A non mouse adapted dengue virus strain as a new model of severe dengue infection in AG129 mice

Journals: Tan GK, Ng JKW, Trasti SL, Schul W, Yip G and Alonso S 2010 A non mouse-adapted dengue virus strain as a new model of severe dengue infection in AG129 mice.. infection with

A NON MOUSE - ADAPTED DENGUE VIRUS STRAIN

AS A NEW MODEL OF SEVERE DENGUE

I would like to express my deepest gratitude to my supervisor, Dr Sylvie Alonso for her guidance and support throughout my master candidature I thank her for all the advices she has given and the trust she has in me to carry out my project independently Also, I am thankful for the many opportunities she has given me to publish various pieces of work with her None of this would have been possible without her expertise, patience and encouragement

I sincerely thank Jowin, for his relentless help and support throughout this period of time To Annabelle, Emily, Weixin, Wenwei: for the advices they have given me in various aspects of my work and being there for me whenever I need them and to the rest of the lab members: for making the lab environment full of life and an enjoyable place to work in

I would also like to thank my parents for all the support they have given me all these while and their understanding for the times when I have to devote my time to my work

Last but not least, I would like to thank my husband, Andrew, for every single thing

he has done for me I wouldn’t have been able to achieve all these without his help and kind understanding

Journals:

Tan GK, Ng JKW, Trasti SL, Schul W, Yip G and Alonso S (2010) A non

mouse-adapted dengue virus strain as a new model of severe dengue infection in AG129

mice PLoS Negl Trop Dis 4(4): e672

Tan GK and Alonso S (2009) Pathogenesis and prevention of dengue virus

infection: state-of-the-art Curr Opin Infect Dis (Invited review) 22(3):302-8

Sim AC, Lin W, Tan GK, Sim MS, Chow VT, Alonso S (2008) Induction of

neutralizing antibodies against dengue virus type 2 upon mucosal administration

of a recombinant Lactococcus lactis strain expressing envelope domain III

antigen Vaccine 26(9): 1145-54

A CKNOWLEDGEMENTS i

P UB ICATIONS ii

C ONT NTS iii

S UMMARY vii

T AB ES AND F IGURES ix

A B REVIATIONS xi

1 I NTRODUCTION 1

1.1 D ENGUE VIRUS 1

1.1.1 C LASSIFICATION 1

1.1.2 D ENGUE VIRUS PROTEINS 1

1.1.2.1 C APSID (C) PROTEIN 1

1.1.2.2 M EMBRANE (M), P RE -M (P R M) AND ENVELOPE (E) PROTEINS 2

1.1.2.3 NS1 PROTEIN 3

1.1.2.4 NS3 PROTEIN 3

1.1.2.5 NS5 PROTEIN – THE VIRAL RNA- DEPENDENT RNA POLYMERASE (R D R P ) 4

1.1.2.6 O THER NON - STRUCTURAL PROTEINS 5

1.2 T RANSMISSION CYCLE OF DENGUE VIRUS 5

1.3 D ENGUE VIRUS REPLICATION IN HUMANS 6

1.4 T HE SPREAD AND THREAT OF DENGUE VIRUS 10

1.5 D ENGUE INDUCED DISEASE 10

1.5.1 C LASSIFICATION 10

1.5.2 C LINICAL MANIFESTATIONS OF DHF/DSS 10

1.6 D ENGUE PATHOGENESIS 14

1.6.1 I MMUNE RESPONSE TO DENGUE VIRUS INFECTION 14

1.6.1.1 P ROTECTIVE ROLE OF IMMUNE RESPONSE AGAINST DENGUE VIRUS INFECTION 14

1.6.1.2 I MMUNOPATHOGENESIS OF DENGUE VIRUS INFECTION : T HE ENHANCEMENT HYPOTHESIS OF DHF/DSS PATHOGENESIS 15

1.6.1.3 N ON - ENHANCING MECHANISMS AS RISK FACTORS FOR DHF/DSS: A UTOIMMUNE AND COMPLEMENT - MEDIATED PATHOGENESIS 20

1.6.2 H OST GENETIC FACTORS 22

1.6.3 V IRUS VIRULENCE 23

1.7 D ENGUE PREVENTION 25

1.7.1 V ECTOR CONTROL 25

1.7.2 D ENGUE VIRUS VACCINES 26

1.7.2.1 L IVE ATTENUATED VACCINE (LAV) 28

1.7.2.2 C HIMERIC VIRUS VACCINE 29

1.7.2.3 S UBUNIT DENGUE VACCINE 30

1.8 M ODELS FOR DENGUE VIRUS INFECTION 33

1.8.1 N ONHUMAN PRIMATE ANIMAL MODEL 33

1.8.2 E STABLISHED MOUSE MODELS 34

1.8.2.1 M OUSE - ADAPTATION OF DENGUE VIRUS 34

1.8.2.2 W ILDTYPE MOUSE STRAINS 35

1.8.2.3 M OUSE - HUMAN CHIMERAS 36

1.8.2.4 K NOCKOUT MOUSE STRAINS 38

2 A IM 41

3 M AT RIALS AND M ETHODS 42

3.1 V IRUS STRAIN AND GROWTH CONDITIONS 42

3.2 V IRUS QUANTITATION 42

3.3 M ICE I NFECTION 43

3.4 E NZYME - LINKED IMMUNOADSORBENT ASSAY (ELISA) 43

3.5 P LAQUE REDUCTION NEUTRALIZATION TEST (PRNT) 44

3.6 V IRUS Q UANTIFICATION IN INFECTED MICE 45

3.7 H ISTOPATHOLOGY 46

3.8 V ASCULAR LEAKAGE ASSESSMENT 46

3.9 C YTOKINE DETECTION 47

3.10 H AEMATOLOGY 47

3.11 S TATISTICAL ANALYSIS 47

4 R ESULTS P ART I 48 I I NTRAP RITONEAL I NF CTION OF D2Y9 P IN AG1 9 48 9 4.1 S URVIVAL RATE IN D2Y98P INFECTED AG129 48

