RVS

INSILICO MODELING AND DOCKING

ANALYSIS OF NEURAMINIDASE 1
OF AVIAN FLU VIRUS (H5N1)

Er. Raghvendra Sachan

2007
 CONTENTS

I. INTRODUCTION…………………………………………………………..1-2
II. INSILICO DRUG DESIGNING……………………………………………3-4
i. Rational Drug Design
ii. Structure Based Drug Design

III. REVIEW OF DISEASE: AVIAN FLU……………………………………..6-18

i. Infectivity
ii. History
iii. Humans & H5N1
iv. Mutation
v. Genetic Structure
vi. Invasion & Replication
vii. Tests & Diagnosis
viii. Prevention
IV. DRUG TARGET: NEURAMINIDASE 1………………………………….19
V. TOOLS AND SERVERS……………………………….………………..…20-29
i. PDB
ii. Blast
iii. Modeller 8v1
iv. Structure Analysis Validation Server
v. Swisspdb Viewer
vi. Ligsite
vii. Hex 4.5
viii. Ligbuilder
ix. Molecular Structure File Converter
x. ChemSketch, Molinspiration, Osiris Property Explorer.
xi. Quantum3.3.0

VI. METHODOLOGY…………………………………………………….……30-31
VII. PROCEDURE OF DRUG DESIGNING……………………………….…..32-42
i. Homology Modeling
ii. Steps Involved

VIII. RESULT AND CONCLUSION………………………………………..…..43-47

Raghvendra Sachan 2
Raghvendra Sachan 3
 INTRODUCTION

Can drug be designed?

The need for ongoing development of new drugs needs no emphasis in light of the
current global situation of health and disease. Traditionally, the process of drug development has
revolved around a screening approach, as nobody knows which compound or approach could
serve as a drug or therapy. Such almost blind screening approach is very time-consuming and
laborious. The shortcoming of traditional drug discovery; as well as the allure of a more
deterministic approach to combating disease has led to the concept of "Rational drug design"
(Kuntz 1992). Nobody could design a drug before knowing more about the disease or infectious
process than past. For "rational" design, the first necessary step is the identification of a
molecular target critical to a disease process or an infectious pathogen. Then the important
prerequisite of "drug design" is the determination of the molecular structure of target, which
makes sense of the word "rational". In fact, the validity of "rational" or "structure-based" drug
discovery rests largely on a high-resolution target structure of sufficient molecule detail to allow
selectivity in the screening of compounds. When the above information is available, the
"rational" design of drug will be more possible.

Drug design is the approach of finding drugs by design, based on their biological targets.
Typically a drug target is a key molecule involved in a particular metabolic or signalling
pathway that is specific to a disease condition or pathology, or to the infectivity or survival of a
microbial pathogen. Some approaches attempt to stop the functioning of the pathway in the
diseased state by causing a key molecule to stop functioning. Drugs may be designed that bind to
the active region and inhibit this key molecule. However these drugs would also have to be
designed in such a way as not to affect any other important molecules that may be similar in
appearance to the key molecules. Sequence homologies are often used to identify such risks.
Other approaches may be to enhance the normal pathway by promoting specific molecules in the
normal pathways that may have been affected in the diseased state. The structure of the drug
molecule that can specifically interact with the biomolecules can be modeled using
computational tools. These tools can allow a drug molecule to be constructed within the
biomolecule using knowledge of its structure and the nature of its active site. Construction of the

Raghvendra Sachan 4
drug molecule can be made inside out or outside in depending on whether the core or the R-
groups are chosen first. However many of these approaches are plagued by the practical
problems of chemical synthesis. Newer approaches have also suggested the use of drug
molecules that are large and proteinaceous in nature rather than as small molecules. There have
also been suggestions to make these using mRNA. Gene silencing may also have therapeutical
application.

Raghvendra Sachan 5
 INSILICO DRUG DESIGNING

Insilico drug designing is a form of computer-based modelling whose technologies are

applied in drug target identification or drug discovery processes. Unlike the historical method of
drug discovery, by trial-and-error testing of chemical substances on animals, and matching the
apparent effects to treatments, Insilico drug design begins with a knowledge of specific chemical
responses in the body or target organism, and tailoring combinations of these to fit a treatment
profile.

Insilico drug Designing uses a variety of computational methods to identify novel

compounds, design compounds for selectivity, efficacy and safety, and develop compounds into
clinical trial candidates. These methods fall into several natural categories – structure-based drug
design, ligand-based drug design, de novo design and homology modeling – depending on how
much information is available about drug targets and potential drug compounds.

Types of drug design:

There are two types of drug design:
a) Rational drug design
b) Structure based drug design

a) Rational Drug Design:

Unlike the historical method of drug discovery, by trial-and-error testing of chemical

substances on animals, and matching the apparent effects to treatments, rational drug design
begins with a knowledge of specific chemical responses in the body or target organism, and
tailoring combinations of these to fit a treatment profile. Due to the complexity of the drug
design process two terms of interest are still serendipity and bounded rationality. Those
challenges are caused by the large chemical space describing potential new drugs without side-
effects.
A particular example of rational drug design involves the use of three-dimensional
information about biomolecules obtained from such techniques as x-ray crystallography and
NMR spectroscopy. This approach to drug discovery is sometimes referred to as structure-based

Raghvendra Sachan 6
drug design. The first unequivocal example of the application of structure-based drug design
leading to an approved drug is the carbonic anhydrase inhibitor dorzolamide which was
approved in 1995.
Rational drug design is a process used in the biopharmaceutical industry to discover and
develop new drug compounds. RDD uses a variety of computational methods to identify novel
compounds, design compounds for selectivity, efficacy and safety, and develop compounds into
clinical trial candidates. These methods fall into several natural categories – structure-based drug
design, ligand-based drug design, de novo design and homology modeling – depending on how
much information is available about drug targets and potential drug compounds.
An important case study in rational drug design is imatinib, a tyrosine kinase inhibitor
designed specifically for the bcr-abl fusion protein that is characteristic for Philadelphia
chromosome-positive leukemias (chronic myelogenous leukemia and occasionally acute
lymphocytic leukemia). Imatinib is substantially different from previous drugs for cancer, as
most agents of chemotherapy simply target rapidly dividing cells, not differentiating between
cancer cells and other tissues

b) Structure based Drug Design:

Structure-based drug design is one of several methods in the rational drug design toolbox.
Drug targets are typically key molecules involved in a specific metabolic or cell signaling
pathway that is known, or believed, to be related to a particular disease state. Drug targets are
most often proteins and enzymes in these pathways. Drug compounds are designed to inhibit,
restore or otherwise modify the structure and behavior of disease-related proteins and enzymes.
SBDD uses the known 3D geometrical shape or structure of proteins to assist in the development
of new drug compounds. The 3D structure of protein targets is most often derived from x-ray
crystallography or nuclear magnetic resonance (NMR) techniques. X-ray and NMR methods can
resolve the structure of proteins to a resolution of a few angstroms (about 500,000 times smaller
than the diameter of a human hair). At this level of resolution, researchers can precisely examine
the interactions between atoms in protein targets and atoms in potential drug compounds that
bind to the proteins. This ability to work at high resolution with both proteins and drug
compounds makes SBDD one of the most powerful methods in drug design.

Raghvendra Sachan 7
 Structure-based design is one of the first techniques to be used in drug design. Structure-
based design refers specifically to finding and complementing the 3D structure (binding and/or
active site) of a target molecule such as a receptor protein. Chemists may be guided to subsets of
compounds with desired features to complement 3-dimensional shape of the site. From the
geometry and functional features of the binding site, complementary structures of a compound
(ligand) are so designed as to have high binding affinity with the target molecule. It is a powerful
technique to design a corresponding ligand specifically interacting with the target, particularly
for the development of a novel therapeutic through stimulation or inhibition of the receptor
protein. As mentioned above, this is keenly so in the advent of genomic approaches and in the
light of ever-increasing 3D structural data (14,680 as of 19 March, 2001) in the Protein Data
Bank.

