Design, synthesis and biological evaluation of inhibitors of flavivirus NS2B NS3 protease

2.2.2 Expression and purification of fusion protein 302.4.3 Release of the target peptide by enterokinase cleavage 39 2.4.5 Mass spectrometric analysis of the target peptide 41 Chapter 3

Design, Synthesis and Biological Evaluation of Inhibitors of

Flavivirus NS2B/NS3 Protease

Gao Yaojun (B.Sc., Soochow University)

A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF CHEMISTRY NATIONAL UNINVERSITY OF SINGAPORE

2009

ACKNOWLEDGEMENTS

First and foremost I offer my sincerest gratitude to my supervisor, Associate Professor Lam Yulin, for her invaluable support, encouragement, supervision and useful suggestions throughout my Ph.D Her moral support and continuous guidance enabled

me to complete my work successfully I am also highly thankful to my co-supervisor,

Dr Cui Taian, Senior Lecturer in Singapore polytechnic for his encouragement and effort and without him this thesis, too, would not have been completed

I gratefully acknowledge the laboratory officers in CMMAC, Dept of Chemistry, Miss Tan Geok Kheng, Mdm Han Yanhui, Mdm Wong Lai Kwan and Mdm Lai Hui Ngee for their assistance and technical support, and all others who have helped in one way or another

I deeply appreciate my group members, Fu Han, Kong Kah Hoe, He Rongjun, Gao Yongnian, Che Jun, Ching Shi Min, Fang Zhanxiong, Wong Ling Kai, William Lin Xijie and Sanjay Samanta, for all their help and encouragement during my research Furthermore, I would like to thank the staff at SP, Miss Ang Cuixia, Dr Puah Chum Mok, Dr Chen Gang and Dr Liew Oi Wah, thank you for all the support given

I am as ever, especially indebted to my parents and my sister for their love and support throughout my life Finally, I thank National University of Singapore for awarding me a research scholarship to pursue my doctorate degree

TABLE OF CONTENTS

SUMMARY v

Chapter 2: Biosynthesis of an Acyclic Permutant of Kalata B1 from a Recombinant Fusion Protein with Thioredoxin

2.2.2 Expression and purification of fusion protein 30

2.4.3 Release of the target peptide by enterokinase cleavage 39

2.4.5 Mass spectrometric analysis of the target peptide 41

Chapter 3: Design and disulfide bond connectivity-activity studies of a kalata B1-inspired cyclopeptide against dengue NS2B/NS3 protease

3.2.1 Design and oxidative refolding of cyclopeptide 1 483.2.2 Determination of disulfide bond connectivity in cyclopeptide

1

50

3.3 Conclusion 56

3.4.2 Oxidative refolding of cyclopeptide 1 and purification of the

3.4.4 Inhibitory activity assay against DEN2 NS2b-NS3 protease 60

Chapter 4: Synthesis and biological evaluation of small molecule inhibitors of West Nile Virus NS2B/NS3 Protease

Chapter 5: Synthesis of Pyrazolo[5,1-d][1,2,3,5]tetrazine-4(3H)- ones

SUMMARY

This thesis is divided into two parts The first part which is the main focus of the thesis involves the design, synthesis and biological evaluation of inhibitors of Dengue and West Nile virus NS2B-NS3 protease

For the design of dengue NS2B-NS3 protease inhibitors, we were inspired by the unique structural and diverse biological activities found in cyclotides to design cyclopeptides as inhibitors of dengue virus protease Firstly, we designed a new approach to obtain some acyclic cyclotides based on the bacterial expression of a thioredoxin-ac kalata B1 fusion protein and subsequent liberation of ac kalata B1 by enterokinase cleavage of the precursor Secondly, using the new approach and chemical synthetic method, we prepared various kalata B1 analogues by varying its amino acid sequence and found the two fully oxidized forms of a cyclopeptide showed potent inhibition with Ki value of 1.39 ± 0.35 and 3.03 ± 0.75 μM, respectively To our best knowledge, these were among the most potent peptide inhibitors achieved for the dengue viral protease

For the design of West Nile virus NS2B-NS3 protease inhibitors, initially, a library of more than 100 compounds was screened for WNV NS3 protease inhibition assays by high throughput screening (HTS) Through HTS, we found several “hits” that inhibited the WNV NS2B/NS3 protease Among these “hits” compounds, a compound showed the best inhibition and was chosen for structure activity

relationship (SAR) exploration on WNV NS3 protease inhibition assays In the studies, a potent, stable molecule with Ki value of 1.82±0.58 μM was identified to be

an uncompetitive inhibitor To our knowledge, this is the most potent compound amongst the stable small molecule inhibitors of WNV NS2B-NS3 protease reported

so for

The second part of this thesis involves the methodology development of the

solid-phase synthesis of pyrazolo[5,1-d][1,2,3,5]tetrazine-4(3H)-ones In the

methodology, a one-pot reaction from 5-aminopyrazoles to the

pyrazolo[5,1-d][1,2,3,5] tetrazine-4(3H)-ones which provided the compounds in good

