Prof. Hans Ågren for
giving me a great space, trusting and accepting me as his student. He has been an inspiration to
me by setting an example of sincerity, activeness and commitment.
I should thank Dr. Arul Murugan for his constant guidance, help and valuable time during my
period of the study. Dr. Yaoquan Tu, for his encouragement and patience to answer all my
thoughtless questions. I am always impressed by Dr. Boris Minaev for his active and sincere
dedication for any kind of work and it was my privilege listening to his lectures.
This stage of mine would not be possible without Prof. Yi Luo, Prof. Faris Gel'mukhanov, Prof.
Olav Vahtras, Dr. Marten Ahlquist and Dr. Zilvinas Rinkevicius, who always found involve in
the scientific discussions which was well encouraging. The energy, the humour sense, and the
extrovert of the faculties gave plenty of liveliness in the working atmosphere. Ms. Nina Bauer
made my task easier for me with all her tireless administrative support..
I cannot stop here without mentioning about my friends and colleagues in the department for
making me to feel better in the work place and fun in the office. Great thanks for Anuar, Y. Ai,
Anand, Weijie, Y. Ai, Lili, Xin Li, F. Qiang, Rocio, S. Duan, Robert, X. Chen, Liqin, Bogdan,
Ce Song, Chunze, Ignat, Irina, Jaime, Jungfeng, Jing, Yongfei, Hongboa, Lijun, Xin Li, Gao Li,
Lu Sun, Miao, Guangjun, Xu, Yan Wang, Wei Hu, Xiao-Fei Li, Yong Ma, Xinrui, Vinicius, Ke-
Yan Lian, Qiang Zhang, Kestutis, Ying Wang, Guanglin especially Xianqiang and Runa, there is
not much we have not discussed! There are lot more in the list to express my gratitude.
Family is something, which helps for any step of one’s development, that’s absolutely true in my
case. I thank all my family members in India, for their moral support, encouragement, love and
understanding throughout my life. Though they do not read this, my love should be expressed.
My parents, without whom my life would not be to this extent. I am very lucky to have my
parents-in-law, who never put me down in front of anybody and treat me more than their own
daughter. The boon of my life is my beloved sister, the heart always that I can trust and think
vi
only good for me, and also my brother-in-law who gave “do it now” attitude to me. The person
who made meaning in my life is my brilliant nephew, Vamsi, love you!
“Behind every man, there is a woman” in my case, it’s none other than my lovable husband,
Pavan and his help in my life cannot be expressed in words. Another very important person in
my life who actually helps always with tons and tons of energy just by an innocent smile and
touch is my sweet baby girl, Mitti. For these two pillars of my life, I cannot stop with thanks.
Because they gave, giving and will give everything and anything that I need and want in my life,
this thesis is part of their love. No words to express my gratefulness and love to both of you.
Thanks for all your help and encouragement.,
Kayathri
vii
Abbreviations
AZT Azadirachtin
B3LYP Becke 3 parameter Lee, Yang-Parr functional
DNA Deoxyribonucleic acid
DOF Degrees Of Freedom
GBSA Generalized Born Surface Area
GSH Gluthathione
GST Gluthathione-S-Transferase
HL Human Leukemic
LDL Low-Density Lipoprotein
MD Molecular Dynamics
MM Molecular Mechanics
MM-GBSA Molecular Mechanics- Generalized Born Surface Area
MM-PBSA Molecular Mechanics- Poisson Bolzmann Surface Area
MO Moringa oleifera
MRSA Methicillin-Resistant Staphylococcus aureus
NAB Nucleic Acid Builder
Nmode Normal Mode
NMR Nuclear Magenetic Resonanace
PBSA Poisson Bolzmann Surface Area
PDB Protein Data Bank
PmGST Proteus mirabilis Gluthathione-S-Transferase
RDA Recommended Daily Allowance
RMSD Root Mean Square Displacement
RNA Ribonucleic acid
SASA Solvent Accessible Surface Area
VLDL Very-low-density lipoprotein
viii
One and three letter abbreviations for amino acids
Alanine Ala A
Arginine Arg R
Asparagine Asn N
Aspartic acid Asp D
Cysteine Cys C
Glutamic acid Glu E
Glutamine Gln Q
Glycine Gly G
Histidine His H
Isoleucine Ile I
Leucine Leu L
Lysine Lys K
Methionine Met M
Phenylalanine Phe F
Proline Pro P
Serine Ser S
Threonine Thr T
Tryptophan Trp W
Tyrosine Tyr Y
Valine Val V
Abbreviations for nucleic acid residues
A Adenine
C Cytosine
G Guanine
T Thymine
ix
List of contents
1 Introduction
1.1 Natural products for biomedical applications
1.2 Nutraceuticals
1.2.1 Types and importance of nutraceuticals
1.2.2 Plant derived medicinal plants
1.3 In silico study in biomedical applications
1.4 Plant-based medicinal products
1.4.1 Medicinal values of Moringa oleifera
1.4.1.1 Coagulant protein from the seeds of Moringa oleifera
1.4.2 Traditional medicinal plant Azadirachta indica
1.4.2.1 The limonoid, azadirachtin as an antimicrobial compound
1.4.3 Palmatine – structural analogue of berberine
1.5 Objectives
2 Computational methods
2.1 Docking
2.1.1 Introduction
2.1.2 Types of docking
2.1.3 Applications
2.2 Molecular dynamics simulation
2.3 Binding free energy
3 Present investigations
3.1 Structure determination of Moringa oleifera coagulation protein (Paper 1)
3.2 Azadirachtin as an antimicrobial compound (Paper II)
3.3 Palmatine, a nucleotide sequence specific molecule for cancer treatment (Paper 3)
Bibliography
1
Chapter 1
Introduction
Over the decades, natural products are considered to be the most productive source for the
development of novel drugs. Serving as the most active ingredients for a wide-range of novel
medicines, natural products are usually analysed through different biochemical techniques to
evaluate their biological competence. Several combinatorial chemistry techniques combined with
in silico methods are applied on natural scaffolds to create ‘drug-like compounds or lead-
molecules’. Such modern combinatorial methods are advent of high-throughput screening and
the post-genomic era to discover more than 80% of drug substances from the natural compounds
of ‘olden times’ [1, 2].
1.1 Natural products for biomedical applications
Traditionally, natural products are being used as dyes, polymers, fibers, oils, glues/waxes,
flavouring agents, perfumes, and drugs like penicillin. There are more than 100,000 low-
molecular-mass natural products, mostly derived from the isoprenoid, phenylpropanoid, alkaloid
or fatty acid/polyketide pathways are being used for a wide variety of biomedical and industrial
purposes. The investigation of natural products, as the source of novel human therapeutics
attained its peak during 1970-1980 and completely revolutionized the area of pharmaceutical
chemistry to develop novel non-synthetic molecules. Based on Pharmaproject database, the
natural products in different stages of drug development from various biological sources are
explained in Table 1. As a result, pharmaceutical industries developed about 63% of 974 small
molecules (new chemical entities) between 1981 and 2006 [2-5]. The empiric recognition of
potent biological activities from numerous natural products has fuelled to develop several novel
biomedicines including antibiotics, insecticides, herbicides, anti-inflammatory and even anti-
cancer drugs. Such promising medicinal advantages triggered the research of natural products to
re-evaluate and identify novel ‘active components or lead molecules’ to address today’s health
problems like cancer, diabetes, neural diseases, antimicrobial resistance, especially in the context
of eco-friendly interactions.
2
Table 1: Drugs based on natural products at different stages of development
Development
stages
Plant Bacterial Fungal Animal Semi-
synthetic
Total
Pre-clinical 46 12 7 7 27 99
Phase I 14 5 0 3 8 30
Phase II 41 4 0 10 11 66
Phase III 5 4 0 4 13 26
Pre-registration 2 0 0 0 2 4
Total 108 25 7 24 61 225
Reproduced from Harvey, 2008 [5; Source: Pharmaprojects database, March 2008]
1.2 Nutraceuticals
The term ‘nutraceuticals’ was coined by Dr. Stephen DeFelice in 1989 to define natural food
materials possessing pharmaceutical medicinal activities and nowadays, it is used as standard
acronym in the field of medicinal chemistry [6]. Nutraceuticals can be defined as "a food (or part
of a food) that provides medical or health benefits, including the prevention and/or treatment of a
disease" [7]. Most of the natural medicinal products derived from plants and animals are edible
in one or the other form (Table 2). Hence, in the recent years the focus of pharmaceutical and
natural product researchers have turned to investigate nutraceuticals through several technical
approaches.
1.2.1 Types and importance of nutraceuticals
The nutraceuticals can be broadly classified as (i) dietary supplements (ii) functional foods.
Dietary supplements derived from food products contain nutrients like vitamins, minerals and
amino acids. Functional foods are designed to provide natural enrichment to the consumer, rather
than manufactured dietary supplements in liquid or capsule form. Generally, a process called
‘nutrification’ is followed to provide required amount of body nutrients. Thus, it is vital that the
functional foods of natural origin should be an essential part of our daily diet and thus regulate
the biological processes, resulting in disease control or prevention [8, 9]. Hence, it is proposed
that the herbal remedies may be classified as nutraceuticals because of their perceived risk with
3
self-medication (e.g. Digitalis); the functional food consumed as a part of normal diet (e.g.
carotenoid, vitamins, etc.). There are several nutraceuticals available in the market like fortified
cereals (rich in vitamins and minerals (muesli, bran flakes etc), nutrient supplements (vitamin A
(β-Carotene), additional supplements (cod liver, oil, primrose oil, glucosamine, garlic etc.,),
energy drinks/tablets (Tropicana, Minute Maid Pulp, Frooti), cholesterol reducers (Abcor), body
building supplements (protein powders and bars), sports products (Glucon-D, glucose D),
probiotics (lactobacillus containing yoghurt and medicines) [7].