Raghvendra Sachan 8
 REVIEW OF DISEASE

AVIAN FLU

Bird Flu, also known as Avian Flu is a respiratory disease infecting both birds and
mammals .It is caused by RNA virus of family Orthomyxoviridae .It is an Influenza A type
virus. Influenza A virus subtype H5N1, also known as A(H5N1) or simply H5N1, is a subtype
of the Influenza A virus which can cause illness in humans and many other animal species.[1] A
bird-adapted strain of H5N1, called HPAI A(H5N1) for "highly pathogenic avian influenza
virus of type A of subtype H5N1", is the causative agent of H5N1 flu, commonly known as
"avian influenza" or "bird flu".
Due to the high lethality and virulence of HPAI A(H5N1), its endemic presence, its
increasingly large host reservoir, and its significant ongoing mutations, the H5N1 virus is the
world's largest current pandemic threat, and billions of dollars are being spent researching H5N1
and preparing for a potential influenza pandemic.

Colorized Transmission Electron Micrograph of Avian Influenza

Infectivity:
H5N1 is easily transmissible between birds facilitating a potential global spread of H5N1.
While H5N1 undergoes specific mutations and reassorting creating variations which can infect
species not previously known to carry the virus, not all of these variant forms can infect humans.
H5N1 as an avian virus preferentially binds to a type of galactose receptors that populate the
avian respiratory tract from the nose to the lungs and are virtually absent in humans, occurring
only in and around the alveoli, structures deep in the lungs where oxygen is passed to the blood.

Raghvendra Sachan 9
Therefore, the virus is not easily expelled by coughing and sneezing, the usual route of
transmission.
H5N1 is mainly spread by domestic poultry, both through the movements of infected
birds and poultry products and through the use of infected poultry manure as fertilizer or feed.
Humans with H5N1 have typically caught it from chickens, which were in turn infected by other
poultry or waterfowl. Migrating waterfowl (wild ducks, geese and swans) carry H5N1, often
without becoming sick. Many species of birds and mammals can be infected with HPAI
A(H5N1), but the role of animals other than poultry and waterfowl as disease-spreading hosts is
unknown.

Virulence:
H5N1 has mutated into a variety of strains with differing pathogenic profiles, some
pathogenic to one species but not others, some pathogenic to multiple species. Each specific
known genetic variation is traceable to a virus isolate of a specific case of infection. Through
antigenic drift, H5N1 has mutated into dozens of highly pathogenic varieties divided into
genetic clades which are known from specific isolates, but all currently belonging to genotype
Z of avian influenza virus H5N1, now the dominant genotype.

History:

H5N1 isolates found in Hong Kong in 1997 and 2001 were not consistently transmitted
efficiently among birds and did not cause significant disease in these animals. In 2002 new
isolates of H5N1 were appearing within the bird population of Hong Kong. These new isolates
caused acute disease, including severe neurological dysfunction and death in ducks. This was the
first reported case of lethal influenza virus infection in wild aquatic birds since 1961. Genotype
Z emerged in 2002 through reassortment from earlier highly pathogenic genotypes of H5N1. That
first infected birds in China in 1996, and first infected humans in Hong Kong in 1997. Genotype
Z is endemic in birds in Southeast Asia, has created at least two clades that can infect humans,
and is spreading across the globe in bird populations. Mutations are occurring within this
genotype that is increasing their pathogenicity. Birds are also able to shed the virus for longer
periods of time before their death, increasing the transmissibility of the virus.

Raghvendra Sachan 10
 Humans and H5N1:

Confirmed human cases and mortality rate of avian influenza (H5N1)

As of May 31, 2007

Report dates

Country 2003 2004 2005 2006 2007 Total

cases deaths cases deaths cases deaths cases deaths cases deaths cases deaths
Azerbaijan 8 5 63% 8 5 63%

Cambodia
4 4 100% 2 2 100% 1 1 100% 7 7 100%
PR China 1 1 100% 8 5 63% 13 8 62% 2 1 50% 25 15 63%
Djibouti 1 0 0% 1 0 0%
1
Egypt 18 0 56% 16 4 25% 34 14 41%
Indonesia 1 4 1
19 2 63% 56 6 82% 22 9 86% 98 78 79%
Iraq 3 2 67% 3 2 67%
Laos 2 2 100% 2 2 100%
Nigeria 1 1 100% 1 1 100%

Thailand 1
17 2 71% 5 2 40% 3 3 100% 25 17 68%
Turkey 12 4 33% 12 4 33%
2 1
Vietnam 3 3 100% 29 0 69% 61 9 31% 93 42 45%

3 4 8 2 18
Total 4 4 100% 46 2 70% 97 2 43% 116 0 69% 43 7 63% 307 7 61%

Source: World Health Organization Communicable Disease Surveillance & Response (CSR)

Transmission & Incubation:

• From infected mammals through the air by coughs or sneezes, creating aerosols
containing the virus.
• From infected birds through their droppings.
• Influenza can also be transmitted by saliva, nasal secretions, feces and blood.
• Flu viruses can remain infectious for about one week at human body temperature, over 30
days at 0 °C (32 °F), and indefinitely at very low temperatures. Most influenza strains can be
inactivated easily by disinfectants and detergents.

Raghvendra Sachan 11
Symptoms:
Symptoms start 24 to 48 hours after infection and can begin suddenly. Chills or a chilly
sensation are often the first indication of influenza. Fever is common during the first few days,
and the temperature may rise to 102 to 103 °F. Many people feel sufficiently ill to remain in bed
for days; they have aches and pains throughout the body, most pronounced in the back and legs.
Common symptoms of the flu such as fever, headaches, and fatigue come from the huge
amounts of proinflammatory cytokines and chemokines (such as interferon or tumor necrosis
factor) produced from influenza-infected cells. Symptoms of influenza may include:
• Body aches, especially joints and throat
• Coughing and sneezing
• Extreme coldness and fever
• Fatigue
• Headache
• Irritated watering eyes
• Nasal congestion
• Nausea and vomiting
• Reddened eyes, skin (especially face), mouth, throat and nose.

Mutation:

Influenza viruses have a relatively high mutation rate that is characteristic of RNA
viruses. The segmentation of the influenza genome facilitates genetic recombination by segment
reassortment in hosts who are infected with two different influenza viruses at the same
time.H5N1 viruses can reassort genes with other strains that co-infect a host organism, such as a
pig, bird, or human, and mutate into a form that can pass easily among humans. This is one of
many possible paths to a pandemic.
Influenza A, B and C are all capable of antigenic drift. Constant point mutations occur
in the amino acid sequences of hemagglutinin and neuraminidase glycoproteins. This 'drift' in
the antigen sequence allows accumulation of mutations to give slightly new strains. Eventually,
only new strains can be detected as old strains disappear from circulation due to selection
pressure.

Raghvendra Sachan 12
 The ability of various influenza strains to show species-selectivity is largely due to
variation in the hemagglutinin genes. Genetic mutations in the hemagglutinin gene that cause
single amino acid substitutions can significantly alter the ability of viral hemagglutinin proteins
to bind to receptors on the surface of host cells. Such mutations in avian H5N1 viruses can
change virus strains from being inefficient at infecting human cells to being as efficient in
causing human infections as more common human influenza virus types. This doesn't mean that
one amino acid substitution can cause a pandemic, but it does mean that one amino acid
substitution can cause an avian flu virus that is not pathogenic in humans to become pathogenic
in humans.

