Cover Page

The handle http://hdl.handle.net/1887/22114 holds various files of this Leiden University dissertation

Author: Jiang, Xiaohong Title: Construction characterization and application of Flavivirus infectious clones Issue Date: 2013-11-06 Construction, Characterization and Application of Flavivirus Infectious Clones

Xiaohong Jiang ISBN: 978-90-8891-733-2

Cover design: Xiaohong Jiang, Joe Zhou and Kerry Zhang (K&A Inc.)

The image on the front cover is a schematic representation of the landmarks in Shanghai, China, which is sketched out by the use of predicted RNA structural elements in the 3’-UTR of flaviviruses as illustrated in the thesis.

Layout and printing: Proefschriftmaken.nl || Uitgeverij BOXPress Published by: Uitgeverij BOXPress, ’s-Hertogenbosch Construction, Characterization and Application of Flavivirus Infectious Clones

Proefschrift

ter verkrijging van de graad van Doctor aan de Universiteit Leiden, op gezag van Rector Magnificus prof. mr. C.J.J.M. Stolker, volgens besluit van het College voor Promoties te verdedigen op woensdag 6 november 2013

klokke 11:15 uur

door

Xiaohong Jiang

geboren te Shanghai, China in 1983 PROMOTIECOMMISSIE

Promotor: Prof. Dr. W.J.M. Spaan Copromotor: Dr. P.J. Bredenbeek Overige leden: Prof. Dr. E.J. Snijder Prof. Dr. R.C. Hoeben Dr. D. Franco (Katholieke Universiteit Leuven) Prof. Dr. J.H. Neyts (Katholieke Universiteit Leuven)

献给我挚爱的父母

5 6 CONTENTS

Chapter 1 General Introduction 9

Chapter 2 Yellow fever 17D-vectored vaccines expressing Lassa virus GP1 and GP2 glycoproteins provide protection against fatal disease in guinea pigs 31

Chapter 3 Molecular and immunological characterization of a DNA launched yellow fever virus 17D infectious clone 55

Chapter 4 An infectious Modoc virus cDNA as a tool to study conserved 3’-UTR RNA elements in flaviviruses with no known vector 79

Chapter 5 Mutagenesis and biochemical analysis of the XRN-1 stalling signal for subgenomic flavivirus (sf)RNA production 105

Chapter 6 Trans-complementation of ellowy fever virus cytopathogenicity by cells expressing subgenomic flavivirus (sf)RNA 127

Chapter 7 Epilogue 145

Summary 159

Samenvatting 163

Curriculum vitae 167

7 8 Chapter 1

Introduction

9 Chapter 1

Prologue: Virus

Viruses, derived from the Greek word ‘poison’, are a group of obligatory parasites that hijack the hosts’ resources to produce their progeny. Indeed, viruses are notorious for the threats they pose to human beings, as the name suggests. From smallpox, one of human kind’s greatest scourges, to HIV, an enemy that tactfully disarms the human immune system; from the 1918 influenza pandemic to the 2003 SARS outbreak; viruses have often been associated with disease and death. This reputation is, in a sense, well deserved; however, the advent of genomics era has started to enrich our perception of these ‘killers’. The omnipresence and abundance of viruses turn out to go far beyond our wildest guess. In fact, viruses are so pervasive that they essentially infect any form of life, and even parasitizing other viruses [1]. Moreover, viruses have been proposed to play major roles in the global ecosystem [2], have influenced and are still shaping the evolution of cellular life forms, with elements of viral origin embedded in host genomes. The extent of their infiltration into the human genome is surprising yet undeniable [3;4]. We are, for better or worse, living in a virosphere with viruses lurking within us. Virus particles, or virions, consist of nucleic acids, the genetic material, as well as a protein coat, known as the capsid; in some cases, this nucleocapsid is further wrapped by an envelope of lipids. Different from ribosome-encoding organisms, i.e., bacteria, archaea, and eukarya, which all have DNA as their genetic material, viruses (defined as capsid-encoding organisms [5]) can also use RNA molecules to convey their genetic information. The genomic RNA for some viruses like reoviruses is double-stranded, while most of the RNA viruses possess single-stranded RNAs of either positive or negative polarity. Viruses with genomic RNA complementary to mRNA are denoted negative-strand RNA viruses, whereas those with genomic RNA of mRNA polarity are classified as positive-strand or plus-strand RNA (+ssRNA) viruses. The topics covered in this thesis are revolving around a group of +ssRNA viruses, the flaviviruses.

Flavivirus: Phylogeny

Flaviviruses form a genus within the Flaviviridae family that also includes the genera Pestivirus and Hepacivirus, as well as the newly proposed Pegivirus genus [6]. The Flavivirus genus comprises more than 70 viruses, which, based on phylogenetic analysis, can be grouped into three clusters of related viruses that are in general distinguished by their route of transmission: (i) mosquito-borne, (ii) tick-borne, and (iii) no known vector (NKV) flaviviruses [7]. The first two clusters of flaviviruses are transmitted by arthropods,

10 Introduction mosquitoes and ticks, respectively. They are exemplified by important human pathogens like dengue virus (DENV), Japanese encephalitis virus (JEV), West Nile virus (WNV), yellow fever virus (YFV) and tick-borne encephalitis virus (TBEV). The third cluster comprises a less-well studied group of viruses for which no arthropod vector has been implicated in transmission. These viruses have been exclusively isolated from bats (such as, Montana myotis leukoencephalitis virus [MMLV] [8] and Rio Bravo virus [RBV] [9]) and rodents (for example, Modoc virus [MODV] [10] and Apoi virus [APOIV]). In addition to the viruses assigned to any of the above-mentioned clusters within the Flavivirus genus, there are a few viruses that are currently considered tentative species of this genus. This group of viruses, such as cell fusing agent virus (CFAV), Kamiti River virus (KRV) and Culex flavivirus (CxFV), have all been exclusively isolated from mosquitoes or insect cell lines [11-13]. There is no evidence that these viruses are able to infect a vertebrate host and therefore they are referred to as insect-specific flaviviruses.

Flavivirus: General Features

Flaviviruses share similarities in virion morphology, genome organization and replication strategy. The enveloped viral particles are composed of a lipid bilayer surrounding a nucleocapsid that consists of a single-stranded, positive-sense genome RNA complexed with multiple copies of a basic capsid (C) protein [14]. 7 The genomic RNA is approximately 11 kb in length with a 5’ type I cap, m GpppAmpN2, while lacking a 3’ poly(A) tail. As other positive-strand RNA viruses, the flavivirus genome plays various roles in the viral life cycle: (i) directing the synthesis of viral proteins by ribosomes, (ii) serving as a template for the production of negative-strand RNA, and (iii) as the genetic material to be encapsidated into progeny virions. The genome is translated into a large polyprotein that is co- and post-translationally processed into functional viral proteins by cellular and viral proteases (Fig.1). The N-terminal third of the polyprotein yields the viral structural proteins: core (C), membrane (prM/M), and envelope (E) that participate in virus entry and assembly. The C-terminal two-thirds of the polyprotein is cleaved into at least seven nonstructural (NS) proteins, namely NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5, which have been implicated directly or indirectly in viral RNA replication, evasion of the innate immune response, as well as virus assembly [14-21]. NS1 is a multifunctional glycoprotein that is localized to the replication complexes, expressed on the plasma membrane of infected cells and secreted into the extracellular milieu. The intracellular form of NS1 was proposed to interact with NS4A and NS4B and plays an important role in regulating negative-strand RNA synthesis, whereas its extracellular

11 Chapter 1

Figure 1. Flavivirus Genome Organization The flavivirus genome encompasses a single open reading frame that is flanked by ′5 and 3′ untranslated regions (UTRs), within which high-ordered RNA structures are formed. Stem loop (SL) structures are indicated with dots representing the loop and short lines for stems. Location of the conserved sequences (CS) and the upstream AUG region (UAR) that are involved in cyclization of the genomic RNA are indicated. The polyprotein is co- and post- translationally cleaved into three structural proteins (in green) and seven nonstructural proteins (in red). Putative functions of these proteins during infection are described. aa, amino acids; C, capsid protein; E, envelope; M, membrane; RdRp, RNA-dependent RNA-polymerase (Adapted from Cell Host & Microbe, 2009 Apr 23;5(4):318-28, Fernandez-Garcia MD, Mazzon M, Jacobs M, Amara A., reprinted with permission)

form was implicated in immune evasion [22-27]. NS2A is a small hydrophobic protein that is essential for virus replication and assembly and may also modulate the host innate immune response [17;18;28;29]. NS2B, another small membrane-associated protein, is a cofactor for the NS2B-NS3 serine protease that mediates cleavage of NS2A/NS2B, NS2B/NS3, NS3/NS4A, NS4A/NS5 junction, and generates the C-termini of mature C and NS4A [14]. In addition to its protease activity, NS3 also contains RNA helicase, nucleotide triphosphatase and RNA triphosphatase activities [30-33]. Both NS4A and NS4B are small hydrophobic proteins that co-localize to ER-derived membrane structures presumed to be the site of replication [34]. NS5, a large and highly conserved flaviviral protein, functions as an RNA-dependent RNA polymerase and contains methyltransferase activity that is required for the formation of a 5’ type I cap as well as internal adenosine 2’-O methylation of the viral genome [14;35-37]. Flaviviruses enter the cell by receptor-mediated endocytosis via clathrin-coated pits [38;39]. The viral particles are subsequently trafficked to prelysosomal endocytic compartments where low pH induces a conformational change in the envelope protein, triggering the fusion between the viral and the endosomal membranes [38;40-43]. The nucleocapsid is thus released into the cytoplasm, where membrane-associated translation-

12 Introduction replication complexes are formed [34;44;45]. Newly produced nucleocapsids further acquire the envelope upon budding into the endoplasmatic reticulum (ER) lumen. The immature viral particles are then transported to the Golgi apparatus, where E/prM heterodimers are dissociated and the formation of homodimers allows the cleavage of prM by the Golgi- resident protease furin [34;46-50]. Mature virions are released via the exocytosis pathway for the next round of infection [14]. (Fig. 2)

Flavivirus: 3’-UTR and sfRNA

The flavivirus genome contains a single open reading frame (ORF) flanked by 5’ and3’ untranslated regions (UTRs). The sequence of the 5’-UTR is not well conserved among different flaviviruses, although common secondary structures have been identified within this region that influence translation and function in viral replication (reviewed in [51]). The 3’-UTR varies in length from about 350 to 800 nucleotides. This heterogeneity in length is mainly due to the variation in the so-called variable region (VR), which is located

Figure 2. Flavivirus Life Cycle (1) receptor-mediated endocytosis; (2) fusion of the virus envelop with the endosomal membrane induced by the acidification (red) of the endosomes; (3) release of the genome into the cytoplasm; (4) topology of the viral proteins at the endoplasmic reticulum (ER) membrane, polyprotein cleavage mediated by host signalases and the viral serine protease NS3 are indicated with black and blue arrows, respectively; (5) assembly of membrane associated replication complex; (6) assembly and budding of the progeny virions into the lumen of the ER; (7) virion maturation during transportation through the trans-Golgi network; (8) virion release via exocytosis (Adapted from Cell Host & Microbe, 2009 Apr 23;5(4):318-28, Fernandez-Garcia MD, Mazzon M, Jacobs M, Amara A., reprinted with permission)

13 Chapter 1 proximal to the translation termination codon and characterized by extensive sequence duplications and deletions, even among strains of the same virus, and by an apparent lack of sequence conservation. The high variability of this region among different natural strains suggests it is non-essential for viral replication. Indeed in the context of infectious cDNA clones, the whole VR can be deleted without significantly compromising viral viability or virulence in either cell culture or mice [52-56]. In contrast, the distal part of the 3’-UTR exhibits more sequence and/or structure similarities among flaviviruses, with even more conservation within each cluster. This “core element” of the 3’-UTR contains RNA motifs and structures that have been shown to be essential for viral translation, replication, and possibly assembly [56-63]. A universal feature that has been identified in every flavivirus studied thus far is the presence of a large, stable stem-loop structure formed by the last 90 to 120 nts of the genome (3’ SL). Sequence conservation within this structure, on the other hand, is only restricted to the 3’ terminal dinucleotide 5’-CU-3’ and the pentanucleotide (PN) motif 5’-CACAG-3’ in the top loop of the 3’ SL in arthropod-borne flaviviruses (reviewed in [51]). This PN motif is also rather conserved among the NKV flaviviruses, albeit with limited sequence variability [64]. For tick-borne flaviviruses, the sequence of the bottom stem of the 3’ SL is well conserved and was proposed to be involved in genome “cyclization” with a complementary sequence present in the 5’-UTR. This cyclization is generally believed to be an essential prerequisite for viral replication [14;65]. Conserved sequence motifs for genomic RNA “cyclization” have also been reported for the mosquito-borne flaviviruses [14;65;66], although in these viruses the 3’ cyclization sequence (3’CS) is located upstream of the 3’ SL. The 3’CS forms a long- range RNA-RNA interaction with a complementary conserved sequence (5’ CS) near the 5’ end of the genome, downstream of the translation initiation codon in the capsid gene [66] that is critical for viral RNA synthesis [52;67-72]. Recent studies have identified an additional long-range RNA-RNA interaction that also plays a role in promoting genome cyclization. This interaction involves complementary sequences at the 5’ end, located immediatelyu pstream of the AUG start codon region (UAR), and at the 3’ end within the bottom part of the 3’ SL (3’ UAR) [72]. This pair of complementary sequences has been shown to be important for viral replication [71;73-77]. In dengue virus and West Nile virus yet a third interaction important for genome circularization and RNA replication was recently identified between nucleotides downstream of the AUG region (5’ DAR) and nucleotides downstream of the 3’CS (3’ DAR) [78-80]. Similar RNA elements involved in long-distance RNA interaction have also been proposed for NKV flaviviruses [81]. Upstream of the cyclization sequence, another stretch of conserved nucleotides approximately 24 nucleotides in length was identified among mosquito-borne flaviviruses, designated as CS2. YFV contains only one copy of CS2 but this motif is duplicated (RCS2) in members of the JEV and DENV subgroups [66]. Deletion of CS2 has little effect on viral

14 Introduction

RNA synthesis but appears to affect cytopathogenicity of at least YFV and DENV, as mutants lacking CS2 form turbid plaques [52;67]. Dengue viruses lacking this sequence are attenuated in rhesus monkeys, suggesting its role in pathogenesis [67]. A region encompassing CS2 within the 3’-UTR of mosquito-borne flaviviruses was predicted to fold into a dumbbell- like structure from which a loop is to be involved in the formation of an RNA pseudoknot with downstream sequences [82]. A similar dumbbell-like structure has also been predicted to be formed in the 3’-UTR of NKV flaviviruses that includes a conserved sequence motif homologous to CS2 [81], whereas the CS2 containing structure appears to be absent in tick- borne flaviviruses. The accumulation of a small RNA species co-linear with the distal part of the viral 3’-UTR was recently discovered in cells infected with arthropod-borne flaviviruses and this feature was later found to be conserved among the whole Flavivirus genus [83-88]. This small flavivirus (sf) RNA results from incomplete degradation of the viral genomic RNA by the host 5’>3’ exoribonuclease, XRN-1 [89], which is the main cytoplasmic ribonuclease associated with 5’ to 3’ mRNA decay that takes place in cytoplasmic processing bodies [90] (reviewed in [91-93]). WNV sfRNA was demonstrated to be associated with viral pathogenicity and contribute to viral evasion of host innate immune response [89;94]. Recent studies proposed sfRNA as a repressor for RNA interference in both mammalian and insect cells [95], and as an antagonist for XRN-1 thus perturbing cellular mRNA stability [96]. A pseudoknot forming stem-loop structure located upstream of the CS2-containing dumbbell-like structures in the mosquito-borne flavivirus 3’-UTR was proposed to be the signal for the stalling of XRN-1 [87;97]. Similar structures were also predicted to be present in the 3’-UTR of tick-borne, NKV, as well as insect specific flaviviruses [88;98], although the molecular basis underlying XRN-1 stalling still remains to be elucidated.

Yellow Fever Virus

Yellow fever virus is the prototype virus of the Flavivirus genus. In fact, the genus name flavi- is derived from the Latin word flavus, meaning “yellow”. It is postulated that the virus originated in Africa and was subsequently introduced from the Old World into the Americas during the slave trade period in the 16th century (reviewed in [99]). In 1900, an American commission headed by Walter Reed demonstrated that yellow fever was caused by a filterable agent that can be transmitted to humans by mosquitoes [100]. In 1927, Mahaffy and Bauer first isolated the virus by inoculation of a rhesus monkey with the blood of an YFV-infected Ghanian male named Asibi [101].

15 Chapter 1

Yellow fever virus is endemic in tropical regions of Africa and South America, where the enzootic transmission cycle involves hematophagous mosquitoes (primarily Aedes or Haemagogus spp.) and non-human primates. Occasional transmission to human beings who enter the tropical rainforests leads to sylvatic (or jungle) yellow fever. Other types of transmission include: intermediate yellow fever caused by semi-domestic mosquitoes (that breed in the wild and around households), leading to small-scale epidemics, and urban yellow fever, large epidemics that occur when infected humans introduce the virus into densely populated areas, where a large number of non-immune people is susceptible to infection mediated by domestic mosquitoes [102]. With over 900 million people at risk, there are an estimated 200,000 cases of yellow fever worldwide each year, causing 30,000 deaths [102]. Small numbers of imported cases are also reported in countries outside the endemic region [102]. YFV infection presents a broad clinical spectrum varying from mild, non-specific febrile illness to a fulminating, sometimes fatal hemorrhagic fever, with a biphasic pattern. Once contracted, the virus incubates in the body for 3 to 6 days, followed by the first, “acute” phase with symptoms like fever, muscle pain, prominent backache, headache, shivers, loss of appetite, and nausea or vomiting. During this phase, patients are viremic and infectious to mosquitoes. Most patients improve and their symptoms disappear after 3 to 4 days. However, 15% of the patients would enter a second, more toxic phase within 24 hours of the initial remission. High fever returns and several body systems are affected. The patient rapidly develops jaundice and complains of abdominal pain with vomiting. Bleeding can occur from the mouth, nose, eyes or stomach. Once this happens, blood appears in the vomit and faeces. Kidney function deteriorates. This is the so-called “period of intoxication”, and up to 50% of the patients who enter the toxic phase die within 10 to 14 days, while the rest recover without significant organ damage. (reviewed in [99;102;103])

Yellow Fever Vaccine

There is no cure for yellow fever; vaccination is the single most important measure for preventing the disease [102]. The YFV-17D vaccine was empirically developed in the 1930s by Theiler and Smith [104], for which Theiler was awarded the 1951 Nobel Prize in medicine. The YFV Asibi strain was attenuated by serial passages in various animal tissues, which resulted in 68 nucleotide and 32 amino acid changes [104;105]. Attenuation was clearly demonstrated by the lack of neurotropism, viscerotropism, and lethality upon intracerebral inoculation of rhesus monkeys and subcutaneous vaccination of naive human subjects [106], although the mechanism underlying attenuation is not fully understood. The 12 amino acid

16 Introduction changes clustered in the envelope gene were thought to alter receptor binding and cellular tropism, possibly affecting the virulence [107]. Since its development, the YFV-17D vaccine has been administered to over 540 million human beings globally, and has stood as a paradigm for a successful vaccine with a great record of both safety and efficacy [108]. Vaccination is well tolerated, with only rare cases of severe adverse reactions (reviewed in [102;109;110]). In 95% of the vaccinees, protective immunity develops within 10 days against all known wildtype strains [109], although the molecular determinants of its immunogenicity are yet to be defined. Vaccination with YFV-17D esultsr in a transient viremia that peaks at days 5-7 post immunization and subsequently dissipates [111]. Viral RNA is undetectable by day 14 in plasma. This restricted viral replication, however, is sufficient to induce robust and polyvalent adaptive immune responses including the production of cytotoxic T cells, a mixed T helper 1 (Th1) and Th2 cell profile and robust neutralizing antibodies that can persist up to 40 years after vaccination [109]. The production of neutralizing antibodies most likely accounts for the effectiveness of the vaccine, since passive immunization in monkeys [112] and mice [113] protects more than 90% of these animals against a challenge with wildtype virus. However, a role for cell-mediated immunity in generating an effective immune response has also been suggested, with both YFV specific effector and memory CD8+ T cell responses detected [114-117]. With the increasing knowledge of innate immunity, attention was also drawn to the interaction between YFV-17D and the innate immune system, as well as the importance of this interaction in determining the breadth, magnitude and duration of the adaptive immune responses. YFV-17D was found to infect human dendritic cells and activate multiple TLRs, priming DCs for the presentation of YFV antigens and coordinating the adaptive immunity [118;119]. Systems biology studies revealed YFV-17D vaccination induced integrated multilineage and polyfunctional immune responses that includes several effector arms of innate immunity, such as complement, the inflammasome, and the interferons [119;120]. In fact, the study of molecular signatures induced by YFV-17D prompted the development of systems vaccinology, i.e., using a vaccine chip to predict the immunogenicity and/or protective capacity of future vaccine candidates [110].

YFV-17D Vectored Recombinant Vaccines

In view of the great success of YFV-17D vaccine and the availability of a full-length cDNA clone [52], possibilities have been explored to utilize YFV-17D backbone as the vector for the construction of vaccine candidates against other diseases.

17 Chapter 1

The most promising YFV-17D based vaccine candidates are chimeric vaccines for related flaviviruses, such as DENV, WNV, and JEV, which are establishing in new habitats and infesting an ever increasing population [121-123]. Various social and environmental factors contribute to the re-emergence and spread of these arbo-flaviviruses that can cause diseases ranging from mild febrile illness to severe meningoencephalitis or hemorrhagic fever. Although vaccines are available to prevent JEV, the expensive and time-consuming vaccination regimen of the inactivated virus vaccine (INV) restricted its application mainly within the developed countries, while a live-attenuated vaccine (LAV) was only approved in a limited number of Asian countries, due to concerns over cell substrate certification. An inactivated WNV vaccine is also available but is only licensed for use in livestocks [124;125]. As for DENV, it is even more challenging to make a safe and effective vaccine, because low levels of cross-reactive immunity to each of the four individual serotypes of DENV have been shown to be a risk factor for developing severe forms of the disease. Therefore, a safe and efficacious DENV vaccine must be tetravalent, and simultaneously inducing strong immunity against all four serotypes (reviewed in [124]). Vaccine candidates have been made for JEV, WNV and DENV using the ChimeriVax platform, which is based on the finding that chimeric viruses encoding the PrM/E gene of other flaviviruses in the context of YFV-17D backbone can be produced without significantly compromising virus replication and production. The expressed prM and E of the donor virus will drive the efficient assembly and budding of an enveloped virion, in which the recombinant RNA is complexed with the YFV C protein. These chimeric viruses will trigger an immune response against the donor virus upon vaccination. IMOJEV(®), a Japanese encephalitis vaccine developed based on the Chimerivax platform has already been licensed for human use in Australia [126], and Chimerivax-WNV and tetravalent Chimerivax-DENV have also shown encouraging results in phase II clinical trials [127;128]. YFV-17D has also been successfully exploited as a vector for the expression of foreign T cell or B cell epitopes, as potential vaccine candidates against malaria [129;130] or as a potential therapeutic anticancer vaccine [131]. Moreover, recombinant YFV-17D viruses expressing heterologous antigens were proposed as bivalent vaccine candidates against Lassa virus [132;133] or as alternative vaccine vectors for malaria [134] and HIV [135;136]. In vitro studies demonstrated that these recombinant viruses were viable and could express both YFV and foreign proteins; besides, small-scale animal experiments showed promising results in terms of immunogenicity and protection. Nevertheless, the genetic stability of these YFV-17D recombinants has been an issue especially with larger inserts, a problem attributed to the limited packaging capacity of the icosahedral capsid [133;134;137].

18 Introduction

Scope and Outline of The Thesis

Full-length infectious clones for +ssRNA viruses have served as valuable tools in studying various aspects of virus biology with reverse genetics approaches. The construction of stable full-length clones for flaviviruses, however, was not always straightforward, as insert instability has been encountered during plasmid DNA propagation in bacterial hosts. Laborious in vitro ligation procedures were therefore used for the production ofDNA templates for the generation of infectious RNA transcripts for certain flaviviruses [138-140], until the instability problem was circumvented by cloning the cDNA into low copy number plasmid vectors [52;141]. The successful construction of infectious clones for Kunjin virus (KUNV) [142], West Nile virus (WNV) [143] and dengue virus (DENV) [144;145], as well as tick-borne encephalitis virus (TBEV) [146] have not only facilitated the research on molecular virology of flaviviruses, but also opened up new possibilities of attenuating these viruses for the development of vaccine candidates against these medically important flaviviruses. The availability of the YFV-17D infectious clone [52;138] had a more profound significance in the sense that it can serve as the backbone for recombinant or chimeric vaccines without raising much controversy over vector safety. In this thesis, an example of a recombinant vaccine based on YFV-17D clone was described in Chapter 2, in which sequences encoding Lassa virus (LASV) glycoproteins were inserted into the YFV-17D background. The recombinant viruses thus constructed could replicate efficiently in cell culture and grow to high titers. Both YFV proteins as well as LASV glycoproteins were expressed and properly released from the polyprotein. YFV/ LASV vaccination could induce specific cellular immune response to LASV, and could protect animals against lethal challenge. To explore the feasibility of incorporating the commendable features of a DNA vaccine with those of the live attenuated YFV-17D vaccine, a DNA-launched YFV-17D infectious clone was constructed by cloning the full-length YFV-17D cDNA in between the cytomegalovirus promoter and the hepatitis delta virus ribozyme. The DNA/LAV (live attenuated vaccine) layered approach could potentially simplify vaccine production, transportation and storage while retaining high immunogenicity characteristic of an LAV. The construction and characterization of the DNA-launched YFV-17D infectious clone was described in Chapter 3. To evaluate the potential of this novel DNA launched YFV-17D vaccine platform, mice were immunized with the construct in parallel with a couple of mutants, and immune responses induced were analyzed in detail. As mentioned above, inaddition to the arthropod-borne flaviviruses, there is a third cluster of flaviviruses that have been exclusively isolated from rodents or bats. Designated as NKV flaviviruses, this group of viruses is less-well studied, which is in part due to the

19 Chapter 1 lack of infectious clones. Chapter 4 describes the construction and characterization of the first NKV flavivirus infectious clone, which was further utilized to study the functional significance of RNA elements present in Modoc virus 3’-UTR, as a representative of the NKV flaviviruses. The differences/similarities between NKV flavivirus 3’-UTR and that of vector- borne flaviviruses were explored and their implications were discussed. A new common feature for flaviviruses, i.e., the production of a small flavivirus RNA (sfRNA) in infected cells or animals, was recently discovered. Attempts were made to unravel the sequence/structure requirements for the production of YFV sfRNA, as described in Chapter 5. Mutations were introduced into the XRN-1 stalling site in the context of YFV- 17D full-length clone, and the influence of these mutations on sfRNA production and viral cytopathogenicity were analyzed in detail. For a more comprehensive understanding of sfRNA function, mammalian cell lines that constitutively express sfRNA from YFV-17D and WNV were established and characterized, as described in Chapter 6. The ability of these cell lines to trans-complement sfRNA-deficient YFV mutants in inducing cytopathic effects was studied, and possible mechanism behind the trans-complementation is discussed. Chapter 7 is an epilogue, in which the findings presented in the previous chapters are summarized and discussed. Hurdles encountered in the development of YFV-17D based vaccine systems are summarized, and possible improvements and applications for the YFV-17D based vaccine platforms are proposed. Findings from the sfRNA expressing cell line approach are discussed in reference to previously published results regarding sfRNA function, and future directions are suggested for further solving the enigma surrounding this once ignored RNA species.

20 Introduction

Reference List

[1] La SB, Desnues C, Pagnier I, et al. The virophage as a unique parasite of the giant mimivirus. Nature 2008 Sep 4;455(7209):100-4.

[2] Suttle CA. Marine viruses--major players in the global ecosystem. Nat Rev Microbiol 2007 Oct;5(10):801- 12.

[3] Lander ES, Linton LM, Birren B, et al. Initial sequencing and analysis of the human genome. Nature 2001 Feb 15;409(6822):860-921.

[4] Horie M, Honda T, Suzuki Y, et al. Endogenous non-retroviral RNA virus elements in mammalian genomes. Nature 2010 Jan 7;463(7277):84-7.

[5] Raoult D, Forterre P. Redefining viruses: lessons from Mimivirus. Nat Rev Microbiol 2008 Apr;6(4):315-9.

[6] Stapleton JT, Foung S, Muerhoff AS, Bukh J, Simmonds P. The GB viruses: a review and proposed classification of GBV-A, GBV-C (HGV), and GBV-D in genus Pegivirus within the family Flaviviridae. J Gen Virol 2011 Feb;92(Pt 2):233-46.

[7] Kuno G, Chang GJ, Tsuchiya KR, Karabatsos N, Cropp CB. Phylogeny of the genus Flavivirus. J Virol 1998 Jan;72(1):73-83.

[8] Bell JF, Thomas LA. a new virus, “MML”, enzootic in bats (myotis lucifugus) of Montana. Am J Trop Med Hyg 1964 Jul;13:607-12.

[9] Constantine DG, WoodallDF. Latent infection of Rio Bravo virus in salivary glands of bats. Public Health Rep 1964 Dec;79:1033-9.

[10] Zarnke RL, Yuill TM. Modoc-like virus isolated from wild deer mice (Peromyscus maniculatus) in Alberta. J Wildl Dis 1985 Apr;21(2):94-9.

[11] Stollar V, Thomas VL. An agent in the Aedes aegypti cell line (Peleg) which causes fusion of Aedes albopictus cells. Virology 1975 Apr;64(2):367-77.

[12] Crabtree MB, Sang RC, Stollar V, Dunster LM, Miller BR. Genetic and phenotypic characterization of the newly described insect flavivirus, Kamiti River virus. Arch Virol 2003 Jun;148(6):1095-118.

[13] Hoshino K, Isawa H, Tsuda Y, et al. Genetic characterization of a new insect flavivirus isolated from Culex pipiens mosquito in Japan. Virology 2007 Mar 15;359(2):405-14.

[14] Lindenbach B, Rice CM. Flaviviridae: The Viruses and Their Replication. In: Knipe D.M, Howley P.M., editors. Fields virology. 5 ed. Philadelphia, Lippincott Williams and Wilkins, 2007: p. 991-1042.

[15] Guo JT, Hayashi J, Seeger C. West Nile virus inhibits the signal transduction pathway of alpha interferon. J Virol 2005 Feb;79(3):1343-50.

[16] Liu WJ, Sedlak PL, Kondratieva N, Khromykh AA. Complementation analysis of the flavivirus Kunjin NS3 and NS5 proteins defines the minimal regions essential for formation of a replication complex and shows a requirement of NS3 in cis for virus assembly. J Virol 2002 Nov;76(21):10766-75.