Table 2: Common nutraceuticals and their drug targets
Crop Latin name Drug/natural products Targets
Cassava Manihot esculenta Linamarin, lotaustralin Cancer
Celery Apium graveolens Psoralen, xanthotoxin,
bergapten
Anticoagulant
Broccoli Brassica napa Sulforaphane Cancer
Potato Solanum tuberosum α-solanine, α-chaconine Cancer
Tomato Lycopersicon
esculentum
Tomatine Cancer
Tomato powder
Common bean Phaseolus vulgaris Lectins Cancer, HIV
Soybean Glycine max Genistein, daidzein Breast cancer
(soybean extract)
Lettuce Lactuca sativa Lactucin, lactucopicrin Inflammation,
malaria
Apple (extracts) Malus domestica Vitamin -C Liver cancer
Pomegranate Punica granatum Tetracycline, gentamcin MRSA*
Turmeric Curcuma longa Curcumall Antibacterial
Sarasaparilla Hemidescus indicus Chloramphenicol MRSA*
* MRSA – Methicillin resistant Staphylococcus aureus
Adopted from B. Schmidt et al., 2008 [9].
4
However, consumption of fruits, vegetables, oils, seeds, leaves, and other sources of the active
ingredients are also highly important as it serves as an alternative to the ‘drugs forms’. Based on
the current survey it is known that the plant-derived products are more popular to contain
medicinal properties than the animal sources. Hence, studies of phytochemicals are increasing
day-to-day.
1.2.2 Plant-derived medicinal products
The plant-derived products are not only used as edible medicinal food but also raw materials for
many of today’s modern medicines [10]. Generally the plant-derived biomedical products
include amino acids, irregular amino acids and their derivatives, alkaloids, flavonoids, steroidals,
terpinoids, oils, carbohydrates (sugars, sugar amines, sugar alcohols, polysaccharides), food
supplements, etc., are considered to be very important raw materials for industries in food,
chemical and pharmaceutical sectors. Moreover, the plant-derived products serve as a valuable
source for biosynthetic drugs (elliptinium, galantamine, huperzine), semi-synthetic compounds
(tigecycline, everolimus, telithromycin, micafungin, caspofungin) and synthetic drugs (aspirin).
Even the first synthetic drug, aspirin (acetylsalicylic acid), synthesized by Arthur Eichengr n and
Felix Hoffmann in 1897, is also an active ingredient of analgesic herbs. Such discoveries,
especially in biomedical sector ushered ‘novel drug discovery’ programmes in several research
institutes and pharmaceutical industries, worldwide. About 60% and 75% of today’s drugs,
especially those in the area of cancer and infectious diseases are derived from the plants,
respectively [11].
Along with the important chemical properties, the empirical studies of natural products paved the
way to develop several biochemical techniques (separation techniques, spectroscopic
approaches, structure elucidation) and synthetic methodologies are now became the foundation
of modern drug discovery platforms. The heterologous production platforms for recombinant
pharmaceutical proteins are usually produced by large-scale fermentation in bacteria, yeast or
animal cells [12, 13]. However, it has been only recently that biotechnology has been used to
generate plants that produce specific therapeutic proteins, products that are traditionally
synthesized using recombinant microbes or transformed mammalian cells [14]. The
5
pharmaceutically important proteins like human serum albumin was first expressed in tobacco
and potato leaves for large-scale production [15]; similarly, monoclonal antibody was expressed
in tobacco leaves [16]. These studies demonstrated that plants could also produce stable and
functional human proteins, leading to the commercial production through recombinant protein
expression [17]. Nowadays, several hundreds of medicinal proteins are being produced in a
variety of plants and plant-based systems. However, many of these compounds now have been
shown to have important adaptive significance in protection against herbivore and microbial
infection, as attractants for pollinators and seed-dispersing animals, and as allelopathic agents
that influence competition among plant species [18]. Cumulatively, the plant-derived products
became the next major commercial development in biotechnology due to advantages like large-
scale production, economy product safety, ease of storage and low-cost drugs and vaccines
production, especially suitable for the developing world [19].
However, the commercialization of plant-derived products is becoming limited by the uncertain
regulatory terrain, particularly for the adaptation of good manufacturing practice regulations to
the field-grown plants. Hence, the plant-derived natural products in drug discovery have been
diminished, which can be supplemented with structure activity–guided organic synthesis,
combinatorial chemistry, and computational (in silico) drug design for better drug targeting.
1.3 In silico study in biomedical applications
The in vitro, in vivo and in situ studies are combined with high risk as these processes are
complicated, expensive and also being done without a restrictive biological outcome. Moreover,
conducting experiments with biological samples including animal models are associated with
very strict ethical procedures and high cost. These disadvantages can be addressed through
computer-aided drug designing by simulating the in situ biological situation or reaction. The
recent technological advances led to renewal of natural products in drug discovery by the
development of new synthetic derivatives. [3]. Since a couple of decades, biochemical systems
theory like ‘computer-aided drug designing’ are helpful to discover that many potent broad-
spectrum antibiotics and anti-cancer drugs available in the market are based on the structural-
activities of the ‘lead compounds’. These activities also helped to even develop dual-activity
compounds based on the chemical characteristics of the natural plant-derived products, which
6
have become an integral part of today’s modern organic chemistry and pharmaceutical
industries. Such proactive methods include modification of the active sites of enzymes,
development of inhibitors for a specific protein target, increasing the propensity of bioactive
molecule to improve its binding to the target, identifying a variety of protein domains and
folding motifs are significantly helpful in defining protein–protein or protein-nucleic acid
(DNA/RNA) interactions. As a result, the designed molecules can be more effective in
modulating cellular processes including the immune response, signal transduction, mitosis and
apoptosis, which are considered to be the critical aspects in designing a novel drug molecule. For
example, an enzyme inhibitor must bind to the active site at the interior of the protein, so that it
can involve or restrict or activate few metabolic interactions. Whereas, a small-molecule
disruptor of a protein–protein interaction must bind to the protein surface (rather interior part),
where the tight binding (affinity) of the molecule depends on multiple interactions surface. Such
molecular interactions are based on modifying the binding sites, size, and selective pressure to
bind gene products to initiate a required function. This allows modulating the particular protein–
protein interactions by structural fine-tuning of the medicinal product [20]. Nevertheless, the
relative ease and low cost to offer an affordable solution fuel the modern computer-aided drug
discovery.
1.4 Plant-based medicinal products
Among the several hundreds of medicinal plants including their natural plant-derived products, it
is always interested to study more common edible plants. So considering the wide-importance
and need, the following medicinal plants were studied.
1.4.1 Medicinal values of Moringa oleifera
The medicinal and chemical values of a tropically cultivated ‘magical plant’, Moringa oleifera
(M. oleifera) is a rich source against a broad-spectrum of diseases and infections [21-23]. As a
whole, the plant moringa serves as an effective prophylactic agent in conventional medicine [23].
The most widely reported use of M. oleifera is employment of the seeds for water purification
and clarification [24-26], also exerts its protective effect by decreasing liver lipid peroxides,
antihypertensive [27, 28]. In addition, seed extracts of moringa are reported to remove organic
pollutants [29]. Bark extracts of various moringa tissues have been used as anti-cancer agents
7
[30], root extracts as antilithic, rubefacient, vesicant, carminative, antifertility, anti-
inflammatory, stimulant in paralytic afflictions; act as a cardiac/circulatory tonic, used as a
laxative, abortifacient, treating rheumatism, inflammations, articular pains, lower back or kidney
pain and constipation, anti-trypanosomal agents [31 -35], leaves as purgative, applied as poultice
to sores, rubbed on the temples for headaches, used for piles, fevers, sore throat, bronchitis, eye
and ear infections, scurvy and catarrh; leaf juice is believed to control glucose levels, applied to
reduce glandular swelling [36-39], flowers as high medicinal value as a stimulant, aphrodisiac,
abortifacient, cholagogue; used to cure inflammations, muscle diseases, hysteria, tumors, and
enlargement of the spleen; lower the serum cholesterol, phospholipid, triglyceride, VLDL, LDL
cholesterol to phospholipid ratio and atherogenic index; decrease lipid profile of liver, heart and
aorta in hypercholesterolaemic rabbits and increased the excretion of faecal cholesterol [40-45].
Leaf extracts have been shown to regulate thyroid and cholesterol levels [46, 47]. Incorporation
of processed M. oleifera leaves into noodles improves nutritional quality by supplying the
recommended daily allowance (RDA) for a number of vitamins [48]. Moreover, there are several
food varieties, food supplements and medicines are available in the market.
1.4.1.1 Coagulant protein from the seeds of Moringa oleifera
Though a vast research has been done using crude extracts of M. oleifera, there are no evident
reports about the specific proteins responsible for the antimicrobial, anticancer, bioadsorbant and
anticholesterol values. However, a 6.5 kDa protein from crude seed extract of M. oleifera is
reported for its antimicrobial and flocculation activities [49], which is being used a natural water
purifier in several countries. However, the molecular mass of this water soluble protein was
reported between 3.5-14 kDa [50-52]; whereas, recent reports suggested that this protein could
have subunits forming a molecular mass of 17-26 kDa [53] or single cationic protein of 66 kDa
[54]. In order to understand further, it is necessary to know the exact molecular weight of the
protein for the structural elucidation as well as to study its mechanism of action. Such issues can
be addressed through the modern computational techniques.
1.4.2 Traditional medicinal plant Azadirachta indica
The Azadirachta indica (neem tree) is one of the multi-purpose plants, whose products as well as
byproducts are being used for several biomedical applications like insecticidal, antiseptic,
8
contraceptive, antipyretic, antiparasitic, fertilizers, etc. [55]. Neem fruits, seeds, oil, leaves, bark
and roots are generally used as antiseptics, antimicrobials, treatment of urinary disorders,
diarrhoea, fever and bronchitis, skin diseases, septic sores, infected burns, hypertension and
inflammatory diseases. Among these triterpenes compounds, limonoids are the most effective
one, which is abundant in neem oil. At least nine limonoids are effective in inhibiting insect
growth, especially some of the most deadly varieties found in human health and agriculture
worldwide. Among these, azadirachitin (AZT) is considered to be the main ingredient for
fighting insects and pests, being about 90% effective. Meliantriol is another feeding inhibitor that
prevents locusts chewing and traditional crop protection. Nimbin and nimbidin, also found in
neem have anti-viral properties and also effective against fungal infections. Gedunin, a lesser
limonoid, is effective in treating malaria through teas and infusion of the leaves.