4.2 V IREMIA PROFILE IN D2Y98P INFECTED AG129 50

4.3 S PECIFIC ANTIBODY TITRES AND PRNT 50 LEVELS AGAINST D2Y98P 52

4.4 T ISSUE TROPISM AND KINETIC OF VIRUS REPLICATION IN D2Y98P- INFECTED AG129 55

4.5 H ISTOLOGICAL EXAMINATION OF ORGANS IN D2Y98P- INFECTED AG129 57

4.6 V ASCULAR LEAKAGE IN D2Y98P- INFECTED AG129 61

4.7 C YTOKINE EXPRESSION LEVELS IN D2Y98P- INFECTED AG129 63

4.8 H AEMATOLOGY IN D2Y98P- INFECTED AG129 65

5 R ESULTS P ART I I 67

S UBCUTANEOUS I NF CTION OF D2Y9 P IN AG1 9 67 9

5.1 C OMPARATIVE ANALYSIS OF SURVIVAL RATE AND MEAN VIRUS TITRES IN IP

VS SC D2Y98P- INFECTED AG129 67

5.2 H ISTOPATHOLOGICAL CHANGES AND VASCULAR LEAKAGE IN SC INFECTED AG129 71

5.3 D OSE - DEPENDENT SURVIVAL RATE IN D2Y98P SC INFECTED AG129 74

5.4 S PECIFIC ANTIBODY TITRES AND PRNT 50 LEVELS AGAINST D2Y98P IN SC I NFECTED AG129 77

5.5 T ISSUE TROPISM AND KINETIC OF VIRUS REPLICATION IN AG129 INFECTED SC WITH 10 4 PFU OF D2Y98P VIRUS 78

5.6 H ISTOPATHOLOGY , VASCULAR PERMEABILITY AND CYTOKINE EXPRESSION LEVELS IN 10 4 PFU D2Y98P SC INFECTED AG129 83

5.7 H AEMATOLOGICAL PARAMETERS IN D2Y98P SC INFECTED AG129 88

6 D ISCUS ION 90

6.1 M OUSE MODELS FOR DENGUE INFECTION 90

6.2 T HE AG129 MOUSE MODEL FOR DENGUE INFECTION 91

6.3 I NFECTION OF D2Y98P IN AG129 MICE 92

6.3.1 I P ROUTE OF D2Y98P INFECTION IN AG129 MICE 93

6.3.2 S C ROUTE OF D2Y98P INFECTION IN AG129 MICE 98

6.4 C LINICAL RELEVANCE TO DENGUE DISEASE IN HUMANS 102

7 C ONCLUSION 104

8 F UTURE W ORK 105

9 R E ERENCES 107

1 A P ENDIX 127

The spread of dengue (DEN) worldwide combined with an increased severity of the DEN-associated clinical outcomes have made this mosquito-borne virus of great global public health importance.Progress in understanding DEN pathogenesis and in developing effective treatments has been hampered by the lack of a suitable small animal model Most of the DEN clinical isolates and cell culture-passaged DENV strains reported so far, either require the need for host adaptation, inoculation with a high dose and/or intravenous (iv.) administration to elicit a virulent phenotype in mice, which results at best in a productive infection with none, few or irrelevant disease manifestations, and with mice dying at the peak of viremia

Here we describe a non mouse-adapted DEN2 virus strain (D2Y98P) that is highly infectious in AG129 mice (lacking interferon-α/β and -γ receptors) upon intraperitoneal (ip.) and subcutaneous (sc.) administration Infection with a high dose (107 PFU) of D2Y98P (via the ip and sc route) induced a cytokine storm, massive organ damage, and severe vascular leakage, leading to hemorrhage and rapid death of the animals at the peak of viremia In contrast, very interestingly and uniquely, ip infection with a low dose of D2Y98P (104 PFU) led to asymptomatic viral dissemination and replication in relevant organs, followed by non-paralytic death of the animals few days after virus clearance, similar to the disease kinetic described in DEN-infected patients Tissue damage and increased vascular permeability but no hemorrhage, were observed only at moribund state of infected animals, suggesting intact vascular integrity, a cardinal feature of DEN shock syndrome Interestingly, whereas high pro-inflammatory (TNF-α, IL-6 and IFN-γ) cytokine levels were detected at the peak of viremia – during which the extent of tissue damage and

vascular leakage was minimal; basal production of these cytokines was instead measured in the animals at moribund state, where tissue damage and vascular leakage were more severe This observation suggests that other cytokines or other inflammation-independent mechanisms may be responsible for the vascular leakage phenomenon observed in the infected animals at moribund state

Strikingly, infection with 104 PFU of D2Y98P by the sc route instead, induced clinical manifestations and disease which closely resembled that of the high dose model, indicating that the sc route of infection is more potent than the ip route at triggering a virulent phenotype of DEN infection in AG129 Altogether, infection with the D2Y98P strain thus offers the opportunity to further decipher some of the aspects of DEN pathogenesis and provides a new platform for drug and vaccine testing

TABLES AND FIGURES

Figure 1.1: The intracellular life cycle of DENV .………9

Figure 1.2: Classification and clinical manifestations in humans infected with DENV … ………12

Figure 1.3: A diagrammatic view of the antibody dependent enhancement (ADE) phenomenon of DENV infection ……… 17

Figure 1.4: Schematic representation of the immunopathogenesis of severe DEN disease mediated by cross reactive anti-DEN antibodies and T cells ……….……….19

Table 1.1: Proposed mechanisms involving the host immune system in mediating severe DEN disease in humans ……… 21

Table 1.2: Characteristics of an ideal DEN vaccine ………….……… 27

Figure 1.5: Summary of candidate DEN vaccines currently in development .…32

Figure 4.1: Survival rate in AG129 mice infected with a dose range of D2Y98P virus ………49