Genetic Structure:
Genetic structure of H5N1 consists of following major proteins:
1. Hemagglutinin:
It is an antigenic glycoprotein found on surface of influenza viruses. It is responsible for
binding the virus to the host cell. There are 16 subtypes of hemagglutinin known to us.
Hemagglutinin protein is the receptor-binding and membrane fusion glycoprotein of influenza
virus and the target for infectivity-neutralizing antibodies. The entire hemagglutinin protein (HA)
from the H5N1 is composed of 568 amino acids, with a molecular weight of 56 kDa. The HA
molecule consists of HA1 and HA2 subunits, with the HA1 subunit mediating initial contact
with the cell membrane and HA2 being responsible for membrane fusion.
2. Neuraminidase:
Neuraminidase is a glycoside hydrolase enzyme (EC 3.2.1.18). It is frequently found as
an antigenic glycoprotein and is best known as one of the enzymes found on the surface of the
Influenza virus. Nine subtypes of influenza neuraminidase are known; many occur only in
various species of duck and chicken. Subtypes N1 and N2 have been positively linked to
epidemics in man. Neuraminidase has functions that aid in the efficiency of virus release from
cells. Neuraminidase cleaves terminal sialic acid residues from carbohydrate moieties on the
surfaces of infected cells. This promotes the release of progeny viruses from infected cells.
Neuraminidase also cleaves sialic acid residues from viral proteins, preventing aggregation of
viruses.

Raghvendra Sachan 13
3. Ribo-Nucleic Protein:
There are 8 negative sense single strands of RNA which code for different components
of virus. These include PB2, PB1, PA, HA, NP, NA, M and NS.
4. Matrix Protein 1:
The M1 protein is a matrix protein of the influenza virus. It forms a coat inside the
envelope. The M1 protein binds to the viral RNA. It also has multiple regulatory functions,
performed by interaction with the components of the host cell. The mechanisms regulated
include the export of the viral ribonucleoproteins from the host cell nucleus, inhibition of
viral transcription, and plays a role in the virus assembly and budding. The protein was
found to undergo phosphorylation in the host cell. The M1 protein forms a layer under the
patches of host cell membrane that are rich with the viral hemagglutinin, neuraminidase and M2
transmembrane proteins, and facilitates budding of the mature viruses.
5. Matrix Protein 2:
The M2 protein is a proton-selective ion channel protein, integral in the viral
envelope of the influenza A virus. The channel itself is a homotetramer, where the units are
helixes stabilized by two disulfide bonds. It is activated by low pH. It enables hydrogen ions to
enter the viral particle (virion) from the endosome, thus lowering pH of the inside of the virus,
which causes dissociation of the viral matrix protein M1 from the ribonucleoprotein RNP.
This is a crucial step in uncoating of the virus and exposing its content to the cytoplasm of the
host cell.

Raghvendra Sachan 14
Invasion & Replication:

• The virus is transmitted by infected birds through their saliva, nasal secretions,
feces and blood.
• The hemagglutinin protein on surface of flu virus binds to sialic acid, a sugar
found on surface of host cell.
• The cell membrane then engulfs the virus to form an endosome. The cell then
attempts to begin digesting the contents of endosome by acidifying its interior and
transforming it into lysosome.
• As acid enters the virus through M2 ion channel, conformational changes occur in
hemagglutinin. The globular heads of hemagglutinin fold back and the exposed
innards binds to the cell membrane resulting in its fusion with virus membrane.
• The fusion results in formation of a pore thus releasing RNA’s inside the cell.
• The nucleocapsid goes to nucleus where transcription and viral RNA replication
of all 8 strands takes place.
• Newly formed viral surface proteins – HA, NA and M2, migrate to cell
membrane. Here Neuraminidase slashes off any sialic acid that is present on cell
membrane to prevent newly formed virus to attach again to infected cell.
• The M1 protein helps in packing new viral RNA’s and internal proteins and
transports them to surface to join viral surface proteins.
• Finally newly formed virus buds off from cell.

Raghvendra Sachan 15
 Invasion and replication of avian flu virus (H5N1)

Pathology:
• Virus attacks the epithelial cells of respiratory tract.
• This results in accumulation of exudates containing inflammatory cells and
necrotic epithelial debris causing primary symptoms like fever, cough, sore throat,
muscle aches, conjunctivitis, and pneumonia.

Immune Response:
As virus enters body interleukins α or β are released. Thus macrophages come to destroy
the virus. But virus is able to infect macrophages also. This further results in release of
interleukins α or β, t cells and interferon. Thus causing CYTOKINE STORM. This results in
Acute Respiratory Distress Syndrome (ARDS) causing many complications like multiple organ
failure, dilation blood vessels, necrosis etc.

Raghvendra Sachan 16
Tests and Diagnosis:
Conformation of influenza a virus is necessary to monitor flu epidemics.
Conformational tests include:-
1. Virus isolation: In it nasopharyngeal swabs are inoculated and grown in the lab. After 3-7
days cell supernatants are checked for the presence of influenza virus by HA assay.
2. Immuno capture ELISA: This is used to detect influenza A and B nucleoprotein
antigens the antibody specific to the antigen is attached to a solid substrate. The virus and
soluble viral antigens are allowed to absorb to the antibody and any excess is washed away.
A detector antibody is added, normally linked to an enzyme, and then any unbound antibody
is washed away. The enzyme will usually cause a visible change which indicates that the
viral antigen is present.
3. RT-PCR: The first step in the analysis of a clinical specimen or a viral isolate in our assay
is the generation of a DNA copy of the viral RNA, which is accomplished by reverse
transcription coupled to PCR (RT-PCR). This was performed using two different biotinylated
combinations of PCR primers specific to the H5N1 regions of interest in order to achieve
optimal sequencing flexibility of H5N1 isolates
4. RT-LAMP: Reverse transcriptase loop-mediated isothermal amplification (RT-LAMP) is
a unique gene amplification method that can be completed within 35 min at 62.5 °C. In the
present study, RT-LAMP was used to develop a rapid and sensitive laboratory diagnostic
system for the H5N1 highly pathogenic avian influenza. The system detected H5-HA genes
in throat swabs collected from humans as well as from wild birds.

Prevention:
Vaccination and infection control:
Vaccination against influenza with a flu vaccine is strongly recommended for high-risk
groups, such as children and the elderly. These vaccines can be produced in several ways; the
most common method is to grow the virus in fertilized hen eggs. After purification, the virus is
inactivated (for example, by treatment with detergent) to produce an inactivated-virus vaccine.
Alternatively, the virus can be grown in eggs until it loses virulence and the avirulent virus given
as a live vaccine. The effectiveness of these flu vaccines is variable. Due to the high mutation

Raghvendra Sachan 17
rate of the virus, a particular flu vaccine usually confers protection for no more than a few years.
Every year, the World Health Organization predicts which strains of the virus are most likely to
be circulating in the next year, allowing pharmaceutical companies to develop vaccines that will
provide the best immunity against these strains. Vaccines have also been developed to protect
poultry from avian influenza. These vaccines can be effective against multiple strains and are
used either as part of a preventative strategy, or combined with culling in attempts to eradicate
outbreaks.
Recently the vaccine for H5N1 manufactured by Sanofi-Aventis at a plant in Swiftwater,
Pennsylvania with US approval on April 17, 2007.
Good personal health and hygiene habits are reasonably effective in avoiding and
minimizing influenza. People who contract influenza are most infective between the second and
third days after infection and infectivity lasts for around 10 days. Children are notably more
infectious than adults, and shed virus from just before they develop symptoms until 2 weeks after
infection.
Influenza spreads through aerosols and contact with contaminated surfaces, it is
important to persuade people to cover their mouths while sneezing and to wash their hands
regularly. Surface sanitizing is recommended in areas where influenza may be present on
surfaces. Alcohol is an effective sanitizer against influenza viruses, while quaternary ammonium
compounds can be used with alcohol, to increase the duration of the sanitizing action. In
hospitals, quaternary ammonium compounds and halogen-releasing agents such as sodium
hypochlorite are commonly used to sanitize rooms or equipment that have been occupied by
patients with influenza symptoms. During past pandemics, closing schools, churches and theaters
slowed the spread of the virus but did not have a large effect on the overall death rate.