1.4.2.1 The limonoid, azadirachtin as an antimicrobial compound
AZT is an oxidized tetranortriterpenoid that helps oxygen functionality in variety of carboxylic
esters in order to increase detoxification process. There are at least six isoforms of AZT that have
been identified: AZT-A, AZT-B, AZT-C, AZT-D, AZT-H and AZT-I [56]. The NMR and X-ray
crystal structures have been determined for all the AZT forms [57]. Among them, AZT-A and
AZT-B are the predominant as well as abundant forms, while the rest are available only few
parts compared to hundred parts of AZT-A, which is a well-studied isoform in terms of activity
among other subtypes [58]. The AZT-H has very good fungicidal activity against the
phytopathogenic fungi R. solani and S. rolfsii along with the nematicidal activity [59], as
detoxification process.
Detoxification is one of the major and essential events for any cell, either prokaryotes or
eukaryotes, for their survival against noxious adverse compounds present in their micro
environment. Thus, GST serves as a significant source for contributing resistance to the bacterial
pathogens like Protease mirabilis (P. mirabilis), pathogenic Escherichia coli, Salmonella,
Pseudomonas, etc, by either enhancing their efflux systems or transporting such compounds to
outer surface of the cell [60]. Among these pathogens, the P. mirabilis is a major problem
causing anatomical abnormalities in patients with urolithiasis or a chronic urinary catheter due to
urinary tract infections [61]. The P. mirabilis produce urease, which generates ammonia that
9
increases the pH of the urine to >7.2. The elevated calcium and magnesium crystallization in the
urine blocks the catheter lumen and causes bacteriuria, pyelonephritis, bacteremia, and shock.
Though P. mirabilis is susceptible to β-lactams, aminoglycosides, fluoroquinolones, and
trimethoprim/sulfamethoxazole, it shows resistance to nitrofurantoin, tetracycline and other
powerful antibiotics [62]. This increased resistance contributes to high morbidity and mortality
of the patients, especially in the hospitals. In this context, using their null mutants [63], reported
that GST protects the bacteria against several toxic chemicals, including antimicrobial drugs.
Hence, the GST protein could be a good target biostructure for treating bacterial infection from
P. mirabilis.
1.4.3 Palmatine – structural analogue of berberine
Palmatine is found in many plant families, which is a close structural analog of berberine.
Palmatine bears the same tetracyclic structure (7, 8, 13, 13atetrahydro-9, 10-dimethoxy-
berberinium) as similar to berberine but differs by the nature of the substituents at positions 2,3
on the benzo ring; being methylene dioxy for berberine and dimethoxy for palmatine. This
alkaloid has been shown to exhibit significant antitumor activity against HL-60 cells and has
antimicrobial properties [64, 65]. However, the binding studies of palmatine to nucleic acid bio-
targets have been scanty and till today no data is available in the literature on the mode,
mechanism and base sequence specificity of binding of palmatine to DNA [66].
Recently, it is identified that DNA as a potential bioreceptor for palmatine and suggested both
intercalation and external stacking of palmatine on complexation with natural DNA having
heterogeneous AT and GC sequences. The noncovalent complexes of several protoberberine
alkaloids including berberine and palmatine with few oligonucleotides showed that the palmatine
has higher binding affinity than berberine to the oligonucleotides and also demonstrate a
remarkable AT base pair specific intercalative mode of DNA binding for palmatine [67]. Hence,
it could be useful if the effect of palmatine is studied with known sequence of DNA templates.
10
1.5 Objectives
With the growing demand for natural medicinal products, we are interested to apply
computational tools to address the biological problems to propose novel drug molecules. With
this overall aim, we frame the following objectives:
(i) Though the (M. oleifera) MO2.1 protein is considered to be the most important
substance for coagulation and antimicrobial activity, its structure is not yet known.
Hence, our first objective is to identify a plausible tertiary structure of its
monomer has been elucidated based on homology modeling. With this structure,
we try to investigate the oligomerization property of this protein using protein-
protein docking experiments.
(ii) Considering the increasing incidence of urinary tract infection due to P. mirabilis,
GST responsible for the antimicrobial resistant is targeted. We aim to propose a
potent antimicrobial agent, AZT to disrupt the detoxification process by inhibiting
GST of P. mirabilis. The binding affinity of AZT with GST is studied by binding
free energy calculations.
(iii) Based on the experimental results, we try to investigate the efficiency of
palmatine with the nucleotide sequences. In order to re-evaluate its efficiency,
palmatine is docked and simulated with DNA to observe its specificity with
different nucleotide sequences.
11
Chapter 2
Computational methods
Understanding the stable three-dimensional complex structures in the living organisms and to
find the ligand binding affinity to the target is very important in computer-aided drug design.
Both these aspects are vital for any kind of application, especially in the field of biomedicine to
accelerate the speedy identification of a new drug candidate. We have used some tools to
understand the property and the binding affinity of the natural products mentioned in Chapter 1.
The docking studies are used to identify the lead or hit molecule from the high-throughput virtual
screening. Here, it is used to find the orientation of the ligand in the binding site (paper 2 and
paper3) and the stable complex structure (paper 1). The more accurate and rigorous methods are
used further to calculate binding free energy using Molecular Mechanics/Poisson Bolzmann
Surface Area (MM-PBSA) and Molecular Mechanics/Generalized Born Surface Area (MM-
GBSA) approaches [68]. And the missing entropy contributions were calculated by NMode
analysis.
The theoretical methods we used are in the following order (1) Docking (2) Molecular Dynamics
(MD) simulation (3) Binding free energy calculation. The brief introduction of the methods used
is discussed below.
2.1 Docking
2.1.1 Introduction
The aim of finding a drug target is to estimate the binding affinity of a ligand with the
macromolecule, which in turn accelerates the process of developing a potent drug. There are
many computer based methods to study the interaction of the receptor-ligand, which could be a
protein-protein, protein-drug, and nucleic acid-drug. The initial step which gives the idea of the
binding orientation of the ligand with respect to the target molecule to form a stable complex is
docking [69]. The strength of binding affinity of the ligand to the receptor is determined by the
orientation of the ligand [70]. This is also referred as lock and key model, where the receptor and
the ligand are rigid, in some cases the protein changes its conformation to fit the ligand and
called as induced-fit model.
12
2.1.2 Types of docking
Depending upon the binding site:
Local docking: the ligand orientation in the binding site is determined where the
binding site of the receptor is known.
Global docking/blind docking: in doing so, the binding site of the receptor is unknown
and the search for the binding site and the orientation of the ligand are determined.
Based on its structure:
Bound docking: Docking is performed to test the known complex, where the receptor and
the ligand are bound to each other.
Unbound docking: Docking is done from the unbound receptor and the ligand.
Depending upon the rigidity of the receptor and the ligand:
Rigid docking: Here, both the receptor and the ligands are kept rigid. This method is
easier due to the less degree of freedom (DOF).
Flexible docking: In this, either the ligand or the receptor or both are kept flexible. This
has more DOF and the difficult to perform.
Depending on the receptor:
Macromolecular docking: In macromolecular docking two or more biological
macromolecules interact to form the quaternary complex structures. The stable
macromolecular structure complex that occurs in the living organism can be found. Eg.
Protein-protein docking or protein-nucleic acids docking.
Protein-ligand/nucleic acids-ligand: This docking is to predict the position, binding
affinity and the orientation of the ligand with the receptor (protein/nucleic acids).
2.1.3 Applications
The different applications of docking which reduce the cost and time to find a potent drug
candidate are listed below:
13
(i) To address the protein-protein interaction for biomedical application and
bioremediation.
(ii) In drug designing to identify the molecule which can bind with the receptor by
screening large set of molecules, is called as hit identification.
(iii) To find the most stable orientation of the ligand to bind with the receptor to form the
most stable complex structure, lead optimization.
The main input structure can either be downloaded from the PDB (determined by X-ray and
NMR) or by homology modeling. The good docking is predicted from scoring criteria and search
strategies/algorithms.
Docking scoring criteria:
The scoring function may be based on the molecular mechanics (MM) force-field and the
statistical potential approach. The force field estimates the good hydrogen bonding; the low
energy indicates the stable minimized system. The statistical potential is from the large set of
data (from PDB) and the charge potential which defines the polarity. The scoring function is
more from the geometry match of the large set of molecules [69].
Docking search algorithm:
The different docking search algorithms are systematic search method, simulated method and
random search method [71]. In the simulated method, the receptor is rigid and the ligand takes
different conformations, and the lowest energy conformation is considered. The systematic
search method is the one in which all the possible DOF are explored for the given molecule in all
the possible combinatorial way. CLUSPRO [72] used in paper1 to study the protein-protein
interaction, is the systematic search method of the Fast Fourier Transform correlation technique.
CLUSPRO is the fully automated program for protein docking.
Lastly, in the random method or stochastic method the pose of the ligand is estimated from the
pre-defined probability function. The examples of the random search algorithm are genetic
algorithm, Monte Carlo and Tabu search. The AUTODOCK 1.5 [73] program was based on the
Lamarckian algorithm, improvisation of the genetic algorithm, which is used in paper 2 and
14
paper 3. AUTODOCK is used mainly for the receptor (protein, nucleic acids)-ligand interaction.
This has two main programs autogrid and autodock.
The conformational changes due to binding of the small molecule with the protein are not easily
predicted in case of rigid docking. So, the most feasible method is by keeping the receptor rigid
and the ligand flexible with more manageable degrees of freedom. The false positive results are
more common by this scoring function because some of the low affinity ligands can form a
stable complex than the high affinity ligands. This can be overcome by the more rigorous
computational methods like GB/PB methods.
Docking gives the binding orientation of the ligand without the explicit solvent molecules.
However, the flexibility of the receptor and the ligand in presence of the explicit solvent can be
well studied by MD simulation. The exact bonding affinity of the ligand to that of the receptor
can be evaluated by binding free energy calculations [74, 75].