yields was demonstrated A representative set of 16

pyrazolo[5,1-d][1,2,3,5]tetrazine-4(3H)-ones was prepared

LIST OF TABLES

Table 1.1 Characteristics and functions of flavivirus proteins 7

Table 1.2 Summary of peptidic inhibitors of NS2B/NS3pro 22

Table 3.1 Peptides designed as potential inhibitors to DEN2 NS2B/NS3

Table 3.3 Inhibition of dengue NS2B-NS3 protease by isomers 1B and 1C 56

Table 4.1 Optimization of cyclization reaction to prepare compound 4-6 67

Table 4.2 WNV NS3 protease inhibitor analogues and their inhibition results 73

Table 4.3 The inhibition results of enantiomers from selected racemic

compounds

74

Table 5.1 Nitrile analogs and their reaction times 110

LIST OF FIGURES

Figure 1.2 Schematic representation of flavivirus genome organization and

polyprotein processing

5

Figure 1.3 Nomenclature for peptide residues (P3-P3’) and their

corresponding binding sites (S3-S3’) in the enzyme

14

Figure 1.4 Crystal structures of WNV NS2B/NS3pro and predicted

substrate and membrane interactions

18

Figure 1.5 Non-peptidic inhibitors of DEN and WNV NS2B/NS3 proteases

and their inhibitory potencies

25

Figure 2.2 Expression and enterokinase-catalyzed cleavage of recombinant

thioredoxin- ackalata B1 fusion protein studied by SDS-PAGE

31

Figure 2.3 Time-dependent and hydrogen peroxide-dependent cleavage of

recombinant thioredoxin-ac kalata B1 fusion protein by enterokinase

34

Figure 2.4 HPLC chromatogram and MS spectra of ac-kalata B1 35

Figure 3.1 Structure of Ciluprevir, a cyclic peptide inhibitor of HCV

NS3pro

44

Figure 3.2 Schematic representation of a cyclotide (kalata B1) structure 46

Figure 3.5 Inhibition of WNV by protease inhibitors 56

Figure 4.1 Structures and IC50 values of WNV NS2B-NS3pro inhibitors

confirmed in the HTS

65

Figure 4.2 Uncompetitive mechanism of inhibition of WNV NS2B-NS3pro

by the (-) enantiomer of compound 4-6o

76

Figure 5.1 Library of synthesized pyrazolo[5,1-d][1,2,3,5]tetrazine

-4(3H)-ones 5-1a-p

105

LIST OF SCHEME

Scheme 1.1 Chemical form of aldehyde inhibitor in water 19

Scheme 3.1 Schematic representation of strategy 1 used for disulfide bond

Scheme 5.2 SPS of Pyrazolo[5,1-d][1,2,3,5]tetrazine-4(3H)-ones 103

CDAP 1-Cyano-4-dimethyl-aminopyridinium tetrafluoroborate

CDC Centers for Disease Control and Prevention

DBU 1,8-Diazabicycloundec-7-ene

DMA Dimethylacetamide

DMF Dimethylformamide

DTT Dithiothreitol

E coli Escherichia coli

HCV Hepatitis C virus

IC50 Half maximal inhibitory concentration

IPTG Isopropyl-β-D-1- thiogalactopyranoside

MALDI TOF Matrix Assisted Laser Desorption /Ionization- Time Of Flight

NEM N-Ethylmaleimide

TLC Thin layer chromatography

Wang resin 4-Hydroxymethylphenoxy resin

protease Bioorg Med Chem 2010, 18, 1331–1336

3 Yaojun Gao, Yulin Lam Synthesis of Pyrazolo[5,1-d][1,2,3,5] tetrazine-4(3H)-

ones J Comb Chem 2010, 12, 69–74

4 Taian Cui, Yaojun Gao, et al Hydrogen peroxide enhances enterokinase-catalysed

proteolytic cleavage of fusion protein Recent Pat Biotechnol 2008, 2, 189-90

5 Taian Cui, Yaojun Gao, et al Efficient preparation of an acyclic permutant of

kalata B1 from a recombinant fusion protein with thioredoxin J Biotechnol 2007,

130, 378-384

PATENT

1 Taian Cui, Chum Mok Paul, Yulin Lam, Yaojun Gao Conpounds for use as anti-virus agents Patent No 10757sg28, Filling date: 17 Dec 2008

Chapter 1 Introduction

1.1 Flavivirus

Flaviviruses (Latin flavus meaning yellow, because of the jaundice induced by yellow fever virus) are a major cause of infectious disease in humans The genus Flavivirus contains more than 70 members1,2, including yellow fever virus (YFV), dengue virus (DEN), West Nile virus (WNV), Japanese encephalitis virus (JEV), and tick-borne encephalitis virus (TBE), etc This number is increasing as more viruses are discovered and will undoubtedly continue to increase for some time yet Some flaviviruses such as YFV, JEV, and TBE were first recognized because they caused major human epidemics involving high fatality rates Others, such as dengue virus, cause about 50–100 million cases of dengue fever occurring in the tropical and sub-tropical regions of the world and the infected cases exhibit a broad spectrum of clinical symptoms ranging from being fully asymptomatic to causing life-threatening conditions like dengue hemorrhagic fever (DHF) or dengue shock syndrome (DSS) West Nile virus, another example, historically occurs in Africa, Europe, the Middle East, Central Asia, and West According to the report published by the Centers for Disease Control and Prevention (CDC), from an outbreak occurred in the USA in