Treatment:
People with the flu are advised to get plenty of rest, drink a lot of liquids, avoid using
alcohol and tobacco and, if necessary, take medications such as paracetamol (acetaminophen) to
relieve the fever and muscle aches associated with the flu. Children and teenagers with flu
symptoms (particularly fever) should avoid taking aspirin during an influenza infection
(especially influenza type B) because doing so can lead to Reye’s syndrome, a rare but
potentially fatal disease of the liver. Since influenza is caused by a virus, antibiotics have no
effect on the infection; unless prescribed for secondary infections such as bacterial pneumonia,

Raghvendra Sachan 18
they may lead to resistant bacteria. Antiviral medication is sometimes effective, but viruses can
develop resistance to the standard antiviral drugs. The two classes of anti-virals are
neuraminidase inhibitors and M2 inhibitors (adamantane derivatives). Neuraminidase inhibitors
are currently preferred for flu virus infections. The CDC recommended against using M2
inhibitors during the 2005–06 influenza season.

Neuraminidase inhibitors:
Neuraminidase inhibitors are antigenic glycoprotein surface enzyme which is essential
for influenza virus’s infectivity cycle. The iterative process of structure based drug design to
produce simple benzoic acid derivatives that are potent inhibitors of viral protein and inhibit viral
reproduction in cell culture. It is also known as Sialidase, is an essential influenza viral coat
enzyme. It cleaves sialic acid from the carbohydrate side chains at the surface of the cells,thus
enabling the virus to penetrate the polar outer cell surface the respiratory tract.
Antiviral drugs such as oseltamivir (trade name Tamiflu) and zanamivir (trade name
Relenza) are neuraminidase inhibitors that are designed to halt the spread of the virus in the
body. These drugs are often effective against both influenza A and B. Different strains of
influenza virus have differing degrees of resistance against these antiviral and it is impossible to
predict what degree of resistance a future pandemic strain might have.
It release the progeny virus from the infected cells. Flu drugs work by inhibiting some
strains of NA. They were developed based on N2 and N9.In the N1 form of the protein a small
segment called 150-loop is inverted, creating a hollow pocket that does not exist in N2 and N9
protein. During researches it is found that existing drugs interacted with the N1 protein, in the
presence of N1 inhibitors, the loop changed, its conformation to one similar to that in the N2 and
N9 proteins.
Peramivir:
Peramivir is an experimental antiviral drug being developed by Biocryst
Pharmaceuticals to treat influenza, acting as a state analogue inhibitor of
influenza neuraminidase and thereby preventing.
(1S,2S,3S,4R)-3-[(1S)-1-Acetamido-2-ethyl-butyl]-4-
(diaminomethylideneamino)-2-hydroxy-cyclopentane -1-carboxylic acid

Zanamivir:
Zanamivir is a neuraminidase inhibitor used in the treatment of and prophylaxis of both
Influenzavirus A and Influenzavirus B. Zanamivir was the first neuraminidase inhibitor

Raghvendra Sachan 19
 commercially developed. It is currently marketed by GlaxoSmithKline
under the trade name Relenza. Zanamivir was the first of the
neuraminidase inhibitors.
5-acetamido-4-guanidino-6-(1,2,3-trihydroxypropyl)-
5,6-dihydro-4H-pyran-2-carboxylic acid

Limitations:
Whilst zanamivir proved to be a potent and effective inhibitor of influenza neuraminidase
and inhibitor of influenza virus replication in vitro and in vivo, this didn't necessarily translate
into a successful clinical treatment for influenza. In clinical trials it was found that zanamivir was
able to reduce the time to symptom resolution by 1.5 days provided therapy was started within
48 hours of the onset of symptoms.
A further limitation concerns the poor oral bioavailability of zanamivir. This meant that
oral dosing was impossible, limiting dosing to the parenteral routes. Zanamivir, therefore, is
administered by inhalation - a route that was chosen for patient compliance with therapy.

Oseltamivir:
Oseltamivir is an antiviral drug that is used in the treatment and prophylaxis of both
Influenzavirus A and Influenzavirus B. Like zanamivir, oseltamivir is a neuraminidase inhibitor.
It acts as a transition-state analogue inhibitor of influenza neuraminidase, preventing new viruses
from emerging from infected cells.
Oseltamivir was the first orally active neuraminidase inhibitor commercially developed.
It is a prodrug, which is hydrolysed hepatically to the active metabolite, the free carboxylate of
oseltamivir (GS4071). It was developed by Gilead Sciences and is
currently marketed by Hoffmann-La Roche (Roche) under the
trade name Tamiflu.
3R,4R,5S)-4-acetylamino-5-amino-3-(1-ethylpropoxy)-1-cyclohexene-
1-carboxylic acid ethyl ester

Adverse effect:
Common adverse drug reactions (ADRs) associated with oseltamivir therapy include:
nausea, vomiting, diarrhea, abdominal pain, and headache. Rare ADRs include: hepatitis and

Raghvendra Sachan 20
elevated liver enzymes, rash, allergic reactions including anaphylaxis, oseltamivir may cause
dangerous psychological side effects in some people.

M2 inhibitors (adamantanes):
M2 inhibitors are matrix protein along with Hemagglutinin (HA), Neuraminidase (NA),
surface protein that make a capsid. It is a proton selective ion channel protein integral in the cell
membrane of influenza A virus. The channel itself is a homotetramer, where the units are helixes
stabilized by two disulfide bonds. It is activated by low pH.
It is located in viral envelope. It enables hydrogen ions to enter the viral particles(virion)
from the endosome, thus lowering pH of the inside of the virus, which causes dissociation of the
viral matrix protein M1 from the ribonucleoprotein RNP. This is crucial step in uncoating of
virus and exposing its content to the cytoplasm of the host cell.
The Amantadine and rimantadine are licensed and approved for both prophylaxis and
therapy. They are primary amines with a carbon ring both of which are essential for their
antiviral effect. They both work by binding to the hydrophobic domain of the M2 protein in the
virus which blocks the viral ion channel. Protons do not flow in and acidification(necessary for
the dissociation of the influenza ribonucleoprotein from the M1 protein) does not occur The
antiviral drugs amantadine and rimantadine are designed to block a viral ion channel and prevent
the virus from infecting cells. These drugs are sometimes effective against influenza A if given
early in the infection, but are always ineffective against influenza B.
Use:
They must be started within 48 hrs of symptoms onset and used for five days to be
effective. Résistance is very high.
Side effects:
Both cause CNS (nervousness, anxiety, difficulty concentrating and light headedness)
and GI effects.CNS effects are more common in Amantadine. Dosages for both drugs must be
lower in the elderly and persons with reduced kidney function.

Raghvendra Sachan 21
 POTENTIAL DRUG TARGET:

NEURAMINIDASE 1
Neuraminidase is a protein that is necessary for virus proliferation, which makes the
protein an ideal target for the development of anti-influenza drugs.
Neuraminidase is a protein that regulates the release of virus from an infected cell. New
infections are made through the release of progeny viruses from the infected cell that searches
for a new host. Inhibiting the activity of neuraminidase will halt the spread of influenza virus.

Neuraminidase is divided into 9 subtypes (N1 to N9). The database contains 1603 three-
dimensional models of all subtypes of the protein. The prediction for the three-dimensional
structure of the protein has been generated through a computational method call homology
modeling. The result of the predictions has been stored in a database to provide a resource to
respond against a potential influenza pandemic.

Raghvendra Sachan 22
 TOOLS AND SERVERS

For the designing of a potential drug for avian flu various web based tools, servers and software

were used. Some of them are listed below:

1. Databases : PDB

2. Template Search : Blast

3. Modelling : Modeller 8v1

4. Verification : Structure Analysis Validation Server

5. Visualization : Swisspdb Viewer

6. Active site Prediction : Ligsite

7. Ligand Screening : Hex 4.5

8. Ligand Generation and Optimization : Ligbuilder

9. File Format Converter : Molecular Structure File Converter

10. Docking : Hex 4.5, Quantum3.3.0

11. Structure Optimization and Validation : ChemSketch, Molinspiration,

Osiris Property Explorer.