[17] Liu WJ, Chen HB, Khromykh AA. Molecular and functional analyses of Kunjin virus infectious cDNA clones demonstrate the essential roles for NS2A in virus assembly and for a nonconservative residue in NS3 in RNA replication. J Virol 2003 Jul;77(14):7804-13.

21 Chapter 1

[18] Liu WJ, Chen HB, Wang XJ, Huang H, Khromykh AA. Analysis of adaptive mutations in Kunjin virus replicon RNA reveals a novel role for the flavivirus nonstructural protein NS2A in inhibition of beta interferon promoter-driven transcription. J Virol 2004 Nov;78(22):12225-35.

[19] Liu WJ, Wang XJ, Mokhonov VV, Shi PY, Randall R, Khromykh AA. Inhibition of interferon signaling by the New York 99 strain and Kunjin subtype of West Nile virus involves blockage of STAT1 and STAT2 activation by nonstructural proteins. J Virol 2005 Feb;79(3):1934-42.

[20] Munoz-Jordan JL, Sanchez-Burgos GG, Laurent-Rolle M, Garcia-Sastre A. Inhibition of interferon signaling by dengue virus. Proc Natl Acad Sci U S A 2003 Nov 25;100(24):14333-8.

[21] Munoz-Jordan JL, Fredericksen BL. How flaviviruses activate and suppress the interferon response. Viruses 2010 Feb;2(2):676-91.

[22] Chung KM, Liszewski MK, Nybakken G, et al. West Nile virus nonstructural protein NS1 inhibits complement activation by binding the regulatory protein factor H. Proc Natl Acad Sci U S A 2006 Dec 12;103(50):19111- 6.

[23] Chung KM, Diamond MS. Defining the levels of secreted non-structural protein NS1 after West Nile virus infection in cell culture and mice. J Med Virol 2008 Mar;80(3):547-56.

[24] Lindenbach BD, Rice CM. trans-Complementation of yellow fever virus NS1 reveals a role in early RNA replication. J Virol 1997 Dec;71(12):9608-17.

[25] Lindenbach BD, Rice CM. Genetic interaction of flavivirus nonstructural proteins NS1 and NS4A asa determinant of replicase function. J Virol 1999 Jun;73(6):4611-21.

[26] Youn S, Cho H, Fremont DH, Diamond MS. A short N-terminal peptide motif on flavivirus nonstructural protein NS1 modulates cellular targeting and immune recognition. J Virol 2010 Sep;84(18):9516-32.

[27] Youn S, Li T, McCune BT, et al. Evidence for a genetic and physical interaction between nonstructural proteins NS1 and NS4B that modulates replication of West Nile virus. J Virol 2012 Jul;86(13):7360-71.

[28] Kummerer BM, Rice CM. Mutations in the yellow fever virus nonstructural protein NS2A selectively block production of infectious particles. J Virol 2002 May;76(10):4773-84.

[29] Leung JY, Pijlman GP, Kondratieva N, Hyde J, Mackenzie JM, Khromykh AA. Role of nonstructural protein NS2A in flavivirus assembly. J Virol 2008 May;82(10):4731-41.

[30] Mancini EJ, Assenberg R, Verma A, et al. Structure of the Murray Valley encephalitis virus RNA helicase at 1.9 Angstrom resolution. Protein Sci 2007 Oct;16(10):2294-300.

[31] Wengler G, Wengler G. The carboxy-terminal part of the NS 3 protein of the West Nile flavivirus can be isolated as a soluble protein after proteolytic cleavage and represents an RNA-stimulated NTPase. Virology 1991 Oct;184(2):707-15.

[32] Wengler G, Wengler G. The NS 3 nonstructural protein of flaviviruses contains an RNA triphosphatase activity. Virology 1993 Nov;197(1):265-73.

[33] Wu J, Bera AK, Kuhn RJ, Smith JL. Structure of the Flavivirus helicase: implications for catalytic activity, protein interactions, and proteolytic processing. J Virol 2005 Aug;79(16):10268-77.

[34] Welsch S, Miller S, Romero-Brey I, et al. Composition and three-dimensional architecture of the dengue virus replication and assembly sites. Cell Host Microbe 2009 Apr 23;5(4):365-75.

22 Introduction

[35] Dong H, Chang DC, Hua MH, et al. 2’-O methylation of internal adenosine by flavivirus NS5 methyltransferase. PLoS Pathog 2012;8(4):e1002642.

[36] Egloff MP, Benarroch D,Selisko B, Romette JL, Canard B. An RNA cap (nucleoside-2’-O-)-methyltransferase in the flavivirus RNA polymerase NS5: crystal structure and functional characterization. EMBO J 2002 Jun 3;21(11):2757-68.

[37] Ray D, Shah A, Tilgner M, et al. West Nile virus 5’-cap structure is formed by sequential guanine N-7 and ribose 2’-O methylations by nonstructural protein 5. J Virol 2006 Sep;80(17):8362-70.

[38] Chu JJ, Ng ML. Infectious entry of West Nile virus occurs through a clathrin-mediated endocytic pathway. J Virol 2004 Oct;78(19):10543-55.

[39] van der Schaar HM, Rust MJ, Chen C, et al. Dissecting the cell entry pathway of dengue virus by single- particle tracking in living cells. PLoS Pathog 2008 Dec;4(12):e1000244.

[40] Harrison SC. The pH sensor for flavivirus membrane fusion. J Cell Biol 2008 Oct 20;183(2):177-9.

[41] Randolph VB, Stollar V. Low pH-induced cell fusion in flavivirus-infected Aedes albopictus cell cultures. J Gen Virol 1990 Aug;71 ( Pt 8):1845-50.

[42] Smit JM, Moesker B, Rodenhuis-Zybert I, Wilschut J. Flavivirus cell entry and membrane fusion. Viruses 2011 Feb;3(2):160-71.

[43] van der Schaar HM, Rust MJ, Waarts BL, et al. Characterization of the early events in dengue virus cell entry by biochemical assays and single-virus tracking. J Virol 2007 Nov;81(21):12019-28.

[44] Gillespie LK, Hoenen A, Morgan G, Mackenzie JM. The endoplasmic reticulum provides the membrane platform for biogenesis of the flavivirus replication complex. J Virol 2010 Oct;84(20):10438-47.

[45] Miller S, Krijnse-Locker J. Modification of intracellular membrane structures for virus replication. Nat Rev Microbiol 2008 May;6(5):363-74.

[46] Heinz FX, Stiasny K, Puschner-Auer G, et al. Structural changes and functional control of the tick-borne encephalitis virus glycoprotein E by the heterodimeric association with protein prM. Virology 1994 Jan;198(1):109-17.

[47] Lorenz IC, Allison SL, Heinz FX, Helenius A. Folding and dimerization of tick-borne encephalitis virus envelope proteins prM and E in the endoplasmic reticulum. J Virol 2002 Jun;76(11):5480-91.

[48] Mackenzie JM, Westaway EG. Assembly and maturation of the flavivirus Kunjin virus appear to occur in the rough endoplasmic reticulum and along the secretory pathway, respectively. J Virol 2001 Nov;75(22):10787- 99.

[49] Yu IM, Zhang W, Holdaway HA, et al. Structure of the immature dengue virus at low pH primes proteolytic maturation. Science 2008 Mar 28;319(5871):1834-7.

[50] Yu IM, Holdaway HA, Chipman PR, Kuhn RJ, Rossmann MG, Chen J. Association of the pr peptides with dengue virus at acidic pH blocks membrane fusion. J Virol 2009 Dec;83(23):12101-7.

[51] Markoff L. 5’- and 3’-noncoding regions in flavivirus RNA. Adv Viruses R 2003;59:177-228.

[52] Bredenbeek PJ, Kooi EA, Lindenbach B, Huijkman N, Rice CM, Spaan WJ. A stable full-length yellow fever virus cDNA clone and the role of conserved RNA elements in flavivirus replication. J Gen Virol 2003 May;84(Pt 5):1261-8.

23 Chapter 1

[53] Bryant JE, Vasconcelos PF, Rijnbrand RC, Mutebi JP, Higgs S, Barrett AD. Size heterogeneity in the 3’ noncoding region of South American isolates of yellow fever virus. J Virol 2005 Mar;79(6):3807-21.

[54] Hoenninger VM, Rouha H, Orlinger KK, et al. Analysis of the effects of alterations in the tick-borne encephalitis virus 3’-noncoding region on translation and RNA replication using reporter replicons. Virology 2008 Aug 1;377(2):419-30.

[55] Khromykh AA, Westaway EG. Subgenomic replicons of the flavivirus Kunjin: construction and applications. J Virol 1997 Feb;71(2):1497-505.

[56] Mandl CW, Holzmann H, Meixner T, et al. Spontaneous and engineered deletions in the 3’ noncoding region of tick-borne encephalitis virus: construction of highly attenuated mutants of a flavivirus. J Virol 1998 Mar;72(3):2132-40.

[57] Mandl CW, Kunz C, Heinz FX. Presence of poly(A) in a flavivirus: significant differences between the 3’ noncoding regions of the genomic RNAs of tick-borne encephalitis virus strains. J Virol 1991 Aug;65(8):4070- 7.

[58] Wallner G, Mandl CW, Kunz C, Heinz FX. The flavivirus 3’-noncoding region: extensive size heterogeneity independent of evolutionary relationships among strains of tick-borne encephalitis virus. Virology 1995 Oct 20;213(1):169-78.

[59] Poidinger M, Hall RA, Mackenzie JS. Molecular characterization of the Japanese encephalitis serocomplex of the flavivirus genus. Virology 1996 Apr 15;218(2):417-21.

[60] Wang E, Weaver SC, Shope RE, Tesh RB, Watts DM, Barrett AD. Genetic variation in yellow fever virus: duplication in the 3’ noncoding region of strains from Africa. Virology 1996 Nov 15;225(2):274-81.

[61] Gritsun TS, Venugopal K, Zanotto PM, et al. Complete sequence of two tick-borne flaviviruses isolated from Siberia and the UK: analysis and significance of the 5’ and 3’-UTRs. Virus Res 1997 May;49(1):27-39.

[62] Proutski V, Gould EA, Holmes EC. Secondary structure of the 3’ untranslated region of flaviviruses: similarities and differences. Nucleic Acids Res 1997 Mar 15;25(6):1194-202.

[63] Rauscher S, Flamm C, Mandl CW, Heinz FX, Stadler PF. Secondary structure of the 3’-noncoding region of flavivirus genomes: comparative analysis of base pairing probabilities. RNA 1997 Jul;3(7):779-91.

[64] Charlier N, Leyssen P, Pleij CW, et al. Complete genome sequence of Montana Myotis leukoencephalitis virus, phylogenetic analysis and comparative study of the 3’ untranslated region of flaviviruses with no known vector. J Gen Virol 2002 Aug;83(Pt 8):1875-85.

[65] Kofler RM, Hoenninger VM, Thurner C, Mandl CW. Functional analysis of the tick-borne encephalitis virus cyclization elements indicates major differences between mosquito-borne and tick-borne flaviviruses. J Virol 2006 Apr;80(8):4099-113.

[66] Hahn CS, Hahn YS, Rice CM, et al. Conserved elements in the 3’ untranslated region of flavivirus RNAs and potential cyclization sequences. J Mol Biol 1987 Nov 5;198(1):33-41.

[67] Men R, Bray M, Clark D, Chanock RM, Lai CJ. Dengue type 4 virus mutants containing deletions in the 3’ noncoding region of the RNA genome: analysis of growth restriction in cell culture and altered viremia pattern and immunogenicity in rhesus monkeys. J Virol 1996 Jun;70(6):3930-7.

[68] Khromykh AA, Meka H, Guyatt KJ, Westaway EG. Essential role of cyclization sequences in flavivirus RNA replication. J Virol 2001 Jul;75(14):6719-28.

24 Introduction

[69] Lo MK, Tilgner M, Bernard KA, Shi PY. Functional analysis of mosquito-borne flavivirus conserved sequence elements within 3’ untranslated region of West Nile virus by use of a reporting replicon that differentiates between viral translation and RNA replication. J Virol 2003 Sep;77(18):10004-14.

[70] Corver J, Lenches E, Smith K, et al. Fine mapping of a cis-acting sequence element in yellow fever virus RNA that is required for RNA replication and cyclization. J Virol 2003 Feb;77(3):2265-70.

[71] Alvarez DE, De Lella Ezcurra AL, Fucito S, Gamarnik AV. Role of RNA structures present at the 3’UTR of dengue virus on translation, RNA synthesis, and viral replication. Virology 2005 Sep 1;339(2):200-12.

[72] Alvarez DE, Lodeiro MF, Luduena SJ, Pietrasanta LI, Gamarnik AV. Long-range RNA-RNA interactions circularize the dengue virus genome. J Virol 2005 Jun;79(11):6631-43.

[73] Thurner C, Witwer C, Hofacker IL, Stadler PF. Conserved RNA secondary structures in Flaviviridae genomes. J Gen Virol 2004 May;85(Pt 5):1113-24.

[74] Filomatori CV, Lodeiro MF, Alvarez DE, Samsa MM, Pietrasanta L, Gamarnik AV. A 5’ RNA element promotes dengue virus RNA synthesis on a circular genome. Genes Dev 2006 Aug 15;20(16):2238-49.

[75] Zhang B, Dong H, Stein DA, Iversen PL, Shi PY. West Nile virus genome cyclization and RNA replication require two pairs of long-distance RNA interactions. Virology 2008 Mar 30;373(1):1-13.

[76] Alvarez DE, Filomatori CV, Gamarnik AV. Functional analysis of dengue virus cyclization sequences located at the 5’ and 3’UTRs. Virology 2008 May 25;375(1):223-35.

[77] Song BH, Yun SI, Choi YJ, Kim JM, Lee CH, Lee YM. A complex RNA motif defined by three discontinuous 5-nucleotide-long strands is essential for Flavivirus RNA replication. RNA 2008 Sep;14(9):1791-813.

[78] Friebe P, Harris E. Interplay of RNA elements in the dengue virus 5’ and 3’ ends required for viral RNA replication. J Virol 2010 Jun;84(12):6103-18.

[79] Friebe P, Shi PY, Harris E. The 5’ and 3’ Downstream of AUG region (DAR) elements are required for mosquito-borne flavivirus RNA replication. J Virol 2010 Dec 1.

[80] Zhang B, Dong H, Ye H, Tilgner M, Shi PY. Genetic analysis of West Nile virus containing a complete 3’CSI RNA deletion. Virology 2010 Dec 20;408(2):138-45.

[81] Leyssen P, Charlier N, Lemey P, et al. Complete genome sequence, taxonomic assignment, and comparative analysis of the untranslated regions of the Modoc virus, a flavivirus with no known vector. Virology 2002 Feb 1;293(1):125-40.

[82] Olsthoorn RC, Bol JF. Sequence comparison and secondary structure analysis of the 3’ noncoding region of flavivirus genomes reveals multiple pseudoknots. RNA 2001 Oct;7(10):1370-7.

[83] Urosevic N, van MM, Mansfield JP, Mackenzie JS, Shellam GR. Molecular characterization of virus-specific RNA produced in the brains of flavivirus-susceptible and -resistant mice after challenge with Murray Valley encephalitis virus. J Gen Virol 1997 Jan;78 ( Pt 1):23-9.

[84] Lin KC, Chang HL, Chang RY. Accumulation of a 3’-terminal genome fragment in Japanese encephalitis virus-infected mammalian and mosquito cells. J Virol 2004 May;78(10):5133-8.

[85] Scherbik SV, Paranjape JM, Stockman BM, Silverman RH, Brinton MA. RNase L plays a role in the antiviral response to West Nile virus. J Virol 2006 Mar;80(6):2987-99.

[86] Liu R, Yue L, Li X, et al. Identification and characterization of small sub-genomic RNAs in dengue 1-4 virus- infected cell cultures and tissues. Biochem Biophys Res Commun 2010 Jan 1;391(1):1099-103.

25 Chapter 1

[87] Silva PA, Pereira CF, Dalebout TJ, Spaan WJ, Bredenbeek PJ. An RNA pseudoknot is required for production of yellow fever virus subgenomic RNA by the host nuclease XRN1. J Virol 2010 Nov;84(21):11395-406.

[88] Silva PA, Dalebout TJ, Gultyaev AP, Olsthoorn RC, Bredenbeek PJ. Characterization of the sfRNAs that are produced in cells infected with no known vector and cell fusing agent. 2013.

unpublished work

[89] Pijlman GP, Funk A, Kondratieva N, et al. A highly structured, nuclease-resistant, noncoding RNA produced by flaviviruses is required for pathogenicity. Cell Host Microbe 2008 Dec 11;4(6):579-91.

[90] Sheth U, Parker R. Decapping and decay of messenger RNA occur in cytoplasmic processing bodies. Science 2003 May 2;300(5620):805-8.

[91] Anderson P, Kedersha N. RNA granules. J Cell Biol 2006 Mar 13;172(6):803-8.

[92] Eulalio A, Behm-Ansmant I, Izaurralde E. P bodies: at the crossroads of post-transcriptional pathways. Nat Rev Mol Cell Biol 2007 Jan;8(1):9-22.

[93] Garneau NL, Wilusz J, Wilusz CJ. The highways and byways of mRNA decay. Nat Rev Mol Cell Biol 2007 Feb;8(2):113-26.

[94] Schuessler A, Funk A, Lazear HM, et al. West Nile virus non-coding subgenomic RNA contributes to viral evasion of type I interferon-mediated antiviral response. J Virol 2012 Feb 29.

[95] Schnettler E, Sterken MG, Leung JY, et al. Noncoding flavivirus RNA displays RNA interference suppressor activity in insect and Mammalian cells. J Virol 2012 Dec;86(24):13486-500.

[96] Moon SL, Anderson JR, Kumagai Y, et al. A noncoding RNA produced by arthropod-borne flaviviruses inhibits the cellular exoribonuclease XRN1 and alters host mRNA stability. RNA 2012 Nov;18(11):2029-40.

[97] Funk A, Truong K, Nagasaki T, et al. RNA structures required for production of subgenomic flavivirus RNA. J Virol 2010 Nov;84(21):11407-17.

[98] Knaap R., Jiang X, Olsthoorn RC, Bredenbeek PJ. RNA structure requirements for XRN1 stalling in the tick- born encephalitis virus 3’untranslated region. 2013. unpublished work

[99] Monath TP. Yellow fever: an update. Lancet Infect Dis 2001 Aug;1(1):11-20.

[100] Reed W, Carroll J, Agramonte A, Lazear JW. The Etiology of Yellow Fever-A Preliminary Note. Public Health Pap Rep 1900;26:37-53.

[101] Stokes A, Bauer JH, Hudson NP. The transmission of yellow fever to Macacus rhesus. 1928. Rev Med Virol 2001 May;11(3):141-8.

[102] WHO (World Health Organization). Yellow Fever fact sheet No.100 http://www.who.int/mediacentre/ factsheets/fs100/en/. 2009. Pamphlet

[103] WHO. Yellow fever vaccine. WHO position paper. Wkly Epidemiol Rec 2003 Oct 3;78(40):349-59.

[104] Theiler M, Smith HH. The use of yellow fever virus modified byin vitro cultivation for human immunization. J. Exp. Med. 65, 787-800 (1937). Rev Med Virol 2000 Jan;10(1):6-16.

[105] Hahn CS, Dalrymple JM, Strauss JH, Rice CM. Comparison of the virulent Asibi strain of yellow fever virus with the 17D vaccine strain derived from it. Proc Natl Acad Sci U S A 1987 Apr;84(7):2019-23.

[106] Theiler M, Smith HH. The use of yellow fever virus modified byin vitro cultivation for human immunization. J. Exp. Med. 65, 787-800 (1937). Rev Med Virol 2000 Jan;10(1):6-16.

26 Introduction

[107] Galler R, Freire MS, Jabor AV, Mann GF. The yellow fever 17D vaccine virus: molecular basis of viral attenuation and its use as an expression vector. Braz J Med Biol Res 1997 Feb;30(2):157-68.

[108] Barrett AD, MonathTP, Barban V, Niedrig M, Teuwen DE. 17D yellow fever vaccines: new insights. A report of a workshop held during the World Congress on medicine and health in the tropics, Marseille, France, Monday 12 September 2005. Vaccine 2007 Apr 12;25(15):2758-65.

[109] Barrett AD, euwenT DE. Yellow fever vaccine - how does it work and why do rare cases of serious adverse events take place? Curr Opin Immunol 2009 Jun;21(3):308-13.

[110] Pulendran B. Learning immunology from the yellow fever vaccine: innate immunity to systems vaccinology. Nat Rev Immunol 2009 Oct;9(10):741-7.

[111] Roukens AH, Soonawala D, Joosten SA, et al. Elderly subjects have a delayed antibody response and prolonged viraemia following yellow fever vaccination: a prospective controlled cohort study. PLoS One 2011;6(12):e27753.

[112] Bauer JH. The duration of passive immunity in yellow ever.f Am.J.Trop.Med.Hyg. 11, 451. 1931.

[113] Brandriss MW, Schlesinger JJ, Walsh EE, Briselli M. Lethal 17D yellow fever encephalitis in mice. I. Passive protection by monoclonal antibodies to the envelope proteins of 17D yellow fever and dengue 2 viruses. J Gen Virol 1986 Feb;67 ( Pt 2):229-34.

[114] Co MD, Terajima M, Cruz J, Ennis FA, Rothman AL. Human cytotoxic T lymphocyte responses to live attenuated 17D yellow fever vaccine: identification of HLA-B35-restricted CTL epitopes on nonstructural proteins NS1, NS2b, NS3, and the structural protein E. Virology 2002 Feb 1;293(1):151-63.

[115] Ahmed R, Akondy RS. Insights into human CD8(+) T-cell memory using the yellow fever and smallpox vaccines. Immunol Cell Biol 2011 Mar;89(3):340-5.

[116] Akondy RS, Monson ND, Miller JD, et al. The yellow fever virus vaccine induces a broad and polyfunctional human memory CD8+ T cell response. J Immunol 2009 Dec 15;183(12):7919-30.

[117] Miller JD, van der Most RG, Akondy RS, et al. Human effector and memory CD8+ T cell responses to smallpox and yellow fever vaccines. Immunity 2008 May;28(5):710-22.

[118] Barba-Spaeth G, Longman RS, Albert ML, Rice CM. Live attenuated yellow fever 17D infects human DCs and allows for presentation of endogenous and recombinant T cell epitopes. J Exp Med 2005 Nov 7;202(9):1179-84.

[119] Querec TD, Pulendran B. Understanding the role of innate immunity in the mechanism of action of the live attenuated Yellow Fever Vaccine 17D. Adv Exp Med Biol 2007;590:43-53.

[120] Gaucher D, Therrien R, Kettaf N, et al. Yellow fever vaccine induces integrated multilineage and polyfunctional immune responses. J Exp Med 2008 Dec 22;205(13):3119-31.

[121] Briese T, Bernard KA. West Nile virus--an old virus learning new tricks? J Neurovirol 2005 Oct;11(5):469-75.

[122] Briese T, Jia XY, Huang C, Grady LJ, Lipkin WI. Identification of a Kunjin/West Nile-like flavivirus in brains of patients with New York encephalitis. Lancet 1999 Oct 9;354(9186):1261-2.

[123] Gould EA, Higgs S. Impact of climate change and other factors on emerging arbovirus diseases. Trans R Soc Trop Med Hyg 2009 Feb;103(2):109-21.

[124] Pugachev KV, Guirakhoo F, Monath TP. New developments in flavivirus vaccines with special attention to yellow fever. Curr Opin Infect Dis 2005 Oct;18(5):387-94.

27 Chapter 1

[125] Halstead SB, Thomas SJ. Japanese encephalitis: new options for active immunization. Clin Infect Dis 2010 Apr 15;50(8):1155-64.

[126] Halstead SB, Thomas SJ. New Japanese encephalitis vaccines: alternatives to production in mouse brain. Expert Rev Vaccines 2011 Mar;10(3):355-64.

[127] Morrison D, Legg TJ, Billings CW, Forrat R, Yoksan S, Lang J. A novel tetravalent dengue vaccine is well tolerated and immunogenic against all 4 serotypes in flavivirus-naive adults. J Infect Dis 2010 Feb 1;201(3):370-7.

[128] Guy B, Guirakhoo F, Barban V, Higgs S, Monath TP, Lang J. Preclinical and clinical development of YFV 17D-based chimeric vaccines against dengue, West Nile and Japanese encephalitis viruses. Vaccine 2010 Jan 8;28(3):632-49.

[129] Bonaldo MC, Garratt C,R Caufour PS, et al. Surface expression of an immunodominant malaria protein B cell epitope by yellow fever virus. J Mol Biol 2002 Jan 25;315(4):873-85.

[130] Tao D, Barba-Spaeth G, Rai U, Nussenzweig V, Rice CM, Nussenzweig RS. Yellow fever 17D as a vaccine vector for microbial CTL epitopes: protection in a rodent malaria model. J Exp Med 2005 Jan 17;201(2):201- 9.

[131] McAllister A, Arbetman AE, Mandl S, Pena-Rossi C, Andino R. Recombinant yellow fever viruses are effective therapeutic vaccines for treatment of murine experimental solid tumors and pulmonary metastases. J Virol 2000 Oct;74(19):9197-205.

[132] Bredenbeek PJ, Molenkamp R, Spaan WJ, et al. A recombinant Yellow Fever 17D vaccine expressing Lassa virus glycoproteins. Virology 2006 Feb 20;345(2):299-304.

[133] Jiang X, Dalebout TJ, Bredenbeek PJ, et al. Yellow fever 17D-vectored vaccines expressing Lassa virus GP1 and GP2 glycoproteins provide protection against fatal disease in guinea pigs. Vaccine 2011 Feb 1;29(6):1248-57.

[134] Stoyanov CT, Boscardin SB, Deroubaix S, et al. Immunogenicity and protective efficacy of a recombinant yellow fever vaccine against the murine malarial parasite Plasmodium yoelii. Vaccine 2010 Jun 23;28(29):4644-52.

[135] Van Epps HL. Broadening the horizons for yellow fever: new uses for an old vaccine. J Exp Med 2005 Jan 17;201(2):165-8.

[136] Franco D, Li W, Qing F, et al. Evaluation of yellow fever virus 17D strain as a new vector for HIV-1 vaccine development. Vaccine 2010 Aug 9;28(35):5676-85.

[137] Bonaldo MC, Mello SM, Trindade GF, et al. Construction and characterization of recombinant flaviviruses bearing insertions between E and NS1 genes. Virol J 2007;4:115.

[138] Rice CM, Grakoui A, Galler R, Chambers TJ. Transcription of infectious yellow fever RNA from full-length cDNA templates produced by in vitro ligation. New Biol 1989 Dec;1(3):285-96.

[139] Kapoor M, Zhang L, Mohan PM, Padmanabhan R. Synthesis and characterization of an infectious dengue virus type-2 RNA genome (New Guinea C strain). Gene 1995 Sep 11;162(2):175-80.

[140] Sumiyoshi H, Hoke CH, Trent DW. Infectious Japanese encephalitis virus RNA can be synthesized from in vitro-ligated cDNA templates. J Virol 1992 Sep;66(9):5425-31.

28 Introduction

[141] Gualano RC, Pryor MJ, Cauchi MR, Wright PJ, Davidson AD. Identification of a major determinant of mouse neurovirulence of dengue virus type 2 using stably cloned genomic-length cDNA. J Gen Virol 1998 Mar;79 ( Pt 3):437-46.

[142] Khromykh AA, Westaway EG. Completion of Kunjin virus RNA sequence and recovery of an infectious RNA transcribed from stably cloned full-length cDNA. J Virol 1994 Jul;68(7):4580-8.

[143] Shi PY, Tilgner M, Lo MK. Construction and characterization of subgenomic replicons of New York strain of West Nile virus. Virology 2002 May 10;296(2):219-33.

[144] Kinney RM, Butrapet S, Chang GJ, et al. Construction of infectious cDNA clones for dengue 2 virus: strain 16681 and its attenuated vaccine derivative, strain PDK-53. Virology 1997 Apr 14;230(2):300-8.

[145] Lai CJ, Zhao BT, Hori H, Bray M. Infectious RNA transcribed from stably cloned full-length cDNA of dengue type 4 virus. Proc Natl Acad Sci U S A 1991 Jun 15;88(12):5139-43.

[146] Mandl CW, Ecker M, Holzmann H, Kunz C, Heinz FX. Infectious cDNA clones of tick-borne encephalitis virus European subtype prototypic strain Neudoerfl and high virulence strain Hypr. J Gen Virol 1997 May;78 ( Pt 5):1049-57.

29 30 Chapter 2

Yellow fever 17D-vectored vaccines expressing Lassa virus GP1 and GP2 glycoproteins provide protection against fatal disease in guinea pigs

Xiaohong Jiang, Tim J. Dalebout, Peter J. Bredenbeek, Ricardo Carrion Jr., Kathleen Brasky, Jean Patterson, Marco Goicochea, Joseph Bryant, Maria S. Salvato, Igor S. Lukashevich

Vaccine. 2011; 29:1248-57

Reprinted with permission

31 Chapter 2

Abstract

Yellow Fever (YF) and Lassa Fever (LF) are two prevalent hemorrhagic fevers co-circulating in West Africa and responsible for thousands of deaths annually. The YF vaccine 17D has been used as a vector for the Lassa virus glycoprotein precursor (LASV-GPC) or their subunits, GP1 (attachment glycoprotein) and GP2 (fusion glycoprotein). Cloning shorter inserts, LASV GP1 and GP2, between YF17D E and NS1 genes enhanced genetic stability of recombinant viruses, YF17D/LASV-GP1 and –GP2, in comparison with YF17D/LASV-GPC recombinant. The recombinant viruses were replication competent and properly processed YF and LASV GP proteins in infected cells. YF17D/LASV-GP1&GP2 induced specific CD8+ T cell responses in mice and protected strain 13 guinea pigs against fatal LF. Unlike immunization with live attenuated reassortant vaccine ML29, immunization with YF17D/LASV-GP1&GP2 did not provide sterilizing immunity. This study demonstrates the feasibility of YF17D-based vaccine to control LF in West Africa.