Figure 4.2: Body weight changes in D2Y98P-infected AG129 ……… 49

Figure 4.3: Viremia profile in D2Y98P-infected AG129 ……….51

Figure 4.4: Specific antibody titres against D2Y98P ……… 53

Figure 4.5: Neutralizing antibody titres specific against D2Y98P ……… 54

Figure 4.6: Gross pathological changes in D2Y98P-infected AG129 ………… 56

Figure 4.7: Quantification of virus titres in the liver, spleen and brain of D2Y98P-infected AG129 ……… 56

Figure 4.8: Histopathology of D2Y98P-infected AG129 ……….59

Figure 4.9: Serum levels of aspartate (AST) and alanine (ALT) transaminases 60

Figure 4.10: Vascular leakage in D2Y98P-infected AG129 ……… 62

Figure 4.11 Pro-inflammatory cytokine expression in D2Y98P-infected AG129 ……… 64

Table 4.1: Haematology in D2Y98P-infected AG129 ………66

Figure 5.1: Gross pathological changes in D2Y98P sc infected AG129 ………67 Figure 5.2: Comparative analysis of the survival rate and viremia titres in AG129

infected with D2Y98P virus via the ip and sc routes ……… 68 Figure 5.3: Comparative analysis of mean virus titres in organs of AG129 infected

with D2Y98P virus via the ip and sc route ……… 70 Figure 5.4: Histopathology and vascular leakage in sc infected AG129 ……….72

Figure 5.5: Survival rate in AG129 sc infected with a dose range of D2Y98P

virus ………75 Figure 5.6: Body weight changes in D2Y98P sc infected AG129 ……… 76

Figure 5.8: Gross pathological changes in D2Y98P sc infected AG129 .…… 78 Figure 5.9: Comparative analysis of virus titres in organs of AG129 infected sc

Figure 5.10: Comparative analysis of mean virus titres in organs of AG129 infected

with 104 PFU of D2Y98P via the sc and ip route ……….……81 Figure 5.11: Histopathology of 104 PFU D2Y98P sc infected AG129 … … ….84

Figure 5.12: Serum levels of aspartate (AST) and alanine (ALT) transaminases

……… 85

Figure 5.13: Comparative analysis of vascular leakage and albumin concentration in

D2Y98P sc infected AG129 .……….86

Figure 5.14: Pro-inflammatory cytokine expression in D2Y98P sc infected AG129

……… 87

ABBREVIATIONS

HRP horse radish peroxidase

p.i post-infection

1 INTRODUCTION

1.1 DENGUE VIRUS

1.1.1 CLASSIFICATION

Dengue virus (DENV) consists of four antigenically related but genetically distinct

serotypes (DEN 1-4) They belong to the Flavivirus genus, family Flaviviridae,

whose other members include Yellow Fever Virus (YFV), West Nile Virus (WNV), Japanese Encephalitis Virus (JEV) and Tick-Borne Encephalitis Virus (TBEV) They are arthropod- borne and are transmitted via infected tick or mosquito vectors

1.1.2 DENGUE VIRUS PROTEINS

The flavivirus genome is a positive single-stranded RNA of approximately 11kb

(Chambers et al., 1990a) It contains a large open reading frame encoding a single

polyprotein precursor flanked by 5' and 3' untranslated regions (UTR) The polyprotein precursor is co-translationally and post-translationally processed by host proteases and viral serine protease to produce ten mature viral proteins: three of which are structural proteins (C, prM, and E) and seven are non-structural (NS1, NS2A,

NS2B, NS3, NS4A, NS4B, NS5) (Chambers et al., 1990a)

1.1.2.1 CAPSID (C) PROTEIN

The viral nucleocapsid is made up of the virus genome surrounded by a 13-16 kDa capsid (C) protein Two highly basic domains identified at the N and C termini of the

C protein have shown to bind specifically to the 5’ and 3’ non-coding region (NCR)

of the viral RNA genome, likely via interactions with conserved stem loop structures (Khromykh and Westaway, 1996) The C protein is found to exist as two forms in an

infected cell: the membrane-anchored form (Canchor) – when cleaved by host cell signalases, and the mature virion-associated form (Cvir) – achieved during the cleavage of the hydrophobic tail by the viral NS2B/3 proteases which remains integrated to the endoplasmic reticulum (ER) membranes and functions as a

membrane anchor (Markoff et al., 1997)

1.1.2.2 MEMBRANE (M), PRE-M (PRM) AND ENVELOPE (E) PROTEINS

The viral envelope consists of a lipid bilayer in which embeds the envelope (E) and membrane (M) proteins (Heinz and Roehrig, 1990) The M protein, depending on the maturity state of the virion, may be found in two forms; during the immature state of the virion within infected cells, the pre-M (PrM) protein predominates and associates

with the E protein to form a stable heterodimer (Zhang et al., 2003) The prM protein

is believed to protect E from pH-induced reorganization and premature fusion during

secretion (Guirakhoo et al., 2001; Guirakhoo et al., 2002) and serves possibly as a

chaperone for proper E folding and assembly (Heinz et al., 2003) Hence the

incorporation of this heterodimer into the virions during budding from the lumen (Mackenzie and Westaway, 2001) may be vital for the maintenance of the E protein in

a stable, fusion-inactive conformation before viral release (Konishi and Mason, 1993)

The E protein is a major constituent of the virus envelope (Rice et al., 1986) and

comprises three domains: domain I - a centrally located β barrel, domain II – containing a dimerization region and the fusion peptide, and domain III – harbouring the receptor-binding activity (Rice, 1996) As suggested by the functions of each individual domains, the E protein serves many purposes including receptor binding

(Anderson et al., 1992; Chen et al., 1996; Wang et al., 1999), membrane fusion

(Schalich et al., 1996; Rice, 1996), virion assembly (Stiasny et al., 2002) and is the

primary target for neutralizing antibodies (Heinz, 1986) Mutations within the E protein have been observed to significantly influence the function and virulence of DEN and other flaviviruses (McMinn, 1997)