PROTEIN DATA BANK:

The Protein Data Bank (PDB) is a website, a depositary files containing experimentally
determined atomic coordinates of biological macro molecules. It is a public domain repository
for 3-D structural data of proteins and nucleic acids. This data, typically obtained by X-ray
crystallography or NMR spectroscopy.The database is the central repository for biological
structural data. It includes placing queries to the PDB and interpreting files.

Raghvendra Sachan 23
BLAST:

In bioinformatics, Basic Local Alignment Search Tool, or BLAST, is an algorithm

for comparing biological sequences, such as the amino-acid sequences of different proteins or the
DNA sequences. A BLAST search enables a researcher to compare a query sequence with a
library or database of sequences, and identify library sequences that resemble the query sequence
above a certain threshold. Nucleotide-nucleotide BLAST (blastn) program, given a DNA query,
returns the most similar DNA sequences from the DNA database that the user specifies. Protein-
protein BLAST (blastp) program, given a protein query, returns the most similar protein
sequences from the protein database that the user specifies. Position-Specific Iterative BLAST
(PSI-BLAST) is the more recent BLAST programs, this program is used for finding distant
relatives of a protein.
Nucleotide 6-frame translation-protein (blastx) program compares the six-frame
conceptual translation products of a nucleotide query sequence (both strands) against a protein
sequence database.Nucleotide 6-frame translation-nucleotide 6-frame translation (tblastx)
program is translates the query nucleotide sequence in all six possible frames and compares it
against the six-frame translations of a nucleotide sequence database. Protein-nucleotide 6-frame
translation (tblastn) program compares a protein query against the six-frame translations of a
nucleotide sequence database. large numbers of query sequences (megablast): When comparing
large numbers of input sequences via the command-line BLAST, "megablast" is much faster than
running BLAST multiple times. BLAST searches for high scoring sequence alignments between
the query sequence and sequences in the database using a heuristic approach that approximates
the Smith-Waterman algorithm.

MODELLER 8V1:

Modeller is used for homology or comparative modeling of protein three- dimensional

structures. The user provides an alignment of a sequence to be modeled with known related
structures and MODELLER automatically calculates a model containing all non-hydrogen
atoms. MODELLER implements comparative protein structure modeling by satisfaction of
spatial restraints (2,3), and can perform many additional tasks, including de novo modeling of
loops in protein structures, optimization of various models of protein structure with respect to a

Raghvendra Sachan 24
flexibly defined objective function, multiple alignment of protein sequences and/or structures,
clustering, searching of sequence databases, comparison of protein structures, etc. Modeller is
written in Fortran-90 and is meant to run on UNIX, Windows and Apple systems.

STRUCTURE ANALYSIS & VALIDATION SERVER:

SAVS is a server for analyzing protein structures for validity and assessing how correct they
are. It utilizes 6 programs for doing this:

• PROCHECK
• WHAT_CHECK
• ERRAT
• VERIFY_3D
• PROVE

Structure Analysis & Validation Server (web page)

Raghvendra Sachan 25
SWISS PDB VIEWER:

Deep view (formerly called Swiss-PDB Viewer) is a friendly but powerful molecular
graphics program. It is designed for use with computing tools available from the Expert Protein
Analysis System, or ExPASy Molecular Biology Server in Geneva, Switzerland. Swiss PDB
Viewer is under continuing development by Nicolas Guex and Manuel C. Peitsch of Geneva
Glaxo Welcome Experimental Research. Deep view allows you to build models from scratch,
simply by giving an amino acid sequence. Deep view can find hydrogen bonds within proteins
and between proteins and ligand. It allows examining electron density maps from
crystallographic structure determination, to judge the quality of maps and models, and to identify
many common types of problems in protein models. It allows viewing several proteins
simultaneously and superimposing them to compare their structures and sequences. It compute
electrostatic potentials and molecular surfaces, and carries out energy minimization.

RAMACHANDRAN PLOT:
A Ramachandran plot (also known as Ramachandran Map or a Ramachandran
Diagram), developed by Gopalasamudram Narayana Ramachandran, and is a way to
visualize dihedral angles φ against ψ of amino acid residues in protein structure. It shows the
possible conformations of φ and ψ angles for the polypeptide. One would expect their larger side
chains would result in more restrictions and consequently a smaller allowable region in the
Ramachandran plot. In practice it does not appear the case, only the methylene group at the β
position has an influence Glycine has a hydrogen atoms with a smaller Vander Walls Radius,
instead of the methyl group at the β position. Hence it is least restricted and this is apparent in the
Ramachandran plot for Glycine for which the allowable area is considerably larger. In contrast,
the Ramachandran plot for Proline shows only a very limited number of possible combinations
of φ and ψ angles.

Raghvendra Sachan 26
 Ramachandran Plot
LIGSITE:
Automatic and efficient detection of potential small molecule-binding sites in proteins.
LIGSITE is a new program for the automatic and time-efficient detection of pockets on the
surface of proteins that may act as binding sites for small molecule ligand. Pockets are identified
with a series of simple operations on a cubic grid. Using a set of receptor-ligand complexes we
show that LIGSITE is able to identify the binding sites of small molecule ligand with high
precision. The main advantage of LIGSITE is its speed. Typical search times are in the range of
5 to 20 s for medium-sized proteins. LIGSITE is therefore well suited for identification of
pockets in large sets of proteins (e.g., protein families) for comparative studies. For graphical
display LIGSITE produces VRML representations of the protein-ligand complex and the binding
site for display with a VRML viewer such as WebSpace from SGI.

Raghvendra Sachan 27
HEX:
Hex was written by Dave Ritchie, computing science, and university of Aberdeen.Hex is
an interactive molecular graphics program for calculating and displaying feasible docking modes
of pairs of protein and DNA molecules. Hex can also calculate small-ligand/protein docking
(provided the ligand is rigid), and it can superpose pairs of molecules using only knowledge of
their 3D shapes.
The main thing which distinguishes Hex from other macromolecular docking programs
and molecular graphics packages is its use of spherical polar Fourier correlations to accelerate
the docking and superposition calculations. The graphical nature of Hex came about largely
because I wanted to visualize the results of such docking calculations in a natural and seamless
way, without having to export unmanageably many (and usually quite big) coordinate files to
one of the many existing molecular graphics packages. For this reason, the graphical capabilities
in Hex are relatively primitive compared to commercial packages, although these days one can
do quite a lot with a few calls to OpenGL.
In Hex's docking calculations, each molecule is modeled using 3D parametric functions
which are used to encode both surface shape and electrostatic charge and potential distributions.
The parametric functions are based on expansions of real orthogonal spherical polar basis
functions. Essentially, this allows each property to be represented by a vector of coefficients.
Hex's surface shape representation uses a novel 3D surface skin model of protein topology,
whereas the electrostatic model is derived from classical electrostatic theory. By writing an
expression for the overlap of pairs of parametric functions, one can derive an expression for a
docking score as a function of the six degrees of freedom in a rigid body docking search. With
suitable scaling factors, this docking score can be interpreted as an interaction energy, which we
seek to minimize. Due to the special orthogonality property of the basis functions, the correlation
(or overlap as a function of translation/rotation operations) between a pair of 3D functions can be
calculated using expressions which involve only the original expansion coefficients. In many
respects, this approach is similar to conventional fast Fourier transform (FFT) docking methods
which use a Cartesian grid to perform the Fourier transforms. However, the FFT approach only
accelerates a docking search in three (translational) degrees of freedom whereas with a spherical
polar approach, we can both translate (with some effort) and rotate (relatively easily) the

Raghvendra Sachan 28
coefficient vectors to generate and evaluate candidate docking orientations in what is effectively
a six dimensional Fourier correlation.
In the spherical polar approach, it is natural to assign the six rigid body degrees of
freedom as five Euler rotation angles and an intermolecular separation. Thus, in complete
contrast to the FFT approach, the rotational part of a docking search is the "easy bit" and
modelling translations becomes the "hard part". Fortunately, however, only a few translations
(typically about 40 steps of 0.75 Angstrom) are required to complete a six dimensional docking
search. A further advantage of the spherical polar approach is that it is easy to constrain the
docking search to one or both binding sites, when this knowledge is available, simply by
constraining one or two of the angular degrees of freedom. This can reduce docking times to a
matter of minutes on modern workstations. So, depending on how well a particular FFT
algorithm is implemented (and on who you believe!), I claim that Hex is somewhere between 10
and 100 times faster than conventional FFT docking algorithms.