Key words: Yellow Fever, Lassa Fever, YF17D vaccine, recombinant vaccines

32 Yellow Fever 17D-Lassa Vaccine

1. Introduction

Yellow Fever 17D is one of the most efficacious and safe vaccines ever developed with a highly favorable benefit-risk profile and excellent cost-effectiveness. In YF endemic areas a vaccinated child is fully protected over a 50-year life time against different genotypes of YF for an investment of a few cents per year [1, 2]. Since 1937 more than 540 million doses have been delivered world-wide and no reversion to wild-type YF virus has been reported. Rare observed adverse effects seem to be associated with host factors and are not due to virus reversion [3-5]. After sub-subcutaneous inoculation, YF17D minimally replicates in local DCs and stimulates DC subsets via multiple Toll-like-receptors (TLRs) to elicit pro-inflammatory cytokines [6-8]. Cytokine-activated and mature dendritic cells (DCs) become protected against YF-induced cytopathogenicity and migrate to regional lymph nodes to elicit a broad spectrum of innate and adaptive immune responses [9-13]. A systems biology approach confirmed that YF17D induces integrated multi-lineage and poly-functional immune responses including activation of effector mechanisms of innate immunity (complement, multiple TLRs, cytokines/cytokine receptors, and interferons), as well as adaptive immune responses through an early T cell activation followed by robust B cell responses [14, 15]. Computational analyses identified a gene signature, B cell growth factor TNFRS17, predicting with up to 100% accuracy the neutralizing antibody response, a well- established correlate of protection [15]. The introduction of active immunization programs and vector control measures in 1937-1938, transformed YF from a major human plague to a medical curiosity by the end of World War II [1, 16, 17]. However, during the past 20 years the absence of effective public health policy, social-ecological factors and vaccine shortage resulted in a resurgence of YF (~ 200,000 cases and 30,000 deaths annually) in South America and Africa. Thirty-two African countries with a total population of 610 million people including more than 219 million urbanites are now considered at risk for YF. The vast majority of cases and deaths take place in sub-Saharan Africa, where YF is a major public health problem occurring in epidemic patterns [1]. In 2007 WHO–UNICEF recommended including YF17D vaccine in routine infant immunization programs and initiated mass vaccination campaigns to rapidly increase population immunity in high-risk areas [18]. Based on outstanding YF17D safety records and recent success in molecular biology of flaviviruses, the genetic backbone of YF17D has been used for construction of chimeric YF17D-based viruses expressing prM and E proteins of closely-related flaviviruses, Japaneses encephalitis, Dengue, and West Nile virus. ChimeriVax™-based vaccines against these flaviviruses are currently undergoing Phase II-III clinical testing (reviewed by [19]). Insertion

33 Chapter 2 of short immunogenic peptides into E surface glycoprotein protein did not compromise the morphological structure and attenuated phenotype of YF17D. Recombinant YF17D- based vectors have also been used in the development of promising new vaccines carrying immunogenic epitopes of flavivirus-unrelated pathogens (influenza, malaria, oncogenes) [6, 20-23]. We recently developed a novel strategy to express relatively large foreign genes using the YF17D backbone. We used this strategy to design a bivalent YF17D-based recombinant vaccine against Lassa Fever (LF) caused by Lassa virus (LASV). LASV circulates in YF endemic areas in West Africa and is responsible for thousands of deaths annually [24, 25]. Prospective serological and epidemiological studies performed in Sierra-Leone, Guinea, and Nigeria revealed the population at risk to be 59 million with an estimated annual incidence of illness of 3 million [25-28]. The sizeable disease burden makes a strong case for LASV vaccine development [29-31]. Like other members of the Arenaviridae, LASV has a bi-segmented (L and S) ambisense RNA genome [32, 33]. The L RNA encodes a large protein (L, or RNA-dependent RNA polymerase) [34], and a small zinc-binding Z protein [35]. The S RNA encodes the major structural proteins, nucleoprotein (NP) and glycoprotein precursor (GPC), cleaved into GP1, GP2, and a stable signal peptide, SSP [36-38]. In contrast to YF, CD8+ T cell responses play a major role in protection against LASV. Live replication-competent vaccine candidates based on vaccinia [31], vesicular stomatitis virus (VSV) [39], and Mopeia virus (MOPV), a non-pathogenic arenavirus relative of LASV [40, 41], have been proposed as potential vaccine candidates. However, vector safety concerns remain a major obstacle for further development of these vaccines. We designed a recombinant LASV vaccine based on the YF17D vector which, as it has been mentioned, was extensively tested in humans. A LASV truncated GPC gene (with SSP deletion) was cloned between the E and NS1 genes of YF17D. The YF17D/LASV∆GPC recombinant virus was replication-competent, deeply attenuated, induced immune responses against both pathogens, YFV and LASV, and protected guinea pigs against fatal LF in proof-of-concept challenge experiments [42]. However, our further attempts to scale-up this technology were complicated by instability of the YF17D/LASV∆GPC recombinant viruses. The LASV∆GPC insert within a monocistronic YF17D vector was stable only during the first 5 passages and the insert was randomly deleted after passage no. 5. In this paper, we describe our approaches to increase stability of LASV inserts cloned into the YF17D vector. The YF17D-based recombinant LASV vaccines with enhanced genetic stability have been tested in challenge experiments in strain 13 guinea pigs.

34 Yellow Fever 17D-Lassa Vaccine

2. Materials and methods

2.1. Viruses and cells

LASV, strain Josiah/SL, was obtained from the Centers for Disease Control (Atlanta, GA). All work with LASV-infectious samples was performed within the maximum containment laboratory at the Southwest Foundation for Biomedical Research in San Antonio, Texas. YF17D-204 was purchased from ATCC (Manassas, VA). The viruses were grown on Vero E6 cells cultured in Dulbecco’s modified minimum Eagle’s medium (DMEM, GIBCO-BRL) with 2% fetal calf serum (FCS, GIBCO-BRL), 1% penicillin-streptomycin, and L-glutamine (2 mM) at 37° C in 5% CO2 [41] by using a multiplicity of infection of 0.01. Supernatants were collected at 72 hours post-infection (LASV) or when cytopathic effect was clearly developed (YF17D), titrated on Vero cells as previously described [42], and virus stocks (1-5 x 107 PFU/ ml) were stored at -70° C.

2.2. Construction of YF17D/LASV recombinant viruses

The YF17D/LASV-GPC plasmid was constructed in the background of the full-length YFV17D cDNA clone by fusion PCR mutagenesis [42]. The recombinant plasmid was linearized by XhoI and used for in vitro RNA transcription. In vitro RNA transcription, electroporation of the BHK-21J cells, in vivo RNA labeling, preparation of virus stocks, immunofluorescence and plaque assays were previously described [42, 43]. LASV-GPC subunits, GP1 (aa 59-259) and GP2 (aa 260-491), were cloned between E and NS1 genes using similar techniques. Recombinant YF17D/LASV-GP1 viruses were replication competent and grew to titers (7-8 log10 PFU/ml) comparable with the YF17D virus. The initial attempts to construct YF17D/ LASV-GP2 did not result in production of infectious virus despite detectable viral RNA synthesis and positive immunofluorescence staining for YF NS1. This was probably due to the lack of cleavage between the YF17D E and LASV GP2 protein, which was supposed to be mediated by cellular signalase. To solve this problem, two new YF17D/LASV-GP2 recombinants with modified YF17D E – LASV GP2 fusion sequences were designed. In the first set of recombinant, YFV17D/LASV-GP2BglII, the naturally occurring YF17D E –NS1 cleavage site was restored by mutating the first amino acid of GP2 from Leu to Asp, which resulted in the production of a properly sized LASV GP2 and created a BglII site as indicated by the name of the construct. In the second sets of recombinants, YFV17D/GP2fus, the sequence encoding the first 14 aa of LASV GP1 was inserted between YF17D E and LASV GP2. This resulted in the production of a GP2 with a small N terminal GP1 extension. Both approaches gave infections recombinant viruses with high titer (1.0-2.0 x 107 PFU/ml) and stable LASV GP2 expression.

35 Chapter 2

2.3. Analysis of genetic stability of recombinant viruses

To test the genetic stability of the recombinant viruses, virus stocks recovered from YF17D/LASV-GP1 or –GP2-electroporated cells were passed 10 times in BHK-21J cells [43]. Supernatants of the infected cells were routinely harvested 24 hours post infection, diluted to 1:10 with phosphate buffered saline (PBS) and used to inoculate fresh cells. Two independent passage experiments were performed for each recombinant virus, YF17D/LASV- GP1 and –GP2. Electroporated and/or infected cells were labeled with 3[H]-Uridine in the presence of actinomycin D, intracellular RNA was isolated and analyzed by electrophoresis as previously described [42].

2.4. Detection of YF17D-specific and LASV-specific proteins in infected cells

To determine whether YF17D proteins were properly processed and released from the recombinant polyprotein-precursor containing LASV-GP1 and -GP2 protein-inserts, infected cells were incubated with radioactively labeled amino acids and processing of YF-specific proteins was monitored in pulse-chase experiments. The virus-specific proteins were immunoprecipitated with mouse polyclonal hyperimmune ascitic fluid to YF (ATCC), and subjected to SDS-PAGE analysis. For detection of LASV-GP1 and -GP2, transfected or infected cell lysates were subjected to SDS-PAGE separation and proteins were electroblotted to PVDF membranes (Hybond-P, Amersham, Piscataway, NJ). Monoclonal anti-GP1 and polyclonal rabbit anti-GP2 antibodies (gift from W. Garten, the Institut für Virologie der Phillips-Universität Marburg, Germany) were used to identify the LASV glycoproteins after incubation with an appropriate secondary goat IgG conjugated to alkaline phosphatase using NBT/BCIP (Invitrogen, Carlsbad, CA) as chromogenic substrate.

2.5. Immunogenicity studies in CBA/J mice

Four-week-old female CBA/J mice from Jackson Laboratories were used for immunogenicity studies. Previously, we showed that CBA/J mice are a useful small animal model to evaluate immunogenicity of a LASV vaccine candidate, the reassortant ML29 [41, 44]. In general, this reassortant encodes major immunogenic proteins, GPC and NP, from LASV and RNA polymerase and Z protein from MOPV. It also contains 5 non-conserved amino acid substitutions, mostly in L protein, that distinguished ML29 from the parental viruses. A single-shot immunization with ML29 induces broad CD8+ T cell-mediated sterilizing immunity in all tested animal models including non-human primates (review in [45-47]). In this study, ML29 was used as a positive vaccination control and for mapping immunodominant LASV GPC H2k-restricted CD8+ T cell epitopes. CBA/J mice were immunized intraperitoneally (i.p.) or subcutaneously (s.c.) with 1000 PFU of ML29 and euthanized on a weekly basis. Bulk

36 Yellow Fever 17D-Lassa Vaccine splenocytes were used for flow cytometry to evaluate the activation status of populations of CD3+CD8+ T lymphocytes using an antibody to CD11b. For detection of virus-specific T cells and mapping GPC H2k-restricted epitopes, immune splenocytes or in some experiments purified CD8+ T cells (MACS, Miltenyi Biotec Inc, Auburn, CA) were stimulated with a peptide library consisting of overlapping 21-mer peptides (MIMOTOPES, Australia) derived from LASV GPC, GP1, and GP2. Purified splenocytes from immunized or naïve mice were plated in a PVDF membrane-bottomed 96-well plate and stimulated with GPC-, GP1-, GP2- derived peptides (100 ng), ConA (10 μg) or RPMI medium. Cells were incubated overnight and processed the next day according to the protocol for mouse IFN-γ ELISPOT (U-CyTech biosciences, Utrecht, The Netherlands), as provided by the manufacturer. In brief, after the stimulation, the cells were washed, re-suspended in the same medium, and 0.3 – 0.4 x 106 cells per well were added to ELISPOT 96-well plates pre-coated with antibodies specific to mouse IFN-γ. The plates were incubated at 37°C for 5 h and the cells were washed away. Biotinylated detection antibodies were then added and the plates were incubated for 1 h at 37°C. Plates were washed, incubated with anti-biotin antibody labeled with gold particles. The spot-forming cells (SFC) secreting IFN-γ were detected and enumerated with C.T.L. Ltd Immunospot® S5 Micro-analyzer and Immunospot® V 4.0 software. For an immunogenicity study, CBA/J mice (4 animals per group) were immunized s.c. or i.p. with YF17D, YF17D/LAS-GPC, or YF17D/LAS-GP2 (1 x 106 PFU/mice in 0.1 ml) and boosted on day 14 with the same amount of the vaccines. Immunized animals were sacrificed on day 14 after the boost and secreted IFN-γ was evaluated by ELISPOT using bulk splenocytes stimulated with GPC-, GP1-, and GP2-derived peptides.

2.6. Vaccination-challenge experiments in strain 13 guinea pigs

Strain 13 guinea pigs (300-500 g, female) were purchased from USAMRIID (Fort Detrick, Frederick, MD). Animals (four to six animals per group) were vaccinated with YF17D/LASV- GP1&GP2 s.c. Both recombinant vaccines expressing GP1 and GP2 with infectious titers of 5 x 106 PFU/ml were equally mixed in PBS and injected in 0.5 ml in two separate sites on the back of animals. As a positive vaccination control 1,000 PFU of reassortant ML29 vaccine was inoculated s.c. As a negative vaccination control, animals were inoculated with PBS in 0.5 ml. On day 14, animals from the experimentally vaccinated group were boosted with the same dose of YF17D/LASV-GP1&GP2. On day 44, animals were challenged s.c. with LASV

(Josiah), 1,000 LD50/ml. The animals were observed twice daily for clinical manifestations according to an approved scoring sheet (temperature, weight, decreased activity, ruffled fur, loss of weight, labored breathing, hunched posture). Temperature was monitored by using implanted chips. Death or survival past 25 days was defined as an endpoint [40, 48]. Previously we

37 Chapter 2 showed that animals past this time point did not change survival outcome during extended follow-up period [41]. Animals which survived to day 25 or animals which met euthanasia criteria (fever, weakness, labored breathing, >25% loss in weight) were euthanized and sacrificed for histological studies as previously described [41]. For detection of viral RNA, tissue samples from vaccinated and challenged animals were submerged in RNAlater and cryopreserved. RNA was extracted later on using RNeasy mini kit (Qiagen, cat. no. 75142). RNA was converted into cDNA and amplified with 36E2 and 80F2 primers targeting LASV- GPC gene [49]. Standards used in qRT-PCR were generated from RNA isolated from serial 10-fold dilutions (101- 10-7 PFU/ml) of LASV Josiah/SL viral stocks that were enumerated in triplicate by conventional plaque assay as previously described [41]. Sensitivity of qRT-PCR was around 100-300 viral RNA copies per g of tissue.

2. 7. Data analysis

Statistical analyses (mean, SD, T-test) and graphics were performed using the Origin6.0 package (Microcal Software, Inc., Northampton, MA).

3. Results

3.1. Further development and improvement of YF17D recombinant technology based on foreign gene insertion between E and NS1 genes

Foreign gene inserts between the YF17D E and NS1 genes must be stable for at least 10 passages to consider this cloning strategy for further scale-up development. Ten passages will be enough to produce master-seed-lots and manufacture non-GMP bulk lots for pre-clinical studies [19]. As mentioned, our recombinant YF17D/ LASV∆GPC viruses were reasonably stable for at least 5 passages [42]. However, some virus stocks lost the inserted cassettes when the passages were extended. Similar observations were made by Bonaldo et al. [50]. To track the LASV∆GPC cassette integrity in Vero cells, a cassette region containing the insert and flanking sequences (Fig. 1A) was amplified (sense: AGGAAAGTTGTTCACTCAGACCATG; antisense: CAAGCTTCACAGG ATCTTCTGGATA) and analyzed by gel electrophoresis. The size of the cassette with the LASV∆GPC insert was 1.8 kb. The 1.8 kb amplicon was detected in passage 3-5 viruses (P3-P5). However, P6 viruses displayed a diverse electrophoresis profile indicating the loss of the LASV∆GPC inserts as confirmed by further sequence analysis (not shown). This loss of the insert during multiple cycles of viral RNA replication was originally attributed to homologous recombination, due to the duplication of the transmembrane region of YF17D E coding sequence flanking the LASV∆GPC insert [42]. In an attempt to

38 Yellow Fever 17D-Lassa Vaccine

Jiang et al. Page 15 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Fig. 1. Construction of genetically stable YF17D/LASV-GP vaccines. (A) Schematic representation of recombinant

YF17D-LASV virus carryingVaccine the . Author ∆GPC manuscript; insert. available In the in PMC initial 2012 March constructs 8. [42] the insert (1,365nt) was flanked by homologous E-derived sequences (red box) to insure the proper processing of recombinant polyprotein. In a new set of recombinant viruses, the fusion sequence between GP2 and NS1 (pink box) was substituted with heterologous sequences derived either from WNV or artificially designed (Art) to reduce homology with YF17D. Arrows indicate the sequences designed for PCR amplification of LASV ∆GPC insert and flanking areas to track the insert size during passages in BHK-21J cells. (B) Genetic stability of recombinant YF17D/LASV-∆GPC. Virus-specific RNA was labeled with 3[H]-Uridine in the presence of actinomycin D and intracellular RNA at passage 1, 5 and 10 was analyzed by gel electrophoresis. The RNA from YF17D-infected cells was used as a control. Arrows indicate recombinant YF17D/LASV∆GPC RNA, 12,227nt; and parental YF17D RNA, 10,862nt. (C), (D) Genetic structure of recombinant YF17D/LASV-GP1 and –GP2 viruses, respectively. Cleavage site between the YF17D E and GP2 protein was designed as described in Materials and Methods. (E) The size of LASV GP inserts after electroporation (Elpo) or after 10 passages in BHK-21J. Two independent passage experiments were performed, Exp1 and Exp2. Arrows indicate recombinant YF17D/LASV-GP1 RNA, 11,609nt; and parental YF17D RNA, 10,862nt. The size of YF17D/ LASV-GP2 RNA is 11,672nt.

39 Chapter 2 reduce recombination and enhance the stability of LASV∆GPC insert, the downstream transmembrane E sequence was substituted with a heterologous sequence with the similar function ensuring insertion of the YF17D into the endoplasmic reticulum. We suggested that the lower homology of the ∆GPC-NS1 fusion sequence to the C-terminal part of E protein can enhance the stability of the recombinant virus. Two new linkage sequences were tested (Fig. 1). The 1st fusion sequence was derived from West Nile virus (WNV). The 2nd one was an artificial heterologous sequence encoding a signal sequence motif for flavivirus NS1 protein. Insertion of these sequences between ∆GPC and NS1 genes improved stability of the GPC cassette, but did not completely resolve the instability problem. As seen in Fig. 1B, the YF17D/LASV∆GPC viruses made according to this strategy had enhanced stability. At least some P10 clones with WNV fusion sequence contained the full ∆GPC insert as was judged by RT/PCR analysis (not shown). Nevertheless, P10 recombinant viruses were still not fully stable in terms of preservation of integrity of LASV∆GPC sequences. The observed heterogeneity of the P10 virus population suggested that the loss of the LASV ∆GPC insert was a gradual process and not the result of the recombinant event. Apparently RNA with a length close to the parental YF17D virus has an advantage in replication and/or packaging over recombinant RNA containing large inserts like the LASV ∆GPC gene. In further experiments, we decided to reduce the size of LASV inserts and separately clone LASV GP1 and GP2 between E and NS1 genes as described in Materials and Methods. Based on our previous observations, we have reasonably assumed that LASV GP1 and GP2 maturation does not interfere with proteosomal degradation and binding with MHC-I molecules. At least LASV GPC, with deleted SSP [42] or with genetically modified GP1-GP2 cleavage site and deleted GP2 transmembrane domain (Lukashevich et al., unpublished), induced protective virus-specific CD8+ T cell responses in guinea pigs and in mice. We do see any concerns associated with the fusion sequences derived from WNV. Nevertheless, in further experiments we used an artificial fusion sequence between GP1 or GP2 and NS1.

3.2. Genetic stability of a new generation of YF17D/LASV-GP recombinant vaccines

As seen in Fig. 1, size reduction of cloning inserts combined with introduction of modified fusion sequences between the GP insert and NS1 gene resulted in recovery of YF17D-based recombinant viruses with enhanced genetic stability. To test the genetic stability of the recombinant viruses, virus stocks recovered from YF17D/LASV-GP1 or –GP2-electroporated cells were passed 10 times in BHK-21J cells. Supernatants of the infected cells were routinely harvested 24 hours post infection, diluted and used to inoculate fresh cells. Two independent passage experiments were performed for each recombinant virus, YF17D/LASV-GP1 and – GP2. Electroporated and/or infected cells were labeled with 3[H]-Uridine in the presence

40 Yellow Fever 17D-Lassa Vaccine of actinomycin D, intracellular RNA was isolated and analyzed by electrophoresis. A single RNA band was detected in passage 10 (P10) cells as well as in transfected cells. This RNA band migrated slightly slower than those from YF17D-infected cells and contained LASV GP1 and GP2 inserts as was confirmed by RT/PCR analysis. Interestingly, radioactive labeling revealed reduced RNA content in cells infected with YF17D/LASV-GP1 virus vs. –GP2 virus. Variations in infectivity during the passages and/or accumulation of defective interference particles can be possibly responsible for this observation.

3.3. Phenotypic features and stability in vitro

Recombinant YF17D/LASV-GP1 and -GP2 viruses induced pronounced cytopathogenic effects in BHK-21J and Vero cells and virus yields in these cells were around 1-2 x 107 PFU/ ml. The recombinant viruses produced clear plaques, which appeared to be smaller than those produced by the parental virus, YF17D (Fig. 2), in line with our previous observation [42]. This plaque phenotype was stable for P1-P10 viruses. Both YF17D/LASV-GP1 and -GP2 RNA-transfected cells, as well as cells infected with recombinant viruses, passage 1 (P1) and passage 10 (P10) viruses, were positively stained with anti-YF17D NS3 antibody, showing typical perinuclear staining. Also, LASV GP1 and YFV NS3 were found to be co-localized when double labeling was performed on YF17D/LASV- GP1-transfected cells (Fig. 3). Staining of cells transfected and/or infected with YF17D/LASV- GP with polyclonal antibodies against GP-derived peptide gave similar results [42]. Cells infected with recombinant YF17D/LASV-GP1 and –GP2 viruses were capable of expressing properly processed YFV antigens. As seen in Fig. 4, YF17D-specific antibodies

Jiang et al. Page 17 NIH-PA Author Manuscript NIH-PA Author Manuscript

Fig. 2. Plaque morphology of YF17D/LASV-GP viruses. YF17D/LASV-GP2 (A) and YF17D (B) viruses recovered after Fig. 2. Plaque morphology of YF17D/LASV-GP viruses. YF17D/LASV-GP2 (A) and YF17D (B) electroporation were platedviruses on recoveredconfluent after electroporation monolayers were plated of onBHK-21J confluent monolayerscells in of 6-well BHK-21J plates and viruses were titrated cells in 6-well plates and viruses were titrated by plaque assay as previously described [42]. by plaque assay as previouslyThe transfecteddescribed cells were [42]. incubated The for transfected 96 hours, fixed with cells 8% formaldehydewere incubated and stained for 96 hours, fixed with 8% with crystal violet. The final titer was 1 × 108 and 2 × 107 PFU/ml for8 YF17D and YF17D/7 formaldehyde and stained withLASV-GP2, crystal respectively. violet. The final titer was 1 x 10 and 2 x 10 PFU/ml for YF17D and YF17D/ LASV-GP2, respectively.

NIH-PA Author Manuscript 41

Vaccine. Author manuscript; available in PMC 2012 March 8. Chapter 2

Jiang et al. Page 18 NIH-PA Author Manuscript NIH-PA Author Manuscript

Fig. 3. Immunofluorescence staining of transfected cells. Cells transfected with YF17D/LAS-GP1 were fixed at 24 Fig. 3. hours post electroporation,Immunofluorescence followed stainingby staining of transfected for YF17D cells. Cells NS1 transfected (green) with and YF17D/LAS-GP1 LASV GP1 (red). Nuclei were stained were fixed at 24 hours post electroporation, followed by staining for YF17D NS1 (green) with Hoechst (blue). Magnification:and LASV GP1 (red). x Nuclei40. were stained with Hoechst (blue). Magnification: × 40.

NIH-PA Author Manuscript Jiang et al. Page 19 NIH-PA Author Manuscript

Vaccine. Author manuscript; available in PMC 2012 March 8. NIH-PA Author Manuscript

Fig. 4. Pulse-chase labelingFig. of 4. YF17D proteins. Cells infected with either YFV17D or the recombinant viruses were Pulse-chase labeling of YF17D proteins. Cells infected with either YFV17D or the radioactively labeled for 4recombinant hours, virusesor pulsed were radioactively 30 min, followedlabeled for 4 hours,by chasing or pulsed 30for min, either followed 1 by hour or 3.5 hours. Cell lysates chasing for either 1 hour or 3.5 hours. Cell lysates were subjected to immunoprecipitation were subjected to immunoprecipitationwith mouse hyperimmune with ascetic mouse fluid to YF17Dhyperimmune (ATCC). The precipitatesascites fluidwere separated to YF17D (ATCC). The precipitates were separated on 12.5% SDS-PAGEon 12.5% SDS-PAGE gel, and gel, and analyzed analyzed by by phosphor-imaging. phosphor-imaging. 1, label 4 hours; 1, 2,label pulse- 4 hours; 2, pulse-labeling, 30 labeling, 30 minutes; 3, 1 hour chase; 4, 3.5 hours chase. Viral proteins: NS5 (103 kDa), minutes; 3, 1 hour chase; 4,NS3 3.5 (70 hourskDa), E (50chase. kDa), NS1 Viral (46 proteins:kDa), NS4B (27 NS5 kDa), (103 PrM (26kDa), kDa), NS3 NS2A (70 (22 kDa), kDa), E (50 kDa), NS1 (46 kDa), NIH-PA Author Manuscript and NS2B (14 kDa) could be seen. Molecular weight markers are indicated on the left side. NS4B (27 kDa), PrM (26 kDa), NS2A (22 kDa), and NS2B (14 kDa) could be seen. Molecular weight markers are indicated on the left side.

42

Vaccine. Author manuscript; available in PMC 2012 March 8. Yellow Fever 17D-Lassa Vaccine precipitated from infected cells YF NS1, NS2A, NS2B, NS3, NS4B, NS5, PrM and E proteins. These results indicate that monocistronic recombinant mRNA translated into a polyprotein- precursor, and the processing of YF17D polyprotein-precursor was not affected by the LASV GP1 and GP2 inserts. Finally, LASV-specific proteins, GP1 and GP2, were detected in YF17D/LASV-GP- infected cells by Western blot using GP1 monoclonal and GP2 polyclonal antibodies as previously described [42]. These results (Fig. 5) demonstrate that both LASV glycoproteins, GP1 and GP2, were translated from recombinant monocistronic mRNA and properly processed and released from the recombinant YF17D/LASV-GP1 and –GP2 polyprotein.

3.4. YF17D/LASV-GP1 and –GP2 recombinant viruses induce specific CD8+ T cell responses in CBA/J mice

It is important to evaluate the immunogenicity of experimental vaccines in small animal models before testing putative vaccine candidates in expensive vaccination-challenge experiments. Currently, there is no mouse model of human LF. The outcome of LASV infection in immunocompetent mice depends on H2 genotype, age, routes of infection, and resembles LCMV-like immunopathology (review in [46]). In contrast, virulence of LASV in man and in non-human primates is directly related to virus load [25, 40, 51] and is not associated

Jiang et al. Page 20 NIH-PA Author Manuscript NIH-PA Author Manuscript

Fig. 5. Western-blot analysisFig. 5. of LASV-specific glycoproteins in cells infected with recombinant YF17D/LAS-GP1 and Western-blot analysis of LASV-specific glycoproteins in cells infected with recombinant –GP2 viruses. The analysisYF17D/LAS-GP1 was carried andout –GP2 with viruses. a 1:2,000 The analysis dilution was carried of outanti-GP1 with a 1:2,000 and dilution anti-GP of antibodies as previously described [42]. Markers inanti-GP1 kilodaltons and anti-GP are antibodies indicated as previously on the described side. The[42]. Markersarrow in in kilodaltons (A) indicates are GP1 and in (B) indicates indicated on the side. The arrow in (A) indicates GP1 and in (B) indicates GP2 GP2 glycoproteins. The recombinantglycoproteins. The YF17D/LASV-GP2BglII recombinant YF17D/LASV-GP2BglII contained contained natural natural YFVYFV E E– NS1– NS1 cleavage site with Leu/ cleavage site with Leu/Asp mutation in the first amino acid of GP2. The recombinant Asp mutation in the firstYF17D/LASV-GP2fus amino acid of containedGP2. The a sequencerecombinant encoding theYF17D/LASV-GP2fus first 14 aa of LASV GP1 inserted contained a sequence encoding between YF17D E and LASV GP2 (see Materials and Methods for details). the NIH-PA Author Manuscript first 14 aa of LASV GP1 inserted between YF17D E and LASV GP2 (see Materials and Methods for details).

43

Vaccine. Author manuscript; available in PMC 2012 March 8. Chapter 2

with immunopathology. We previously showed that young adult CBA/J mice inoculated s.c. or i.p. with the MOP/LAS (ML29) reassortant induced protective CD8+ T cell responses [41, 45, 47] and can be used as an economical assay for vaccine immunogenicity to access capacity to elicit specific CD8+ T cell responses that play a major role in LASV protective immunity [45, 47]. After ML29-immunization, activation of+ CD8 T cells peaked on day 14 and this time-point was used for collection of splenocytes for mapping immunodominant epitopes. Immune splenocytes stimulation with pools of peptides (Fig. 6B) or with individual peptides (not shown) reveled that LASV GPC H2k-restricted epitopes were located within GP1 glycoprotein at aa positions 57-83, 218-244 and within GP2 glycoprotein at positions 400-426. Immunization of CBA/J mice with recombinant YF17D/LASV-GP1 or -GP2 vaccines, but not with parental YF17D vaccine, induced LASV antigen-specific CD8+ T cell immune responses detected in IFN-γ ELISPOT (Fig. 6C). Extracellular secretion of IFN-γ was detected after stimulation of immune GP1- or GP2-specific splenocytes either with peptide cocktails derived from LASV GPC or from GP1 and GP2, respectively. The level of extracellular IFN-γ production after stimulation of immune splenocytes with LASV GP-specific peptides was approximately 6-fold lower after immunization of mice with YF17D-based recombinant GP1

Jiang et al. Page 21 NIH-PA Author Manuscript

Fig. 6. Evaluation of immunogenicity of LASV vaccine candidates in CBA/J mice. (A) Vaccination control group: Fig. 6. activated CD3+CD8Evaluation+ cells peaked of immunogenicity on day 14 afterof LASV immunization. vaccine candidates (B) inMapping CBA/J of mice. GPC (A) kH2 -restricted Vaccination epitopes: on day NIH-PA Author Manuscript 14 after vaccinationcontrol group:splenocytes activated were CD3+CD8+ stimulated cells with peaked a peptideon day 14 library after immunization. consisting of (B) overlapping 21-mer peptides Mapping of GPC H2k-restricted epitopes: on day 14 after vaccination splenocytes were derived from LASVstimulated GPC. Stimulated with a peptide splenocytes library consisting were incubatedof overlapping in wells21-mer of peptidesthe ELISPOT derived plates from pre-coated with Abs against mouse IFN-γLASV and GPC. washed Stimulated away. splenocytes Cytokine-secreting were incubated cells in wellsor spot-forming of the ELISPOT cells (SFC)plates cellspre- were detected and coated with Abs against mouse IFN-γ and washed away. Cytokine-secreting cells or spot- enumerated according to the manufacturer’s instructions for mouse IFN-γ ELISPOT (U-CyTech biosciences, Utrecht, forming cells (SFC) cells were detected and enumerated according to the manufacturer’s The Netherlands).instructions Responses for mouse to peptides IFN-γ ELISPOT derived (U-CyTechfrom GP1 biosciences,and GP2 are Utrecht, indicated The byNetherlands). black bars. (C) CD8+ T cell responses in miceResponses immunized to peptides with derivedrecombinant from GP1 YF17D/LAS-GP1 and GP2 are indicated and –GP2 by blackviruses. bars. Splenocytes (C) CD8+ T from immunized cell responses in mice immunized with recombinant YF17D/LAS-GP1 and –GP2 viruses. mice were stimulatedSplenocytes with from peptide immunized cocktails mice were derived stimulated from withfull peptideGPC (black cocktails boxes) derived or from from fullGP1 or GP2 subunits (sparse boxes).GPC (black boxes) or from GP1 or GP2 subunits (sparse boxes).