2.2 Molecular dynamics simulation
MM approach was developed to study the bigger system consisting of thousands to millions of
atoms. Newton’s second law of motion (Eq. 1) paved the way to study the larger systems by MM
approach. By representation of classical spheres for the atoms, the motion of the nuclei is
modeled without including the electrons explicitly in MM approach. This is the basic principle of
MD simulation of the atoms and molecules of the larger system.
(1)
where, F is the force exerted on the particle, m is the mass and a is the acceleration.
The acceleration of each atom in the system is determined by the force acting on each atom. The
trajectory gives the average value of properties and it is the integration of the equation of motion
to define the velocities, acceleration and the position of the particles. From the velocity and
position of each atom, the state of the system can be predicted with respect to time.
Thus,
(2)
15
(3)
which explains the movement of the particle mass m along the direction x.
(4)
The accuracy of the calculation depends much on the force from the force field which in turn
explains the potential energy of the system. The potential energy is explained in terms of
covalent and non-covalent interactions by the force-field.
+ (5)
where, and interactions are defined as Eq. 6 and Eq. 7, respectively.
(6)
(7)
Each of the bonded and the non-bonded interaction terms are expressed as below:
(8)
(9)
[ ] (10)
(11)
[(
)
(
)
] (12)
(13)
16
Molecular dynamics simulations were carried out for all the three systems (paper 1-3) were
carried out using the Amber 11 software [76].
In paper 1:
The MO2.1 polypeptide sequence, QGPGRQPDFQRCGQQLRNISPPQRCPSLRQAVQL
THQQQGQVGPQQVRQMYRVASNIPST (UniProt ID: P24303) was BLAST (Basic Local
Alignment Search Tool) against the PDB database to identify a structural homologue. The
homologue structure after docking was used as the starting structure for the simulation.
In paper 2:
The crystal structure of GST-GSH complex from 1PMT was downloaded from the protein data
bank and the stable GST-AZT complex from docking were used as the initial configuration for
the MD simulation.
In paper 3:
The B-DNA structure of four different sequences of poly AA, poly AT, poly GG, poly GC were
generated by the NAB program in AMBER 11 package. The equilibrated structure was used for
docking; the stable complex is used further for the MD simulation.
The ESP (electrostatic potential) charges of the small molecules (paper 2 and 3) were calculated
using the B3LYP/6-31+G(d) basis set in Gaussian 09 package [77]. The generalized Amber
Force-Field (GAFF) was used for all the ligands, GSH, AZT and palamtine; whereas, PARM99
was used for all the receptors (paper 1-3). The systems were neutralized by Na+ or Cl
- ions and
solvated by TIP3P water model. Equilibration was carried out for the solvated complexes after
running a minimization of 500 steps in each case. The isothermal-isobaric ensemble of
temperature 298K and 1 atmospheric pressure was maintained throughout the Langevin
dynamics with weak restraints under periodic boundary condition. The SHAKE algorithm [78]
was used to constrain the hydrogen bond lengths with a non-bonded cut-off of 8Å for the van der
Waals and electrostatic interactions. The production run was performed until the convergence of
structural quantities like density and total energies of the complex. The trajectories of the MD
were used for the binding free energy calculation.
17
Root mean square displacement (RMSD) was assessed for the ligand to know the mobility of the
ligands during the simulation by Eq. 13.
√∑
(14)
2.3 Binding free energy
All the reactions or changes in the biological reaction take place in the presence of solvent
molecules. Thus, it is important to study the dissolvation of the solute in a solvent that depends
on the free energy change. There are different types of interactions which is of short and long
range effects like hydrogen bonding, ion-dipole/electrostatic interactions, Van der Waals/dipole-
dipole, ion-ion interactions. The favourable solvation process is based on the thermodynamic
properties and associated with the decrease of Gibbs free energy. If Gibbs free energy is
negative, it indicates the difference of the enthalpy and the entropy should be a negative value.
The two major types of commonly used solvent models are the explicit solvent model (Fig. 2.1)
and the implicit solvent model (Fig. 2.2). The explicit model is where all the individual water
molecules are included explicitly. Thought, it is a precise method but the calculation of the free
energy of solvation by the simulation methods is time consuming even in the classical approach.
In the implicit model the solvent is treated as the polarizable continuum with a dielectric
constant. The charge distribution of the solute polarizes the solvent by creating a reaction
potential, this in turn changes the solute’s reaction potential. This interaction is introduced into
the Hamiltonian is represented by solvent reaction potential and computed self-consistently.
18
Figure 2.1: Explicit water model Figure 2.2: Implicit water model
The widely used implicit model is generalized Born/surface area model [79-81] and Poisson
Boltzmann/surface area model. These models combined with molecular mechanics [82] are MM-
GBSA/MM-PBSA method which is widely used to study the receptor (biomolecules)-ligand
binding free energy is an endpoint method [73, 83].
In the MM-PB(GB)SA, the difference of the binding free energy of a ligand (L) to a receptor (R)
to form the complex (RL) [84] is represented as:
(15)
The free energy of the receptor, ligand and the complex is written as:
(16)
- the total molecular mechanics energy of molecular system X in the gas phase
- solvation free energy for X in solvent
S - entropy of X.
The terms appearing in Eq. 16 are explained below:
(17)
The energy contributions for the bound and unbound state are calculated using the molecular
mechanics or the force-field method.
19
(18)
The solvation free energy is the sum of polar and non-polar contributions. The polar terms
are by the generalized Born, Poisson, or Poisson-Boltzmann model and the non-polar terms are
by solvent-accessible surface area (SASA).
Where, (19)
and depends on the solvation model and the method (PBSA or GBSA) [85].
Increase in the SASA in turn increases the solvation free energy, which is unfavourable and
excludes the nonpolar terms in the solute. The polar contribution ( ) is the solvation free
energy contribution to the electrostatic interactions of the solute surrounded by the solvent. In PB
and GB models, the solvent is treated as a high dielectric medium and the solute as low dielectric
medium.
∑
(20)
The
, is the interaction between the charge i and j is determined using the GB model by
an analytical expression form [86, 87] or by the numerical equation of PB model.
(
) [
(
)]
⁄
(21)
- distance from the solute-solvent dielectric boundary of atoms i
ε - solvent dielectric constant
- distance between i and j
The constants n and A were set to 2 and 4, respectively by Still and coworkers (84).
The difference between the PB and GB model is the solute dielectric constant ( ) term. In PB,
varies the
in Eq. 20. The is excluded in GB model (Eq. 21) due to the
approximations for the analytical expression. Still can be expressed other than 1 and in Eq.
21 it becomes( ⁄ ⁄ ).
20
The Eq. 20 is applied on the receptor, ligand and complex which the expresses the equation
below:
(22)
The electrostatic free energy in Eq. 22 can be divided as the desolvation and the interaction
components [16, 17] due to the summation of atomic charges in Eq. 20.
∑
[
∑
∑
] [
∑
∑
] (23)
There are three terms on the right hand side of the Eq. 23, the first term is the interaction term
from the receptor and ligand charges which are the contributions from the specific residues of the
receptor [88, 89]. Therefore, the interaction term can be written for the specific residue T as,
∑
(24)
The last two terms on the right hand side of the Eq. 23 are the desolvation terms corresponding
to the receptor and ligand, respectively. The desolvation contribution is due to the change in the
charge distribution and by replacing the high dielectric solvent medium by the low dielectric
medium (paper).
The equilibrium conformations after discarding the explicit solvent for the receptor, ligands, and
the complex are used to calculate the MM-GB(PB)SA, where the solvent is expressed as
continuum dielectric. These models are implemented in packages of Amber [90], Delphi [91]
and Schrödinger [92].
The last term TS in Eq. 16, is estimated by the normal mode analysis of harmonic or
quasiharmonic method [93-95]. In the MM-PB(GB)SA methods, the entropy change with the
ligand is estimated by normal mode analysis [96, 97]. The entropy S is the sum of the
translational, rotational and vibrational contributions. For the optimized configurations the
Nmode analysis was performed to obtain the 3N modes (including the 3N-6 vibrational, 3
translational and 3 rotational modes). All the three contributions are calculated by the current
21
software implementations. The entropy contribution is necessary for calculating the absolute
binding free energy, but it can be neglected in case of relative binding free energy due to its high
computational cost.
In this thesis, the MM-PB(GB)SA and normal mode analysis were computed form MMPBSA.py
module in Amber tools 1.5. The MM-PB(GB)SA was computed for several 100 snapshots and
Nmode for very few snapshots because of the computational cost.
22
23
Chapter 3
Present investigations
In this thesis, a few nutraceuticals were studied either to propose it as an active drug molecule or
to investigate their binding affinity with protein or nucleic acid targets. A multi-function seed
protein, MO2.1 from M. oleifera, was investigated in paper 1 to propose it as an efficient
antimicrobial target along with its coagulation property. In the second paper, the antimicrobial
efficiency of a plant limonoid called AZT of Azadirachta indica was investigated against a
urinary tract pathogen, P. mirabilis. Similarly, a structural analogue of berberine called
palmatine, was investigated to find its sequence specificity with different nucleotide sequences to
propose it as an anticancer agent against DNA target. The work presented in each of these papers
is described below:
3.1 Structure determination of Moringa oleifera coagulation protein (Paper 1)
The coagulant protein (MO2.1) from the seeds of an ancient medicinal plant, M. oleifera is
considered to be a water purification agent in several tropical countries [23]. Though there are
several research studies to support its water purification and antimicrobial activities, researchers
have reported different molecular weights for this protein ranging from 3.5 kDa to 6.6 kDa. In
this regard, a 6.5 kDa protein was designated for the antimicrobial and coagulant activities by a
couple of studies with recombinant or synthetic protein experiments [98, 99]. Having no crystal
structure elucidated, in 2003, Suarez et al. proposed a homology model for this 6.5 kDa protein,
which was 70% homologous to another plant protein called napin (a homologous 2S-albumin),
but they have not reported its amino acid sequence. This sequence is not effective to study its
interactions with other molecules (ligands). So, we wanted to establish a secondary structure
based on homology modeling, considering its amino acid sequence that can be used for
computational analysis to study the protein’s interaction to address the pharmaceutical
advantages.