LIGBUILDER:
Ligbuilder is a general-purposed program package written for structure-based drug design
procedure. Based on the 3D structure of the target protein, it can automatically build Ligand
molecules within the binding pocket. Manuscripts describing Ligbuilder in details are currently
under preparation. In brief, main features of Ligbuilder are:
The program analyzes the binding pocket of the target protein and derives the key
interaction sites. A pharmacophore model is suggested and it could be applied to 3D database
search for finding novel ligand molecules. Molecules are constructed by using fragments as
building blocks. Various kinds of structural manipulation are provided, such as growing, linking,
and mutation. On-the-fly minimization of conformation is performed during the building-up
procedure. While the target protein is kept rigid, flexibility of the ligand molecules is considered.
User can choose either growing strategy or linking strategy to develop ligand molecules.
Molecules are evolved by Genetic Algorithm. The fitness score of a molecule is evaluated by
considering its chemical viability as well as binding affinity. Chemical rules are adopted for
evaluating "drug-likeness" of the resultant molecules. Chemical stability, synthesis feasibility,
and toxicity can also be taken into account by defining "forbidden structure" libraries.
All the input and output molecules are in popular format, i.e. protein in PDB format and
ligand in Sybyl Mol2 format. Ligbuilder is written in ANSI C++ and has been tested on UNIX

Raghvendra Sachan 29
and LINUX platforms. Running Ligbuilder itself does not need any other software. However,
you may need graphical molecular modeling software to prepare the input files and check the
resultant molecules.

CHEMSKETCH:

ChemSketch is a free computer program that can be downloaded from the Internet by
teachers or pupils and used to construct chemical equations, molecular structures and laboratory
diagrams with comparative ease. The program incorporates many advanced features such as the
ability to view coloured, rotating molecules in space-filling mode. Ions, functional groups,
complete molecules and apparatus can be selected from a large menu of templates and re-sized
or moved to any point on a page. The package includes text and drawing tools. Here the author
shows how to construct some sample molecules and mathematical equations using ChemSketch.
Computer programs for use in chemistry can be expensive, time-consuming to learn,
exasperating to use, and sometimes lacking in the combined text, mathematical and drawing
tools needed to produce quality hand-outs. Even the type of computer available can limit access
to good programs.

MOLINSPIRATION:

Molinspiration miscreen toolkit allows fast virtual screening of large collections of

molecules. The screening is based on identification of substructure features / fragments typical
for the training set of active molecules. No information about the 3D structure of receptor is
necessary, SMILES codes of active molecules are sufficient for the training, therefore the
procedure may be applied also in the very early project stage. The fragment-based virtual
screening is very fast (ca 100'000 molecules may be screened in less than an hour) and therefore
allows processing of very large molecular libraries. Validation tests performed at Molinspiration,
as well as by our customers on various target classes (such as kinase inhibitors, GPCR ligand or
pesticides) show 10 to 20-fold increase in hit rate in comparison with standard / random selection
of molecules for screening. Another advantage of Molinspiration fragment-based screening
procedure is that it is able to identify also novel active scaffolds which are not present in the
training set.

Raghvendra Sachan 30
OSIRIS PROPERTY EXPLORER:

The OSIRIS Property Explorer is an integral part of Actelion's inhouse substance

registration system. It helps to draw chemical structures and calculates on-the-fly various drug-
relevant properties whenever a structure is valid. Prediction results are valued and color coded.

 logP value- The logP value of a compound, which is the logarithm of its partition
coefficient between n-octanol and water log(Coctanol/Cwater), is a well established measure of
the compound's hydrophilicity. Low hydrophilicities and therefore high logP values cause
poor absorption or permeation. It has been shown for compounds to have a reasonable
probability of being well absorbed their logP value must not be greater than 5.0

 Toxicity risk Assessment- It gives the tumerogenic nature, reproductive effect,

mutagenecity and irritant nature of the drug.

 Molecular weight- Optimizing compounds for high activity on a biological target almost
often go along with increased molecular weights. However, compounds with higher
weights are less likely to be absorbed and therefore to ever reach the place of action.

 Drug likeness- The drug likeness is calculated with the following equation summing up
score values of those fragments that are present in the molecule under investigation:

 Drug score- The drug score combines drug likeness, clogP, logS, molecular weight
and toxicity risks in one handy value than may be used to judge the compound's overall
potential to qualify for a drug. This value is calculated by multiplying contributions of the
individual properties with the first equation, where ds is the drug score. si are the
contributions calculated directly from of cLogP, logS, molweight and druglikeness.
Parameters a and b are (1, -5), (1, 5), (0.012, -6) and (1, 0) for cLogP, logS, molweight and

Raghvendra Sachan 31
 druglikeness, respectively. These are the contributions taken from the 4 toxicity risk types.

QUANTUM DRUG HIT IDENTIFICATION TOOL:

Drug Hit Identification Tool calculates the IC50 (Kd, Ki, pKd) value of any protein-
ligand complex, docks a small molecule in the active site of a protein and performs screening a
library of compounds against a target-protein or DNA/RNA. Hit Identification Tool consists of
three modules: The IC50 module, 2) Ligand Docking and 3) Library Screening modules. Drug
Hit Identification Tool can run both in Windows and Linux environments.
In QUANTUM binding affinity of a protein-ligand complex is estimated on the basis of
free binding energy calculations. A successful free binding energy calculation consists of three
key parts: 1) a vacuum force field for internal molecular energy calculations (conformational
changes), 2) a model of intermolecular interactions in solution (water model), and 3) a fast and
accurate machinery for thermodynamic calculations (entropy losses evaluation).
Hit Identification Tool requires the 3D structure of a given macromolecule (protein or
DNA/RNA) with the resolution better than 2 Angstrom. The ligand structures can be provided in
any of commonly used file formats such as SMILES, .sdf, .mol2, .hin, .pdb or drawn by using
QUANTUM’s molecule drawing tool.
IC50 module:
Calculates the free binding energy and thus predicts the IC50 of a given protein-ligand
complex.
Ligand docking module:
QUANTUM docking algorithm finds the best positions of a given small molecule in the
active site of a protein with the minimum value of the free binding energy.
Library screening module:
Library screening module performs docking of a library of compounds in the active site
of a given protein or DNA/RNA.

Raghvendra Sachan 32
 METHODOLOGY

Neuraminidase protein was used as target for our disease.

The first and foremost step is searching for the protein through EXPASY, NCBI that provide us
with the basic properties and the Genbank id for the protein along with the FASTA format.

In NCBI the BLAST option is clicked and protein-protein blast is selected. The FASTA format is
selected and clicks the BLAST option.

The FASTA format of protein aligned with tool protein-protein BLAST. The result allows the
similarity in scores. The highest score result was selected as template sequence for the modeling
of Protein.

The program modeller is then used to make the best five models of the query protein by
preparing the Atom, Alignment and Python files.

All the five models were then uploaded to SAVS Server for PROCHECK of the protein models.
The results from the PROCHECK shows the disallowed regions according to the
RAMACHANDRAN Plot and the bad contacts of the model.

Deep View Swiss PDB viewer was then used to rectify the bad contacts by Energy Minimization
after removing the disallowed regions.

The Ligsite, an online tool, is used to predict Active Sites of the protein.

The seed (lead) molecule or the ligand molecule is sketched by using ChemSketch.