1.1.2.3 NS1 PROTEIN

The NS1 protein is a variably glycosylated protein and is expressed in three forms: an

ER resident form that colocalizes with the viral replication complex, a

membrane-anchored form and the soluble form (sNS1) (Lindenbach and Rice, 2003)

Immunolocalization of NS1 with the double stranded replicative form (RF) RNA by both light and electron microscopy has suggested its involvement in viral RNA

replication (MacKenzie et al., 1996) Furthermore, mutations within NS1 have been

shown to inhibit or abolish viral replication significantly and as a result, lead to RNA

accumulation within the infected cell (Crabtree et al., 2005) NS1 may also play a role

in virion assembly or maturation, as it has been found to be associated with the immature E protein in the lumen of the ER (Fan and Mason, 1990) sNS1 is also another dominant target of the humoral immunity and may play a significant role in the pathogenesis of DEN disease (Falconar and Young, 1991)

1.1.2.4 NS3 PROTEIN

The NS3 protein - the second-largest protein encoded in the DEN genome containing the most conserved amino acid sequence between DEN strains - is a viral protease

involved in the cleavage of the translated viral polyprotein (Falgout et al., 1991) The

protease activity of NS3 has been demonstrated for several flaviviruses, including

DEN but requires the presence of the NS2B protein for efficient activity (Falgout et

al., 1991) Structural and mutational studies of the NS2B-NS3 protease complex have

determined that the NS2B cofactor is critical for protease activity and acts to both stabilize the NS3 structure and form part of the substrate binding site

(Niyomrattanakit et al., 2004; Erbel et al., 2006) The NS3 protein also plays roles in

viral RNA replication through the nucleotide triphosphatase (NTPase), RNA 5’ triphosphatase (RTPase) and helicase activities (Wengler and Wengler 1991; Wengler and Wengler 1993) It has been identified that the C terminal portion of NS3 contains domains that allow RNA binding during RNA replication, as well as NTPase and

helicase activities (Westaway et al., 1997)

1.1.2.5 NS5 PROTEIN – THE VIRAL RNA-DEPENDENT RNA POLYMERASE

Post-translational modification of NS5 has shown to be important in both the

regulation of its function and cellular distribution (Kapoor et al., 1995) The degree of

phosphorylation affects both the interaction of NS5 with NS3, critical to the formation

of the viral replication complex, as well as the localization of NS5 within the infected cell A highly phosphorylated NS5 has been noted to dissociate from NS3 in the cell cytoplasm and be translocated to the nucleus, via the action of the nuclear localization

sequences (NLS) at the central region of NS5 (Forwood et al., 1999) Phosphorylation

of the NLS region instead, has been reported to inhibit nuclear transport of the NS5

protein (Forwood et al., 1999)

1.1.2.6 OTHER NON-STRUCTURAL PROTEINS

DENV also encodes four small hydrophobic proteins –NS2A, NS2B, NS4A and NS4B – all of which do not have well conserved primary sequences between viruses

of different serotypes (Chambers et al., 1990) NS2B, as aforementioned, is a cofactor required for the proteolytic activity of the NS3 protease (Falgout et al., 1991) NS2A

and NS4A, as documented in Kunjin Virus (KUNV)-infected cells via immunoelectron microscopy, is closely associated with other molecules that define the viral replicase complex, including both the NS3/NS5 and double stranded

‘replicative form’ viral RNA (MacKenzie et al., 1998) These proteins have also been

shown to bind specifically to the 3’ NCR of viral template RNA, suggesting the direct role of NS2A in the replication of viral RNA and that NS4A acts to target or anchor

the replication complex (MacKenzie et al., 1998) In the YFV NS2A protein,

mutations at an internal cleavage site have also been reported to abrogate effective production of infectious virus despite normal RNA replication and protein synthesis, hence suggesting the importance of this protein in virion assembly and/or release (Kummerer and Rice, 2002) Less well characterized however, is the NS4B protein Nevertheless, it has been identified to be involved with viral replication via its

interaction with a series of internal transmembrane domain sequences (Westaway et al., 1997) NS4B in the DEN2 also acts as an interferon antagonist via the blockage of

Signal Transducers and Activators of Transcription (STAT)-1 activation

(Munoz-Jordan et al., 2005)

1.2 TRANSMISSION CYCLE OF DENGUE VIRUS

The transmission cycle of DEN by the mosquito Aedes aegypti begins with a viremic DEN-infected human (Whitehead et al., 2007) During the viremic period that lasts

for about five days, an uninfected female Aedes aegypti mosquito takes a blood meal

from the DEN-infected person The virus undergoes an extrinsic incubation period of eight to twelve days within the mosquito, replicating within the midgut, nerve tissue, ovaries and body fat The virus subsequently moves to the mosquito’s salivary glands where it continues to replicate Through its saliva, the infected mosquito is capable, during probing and blood feeding, of transmitting the virus to susceptible individuals for the rest of its life The virus proceeds to replicate in the second person and symptoms manifest at an average of four to seven days after the mosquito bite in the second infected human This is known as the intrinsic incubation period (Whitehead

et al., 2007)

1.3 DENGUE VIRUS REPLICATION IN HUMANS

The replication cycle of DENV begins in the human host when the viron infects a permissive host cell In humans, langerhan cells, dermal and interstitial dendritic cells (DCs) have long been proposed as initial targets for DENV infection at the site of the

mosquito bite (Wu et al., 2000) Among the skin resident immunocytes are also the

dermal macrophages which exhibit potential in DENV internalization but confer

resistance to viral propagation (Kwan et al., 2008) This suggests the possibility of a

first line defence against DENV spread at the site of infection Downstream host target cells permissive for DENV replication are invariably identified over decades of research, including cells of the mononuclear phagocytic lineage such as monocytes,

macrophages and dendritic cells (Jessie et al., 2004) It is currently still debatable

whether hepatocytes, lymphocytes, endothelial cells as well as neuronal cells are susceptible to DENV infection (Halstead, 2007) although cell lines of similar lineages

have been shown to permit DENV replication in vitro (Thongtan et al., 2003; Lin et

al., 2002; Balsitis et al., 2009; Liu et al., 2009)