In this study, homology modeling for the polypeptide sequence of MO2.1 was performed with
MabinlinB protein (Fig. 3.1(A)) and obtained 55 residues homologous sequence of MO2.1 instead
of 60 residues. So, we supplemented the remaining 5 residues (QGPGR) at the N terminal to
24
establish an accurate model for MO2.1. The final sequence revealed a 100% sequence similarity
with the sequence reported by Broin et al. (2002) [98], while 97% with the polypeptide sequence
of MO2.1 (P24303). In comparison with the sequence of P24303, the model peptide showed
differences at 13th and 23rd
residues. The proposed tertiary structure of the monomeric form of
MO2.1 was depicted in Fig.3.1 (B), where the region of five residues (QGPGR) that were not
obtained from homology modeling with MabinlinB was highlighted in red.
Figure 3.1. Homology modeling of MO2.1.
(A) Predicted structure for the monomer MO2.1, red coloured region indicates the five amino acid
residues (QGPGR) that were supplemented at N-terminal. The remaining cyan coloured helices
and loops were obtained from homology modeling. (B) The minimum energy structures of dimer
of MO2.1 protein. The Gly of monomer 1 (40th residue) and Val (42nd residue) shown as ball
and stick model in cyan make contacts of distance less than 4 Å, whereas the residues making
contacts within a distance range 4–5 Å were depicted in yellow and colors. The blue and red
distinguish the two monomers with its respective residues.
The model peptide had only 10 three helices and loop regions when compared to four helical
structures of MabinlinB. To understand the dimer formation of MO2.1 protein, a protein–protein-
docking study was carried out with two monomers using CLUSPRO software. As can be seen
from Fig.3.1 (B), the dimer was formed by both loop-loop and helix-helix contacts of the
monomers, while the binding free energy calculations revealed a stable dimerization property.
Our collaborators have evaluated this phenomenon through their experiments with recombinant
protein of MO2.1 and mass spectrometric analysis. Conclusively, we have developed a tertiary
structure for MO2.1 by homology modeling that can interact with other proteins and ligands.
(A) (B)
25
3.2 Azadirachtin as an antimicrobial compound (Paper II)
In the recent years, the incidence of urinary tract infections due to P. mirabilis is considerably
increasing worldwide. This pathogen not only affects the kidneys, but also cause bacteriuria,
pyelonephritis, bacteremia, and shock, which are very serious problems in clinical world. In this
context, the GST contributes to resistance by either enhancing their efflux systems or
transporting such compounds to outer surface of the cell. In order to address this problem, we
chose to analyze the limonoid, AZT (Fig.3.2), obtained from neem tree (Azadirachta indica).
Either the whole parts or their crude extracts of different parts of neem tree are used as an
antimicrobial agent from the ancient days, traditionally. The liminoid, AZT has significant
properties to be treated as an antimicrobial agent, because of the oxygen functionality in various
carboxylic esters that are responsible for the detoxification process. Among several
pharmaceutically important compounds of neem tree, we proposed AZT as a better inhibitor to
overcome the antimicrobial resistant property for P. mirabilis. In this context, we investigate the
interactions between GST and AZT because the GST detoxifies toxic chemicals and
antimicrobial drugs to protect P. mirabilis [63]. The detoxification process involves the binding
of GSH to G-site of GST reducing the pKa value of the thiols from ~9.2 to ~6.5 pH units [100].
This led us to develop an inhibitor for GST that can bind efficiently to GST than GSH, as a
whole.
Figure 3.2: Optimized Structure of AZT by B3LYP/6-31+G(d) method using Gaussian09
26
Our preliminary analysis revealed an efficient binding AZT and GST structures, as shown in
Fig.3.3. We also analyzed the stability of AZT-GST complex and the relative binding strength of
GSH and AZT with GST. The binding free energy of GST-GSH and GST-AZT complexes by
MM-PB(GB)SA method are -14.43/-17.59 kcal/mol and -26.24/-29.90 kcal/mol. From both these
results, we predict that GST-AZT complex is more stable than GST-GSH complex around 15-10
kcal/mol.
Furthermore, the Nmode analysis were performed to calculate the entropy contributions of the
complexes and unbound individual components. The ΔS value obtained of GST-GSH and GST-
AZT complexes are not to be the real binding free energy; but the values of GST-AZT complex -
21.92 kcal/mol and favorable than GST-GSH complex -20.20 kcal/mol. Analysis of this study
concludes that the AZT can bind instead of GSH to GST to initiate the detoxification process on
the bacterial cell and stop the growth. Hence the AZT, a phytochemical can be considered as a
pharmaceutically advantageous antimicrobial agent for the treatment of bacterial infections,
especially P. mirabilis. However, further in vivo and in situ studies are necessary for
pharmological assessment of AZT to describe its antimicrobial property.
Figure 3.3: Docked structure of the GST-AZT complex by Autodock
27
3.3 Palmatine, a nucleotide sequence specific molecule for cancer treatment (Paper III)
The protoberberine alkaloid, palmatine is a found in several plants like Phellodendron amurense,
Rhizoma coptidis/Coptis Chinensis and Corydalis yanhusuo and this compound is also useful in
the treatment of jaundice, dysentery, hypertension, inflammation, liver diseases and viral
infections. The tetracyclic structure of palmatine (7, 8, 13, 13atetrahydro-9, 10-dimethoxy-
berberinium) is reported to have anticancer and antimicrobial activities, especially by binding
with various polynucleotides that were sequence dependent. As the pharmaceuticals are
enthusiastic in developing dual-activity molecules especially that act on specific nucleotide-rich
sequences, we are proposing palmatine as one of such candidates, based on the available
laboratory reports.
Among several reports, the experiments of Bhadra et al., [67] Fig. 3.4 prompted us to investigate
in this direction. It is reported that the base dependent binding of the palmatine to four synthetic
polynucleotides, poly(dA).poly(dT), poly(dA–dT).poly(dA–dT), poly(dG).poly(dC) and
poly(dG–dC).poly(dG–dC) of a DNA molecule via intercalation and external stacking properties
with heterogeneous AT and GC sequences. The absorbance measurements of intrinsic binding
constants decreased in the order poly(dA).poly(dT)>poly(dA–dT).poly(dA–dT)>poly(dG–
dC).poly(dG–dC)>poly(dG).poly(dC) and this could be attributed to the different helical
parameters of the DNA helix. The noncovalent complexes of this cytotoxic alkaloid exhibited the
highest binding affinity (much higher than berberine) to the AT rich oligonucleotides and
demonstrated a remarkable base pair specific DNA binding. This specific activity prompted us to
think about utilizing palmatine against sequence-specific DNA molecules, especially during the
uncontrolled cancer conditions. Our idea was to propose palmatine as a lead molecule for either
sequence specific or mutation specific DNA target that could have been aroused from either
cancer or other form of horizontal mutations.
In this study, we designed four polynucleotides of poly(dA).poly(dT), poly(dA–dT).poly(dA–
dT), poly(dG).poly(dC) and poly(dG–dC).poly(dG–dC), which aimed to understand their
molecular mechanisms to check if they form stable structures. As depicted in the Fig. 3.5, the
equilibrated structures from the above sequences were generated and that was used for further
studies to analyze the interactions with the ligand (palmatine). Binding free energies of these
28
DNA structures indicated that the ligand is more specific to AT-rich sequence [(-45.48 kcal/mol
(GBSA) & -38.12 kcal/mol (PBSA)] and weak binding to GC-rich sequence [(-26.47 kcal/mol
(GBSA) & -21.13 kcal/mol (PBSA)], indicating the efficacy of this nucleotide-specific target,
which agrees with the experimental data. These results suggest that the palmatine is a sequence
specific inhibitor molecule for DNA based therapeutic agents, especially in cancer treatment.
Figure 3.4: The concentration of palmatine bound to different nucleic acid sequences
(Adopted from Bhadra et al., 2007 [67]).
Figure 3.5: Molecular simulations of sequence specific oligonucleotides with palmatine.
29
Future study
These results can be used for further studies to investigate the helical parameters of such DNA
sequences to understand the binding interactions of palmatine. Such approaches will not only
explain the intercalating properties of palmatine, but would certainly be helpful to understand
similar phenomena of other DNA intercalators.
30
31
Bibliography
1 Sneader, W. (1996) Drug Prototypes and Their Exploitation. Wiley, UK.
2 Newman, D. J. and Cragg, G. M.; J. Nat. Prod., 2007, 70: 461–477.
3 Koehn, F. E. and Carter, G. T.; Nature Reviews Drug Discovery, 2005, 4: 206-220.
4 Newman, D. J and Cragg, G. M; Curr Drug Targets, 2006, 7: 279-304.
5 Harvey, A. L.; Drug Discovery Today, 2008, 13: 19-20.
6 Brower, V.; Nat Biotechnol., 1998, 16: 728-731.
7 Kalra, E. K.; AAPS PharmSci., 2003, 5: 27-28.
8 Hardy, G.; Nutrition, 2000, 16: 688.
9 Barbara, S., Ribnicky, D. M., Poulev, A., Logendra, S. C., William T. and Raskin, I; A
natural history of botanical therapeutics, Metabolism Clinical and Experimental 57
(Suppl 1) (2008) S3–S9
10 Raskin, I. and Ripoll, C.; Curr Pharm Des., 2004, 10: 3419-3429.
11 Cragg, G. M., Newman, D. J. and Snader, K. M.; J. Nat. Prod., 1997, 60: 52-60.
12 Twyman, R. M., Schillberg, S. and Fischer, R.; Expert Opin Emerg Drugs, 2005, 10:
185–218.