Raghvendra Sachan 33
After calculating the nearest residue to the active site by spdbv, the seed molecule is then
directed for the Docking via HEX.

The program HEX directs the path for seed molecule to Dock for Blocking of the protein.

Ligbuilder is then used for generating the suitable Ligand Population for Blocking the Protein
Active Site directed by the HEX Docking.

Ligbuilder provides the best 10 ligand molecules.

The Ligand molecule from the Ligbuilder is used for Quantum or AUTODOCK as input Ligand
molecule file for final Docking.

Quantum performs the final Docking according to the Grid input. After performing the Docking
for each 10 Ligand molecules, the result is displayed in Docking Log File (.dlg).

The least Docking Energy molecule among 10 *10 was then selected as the final Drug molecule.

The final Drug molecule was then tested among various Online Validation Server (Molsoft,
Molinspiration, and OSIRIS) for Drug Likeness.

Raghvendra Sachan 34
 PROCEDURE OF DRUG DESIGNING

Homology modeling:

Prediction of the 3-dimentional structure of the target protein is an important step in

insilico drug designing as it helps modeling a drug molecule which is specific for that disease. A
large number of servers are available which provide protein structures but they may not have
novel protein structures. One method that can be applied to generate reasonable models of
protein structures is homology modeling. This procedure, also termed comparative modeling or
knowledge-based modeling, develops a three-dimensional model from a protein sequence based
on the structures of homologous proteins.
Homology has a precise definition: having a common evolutionary origin. Thus,
homology is a qualitative description of the nature of the relationship between two or more
things, and it cannot be partial. Either there is an evolutionary relationship or there is not. An
assertion of homology usually must remain a hypothesis. Supporting data for a homologous
relationship may include sequence or three-dimensional similarities, the relationships between
which can be described in quantitative terms.
The steps to creating a homology model are as follows:
 Identify homologous proteins and determine the extent of their sequence similarity with
one another and the unknown
 Align the sequences
 Identify structurally conserved and structurally variable regions
 Generate coordinates for core (structurally conserved) residues of the unknown structure
from those of the known structure(s)
 Generate conformations for the loops (structurally variable) in the unknown structure
 Build the side-chain conformations
 Refine and evaluate the unknown structure.
Though the structure of protective antigen had already been reported in some of the
structural databases, the structure of protective antigen was exclusively modeled through
homology modeling. Several computerized search methods are available to assist in identifying

Raghvendra Sachan 35
homologues. In most cases of homology modeling, amino acid sequence of the protein is
available.
Steps Involved:
1. Select the target protein:
Neuraminidase 1 of Influenza A (H5N1):
Neuraminidase is a glycoside hydrolase enzyme. Neuraminidase has functions that aid
in the efficiency of virus release from cells. Neuraminidase cleaves terminal sialic acid residues
from carbohydrate moieties on the surfaces of infected cells. This promotes the release of
progeny viruses from infected cells. Neuraminidase also cleaves sialic acid residues from viral
proteins, preventing aggregation of viruses.
Administration of chemical inhibitors of neuraminidase is a treatment that limits the
severity and spread of viral infections.

2. Find the sequence of target protein in FASTA format from NCBI:

NCBI i.e. National center for Biotechnology Information was used to find out the
sequence of Neuraminidase 1(N1) protein. FASTA format sequence of the protein was taken.

FASTA format of target protein N1:

>gi|28849545|gb|AAO52955.1|AF509112_1 neuraminidase [Influenza A virus

(A/Chicken/Hong Kong/893.2/01 (H5N1))]
MNPNQKIITIGSICMVIGIVSLMLQIGNIISIWVSHSIQTGNQHQAEPIGNTNFLTEKAVAS
VTLAGNSS
LCPISGWAVHSKDNGIRIGSKGDVFVIREPFISCSHLECRTFFLTQGALLNDKHSNGTVKD
RSPHRTLMS
CPVGEAPSPYNSRFESVAWSASACHDGTSWLTIGISGPDNGAVAVLKYNGIITDTIKSWR
NNILRTQESE
CACVNGSCFTVMTDGPSNGQASYKIFKMEKGKVVKSVELDAPNYHYEECSCYPDAGEIT
CVCRDNWHGSN
RPWVSFNQNLEYQIGYICSGVFGDNPRPNDGTGSCGPVSPNGAYGVKGFSFKYGNGVW
IGRTKSTNSR

3. Perform blastp to get sequence similar to the Neuraminidase 1:

The FASTA format of the protein was noted and was given as input for NCBI protein-
protein BLAST (blast-p). The results showed a number of proteins which are similar in their

Raghvendra Sachan 36
nucleotide sequence with the target protein. The extent of similarity was judged on the basis of
certain parameters like E value, Identities and Positives. Taking this into account the protein
with the best similarity was selected from the BLAST results. The PDB ID of the protein
(2HTY) was also noted.

Raghvendra Sachan 37
BLAST Result:

>pdb|2HTY|A Chain A, N1 Neuraminidase

pdb|2HTY|B Chain B, N1 Neuraminidase
pdb|2HTY|C Chain C, N1 Neuraminidase
pdb|2HTY|D Chain D, N1 Neuraminidase
pdb|2HTY|E Chain E, N1 Neuraminidase
pdb|2HTY|F Chain F, N1 Neuraminidase
pdb|2HTY|G Chain G, N1 Neuraminidase
pdb|2HTY|H Chain H, N1 Neuraminidase

Score = 571 bits (1472), Expect = 1e-163, Method: Composition-based stats.

Identities = 280/286 (97%), Positives = 283/286 (98%), Gaps = 0/286 (0%)

Query 63 VTLAGNSSLCPISGWAVHSKDNGIRIGSKGDVFVIREPFISCSHLECRTFFLTQGALLND 122

V LAGNSSLCPI +GWAV+SKDN IRIGSKGDVFVIREPFISCSHLECRTFFLTQGALLND
Sbjct 1 VKLAGNSSLCPINGWAVYSKDNSIRIGSKGDVFVIREPFISCSHLECRTFFLTQGALLND 60

Query 123 KHSNGTVKDRSPHRTLMSCPVGEAPSPYNSRFESVAWSASACHDGTSWLTIGISGPDNGA 182

KHSNGTVKDRSPHRTLMSCPVGEAPSPYNSRFESVAWSASACHDGTSWLTIGISGPDNGA
Sbjct 61 KHSNGTVKDRSPHRTLMSCPVGEAPSPYNSRFESVAWSASACHDGTSWLTIGISGPDNGA 120

Query 183 VAVLKYNGIITDTIKSWRNNILRTQESECACVNGSCFTVMTDGPSNGQASYKIFKMEKGK 242

VAVLKYNGIITDTIKSWRNNILRTQESECACVNGSCFTVMTDGPSNGQASYKIFKMEKGK
Sbjct 123 VAVLKYNGIITDTIKSWRNNILRTQESECACVNGSCFTVMTDGPSNGQASYKIFKMEKGK 180

Query 243 VVKSVELDAPNYHYEECSCYPDAGEITCVCRDNWHGSNRPWVSFNQNLEYQIGYICSGVF 302

VVKSVELDAPNYHYEECSCYP +AGEITCVCRDNWHGSNRPWVSFNQNLEYQIGYICSGVF
Sbjct 181 VVKSVELDAPNYHYEECSCYPNAGEITCVCRDNWHGSNRPWVSFNQNLEYQIGYICSGVF 240

Query 303 GDNPRPNDGTGSCGPVSPNGAYGVKGFSFKYGNGVWIGRTKSTNSR 348

GDNPRPNDGTGSCGPVS NGAYGVKGFSFKYGNGVWIGRTKSTNSR
Sbjct 241 GDNPRPNDGTGSCGPVSSNGAYGVKGFSFKYGNGVWIGRTKSTNSR 286

4. Design protein structure through homology modelling using

Modeller8v1:

Raghvendra Sachan 38
 First download pdb file of similar protein from pdb bank. To find structure of target
protein we use Modeller. It is used for performing comparative or 3d modelling of protein
structure.
In Modeller we take 3 files as an input.
 Atom file: saved with extension (.atm) as filename.atm is, obtained from pdb.
 Alignment file: Modeller uses a special form of PIR format where information about
sequence numbering and chain codes is written into the description. Line between the PIR
protein tag and protein alignment entry.
 Script file or Python file: This is an ordinary python 2.3 scripts to specify the alignment
and the atom file for modeler.
The Modeller generates models using input provided. The best model amongst these is
selected for further process.