Crucial to the initial step of DENV entry into target cells is the interaction of the viral envelope glycoprotein (E) with putative host receptors such as heparan sulfate (Chen

et al., 1997), heat shock protein (Hsp) 70, Hsp 90 (Valle et al., 2005), GRP78 (BiP) (Jindadamrongwech et al., 2004), CD14 (Chen et al., 1999), 37-kDa/67-kDa high

affinity laminin receptor (Thepparit and Smith, 2004), liver/lymph node-specific intercellular adhesion molecule 3 (ICAM-3)-grabbing non-integrin, DC-specific

ICAM-3-grabbing non-integrin (DC-SIGN) (Tassaneetrithep et al., 2003; Sanchez et al., 2003) and the mannose receptor (Miller et al., 2008) for virus-cell

Navarro-membrane fusion These interactions trigger access for the virus to the endocytic pathway via clathrin or Rab-5 mediated endocytosis and transport processes (van de

Schaar et al., 2007; Krishnan et al., 2007)

Upon internalization of the virus by receptor-mediated endocytosis (RME) and acidification of the endosome, the E protein – that exists as homodimers in the mature

virion with an inaccessible fusion peptide (Modis et al., 2004) – then undergoes

irreversible trimerization to expose the fusion peptide and mediate endosomal fusion

(Modis et al., 2004), resulting in the release of the viral RNA into the cytoplasm and genome uncoating (Kimura and Ohyama, 1988; Guirakhoo et al., 1989) Following

the release of viral RNA into the cytoplasm, translation of the input strand first takesplace to generate the viral RNA polymerase (NS5) in order to establish a productive infection (Grun and Brinton, 1987) Following the synthesis of the replication complex - including NS5 along with the other three structural and six non-structural

proteins, the virus then switches from translation to synthesis of a negative-strand intermediate, which serves as a template for the production of multiple copies of positive-strand viralRNA (vRNA) (Fields et al., 2001)

Subsequently, the viral mRNA is translated and the structural C, prM, and E proteins, along with the vRNA, are assembled into progeny virions, which are transported through the Golgi compartment The immature virions are transported via the secretion pathway and undergoes maturation where the prM protein is processed to

the mature M protein in the trans Golgi compartment by furin, hence allowing the E

receptor binding domain to be exposed for infection It is this fully mature state of the

viron that makes it competent for infection (Heinz et al., 2003) Finally, release of the progeny virus from the host cell proceeds via secretory exocytosis (Hase et al., 1987)

(Fig 1.1)

Figure 1.1: The intracellular life cycle of DENV DENV binds (step 1) and enters (step 2)

cells via RME Endosomal acidification (step 3) results in an irreversible trimerization of the viral E protein, exposing the fusion domain After being uncoated, the vRNA is translated (step 4) at ER-derived membranes, where it is processed into three structural and seven NS proteins After the viral replication complex is synthesized, vRNA translation switches off and RNA synthesis (step 5) begins Subsequently, successive rounds of translation (step 6) are followed by assembly in the ER The virion is maturated in the Golgi compartment (step 7) and exits via the host secretory pathway L-SIGN, liver/lymph node-specific ICAM-3- grabbing nonintegrin; PTB, polypyrimidine tract binding protein; EF-1α, elongation factor 1α; hnRNP L, heterogeneous nuclear ribonucleoprotein L; PDI, protein disulfide isomerise

(Adapted from Clyde et al., 2006)

1.4 THE SPREAD AND THREAT OF DENGUE VIRUS

DEN is endemic in subtropical and tropical countries (Gubler, 2002) The highly

domesticated Aedes aegypti mosquito represents the main vector for DEN

transmission to humans in those regions However, the strong ecological plasticity of

the closely related mosquito Aedes albopictus, has allowed a further spread of DEN

throughout the globe With approximately half the world's population residing in DEN endemic regions (Tolle, 2009) and more than 50 million new infections projected to occur annually (Halstead, 2007), DEN certainly poses as a global health threat

1.5 DENGUE INDUCED DISEASE

1.5.1 CLASSIFICATION

DEN infection in humans can be asymptomatic or range from mild acute febrile illness associated with the classical DEN fever (DF) to severe DEN hemorrhagic fever/ DEN shock syndrome (DHF/DSS) According to the World Health Organization guidelines (WHO guidelines, 2008), all four clinical/laboratory criteria (fever, hemorrhagic tendency, thrombocytopenia, and capillary leakage) must be fulfilled for DHF classification Increased vascular permeability may progress to vascular collapse (DSS) and death

hrs of fever - resolve Pleural effusion, ascites and hemoconcentration (hematocrit

increased by more than 20%) are indicative of leakage (Bhamarapravati et al., 1967)

This can quickly progress to shock if volumic loss is not remedied with proper fluid therapy Bleeding tendencies are also observed in DSS patients: hemorrhagic manifestations range from a positive tourniquet test to the presence of petechiae, ecchymosis or purpura and/or spontaneous bleeding from the mucosa, gastrointestinal tract or any body orifice Hemoconcentration and marked thrombocytopenia (platelet count <100 x 109/L) are the two hallmarks of DHF/DSS (Kurane, 2007)

Hepatic involvement has also been implicated in DEN disease; hence elevations in aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels (Burke

et al., 1988), - indicative of liver damage - are noted in DEN-infected patients While

the three organ systems, haematological, vascular and hepatic, are well documented to

be involved in the pathological changes in DHF/DSS (Leitmeyer et al., 1999), a

continuum of atypical clinical presentations implicated in DEN infection that involve other organs, resulting in DEN associated encephalitis, myocarditis and cholecystitis,

is increasingly reported (Gulati et al., 2007; Wasay et al., 2008; Park et al., 2008)

These manifestations can increase the severity of the disease but may fall short of fulfilling the criteria for DHF Such presentations should be categorized separately (Fig 1.2A)