1999 to 2008, with largest ever WNV outbreaks occurring in 2002 and 2003, a total of 29,000 around cases of WNV infection, more than 1100 of whom died, had confirmed infections with the virus (see: http://www.cdc.gov/ncidod/dvbid/westnile/index.htm) Although licensed vaccines3 are available for YFV, JEV and TBE, none have been

developed for other flaviviral diseases Efforts for vaccine development for dengue have been a continuous challenge for decades, the main issue being the inability of vaccines to protect simultaneously against all four antigenically distinct serotypes A further barrier to vaccine development is the sporadic nature of infections caused by agents such as WNV, JEV and TBE, which could only be completely prevented by carrying out universal immunization across huge geographic regions In the absence

of vaccines, drugs for specific therapy are needed, but no antiviral medications have been approved for use against the flaviviruses

1.2 Flavivirus virion and viral life cycle

The mature flavivirus virions are smooth and spherical, with a diameter of 500 Å4

(Figure 1.1) They possess an icosahedral nucleocapsid (NC) of approximately 30 nm

consisting of single-stranded positive-sense RNA genome and several copies of a small, basic capsid protein The nucleocapsid is surrounded by a lipid envelope in which two envelope proteins are embedded: the envelope protein (E) and the membraneprotein (M) or its precursor M5 Intracellular virions, which only contain prM, as well as released extracellular virons, which predominantly contain M protein, have been described In comparison with intracellular particles, extracellular particles have a greater infectivity than those remaining intracellular For West Nile virus, the specific infectivity of intracellular virus was demonstrated to be 60-fold lower than the one of extracellular particles6 Recently, the structure of immature and mature,

pr-M containing YFV7 and dengue particles7,8 was determined to 25 Å and 24 Å resolutions, respectively, by cryoelectron microscopy and image reconstruction techniques The structure suggests that flaviviruses employ a fusion mechanism in which the distal β barrels of domain II of the glycoprotein E are inserted into the cellular membrane8

Figure 1.1 Image of the flavivirus virion

In the flavivirus replication cycle4,9, virions bind to cell-surface attachment molecules and receptors and are internalized through endocytosis In the low pH of the endosome, viral glycoproteins mediate fusion of the viral and cellular membranes, allowing disassembly of the virion and release of its RNA into the cytoplasm The viral RNA is translated into a polyprotein that is processed by viral and cellular proteases Genome replication occurs on intracellular membranes Virion assembly occurs on the surface of the endoplasmic reticulum (ER) membrane Capsid protein and viral RNA are enveloped by the ER membrane and its embedded glycoproteins to form immature virus particles, which are then transported through the secretory

pathway In the low pH of the trans-Golgi network (TGN), prM is cleaved by furin Mature virions are then released into the cytoplasm

1.3 Flavivirus genome structure and polyprotein processing

The genomic RNA of flavivirus is infectious It consists of a single stranded RNA of positive polarity with length of approximately 11kb10 The genome encodes one large polyprotein which is flanked by a short 5′ untranslated region (UTR) and 3′ UTR11

(Figure 1.2) which have secondary structures that are essential for the initiation of

translation and for replication12 The 5′ UTR is about 120 nucleotides in length and the 3′ UTR comprises about 500 nucleotides Similar to eukaryotic RNAs, the flaviviral genome contains a 5′ cap structure, but the 3′ end lacks a poly-A tail Translation of the genome by the host cell machinery produces a polyprotein comprising the viral structural and non-structural proteins that are required for replication and assembly of new virions

The large polyprotein is cleaved co- and posttranslationally by host and viral

proteases to release the single viral protease (Figure 1.2 and Table 1.1) The

structural proteins are encoded from the 5′-terminal quarter of the genome whereas the remaining two-thirds encode the nostructural proteins The order of the proteins within the polyprotein is NH2-C-prM(M)-E-NS1-NS2A-NS2B-NS3-NS4A-NS4B- NS5-COOH In addition, two small hydrophobic proteins are released from the

polyprotein, one is derived from the C-terminus of the anchored capsid protein after cleavage the mature capsid is released13 The second one represents a small fragment between the NS4A and the NS4B protein and is called 2K based on its predicted size14

Structural Nonstructural

Flavivrus RNA genome and polyprotein

Postive-strand RNA genome

M pr

? Unknown protease

?

◆

Structural Nonstructural

Flavivrus RNA genome and polyprotein

Postive-strand RNA genome

M pr

? Unknown protease

?