44 NIH-PA Author Manuscript

Vaccine. Author manuscript; available in PMC 2012 March 8. Yellow Fever 17D-Lassa Vaccine and GP2 vaccines in comparison with ML29 reassortant expressing naturally processed LASV GPC.

3.5. Recombinant YF17D/LASV-GP1 and –GP2 vaccines protected strain 13 guinea pigs against fatal LF disease but did not prevent LASV infection

Individually expressed in vaccinia virus, LASV GPC subunits, GP1 and GP2, did not protect non-human primates against fatal LF [30]. Based on this observation, recombinant YF17D- based viruses expressing GP1 and GP2 were equally mixed and used for immunization of strain 13 guinea pigs. These animals are extremely sensitive to LASV with LD50 < 1 PFU/ ml [40, 52] and all non-vaccinated control animals met euthanasia criteria 2 weeks after challenge. The most prominent manifestations of the disease included fever, weight loss, and elevated liver enzymes; pathohistological findings included intestinal pneumonia with septal and alveolar edema, and hepatitis with hepatocytic vacuolization and lipidosis [46, 53]. Vaccination control group animals were fully protected against LASV challenge and had no signs of the disease. In these animals LASV was only marginally detected by RT-PCR in peritoneal lymph nodes in good confirmation with our previous results [41, 53]. Fiveof six guinea pigs vaccinated with YF17D/LASV-GP1&GP2 were also protected against fatal LF. However, surviving animals expressed clinical symptoms of LF, had an elevated body temperature, lost body weight, and LASV was detected in tested tissues. Still, viral titers as well as clinical symptoms were lower than in non-vaccinated control group and animals with disease signs recovered at the end of the observation period (Fig. 7). These results indicate that immunization with recombinant YF17D/LASV-GP1&GP2 vaccine protected guinea pigs against fatal LASV disease but did not prevent LASV infection.

4. Discussion

Due to its outstanding safety record, YF17D has been used as a promising vector for the development of vaccines against arthropod-borne flaviviruses and against non-flavivirus- related pathogens [6, 22, 54-57]. The YF17D-based ChimeriVax™ technology [58] replaces the prM and E genes of YF17D with the corresponding genes of the closely-related flaviviruses, Dengue, WNV, and Japanese encephalitis. These vaccines are in advanced stages of development and ChimerixVax™-WN vaccine (PreveNile) was licensed in the U.S. for veterinary applications (rev. [19]). Various insertion locations in structural and non-structural genes and in the′ 3 UTR of YF17D were used to clone and express flavivirus-unrelated foreign inserts. The E protein is the major flavivirus antigen inducing neutralizing antibodies and represents an attractive

45 Chapter 2

Jiang et al. Page 22 NIH-PA Author Manuscript NIH-PA Author Manuscript

Fig. 7. YF17D/LASV-GP1&GP2Fig. 7. vaccination protects guinea pigs from fatal outcome but does not prevent LASV infection. (A) SurvivalYF17D/LASV-GP1&GP2 rate and clinical vaccinationscores (insert). protects Clinical guinea scoring pigs from was fatal based outcome on evaluation but does not of responsiveness, prevent LASV infection. (A) Survival rate and clinical scores (insert). Clinical scoring was edema, dehydration,based on rash, evaluation appearance, of responsiveness, petechia, feededema, consumption, dehydration, rash, anorexia, appearance, stool, petechia,urine, fluid feed uptake (0, normal; 1, mild; 2, moderate;consumption, 3, severe. anorexia, (B) Temperature stool, urine, fluid(°F) was uptake monitored (0, normal; by 1,sensor-implants mild; 2, moderate; and 3, recorded severe. twice per day for each animal.(B) (C) Temperature Weight of (°F)animals was monitoredon day of bychallenge sensor-implants was assayed and recorded as 100%. twice Open per circles,day for eachanimals vaccinated animal. (C) Weight of animals on day of challenge was assayed as 100%. Open circles, NIH-PA Author Manuscript with YF17D/LASV-GP1&GP2;animals vaccinated open with squares, YF17D/LASV-GP1&GP2; animals vaccinated open with squares, ML29 animals (positive vaccinated control); with closed squares, negative controlML29 (animals (positive control);inoculated closed with squares, media). negative (D) control Viral (animals loads (LASVinoculated RNA with copies) media). in vaccinated-challenged (D) Viral loads (LASV RNA copies) in vaccinated-challenged animals. Black boxes, animals. Black boxes, animals challenged with LASV (no vaccine); white boxes, animals vaccinated with YF17D/ animals challenged with LASV (no vaccine); white boxes, animals vaccinated with YF17D/ LASV-GP1&GP2;LASV-GP1&GP2; spare boxes, animals spare boxes,vaccinated animals with vaccinated ML29 (positive with ML29 control). (positive control).

site for insertion of foreign epitopes. However, the structural limitations of the flavivirus E protein allowed expression of relatively short inserts, ~ 30-35 aa [21]. Importantly, these inserts resulted in additional attenuation of recombinant YF17D viruses concomitant with enhanced safety profiles of chimeric YF17D-based vaccines (rev. [19]). The NS1 protein also induces antibody responses contributing to protection against flaviviruses. In a recent study the M2e peptideVaccine (23 . Authoraa) of manuscript; influenza available virus in PMC 2012was March successfully 8. cloned into the NS1 gene using transposon-mediated insertion [59]. Non-flavivirus CTL epitopes have been also cloned between the NS2B-NS3 cleavage site. The insertion of the H-2Kd-restricted CTL epitope of the circumsporozoite protein of P. yoelii at this site resulted in a recombinant virus which was replication competent, stable, and diminished the parasite burden in the liver by ~70% [22]. Immunization of mice with

46 Yellow Fever 17D-Lassa Vaccine recombinant YF17D carrying ova-specific T-cell epitope, SIINFEKL, between NS2B and NS3, induced a specific CD8+ T-cell population and protected animals from lethal challenge with an aggressive lethal melanoma cell line [54]. We cloned a relatively large gene encoding LASV ∆GPC glycoprotein (433 aa) between YF17D E and NS1 genes in attempts to design bivalent YF17D-LASV vaccine to control both infections in co-endemic areas of Africa [42]. The recombinant virus induced humoral and cell-mediated immune responses against YF and LASV, respectively, and protected guinea pigs against fatal LF disease. In spite of these encouraging results, we [42] and others [50] faced instability problems when large inserts were cloned at the E-NS1 site. It has been suggested that duplication of flavivirus sequences flanking an inserted gene contributed to the genetic instability of the virus. In the current study two approaches have been used to address this problem: (i) insertion of modified fusion sequences between the C-terminal part of foreign inserts and NS1 gene to prevent possible recombination events resulting in deletion of the foreign gene; and (ii) reduction of insert size. We have shownthat modification of fusion sequences improved stability but did not preserve the large ∆GPC insert from the deletion during 10 passages in tissue culture. When the size of the insert was reduced almost 2-fold, GP1 and GP2 sub-units of LASV GPC were successfully cloned at the E-NS1 site and these inserts were stable during 10 passages. The size of the LASV GP1 is 200 aa and the size of GP2 is 231 aa residues. In accord with our results, a similar sized insert, 224 aa residues, from SIVmac239 Gag was cloned between the genes encoding the viral proteins E and NS1. The resulting recombinant virus, the YF17D/SIVGag45-269, was stable in vivo and able to induce SIV-specific CD8+ T cell responses in rhesus macaques [60]. The size limitation of foreign gene inserts in YF17D vector seems not to be unique for the E-NS1 site. Recently recombinant YF17D vaccine was engineered by cloning CSP (aa 57-344) from P. yoelii using a novel insertion site, the capsid gene. The insert was cloned in frame between the highly conserved cyclization sequences (nt 1-75) and FDMV 2A and Ubi genes which were inserted to ensure appropriate cleavage [61]. Notably, the CSP insert was intact up to 6 passages in tissue culture and became undetectable by RT/PCR by passage 7. We have used a similar strategy with limited success in attempts to clone and express overlapping fragments of LASV NP resembling in some ways cytoplasmic YF core protein (Bredenbeek, unpublished). Taken together these results indicate that a monocistronic YF17D genome encoding a relatively large flavivirus-nonrelated gene inserts at the E-NS1 site or in the C gene can produce viable recombinant viruses in cell cultures. However, genetic stability of recombinant viruses is the major problem because these viruses lose genetic material after more than 10 passages. We have shown here that cloning of inserts of moderate size (200-230 aa residues) between the genes encoding the viral proteins E and NS1 resulted in genetically stable recombinant viruses during 10 passages in tissue culture. The YF17D genome itself is

47 Chapter 2 remarkably stable [62]. Ten passages of recombinant YF17D-based viruses will be sufficient to manufacture non-GMP lots for toxicology and pre-clinical studies and for making Phase I lots during vaccine development [19]. Insertion of foreign genes into the YF17D genome resulted in additional attenuation of chimeric or recombinant YF17D-based viruses. In plasma and tissue of vaccinated guinea pigs [42] and marmosets (Lukashevich, unpublished) YF17D/LASV-GPC was not detectable by plaque assay and only transient and limited replication was confirmed by co-cultivation in Vero and/or by RT/PCR in line with previously published data [19, 63, 64]. Still, we were able to detect anti-YF antibodies in YF17D/LASV∆GPC-immunized animals. Nevertheless, for bivalent recombinant YF17D-based vaccines anti-YF efficacy must be additionally evaluated in challenge experiments with wild-type YF virus to confirm efficacy of the parental vaccine, YF17D. In the recent review [19] theoretical safety (reversion to virulence and potential viscero- and neurotropism, recombinantion with different flaviviruses or distantly-related RNA viruses) and enviromental concerns (transmission by arthropod vectors) of YF17D- based vaccines have been carefully addressed and evaluated in support of their further late development and registration. Importantly, YF17D is safe and effective in HIV-infected individuals with CD4 cell counts above 200 cell/mm3, even if they have previously developed AIDS or if their viral load is above 10,000 copies/ml [65, 66]. Moreover, flaviviruses decrease CD4 expression and inhibit HIV-1 replication in human CD4+ T cells [67]. We have recently showed that our live attenuated reassortant ML29 expressing LASV GP and NP is safe and immunugenic in SIV-positive rhesus macaques [68]. Takent together these results suggest that YF17D-vectored LASV vaccines will be safe in HIV-infected individuals representing the most vulnerable group among immunocopromized populations in West Africa. As mentioned above, w lo or marginal transient viremia in animals immunized with recombinant YF17D-based vaccines seems to be associated with relatively low immunogenicity. As a result, for recombinant YF17D-based vaccines, in contrast to parental YF17D, a prime-boost immunization protocol is often required to induce antigen-specific immune responses. To avoid anti-vector immunity and stimulate specific immunity the optional prime-boost strategy has to include different vectors. For example, priming with recombinant BCG expressing SIV antigens increased the frequency of the SIV-specific CD8+

T-cell responses after boosting of rhesus monkeys with recombinant rYF17D/SIVGag45-269 [60]. However, we and others [61] showed that prime-boost immunization with YF17D- based recombinant vaccines does not necessarily induce specific responses against cloned foreign antigens and/or undesirable anti-vector immunity. In our experiments with YF17D/ LASV-GPC [42] or –GP&GP2 vaccines (this study) we did not see differences in protective efficacy after single shot or prime-boost vaccination. Based on the nature of protective immune responses against certain pathogens, it is reasonable to consider other possible

48 Yellow Fever 17D-Lassa Vaccine approaches to enhance specific immune responses against protective antigens cloned into YF17D vector. At least in the case of anti-LASV immunity, rational design of the GP antigen seems to be a promising approach. Preliminary experiments with structurally stable trimeric LASV GP showed that CD8+ T cell cross-priming can additionally contribute to effects of immunization based on conventional MHC type I antigen presentation.

Acknowledgments

This work was supported by grant RO1 AI068961 (to I.S.L.) from the National Institutes of Health.

49 Chapter 2

References

1. Galbraith SE, Barrett ADT: Yellow Fever. In ADT Barrett, LR Stanberry (eds), Vaccines for Biodefense and Emerging and Neglected Diseases Academic Press, Elsevier, Amsterdam, The Netherlands 2009:753-785.

2. Monath TP: Yellow fever vaccine. Expert Review of Vaccines 2005, 4:553-574.

3. Hayes EB: Acute viscerotropic disease following vaccination against yellow fever. Trans R Soc Trop Med Hyg 2007, 101:967-971.

4. Lindsey NP, Schroeder BA, Miller ER, Braun MM, Hinckley AF, Marano N, Slade BA, Barnett ED, Brunette GW, Horan K, et al: Adverse event reports following yellow fever vaccination. Vaccine 2008, 26:6077- 6082.

5. Silva ML, Espirito-Santo LR, Martins MA, Silveira-Lemos D, Peruhype-Magalhaes V, Caminha RC, de Andrade Maranhao-Filho P, Auxiliadora-Martins M, de Menezes Martins R, Galler R, et al: Clinical and Immunological Insights on Severe, Adverse Neurotropic and Viscerotropic Disease following 17D Yellow Fever Vaccination. Clin Vaccine Immunol 2010, 17:118-126.

6. Barba-Spaeth G, Longman RS, Albert ML, Rice CM: Live attenuated yellow fever 17D infects human DCs and allows for presentation of endogenous and recombinant T cell epitopes. J Exp Med 2005, 202:1179- 1184.

7. Querec T, Bennouna S, Alkan S, Laouar Y, Gorden K, Flavell R, Akira S, Ahmed R, Pulendran B: Yellow fever vaccine YF-17D activates multiple dendritic cell subsets via TLR2, 7, 8, and 9 to stimulate polyvalent immunity. J Exp Med 2006, 203:413-424.

8. Palmer DR, Fernandez S, Bisbing J, Peachman KK, Rao M, Barvir D, Gunther V, Burgess T, Kohno Y, Padmanabhan R, Sun W: Restricted replication and lysosomal trafficking of yellow fever 17D vaccine virus in human dendritic cells. J Gen Virol 2007, 88:148-156.

9. Akondy RS, Monson ND, Miller JD, Edupuganti S, Teuwen D, Wu H, Quyyumi F, Garg S, Altman JD, Del Rio C, et al: The Yellow Fever Virus Vaccine Induces a Broad and Polyfunctional Human Memory CD8+ T Cell Response. J Immunol 2009, 183:7919-7930.

10. Martins MÂ, Silva ML, Elói-Santos SM, Ribeiro JGL, Peruhype-Magalhães V, Marciano APV, Homma A, Kroon EG, Teixeira-Carvalho A, Martins-Filho OA: Innate immunity phenotypic features point toward simultaneous raise of activation and modulation events following 17DD live attenuated yellow fever first-time vaccination. Vaccine 2008, 26:1173-1184.

11. Martins MÂ, Silva ML,Marciano APV, Peruhype-Magalhães V, Eloi-Santos SM, Ribeiro JGL, Correa-Oliveira R, Homma A, Kroon EG, Teixeira-Carvalho A, Martins-Filho OA: Activation/modulation of adaptive immunity emerges simultaneously after 17DD yellow fever first-time vaccination: is this the keyto prevent severe adverse reactions following immunization? Clinical and Experimental Immunology 2007, 148:90-100.

12. Santos AP, Matos DCS, Bertho AL, Mendonça SCF, Marcovistz R: Detection of TH1/TH2 cytokine signatures in yellow fever 17DD first-time vaccinees through ELISpot assay. Cytokine 2008, 42:152-155.

13. Miller JD, van der Most RG, Akondy RS, Glidewell JT, Albott S, Masopust D, Murali-Krishna K, Mahar PL, Edupuganti S, Lalor S, et al: Human Effector and Memory CD8+ T Cell Responses to Smallpox and Yellow Fever Vaccines. Immunity 2008, 28:710-722.

50 Yellow Fever 17D-Lassa Vaccine

14. Gaucher D, Therrien R, Kettaf N, Angermann BR, Boucher G, Filali-Mouhim A, Moser JM, Mehta RS, Drake DR, III, Castro E, et al: Yellow fever vaccine induces integrated multilineage and polyfunctional immune responses. J Exp Med 2008, 205:3119-3131.

15. Querec TD, Akondy RS, Lee EK, Cao W, Nakaya HI, Teuwen D, Pirani A, Gernert K, Deng J, Marzolf B, et al: Systems biology approach predicts immunogenicity of the yellow fever vaccine in humans. Nat Immunol 2009, 10:116-125.

16. Monath TP: Yellow fever vaccine. In: Vaccines, 4th edition (Plotkin SA, Orenstein WA), Saunders2004 :1095- 1176.

17. Monath TP: Yellow fever as an endemic/epidemic disease and priorities for vaccination. Bull Soc Pathol Exot 2006, 99:341-347.

18. WHO: The yellow fever initiative, 2007. Geneva, WHO (www.who.int/csr/disease/yellowfev/ introduction/en/index/html). 2007.

19. Guy B, Guirakhoo F, Barban V, Higgs S, Monath TP, Lang J: Preclinical and clinical development of YFV 17D-based chimeric vaccines against dengue, West Nile and Japanese encephalitis viruses. Vaccine 2010, 28:632-649.

20. Bonaldo MC, Garratt C, R Caufour PS, Freire MS, Rodrigues MM, Nussenzweig RS, Galler R: Surface expression of an immunodominant malaria protein B cell epitope by yellow fever virus. J Mol Biol 2002, 315:873 - 885.

21. Bonaldo MC, Garratt C,R Marchevsky RS, Coutinho ESF, Jabor AV, Almeida LFC, Yamamura AMY, Duarte AS, Oliveira PJ, Lizeu JOP, et al: Attenuation of Recombinant Yellow Fever 17D Viruses Expressing Foreign Protein Epitopes at the Surface. J Virol 2005, 79:8602-8613.

22. Tao D, Barba-Spaeth G, Rai U, Nussenzweig V, Rice CM, Nussenzweig RS: Yellow fever 17D as a vaccine vector for microbial CTL epitopes: protection in a rodent malaria model. J Exp Med 2005, 201:201-209.

23. Stoyanov CT, Boscardin SB, Deroubaix S, Barba-Spaeth G, Franco D, Nussenzweig RS, Nussenzweig M, Rice CM: Immunogenicity and protective efficacy of a recombinant yellow fever vaccine against the murine malarial parasite Plasmodium yoelii. Vaccine 2010, In Press, Corrected Proof.

24. Khan SH, Goba A, Chu M, Roth C, Healing T, Marx A, Fair J, Guttieri MC, Ferro P, Imes T, et al: New opportunities for field research on the pathogenesis and treatment of Lassa fever. Antiviral Research 2008, 78:103-115.

25. McCormick JB, Fisher-Hoch SP: Lassa fever. Curr Top Microbiol Immunol 2002, 262:75-109.

26. Anonymous: WHO. Update on Lassa fever in West Africa. Weekly Epidemiol Record 2005, 80:86-88.

27. Richmond JK, Baglole D: Lassa fever: epidemiology, clinical features, and social consequences. BMJ 2003, 327:1271-1275.

28. Fichet-Calvet E, Rogers DJ: Risk Maps of Lassa Fever in West Africa. PLoS Negl Trop Dis 2009, 3:doi: e38810.31371/journal.pntd.0000388.

29. Borio L, Inglesby T, Peters CJ, Schmaljohn AL, Hughes JM, Jahrling PB, Ksiazek T, Johnson KM, Meyerhoff A, O’Toole T, et al: Hemorrhagic fever viruses as biological weapons: medical and public health management. JAMA 2002, 287:2391-2405.

30. Fisher-Hoch SP, Hutwagner L, Brown B, McCormick JB: Effective Vaccine for Lassa Fever. J Virol 2000, 74:6777-6783.

51 Chapter 2

31. Fisher-Hoch SP, McCormick JB: Lassa fever vaccine. Expert Rev Vaccines 2004, 3:103-111.

32. Lukashevich IS, Stelmakh TA, Golubev VP, Stchesljenok EP, Lemeshko NN: Ribonucleic acids of Machupo and Lassa viruses. Arch Virol 1984, 79:189-203.

33. Lukashevich IS, Salvato MS: Lassa Virus Genome. Current Genomics 2006, 7:351-379.

34. Lukashevich I, Djavani M, Shapiro K, Sanchez A, Ravkov E, Nichol S, Salvato M: The Lassa fever virus L gene: nucleotide sequence, comparison, and precipitation of a predicted 250 kDa protein with monospecific antiserum. J Gen Virol 1997, 78:547-551.

35. Djavani M, I. S. Lukashevich, A. Sanchez, S. T. Nichol, Salvato MS: Completion of the Lassa fever virus sequence and identification of a RING finger open reading frame at the L RNA 5’ end. Virology 1997, 235:414-418.

36. Auperin DD, Sasso DR, McCormick JB: Nucleotide sequence of the glycoprotein gene and intergenic region of the Lassa virus S genome RNA. Virology 1986, 154:155-167.

37. Clegg JC, Wilson SM, Oram JD: Nucleotide sequence of the S RNA of Lassa virus (Nigerian strain) and comparative analysis of arenavirus gene products. Virus Res 1991, 18:151-164.

38. Lenz O, ter Meulen J, Klenk H-D, Seidah NG, Garten W: The Lassa virus glycoprotein precursor GP-C is proteolytically processed by subtilase SKI-1/S1P. PNAS 2001, 98:12701-12705.

39. Feldmann H, Geisbert TW, Jahrling PB, Klenk HD, Netesov SV, Peters CJ, Sanchez A, Swanepoel R, Volchkov VE: Filoviridae. Virus Taxonomy: VIIIth Report of the International Committee on Taxonomy of Viruses 2005:645 - 653.

40. Peters CJ, P. B. Jahrling, C. T. Liu, R. H. Kenyon, K. T. Jr McKee, Oro. JGB: Experimental studies of arenaviral hemorrhagic fevers. Curr Top Microbiol Immunol 1987, 134:5-68.

41. Lukashevich IS, Patterson J, Carrion R, Moshkoff D, Ticer A, Zapata J, Brasky K, Geiger R, Hubbard GH, Bryant J, Salvato MS: A live attenuated vaccine for Lassa fever made by reassortment of Lassa and Mopeia viruses. J Virol 2005, 79:13934-13942.

42. Bredenbeek PJ, Molenkamp R, Spaan WJM, Deubel V, Marianneau P, Salvato MS, Moshkoff D, Zapata J, Tikhonov I, Patterson J, et al: A recombinant Yellow Fever 17D vaccine expressing Lassa virus glycoproteins. Virology 2006, 345:299-304.

43. Bredenbeek PJ, Kooi EA, Lindenbach B, Huijkman N, Rice CM, Spaan WJM: A stable full-length yellow fever virus cDNA clone and the role of conserved RNA elements in flavivirus replication. J Gen Virol 2003, 84:1261-1268.

44. Lukashevich IS: Generation of eassortantsr between African arenaviruses. Virology 1992, 188:600-605.

45. Lukashevich IS: A live reassortant vaccine for Lassa Fver. In: E M Vela (ed): Molecular Pathogenesis of Hemorrhagic Fever Viruses, Transw Res Net, Kerala, India 2010:pp 143-170.

46. Lukashevich IS, Vela EM: Pathogenesis of Lassa Virus Infection in Experimental Animals. In: E M Vela (ed): Molecular Pathogenesis of Hemorrhagic Fever Viruses, Transw Res Net, Kerala, India 2010:pp 101-142.

47. Salvato MS, Lukashevich IS: Vaccines against Lassa fever. In: Myron M Levine (ed): New Generation Vaccines, Forth Edition, by Marcel Dekker, Inc 2009:pp. 895-904.

52 Yellow Fever 17D-Lassa Vaccine

48. Pushko P, Geisbert J, Parker M, Jahrling P, Smith J: Individual and bivalent vaccines based on alphavirus replicons protect guinea pigs against infection with Lassa and Ebola viruses. J Virol 2001, 75:11677- 11685.

49. Demby AH, Chamberlain J, Brown DW, Clegg CS: Early diagnosis of Lassa fever by reverse transcription- PCR. J Clin Microbiol 1994, 32:2898-2903.

50. Bonaldo M, Mello S, Trindade G, Rangel A, Duarte A, Oliveira P, Freire M, Kubelka C, Galler R: Construction and characterization of recombinant flaviviruses bearing insertions between E and NS1 genes. Virology Journal 2007, 4:115.

51. McCormick JB, D. H. Walker, I. J. King, P. A. Webb, L. H. Elliott, S. G. Whitfield, K. M. Johnson.: Lassa virus hepatitis: a study of fatal Lassa fever in humans. Am J Trop Med Hyg 1986, 35:401-407.

52. Jahrling PB, Smith S, Hesse RA, Rhoderick JB: Pathogenesis of Lassa virus infection in guinea pigs. Infect Immun 1982, 37:771-778.

53. Carrion RJ, Patterson JL, Johnson C, Gonzales M, Moreira CR, Ticer A, Brasky K, Hubbard GB, Moshkoff D, Zapata J, et al: A ML29 reassortant virus protects guinea pigs against a distantly-related Nigerian strain of Lassa virus and can provide sterilizing immunity. Vaccine 2007, 25:4093-4102.

54. McAllister A, Arbetman AE, Mandl S, Pena-Rossi C, Andino R: Recombinant Yellow Fever Viruses Are Effective Therapeutic Vaccines for Treatment of Murine Experimental Solid Tumors and Pulmonary Metastases. J Virol 2000, 74:9197-9205.

55. Van Epps HL: Broadening the horizons for yellow fever: new uses for an old vaccine. J Exp Med 2005, 201:165-168.

56. Bonaldo M, Garratt RC, Caufour PS, Freire MS, Rodrigues MM, Nussenzweig RS, Galler R.: Surface expression of an immunodominant malaria protein B cell epitope by yellow fever virus. J Mol Biol 2002, 315:873-885.

57. Xiaowu P, Fu WC, Guo YH, Zhang LS, Xie TP, G GX: Construction of recombinant yellow fever virus 17D containing 2A fragment as a vaccine vector. Sheng Wu Gong Cheng Xue Bao 2006, 22:492-498.

58. Chambers TJ, Monath TP, Guirakhoo F: US6696281. (2004).

59. Rumyantsev AA, Zhang Z-x, Gao Q-s, Moretti N, Brown N, Kleanthous H, Delagrave S, Guirakhoo F, Collett MS, Pugachev KV: Direct random insertion of an influenza virus immunologic determinant into the NS1 glycoprotein of a vaccine flavivirus. Virology 2010, 396:329-338.

60. Bonaldo MC, Martins MA, Rudersdorf R, Mudd PA, Sacha JB, Piaskowski SM, Costa Neves PC, Veloso de Santana MG, Vojnov L, Capuano S, III, et al: Recombinant Yellow Fever Vaccine Virus 17D Expressing Simian Immunodeficiency Virus SIVmac239 Gag Induces SIV-Specific CD8+ T-Cell Responses in Rhesus Macaques. J Virol 2010, 84:3699-3706.

61. Stoyanov CT, Boscardin SB, Deroubaix S, Barba-Spaeth G, Franco D, Nussenzweig RS, Nussenzweig M, Rice CM: Immunogenicity and protective efficacy of a recombinant yellow fever vaccine against the murine malarial parasite Plasmodium yoelii. Vaccine 2010, 28:4644-4652.

62. Barban V, Girerd Y, Aguirre M, Gulia S, Pétiard F, Riou P, Barrere B, Lang J: High stability of yellow fever 17D-204 vaccine: A 12-year restrospective analysis of large-scale production. Vaccine 2007, 25:2941- 2950.

53 Chapter 2

63. Monath TP, Myers GA, Beck RA, Knauber M, Scappaticci K, Pullano T, Tad Archambault W, Catalan J, Miller C, Zhang Z-X, et al: Safety testing for neurovirulence of novel live, attenuated flavivirus vaccines: Infant mice provide an accurate surrogate for the test in monkeys. Biologicals 2005, 33:131-144.

64. Monath TP, Levenbook I, Soike K, Zhang Z-X, Ratterree M, Draper K, Barrett ADT, Nichols R, Weltzin R, Arroyo J, Guirakhoo F: Chimeric Yellow Fever Virus 17D-Japanese Encephalitis Virus Vaccine: Dose- Response Effectiveness and Extended Safety Testing in Rhesus Monkeys. J Virol 2000, 74:1742-1751.

65. Receveur M, Thiébaut R, Vedy S, Malvy D, Mercié P, Bras ML: Yellow fever vaccination of human immunodeficiency virus-infected patients: report of 2 cases. Clin Infect Dis 2000, 31:E7-8.

66. Tattevin P, Depatureaux AG, Chapplain JM, Dupont M, Souala F, Arvieux C, Poveda JD, Michelet C: Yellow fever vaccine is safe and effective in HIV-infected patients. AIDS 2004, 18:825-827.

67. Xiang J, McLinden JH, Rydze RA, Chang Q, Kaufman TM, Klinzman D, Stapleton JT: Viruses within the Flaviviridae Decrease CD4 Expression and Inhibit HIV Replication in Human CD4+ Cells. J Immunol 2009, 183:7860-7869.

68. Zapata J. C. PB, Goicochea M, Bryant J, Davis H, Pauza C. D., Lukashevich I. S., Salvato M. S.: Arenavirus vaccines in SIV-positive Rhesus macaques. ASV, 28th Meeting, Scientific Program & Abstracts, Vancouver, Canada 2009, W7-7:94.