13 Desai, P., Shrivastava, N. and Padh, H.; Biotechnol Adv., 2010, 28: 427–435.
14 Ma, J.K-C. et al., Nat. Rev. Genet., 2003, 4: 794–805.
15 Sijmons, P. C., Dekker, B. M. M, Schrammeijer, B., Verwoerd, T. C., Van Den Elzen, P.
J. M. and Hoekema, A.; BioTechnol., 1990, 8: 217–221.
16 Hiatt, A., Cafferkey, R, and Bowdish, K.; Nature, 1989, 342: 76–78.
17 Schillberg, S, Fischer, R, and Emans, N.; Cell Mol Life Sci., 2003, 60: 433–445.
18 Richard, M. T., Stefan, S., and Rainer Fischer, , A. Wang and S. Ma (eds.), Molecular
Farming in Plants: Recent Advances and Future Prospects, DOI 10.1007/978-94-007-
2217-0_7.
19 Julian K-C. Ma1, Rachel Chikwamba, Penny Sparrow, Rainer Fischer, Richard Mahoney,
and Richard M. Twyman, TRENDS in Plant Science, 2005, 10: 580-585.
20 Peczuh, M. W. and Hamilton, A. D. Chem. Rev., 2000, 100: 2479–2494.
21 Bennett, R. N.; Mellon, F. A.; Foidl, N.; Pratt, J. H.; Dupoint, S. M.; Perkins, L. And
Kroon, P. A.; J. Agri. Food Chem., 2003, 51: 3546-3553.
22 Fahey, J. W.; Trees for Life, 2005, 1: 5.
32
23 Anwar, F., Latif, S., Ashraf, M., and Gilan, A.; Phytotherapy Research, 2007, 21: 17–25.
24 Ndabigengesere, A., and Narasiah, K.S.; Wat. Res., 1998, 32: 781- 791.
25 Okuda, T., Baes, A. U., Nishijimam, W., and Okadam, M.; Water Research, 2001, 35:
830–834.
26 Ghebremicahel, K. A., Gunaratna, K. R., Henriksson, H., Brumer, H., and Dalhammar,
G.; Water Research, 2005, 39: 2338–2344.
27 Faizi, S.; Siddiqui, B. S.; Saleem, R.; Siddiqui S.; Shaheen, K.; Aftab K.; and Giliani
A.H; New trends in natural products chemistry, 1998, 175-183.
28 Lalas S, Tsaknis J.; J Am Oil Chem Soc., 2002, 79: 677–683.
29 Akhtar, M, Hasany, S. M., Bhanger, M. I., and Iqbal, S.; Journal of Hazardous Materials,
2007, 141: 546-556.
30 Guevara, A. P., Vargas, C., Sakurai, H., Fujiwara, Y., Hashimoto, K., Maoka, T.,
Kozuka, M., Ito, Y., Tokuda, H. and Nishino, H.; Mutation Research, 1999, 440: 181-188
37.
31 Mekonnen, Y., Yardley, V., Rock, P. and Croft, S.; Phototherapy Research, 1999, 13:
538–539.
32 The Wealth of India 1962. Raw Materials, Vol. VI: L-M; Council of Scientific and
Industrial Research: New Delhi, 425–429.
33 Padmarao, P., Acharya, B. M., and Dennis, T. J.; Bulletin of Medico-Ethno-Botanical
Research, 1996, 17: 141–151.
34 Dahot, M. U.; Pak J Biochem, 1988, 21: 1–24.
35 Ruckmani, K. Kavimani, S. Anandan, R. and Jaykar, B.; Indian J Pharm Sci., 1998, 60:
33–35.
36 Morton, J.F.; Economic Botany, 1991, 45: 318-333.
37 Fuglie, L.J.; (ed.) 2001. Dakar: Senegal 63.
38 Makonnen, E, Hunde, A and Damecha, G.; Phytother Res., 1997, 11: 147–148.
39 Kurma, S. R. and Mishra, S. H.; Pharmaceutical Biology, 1998, 36: 295-300.
40 Nair, A. G. R. and Subramanian, S. S.; Curr Sci., 1962, 31: 155–156.
41 Bhattacharya, S. B., Das, A. K., and Banerji, N.; Carbohydr Res., 1982, 102: 253–262.
42 Dahot, M. U.; J. Islamic Acad. Sci., 1998, 11: 1-6.
33
43 Tsaknis, J., Lalas, S., Gergis, V., Dourtoglou, V. and Spiliotis, V.; J Agric Food Chem.,
1999, 47: 4495–4499.
44 Siddhuraju, P, and Becker, K. J Agric Food Chem., 2003, 15: 2144–2155.
45 Mehta, L. K , Balaraman, R , Amin, A. H., Bafna, P. A., and Gulati, O. D.; J
Ethnopharmacol., 2003, 86: 191–195.
46 Tahiliani, P, and Kar, A.; Pharmacol Res., 2000, 41: 319–323.
47 Ghasi, S., Nwobodo, E., and Ofili, J. O.; J. Ethnopharmacol., 2000, 69: 21-25.
48 Abilgos, R. G. and Barba, C. V. C.; J. Science, 1999, 128: 79-84.
49 Ghebremicahel, K. A., Gunaratna, K. R., Henriksson, H., Brumer, H., and Dalhammar,
G.; Water Research, 2005, 39: 2338–2344.
50 Ndabigengesere, A., Narasiah, K. S., and Talbot, B. G.; Water research, 1995, 29: 703–
710.
51 Gassenschmidt, U., Jany, K. D., Tauscher, B. and Niebergall, H.; Biochimica et
Biophysica Acta, 1995, 1243, 477–481.
52 Okuda, T., Baes, A. U., Nishijimam, W. and Okadam, M.; Water Research, 2001, 35:
830–834.
53 García-Fayos, B., Arnal, J. M., Verdú, G., and Rodrigo, I.; International conference on
food innovation. Universidad Politecnica de Valencia, Valencia, Spain, (2010, October).
54 Agrawal, H., Shee, C. and Sharma, A. K.; Global Biotechnology Biochemistry, 2007, 2:
36–39.
55 Thoh, M., Kumar, P., Nagarajaram, H. A. and Manna, S. K.; J. Biol. Chem. 2010, 285:
5888–5895.
56 Kraus, W., Bokel, M., Bruhn, A., Cramer, R., Klaiber, I., Klenck, A., Nagl, G., Pöhnl, H.,
Sadio, H. and Vogler, B.; Tetrahedron, 1987, 43: 2817-2830.
57 Govindachari, T. R., Gopalakrishnan, A. G., Rajan, A. S. S., Kabaleeswaran, B. V., and
Lessinger, L.; Acta. Cryst., 1996, B52: 145-150.
58 Govindachari, T. R. Sandhya, G. and Ganeshraj, S. R.; Chromatographia, 1991, 31: 303-
305
59 Sharma, V., Walia, S., Kumar, J., Muraleedharan, G. N. and Parmar, B. S.; J. Agric. Food
Chem., 2003, 51: 3966-3972.
34
60 Jakobsson, P. J., Morgenstern, R., Mancini, J., Ford-Hutchinson, A. and Persson, B.;
Protein. Sci., 1999, 8: 689–692.
61 Chen, C. Y., Chen, Y. H., Lu, P. L., Lin, W. R., Chen, T. C. and Lin, C. Y.; J Microbiol.
Immunol. Infect., 2012, 45: 228- 236.
62 Lockhart, S. R., Abramson, M. A., Beekmann, S. E., Gallagher, G., Riedel, S., Diekema,
D. J., Quinn, J. P. and Doern, G. V.; J. Clin. Microbiol., 2007, 45: 3352-3359.
63 Allocati, N., Favaloro, B., Masulli, M., Alexeyev, M. F. and Di Ilio, C.; Biochem. J.,
2003, 373: 305–311.
64 Kuo, C.L., Chou, C.C., and Yung, B.Y.M.; Cancer Lett., 1995, 93: 193–200.
65 Schmeller, T., atz-Br ning B. and Wink M.; Phytochemistry, 1997, 44: 257–266.
66 Kluza, J. Baldeyrou, B. Colson P., Rasoanaivo P., Mambu L., Frappier F. and Bailly, C.;
Eur. J. Pharm. Sci., 2003, 20: 383–391.
67 Bhadra, K., Maiti, M. and Kumar G. S; Biochimica et Biophysica Acta, 2007, 1770:
1071–1080.
68 Hayes, J. M. and Archontis, G.; Molecular Dynamics - Studies of Synthetic and
Biological Macromolecules, edited by Wang, L.; 2012, 9: 171.
69 Leach, A. R., Shoichet, B. K. and Peishoff, C. E.; J Med Chem., 2006, 49: 5851.
70 Fischer, E.; Ber. Dtsch. Chem. Ges., 1894, 27: 2985.
71 Kitchen, D. B., Decornez, H., Furr, J. R., and Bajorath, J.; Nature reviews, 2004, 3: 935.
72 Comeau, S. R., Gatchell, D. W., Vajda, S. J. and Camacho, C. J.; Nucleic Acids
Research, 2004, 32.
73 Sanner, M. F. J.; Mol. Graphics Mod., 1999, 17: 57-61.
74 Gilson, M. K. and Zhou, H-X.; 2007, 36: 21–42.
75 Hayes, J. M. and Leonidas, D. D.; Mini Reviews in Medicinal Chemistry, 2010, 10:
1156-1174.
76 Case, D. A., Darden. T. A., Cheatham. I. T. E., Simmerling, C. L., Wang, J., Duke, R. E.,
Luo, R. and Kollman, P. A.; Amber tools 11; University of California, San Francisco,
2010 (http://ambermd.org/).
77 Frisch, M. J., et al., Gaussian09, Revision A.1; Gaussian, Inc., Wallingford CT, 2009
(http://www.gaussian.com/).
78 Ryckaert, J. P., Ciccotti, G. and Berendsen, H. J. C.; J. Comput. Phys. 1977, 23: 327-341.
35
79 Qiu, D., Shenkin, P.S., Hollinger, F.P. and Still, W.C.; J. Phys. Chem. A., 1997, 101:
3005-3014.
80 Still, W.A., Tempczyk, A., Hawley, R.C. and Hendrickson, T.; J. Am. Chem. Soc., 1990,
112: 6127-6129.
81 Watanabe, Y. S., Kim, J. G., Fukunishi, Y. and Nakamura, H.; Chem. Phys. Lett., 2004,
400: 258- 263.