◆

Figure 1.2 Schematic representation of flavivirus genome organization and

polyprotein processing Top, the flaviviral genome with the structural and nonstructural proteins coding region, the 5′ UTR with 5′ cap structure and the 3′ UTR with the potential 3′ secondary structure are shown Below, the mature flaviviral proteins generated by polyteolytic processing of the polyprotein are demonstrated Gray boxes represent the structural proteins (capsid (C), precursor membrane (prM) and envelope (E)), white boxes represent the nonstructural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, NS5) In addition, two small hydrophobic fragments are cleaved from the polyprotein (black bars) Cleavage sites for viral serine protease ( ), the host signalase (◆), furin(↓), or unknown protease ( ? ) are indicated

The structural proteins prM and E as well as the following nonstructural protein NS1 are translocated into the endoplasmic reticulum Cleavages that generate the N-termini of prM and E as well as the C-terminus of E are mediated by host cell

signal peptidase15 In contrast, the capsid protein remains in cytoplasm Processing to generate the C-terminus of the mature capsid proteins is performed by the viral NS2B-NS3 protease13 This cleavage is a prerequisite for efficient processing to generate the N-terminus of prM by the signal peptidase16 Therefore, mutations that abolish cleavage to produce the C-terminus of the mature capsid proteins also prevent the production of infectious particles16 Furthermore, mutations that enhance signalase cleavage to generate the N-Terminus of prM are lethal for virus production17

The protease responsible for processing at the NS1-2A site is assumed to localize in endoplasmic reticulum but has not been identified yet However, for dengue virus it is known that the eight last amino acids of NS1 are required for cleavage at the NS1-2A

by the NS2B-NS3 protease15 This viral protease is responsible for cleavage at the NS2A/2B, NS2B/3, NS3/4A, NS4A/2K, and NS4B/5 sites For YFV, additional cleavage site in the C-terminal region of NS2A has been described (NS2Aa site)20,21 Similar to dengue virus, for which a minor cleavage within NS3 has been described22 The viral serine protease cleavage sites usually consist of two basic amino acids followed by an amino acid with a short side chain In the case of DEN or WNV the

cleavage sites usually consist of RR↓G/S (Table 1.1)15,23 Little variation is observed for the NS2Aa site (QKT) and the NS4A-2K site (QRS) in YFV In contrast to the majority of cleavage events in the NS region, the N-terminus of NS4B is generated by

host cell signal peptidase, but prior cleavage at the NS2A/2K site is required14 Other than the case just mentioned, processing at other sites within the NS region does not take place in an obligate order

Table 1.1 Characteristics and functions of flavivirus proteins

Cleavage site at N-terminus, protease responsible for cleavage

Function/enzymatic activity

RR↓S, NS2B-NS3 Protease

Capsid protein, Interaction with genomic RNA prM/M 26 kD/8 kD PrM: TGG↓V, Signalase

M: ARR↓A, Furin

Membrane protein

Envelope protein, Hemagglutination activity, Mediates binding to cell surface

Role in neurovirulence

particles NS2B 14 kD RR↓S, NS2B-NS3 Protease Cofactor of viral serine protease

1.4 Features of the structural and non-structural proteins

The capsid (C) protein

The C protein which contains a conserved hydrophobic domain is a highly basic

protein of ≈11 kD The hydrophobic domain is cleaved from mature C by the viral serine protease24 Owing to the basic character, C protein binds strongly to RNA, together with the viral RNA, several copies of the C protein form the nucleocapsid (NC) Analysis of purified C protein expressed in Escherichia coli revealed that it is largely alpha-helicl and forms dimers25 It is not yet clear how C protein dimers are organized within nucleocapsids, but interaction with RNA can induce isolated C protein dimers to assemble into nucleocapsidlike particles26.

The membrane (prM/M) protein

The glycoprotein precursor of the mature M protein (≈8 kD), prM (≈26 kD), is translocated into the endoplasmic reticulum (ER) by the C-terminal hydrophobic domain of C Signal peptidase cleavage is delayed, however, until the viral serine protease cleaves upstream of the signal sequence to generate the mature form of C protein13,24,27 The N-terminal region of prM contains one to three N-linked glycosylation sites11 and six conserved cysteine residues, all of which are disulfide linked28 The prM protein folds rapidly and assists in the proper folding of E protein29,30 A major function of prM is to prevent E from undergoing acid-catalyzed rearrangement to the fusogenic form during transit through the secretory pathway31,32 The conversion of immature virus particles to mature virions occurs in the secretory pathway and coincides with cleavage of prM into pr and M fragments by the Golgi-resident protease furin or a related enzyme33

The envelope (E) protein

The E protein (≈53 kD), the major protein on the surface of flavivirus virions, mediates receptor binding and membrane fusion E is synthesized as a type I membrane protein containing 12 conserved cysteines that form disulfide bonds34, and, for some viruses, E is N-glycosylated35,36 As mentioned, proper folding, stabilization

in low pH, and secretion of E depends on coexpression with prM29,30 The native form

of E folds into an elongated structure rich in β-sheets and forming head-to-tail homodimers that lie parallel with the virus envelope4,37 Each E protein subunit is composed of three domains: I, which forms a β-barrel; II, which projects along the virus surface between the transmembrane regions of the homodimer subunits; and III, which maintains an immunoglobulin-like fold