Classical With unusual

hemorrhage

Dengue Hemorrhagic Fever

(DHF) Expanded Dengue Syndrome

Plasma Leakage

No Shock With Shock

(DSS)

CNS encephalitis Hepatitis Myocarditis Cholecystitis

Hypovolumic shock/

Severe bleeding/ DIC

Timeline (Days post- infection)

DF DHF/DSS

Spontaneous bleeding

from mucosa / G.I tract /

any body orifice

Figure 1.2: Classification and clinical manifestations in humans infected with DENV (A)

Case definitions for dengue fever (DF), dengue haemorrhagic fever (DHF) and dengue shock syndrome (DSS) as provided by WHO in 2008 (B) A generalized time course of the clinical manifestations associated with DF, DHF and DSS in humans (C) Grades of disease severity in

DHF/DSS patients (adapted from Whitehead et al., 2007)

1.6 DENGUE PATHOGENESIS

Severe DEN may result from the interplay of multiple factors contributed by both host and virus In the host, immune status, age, sex and genetic factors may contribute to disease severity Virus genetics contributing to its inherent virulence within the host may also influence the disease outcome Typically, classical DF is self-limiting and seldom life-threatening; risk factors are certainly involved in exacerbating disease during DEN infection, amounting to a severe disease state (Tan and Alonso, 2009)

1.6.1 IMMUNE RESPONSE TO DENGUE VIRUS INFECTION

1.6.1.1 PROTECTIVE ROLE OF IMMUNE RESPONSE AGAINST DENGUE VIRUS INFECTION

In contrast to the plethora of studies that focus on elucidating the immunopathogenesis of DEN infection, considerably less effort has been directed at determining what constitutes a protective anti-DEN response Interferon (IFN)-dependent immunity, including the IFN-α, IFN-β and IFN-γ, plays important non-overlapping roles in the host innate response against DEN infection by limiting initial DENV replication as well as subsequent viral spread (Johnson and Roehrig, 1999;

Diamond et al., 2000; Shresta et al., 2004) While neutralizing antibodies have also

proven to be crucial in protection against a re-infection of the same DEN serotype, the contribution of T cells to protection remains largely unexplored until recently where CD8+ T cells were found to be implicated in a protective function against DEN infection via the reduction of viral load in the IFN-α/β and IFN-γ receptor knockout

mouse model (Yauch et al., 2009) The protective role of cell mediated immunity

dependent on CD4+ and CD8+ cells without the induction of neutralizing antibodies

was similarly established in a mouse encephalitis model (Gil et al., 2009a; Gil et al.,

or mild/no disease or complete protection – was first proposed in 1970 by Halstead The concept of antibody-dependent enhancement (ADE) was formulated to explain

the severe manifestations of DHF/DSS occurring in Thai children (Halstead et al.,

1970) that suffered from secondary DEN infection of a heterologous serotype The ADE hypothesis postulates that the antibodies raised against one DEN serotype cannot neutralize but instead, could enhance a secondary infection by another DEN serotype Following that seminal study, extensive epidemiological evidence has subsequently supported the hypothesis that pre-existing, sub-neutralising levels of cross-reactive antibodies can be a major risk factor for developing DHF/DSS in both infants and adults (Halstead, 2007) Babies less than 1 year old who acquired maternal anti-DEN antibodies have also been shown also to be susceptible to DHF/DSS

following primary DEN infection (Kilks et al., 1988; Simmons et al., 2007)

Nevertheless, a recent field study conducted in the Philippines does not support the ADE hypothesis: it was demonstrated that infants who developed DHF did not have significantly higher frequencies or levels of ADE activity as compared to

symptomatic infants without DHF (Libraty et al., 2009) These results suggest that

measurable ADE activity is common in infants where DEN is endemic and varied across all symptomatic DEN infected infants Taken together, no significant

association between ADE and incidence of severe DEN disease was reflected from this case study

Hypothetically, if cross-reactive antibodies posed as risk factors for DHF/DSS manifestation, Fcγ receptors would then represent another important class of ‘first

line’ receptors for DENV entry Enhancing antibodies have been shown largely in vitro, to increase the efficiency of virus attachment and internalization in target cells,

typically a member of the mononuclear phagocytic lineage (Gollins and Porterfield,

1986; Halstead, 2003) and recently in mature DCs (Boonnak et al., 2008) Enrichment

of viral attachment to the cell surface in the presence of antibodies attributes to the extrinsic antibody dependent enhancement (exADE) phenomenon (Fig 1.3)

Fcγ receptor signalling in macrophages and mature DCs triggered by intrinsic ADE (inADE) may suppress the intracellular antiviral mechanisms, such as the release of free radicals and antiviral cytokines, and enhance cellular infectivity and viral output

(Boonnak et al., 2008; Chareonsirisuthigul et al., 2007) Besides the influence of

ADE on the intensity of cellular infection, it is also linked to heightened production of

potent vasoactive cytokines and chemokines in mast cells (Brown et al., 2006), and induce mast cell apoptosis (Brown et al., 2009)

The enhancement phenomenon is also reflected in the T cell response during

secondary infections (Mongkolsapaya et al., 2003; Mongkolsapaya et al., 2006;

Rothman and Ennis, 1999) Increased DEN viral antigen presentation to T lymphocytes during secondary infection causes rapid activation of CD4+ and CD8+memory T cells which induces the release of high levels of pro-inflammatory

cytokines (Mongkolsapaya et al., 2006; Rothman and Ennis, 1999) Such responses,

observed in secondary infections, are further amplified by DEN-specific T cells that trigger the maturation of DEN-infected DCs, leading to increased production of CXCL9, 10 and 11

Figure 1.3: A diagrammatic view of the antibody dependent enhancement (ADE) phenomenon of DENV infection The presence of heterotropic non-neutralizing antibody

does not neutralize the virus but instead, the antibody-virus complex attaches to the Fcγ receptor and enables enhanced entry of the virus into target cells (often a monocytic cell type) Fcγ receptor signaling also induces the activation of a signaling cascade that suppresses intracellular anti-viral mechanisms, allowing for enhanced cellular infectivity and viral output