82 Kollman, P. A., Massova, I., Reyes, C., Kuhn, B., Huo, S., Chong, L., Lee, M., Lee, T.,
Duan, Y., Wang, W., Doninni, O., Cieplak, P., Srinivasan, J., Case, D. A. and Cheatham
III, T. E.; Acc. Chem. Res., 2000, 33: 889-97.
83 Deng, Y. and Roux, B.; Journal of Physical Chemistry B, 2009, 113: 2234-2246.
84 Pearlman, D. A.; Journal of Medicinal Chemistry, 2005, 48: 7796-7807.
85 Rastelli, G., Del Rio, A., Degliesposti, G. and Sgobba M.; Journal of Computational
Chemistry, 2010, 31: 797-810.
86 Simonson, T.; Reports of Progress in Physics, 2003, 66: 737-787.
87 Still, W. C., Tempczyk, A., Hawley, R. and Hendrickson, T.; Journal of the American
Chemical Society, 1990, 112: 6127-6129.
88 Archontis, G., Simonson, T. and Karplus, M.; Journal of Molecular Biology, 2001, 306:
307-327.
89 Hendsch, Z. and Tidor, B.; Protein Science, 1999, 8: 1381-1392.
90 Case, D. A., Cheathham, T. E., III; Darden, T., Gohlke, H., Luo, R., Merz, K. M., Jr.;
Onufriev, A., Simmerling, C., Wang, B. and Woods, R.; Journal of Computational
Chemistry, 2005, 26: 1668-1688.
91 Rocchia, W., Alexov, E. and Honig, B.; Journal of Physical Chemistry B, 2001, 105:
6507-6514.
92 Du, J., Sun. H., Xi, L., Li, J., Yang, Y., Liu, H. and Yao, X.; Journal of Computational
Chemistry, 2011, 32: 2800-2808.
93 Brooks, B.R., Janezic, D. and Karplus M.; Journal of Computational Chemistry, 1995,
16: 1522-1542.
94 Karplus, M. and Kushick, J. N.; Macromolecules, 1981, 14: 325-332.
95 Tidor, B. and Karplus, M.; Journal of Molecular Biology, 1994, 238: 405-414.
96 Brooks, B. and Karplus, M.; Proc. Natl. Acad. Sci. U S A, 1983, 80: 6571–6575.
36
97 Srinivasan, J., Cheatham, T. E., Cieplak, P., Kollman, P. A. and Case, D.A.; J. Am.
Chem. Soc., 1998, 120: 9401-9.
98 Broin, M., Santaella, C., Cuine, S., Kokou, K., Peltier, G., and Joe,t T.; Appl Microbiol
Biotechnol., 2002, 60: 114–119.
99 Suarez, M., Entenza, J. M., and Doerries, C.; Biotechnol Bioeng., 2003, 81: 13–20.
100Caccuri, A. M., Antonini, G., Board, P. G., Parker, M. W., Nicotra, M., Bello, M. L.,
Federici, G. and Ricci, G.; Biochem. J., 1999, 344: 419–425.
Paper I
Dimerization of a flocculent protein from Moringa oleifera: experimental evidence and
in silico interpretation
Asalapuram R. Pavankumar, Rajarathinam Kayathri, Natarajan A. Murugan, Qiong Zhang,
Vaibhav Srivastava, Chuka Okoli, Vincent Bulone and Hans Ågren
Journal of Biomolecular Structure and Dynamics, 2013 (In press)
Paper II
A plant alkaloid azadirachtin as an inhibitor for glutathione-S-transferase from
Proteus mirabilis
Rajarathinam Kayathri, Natarajan A. Murugan, Premkumar Albert, Hans Ågren
(Journal of Chemical Information and Modeling - communicated)
Paper III
Palmatine, a nucleotide sequence specific inhibitor for cancer therapy
Rajarathinam Kayathri, Natarajan A. Murugan, Hans Ågren
(Manuscript)
Palmatine, a nucleotide sequence specific inhibitor for cancer therapy
Rajarathinam Kayathri1, Natarajan A. Murugan
1, Hans Ågren
1
1Division of Theoretical Chemistry and Biology, School of Biotechnology, Royal Institute of
Technology (KTH), 10691 Stockholm, Sweden.
Key words: natural plant alkaloids; palmatine; sequence-specific DNA intercalator; helical
parameters; anticancer.
------------------------------------
#Corresponding author: Prof. Hans Ågren, Division of Theoretical Chemistry and Biology,
School of Biotechnology, Royal Institute of Technology (KTH), 10691 Stockholm, Sweden.
Email: [email protected]; Tel: +46 8 5537 8416.
Introduction
The palmatine is a protoberberine alkaloid, which is found in several plants like Phellodendron
amurense, Rhizoma coptidis/Coptis Chinensis and Corydalis yanhusuo. This plant alkaloid,
palmatine is being used as a natural medicinal compound in the treatment of jaundice, dysentery,
hypertension, inflammation, liver diseases and viral infections. Palmatine bears the same
tetracyclic structure (7, 8, 13, 13atetrahydro-9, 10-dimethoxy-berberinium) as similar to
berberine but differs by the nature of the substituents at positions 2, 3 on the benzo ring; being
methylene dioxy for berberine and dimethoxy for palmatine. However, the binding studies of
palmatine to nucleic acid bio-targets have been scanty and till today no data is available in the
literature on the mode, mechanism and base sequence specificity of binding of palmatine to
DNA (1).
The experiments of Kluza et al. (2003) (1) identified that DNA as a potential bioreceptor for
palmatine and suggested both intercalation and external stacking of palmatine on complexation
with natural DNA having heterogeneous AT and GC sequences. The noncovalent complexes of
several protoberberine alkaloids including berberine and palmatine with few oligonucleotides
showed that the palmatine exhibited the highest binding much higher than berberine to the
oligonucleotides studied and demonstrate a remarkable AT base pair specific intercalative mode
of DNA binding for palmatine (2). This alkaloid has been shown to exhibit significant antitumor
activity against HL-60 cells and has antimicrobial properties (3, 4). Hence, it could be very
useful if the effect of palmatine is studied with a known sequence of DNA templates, which
could be used against nucleotide-sequence specific sequences, like cancer. Apart from cancer,
several genetic disorders or diseases like cancer, ADHD disorders including Alzheimer's,
remains under functional characterization in the physiological context of disease. In spite of
high-end molecular techniques, in vivo determinations of such genes are significantly
challenging. Hence the modern computational analysis could offer reliable solution to understand
the gene functionality and also to develop novel drugs. As the pharmaceuticals are interested to
develop dual-activity molecules, especially which act on specific nucleotide-rich sequences, we
wanted to propose palmatine as one of such candidates based on the available laboratory reports.
In this scenario, the experiments of Badhra et al. (2007) (2) inspired us to investigate the
intercalative binding abilities of palmatine against small sequence-specific DNA nucleotides.
They have reported that the base dependent binding of the palmatine to four synthetic
polynucleotides, poly(dA).poly(dT), poly(dA–dT).poly(dA–dT), poly(dG).poly(dC) and
poly(dG–dC).poly(dG–dC) of a DNA molecule via intercalation and external stacking properties
with heterogeneous AT and GC sequences. The absorbance measurements of intrinsic binding
constants decreased in the order poly(dA).poly(dT)>poly(dA–dT).poly(dA–dT)>poly(dG–
dC).poly(dG–dC)>poly(dG). poly(dC) and this could be attributed to the different helical
parameters of the DNA helix. The noncovalent complexes of this cytotoxic alkaloid exhibited the
highest binding efficiency (much higher than berberine) to the AT rich oligonucleotides and
demonstrated a remarkable base pair specific intercalative mode of DNA binding. So, we
thought to explore the sequence specific binding affinity of palmatine against DNA on the minor
groove-bound state (MNS).
In this study, we designed four polynucleotides of poly(dA).poly(dT), poly(dA–dT).poly(dA–
dT), poly(dG).poly(dC) and poly(dG–dC).poly(dG–dC), and docked them with palmatine. It is
not straightforward to study the intercalation process by the computational methods. The MNS
complex is used to study the binding affinity of palmatine on the different nucleotide sequences.
Methods
Molecular docking of sequence-specific DNA and palmatine
Here, we are most likely interested on the specificity of a ligand on different polynucleotide
sequences. The nucleic acid builder (NAB) program (5) was used to generate 18 base pair DNA
sequences of poly d(A).poly d(T), poly d(G).poly d(C), poly(dA-dT).poly(dA-dT), poly(dG-
dC).poly(dG-dC). The ligand structure palmatine was optimized by B3LYP/6-31G basis set
using Gaussian 09 (6). The equilibrated structure of DNA was considered for the docking
studies. Autodock 1.5 (7) was used to create the minor groove binding of the ligand to the
different sequences. The docked structures were used further to study the binding specificity of
the ligand. The AMBER charges of the DNA sequences and the ESP charge of palmatine was
mainatained during the docking procedure. During the docking the DNA sequences were kept
rigid and the palmatine was kept flexible. The grid box of 60x60x60 was generated with a
spacing of 0.375A by including the polar hydrogens. The LMA was used for all the runs of
population size of 150 includes 2.5x106 energy evaluations. The best-docked conformation was
chosen from the 10 docked structures based on the lowest binding energy. The intermolecular
and intramolecular energies contribute to the binding energy. The various interactions like van
der Waals, electrostatic, hydrogen bonding and desolvation energy between the receptor and the
ligand contributes to the decomposition energy of the total intermolecular energy. From the
docking studies we could infer the binding affinity, the interaction nature and the initial mode of
binding of the ligand and DNA. In this case we consider the MNS binding of palmatine with the
sequence to know its specificity to the sequences.
Molecular dynamic simulations oligonucleotides and palmatine
Amber 11 software (8) used for the MD simulation of MNS complex of DNA-palmatine. The
PARM99 was used for all the four different sequences and GAFF for the ligand, palmatine. The
MNS complex was solvated by TIP3P water molecules in a cubic box size of 15 and neutralized
by adding 22NA ions. The final solvated complex was minimized for the time step of 500 steps
followed by heating upto 300K by Berendesen thermostat for 50ps. The energy minimized
structure was further used for the Langevin dynamics with weak restraints with the maintenance
of the pressure until the convergence of the density and the total energy. The production run of
30ns was performed to extract the trajectories for the binding free energy calculation using the
SHAKE algorithm (9). The electrostatic and VDW was considered with the non-bonded cutoff of
8. The PBC was continued throughout the simulation in an isothermal-isobaric ensemble, of
300K and 1 atmospheric pressure.