The NS1 protein

The NS1 protein (≈46 kD) is translocated into the ER during synthesis and cleaved from E protein by host signal peptidase, whereas an unknown ER-resident host enzyme cleaves the NS1/2A junction.38,39 NS1 is largely retained within infected cells but can localize to the cell surface and is slowly secreted from mammalian cells40 NS1 contains two or three N-linked glycosylation sites and 12 conserved cysteines that form disulfide bonds41-43 NS1 has an important unclear role in RNA replication

It localizes to sites of RNA replication44,45, and mutation of the N-linked glycosylation sites in NS1 can lead to dramatic defects in RNA replication46 and virus

production46,47 Furthermore, the function of the extracellular forms of NS1 is not yet

clear

The NS2A protein

NS2A is a relatively small hydrophobic protein of about 24 kD It contains a serine

protease-dependent cleavage site, which results in the release of a C-terminal

truncated NS2A product of about 22 kD (NS2Aα)21 Mutations at this cleavage site

block the production of infectious particles while the release of subviral particles

remains unimpaired48 Interestingly, not the inhibition of the processing event at the

NS2Aα site but the identity of the amino acids at the NS2Aα seems to be important

for this block For some flavivrus, like the Kunjin virus, it demonstrates that NS2A

localizes to presumed sites of RNA replication and that it binds to NS3, NS5 and the

3′ UTR49 This data suggest that NS2A might help to localize viral RNA to the

membrane-bound replication complex

The NS2B protein

NS2B is also a small (≈14 kD) membrane-associated protein50 and encodes three

hydrophobic membrane domains (two at the N-terminus and a single one at the

C-terminus).NS2B forms a stable complex with NS3 and acts as a cofactor for the

NS2B-NS3 serine protease51 The cofactor activity lies in a central peptide that

intercalates within the fold of the serine protease domain52, similar to the hepatitis C

virus (HCV) NS4A cofactor Mutation of conserved residues in NS2B can have

dramatic effects on autoproteolytic cleavage at the NS2B/NS3 junction and

transcleavage activities53,54

The NS3 protein

The NS3 is a large (≈70 kD) multifunctional protein, containing several activities

required for polyprotein processing and RNA replication The N-terminal third of the

protein is the catalytic domain of the NS2B-NS3 serine protease complex55-57 In

addition to cleaving the NS2A/NS2B, NS2B/NS3, NS3/NS4A, and NS4B/NS5

junctions, the protease generates the C-termini of mature capsid protein13,27 and

NS4A14, and can cleave at internal sites within NS2A and NS3 (Figue 1.2 and Table

1.1) Since the polyprotein processing is a prerequisite for assembly of viral replicase

complex, the viral NS3 protease represents an attractive therapeutic target The

protease preferentially cleaves after adjacent basic residues11 and the crystal structures

for the DENV-2 NS3 protease lacking the NS2B cofactor, with or without a substrate

inhibitor, have been solved58,59 These studies confirm the overall similarity to other

members of this enzyme family, but reinforce an unusually flexible mode of substrate

binding in the S1 pocket Single chain proteases have been recently created by

genetically fusing the NS2B cofactor region with the NS3 protease domain60,61 The

structures of the WNV and DENV-2 NS2B-3 proteases reveal that the cofactor region

of NS2B contributes a β-strand to forming the chymotrypsin-like fold, similar to what

has been seen with HCV52

The NS4A protein

Little is known about the hydrophobic NS4A protein with a size of approximately 16

kD Genetic studies indicated that NS4A interacts with NS1 and that this interaction is

important for RNA replication62

The NS4B protein

NS4B is relative small (≈27 kD) hydrophobic proteins Owing to its hydrophobic

character, it is associated to membranes Studies63 showed that NS4B initially appears

as a 30 kD protein that decreases to approximately 28 kD However, the nature of this

post-translationally modification is not known In addition, NS4B colocalizeds with

double-stranded RNA at putative sites of viral RNA replication45

The NS5 protein

With a predicted MW of 103kD, NS5 is a highly conserved, multifunctional and the

largest protein among the flavivirus proteins It contains a glycine-glycine-aspartic

acid [Gly-Gly-Asp (GDD)] motif, which is present in all RNA-dependent RNA

polymerase, which was demonstrated for recombinant dengue type 1 virus NS5

expressed in E coli64 The N-terminal part of NS5 contains a sequence element that is

homologous to methyltransferase65 The methyltransferase activity is probably

involved in methylation of 5′ cap structure Mutations destroying the GDD or

methyltransferase motif are lethal for RNA replication66

1.5 DEN/WNV NS2B/NS3 protease as drug target

We can know from the above reviews, the two-component NS2B/NS3 viral serine

protease activity plays a key role in flaviviral polyprotein processing and disruption of

this function has been shown to be lethal to virus replication56 This is an obligatory

step prior to viral RNA replication, thus identifying the viral serine protease as an

excellent therapeutic target NS2B/NS3 recognises and cleaves to the C-terminal side

of two highly conserved consecutive basic amino acid residues This is an unusual

specificity not shared by many host proteases, suggesting that inhibitors designed to

recognize this site may also be highly specific11,67 Due to the high level of sequence

conservation in NS3 within the flavivirus genus and the strong conservation of the

dibasic recognition sequence in polypeptide substrates, an antiviral inhibitor against