5. Analysis and validation of the models:

The structures submitted in the SAVS are validated. After validation model with the
maximum core value and zero bad contacts are used for the further process at lead target
prediction.
The Ramachandran plot is also verified for amino lying outside the allowed region. After
the final analysis and validation process, the best protein model of the target protein is used for
active site prediction.
To move residue from disallowed region to allowed region we perform loop formation
using Ramachandran Plot. We do energy minimization to remove bad contacts.
The structures submitted in STRUCTURE ANALYSIS AND VALIDATION SERVER
(Procheck).

Raghvendra Sachan 39
Procheck Result:

Structure of Modeled Target Protein:

Ribbon Structure

Raghvendra Sachan 40
 6. Then we use an online tool Ligsite to find the pocket of modeled
protein:
LIGSITE is used for prediction and visualization of the protein binding pockets. In it
pdb format of final protein structure is used as input file. The outputs obtained are following:
• PDB file with active site proteins
• PDB file of original molecule

7. Selection of Lead molecule:

Lead molecule is selected for generation of drug.
Criteria to choose lead molecule:-
• Size should be small.
• No metal atom.
• Maximum no. of possible growing sites.
• Should not contain any kind of unsaturation i.e. Double bond.

8. Designing of lead molecule:

ChemSketch is used to design drug molecule on computer to define its structure.

9. Rigid Docking:
Hex 4.5 is used for the purpose of docking of the lead with the target molecule. The
protein and ligand are input in the Hex and ligand is docked to the residue on the minimum
distance position to the active site position. After docking is completed the position is saved in
Pdb format for the subsequent steps.

Raghvendra Sachan 41
10. Conversions of pdb file to mol2 file:
Molecular Structure File Convertor which is an online tool is used to convert docked lead
molecule from .pdb format to .mol2 file. The SybylMOL2 format of the lead molecule used as
input to develop through Ligbuilder.

11. Generation and Optimization of Ligand:

Ligbuilder is a tool used for generation of the ligands. Ligbuilder builds up ligand step
by step using a library of organic fragments. In it basically 3 processes are performed:
 Pocketing:
In it docked complex in pdb format and lead molecule in .mol2 format is given as input.
This generates 5 files namely:
• keysite.txt
• pharmacophore.txt
• pharmacophore.pdb
• pocket.txt
• grid.txt
 Growing:
The result from pocket generation i.e. pocket and grid and .mol2 format of ligand are
given as input. It generates populations and ligands file in .lig format. The molecules collected
in the ligands.lig files, obey the Lipinski’s rule of five.
• Final ligand.lig
• Final population.lig
Rule of 5" is set of simple molecular descriptors used by Lipinski in formulating his
"Rule of 5”. The rule states, that most "drug-like" molecules have logP <= 5, molecular weight
<= 500, number of hydrogen bond acceptors <= 10, and number of hydrogen bond donors <= 5.
Molecules violating more than one of these rules may have problems with bioavailability. The
rule is called "Rule of 5", because the border values are 5, 500, 2*5, and 5.
 Processing:
In it input file is ligands.lig and it generates 10 structures of ligand in .mol2 format.
These are then converted into .pdb format.
• Results.mdb

Raghvendra Sachan 42
 Ligbuilder Results:

Result 9 Result 10

Raghvendra Sachan 43
12. Selection of Ligand molecule:
All the 10 structures are drawn in 3 separate software namely ChemSketch,
Molinspiration, and Osiris property explore.
ChemSketch: It is mainly used for property calculation. In it parachore value is major
determining factor.

Raghvendra Sachan 44
Molinspiration: It gives us other information clog p value, number of rotatable bonds and
verifies Lipinski’s rule of five etc.

Osiris Property Explorer: It gives us drug score which is major deciding criteria for selection
of lead molecule.

Raghvendra Sachan 45
 Osiris result:

13. Flexible docking:

First of we select the molecule with highest drug score and parachore value near 440.
Then docking is performed using software Quantum Dock. It is used to predict the binding
pattern of small molecules to the receptor of known 3D structure.
In it, protein molecule file and ligand molecule file in .pdb format are given as input. In
the output we get processed ligand molecule file, protein molecule file, protein molecule with
grids in .pdb format. Also it gives us a number of other parameters like binding energy, torsion
energy, Vander walls energy number of rotatable bonds, IC50 value etc.

Raghvendra Sachan 46
DOCKING RESULT:

Results of docking are mentioned below:

Name of protein : n1_1hb
Ligand : lig_lead
IC50 value (mol/lt) : 1.81e-0.03
Ggrid (KJ/mol) : -15.96
Ees (KJ/mol) : -10.91
Evdw (KJ/mol) : -17.91
TdS (KJ/mol) : -9.28
Etor (KJ/mol) : 3.58
Gprot : 0
Charge : 0
Flexible bond : 0
RMS, A : 32.81

Raghvendra Sachan 47
Protein with docked inhibitor molecule:

Raghvendra Sachan 48
Neuraminidase 1 with inhibitor molecule attached at active site:

Raghvendra Sachan 49
INHIBITOR MOLECULE:

Raghvendra Sachan 50
 CONCLUSION
Avian flu is a disease concerning respiratory disorders. Today it poses pandemic threat.
So detail analysis regarding its pathogenicity is required.
Out of several possible therapies of bird flu, the best suitable is inhibiting neuraminidase.
The protein is modeled and refined in all the best possible ways. With help of Ligbuilder 10
ligands are generated out of which best ligand is taken. Flexible docking is performed against
ligand and receptor. The docked complex with minimum energy -15.96 KJ/mol is taken. This is
most suitable ligand that can be used for inhibition of neuraminidase and hence for prevention of
bird flu.
From the results obtained from Molinspiration and OSIRIS property explorer, the
drug likeness, the effectiveness of the drug molecule was estimated. The drug molecule
developed for treating avian flu (H5N1) has the following characteristics:
 The drug molecule developed has a low logP value 1.356 indicating that it is easily
absorbed.
 With a molecular weight 232 of this drug molecule is highly soluble.
 The drug molecule also has no reproductive effectiveness property and is neither tumor
causing in nature.
 Drug is neither irritant nor mutagenic.
 Taking various parameters like logP value, logS value, drug likeness and solubility,
the drug score (0.63) predicted was very good score.

This drug should be further tested for stability in invitro conditions, further it should be
tested for on probable host top find out its efficiency. It’s hoped that the drug can be proved to be
a good weapon in fight against avian flu.

Raghvendra Sachan 51
References:

Websites:

∗ www.ncbi.nlm.nih.gov/

∗ www.ncbi.nlm.nih.gov/blast/

∗ www.rcsb.org/pdb/files/2hty.pdb

∗ www.scoppi.biotec.tu_dresden.de/pocket/

∗ www.molinspiration.com/cgi-bin/properties

∗ www.en.wikipedia.org/wiki/Avian_flu

∗ www.rkm.com.au

∗ www.kubinyi.de

∗ www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1397843&rendertype=abstract

∗ www.portfolio.mvm.ed.ac.uk/studentwebs/session2/group21/virology.htm

∗ www.cdc.gov/

∗ www.content.nejm.org/

∗ www.protein.gsc.riken.go.jp/Research/insilico/index.html

Books:

∗ BIOINFORMATICS- A Practical Guide to the Analysis of Genes and Proteins

-Wiley & Sons Publication

Raghvendra Sachan 52