(adapted from Whitehead et al., 2007)

These chemokines promote chemotaxis and allow the enrichment of memory and

activated T cells at the site of infection (Dejnirattisai et al., 2008)

High levels of pathological T cell responses are also postulated to be driven by reactive memory T cells (mainly NS3 specific) The ‘original antigenic sin’ model posits that low affinity T cells generated during a primary DEN infection are expanded selectively during a secondary heterologous infection, preceding the generation of higher avidity nạve T cells for the new DEN serotype (Mongkolsapaya

cross-et al., 2003; Mongkolsapaya cross-et al., 2006) The result of this anamnestic T cell

response is an apoptotic phenotype with a suboptimal capacity for elimination of

DENV infected cells, but enhanced cytokine production (Mongkolsapaya et al., 2003; Mongkolsapaya et al., 2006; Mangada and Rothman, 2005; Imrie et al., 2007) High

amplitude cytokine release can act synergistically to further reinforce T cell response, ultimately creating a cytokine storm This wave of cytokine action is believed to enhance vascular permeability, contributing to the pathogenesis of DHF/DSS

(Chaturvedi et al., 2000; Chaturvedi et al., 2007; Basu and Chaturvedi, 2008)

Consistently, elevated levels of pro-inflammatory cytokines IFN-γ, TNF-α, IL-10

have been reported in sera of DHF/DSS patients (Chaturvedi et al., 2000; Chaturvedi

et al., 2007)

A diagrammatic view of the enhancement phenomenon induced by the presence of cross-reactive anti-DEN antibodies and T cells is presented in Fig 1.4

Figure 1.4: Schematic representation of the immunopathogenesis of severe DEN disease

mediated by cross reactive anti-DEN antibodies and T cells (adapted from Webster et al.,

2009)

1.6.1.3 NON-ENHANCING MECHANISMS AS RISK FACTORS FOR DHF/DSS:

AUTOIMMUNE AND COMPLEMENT-MEDIATED PATHOGENESIS

The immune system has also been implicated in the pathogenesis of DHF/DSS under

‘non-enhancing’ circumstances Anti-DEN NS1 antibodies have been documented to

cross react with human platelets and endothelial cells (Lin et al., 2006) Autoimmune

responses against the cross-reactive components of DENV can induce platelet lysis and nitric oxide (NO) - mediated apoptosis of endothelial cells, contributing to

thrombocytopenia and vascular damage (Oishi et al., 2007) Proteomic analysis has

recently revealed several candidate autoantigens on endothelial cells recognized by anti-DEN NS1 antibodies bearing sequence homology with NS1 at amino acid resides

311-330 (Cheng et al., 2009)

DEN NS1 antigen and antibodies against NS1 may also activate complement and

trigger plasma leakage (Avirutnan et al., 2006) Furthermore, NS1 has been found to

interact with the Complement inhibitory factor - clusterin (Clu) Formation of the NS1/Clu complex does not contribute to complement activation, but instead, interferes with the inhibition of activated complement system and may contribute to

vascular leakage (Kurosu et al., 2007)

Table 1.1: Proposed mechanisms involving the host immune system in mediating severe DEN disease in humans

Hypothesis Mediators Mechanism Consequence

dependent enhancement

Antibody-Cross-reactive antibodies

Enhances virus entry into target cells via Fcγ receptor

Enhanced cellular infectivity and viral output via Fcγ signalling

Increased viral burden augments disease severity

Cross-reactive memory T cells

Selective expansion of low affinity T cells (from primary DEN infection) during a

secondary heterologous infection

Cross reactive memory CD8+

T cells produce higher levels of pro-inflammatory cytokines leading to enhanced vascular permeability

mediated pathogenesis

Autoimmune-Anti- NS1 antibodies

Cross react with human platelets and endothelial cells

Induce platelet lysis and nitric oxide mediated apoptosis

of endothelial cells, contributing to thrombocytopenia and vascular damage

Complement-soluble (s)NS1 antigen; Anti-NS1 antibodies

Complement can

be activated by sNS1 alone and enhanced further

by anti-NS1 antibodies

Induces cell lysis and vascular permeability at sites

of complement activation

1.6.2 HOST GENETIC FACTORS

Genetically determined host susceptibility, hypothesized to predispose an individual

to DHF/DSS has become an emerging area of research Major host susceptibility genes implicated in the development of severe DEN include the human leucocyte

antigens (HLA) class I loci (Chiewsilp et al., 1981; Stephens et al., 2002; Mestre et al., 2004; Lan et al., 2008), the HLA class III loci and the non-HLA genes (Chaturvedi et al., 2006) Polymorphisms in the genes of tumor necrosis factor (TNF)-

Fernández-α, promoter of the CD209 gene (Sakuntabhai et al., 2005), receptors of vitamin D and Fcγ IIA (Chaturvedi et al., 2006), transforming growth factor (TGF)-β (Chen et al., 2009) and CTLA-4 (Chen et al., 2009) as well as transporters associated with antigen

presentation (TAP) and human platelet antigen (HPA) have been associated with increased risk for developing DHF in different populations

Deficiency in G6PD is reported to contribute to increased replication of DEN in monocytes, hence predisposing the African population - with high prevalence of

G6PD deficiency - to development of severe DEN disease (Chao et al., 2008)

Restriction of ADE by complement component C1q also suggests that individuals with C1q deficiency are more prone to developing DHF under ADE conditions

(Mehlhop et al., 2007) Partial deficiency of C4, another member of the complement

system, may also be predisposed to atypical ocular complications in DEN infection

Cuban and Mexican populations, specifically in the DRB1*07, DRB1*04 and

DRB1*0901 alleles (Lan et al., 2008) This presents the first novel evidence that HLA

class II can control disease severity in DEN infection and contributes significantly to the hypothesis that genetic makeup can skew the outcome of disease in DEN infections