DEN or WNV may also be effective against other flaviviruses

Based on sequence and structure pattern analysis55, flaviviral NS3 is a trypsin-like

serine protease with a catalytic triad (His, Asp, Ser) This protease recognizes a highly

conserved cleavage site sequence consisting of two basic amino acids on the

N-terminal side and a short chain amino acid on the C terminal side Alignments of

various flavivirus cleavage sequences shows them to predominantly recognise

Lys-Arg, Arg-Arg, or occasionally Gln-Arg or Arg-Lys in positions P2 and P1 (see

Figure 1.3 for nomenclature), before the cleavage site, followed by a Gly, Ala, Ser or

Thr in position P1’ Mutagenesis of the native cleavage sequences of YFV has

confirmed the highly specific nature of this recognition sequence While only

conservative substitutions were tolerated within P2-P1’, mutagenesis of P3 or P4

generally had only a subtle or negligible effect.21,68,69

N

H O

Figure 1.3 Nomenclature for peptide residues (P3-P3’) and their corresponding

binding sites (S3-S3’) in the enzyme

Further biochemical analysis localised viral protease activity to the N-terminal 184

amino acids of NS3 and showed the protease activity to be dependent upon

association with a hydrophilic domain within NS2B56,70-72 An advance in the

understanding of the active protease was provided by the crystal structures of the

NS2B/NS3 proteases for both WNV and DEN252,73, as well as recent crystal

structures for the substrate-free and inhibitor-bound WNV NS2B/NS3 protease73

Kinetic parameters and substrate specificity of DENV2 protease were reported74,75

Recently, it was shown that shortening the linker to five amino acid residues from the

associated heterodimeric WNV protease which was active in cleaving a fluorogenic

peptide substrate, Boc-Gly-Lys-Arg-AMC76 Leung et al77 showed that the linker

between the cofactor, NS2B hydrophilic region, and the NS3 protease domain

(NS3-pro) could be substituted with G4-S-G4 linker and the precursor could be

expressed in E coli as a very active protease in a soluble, non-cleavable form, thus

obviating the denaturation and refolding steps in the purification of the protease

Using this active non-cleavable form of DENV2 and WNV proteases, a number of

groups reported the substrate specificity, kinetic parameters, and profiles of

peptide-based viral protease inhibitors A suitable enzymatic substrate was identified

by functional profiling using tetra peptide and octapeptide libraries comprising

~13,000 substrates78 Detailed specificity studies have led to the design of robust

screening assays in high-throughput formats, employing both colorimetric and

fluorescent readouts67,78-86

1.6 Existing inhibitors of DEN/WNV NS2B/NS3 protease

According to the characteristics and functions of the flavivirus NS2B/NS3 protease,

two possible strategies were performed for inhibiting the protease One strategy is to

block the interactions with its substrate whilst the other strategy is to block the

essential association between NS3 and its cofactor NS2B To date most attention has

been focused on the development of inhibitors that compete for the substrate-binding

cleft The preference of the substrate binding cleft of flaviviral proteases for ligands

with consecutive basic residues at P1 and P2 is not usual for mammalian proteases

and therefore might be exploited to provide inhibitors with specificity for

NS2B/NS3pro However, the charged nature of the interactions of such basic residues

makes the design of nonpeptidic inhibitors extremely challenging

The recently reported crystal structures of the NS2B/NS3 protease, together with the

results of mutagenesis studies and solution structure-activity relationships of

substrates/inhibitors, provides a basis for rational drug design and for structural

optimization of inhibitors that might target the shallow and highly solvent exposed

substrate binding cleft However, neither computational docking of virtual compound

libraries into the substrate-binding cleft nor high throughput screening of millions of

compounds by various groups has led to potent lead compounds directed towards the

substrate binding cleft So far, most studies have focused on the optimization of

substrate-based inhibitors and, while there has been some progress, reduction of

peptidic character and removal of positive charge without reduction in potency have

been problematic

The alternative strategy of blocking the association of NS2B is novel and remains to

be rigorously tested But it may avoid the problems facing development of substrate

based competitive inhibitors for the active site This strategy has recently been given

credence by the identification of a 5- amino-1-(phenyl) sulfonyl-pyrazol-3-yl class of

compounds (Figure 1.5, 36, 37) that behave as uncompetitive inhibitors, and are

suggested by in silico docking to interfere with the association of NS2B87 The

interaction between NS2B and NS3 has been characterised by the crystal structures of