54 Chapter 3

Molecular and immunological characterization of a DNA launched yellow fever virus 17D infectious clone

Xiaohong Jiang, Tim J. Dalebout, Igor S. Lukashevich, David D. Ho, Charles M. Rice, Peter J. Bredenbeek and David Franco

55 Chapter 3

Abstract

Yellow fever virus (YFV)-17D is an empirically developed, highly effective live-attenuated vaccine that has been administered to human beings for almost a century. YFV-17D has stood as a paradigm for a successful viral vaccine, and has been exploited as a potential virus vector for the development of recombinant vaccines against other diseases. In this study, a DNA launched YFV-17D construct (pBeloBAC-FLYF) was explored as a new modality to the standard vaccine to combine the commendable features of both DNA vaccine and live attenuated viral vaccine. The DNA launched YFV-17D construct was characterized extensively both in cell culture and in a small animal model. High titers of YFV-17D were generated upon transfection of the DNA into cells, whereas a mutant with deletion in the capsid coding region (pBeloBAC-YF/ΔC) was restricted to a single round of infection, with no release of progeny virus. Homologous prime-boost immunization of mice with both pBeloBAC-FLYF and pBeloBAC-YF/ΔC elicited specific dose-dependent cellular immune response against YFV-17D. Neutralizing antibodies were only detected when A129 mice that were deficient in innate immunity were vaccinated with pBeloBAC-FLYF. These promising results underline the potential of the DNA launched YFV both as an alternative to standard YFV-17D vaccination but also as a vaccine platform for the development of DNA-based recombinant YFV vaccines.

56 DNA launched YFV-17D

Introduction

Flaviviruses are a group of small enveloped viruses that contain a plus-strand RNA genome of around 11kb in length. The genome encompasses a 5’ untranslated region (5’-UTR), followed by a single large open reading frame and a highly structured 3’-UTR [1]. Many flaviviruses, especially the ones causing diseases in humans, are arthropod- borne pathogens (arboviruses) that are transmitted by either mosquitoes or ticks. Clinical presentation of human infection ranges from mild febrile illness to severe haemorrhagic fever or meningoencephalitis [2]. Yellow fever virus (YFV) is the prototype of the Flavivirus genus and causes haemorrhagic fever featured by liver dysfunction and jaundice [3]. There is no cure for yellow fever, and vaccination is the single most important preventive measure against the disease [3]. The YFV-17D vaccine is a live attenuated vaccine empirically developed in the 1930s by Theiler and Smith [4], which granted Theiler the 1951 Nobel Prize in medicine. Ever since its development, YFV-17D vaccine has been administered to over 540 million humans globally, and has stood as a paradigm for a successful vaccine with a great record of both safety and efficacy [5]. The vaccine is well tolerated, with only rare cases of severe adverse reactions (reviewed in [3;6;7]). Over 95% of the vaccinees will develop protective immunity against all known wildtype YFV strains within 10 days after vaccination [6], although the molecular determinants of its attenuation and immunogenicity are not yet fully understood. A single immunization induces a broad spectrum of immune responses, including cytotoxic T lymphocytes (CTLs), a mixed T helper type I (TH1)-TH2 profile, and neutralizing antibodies that can persist for up to 30 years [8]. These adaptive immune responses are preceded by the activation of several effector arms of innate immunity and the induction of a network of antiviral genes, including various cytokines (IP-10, IL-1α), molecules involved in sensing viruses (TLR-7, RIG-I and MDA-5), as well as transcription factors that regulate type I interferons (IRF7 and STAT1) [9;10]. The broad spectrum of the immune response triggered by YFV vaccine renders it an interesting candidate as a vector for the development of recombinant vaccines. Indeed, YFV has been used as the backbone in chimeric vaccines against other flaviviruses of clinical significance, such as West Nile virus (WNV) [11;12], Japanese encephalitis virus (JEV) [13- 16] and Dengue virus (DENV) [17-19]. These YFV-based chimeric vaccines were found to be safe and highly immunogenic in both preclinical and clinical trials, with IMOJEV(®), a YFV-17D-based Japanese encephalitis vaccine passing phase III clinical trials and licensed for human use in Australia [20]. YFV-17D has also been successfully exploited as a vector for the expression of foreign T cell or B cell epitopes, as potential vaccine candidates against malaria [21;22] or as therapeutic anticancer vaccine [23]. Moreover, recombinant YFV-17D

57 Chapter 3 viruses expressing heterologous antigens were proposed as bivalent vaccine candidates against Lassa virus [24;25] or as alternative vaccine vectors for malaria [26] and HIV [27;28]. In vitro characterization demonstrated that these recombinant viruses were viable and could express both YFV and foreign proteins. Experiments in small animal models showed promising results in terms of immunogenicity and protection. Despite these encouraging results, genetic instability has been encountered for these recombinant viruses, especially with larger inserts, which is attributed to the limited packaging capacity of the icosahedral virion [25;26;29]. Despite the availability of a highly effective YFV vaccine, the number of yellow fever cases has increased over the past two decades, due to the declining immunity to YFV infection in the population, deforestation, urbanization, population movements, poor vector control and climate change [3]. Until recently the vaccine was still in short supply, and vaccine coverage is low in high-risk areas [30]. Alternative strategies to facilitate vaccine production, transportation or storage would therefore help accommodate the increasing needs for YFV vaccine. Compared to conventional live-attenuated vaccines, DNA vaccines are temperature- stable, easily stored and can be manufactured on a large scale [31]. Although the concept of DNA vaccination has generated a great deal of excitement ever since its introduction in the early 1990s, no DNA vaccine has been approved for human use. Significant challenges will need to be overcome before DNA vaccines achieve mainstream acceptance, especially regarding effective delivery and augmenting antigen expression level for a sufficiently potent immune response [31;32]. With the objective of combining the commendable qualities of DNA vaccines with those of a live-attenuated vaccine, a DNA launched YFV-17D vaccine was constructed using a similar strategy as described for a Kunjin virus based DNA vaccine [33]. This infectious YFV- 17D DNA was stable in E.coli host, and was extensively characterized both in cell culture and in mice to evaluate its potential as an alternative vaccine platform. In addition, deletions/ mutations were introduced into the sequence encoding the viral capsid and RNA dependent RNA polymerase (RdRP), rendering the virus deficient in either virus packaging or genome replication. Murine immune responses induced by these DNA launched YFV mutants were determined and compared to those triggered by DNA launched, replication and packaging competent YFV-17D.

58 DNA launched YFV-17D

Materials and Methods

Cell cultures and YFV-17D stocks

BHK-21J cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Life Technologies, Carlsbad, USA) supplemented with 8% fetal calf serum (FCS). YFV-17D stocks were obtained by harvesting the medium of BHK-21J cells that were transfected with in vitro-transcribed RNA of an infectious YFV-17D cDNA clone [34]. Virus titers were determined by plaque assays on BHK-21J cells, as described previously [35].

Construction of infectious YFV-17D DNA clones

General standard nucleic acid methodologies were used throughout this study [36], unless described in more detail. Chemically competent Escherichia coli DH5α cells were used for cloning [37], except for BACmid construction, for which DH10B competent cells (Invitrogen) were used. Nucleotide numbering of the YFV clones was based on GenBank accession no. X03700 [38] The CMV promoter and hepatitis delta virus ribozyme (HDVr) sequences were amplified from pcDNA3.0 (Invitrogen) and RlucRep-HDVr [39] plasmids and fused to the penultimate nucleotides at the 5’ and 3’ untranslated regions of the viral genome by fusion PCR [40]. Primers used in PCR were listed in Table 1. The resulting PCR products were cloned into pACNR-FLYFX [34] to yield pACNR-CMVp-FLYF-HDVr in which the 5’ end of the full-length YFV insert was directly fused to a CMV promoter and the HDVr was engineered to ensure the production of a YFV RNA with an authentic 3’ end. For the construction of pBeloBAC- FLYF, the complete insert of pACNR-CMVp-FLYF-HDVr encompassing the CMV promoter, YFV-17D cDNA and the HDVr was moved as a NotI – AflII fragment into a modified version of the BACmid pBeloBAC11 with unique NotI and AflII sites in the polylinker sequence. A Quick-Change mutagenesis based strategy (Stratagene, La Jolla, USA) was used to create two mutants of pBeloBAC-FLYF. The pBeloBAC-YF/∆C contains an in-frame deletion of 49 amino acids (nt 185 to 331) in the capsid gene of YFV-17D [41]. In pBeloBAC-YF/GSA, the GDD motif (amino acids 3171 to 3173) of the viral RNA dependent RNA polymerase NS5 was mutated to GSA [41;42]. Primers used for the construction of these two derivatives are listed in Table 1.

Transfection of BHK-21J cells with the YFV-17D DNA driven constructs

BACmid DNA was prepared using Nucleobond Xtra Maxi EF (Macherey-Nagel, Düren, Germany), followed by CsCl gradient purification. DNA purity and quality was routinely

59 Chapter 3 Comment PCR, NotI+CMV promoter 5’ PCR, NotI+CMV promoter 3’ (with SacI site) PCR, CMV promoter 1-28 nt CMV 3’end+YFV-17D 2491-2513 nt YFV-17D poly A AflII+SV40 10657-10677 nt YFV-17D 9625-9654 nt YFV-17D 9622-9651 nt YFV-17D 0831-10862+HDVr 5’ 0831-10862+HDVr nt YFV-17D Quickchange mutagenesis, YFV-17D nt 167-203 nt YFV-17D mutagenesis, Quickchange 163-193 nt YFV-17D mutagenesis, Quickchange PCR, YFV-17D nt 324-348 nt PCR, YFV-17D 1675-1699 nt PCR, YFV-17D Polarity + - + - - + + - + - + - Sequence CTCCTCGTCGTACC ggcgcgcctgacattgac GGAtcCGcCTGTGTGGTCCGGCC gagctctctggctaactaga TTCGaagtCTGGACCCAAGACAAGG gcggccgc cttaagacatgataagatacattg ccg tgcttatatagacc ccagagagctc gca TGAAGATACCATCTCCGCACTTG ccg AAGAACGGAGCCTCCGCTACC GGTGAGT CGGACCACACAGgCGgaTCCACTCACCGCC CGACGAGGAGTTCGaagtTTGTCAAACAAAATAAAACttCGAA GTTTGACAAac + gcg AGTAAATCCTGTGTGCTAATTGAGGTGC GCAGACCTTTGGATGACAAACACAAAACCACT gggtcggcatggcatctccacc GATAGTGGCGGCATGCGGAGGTTCA Primer YFV 1596 YFV 1597 YFV 1507 YFV 1315 YFV 1617 YFV 1598 YFV 1599 YFV 1701 YFV 1702 YFV 1704 RU-O-14941 Purpose Introduction of CMV promoterIntroduction YFV 1595 Introduction of HDVr and SV40 and of HDVr Introduction polyA signal the GDD motif in RdRPMutate YFV 1616 185-331 within the nt Delete region coding capsid Primers used in the construction of DNA launched YFV and its derivatives used in the construction 1. Primers Table

60 DNA launched YFV-17D checked with electrophoresis, and the concentration was determined using a NanoDrop photospectrometer. BHK-21J cells were transfected with 5 µg of BACmid DNA by electroporation as described previously [43]. Transfected cells were fixed with 3% paraformaldehyde at various time points after electroporation for immunofluorescense assay, and supernatants were collected for virus titration by plaque assay, as described above.

Immunofluorescence Assay

Transfected cells seeded on coverslips were fixed with 3% paraformaldehyde in PBS (pH 7·4) at 24 h, 48h, and 72h post-electroporation (p.e.) for at least 30 min and washed with PBS containing 10 mM glycine. Following permeabilization with 0·1% Triton X-100 in PBS, indirect immunofluorescence was carried out with rabbit anti-YFV NS3 antiserum diluted 1:2000 in PBS/5% FSC, and visualized with a secondary Alexa488-conjugated goat anti-rabbit IgG (Invitrogen, Carlsbad, USA) antibody with a dilution of 1:300 in PBS/5%FCS.

Animals and immunization schemes

Female, 6 – 8 week old, transgenic AAD mice expressing the α1 and α2 domains from the human HLA-A2.1 and the α3 domain of the murine H-2Db in the C57BL/6 background were obtained from Dr. Ralph Steinman’s lab and were used in groups of 5-6 mice for immunization studies. Mice were immunized intramuscularly by electroporation [44] with 10ng, 100ng, or 1 μg of endotoxin-free pBeloBAC-FLYF-17D or derivatives per mouse on days 0 and 14. For additional studies on emiavir and the production YFV neutralizing antibody titers, 6-8 week old A129 mice that are deficient in interferon (IFN) a and b receptors and congenic 129 mice (obtained from Dr. Charles Rice lab) were immunized intramuscularly with 106 PFU of YFV-17D or 5 μg of pBeloBAC-FLYF. Blood was drawn every other day up till 29 days post immunization. Plaque assays and plaque reduction neutralization assays were used to determine the titers of YFV-17D and virus neutralizing antibody levels in the serum. All animal work was in compliance with approved institutional protocols.

ELISpot assays for enumeration of IFN-γ spot-forming cells (SFC)

Two weeks after the final immunization, AAD mice were sacrificed and their spleens were removed to prepare splenocytes as described previously [28]. ELISpot assays (BD-Biosciences PharMingen, San Diego, USA) were performed to determine the frequency of IFN-γ excreting cells, according to the manufacturer’s protocol. Briefly, splenocytes were cultured in the presence of concanavalin A (2.5 μg/ml; Sigma, St. Louis, USA) or YFV-specific CD8+ and CD4+ T cells peptides [45] at a concentration of 1 μg/ml. A YFV-17D-specific peptide (CD4/E15L; ERWFVRNPFFAVTAL) with a greatly reduced capacity to stimulate YF17D-specific CD4+ T

61 Chapter 3 cells [45] was used as a negative control. After 16h of culture, the plates were processed to determine the frequency of IFN-γ SFC [28]. Briefly, the plates were washed and incubated with biotinylated anti-IFN-γ for 2 h at room temperature, followed by HRP-conjugated avidin for 1 h at room temperature. Reactions were developed with AEC substrate (Calbiochem- Novabiochem Corporation, San Diego, CA). Final enumeration of IFN-γ SFC was performed using the Immunospot Analyzer ELISpot reader with the aid of the Immunospot software version 3.0 (Cellular Technologies Ltd., Shaker Heights, OH). Data are presented as SFC/106 cells. Results were considered positive if the number of SFC was above 20 and higher than the background (culture with medium alone). The results are presented after subtraction of the background, which was consistently found to be 10–25 spots/106 cells throughout the experiments.

Intracellular cytokine staining (ICCS) and flow cytometry

Splenocytes from immunized mice were analyzed by ICCS for Interleukin 2 (IL-2) and IFN-γ production after stimulation with YF17D-specific, 15 aa long peptides as described previously [28]. Cultures without stimulation or those treated with concanavalin A served as negative or positive control, respectively. Stained cells were analyzed with the BD LSR-II flow cytometer using FACSDiva software (BD Biosciences, San Jose, USA). The data were processed with FlowJo 8.6.1 software (Tree Star, Ashland, USA).

Plaque reduction neutralization test (PRNT)

YFV-17D-specific neutralizing antibody titers were determined by 50% endpoint PRNT in 6-well plates of BHK-21J cells using a standard PRNT protocol. Briefly, a two-fold serial dilution of mouse serum was prepared in PBS/2%FCS; 5 μl serum of each dilution was incubated with 5 μl of PBS/2%FCS containing approximately 80 pfu of YFV-17D virus. After 1h of incubation on ice, the serum-virus mixtures were diluted with 300 μl PBS/2%FCS and inoculated onto BHK-21J cells. After an hour of incubation at 37°C the inocula were removed, and an overlay of DMEM/2% FCS with1.2% Avicel was added. Plaques were visualized by staining with 0.1% crystal violet after 4 days of incubation.

62 DNA launched YFV-17D

Results

Construction of an infectious DNA clone of YFV-17D

The characteristics of the DNA construct that was created to launch infection of YFV-17D from DNA are shown in Figure 1. The 5’-UTR of the viral RNA was fused to a cytomegalovirus (CMV) promoter so that cellular RNA polymerase II would initiate transcription of the YFV- 17D genome. The hepatitis delta virus ribozyme (HDVr) was engineered precisely after the last nucleotide of YFV genome to ensure the production an authentic 3’-end ofthe transcribed viral RNA. Initially the CMV promoter and HDVr cassettes were cloned into the pACNR-FLYFx [34]. However, the resulting plasmid pACNR-CMVp-FLYF-HDVr was found to be genetically instable during propagation in E. coli DH5α. Attempts to stabilize the plasmid by using different bacteria strains or alternative culture conditions were unsuccessful. To further decrease the copy number for the stabilization of the construct, the BACmid pBeloBAC11 was evaluated as a vector for the DNA launched YFV-17D cassette. The resulting BACmid pBeloBAC-FLYF was shown to be genetically stable at least up to 20 passages in eitherE.coli DH5α or DH10B (data not shown). To determine the contribution of viral replication and virus spread tothe immunological characteristics of a DNA launched YFV-17D vaccination regime, two other pBeloBAC-based DNA launched YFV plasmids were constructed using pBeloBAC-FLYF as a template. In pBeloBAC-YF/GSA, the GDD motif of the viral RNA dependent RNA polymerase (RdRP) was mutated to GSA, which inactivated the RdRP [42], whereas in pBeloBAC-YF/∆C, nt 185 to 331 in the nucleocapsid gene were deleted [41]. The latter should result in the CMV promoter driven transcription of replication competent YFV-17D RNA that is unable to produce infectious virus.

Characterization of DNA launched YFV-17D in cell culture

The pBeloBAC-FLYF, pBeloBAC-YF/∆C and pBeloBAC-YF/GSA DNA were transfected into BHK-21J cells by electroporation. The transfected cells were fixed at 24hrs, 48hrs, and 72hr post electroporation (p.e.) and the expression of YFV NS3 was analyzed by indirect immunofluorescence. At 24 hr p.e., a small percentage of NS3 positive cells were observed in the cells that were transfected with pBeloBAC-FLYF and pBeloBAC-YF/∆C (Fig. 2A), respectively. As anticipated, the proportion of NS3 positive cells increased for thecells transfected with pBeloBAC-FLYF BACmid, while remaining at a low percentage for the YFV capsid deletion mutant (Fig. 2A). No positive signal for YFV NS3 was detected in the cells transfected with pBeloBAC-YF/GSA DNA (Fig. 2A). Medium of the DNA transfected cells was sampled at 24 hr intervals and used to determine virus production by plaque assays. As shown in Fig. 2B, increasing amounts of YFV-17D were detected in the medium of the cells

63 Chapter 3

5’-UTR 3’-UTR CMVp struc. genes non-struc. genes HDVr pA

RNA pol II Full-length YFV-17D cDNA ribozyme promoter RNA pol II transcription terminator NNNNAG A) 24h …..48h CTNNNN72h

YF/∆C

start transcription ribozyme cleavage AG first two nts YFV. CT last two nts YFV.

Figure 1. Schematic representation of the geneticFLYF structure of pBeloBAC11-FLYF Colored boxes indicate from left to right the CMV promoter (CMVp) (red), the YFV-17D ORF (blue), the hepatitis delta virus ribozyme (yellow) and the RNA polymeraseFig. 1 II transcription terminator (red). The 5’ and 3’ YFV UTR sequences are depicted as a black line. The most 5’YF/GSA YFV nucleotides that originate from the CMV promoter driven RNA polymerase II transcription and the last two nucleotides of YFV genome that are produced by cleavage of the ribozyme are also indicated.

A) 24h 48h 72h B) Growth Kinetics

YF/∆C 1.00E+08

1.00E+07

1.00E+06

FLYF 1.00E+05 FLYF 1.00E+04 titer (pfu/ml) titer YF/GSA

1.00E+03 YF/ C YF/GSA 1.00E+02 △

1.00E+01

0 24h 48h 72h

time point post electroporation B) Growth Kinetics Figure 2. Characterization of DNA launched YFV-17D clones in BHK cells Fig. 2 BHK1.00E+08 cells were electroporated with pBeloBAC-FLYF, pBeloBAC-YF/ΔC and pBeloBAC-YF/GSA and analyzed for the expression1.00E+07 of YFV antigen (A) and virus production (B). A) 1.00E+06Indirect immunofluorescence staining of YFV-17D NS3 protein. Cells were fixed at the indicated times p.e. and

1.00E+05stained as described in the Materials and Methods section. YFV NS3 shows up as green (Alexa 488) and the FLYF 1.00E+04nuclei of the cells were stained with Hoechst and show up as blue. titer (pfu/ml) titer YF/GSA

B) 1.00E+03Kinetics of virus production in BHK-21J cells transfected YF/withC pBeloBAC-FLYF (green), pBeloBAC-YF/ΔC (yellow)

1.00E+02and pBeloBAC-YF/GSA (pink) as determined by plaque assays△ of the supernatants collected at the indicated

1.00E+01times p.e.

0 24h 48h 72h

time point post electroporation

Fig. 2

64 DNA launched YFV-17D transfected with pBeloBAC-FLYF, reaching a titer of over 107 PFU/ml at 72 hr p.e., whereas no virus production was detected in the supernatants collected from cells transfected with pBeloBAC-YF/∆C or pBeloBAC-YF/GSA DNA. The results of both the immunofluoresence assay and the plaque assay confirmed that as expected, only the cells that were transfected with pBeloBAC-FLYF DNA were able to spread and produce infectious virus. The RNA transcribed from pBeloBAC-YF/∆C was able to replicate as demonstrated by the readily detectable expression of YFV NS3, but was incapable of producing infectious progeny virus due to the lack of a functional capsid protein. To further investigate the YFV RdRP driven intracellular RNA synthesis, cells transfected with pBeloBAC-FLYF were labeled with 3[H]-uridine in the presence of actinomycine D (Act.D) with 6-hour intervals. 3[H]-labeled viral RNA was first detected at 33 h p.e. and the amount of labeled RNA gradually increased over time, probably reflecting the spread of infection to cells that were initially not transfected (Fig. 3A). Careful inspection of the fluorogram revealed a faint and faster migrating RNA species in samples collected at later time points, suggesting the presence of a YFV replicon RNA. Flavivirus replicons are RNAs that are able to replicate autonomously, but cannot form virus particles due to an in-frame deletion within the region encoding the structural genes. Trans-complementation by functional structural proteins can lead to the formation of replicon containing virus particles [46]. To test for the presence of YFV replicon RNA in the virus stocks derived from the DNA launched system, BHK-21J cells were infected with YFV-17D virus derived from in vitro transcribed infectious RNA or from pBeloBAC-FLYF transfection. The infected cells were labeled with 3[H]-uridine in the presence of Act.D and the intracellular RNA was analyzed by electrophoresis of denatured RNA. A band, suggestive of YFV replicon RNA was easily detected in the cells infected with the virus stock obtained from DNA transfected cells, whereas this band was not seen in cells infected with virus derived from in vitro transcripts (Fig. 3B). Different from -21JBHK cells transfected with in vitro transcribed YFV RNA, the infectious YFV DNA pBeloBAC-FLYF has to be transcribed in the nucleus of cells, exposing the viral genome RNA to splicesomes. Computer-aided prediction revealed several potential splice donor and acceptor sites within the sequence of the YFV structural genes (Fig. 3C). A splicing event with the use of the predicted splice donor site at nt. 708 (score 1) and splice acceptor site at nt. 2440 (score 0.99) would result in an in-frame deletion of nt 708 to 2440 that could yield an RNA transcript that functioned as a YFV replicon. RT-PCR was performed on total intracellular RNA isolated from BHK-21J cells infected with the DNA launched virus to determine whether splicing at these predicted, favorable splice donor and acceptor sites was the origin of the YFV replicon-like RNA. Based on the predicted splice donor and acceptor sites, a forward primer corresponding to nt 40-64 within viral 5’ UTR and a reverse primer complementary to nt 2491-2513 in the NS1 gene

65 Chapter 3

15h 21h 27h 33h 39h 45h 1 2 3 4 A) B)

C)

Start End Score Exon Intron Start End Score Intron Exon 157 171 0,91 caatatg gtacgacg 858 898tgaggaaccccttttttgc 0,72 ag tgacggctctgaccattgcc 701 715 1 gcatatg gtaagtgt 2282 2322gtgtttggctctgcctttc 0,93 ag gggctatttggcggcttgaa 1132 1146 0,98 tgctgag gtgaggaa 2421 2461tcatgatgtttttgtctct 0,99 ag gagttggggcggatcaagga

D)

E) Growth Curve 1.00E+09 1.00E+08 1.00E+07 1.00E+06 1.00E+05 1.00E+04 r (pfu/ml) e 1.00E+03

Tit FLYF 1.00E+02 1.00E+01 G→C 1.00E+00 12h 24h 36h 48h 60h 72h 84h 96h

Time point

Figure 3. Characterization of YFV-17D replicon RNA that is produced by RNA splicing in pBeloBAC-FLYF transfected BHK cells Fig. 3 A) Analysis of intracellular YFV RNA synthesis by 3[H]-uridine labeling and gel electrophoresis. BHK cells transfected with pBeloBAC-FLYF were labeled for 6 hrs in the presence of Act.D with 6 hr interval from 15 hr p.t. onwards. B) Analysis of viral RNA synthesis in BHK cells infected with YFV-17D derived from in vitro transcribed RNA (lane 1), viruses derived from pBeloBAC-FLYF DNA (lane 3), or pBeloBAC-FLYF-G709C DNA (lane 4). Lane 2 is RNA from mock infected cells. C) Prediction of splice donor andacceptor sites in the sequence of the YFV structural genes. The splice donor and acceptor sequences that were used based on sequencing result shown in Fig. 3D are indicated in yellow and pink respectively. D) Part of the nucleotide sequence of the 750 bp PCR product demonstrating the in-frame deletion in the YFV structural genes resulting in the production of the replicon RNA. Nt 701-707 upstream of the splice donor site at nt. 708 are highlighted in yellow, while nt 2441-2450 that are downstream of the splice acceptor site at nt 2440 are highlighted in pink. E) Growth kinetics ofviruses derived from pBeloBAC-FLYF and pBeloBAC-FLYF-G709C. BHK-21J cells were infected at M.O.I. 1 and virus production was determined by plaque assays with supernatants collected at the indicated times p.i.

66 DNA launched YFV-17D were selected for this analysis. In addition to the approximately 2.5kb product that was expected to be produced with this primer set from full-length viral RNA, a smaller fragment with a length of around 750bp was amplified, among other less prominent bands (data not shown). The 750 bp product was gel purified and sequenced. Sequence analysis revealed an in frame deletion of nucleotides 708 – 2440 (Fig. 3D), strongly suggesting that the predicted splice donor and acceptor sites were indeed used on a fraction of the CMV driven transcripts in the nucleus which resulted in the production of a YFV replicon RNA. Considering the possible influence of this splicing event on the efficiency of this DNA launched system, a silent mutation (G→C) was introduced at position 709 to disrupt the splice donor site, and the resulting construct pBeloBAC-FLYF-G709C was transfected into BHK-21J cells in parallel with the original construct. As predicted, this mutation abolished the production of replicon RNA in the DNA-launched YFV-17D transfected cells, as the faster migrating band was not visible in the 3[H]-uridine gel (Fig. 3B, lane 4). Surprisingly, when virus production in BHK-21J cells transfected with pBeloBAC-FLYF-G709C and pBeloBAC-FLYF were compared, this silent point mutation was found to have an unexpected negative effect on viral production (Fig. 3E). Therefore, the immunogenicity studies were all conducted with the original construct.

In vivo characterization of DNA launched YFV-17D constructs

The induction of cellular and humoral immune responses against YFV-17D was investigated by delivering pBeloBAC-FLYF, pBeloBAC-YF/∆C and pBeloBAC-YF/GSA with an electroporation device to transgenic AAD mice using a homologous prime-boost immunization regimen. Three doses were used: 10ng, 100ng, or 1 μg. Two weeks after the boost immunization, serum was collected from the DNA or mock electroporated animals and neutralizing antibody titers against YFV were determined by PRNT. Mice were subsequently sacrificed and their spleens were collected and used to determine the frequency of IFN-γ producing cells after exposure to YFV peptides identified as dominant HLA-A2-restricted CD8+ and CD4+ T-cell epitopes in an ELISpot assay. A dose-dependent cellular immune response against YFV was clearly detectable in mice vaccinated with pBeloBAC-FLYF or pBeloBAC-YF/∆C (Fig. 4A and 4B). Irrespective of the amount of DNA used, mice electroporated with pBeloBAC-YF/GSA were unable to produce significant response when compared to PBS treated mice. Incubation of the splenocytes with the control peptide induced no significant response (Fig. 4C), confirming the specificity of the T cell response induced by DNA vaccination with pBeloBAC-FLYF or pBeloBAC-YF/∆C. Both CD8+ T-cell and CD4+ T-cell responses were highest following a dose of 1μg of priming and boosting, with levels comparable between the full-length construct and the capsid mutant (Fig. 4A and 4B).