1.6.3 VIRUS VIRULENCE

Virus virulence has also been proposed to contribute to the occurrence of severe DEN disease The hypothesis that viral factors play a role in inherent DENV virulence was first proposed by Barnes and Rosen based on the DEN epidemic which occurred in Niue Island in the South Pacific region in 1972 (Barnes and Rosen, 1974) This epidemic was characterized by a high incidence of haemorrhagic manifestations and deaths Extensive epidemiological and serological investigation indicated that this epidemic was exclusively due to DENV2 and no sign of any DEN activity was evident 25 years prior to the epidemic (Rosen, 1977) This clearly indicated that the DHF/DSS manifestations in the 1972 Niue Island epidemic did not arise out of sequential infection with heterologous serotypes, but was the result of a primary DEN infection Hence, it is proven that prior DEN infection was not a prerequisite to the pathogenesis of DHF/ DSS This notion is strengthened by the field data obtained in

Tonga by Gubler and coworkers in 1974 and 1975 (Gubler et al., 1978) Contrary to

the proposed ADE hypothesis, the DENV1 outbreak in the Kingdom of Tonga in

1975, which followed a DENV2 epidemic the year before was largely the result of a

primary instead of a secondary infection (Gubler et al., 1978) As the severity of the

1975 outbreak did not correlate with prior immune status, these authors proposed that the DENV1 strain responsible for the severe outbreak in 1975 must be inherently

more virulent that the DENV2 strain that caused the DEN outbreak during the

preceding year (Gubler et al., 1978) A further phylogenetic analysis of the DENV2

responsible for the outbreaks in the South Pacific from 1971 to 1974 demonstrated that the DENV2 viruses that initially appeared in 1971 in Tahiti and Fiji had a distinctly virulent character that subsequently caused epidemics in American Samoa, New Caledonia and Niue Island in 1972 However, upon reaching Tonga in 1973, it was found to be considerably attenuated, with a near-silent transmission for over a year Sequencing of these viruses revealed that the DENV2 strains circulating in Tonga were genetically unique, clustered within a single clade and contained substitutions in the prM, NS2A and NS4A that correlated with an attenuated

phenotype (Steel et al., 2010)

Subsequently, accumulating evidence has supported the role for viral genetics in DEN pathogenesis and undoubtedly, the evolution of DENV has had a great impact on its virulence in humans and on the epidemiology of DEN disease (Rico-Hesse, 2003) One of the many field studies that reported a viral genotype association with DEN epidemic emerged from Peurto Rico The DENV2 when first introduced into Peurto Rico in 1982, did not cause an outbreak until 1994 when it was associated with a

clade change genetically (Bennett et al., 2003) Similar observations were made in Sri Lanka (Messer et al., 2003), Singapore (Lee et al., 2010) and Vietnam (Vu et al.,

2010) Further evidence based on molecular epidemiological studies in conjunction with clinical data demonstrated that risk of DHF by DEN of the Asian genotype is greater in children than adults, with a worse outcome in younger children In contrast, American strains of DEN largely produce milder disease in the adult population

(Pandey et al., 2000; Watts et al., 1999) Full length sequencing of the Asian and

American genotypes revealed various nucleotide differences, particularly located in

the E protein coding ORF and in the 5’ and 3’ UTR (Leitmeyer et al., 1999; Kurane,

2007)

Subsequent in vitro studies have demonstrated phenotypic distinctions in relation to

plaque size and virus titers that paralleled virulence (Kurane, 2007) Specific mutations associated with changes in virulence have been identified using infectious

clones (Gualano et al., 1998; Kinney et al., 1997), chimeric constructs (Bray et al., 1998) and by directly comparing wild-type and attenuated DENV strains (Kinney et al., 1997; Sanchez and Ruiz, 1996) with the use of specific mouse models While

animal studies have provided proof-of-principle that viral sequence can contribute to

its virulence in vivo and various molecular determinants of DEN virulence have been

identified progressively (Yauch and Shresta, 2008), whether or not they contribute to virulence in humans remains to be determined

1.7 DENGUE PREVENTION

With the odds of severe DEN on the rise, there is an urgent need to control DEN spread Moreover, since many endemic areas are also popular travel destinations, DEN threat may substantially impact tourism-related businesses and directly affect the countries’ economy Clearly, there is an urgent need for a safe and effective vaccine against DEN Vector control and vaccination represent the two major approaches to curb DEN spread

1.7.1 VECTOR CONTROL

Vector control is one of the means to reduce the incidence of diseases such as DEN, but to date, the government of developing countries have failed to implement sustainable effective vector control programs Moreover, the emergence of insecticide-resistance and the potential environmental issues associated with some powerful insecticides have hampered the deployment of such approaches However,

lectins isolated from Myracrodruon urundeuva bark and heartwood were recently shown to display some larvicidal activity on Aedes aegypti (Ra et al., 2008) and might

well represent a novel environmentally friendly insecticide to fight DEN

The development of mosquitocidal vaccines may be a viable alternative to existing

vector control measures (Billingsley et al., 2008) With concerted efforts to decipher the genome of the Aedes aegypti, it is now possible to genetically manipulate

mosquitoes to impair their ability to be infected by DEN and hence limit the viruses’

transmissibility This has also emerged as a new vector control strategy (Speranca et al., 2007) However, in addition to the technical challenges, ecology- and public

health–related issues still need to be addressed before a pathogen-refractory transgenic mosquito can be released in nature

1.7.2 DENGUE VIRUS VACCINES

The development of a safe and effective vaccine remains the cornerstone of DEN prevention The characteristics of an ideal DEN vaccine have been summarized in table 1.2 The various DEN vaccine strategies currently being explored and developed

have recently been reviewed in detail elsewhere (Webster et al., 2009; Tan and

Alonso, 2009; Guy and Almond, 2008; Wilder-Smith and Deen, 2008) They include live attenuated, chimeric, subunit and DNA vaccines (summarized in Fig 1.5)