NS2B/NS3pro52,73 with mutagenesis having identified specific residues within NS2B

as being essential for binding to NS3 (Figure 1.4 A, B)54,83,88-90 This information

could potentially be exploited for the design of an allosteric inhibitor capable of

blocking the interaction between NS2B and NS3 (or NS2B and substrate) The area

(cofactor target site 1, Figure 1.4A) binds into a deep hydrophobic trench in NS3,

which could be targeted by small aromatic, drug-like compounds.54,90 However,

because this region of the cofactor remains tightly associated in both inhibitor-bound

and substrate-free crystal structures, it is unknown if an allosteric inhibitor would be

capable of binding with high enough affinity to displace the bound cofactor The

region (cofactor target site 2, Figure 1.4B) binds into a deep hydrophobic pocket in

close proximity to the substrate binding cleft of NS3 and forms some interactions with

the bound substrate89 The 5-amino-1-(phenyl) sulfonylpyrazol- 3-yl class of

compounds (Figure 1.5, 36, 37) are suggested to bind the site complementary to

Leu79 and Phe85 cofactor residues and to interfere with attachment of this region87

As this flexible region of NS2B forms part of the substrate binding cleft, blocking of

its association is thereby likely to prevent substrate binding and cleavage Further

optimization and development of inhibitors targeted to these sites could potentially

lead to the generation of a novel antiviral drug candidate

Figure 1.4 Crystal structures of WNV NS2B/NS3pro and predicted substrate and

membrane interactions NS2B (red), NS3pro (blue) and catalytic triad (magenta in B)

A and B, polypeptide backbone and side chains of NS2B residues identified as

important for proteolytic activity are in yellow and correspond to A, site 1,

NS2B59-62 and B, site 2, NS2B75-87 Potential target sites for blocking cofactor

association with NS3pro are designated sites 1 and 2 Crystal structures used in

schemes is aprotinin-bound WNV NS2B/NS3pro (pdb: 2IJO)91

1.6.1 Peptidic inhibitors

The first Dengue 2 NS2B/NS3pro substrate-based peptide inhibitors with a Ki values

at micromolar concentrations were reported by Leung et al.77 These inhibitors were

designed based on native substrate sequences and replacing the cleavable amide bond

with an α-keto amide transition state isostere or replacing C-terminal carboxylic group

with an aldehyde group (Table 1.2, 1 and 2) Subsequently, the inhibition of

substrate-based peptides derived from the P6–P1 and the P1’–P5’ regions of the

natural polyprotein substrate have been investigated75 N-terminal cleavage site

peptides corresponding to the P6–P1 region of the polyprotein (3- 6) were found to

act as competitive inhibitors with Ki values ranging from 67 to 12μM The lowest Ki

value was found for the peptide representing the NS2A/NS2B cleavage site, RTSKKR

10), displaying Ki values in the range from 188 to 22 μM Peptides corresponding to

the P1’–P5’ region of the polyprotein cleavage sites (11) had no effect on enzymatic

activity even at a concentration of 1 mM

N N

H

OH NH 2

NH 2 FASGK

N H O

H FASGK

NH

NH 2

H 2 N

N H OH

OH FASGK

Scheme 1.1, Chemical form of aldehyde inhibitor in water

Since substrate-based peptide inhibitors with aldehyde as a warhead (2) showed

higher inhibition compared to the equivalent α-ketoamide inhibitor (1), the inhibition

of the NS2B-NS3 protease by aldehyde inhibitors have been investigated in detail by

a number of groups81,85,86,92 Modeling study showed that the aldehyde inhibitors bind

to the substrate-binding cleft by forming a covalent bond with the catalytic Ser135

While the true level of inhibition by the actual aldehyde form of such inhibitors may

be underestimated Fairlie and co-workers67 have reasoned that aldehyde inhibitors

containing arginine residue at P1 show only modest activity against flavivirus NS3

proteases, because they are in equilibrium with their hydrate and cyclic forms, with

only about 5% of the active free aldehyde functionality exposed for the interaction

with the active site serine hydroxyl group (see Scheme 1.1) Beside aldehyde warhead,

Yin et al85 also showed that using other electrophilic warheads (12 and 13) were able

to greatly increase inhibition However, these warheads are unlikely to be

incorporated into drug candidates as they are able to interact indiscriminately with

multiple receptors and organic compounds in a cell and are therefore likely to be toxic

in vivo

Among these aldehyde inhibitor, the tetrapeptide aldehyde inhibitor, Bz-Nle-KRR-H

(14), was found to have a relatively high level of inhibition against the DEN2 and

WNV proteases (Ki 5.8μM, 4.1μM, respectively), suggesting a good lead compound

for generating a broad spectrum inhibitors In general, structure-activity relationships

(SAR) observed that S1 and S2 pockets of protease are the key peptide recognition

sites For WNV protease inhibitors, a peptide side chain residue capable of both

σ-stacking and hydrogen bonding is favored in the S1 pocket, while a positively

charged residue is preferred in the S2 pocket For Den protease inhibitors, the

interactions of P2 side chain are more important than P1 followed by P3 and P4 For

example, the inhibitor (15), whose P2 Arg was substituted by Lys comparing with the

inhibitor (14), has 2-fold improvement in Ki against WNV protease, but an 8-fold

increase in Ki against DEN2 protease81,82,86 This is consistent with the observed

difference in substrate specificity between DEN2 and WNV proteases78,89 At the

same time, SAR results, together with the modeling study, showed that the two highly

positive charged P1 and P2 residues within these inhibitors are the major contributing

factor to binding affinity Truncation of such DEN2 or WNV tetrapeptide inhibitors