67 A) CD8+ T cell response 300 10ng A) CD8+ T cell response 100ng µ 300 200 1 g 10ng 100ng splenocytes

6 µ 200 1 g 100 splenocytes 6 SFC/10 100 0 C

∆ SFC/10 PBS FLYF YF/ YF/GSA 0

∆C PBS FLYF YF/ Chapter 3 YF/GSA B) CD4+ T cell response 100 + 10ng A) CD8 T cell response B) CD4+ T cell response 80 100ng 300 µ 100 10ng1 g 10ng 60 100ng 80 100ng

splenocytes µ 6 20040 1 g 1µg 60

splenocytes 20 splenocytes 6 6 SFC/10 40 100 0 C 20 SFC/10

∆ SFC/10 PBS FLYF YF/ 0 YF/GSA 0 C ∆ ∆C PBS FLYF PBS FLYF YF/ F/GSA YF/ F/GSA Y Y C) Negative Control 100 10ng B) CD4+ T cell response C) Negative Control 80 100ng 100 µ 100 10ng1 g 10ng 60 80 100ng 80 100ng splenocytes 6 40 1µg 1µg 60 60 20 splenocytes splenocytes 6 6 SFC/10 40 40 0 20 C 20 SFC/10 SFC/10 ∆ PBS FLYF YF/ F/GSA 0 Y 0 Figure 4. YFV-specific T-cell responsesC C Fig. 4 ∆ ∆ PBS PBS FLYF YF/ FLYF YF/ Splenocytes from individual mice wereYF/GSA prepared for IFN-γ ELISpot analysis. Results are presentedYF/GSA as mean SFCs responding to a YFV-specific CD8+ T-cell peptide (CD8/C9V; CYNAVLTHV) (A),Fig. CD4+4 T-cell peptide (CD4/A15R; AIDLPTHENHGLKTR) (B), or a CD4+T-cell peptide (CD4/E15L; ERWFVRNPFFAVTAL) with greatly reduced capacity to stimulateC) YF17D-specificNegative Control CD4+ T cells as a negative control (C). 100 10ng 80 100ng 1µg 60 Intracellular cytokine staining (ICCS) was performed on CD8+ T cells to evaluate the splenocytes 6 40 quality of cellular immune response against YFV upon DNA immunization. Again, a dose- 20 dependentSFC/10 response was observed for YFV-specific CD8+ T-cells producing IL-2 and/or IFN-γ 0

(Fig. 5). For the dose-range∆C tested in this study, it was obvious that mice immunized with PBS FLYF YF/ the highest dose (1 µg) of eitherYF/GSA pBeloBAC-FLYF or pBeloBAC-YF/∆C produced the highest Fig. 4 percentage of cytokine secreting T-cells (Fig. 5).The ICCS analysis also revealed that although the majority of the CD8+ T-cells were producing only IFN-γ, there were also a fraction of CD8+ T-cells that secreted either only IL-2, or both IFN-γ and IL-2 (Fig. 5). In contrast to the induction of specific and mutli-functional T cell response, results of the PRNT for the evaluation of humoral immune response were rather disappointing. Irrespective of the vaccination dosage, no production of neutralizing antibodies was detected in mouse sera obtained two weeks after the boost (data not shown). To determine whether the murine innate immune response was restricting the production of YFV neutralizing antibodies, an additional experiment was performed using A129 mice that are deficient in expressing the IFNα/β receptor. The A129 and the congenic 129 mice were either electroporated with 5 μg of pBeloBAC-FLYF or injected i.m. with 106 PFU of YFV-

68 A) 0.200 IFN 0.175 IL-2 IFN + IL-2 0.150 A) 0.200 IFN

0.125 0.175 IL-2 IFN + IL-2

0.100 0.150

0.125 %CD8 T cells 0.075

0.050 0.100

0.025 %CD8 T cells 0.075

0.000 0.050 PBS FLYF YF/∆C YF/GSA DNA YFV constructs 0.025

0.000 PBS FLYFDNA l YF/a∆uCnched YF/GSA YFV-17D DNA YFV constructs B) 0.200 IFN 0.175 IL-2 IFN + IL-2 A) 0.2000.150 B) 0.200 IFN IFN 0.1750.125 IL-2 0.175 IL-2 IFN + IL-2 IFN + IL-2 0.1500.100 0.150

0.125 %CD8 T cells 0.1250.075

0.1000.050 0.100 %CD8 T cells %CD8 T cells 0.0750.025 0.075

0.0500.000 0.050 PBS FLYF YF/∆C YF/GSA 0.025 DNA YFV constructs 0.025

0.000 0.000 PBS FLYF YF/∆C YF/GSA PBS FLYF YF/∆C YF/GSA DNA YFV constructs DNA YFV constructs C) 0.5 IFN IL-2 IFN + IL-2 B) 0.2000.4 IFN C) 0.5 IFN 0.175 IL-2 IL-2 IFN + IL-2 0.3 IFN + IL-2 0.150 0.4

0.125 0.2 %CD8 T cells 0.3 0.100

0.1 %CD8 T cells 0.075 0.2 %CD8 T cells

0.050 0.0 PBS FLYF YF/∆C YF/GSA 0.1 0.025 DNA YFV constructs

0.000 Figure 5. IntracellularPBS cytokineFLYF staining YF/∆C analysis YF/GSA (ICCS) 0.0 Fig. 5 PBS FLYF YF/∆C YF/GSA DNA YFV constructs Splenocytes from mice vaccinated with 10ng of DNA (A), 100ng of DNA (B), or 1μgDNA of DNA YFV constructs (C) were prepared, and

ICCS results are presented as the mean percentage of splenocytes positive for IFN-γ,Fig. IL-2, 5 or both cytokines after in vitro incubation with a YFV-17D specific CD8+ T-cell peptide pool. Results are representative of two independent C) 0.5 IFN experiments. Means and standard errors are plotted.IL-2 IFN + IL-2 0.4

0.3 17D respectively. Serum was drawn every other day for the analysis of viremia and YFV 0.2 antibody%CD8 T cells production. No virus was detected in the sera of 129 mice that were transfected with pBeloBAC-FLYF,0.1 or infected with YFV-17D at any time during the experiment. For the

YFV-17D0.0 infected A129 mice, only serum samples collected on day 3 showed low levels of PBS FLYF YF/∆C YF/GSA viremia (103 and 3 xDNA 10 YFV3 PFU/mL). constructs Despite the transient and low level of viremia, YFV-17D virus vaccinated A129Fig. mice 5 produced significant amounts of neutralizing antibodies against YFV. The PRNT50 value of sera collected on day 9 was determined to be above 1280, which remained high till the end of the experiment (day 30). Despite the fact that no virus was detected at any time-point in the sera of A129 that were immunized with pBeloBAC-FLYF, these animals produced specific YFV-17D neutralizing antibodies. The PRNT50 value peaked on day 11 post immunization for these animals with values of 320 and 640 for the two subjects respectively (Fig. 6), although these were clearly lower than those observed for the YFV-17D infected A129 mice. On the contrary, the congenic, fully immuno-competent 129 mice did not produce detectable YFV neutralizing antibodies after immunization with pBeloBAC-FLYF.

69 Chapter 3

100%

90%

80%

70% D1 60% D3 50% D5 D7 40% D9 30% D11 D29 20% D14 D14 10% D11 D29 D9 0% D7 D5 D3 D1

Figure 6. Plaque reduction neutralization test (PRNT) results of A129 mice vaccinated with pBeloBAC-FLYF DNA Sera drawn from A129 mice at day 1, day 3, day 5, day 7, day 9, day 11, day 14 and day 29 post pBeloBAC-FLYF DNA vaccination were diluted and tested for YFV plaque reductionFig. neutralization 6 capacity as described in the Materials and Methods section. Percentages of subjects (mice) are plotted against PRNT50 value.

Discussion

Introduced in the early 1990s, the concept of vaccination using naked DNA has gained widespread recognition, due to its apparent advantages over conventional vaccines, such as ease of production and quality control as well as low cost of propagation [31;32;47]. The low immunogenicity in humans, however, is a major obstacle that has stymied the field [31;32;47]. In contrast, live-attenuated virus vaccines induce strong humoral and/or cellular immune responses. YFV-17D, the attenuated YFV strain that is used as a vaccine to protect against potentially lethal YFV infection is one of the most successful attenuated vaccines ever developed (reviewed in [7]). Attenuation was achieved by repeated passaging of the highly virulent YFV Asibi strain in suckling mice and chicken eggs [4]. In this report, the construction, molecular and immunological characterization of a DNA launched YFV-17D in the artificial bacterial chromosome pBeloBAC11 is described. The initial attempt was aimed at constructing a CMV promoter driven expression plasmid containing a full-length YFV-17D insert followed by a hepatitis delta virus ribozyme (HDVr) in the low copy number plasmid pACNR1180 [34;48]. Despite the success of using this plasmid as the vector for the stable propagation of YFV-17D full-length cDNA, incorporation of the

70 DNA launched YFV-17D

CMV promoter and the HDVr rendered this recombinant YFV plasmid instable in E. coli, as reflected by random deletions within the YFV insert (data not shown). The genetic instability of the insert was overcome by using pBeloBAC11 as a vector and the recombinant plasmid was proven to be stable for at least 20 passages in E. coli strains DH5α and DH10B. In addition to pBeloBAC-FLYF, two derivatives were constructed. In a replication incompetent pBeloBAC-YF/GSA, the viral RNA dependent RNA polymerase (RdRP) was inactivated by mutating the crucial GDD motif to GSA [41;42], whereas pBeloBAC-YF/∆C expresses a YFV replicon RNA that is capable of autonomously replicating its RNA, but requires in trans complementation of the capsid protein for virion formation [41]. These three BACmids were initially characterized in cell culture. Cells transfected with pBeloBAC-FLYF were able to produce infectious YFV virus, although the growth kinetics of virus production were approximately 24 hours delayed when compared to the kinetics of YFV production in BHK- 21J cells transfected with infectious YFV-17D RNA trancripts [34]. A severe handicap in YFV research is the lack of readily affordable small animal models that manifest clinical symptoms representative of what are seen in primates without the necessity of host adaptation of the virus [49]. Mice possessing functional innate immune systems are disease-resistant to both wildtype and vaccine strains of YFV after subcutaneous inoculation, with restricted viral dissemination. Neither morbidity nor mortality was exhibited [49]. Intracerebral inoculation can cause encephalitic disease in mice [50], which nonetheless lacks relevance to human infection that exhibits viscerotropism with possible lethal outcomes. Despite these drawbacks, mice are the only affordable small animal model for the initial testing of a DNA-launched YFV vaccination approach as presented inthis report. To compensate for a possible under-estimation of the immunogenicity of the YFV- 17D infectious DNA in mice due to murine restriction on YFV replication and dissemination [49;51], in vivo electroporation (EP) was employed to deliver DNA into HLA-A2.1 transgenic C57BL/6 mice, a method proven to logarithmically increase antigen expression and the resulting immune responses in various animals [44;52-58]. The same EP device was also successfully used in HIV-1 DNA vaccine human trials [59;60]. Since previous studies [19;24;25] with recombinant YFV-17D-based vaccines showed that anti-vector immunity was a minor concern during prime-boost immunization, a prime-boost vaccination regimen was adopted to further augment the immune responses. Nevertheless, in vivo delivery of YFV infectious DNA by EP failed to induce YFV neutralizing antibodies in HLA-A2.1 transgenic C57BL/6 mice, despite readily detectable CD8+ T cell responses, whereas YFV-17D vaccination of BALB/c mice was reported to yield a balanced Th1/Th2 immune response [28]. The mouse strain used could be one of the reasons for the skewed immune responses observed, since C57BL/6 mice and BALB/c mice are genetically predisposed to develop Th1 or Th2 immune responses, respectively [61]. Therefore, it will be interesting to test the YFV infectious DNA

71 Chapter 3 in BALB/c mice, or to try gene-gun for DNA delivery, which was demonstrated to direct the immune responses toward a predominantly Th2 profile, regardless of the genetic background of the mouse strain [62]. Another factor that could contribute to the Th1-biased immune response is the nature of the plasmid DNA that contains unmethylated CpG motifs, which was shown to induce lymphoid cells to release cytokines such as IFN-γ, IL-12, IL-18 [63;64]. Nevertheless, YFV infectious DNA is expected to produce virions in susceptible cells that should change the way of immune priming, and thus the cytokine profile. Unfortunately, we did not analyze the biodistribution of YFV antigens at the site of injection or the spread of virus to draining lymph nodes or other organs, so it could only be hypothesized that the murine myocytes and mononuclear cells might not support productive YFV-17D infection. The observation that the capsid mutant induced cellular immune responses with similar magnitudes compared to those induced by the wildtype construct further supported the aforementioned hypothesis, since the deficiency in virus assembly didn’t seem to influence the cellular immune responses elicited by vaccination. To further confirm whether the Th1-biased immune response was somehow correlated with the restriction of YFV production and dissemination by murine innate immune responses and the inability of YFV to evade these antiviral countermeasures in a non-natural host [49], humoral immune responses produced in A129 mice was analyzed in parallel with those produced in congenic 129 mice after YFV-17D or YFV infectious DNA vaccination. A129 mice that have deficiencies in innate immunity were proposed to represent a biologically more relevant animal model for studying viscerotropic infection and disease development following wildtype YFV inoculation, as well as mechanisms of 17D-204 vaccine attenuation [49]. In consistent with previous findings, viremia was detected in A129 mice inoculated with YFV-17D, whereas no viremia was detected in serum samples at any time point in YFV-17D vaccinated 129 mice, strongly indicating that murine type-I interferon signaling pathway is important in restricting YFV replication. Despite the lack of obvious viremia, A129 mice were able to mount an efficient humoral immune response after YFV DNA vaccination, which laid down the basis for carrying out more extensive immunization studies with these small animal models. To our knowledge, this is the first publication describing the successful launch of YFV-17D infection in vivo and the induction of YFV-specific cell-mediated immunity and neutralizing antibodies in experimental animals. Of course, to better evaluate the potential of this DNA launched YFV vaccine platform and its possible applications in human beings, the immunogenicity and efficacy as well as the safety profile of the DNA-based YFV vaccine should be more appropriately assessed in non-human primates. Besides, vaccination dosage needs to be adjusted based on the size and the susceptibility to YFV infection of the vaccinated species. Furthermore, novel DNA formulations and delivery routes could be explored to achieve better immunogenicity and acceptability. Although the EP method used

72 DNA launched YFV-17D in this study has been widely tested in a broad range of animals and has shown promising results in human clinical trials with DNA vaccines [59;60], intradermal EP delivery of DNA vaccines has started to attract more attention, due to the accessibility of the skin and the abundance of antigen-presenting cells (APCs) in the epidermis and dermis [65-67]. With the advances in DNA formulation and delivery technologies, the DNA launched YFV-17D vaccine may prove to be a stable, facile-to-produce and effective alternative for the current YFV live attenuated vaccine, and further holds promises as a platform for the development of YFV- 17D based recombinant DNA vaccines.

73 Chapter 3

Reference List

[1] Lindenbach B, Rice CM. Flaviviridae: The Viruses and Their Replication. In: Knipe D.M, Howley P.M., editors. Fields virology. 5 ed. Philadelphia, Lippincott Williams and Wilkins, 2007: p. 991-1042.

[2] Widman DG, Frolov I, Mason PW. Third-generation flavivirus vaccines based on single-cycle, encapsidation- defective viruses. Adv Virus Res 2008;72:77-126.

[3] WHO (World Health Organization). Yellow Fever fact sheet No.100 http://www.who.int/mediacentre/ factsheets/fs100/en/. 2009.

[4] Theiler M, Smith HH. The use of yellow fever virus modified byin vitro cultivation for human immunization. J. Exp. Med. 65, 787-800 (1937). Rev Med Virol 2000 Jan;10(1):6-16.

[5] Barrett AD, MonathTP, Barban V, Niedrig M, Teuwen DE. 17D yellow fever vaccines: new insights. A report of a workshop held during the World Congress on medicine and health in the tropics, Marseille, France, Monday 12 September 2005. Vaccine 2007 Apr 12;25(15):2758-65.

[6] Barrett AD, euwenT DE. Yellow fever vaccine - how does it work and why do rare cases of serious adverse events take place? Curr Opin Immunol 2009 Jun;21(3):308-13.

[7] Pulendran B. Learning immunology from the yellow fever vaccine: innate immunity to systems vaccinology. Nat Rev Immunol 2009 Oct;9(10):741-7.

[8] Monath TP. Yellow fever vaccine. Expert Rev Vaccines 2005 Aug;4(4):553-74.

[9] Querec TD, Akondy RS, Lee EK, et al. Systems biology approach predicts immunogenicity of the yellow fever vaccine in humans. Nat Immunol 2009 Jan;10(1):116-25.

[10] Gaucher D, Therrien R, Kettaf N, et al. Yellow fever vaccine induces integrated multilineage and polyfunctional immune responses. J Exp Med 2008 Dec 22;205(13):3119-31.

[11] Arroyo J, Miller C, Catalan J, et al. ChimeriVax-West Nile virus live-attenuated vaccine: preclinical evaluation of safety, immunogenicity, and efficacy. J Virol 2004 Nov;78(22):12497-507.

[12] Monath TP, Liu J, Kanesa-Thasan N, et al. A live, attenuated recombinant West Nile virus vaccine. Proc Natl Acad Sci U S A 2006 Apr 25;103(17):6694-9.

[13] Guirakhoo F, Zhang ZX, Chambers TJ, et al. Immunogenicity, genetic stability, and protective efficacy of a recombinant, chimeric yellow fever-Japanese encephalitis virus (ChimeriVax-JE) as a live, attenuated vaccine candidate against Japanese encephalitis. Virology 1999 May 10;257(2):363-72.

[14] Monath TP, Soike K, Levenbook I, et al. Recombinant, chimaeric live, attenuated vaccine (ChimeriVax) incorporating the envelope genes of Japanese encephalitis (SA14-14-2) virus and the capsid and nonstructural genes of yellow fever (17D) virus is safe, immunogenic and protective in non-human primates. Vaccine 1999 Apr 9;17(15-16):1869-82.

[15] Chambers TJ, Nestorowicz A, Mason PW, Rice CM. Yellow fever/Japanese encephalitis chimeric viruses: construction and biological properties. J Virol 1999 Apr;73(4):3095-101.

[16] Appaiahgari MB, Vrati S. IMOJEV((R)): a Yellow fever virus-based novel Japanese encephalitis vaccine. Expert Rev Vaccines 2010 Dec;9(12):1371-84.

[17] Galler R, Marchevsky RS, Caride E, et al. Attenuation and immunogenicity of recombinant yellow fever 17D-dengue type 2 virus for rhesus monkeys. Braz J Med Biol Res 2005 Dec;38(12):1835-46.

74 DNA launched YFV-17D

[18] Guirakhoo F, Arroyo J, Pugachev KV, et al. Construction, safety, and immunogenicity in nonhuman primates of a chimeric yellow fever-dengue virus tetravalent vaccine. J Virol 2001 Aug;75(16):7290-304.

[19] Guy B, Saville M, Lang J. Development of Sanofi Pasteur tetravalent dengue vaccine. Hum Vaccin 2010 Sep 16;6(9).

[20] Halstead SB, Thomas SJ. New Japanese encephalitis vaccines: alternatives to production in mouse brain. Expert Rev Vaccines 2011 Mar;10(3):355-64.

[21] Bonaldo MC, Garratt C,R Caufour PS, et al. Surface expression of an immunodominant malaria protein B cell epitope by yellow fever virus. J Mol Biol 2002 Jan 25;315(4):873-85.

[22] Tao D, Barba-Spaeth G, Rai U, Nussenzweig V, Rice CM, Nussenzweig RS. Yellow fever 17D as a vaccine vector for microbial CTL epitopes: protection in a rodent malaria model. J Exp Med 2005 Jan 17;201(2):201- 9.

[23] McAllister A, Arbetman AE, Mandl S, Pena-Rossi C, Andino R. Recombinant yellow fever viruses are effective therapeutic vaccines for treatment of murine experimental solid tumors and pulmonary metastases. J Virol 2000 Oct;74(19):9197-205.

[24] Bredenbeek PJ, Molenkamp R, Spaan WJ, et al. A recombinant Yellow Fever 17D vaccine expressing Lassa virus glycoproteins. Virology 2006 Feb 20;345(2):299-304.

[25] Jiang X, Dalebout TJ, Bredenbeek PJ, et al. Yellow fever 17D-vectored vaccines expressing Lassa virus GP1 and GP2 glycoproteins provide protection against fatal disease in guinea pigs. Vaccine 2011 Feb 1;29(6):1248-57.

[26] Stoyanov CT, Boscardin SB, Deroubaix S, et al. Immunogenicity and protective efficacy of a recombinant yellow fever vaccine against the murine malarial parasite Plasmodium yoelii. Vaccine 2010 Jun 23;28(29):4644-52.

[27] Van Epps HL. Broadening the horizons for yellow fever: new uses for an old vaccine. J Exp Med 2005 Jan 17;201(2):165-8.

[28] Franco D, Li W, Qing F, et al. Evaluation of yellow fever virus 17D strain as a new vector for HIV-1 vaccine development. Vaccine 2010 Aug 9;28(35):5676-85.

[29] Bonaldo MC, Mello SM, Trindade GF, et al. Construction and characterization of recombinant flaviviruses bearing insertions between E and NS1 genes. Virol J 2007;4:115.

[30] WHO. The yellow fever initiative, 2007. www.who.int/csr/disease/yellowfev/introduction/en/index/html

[31] Kutzler MA, Weiner DB. DNA vaccines: ready for prime time? Nat Rev Genet 2008 Oct;9(10):776-88.

[32] Weiner DB, Sardesai NY, Schmaljohn C. Introduction to DNA vaccines - Las Vegas. Vaccine 2010 Feb 23;28(8):1893-6.

[33] Hall RA, Nisbet DJ, Pham KB, Pyke AT, Smith GA, Khromykh AA. DNA vaccine coding for the full-length infectious Kunjin virus RNA protects mice against the New York strain of West Nile virus. Proc Natl Acad Sci U S A 2003 Sep 2;100(18):10460-4.

[34] Bredenbeek PJ, Kooi EA, Lindenbach B, Huijkman N, Rice CM, Spaan WJ. A stable full-length yellow fever virus cDNA clone and the role of conserved RNA elements in flavivirus replication. J Gen Virol 2003 May;84(Pt 5):1261-8.

75 Chapter 3

[35] Silva PA, Pereira CF, Dalebout TJ, Spaan WJ, Bredenbeek PJ. An RNA pseudoknot is required for production of yellow fever virus subgenomic RNA by the host nuclease XRN1. J Virol 2010 Nov;84(21):11395-406.

[36] Sambrook J, Fritsch T, Maniatis T. Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1989.

[37] Inoue H, Nojima H, Okayama H. High efficiency transformation of Escherichia coli with plasmids. Gene 1990 Nov 30;96(1):23-8.

[38] Rice CM, Lenches EM, Eddy SR, Shin SJ, Sheets RL, Strauss JH. Nucleotide sequence of yellow fever virus: implications for flavivirus gene expression and evolution. Science 1985 Aug 23;229(4715):726-33.

[39] Tilgner M, Shi PY. Structure and function of the 3’ terminal six nucleotides of the west nile virus genome in viral replication. J Virol 2004 Aug;78(15):8159-71.

[40] Charlier N, Molenkamp R, Leyssen P, et al. A rapid and convenient variant of fusion-PCR to construct chimeric flaviviruses. J Virol Methods 2003 Mar;108(1):67-74.

[41] Seregin A, Nistler R, Borisevich V, et al. Immunogenicity of West Nile virus infectious DNA and its noninfectious derivatives. Virology 2006 Dec 5;356(1-2):115-25.

[42] Khromykh AA, Varnavski AN, Sedlak PL, Westaway EG. Coupling between replication and packaging of flavivirus RNA: evidence derived from the use of DNA-based full-length cDNA clones of Kunjin virus. J Virol 2001 May;75(10):4633-40.

[43] van Dinten LC, den Boon JA, Wassenaar AL, Spaan WJ, Snijder EJ. An infectious arterivirus cDNA clone: identification of a replicase point mutation that abolishes discontinuous mRNA transcription. Proc Natl Acad Sci U S A 1997 Feb 4;94(3):991-6.

[44] Gardiner DF, Rosenberg T, Zaharatos J, Franco D, Ho DD. A DNA vaccine targeting the receptor-binding domain of Clostridium difficile toxin A. Vaccine 2009 Jun 2;27(27):3598-604.

[45] Maciel M, Jr., Kellathur SN, Chikhlikar P, et al. Comprehensive analysis of T cell epitope discovery strategies using 17DD yellow fever virus structural proteins and BALB/c (H2d) mice model. Virology 2008 Aug 15;378(1):105-17.

[46] Khromykh AA, Westaway EG. Subgenomic replicons of the flavivirus Kunjin: construction and applications. J Virol 1997 Feb;71(2):1497-505.

[47] Webster RG, Robinson HL. DNA vaccines: a review of developments. BioDrugs 1997 Oct;8(4):273-92.

[48] Ruggli N, Tratschin JD, Mittelholzer C, Hofmann MA. Nucleotide sequence of classical swine fever virus strain Alfort/187 and transcription of infectious RNA from stably cloned full-length cDNA. J Virol 1996 Jun;70(6):3478-87.

[49] Meier KC, Gardner CL, Khoretonenko MV, Klimstra WB, Ryman KD. A mouse model for studying viscerotropic disease caused by yellow fever virus infection. PLoS Pathog 2009 Oct;5(10):e1000614.

[50] Barrett AD, Gould A.E Comparison of neurovirulence of different strains of yellow fever virus in mice. J Gen Virol 1986 Apr;67 ( Pt 4):631-7.

[51] Sangster MY, Mackenzie JS, Shellam GR. Genetically determined resistance to flavivirus infection in wild Mus musculus domesticus and other taxonomic groups in the genus Mus. Arch Virol 1998;143(4):697-715.

[52] Khan AS, Pope MA, Draghia-Akli R. Highly efficient constant-current electroporation increases in vivo plasmid expression. DNA Cell Biol 2005 Dec;24(12):810-8.

76 DNA launched YFV-17D

[53] Medi BM, Hoselton S, Marepalli RB, Singh J. Skin targeted DNA vaccine delivery using electroporation in rabbits. I: efficacy. Int J Pharm 2005 Apr 27;294(1-2):53-63.

[54] Otten G, Schaefer M, Doe B, et al. Enhancement of DNA vaccine potency in rhesus macaques by electroporation. Vaccine 2004 Jun 23;22(19):2489-93.

[55] Tollefsen S, Vordermeier M, Olsen I, et al. DNA injection in combination with electroporation: a novel method for vaccination of farmed ruminants. Scand J Immunol 2003 Mar;57(3):229-38.

[56] Luxembourg A, Hannaman D, Ellefsen B, Nakamura G, Bernard R. Enhancement of immune responses to an HBV DNA vaccine by electroporation. Vaccine 2006 May 22;24(21):4490-3.

[57] Otten GR, Schaefer M, Doe B, et al. Potent immunogenicity of an HIV-1 gag-pol fusion DNA vaccine delivered by in vivo electroporation. Vaccine 2006 May 22;24(21):4503-9.

[58] Widera G, Austin M, Rabussay D, et al. Increased DNA vaccine delivery and immunogenicity by electroporation in vivo. J Immunol 2000 May 1;164(9):4635-40.

[59] Kopycinski J, Cheeseman H, Ashraf A, et al. A DNA-based candidate HIV vaccine delivered via in vivo electroporation induces CD4 responses toward the alpha4beta7-binding V2 loop of HIV gp120 in healthy volunteers. Clin Vaccine Immunol 2012 Sep;19(9):1557-9.

[60] Vasan S, Hurley A, Schlesinger SJ, et al. In vivo electroporation enhances the immunogenicity of an HIV-1 DNA vaccine candidate in healthy volunteers. PLoS One 2011;6(5):e19252.

[61] Hsieh CS, Macatonia SE, O’Garra A, Murphy KM. T cell genetic background determines default T helper phenotype development in vitro. J Exp Med 1995 Feb 1;181(2):713-21.

[62] Feltquate DM, Heaney S, Webster RG, Robinson HL. Different T helper cell types and antibody isotypes generated by saline and gene gun DNA immunization. J Immunol 1997 Mar 1;158(5):2278-84.

[63] Chu RS, Targoni OS, Krieg AM, Lehmann PV, Harding CV. CpG oligodeoxynucleotides act as adjuvants that switch on T helper 1 (Th1) immunity. J Exp Med 1997 Nov 17;186(10):1623-31.

[64] Roman M, Martin-Orozco E, Goodman JS, et al. Immunostimulatory DNA sequences function as T helper- 1-promoting adjuvants. Nat Med 1997 Aug;3(8):849-54.

[65] Brave A, Gudmundsdotter L, Sandstrom E, et al. Biodistribution, persistence and lack of integration of a multigene HIV vaccine delivered by needle-free intradermal injection and electroporation. Vaccine 2010 Nov 29;28(51):8203-9.

[66] Brave A, Nystrom S, Roos AK, Applequist SE. Plasmid DNA vaccination using skin electroporation promotes poly-functional CD4 T-cell responses. Immunol Cell Biol 2011 Mar;89(3):492-6.

[67] Lin F, Shen X, Kichaev G, et al. Optimization of electroporation-enhanced intradermal delivery of DNA vaccine using a minimally invasive surface device. Hum Gene Ther Methods 2012 Jun;23(3):157-68.

77 78 Chapter 4

An infectious Modoc virus cDNA as a tool to study conserved 3’-UTR RNA elements in flaviviruses with no known vector

Xiaohong Jiang, Patrícia A. G. C. Silva, Tim J. Dalebout, Charles M. Rice and Peter J. Bredenbeek

79 Chapter 4

Abstract

The Flavivirus genus can be divided into three different groups depending on the vector of transmission: i) mosquito-borne, ii) tick-borne, and iii) no known vector (NKV) flaviviruses. The third group is less-well studied, partly due to the lack of full-length cDNA clones. In this paper, the construction and characterization of the first infectious cDNA clone of an NKV flavivirus, Modoc virus (MODV) is described. The full-length MODV clone pACNR-FLMODV6.1 was shown to be genetically stable in E. coli. The viruses derived from this clone exhibited similar plaque morphology and growth kinetics as the parental MODV. Chemical probing was performed on the 3’-UTR region of the MODV genome, to corroborate the highly- ordered RNA structures predicted to be formed in its 3’-UTR. Subsequently, deletions and substitutions were introduced into the full-length clone to study the functional significance of these conserved RNA motifs and structures in the 3’-UTR of the NKV flaviviruses. The Y-shaped structure (region III) and the 3’ stem-loop structure (region IV) were found to be indispensable for viral replication, whereas more upstream regions were not absolutely required for viral RNA synthesis, although they are probably involved in other aspects of virus life cycle, such as cytopathogenicity and/or virus dissemination.