Binding free energy calculations
To know the binding free energy of palmatine with the different DNA sequences in presence of
the solvent is studied by a more rigorous method like molecular mechanics Poisson-Boltzmann
surface area and the generalized-Born surface area methods (MM-PBSA, MM-GBSA).
MMPBSA.PY module in Amber tools 1.5 (10, 11) in which the implicit continuum solvent
model was used to calculate the binding free energy. The MM-PBSA and MM-GBSA models
differ from the electrostatic interaction between the solvent and the solute. The binding free
energy was calculated from the 500 snapshots has been extracted from the last 3ns from the 30ns
simulation of the production run. The formula to explain the binding free energy by:
ΔG = Gcomplex – Greceptor – Gligand
where,
Gcomplex - Binding energy of DNA/palmatine complexes
Greceptor - Binding energy of DNA (poly aa and poly gg sequences)
Gligand – Binding energy of palmatine
The binding free energy of the system is expressed by:
G= EMM + GPB/GGB + Gnon-polar – TS
EMM - Total molecular mechanics energy of the system in the gas phase
S – Entropy of the system
The solvent accessible surface area (SASA) and polar/non-polar terms are calculated using the
PB/GB model. The molecular mechanical energy, the polar and non-polar solvation free energy
and the entropy terms contributes the binding free energy. The free energy average of the
trajectories of the receptor, ligand and the complex were calculated. The more stable complex is
determined by the large value of the binding free energy and the negative value shows the
favorable complex formation.
Results and discussion
In order to address this problem, we designed sequence specific 18 base pair nucleotides of
poly(dA).poly(dT), poly(dA–dT).poly(dA–dT), poly(dG).poly(dC) and poly(dG–dC).poly(dG–
dC) through NAB programme. Only these sequences alone were simulated to check their
stability (Figure. 1).
As depicted in the figure 1, the equilibrated structures for the above sequences were generated
and that was used for further studies to analyze the interactions with the ligand (palmatine).
These differences in the structure of the DNA helices can be attributed to the change in helical
parameters. The DNA helix gains absolute and relative conformations through the helical
parameters, grouped in two categories: base-pair and base-step parameters, which defines the
point of distortion of a wobble base pair, the distance of helix axis, and non-linear DNA helical
axis. This can be explained by the figure 2.
Figure 1: The longitudinal view of the sequence was captured and depicted, to describe the
change in the bonding angles, and variations in the positions of nucleotides between normal and
simulated sequence.
Figure 2: The different helical parameters of DNA.
Figure 3: Molecular simulations of sequence specific oligo nuclucleotides with palmatine
Table 1: Poisson Boltzmann and Generalized Born model calculations for palmatine with AT
rich sequence.
Poisson Boltzmann
Contributions Complex Receptor Ligand Delta
Mean Mean Mean Mean σ
vdW -670.78 -622.92 -2.33 -45.54 0.45
Eel 5644.86 6365.14 -21.79 -698.50 1.62
Epb -11134.44 -11797.81 -45.28 708.65 1.56
Ecavity 37.82 38.13 2.43 -2.73 0.02
Ggas 4974.08 5742.23 -24.11 -744.04 1.99
Gsol -11096.62 -11759.69 -42.85 705.92 1.54
Gtotal -6122.54 -6017.46 -66.97 -38.12 0.56
Generalized Born
Contributions Complex Receptor Ligand Delta
Mean Mean Mean Mean σ
vdW -670.78 -622.92 -2.33 -45.54 0.45
Eel 5644.86 6365.14 -21.79 -698.50 1.62
Epb -10708.97 -11369.11 -43.27 703.40 1.56
Ecavity 51.74 52.13 4.46 -4.85 0.02
Ggas 4974.08 5742.23 -24.11 -744.04 1.99
Gsol -10657.23 -11316.98 -38.80 698.56 1.54
Gtotal -5683.15 -5574.75 -62.92 -45.48 0.56
vdW – van der Waals from MM force field; Eel – electrostatic energy from MM force field;
Epb/Egb – solvation free energy from the electrostatic contribution by PB/GB ; Ecavity/Esurf –
solvation free energy from the non-polar contribution calculated by PB/GB, respectively. Ggas –
total gas phase energy; Gsol - contributions to solvation; Gtotal – the final total binding free energy
(kcal/mol)
Table 2: Poisson Boltzmann and Generalized Born model calculations for palmatine with GC
rich sequence.
Poisson Boltzmann
Contributions Complex Receptor Ligand Delta
Mean Mean Mean Mean σ
vdW -636.75 -606.94 -2.23 -27.58 0.38
Eel 8046.33 8673.59 -14.55 -612.70 1.52
Epb -11177.00 -11753.47 -44.48 620.95 1.61
Ecavity 38.41 37.75 2.46 -1.80 0.03
Ggas 7409.58 8066.64 -16.78 -640.28 1.78
Gsol -11138.59 -11715.72 -42.01 619.15 1.59
Gtotal -3729.01 -3649.08 -58.80 -21.13 0.34
Generalized Born
Contributions Complex Receptor Ligand Delta
Mean Mean Mean Mean σ
vdW -636.75 -606.94 -2.23 -27.58 0.38
Eel 8046.33 8673.59 -14.55 -612.70 1.52
Epb -10754.11 -11328.83 -42.63 617.35 1.54
Ecavity 52.58 51.66 4.46 -3.54 0.04
Ggas 7409.58 8066.64 -16.78 -640.28 1.78
Gsol -10701.53 -11277.17 -38.16 613.81 1.51
Gtotal -3291.95 -3210.52 -54.95 -26.47 0.38
The binding affinity of the complex is explained in terms of electrostatic, van der Waals, polar
and non-polar solvation energy contribution. The contributions of these terms for palmatine with
poly(dA-dT).poly(dA-dT) and poly(dG-dC).poly(dG-dC), are shown in Table 1 and 2. The
binding specifity of palmatine with poly(dA-dT).poly(dA-dT) is due to the large contribution of
van der Waals (-45.54 kcal/mol) and non-bonded interaction terms (-744.04 kcal/mol) (Table 1)
and in case of poly(dG-dC).poly(dG-dC)-palmatine -27.58 kcal/mol and -640.28 kcal/mol. The
difference in the above contributions are well observed in the complex. The binding free energies
of these DNA structures indicate that the ligand is more specific to AT-rich sequence [(-45.48
kcal/mol (GBSA) & -38.12 kcal/mol (PBSA)] and weak binding to GC-rich sequence [(-26.47
kcal/mol (GBSA) & -21.13 kcal/mol (PBSA)], indicating the efficacy of this nucleotide-specific
target, which explains the AT- rich sequence specificity from the experiments (2). The binding
affinity of palmatine reported here is based on the MNS, because the direct intercalation for
small ligands is milliseconds and it is much more challenging to study by computational
methods. (12) These results suggest about the ligands sequence specificity for developing DNA
based therapeutic agents, especially in cancer treatment.
References
1. J. Kluza, B. Baldeyrou, P. Colson, P. Rasoanaivo, L. Mambu, F. Frappier, C. Bailly,
Cytotoxicity and DNA binding properties of the plant alkaloid burasaine, Eur. J. Pharm. Sci.
20 (2003) 383–391.
2. Kakali Bhadra, Motilal Maiti, Gopinatha Suresh Kumar, Molecular recognition of DNA by
small molecules: AT base pair specific intercalative binding of cytotoxic plant alkaloid
palmatine, Biochimica et Biophysica Acta 1770 (2007) 1071–1080.
3. C. L. Kuo, C.C. Chou, B.Y.-M. Yung, Berberine complexes with DNA in the berberine-
induced apoptosis in human leukemic HL 60 cells, Cancer Lett. 93 (1995) 193–200.
4. T. Schmeller, B. Latz-Brüning, M. Wink, Biochemical activities of berberine, palmatine and
sanguinarine mediating chemical defence against microorganisms and herbivores,
Phytochemistry 44 (1997) 257–266.
5. Macke, T.; Case., D. A. Modeling Unusual Nucleic Acid Structures; Leontes, N.B., Santa-
Lucia, J., Jr., Eds.; American Chemical Society: Washington, DC, 1998
6. Frisch, M. J.; et al., Gaussian09, Revision A.1; Gaussian, Inc., Wallingford CT, 2009
(http://www.gaussian.com/).
7. Sanner, M. F. Python: A Programming Language for Software Integration and
Development. J. Mol. Graphics Mod. 1999, 17, 57-61.
8. Case, D. A.; Darden. T. A.; Cheatham. I. T. E.; Simmerling, C. L.; Wang, J.; Duke, R. E.;
Luo, R.; Kollman, P. A. Amber tools 11; University of California, San Francisco, 2010
(http://ambermd.org/).
9. Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. Numerical Integration of the Cartesian
Equations of Motion of a System with Constraints: Molecular Dynamics of n-Alkanes. J.
Comput. Phys. 1977, 23, 327-341.
10. Brandsdal, B.O.; Österberg, F.; Almlöf, M.; Feierberg, I.; Luzhkov, V. B.; Åqvist, J. Free
energy calculations and ligand binding. Adv. in Protein Chem. 2003, 66, 123-158.
11. Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S.; Chong, L.; Lee, M.; Lee, T.;
Duan, Wang, W.; Donini, O.; Cieplak, P.; Srinivasan, J.; Case, D.A.; Cheatham, T. E.
Calculating structures and free energies of complex molecules: combining molecular
mechanics and continuum models. Acc. Chem. Res. 2000, 233, 889-897.
12. Li, H. J.; Crothers, D. M.; Relaxation studies of the proflavine-DNA complex: The kinetics
of an intercalation reaction, J. Mol. Biol. 1969, 39, 461−477.