(15) to tripeptide inhibitors (16) or even dipeptide inhibitors (17) apparently had little

effect on inhibitor potency However, the incorporation of multiple D-arginine

residues into amide inhibitors (18-24) has been shown to enhance inhibitor potency

for the WNV protease in vitro93 The level of inhibition was found to be improved

500-fold by increasing the number of D-Arg residues from 6 to 12 It is unlikely that

the enzyme has specificity for binding to multiple arginine residues outside of the

P2-P1 recognition site and so another explanation is that multiple arginine residues

have a cumulative effect on the association at S1 and S2

Since homology models of the protease showed the S1 and S2 pockets are the key

peptide recognition sites, for substrate recognition, the side chain of P1 and P2 are

important for inhibitor binding Changing the positive charged P1 and P2 residues of

peptide inhibitor to other residues can severely decrease inhibitor potency For

example, replacement by Ala or Phe at P1 or P2 position (25-28) severely decreases

inhibitor potency However, there are two exceptions The one exception is P1 Ala

replacement (25), which still showed good inhibition on WNV protease This can be

explained that the dibasic recognition site binding to the Lys and Arg in P3 and P2

respectively, before the aldehyde interacts with the catalytic Ser135 The other

exception, the replacement of the P1 Arg to Phe (27) showed only a 3- fold increase

in the Ki for DEN2 NS2B/NS3pro but 27-fold increases for WNV NS2B/NS3pro

This suggests that while the S1 pocket of DEN2 NS2B/NS3pro is large enough to

accommodate an aromatic compound, the S1 pockets of the WNV proteases may be

slightly smaller

Table 1.2, Summary of Peptidic Inhibitors of NS2B/NS3pro

Recently, Stoermer et al92 reported a class of small and highly potent inhibitors

against WNV NS2B/NS3pro These inhibitors are cationic tripeptides with

nonpeptidic caps at the N-terminus and aldehyde at the C-terminus By incorporating

an aromatic phenylacetyl or 4-phenylphenylacetyl group at the putative P4 position

(29 and 30), the inhibitory potency increased to the low nanomolar range

1.6.2 Nonpeptidic inhibitors

Compared to the peptidic inhibitor, only a few nonpeptidic inhibitors showing

satisfactory inhibition have been identified against DEN and WNV NS2B/NS3pro

(Figure 1.5) Ganesh et al.94 identified a few small nonpeptidic compounds (31-33)

which were found to be competitive inhibitors of both the WNV and DEN2

NS2B/NS3pro at micro-molar concentrations These compounds were identified by

computational screening for potential mimics of the bifurcated P1 Arg side chain

observed in the crystal structure of DEN NS3 protease Modeling studies on

compounds 32 and 33 showed that the amide oxygen at the indolinone ring in

compound 32 and oxygen of the phosphonic acid group in compound 33 could make

hydrogen bond with active site Ser 135 in DEN2 NS3-pro This can explain that

compounds 32 and 33 showed better inhibition than compound (31) Two

cyclohexenyl chalcone derivatives95, panduratin A (34) and 4-hydroxypanduratin A

(35), also showed competitive inhibitory activities towards dengue 2 virus NS3

protease with the Ki values of 21 μM and 25 μM, respectively However their mode of

binding has not been investigated Besides the above rational designed inhibitors of

WNV NS2B/NS3pro, some inhibitors of this protease were identified by high

throughput screening which provides a powerful complement to structure-based

rational design of small-molecule inhibitors of proteases Johnston et al87 reported

some novel, uncompetitive inhibitors of WNV NS2B-NS3pro, 5-amino-1- (phenyl)

sulfonyl- pyrazol-3-yl class of compound (36 and 37), that appear to interfere with the

productive interactions of the NS2B cofactor with the NS3pro domain These

compounds showed relatively high potencies (IC50 < 200 nM) against recombinant

NS2B/NS3pro in vitro and provided another approach that blocks the association of

NS2B with NS3 for designing inhibitors of falvivirus NS3 protease Another

high-throughput screening assay for the WNV NS2B/NS3 protease carried out by

Mueller et al 96identified 3 compounds (38-40) that showed competitive inhibition with

Ki values of 3.2 μM, 3.4 μM and 37.3 μM, respectively These 3 compounds were

also tested on the dengue virus type 2 protease showing Ki values of 28.6 μM, 30.2

μM and 17.0 μM respectively Although all of the nonpeptidic small molecule

inhibitors of NS2B/NS3pro currently identified (31-40) are not sufficiently active to

be viable drug candidates, analysis of their binding modes may provide useful

information for rational drug design of a nonpeptidic drug without the problems

encountered by substrate-based inhibitors