80 MODV infectious clone

Introduction

Based on phylogenetic analysis, members of theFlavivirus genus can be grouped into three clusters of related viruses that largely reflect their route of transmission. The clusters of mosquito-borne and tick-borne flaviviruses include important human pathogens such as dengue virus (DENV), yellow fever virus (YFV), and tick-borne encephalitis virus (TBEV). The third cluster, on the other hand, comprises less-well studied flaviviruses that have been exclusively isolated from bats (Montana myotis leukoencephalitis virus [MMLV] and Rio Bravo virus [RBV] [1]) and rodents (Modoc virus [MODV] [2] and Apoi virus [APOIV] [3]), and for which no arthropod vector has been implicated in transmission. An increasing number of these no known vector (NKV) flaviviruses have been isolated and sequenced [4-6]. Like arthropod-borne flaviviruses, NKV flaviviruses encode a single large polyprotein that is co- and post-translationally cleaved into three structural and seven non-structural proteins: C, prM/M, E, NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5 [4-6]. Modoc virus (MODV) was initiallyisolated from white-footed deer mouse (Peromyscus maniculatis) in Modoc County, California [7], and was shown to cause a persistent infection in rodents [8]. The virus has not been implicated in human disease, although a serological survey provided evidence for the occurrence of natural human infection without disease among inhabitants of Alberta, Canada [2]. MODV is neuroinvasive and causes lethal encephalitis in SCID mice and hamsters, rendering it a potential model for the study of flavivirus-induced encephalitis in humans [9]. The viral prM and/or E proteins were shown to be important for the neuroinvasive properties of MODV in SCID mice [10]. Protease and NTPase/helicase/RNA triphosphatase domains were identified in the NS3 of MODV [6], whereas conserved motifs associated with methyltransferase activity and RNA-dependent RNA polymerase (RdRp) domains were mapped to the MODV NS5 protein [6]. The crystal structure of the methyltransferase domain of MODV NS5 was recently solved [11]. The single ORF of the flavivirus genome is flanked by 5’- and 3’-untranslated regions (UTRs). Conserved RNA motifs and structures within the UTRs of mosquito-borne flaviviruses have been recognized as essential elements for genome replication, translation and/or pathogenesis, such as the well conserved stem-loop structure formed by the terminal ~100 nucleotides of the 3’-UTR (3’SL) and the pentanucleotide (PN) motif located in the top loop of the 3’SL, the conserved sequences CS1 and CS2, the RNA pseudoknot interactions PSK1 and PSK2, as well as the RNA structure involved in the production of subgenomic flavivirus RNA (sfRNA) [12-23]. Computer-aided sequence analysis and RNA folding of MODV, APOIV, RBV and MMLV genome revealed the presence of four regions within the distal part of the viral 3’-UTR that appeared to be conserved in RNA structures [5]. Region I was predicted to form a long hairpin

81 Chapter 4 with a branching stem–loop. The loop sequence within this region (nt 10,294 – 10,298) was proposed to be involved in a pseudoknot interaction with downstream nucleotides [24]. The formation of this particular type of pseudoknot was reported to be essential for the stalling of the host ribonuclease XRN-1 in the production of small flavivirus RNA (sfRNA) [21;22;24;25]. Region II was predicted to form a dumbbell like structure, the 3’ arm of which contains the CS2 sequence [5’-G(A/U)CUAGAGGUUAGAGGAGACCC-3’] that is well conserved among NKV flaviviruses and mosquito-borne flaviviruses [5;12]. CS2 was shown to be nonessential for replication of mosquito-borne flaviviruses in cell culture, although YFV-17D and DENV viruses lacking CS2 appear to be less pathogenic [13;14]. Region III was proposed to fold into a Y-shaped structure that bears resemblance to the Y-shaped structure present in tick-borne flaviviruses [26], while mosquito-borne flaviviruses seem to lack similar structures. The two loops of the “Y” structure are formed by a conserved stretch of nucleotides. However, the sequences of the stems carrying these loops are not conserved, but rather show a large number of compensatory base changes [5]. The 3’-terminal nucleotides of the 3’-UTR of the NKV flaviviruses were predicted to form a long stem-loop (3’ SL), a feature conserved among all flaviviruses. At the 5’ side of the 3’ SL, a small stem–loop was predicted which, together with 3’SL, belongs to region IV. Despite the predicted similarity in the RNA folding of this region, sequence conservation is restricted to the pentanucleotide (PN) motif and terminal dinucleotide only. The predicted structures for the 3’-UTR of the NKV flaviviruses have not been verified by RNA structure probing and investigation into their function(s) was hindered by the lack of infectious cDNA clones for the NKV flaviviruses. This chapter describes the construction and characterization of the first stable full- length NKV cDNA clone, which serves as a versatile tool to address basic questions concerning e.g. replication, tropism, and host-virus interactions of the NKV flaviviruses. In this study, the biological relevance of conserved RNA sequences and predicted RNA structures in NKV 3’-UTR was analyzed in detail. The predicted RNA topology of the MODV 3’-UTR was first corroborated by using chemical probing. Selected mutations were subsequently introduced into the full-length clone to study the functional significance of these conserved RNA elements in the MODV replicative cycle.

82 MODV infectious clone

Material and Methods

Cell culture and virus

The origin and culture conditions of the BHK-21J cells that were used throughout this study were described before [13]. The Modoc virus strain M544 was obtained from Prof. J. Neyts (Leuven, Belgium) during collaborative research [10] and was originally purchased from the American Tissue Culture Collection (ATCC, Manassas, USA). Stocks of MODV M544 were produced by infecting BHK-21J cells at a multiplicity of infection (MOI) of 0.1 in PBS containing 2% fetal calf serum (FCS) for 1 hr and subsequent incubation at 37°C with 5%

CO2 in DMEM/2%FCS. After 3 to 4 days, when cytopathic effect (CPE) became visible, the medium was harvested and centrifuged at 3,000 x g for 5 min to remove cellular debris. The supernatant was used as a virus stock. Stocks of the cDNA-derived viruses were obtained by electroporating BHK-21J cells with full-length RNA transcripts [27].

Recombinant DNA techniques and plasmid constructions

General standard nucleic acid methodologies were used throughout this study [28] unless described in more detail. Chemically competent E. coli DH5α cells were used for cloning [29]. Nucleotide numbering of the various constructs containing MODV-derived inserts and the resulting full-length clones was based on the MODV sequence deposited in GenBank (AJ242984 [6]). MODV cDNA was prepared using a one-step RT-PCR system containing a modified M-MLV reverse transcriptase for the cDNA reaction and a mixture of Taq polymerase and Pyrococcus GB-D polymerase (Invitrogen, Carlsbad, USA) for the PCR. RT-PCR reactions contained 1 µg of total RNA from MODV-infected cells. Reaction conditions were as suggested by the supplier. Oligonucleotides were designed based on the published MODV sequence. The most 5’ oligonucleotide (NKV41,fig. 1) contained the T7 Ф2.5 promoter [30], so that T7 RNA polymerase driven transcription would start on the “A” residue that is the first nucleotide of the MODV genome. In the oligonucleotide that hybridized to the extreme 3’ end of MODV (NKV40, Fig. 1) the complement of the two last viral nucleotides were fused to 3’ TAAG 5’ to yield an unique AflII restriction enzyme site. Relative positions of the oligonucleotides used for the cDNA reconstruction of MODV are depicted in Fig. 1. 3’-UTR mutants were constructed in the background of pACNR-FLMODV6.1. Quickchange mutagenesis was used for nucleotide deletion/substitution in the construction of pACNR-FLMODV-YFVpk3’, pACNR-FLMODV-YFVpk3, pACNR-FLMODV-YFVpk3’pk3, pACNR-FLMODV-Δ10370-10393, pACNR-FLMODV-ΔCS2, pACNR-FLMODV-AACC, pACNR- FLMODV-BB, pACNR-FLMODV-AA, pACNR-FLMODV-BA as well as pACNR-FLMODV-Δpn; while deletion of the predicted RNA secondary structures, i.e., region I, region II, region III,

83 Chapter 4 region IV, was achieved by engineering unique restriction enzyme recognition sites at the ends of the PCR products that were joined with the use of these restriction sites.

RNA transcription

Plasmid DNA for in vitro run-off RNA transcription was purified using the Nucleobond AX DNA isolation kit (Macherey-Nagel, Düren, Germany). pACNR-FLMODV plasmids were linearized with AflII, followed by proteinase K treatment and phenol/chloroform extraction. Approximately 1 µg of linearized DNA was used as a template for in vitro transcription using the Ampliscribe™ T7 high-yield transcription kit (Epicentre, Madison, USA). For the production of 5’-capped full-length MODV transcripts, the UTP, GTP and CTP concentration was 7.5 mM, whereas the ATP concentration was adjusted to 2 mM. G(5’)ppp(5’)A (NEB, Ipswich, USA) was added as RNA cap analog to a final concentration of 6 mM. After a 2 hr incubation at 37°C, DNase I was added and the incubation was continued for another 15 min. The RNA transcripts were subsequently purified by LiCl precipitation and the concentration was determined by spectrophotometry.

RNA transfection and analysis of viral RNA synthesis

BHK-21J cells were transfected with 5 µg of full-length MODV RNA as described previously [27]. For RNA analysis, 2.5 ml (approximately 1.5 x 106 cells) of the transfected BHK-21J cell suspension was seeded in a 10 cm2 plate. Total RNA was isolated from the transfected cells at 30 h post electroporation (p.e). Analysis of RNA synthesis by 3[H]-uridine labeling was performed as described before [13]. Tripure (Roche, Mannheim, Germany) was used for cell lysis and subsequent RNA isolation.3 [H]-Uridine labelled RNAs were denatured with glyoxal and analyzed on 0.8% agarose gels [28].

Northern blotting and hybridization

Samples containing 10 µg of total RNA from electroporated cells were denatured using formaldehyde, separated in a formaldehyde-containing 1.5% agarose gel, and blotted onto a Hybond-N+ membrane (GE Healthcare, Buckinghamshire, UK). The blots were hybridized with 32P-labeled oligonucleotides as described previously [31;32].

RNA structure determination by selective 2’-hydroxyl acylation and primer extension (SHAPE) probing

A pBluescript plasmid containing MODV nucleotides 9651-10506 was used to generate a PCR fragment of 383 nts, with oligonucleotide NKV45 that contained a T7 promoter sequence fused to nts complementary to MODV 10244-10269 and M13 reverse oligonucleotide as

84 MODV infectious clone primers. The resulting PCR product was used as template for in vitro RNA transcription (T7 MEGAscript kit; Ambion). Transcripts were treated with DNase I and purified by LiCl precipitation as recommended by Ambion, and used for SHAPE probing as described previously [22;33]. NMIA-induced modifications in the MODV transcript were identified by primer extension using 32P-labeled oligonucleotide T3 primer, NKV 20, NKV 53 and NKV 161, which binds to the T3 promoter, nt 10332-10356 (linker between region I and II), nt 10399- 10420 (the CS2 motif), and nt 10433-10452 (linker between region II and III and the 5’ arm of the bottom stem of region III) respectively.

Immunofluorescence microscopy

At 30 h p.e., control and infected cells were washed once with PBS and prepared for immunofluorescence microscopy as described previously [13]. Commercially available immune ascitic fluid obtained from mice infected with MODV (ATCC, Manassas, USA) was used as primary antibody in a 1:1000 dilution. Alexa Fluor® 488 goat anti-mouse IgG (Invitrogen) was used in a 1:500 dilution as secondary antibody.

Immunostaining of plaque assays

After the regular 4 days of incubation, the Avicel overlay was removed and the cells were fixed with 3.5% formaldehyde for 20 min, followed by 3 washes with PBS-Glycine (10mM), and permeabilized with 0,5% Triton X-100 in PBS for 15 min. After 3 additional washes with PBS, the individual 10 cm2-wells were incubated with 200 μl of a 1:500 dilution of the MODV hyperimmune serum in PBS/10% horse serum/0,05% Tween-80 for 1 h, followed by 3 x 5 min washes with PBS/0.05% Tween-80, and subsequent incubation with Goat-anti-Mouse-HRP (Dako, Glostrup, Denmark), diluted 1:1000 in PBS/10% horse serum/0.05% Tween-80 for 1h. After 3 washes with PBS/0.05% Tween-80 and a short rinse with 0.1 M Sodium Acetate buffer (pH 5.0), cells were incubated with 3-amino-9-ethylcarbazole as substrate in the dark at RT for >30 min. The peroxidase reaction was stopped by washing with milli-Q water when foci of infected cells were clearly visible.

Viral growth kinetics determination by qRT-PCR

For analysis of the viral growth kinetics, BHK-21J cells were infected at an MOI of 1 and the medium was subsequently collected and replaced by the same volume of fresh medium at 6-h intervals. Plaque assays were used when comparing the growth kinetics of parental MODV and the MODV-6.1(wt) viruses, whereas qRT-PCR was used to determine the growth kinetics of the MODV 3’-UTR mutants. Briefly, Triton X-100 was added to the collected medium to a final concentration of 0.1% for the disruption of virion structure and thus releasing the viral genome. 5 μl of the treated samples were used without any additional

85 Chapter 4 purification steps in a qRT-PCR reaction using a one-step RT-PCR kit (Qiagen) and primers as previously described [9].

Results

Construction and characterization of a full-length MODV cDNA clone

The construction of full-length cDNAs for the transcription of infectious flavivirus RNA was often hampered by genetic instability of these clones in E. coli [34-39]. Given the positive experience with a pACNR1180-derived vector [40] for the construction of a stable full- length YFV-17D clone [13], this low-copy number vector was selected as the vector for the construction of an infectious MODV cDNA.

The full-length MODV cDNA was assembled in a modified pACNR-FLYF17Da vector [13], in which most of the YFV sequences were deleted while simultaneously creating uniqueNot I, HindIII, and SalI restriction sites. Three RT-PCR products (Fig. 1A) were sequentially inserted in the 5’ to 3’ direction to create the full-length MODV cDNA. Primer NKV41 contained the T7 RNA polymerase φ2.5 promoter sequence fused to the first 19 nucleotides of MODV 5’-UTR, whereas in oligonucleotide NKV40, nucleotides complementary to the last 28 nt of the MODV genome were fused to an AflII recognition sequence, which served as run-off site for in vitro transcription. Upon transfection into BHK-21J cells, in vitro-generated RNA transcripts of several plasmids harbouring the full-length MODV cDNA were analyzed for viral RNA synthesis and protein expression by 3[H]-uridine labeling and immunofluorescence assays, respectively. Among all the analyzed clones, the transcripts of pACNR-FLMODV 6.1 yielded higher virus titers compared to the other tested clones (data not shown) and virus derived from this clone was characterized and compared in more detail to the parental MODV. The genetic stability of pACNR-FLMODV6.1 in E. coli strain DH5α was evaluated by repeated passaging. One passage was defined as growing the bacteria for more than 12 h in 2 ml of LB medium containing 50 μg/ml ampicillin, followed by a streak on selective medium to obtain a single colony for the next cycle. After the 10th streak, a bacterial colony was picked and used to prepare plasmid DNA (pACNR-FLMODV6.1-p10). No differences were observed in immunofluorescence and 3[H]-RNA labeling (fig.1B and 1C) of cells transfected with RNA transcripts of the original plasmid and the passage 10 plasmid. More importantly, viruses derived from both pACNR-FLMODV6.1 and pACNR-FLMODV6.1-p10 showed very similar growth kinetics and plaque morphology in BHK-21J cells when compared to the parental MODV (Fig. 1D and 1E).

86 MODV infectious clone

A)

B) C) Mock MODV MODV6.1 MODV6.1-P10

Mock MODV wt

MODV 6.1 MODV 6.1 p10

D) Growth Curve E) 1.0E+08 Plaque Morphology

1.0E+07

1.0E+06 parental MODV MODV6.1-P0 MODV6.1-P10

1.0E+05

1.0E+04

Log (Pfu/ml) Log 1.0E+03

1.0E+02 MODV 6.1-P10 1.0E+01 MODV 6.1 MODV WT 1.0E+00 6 12 18 24 30 36 42 48

hrs post electroporation Figure 1. Construction and characterization of the pACNR-MODV6.1 infectious cDNA clone A) Schematic representation of theonstruction c of the MODV full-length clone. The large boxes represent the viral ORF encoding theFig. structural1 and non-structural proteins. The oligonucleotides that were used to generate the cDNA fragments for constructing the clone are indicated by triangles. The open circle upstream of the MODV insert represents the T7 Ф2.5 promoter. The restriction sites that were utilized for the assembly of the clone as well as the Sal I site that was deleted are indicated. B) Immunofluorescence staining of BHK-21J cells infected with wt-MODV or transfected with in vitro transcribed full-length MODV RNA derived from pACNR-MODV6.1 and the passage 10 plasmid respectively. Cells were fixed at 30 hrs p.i. or p.e. and stained with the MODV hyperimmune serum. C) Viral RNA synthesis in BHK-21J cells transfected with in vitro transcribed RNA of the pACNR-MODV6.1 (lane2), pACNR-MODV6.1-p10 (lane 3). RNA isolated from cells infected with parental MODV virus (lane 2) and that from mock infected cells (lane 1) were loaded as controls. D) Viral growth kinetics. BHK-21J cells were infected with the parental MODV (diamond), virus derived from pACNR-MODV6.1 (square) or virus derived from pACNR-MODV6.1-p10 (triangle) at an MOI of 1; the medium of the infected cells was harvested at the indicated time points p.i. and the viral titer was determined by plaque assays. E) Plaque morphology of the parental MODV and virus derived from pACNR-MODV6.1 or pACNR-MODV6.1-p10 in BHK-21J cells.

87 Chapter 4

The nucleotide sequence of pACNR-FLMODV6.1 was determined and compared to the only other full-length MODV sequence (accession number AJ242984) that was available in GenBank. As summarized in table 1, the 22 nucleotide differences detected when comparing pACNR-FLMODV6.1 and the Genbank entry were scattered throughout the coding sequence. Most amino acid substitutions (8) were present in NS5, the largest virus-encoded protein. Surprisingly, the rather small (254-amino acid) NS4B protein, which is generally well conserved among flaviviruses, contained 4 amino acid substitutions. None of the mutations found in NS3 or NS5 were located within the functional domains that were previously proposed for these viral proteins [6].

MODV 3’-UTR folds into four discrete stem-loop regions.

As schematically depicted in Fig. 2A, MODV 3’-UTR as well as the distal part of the 3’-UTR of other NKV viruses, was proposed to fold into four highly ordered RNA structures [5]. To verify the predicted RNA topology of the MODV 3’-UTR, nt 9,651-10,,506 were subjected to selective 2’-hydroxyl acylation and primer extension (SHAPE) probing. This region covers the complete 3’-UTR except for the terminal ~100 nt, which are predicted to fold into the well conserved long stem-loop structure (3’SL) designated as region IV [5] (Fig. 2A). The susceptibility of the nucleotides to N-methylisatoic anhydride (NMIA) treatment is summarized in Fig. 2C, with colors indicating strong (red), weak (yellow) or noisy (blue) reactivity. Apart from nucleotides comprising the CS2 motif, the bases predicted tobe involved in basepairing showed little or no reactivity to NMIA, supporting their participation in higher-order RNA structures. Nucleotides predicted to be present in loop structures also correlated rather well with the chemical probing data, although their reactivity varied. Strong reactivity was observed for most of the nucleotides positioned in between region I and II, confirming the presence of a non-structured linker sequence between the depicted stem-loop structures. Discrepancy between chemical probing result and the RNA structure prediction, however, was noticed for nucleotides separating region II and III, which showed weak or no reactivity, instead of high reactivity that was expected for the predicted single- stranded conformation. To investigate the roles that these primary and secondary RNA structures play in viral life cycle, deletion and substitution were introduced into these regions. The constructed mutants are listed in Table 2, where detailed information is provided on the nucleotides that were deleted or mutated.

88 MODV infectious clone

Table 1. Summary of the nucleotide differences between the MODV AJ242984 and the MODV clone 6.1 sequence Changes are grouped by encoded viral proteins. Positions, actual nucleotide change as well as the amino acid substitutions are shown. Nucleotide or amino acid to the right indicates MODV NCBI AJ242984 followed by the nucleotide or amino acid encountered in the MODV6.1 genome.

Gene Position Nucleotide Amino Acid prM 610 U  C silent

Env 1543 U  G Phe  Leu

NS1 2776 C  U silent 3089 C  A Gln  Arg

NS2A 3529 A  G Ile  Met

NS2B 4410 A  G Glu  Gly

NS3 4861 A  G silent 6312 U  A Leu  Gln

NS4B 6837 U  C Ile  Thr 7098 G  A Ser  Asn 7444 C  G silent 7445 C  G Leu  Val 7503 U  A Leu  His

NS5 7756 G  A silent 7767 G  C Ser  Thr 7938 G  C Arg  Thr 8141 C  G Arg  Ala 8142 G  C Arg  Ala 8612 C  G Gln  Glu 8920 U  A Ser  Arg 9120 A  G Lys  Arg 9990 A  G Asp  Gly

89 ChapterA) 4 B) + - A C G T

A) RegionIV B) + - A C G T

PNRegionIV

RegionI RegionII RegionIII

PN PK’ UUGG RegionI RegionII RegionIII

PK’ UUGG

CS2

PK

CS2

PK

C)

C U A G G GC strong reactivity C) GC weak reactivity UA 10463 GC noisy UA A U 10400 CCCA AU G G G U G G G G G C A A ACU A G AGG C strong reactivity ACGC A C C C A C U U A GCG C C G G U GC G A U weak reactivity UA A 10463A A GGCC G CG noisy AU 10400A UA CA A 10300A U U U C GC U G G GU G U G C UA G U G UU GG GC GC A U U A A C C U C UG GG G ACU A G AGG C AC A C C C A C U GU A GC• CG U UG• C G U G U UG CG G C C C GC C A U GC A G G G A CGA C C C AGA C CAGA GC U U A U A GC A A AU G C G CG GC U GUAU• A GC UA GC 10300 10420U U AUGC 10370U CG 10450UA GC G U G UU G GC GC UGCU A C C U C GU GG G CG GC AU AU GC G • CG CG U UG• U A UG CG AUGC A G G A GA C C C A C C C GCAUC CAG GC AUU UGA GUAAAUACUUUGGCCAGUCAUUGUG U AAAUAGGUUAGGGC A AU U A GCUUUGAAGUUAGGGCAAC UAUUUUUU GU• GC 10420 GCGC AU 10370 CG 10450 GC 10252GC10320 10336 10350 CG AU 10492 AU GC CG AU AU AU FigureUGA 2. PredictionGUAAAUACUUUGGCCAGUCAUUGU and verification of RNA secondaryAAAUAGGUUAGGG structure ofUUUGAAGUUAGGGCAAC MODV 3’-UTR UAUUUUU A) Schematic representation of predicted RNA structure of MODV 3’-UTR. Sequences conserved in the 3’-UTR 10336 10350 10492 10252of NKV flaviviruses10320 are highlighted: CS2 (orange), the Pentanucleotide motif (PN) (red) and the UUGG loop in Region III (purple). The sequences (PK’ and PK) involvedFig. 2 in the formation of the predicted RNA pseudoknot are highlighted in blue. B) A representative radiograph of a SHAPE probing experiment. Template RNA was produced and treated with NMIA and analyzed as described in the MaterialFig. and 2Method section. Primer extension was performed using primer NKV 161 that binds to nt10,433-10,452 of the MODV genome. Primer extension was performed on NMIA treated (+) or not treated (-) RNA. C) Results of SHAPE probing for nt 10,252–10,492 within MODV 3’-UTR. NMIA reactivity is summarized with colors indicating differences in reactivity: red, strong reactivity; yellow, weak reactivity; blue, similar reactivity in mock and NMIA treated RNA; no color indicates that these nucleotides were not reactive to NMIA.

90 MODV infectious clone

Table 2. Characteristics of MODV 3’-UTR mutants constructed and analyzed in this study Column one lists all the mutants constructed. Details of the nucleotides substituted or deleted are shown in column two, and the positions of changed nucleotides within the full-length genome are indicated in column three.

MODV mutants Number of deleted (Δ) or Position of the mutated nts. in Construct substituted (x→y) nts. the MODV genome 1 pACNR-FLMODV-Δregion IV Δ 93 10506 – 10599 2 pACNR-FLMODV-Δpn Δ 5 10552 – 10556 3 pACNR-FLMODV-Δregion III Δ 39 (+AGAT) 10447 – 10485 4 pACNR-FLMODV-AACC 4 (TTGG → AACC) 10460 – 10463 5 pACNR-FLMODV-BB 6 (TTGGTG → CGCCAA) 10451 – 10456 6 pACNR-FLMODV-AA 6 (CGCCAA → TTGGTG) 10468 – 10473 7 pACNR-FLMODV-BA 6 (TTGGTG → CGCCAA) 10451 – 10456 6 (CGCCAA → TTGGTG) 10468 – 10473 8 pACNR-FLMODV-Δregion II Δ 64 10360 – 10423 9 pACNR-FLMODV-Δ10370-10393 Δ 24 10370 – 10393 10 pACNR-FLMODV-ΔCS2 Δ 20 10400 – 10419 11 pACNR-FLMODV-Δregion I Δ 67 (+ AAGCTT) 10253 – 10319 12 pACNR-FLMODV-YFVpk3’ 5 (ATGAC → GCTGT) 10294 – 10298 13 pACNR-FLMODV-YFVpk3 5 (GTCAT → ACAGC) 10336 – 10340 14 pACNR-FLMODV-YFVpk3’pk3 5 (pk3’; ATGAC → GCTGT) 10294 – 10298 5 (pk3; GTCAT → ACAGC) 10336 – 10340

The 3’-terminal stem-loop structure is indispensable for MODV replication.

Deletion of the 3’-terminal SL structure (region IV according to the nomenclature of [5]) was lethal to the virus, as no viral RNA synthesis was detected when 3[H]-RNA labeling was performed on BHK-21J cells transfected with the construct in which the whole region IV was deleted (Fig. 3A, lane 8), nor could any virus be detected by plaque assays with the medium collected from the transfected cells. In fact, the mere deletion of the MODV 5’ CUCAG 3’ pentanucleotide (PN) motif (nt 10552–10556) was sufficient to completely abolish viral RNA synthesis (Fig. 3A, lane 9).

The Y-shaped RNA structure upstream of the 3’ SL is required for MODV replication.

The proposed RNA folding of region III into a Y-shaped structure correlated well with the results of the SHAPE probing (Fig. 2C). Nucleotides in the loop region were quite accessible

91 A) 1 2 3 4 5 6 7 8 9

Chapter 4

A) 1 2 3 4 5 6 7 8 9

B) MODV-6.1 (wt) MODV-AACC MODV-BA

B) MODV-6.1 (wt) MODV-AACC MODV-BA

Figure 3. Characteristics of the MODV RegionIII and Region IV mutants A) Viral RNA synthesis was analyzed by 3[H]-RNA labeling as described in the Material and Methods section. BHK- 21J cells were infected (M.O.I. 1) with MODV-6.1 (wt) (lane 2), MODV-Δregion III (lane 3), FLMODV-AACC (lane 4), MODV-BB (lane 5), MODV-AA (lane 6), MODV-BA (lane 7), MODV-Δregion IV (lane 8), MODV-ΔPN (lane 9) and mock infected (lane 1). B) Plaque morphology (upper panel) and foci formation (lower panel) of MODV-6.1(wt) (column 1), MODV-AACC (column 2), and MODV-BA (column 3) on BHK-21J cells. Plaque assay and immunostaining methods were performed as described in the Materials and Methods section. Fig. 3 to modification by NMIA, while nucleotides in the predicted stems were in general non- reactive, with discrepancies being observed for only a few nucleotides. This structure was shown to be absolutely required for viral replication, since no RNA synthesis was detected in cells transfected withFig. MODV-ΔregionIII 3 RNA (Fig. 3A, lane 3). To dissect the functional elements within this region, additional mutations were introduced to either change the loop sequence 5’ UUGG 3’ that is conserved in all NKV flaviviruses,

92 MODV infectious clone to 5’ AACC 3’ (highlighted in purple, Fig. 2A), or to disrupt the base-pairing of the stem (highlighted in green, Fig. 2A). Substituting the sequence 5’ GUGGUU 3’ (nt. 10,451 – 10,456) with 5’ CGCCAA 3’, which is present at position 10,473-10,468, or vice versa, resulted in the construction of mutants MODV-BB and MODV-AA, respectively. Base pairing of the 5’ stem of region III was predicted to be disrupted in these two mutants, whereas the base pairing potential was restored in mutant MODV-BA, in which the 5’ and 3’ arm of the duplex were swapped. As shown in Fig. 3A, no viral RNA synthesis was detected in cells transfected with either MODV-BB (lane 5) or MODV-AA RNA (lane 6). In contrast, cells that were electroporated with the MODV-BA transcript showed efficient viral RNA synthesis (Fig. 3A, lane 7), underlining the importance of this stem-loop structure for MODV replication. Accordingly, no virus was detected in the medium of the cells transfected with either the MODV-BB or the -AA transcript, whereas cells electroporated with the MODV-BA transcript produced similar amounts of virus with identical plaque and foci morphology as the wt MODV-infected cells (Fig. 3B). Although the conserved loop sequence 5’ UUGG 3’ was dispensable for viral replication, its mutagenesis dramatically influenced the efficiency of RNA synthesis (Fig. 3A, lane 4). Virus derived from MODV-AACC RNA transfection produced smaller plaques and foci on BHK-21J cells compared to wtMODV (Fig. 3B). From these results it can be concluded that this particular RNA stem-loop structure is important for optimal MODV replication.

The CS2-containing dumbbell-like RNA structure is nonessential for viral replication.

The SHAPE probing results (Fig. 2C) did not fully support the predicted RNA structure for region II. Although the lack of reactivity to NMIA of the nucleotides that formed the bottom and the longer 5’ stem structure of region II was in agreement with the predicted RNA structure, most nucleotides within the CS2 motif reacted strongly with NMIA, indicating that this part of region II was not involved in the formation of secondary or any higher-order RNA structures. The CS2 motif was therefore depicted as a single-stranded loop in Fig. 2C. Three deletion mutants erew constructed to analyze the role of region II in the viral life cycle. In MODV-ΔregionII, the complete sequence of region II (nt 10360-10423) was deleted, while in MODV-Δ10370-10393 and MODV-ΔCS2 the 5’ arm of the dumbbell-like structure and the CS2 motif were deleted respectively. Surprisingly, cells transfected with RNA from these three mutants all showed a similar level of viral RNA synthesis as cells electroporated with the wt MODV control transcript (Fig. 4A). Crystal violet staining of MODV- ΔregionII plaque assays did not show visible plaques on BHK-21J cells whereas plaque assays for MODV-Δ10370-10393 and MODV-ΔCS2 revealed the formation of rather small and not well-defined plaques on BHK-21J cells (Fig. 4B). However, these viruses were able to infect

93 A) 1 2 3 4 5 Chapter 4 A) 1 2 3 4 5

A) 1 2 3 4 5

B) MODV-6.1(wt) MODV-∆RegionII MODV-∆10370-10393 MODV-∆CS2

B) MODV-6.1(wt) MODV-∆RegionII MODV-∆10370-10393 MODV-∆CS2

B) MODV-6.1(wt) MODV-∆RegionII MODV-∆10370-10393 MODV-∆CS2

C) MODV Region II mutants C) 7.0 MODV Region II mutants 6.0 C) 7.0 5.0 6.0 MODV Region II mutants 7.0 4.0 5.0 6.0 3.0 Log pfu/ml 4.0

5.0 Wildtype 2.0 3.0 Log pfu/ml deltaΔ Region Region II II 4.0 1.0 Wildtypedelta10370-10391Δ10370-10391 2.0 deltaΔ CS2 CS2Region II 3.0 Log pfu/ml 0.0 1.0 delta10370-10391 12 18 24 30 36Wildtype 42 48 2.0 delta CS2 0.0 hrs p.i. delta Region II 1.0 12 18 24 30 36delta10370-10391 42 48 Figure 4. Characteristics of the MODV Region II mutantshrs p.i. delta CS2 0.0 A) Analysis of MODV RNA synthesis by 3[H]-RNA labeling. BHK-21J cells were infected (M.O.I. 1) with MODV-6.1(wt) 12 18 24 30 36 42 48

(lane 1), MODV-Δregion II (lane 3), MODV-Δ10370-10393hrs p.i. (lane 4), MODV-ΔCS2 (lane 5) and mock infected (lane 2). RNA was labeled, isolated, and analyzed as described in the Materials and Methods section. B) Analysis of plaque (upper panel) and infectious center formation (lower panel) on BHK-21J cells by MODV- 6.1(wt) (column 1), MODV-Δregion II (column 2), MODV-Δ10370-10393 (column 3) and MODV-ΔCS2 (column 4). Plaque assays and immunostaining methods were performed as described in the Materials and Methods section. C) Viral growth kinetics of the MODV 3-UTR region II mutants. BHK-21J cells were infected with MODV-6.1(wt), MODV-Δregion II, MODV-Δ10370-10393 and MODV-ΔCS2 at an MOI of 1; the medium of the infected cells was harvested at the indicated time points p.i. and the amount of viral genome copies was determined by qRT-PCR as described in the Materials and Methods section.

94 MODV infectious clone

BHK-21J cells efficiently as was demonstrated by the clearly visible foci of virus-infected cells upon immunostaining of plaque assays on BHK-21J cells. The diameter of the foci formed by MODV-ΔCS2 was similar to those produced by wt MODV, and MODV-Δ10370-10393 produced foci were only slightly smaller than those produced by wt MODV. Compared to the wt virus, significantly smaller yet still clearly visible foci were produced by ΔregionII mutant (Fig. 4B). The differences in plaque forming capability and foci morphology, however, were not associated with virus production efficiency. Similar growth kinetics was observed for all these three mutants, which produced marginally lower titers of viruses from 18 to 36 h post infection compared to wtMODV (Fig. 4C).

Region I is involved in sfRNA production.

The results of the SHAPE probing correlated rather well with the RNA structure predicted for region I, especially for the nucleotides that were predicted to form the stems of this particular RNA structure. Surprisingly little or even no reactivity to NMIA was observed for the nucleotides that were predicted to form the loop regions of the predicted structure. For instance the sequence 5’ CAACC 3 ’ (nt 10269-10273) that was predicted to form a single-stranded bulge in between two stems did not show reactivity to NMIA (Fig. 2C). Interestingly, the sequence 5’ AUGAC 3’ (nt 10,294-10,298) that was predicted toform an RNA pseudoknot [24] by baseparing with the downstream sequence 5’ GUCAU 3’ (nt 10,336-10,340) showed a weak but reproducible reactivity to NMIA. For the predicted pseudoknot partner 5’ GUCAU 3’, three (GUC) out of the five nucleotides did not react to NMIA, while the other two nucleotides reacted strongly. These results suggested that the predicted RNA pseudoknot was at least not the thermodynamically favored RNA structure under the conditions used for these probing experiments. Despite the ambiguous probing results for these particular nucleotides, these sequences were found to be essential for MODV sfRNA production. No MODV sfRNA could be detected when the pk’ sequence (Fig. 2A, nomenclature according to [20]) AUGAC was mutated to GCUGU (sequence involved in pseudoknot formation and sfRNA production in YFV 3’-UTR [20;22]; Fig. 5B). Likewise, mutagenesis of the predicted interacting pk sequence GUCAU to ACAGC (interaction partner for GCUGU in the YFV 3’-UTR) also abolished sfRNA production (Fig. 5B). Surprisingly, MODV-YFVpk3’pk3 in which the above-mentioned mutations were combined to restore the predicted RNA pseudoknot formation, was still unable to produce detectable amounts of MODV sfRNA, as was MODV Δregion I mutant. Viral genomic RNA synthesis, on the other hand, appeared to be only slightly compromised for all of the above-mentioned mutants, as judged by the intensity of the MODV genomic RNA band in the Northern blot and the results of the intracellular 3[H]-uridine labeling of viral RNA (Fig. 5A and 5B). Interestingly, BHK- 21J cells electroporated with in vitro transcribed RNA of MODV-DregionI and the MODV

95 Chapter 4

A) 1 2 3 4 5 6 B) 1 2 3 4 5 6

C) MODV-6.1(wt) MODV-∆RegionI

D)

MODV Region I mutants 7.0

6.0

5.0

4.0

3.0 Log pfu/mlLog Wildtype ∆RI 2.0 YFV pk3'

1.0 YFV pk3 YFV pk3'3 0.0 12 18 24 30 36 42 48

hrs p.i.

Figure 5. Characteristics of the MODV Region I mutants 3 A) Analysis of viral RNA synthesis was analyzed by [H]-RNAFig. 5labeling. BHK-21J cells were infected (M.O.I. 1) with MODV-6.1(wt) (lane 1), MODV-Δregion I (lane 3), MODV-YFVpk3’ (lane 4), MODV-YFVpk3 (lane 5), pACNR- FLMODV-YFVpk3’pk3 (lane 6) or mock infected (lane 2). RNA was labeled and analyzed as described in the Materials and Methods section. B) Analysis of MODV sfRNA production by Northern blotting. BHK-21J cells were infected (M.O.I. 1) with MODV- 6.1(wt) (lane 1), MODV-Δregion I (lane 2), MODV-YFVpk3’ (lane 3), MODV- YFVpk3 (lane 4), MODV-YFVpk3’pk3 (lane 5) and FLMODV-Δregion II (lane 6). Intracellular RNA was isolated at 30hrs p.i. and analyzed by Northern blotting as described in the Materials and Methods section. C) Analysis of plaque (upper panel) and infectious center formation (lower panel) on BHK-21J cells by MODV- 6.1(wt) (column 1) and MODV-Δregion I (column 2). Plaque assays and immunostaining were done as described in the Materials and Methods section. D) Viral growth kinetics. BHK-21J cells were infected with MODV-6.1(wt), MODV-Δregion I, MODV-YFVpk3’, MODV- YFVpk3, or MODV- YFVpk3’pk3 at an MOI of 1; the medium of the infected cells was harvested at the indicated time points p.i. and the amount of viral genome copies was determined by qRT-PCR as described in the Materials and Methods section.

96 MODV infectious clone pk mutants did not show any visual CPE at time points when CPE was clearly detectable in wtMODV RNA transfected cells (data not shown). In addition, no plaques were observed by crystal violet staining on BHK-21J cells for any of these MODV mutants that were unable to produce sfRNA. Nevertheless, immunostaining of cells infected with these MODV region I mutants showed clearly detectable yet small foci (Fig. 5C). The viral growth kinetics of MODV region I mutants on BHK-21J cells revealed that these viruses produced virus titers 5- to 10-fold lower compared to those of wtMODV during the exponential phase of the growth curve (Fig. 5D).

Discussion

Unlike for the mosquito- and tick-borne flaviviruses, no arthropod vectors have been identified for NKV flaviviruses thus far. It is therefore interesting to study the similarities of this group of flaviviruses compared to their vector-borne counterparts, as well as to identify unique features that distinguish them from arbo-flaviviruses. Unfortunately, these basic studies have been hindered by the lack of infectious full-length cDNA clones for NKV flaviviruses, while those of vector-borne flaviviruses have significantly facilitated research on yellow fever virus (YFV) [13;36], Kunjin virus (KUNV) [41], West Nile virus (WNV) [42], dengue virus (DENV) [43;44] and tick-borne encephalitis virus (TBEV) [45]. Construction of stable full-length clones for flaviviruses, however, has not always been straightforward, as genetic instability in E. coli hosts was often encountered. For certain flaviviruses [35- 37], labor-intensive in vitro ligation procedures were used initially for the production of full-length cDNA templates for infectious RNA transcripts, until the instability problem was circumvented by cloning the cDNA into low-copy number plasmid vectors [13;34]. The low-copy number plasmid pACNR1180, previously used for the cloning of viral sequences that are not well tolerated by E. coli [13;40], was therefore chosen as the vector for the construction of full-length MODV clone. Clone pACNR-FLMODV6.1 was proven to be genetically stable upon repeated passages inE. coli. Attempts to clone the full-length MODV insert from pACNR-FLMODV6.1 into high copy number plasmids like pBluescript or pUC met no success (unpublished results). Viruses derived from the full-length clone (pACNR- FLMODV6.1) possessed similar characteristics as the parental MODV, in terms of genome replication efficiency, growth kinetics, and plaque morphology, establishing a reverse genetics system for the study of MODV molecular biology. Besides the functional analysis of RNA structures present in the 3’-UTR as described in this chapter, the construction of chimeric flaviviruses by exchanging the prM and E genes of NKV and arthropod-borne flaviviruses could create valuable tools to elucidate the molecular basis for flavivirus host-

97 Chapter 4 range restrictions and pathogenicity, as illustrated by studies using a yellow fever virus in which prM and E were replaced by their counterparts of MODV [10;46]. These studies concluded that the neuroinvasive character of MODV in mice was mediated by prM and E, while the inability of MODV to replicate in arthropod hosts was determined by a post-entry event [10;46]. The availability of an infectious MODV cDNA allows construction of similar MODV chimeric viruses in which the MODV prM and E are replaced by the comparable genes of YFV or other flaviviruses to corroborate the results obtained with the YFV-MODVprM-E chimera. Flavivirus genomes contain a single open reading frame (ORF) flanked by 5’-UTR and 3’-UTR. The 3’-UTR varies in length from about 350 to 800 nucleotides. This heterogeneity in length is mainly due to the variation in the so-called variable region (VR), which is located proximal to the stop codon. It is characterized by extensive sequence duplications and deletions, even among strains of the same virus, and by an apparent lack of sequence conservation. In contrast, the distal part of the 3’-UTR exhibits more sequence and/or structure similarities among flaviviruses. Since this part of the 3’-UTR contains RNA elements essential for viral translation, replication, and possibly assembly [47-54], it is referred to as the ‘core element’. With a length of only 366 nucleotides, MODV has the shortest 3’-UTR among all flaviviruses sequenced thus far. In comparison to other flaviviruses, the MODV 3’-UTR comprises only the core element and probably represents the minimal requirements for the 3’-UTR of NKV flaviviruses. The region was predicted to fold into four well-conserved RNA structures [5]. In this study, SHAPE RNA probing was used to validate three out of the four predicted RNA structures for the MODV 3’-UTR. Except for the CS2 sequence in region II, probing results were generally in agreement with the predicted structures. These RNA structures were deleted and mutated accordingly, and the effect of these mutations on viral RNA synthesis and virus productions was analyzed to determine the role of these RNA elements in the viral life cycle. The well-conserved 3’-SL and the PN motif (within region IV) were indispensable for MODV replication, as was shown for vector-borne flaviviruses [13;42;55-58]. Using a primer extension analysis and an alphavirus-based expression system [24], region I was previously predicted to serve as the stalling site for the host ribonuclease XRN-1, resulting in the production of the MODV sfRNA [21;22;24]. As demonstrated in this study, region I is non-essential for viral RNA synthesis or virus production but plays a critical role in stalling the host ribonuclease XRN-1, resulting in the production of sfRNA that is characteristic for every flavivirus [21;22;24]. Although mutation of the sequences proposed to be involved in pseudoknot formation [24] illustrated the importance of these nucleotides in sfRNA production, SHAPE probing results did not confirm their participation in pseudoknot formation. This discrepancy between the biochemical probing results and the model proposed could be due to the fact that the RNA structure probing involved only part of the

98 MODV infectious clone viral 3’-UTR and that folding of this particular region might be different in the context of the whole viral genome and the presence of viral and/or host factors interacting with the viral 3’-UTR. An alternative explanation which poses more challenges for future research is that the current model of the folding of region I or in more general terms the XRN-1 stalling site in the 3’-UTR of the flavivirus genome is too simplistic and might be incorrect. After all it is rather surprising that such a relatively simple RNA structure is able to stall the highly processive XRN-1. In vivo probing with a recently described method that allowed structural determination of RNA molecules in a physiologically more relevant setting would hopefully provide more insights into the RNA structure of this region in living cells [59], and it would be interesting to determine the 3D structure of the MODV or a related flavivirus XRN-1 stalling site by NMR. In between region I (XRN-1 stalling site) and region IV (3’SL), two other conserved highly-ordered RNA structures were predicted for all NKV flaviviruses [5]. Interestingly, the structure of region II is predicted to be similar to the CS2-containing structure in mosquito- borne flaviviruses which appears to be absent in tick-borne flaviviruses, whereas a conserved Y-shaped RNA structure similar to the region III of NKV flaviviruses has been identified in tick-borne flaviviruses and has no counterparts in mosquito-borne flaviviruses. It appeared that the conserved structures of region II and III in NKV flaviviruses also bear functional similarities to their counterparts of the vector-borne flaviviruses. Region II encompassing the CS2 containing dumbbell-like RNA structure is not required for RNA synthesis and virus production. Given the effect of mutations in MODV region II on cytopathogenicity in cell culture, the RNA elements within this region are probably involved in virus-host interactions. Similar observations have been reported for mutations involving the CS2-containing region of DENV and YFV [13;14;60]. The Y-shaped structure (region III) that has a structurally and positionally similar counterpart in TBEV is, just like in TBEV, indispensable for viral replication [13;14;60;61]. Using the first infectious cDNA clone for a NKV flavivirus, this study provides valuable information on the functional significance of conserved RNA sequences/structures in the life cycle of these flaviviruses. It still remains to be elucidated at which stage of the viral life cycle the changes of these motifs and/or structures exert their influence and which viral or cellular proteins are involved in the interaction. Besides, it is still beyond our understanding how an element promoting cytopathology could benefit NKV flaviviruses that in general establish persistent infection in mammalian hosts. Moreover, the identification of insect-specific flaviviruses [62-64] has added a novel angle to the analysis of flavivirus host restriction. Hopefully the access to full-length clones representing each cluster of flaviviruses will help to solve these basic questions in flavivirus biology.

99 Chapter 4

Acknoledgements

We thank Prof. J. Neyts (Leuven, Belgium) for providing the Modoc virus strain M544.

100 MODV infectious clone

Reference List

[1] Constantine DG, WoodallDF. Latent infection of rio bravo virus in salivary glands of bats. Public Health Rep 1964 Dec;79:1033-9.

[2] Zarnke RL, Yuill TM. Modoc-like virus isolated from wild deer mice (Peromyscus maniculatus) in Alberta. J Wildl Dis 1985 Apr;21(2):94-9.

[3] Varelas-Wesley I CCH. Antigenic relationships of flaviviruses with undetermined arthropod-borne status. Am J Trop Med Hyg. 1982 Nov;31(6), 1273-1284. 1982.

[4] Billoir F, de CR, Tolou H, de MP, Gould EA, de L, X. Phylogeny of the genus flavivirus using complete coding sequences of arthropod-borne viruses and viruses with no known vector. J Gen Virol 2000 Sep;81 Pt 9:2339.

[5] Charlier N, Leyssen P, Pleij CW, et al. Complete genome sequence of Montana Myotis leukoencephalitis virus, phylogenetic analysis and comparative study of the 3’ untranslated region of flaviviruses with no known vector. J Gen Virol 2002 Aug;83(Pt 8):1875-85.

[6] Leyssen P, Charlier N, Lemey P, et al. Complete genome sequence, taxonomic assignment, and comparative analysis of the untranslated regions of the Modoc virus, a flavivirus with no known vector. Virology 2002 Feb 1;293(1):125-40.

[7] Johnson HN. Ecological implications of antigenically related mammalian viruses for which arthropod vectors are unknown and avian associated soft tick viruses. Jpn J Med Sci Biol 1967 Dec;20 Suppl:160-6.

[8] Davis JW, Hardy JL. Characterization of persistent Modoc viral infections in Syrian hamsters. Infect Immun 1974 Aug;10(2):328-34.

[9] Leyssen P, Van LA, Drosten C, Schmitz H, De CE, Neyts J. A novel model for the study of the therapy of flavivirus infections using the Modoc virus. Virology 2001 Jan 5;279(1):27-37.

[10] Charlier N, Molenkamp R, Leyssen P, et al. Exchanging the yellow fever virus envelope proteins with Modoc virus prM and E proteins results in a chimeric virus that is neuroinvasive in SCID mice. J Virol 2004 Jul;78(14):7418-26.

[11] Jansson AM, Jakobsson E, Johansson P, et al. Structure of the methyltransferase domain from the Modoc virus, a flavivirus with no known vector. Acta Crystallogr D Biol Crystallogr 2009 Aug;65(Pt 8):796-803.

[12] Hahn CS, Hahn YS, Rice CM, et al. Conserved elements in the 3’ untranslated region of flavivirus RNAs and potential cyclization sequences. J Mol Biol 1987 Nov 5;198(1):33-41.

[13] Bredenbeek PJ, Kooi EA, Lindenbach B, Huijkman N, Rice CM, Spaan WJ. A stable full-length yellow fever virus cDNA clone and the role of conserved RNA elements in flavivirus replication. J Gen Virol 2003 May;84(Pt 5):1261-8.

[14] Men R, Bray M, Clark D, Chanock RM, Lai CJ. Dengue type 4 virus mutants containing deletions in the 3’ noncoding region of the RNA genome: analysis of growth restriction in cell culture and altered viremia pattern and immunogenicity in rhesus monkeys. J Virol 1996 Jun;70(6):3930-7.

[15] Khromykh AA, Meka H, Guyatt KJ, Westaway EG. Essential role of cyclization sequences in flavivirus RNA replication. J Virol 2001 Jul;75(14):6719-28.

101 Chapter 4

[16] Lo MK, Tilgner M, Bernard KA, Shi PY. Functional analysis of mosquito-borne flavivirus conserved sequence elements within 3’ untranslated region of West Nile virus by use of a reporting replicon that differentiates between viral translation and RNA replication. J Virol 2003 Sep;77(18):10004-14.

[17] Corver J, Lenches E, Smith K, et al. Fine mapping of a cis-acting sequence element in yellow fever virus RNA that is required for RNA replication and cyclization. J Virol 2003 Feb;77(3):2265-70.

[18] Alvarez DE, De Lella Ezcurra AL, Fucito S, Gamarnik AV. Role of RNA structures present at the 3’UTR of dengue virus on translation, RNA synthesis, and viral replication. Virology 2005 Sep 1;339(2):200-12.

[19] Alvarez DE, Lodeiro MF, Luduena SJ, Pietrasanta LI, Gamarnik AV. Long-range RNA-RNA interactions circularize the dengue virus genome. J Virol 2005 Jun;79(11):6631-43.

[20] Olsthoorn RC, Bol JF. Sequence comparison and secondary structure analysis of the 3’ noncoding region of flavivirus genomes reveals multiple pseudoknots. RNA 2001 Oct;7(10):1370-7.

[21] Pijlman GP, Funk A, Kondratieva N, et al. A highly structured, nuclease-resistant, noncoding RNA produced by flaviviruses is required for pathogenicity. Cell Host Microbe 2008 Dec 11;4(6):579-91.

[22] Silva PA, Pereira CF, Dalebout TJ, Spaan WJ, Bredenbeek PJ. An RNA pseudoknot is required for production of yellow fever virus subgenomic RNA by the host nuclease XRN1. J Virol 2010 Nov;84(21):11395-406.

[23] Manzano M, Reichert ED, Polo S, et al. Identification of cis-acting elements in the 3’-untranslated region of the dengue virus type 2 RNA that modulate translation and replication. J Biol Chem 2011 Jun 24;286(25):22521-34.

[24] Silva PA, Dalebout TJ, Gultyaev AP, Olsthoorn RC, Bredenbeek PJ. Characterization of the sfRNAs that are produced in cells infected with no known vector and cell fusing agent. unpublished.

[25] Funk A, Truong K, Nagasaki T, et al. RNA structures required for production of subgenomic flavivirus RNA. J Virol 2010 Nov;84(21):11407-17.

[26] Gritsun TS, Gould EA. Origin and evolution of 3’UTR of flaviviruses: long direct repeats as a basis for the formation of secondary structures and their significance for virus transmission. Adv Virus Res 2007;69:203- 48.

[27] Silva PA, Molenkamp R, Dalebout TJ, et al. Conservation of the pentanucleotide motif at the top of the yellow fever virus 17D 3’ stem-loop structure is not required for replication. J Gen Virol 2007 Jun;88(Pt 6):1738-47.

[28] Sambrook J, Fritsch T, Maniatis T. Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1989.

[29] Inoue H, Nojima H, Okayama H. High efficiency transformation of Escherichia coli with plasmids. Gene 1990 Nov 30;96(1):23-8.

[30] Coleman TM, Wang G, Huang F. Superior 5’ homogeneity of RNA from ATP-initiated transcription under the T7 phi 2.5 promoter. Nucleic Acids Res 2004;32(1):e14.

[31] Meinkoth J, Wahl G. Hybridization of nucleic acids immobilized on solid supports. Anal Biochem 1984 May 1;138(2):267-84.

[32] van der Most RG, Bredenbeek PJ, Spaan WJ. A domain at the 3’ end of the polymerase gene is essential for encapsidation of coronavirus defective interfering RNAs. J Virol 1991 Jun;65(6):3219-26.

102 MODV infectious clone

[33] Merino EJ, Wilkinson KA, Coughlan JL, Weeks KM. RNA structure analysis at single nucleotide resolution by selective 2’-hydroxyl acylation and primer extension (SHAPE). J Am Chem Soc 2005 Mar 30;127(12):4223- 31.

[34] Gualano RC, Pryor MJ, Cauchi MR, Wright PJ, Davidson AD. Identification of a major determinant of mouse neurovirulence of dengue virus type 2 using stably cloned genomic-length cDNA. J Gen Virol 1998 Mar;79 ( Pt 3):437-46.

[35] Kapoor M, Zhang L, Mohan PM, Padmanabhan R. Synthesis and characterization of an infectious dengue virus type-2 RNA genome (New Guinea C strain). Gene 1995 Sep 11;162(2):175-80.

[36] Rice CM, Grakoui A, Galler R, Chambers TJ. Transcription of infectious yellow fever RNA from full-length cDNA templates produced by in vitro ligation. New Biol 1989 Dec;1(3):285-96.

[37] Sumiyoshi H, Hoke CH, Trent DW. Infectious Japanese encephalitis virus RNA can be synthesized from in vitro-ligated cDNA templates. J Virol 1992 Sep;66(9):5425-31.

[38] Yamshchikov V, Mishin V, Cominelli F. A new strategy in design of +RNA virus infectious clones enabling their stable propagation in E. coli. Virology 2001 Mar 15;281(2):272-80.

[39] Yamshchikov VF, Wengler G, Perelygin AA, Brinton MA, Compans RW. An infectious clone of the West Nile flavivirus. Virology 2001 Mar 15;281(2):294-304.

[40] Ruggli N, Tratschin JD, Mittelholzer C, Hofmann MA. Nucleotide sequence of classical swine fever virus strain Alfort/187 and transcription of infectious RNA from stably cloned full-length cDNA. J Virol 1996 Jun;70(6):3478-87.

[41] Khromykh AA, Westaway EG. Completion of Kunjin virus RNA sequence and recovery of an infectious RNA transcribed from stably cloned full-length cDNA. J Virol 1994 Jul;68(7):4580-8.

[42] Shi PY, Tilgner M, Lo MK. Construction and characterization of subgenomic replicons of New York strain of West Nile virus. Virology 2002 May 10;296(2):219-33.

[43] Kinney RM, Butrapet S, Chang GJ, et al. Construction of infectious cDNA clones for dengue 2 virus: strain 16681 and its attenuated vaccine derivative, strain PDK-53. Virology 1997 Apr 14;230(2):300-8.

[44] Lai CJ, Zhao BT, Hori H, Bray M. Infectious RNA transcribed from stably cloned full-length cDNA of dengue type 4 virus. Proc Natl Acad Sci U S A 1991 Jun 15;88(12):5139-43.

[45] Mandl CW, Ecker M, Holzmann H, Kunz C, Heinz FX. Infectious cDNA clones of tick-borne encephalitis virus European subtype prototypic strain Neudoerfl and high virulence strain Hypr. J Gen Virol 1997 May;78 ( Pt 5):1049-57.

[46] Charlier N, Davidson A, Dallmeier K, Molenkamp R, De CE, Neyts J. Replication of not-known-vector flaviviruses in mosquito cells is restricted by intracellular host factors rather than by the viral envelope proteins. J Gen Virol 2010 Jul;91(Pt 7):1693-7.

[47] Mandl CW, Kunz C, Heinz FX. Presence of poly(A) in a flavivirus: significant differences between the 3’ noncoding regions of the genomic RNAs of tick-borne encephalitis virus strains. J Virol 1991 Aug;65(8):4070- 7.

[48] Wallner G, Mandl CW, Kunz C, Heinz FX. The flavivirus 3’-noncoding region: extensive size heterogeneity independent of evolutionary relationships among strains of tick-borne encephalitis virus. Virology 1995 Oct 20;213(1):169-78.

103 Chapter 4

[49] Poidinger M, Hall RA, Mackenzie JS. Molecular characterization of the Japanese encephalitis serocomplex of the flavivirus genus. Virology 1996 Apr 15;218(2):417-21.

[50] Wang E, Weaver SC, Shope RE, Tesh RB, Watts DM, Barrett AD. Genetic variation in yellow fever virus: duplication in the 3’ noncoding region of strains from Africa. Virology 1996 Nov 15;225(2):274-81.

[51] Gritsun TS, Venugopal K, Zanotto PM, et al. Complete sequence of two tick-borne flaviviruses isolated from Siberia and the UK: analysis and significance of the 5’ and 3’-UTRs. Virus Res 1997 May;49(1):27-39.

[52] Proutski V, Gould EA, Holmes EC. Secondary structure of the 3’ untranslated region of flaviviruses: similarities and differences. Nucleic Acids Res 1997 Mar 15;25(6):1194-202.

[53] Rauscher S, Flamm C, Mandl CW, Heinz FX, Stadler PF. Secondary structure of the 3’-noncoding region of flavivirus genomes: comparative analysis of base pairing probabilities. RNA 1997 Jul;3(7):779-91.

[54] Mandl CW, Holzmann H, Meixner T, et al. Spontaneous and engineered deletions in the 3’ noncoding region of tick-borne encephalitis virus: construction of highly attenuated mutants of a flavivirus. J Virol 1998 Mar;72(3):2132-40.

[55] Khromykh AA, Kondratieva N, Sgro JY, Palmenberg A, Westaway EG. Significance in replication ofthe terminal nucleotides of the flavivirus genome. J Virol 2003 Oct;77(19):10623-9.

[56] Elghonemy S, Davis WG, Brinton MA. The majority of the nucleotides in the top loop of the genomic 3’ terminal stem loop structure are cis-acting in a West Nile virus infectious clone. Virology 2005 Jan 20;331(2):238-46.

[57] You S, Falgout B, Markoff L, Padmanabhan R. In vitro RNA synthesis from exogenous dengue viral RNA templates requires long range interactions between 5’- and 3’-terminal regions that influence RNA structure. J Biol Chem 2001 May 11;276(19):15581-91.

[58] Yu L, Markoff L.The topology of bulges in the long stem of the flavivirus 3’ stem-loop is a major determinant of RNA replication competence. J Virol 2005 Feb;79(4):2309-24.

[59] Spitale RC, Crisalli P, Flynn RA, Torre EA, Kool ET, Chang HY. RNA SHAPE analysis in living cells. Nat Chem Biol 2013 Jan;9(1):18-20.

[60] Lo MK, Tilgner M, Bernard KA, Shi PY. Functional analysis of mosquito-borne flavivirus conserved sequence elements within 3’ untranslated region of West Nile virus by use of a reporting replicon that differentiates between viral translation and RNA replication. J Virol 2003 Sep;77(18):10004-14.

[61] Hoenninger VM, Rouha H, Orlinger KK, et al. Analysis of the effects of alterations in the tick-borne encephalitis virus 3’-noncoding region on translation and RNA replication using reporter replicons. Virology 2008 Aug 1;377(2):419-30.

[62] Stollar V, Thomas VL. An agent in the Aedes aegypti cell line (Peleg) which causes fusion of Aedes albopictus cells. Virology 1975 Apr;64(2):367-77.

[63] Crabtree MB, Sang RC, Stollar V, Dunster LM, Miller BR. Genetic and phenotypic characterization of the newly described insect flavivirus, Kamiti River virus. Arch Virol 2003 Jun;148(6):1095-118.

[64] Hoshino K, Isawa H, Tsuda Y, et al. Genetic characterization of a new insect flavivirus isolated from Culex pipiens mosquito in Japan. Virology 2007 Mar 15;359(2):405-14.

104 Chapter 5

Mutagenesis and biochemical analysis of the XRN-1 stalling signal for subgenomic flavivirus (sf)RNA production

Xiaohong Jiang, Tim J. Dalebout and Peter J. Bredenbeek

105 Chapter 5

Abstract

Cells and mice infected with arthropod-borne flaviviruses produce a small subgenomic RNA that is co-linear with the viral 3’-untranslated region (3’-UTR). This small subgenomic flavivirus RNA (sfRNA) is generated through incomplete degradation of the viral genome by the host 5’→3’ exoribonuclease XRN-1, and has been implicated in cytopathogenicity and pathogenesis of the virus. A stem-loop structure stabilized by a pseudoknot interaction was determined to be essential for sfRNA production in yellow fever virus (YFV) and West Nile virus (WNV). Here, we show that a similar RNA structure with an equivalent function in the 3’-UTR of other members of the Flavivirus genus could replace that of YFV in stalling XRN-1 for the production of sfRNA. Compared to typical H-type pseudoknots, these XRN-1 stalling signals of mosquito-borne flaviviruses are featured with a side stem-loop structure and a short linkage sequence between the stem-loop and the downstream pseudoknotting partner. In this study, mutations were introduced within the context of the YFV infectious clone so as to change the nucleotides conserved among mosquito-borne flaviviruses or to disrupt predicted secondary structures. All mutants constructed in this manner were replication competent. The side stem-loop was found to be indispensable for XRN-1 stalling, albeit with no sequence specificity. The short linker also contributes to the stability of the complex RNA structure that is responsible for XRN-1 stalling. In addition, mutation of the conserved nucleotides in the bottom stem (e1) region either diminished sfRNA production, or decreased cytopathogenicity of the virus in Vero cells, suggesting their involvement in stabilizing the XRN-1-stalling structure and/or in cytopathogenicity.

106 RNA structural features for sfRNA production

Introduction

The flaviviruses form a genus within the Flaviviridae that comprises more than 80 viruses. Phylogenetically, flaviviruses can be grouped into three clusters of related viruses that are in general also unified by their route of transmission: (i) mosquito-borne (e.g., West Nile virus [WNV] and yellow fever virus [YFV]), (ii) tick-borne, and (iii) no known vector (NKV) flaviviruses (such as Modoc virus [MODV]) [1]. In addition to the viruses assigned to the above-mentioned clusters within the Flavivirus genus, there is a rapidly expanding group of viruses that are currently considered tentative species of the Flavivirus genus. This group includes cell fusing agent virus (CFAV), Kamiti River virus (KRV) and Culex flavivirus (CxFV) that have all been exclusively isolated from mosquitoes or insect cell lines [2-4]. There is no evidence that these viruses are able to infect a vertebrate host and therefore they are often referred to as insect-specific flaviviruses. Flaviviruses are small enveloped viruses with a positive-sense single-stranded RNA genome of approximately 11 kb in length, which is 5’ capped, yet not polyadenylated at its 3’ terminus. The RNA encodes a single polyprotein of approximately 3500 amino acids that is co- and post-translationally processed by viral and cellular proteases into three structural proteins (C, prM, and E) and seven nonstructural proteins (NSs). This single ORF is flanked by 5’- and 3’-untranslated regions (UTRs) (reviewed in [5]). The flavivirus 3’-UTR comprises a variable region (VR) that is located proximal to the translation termination codon and is characterized by low sequence similarity and the presence of the distal ‘core element’, which contains conserved RNA elements essential for viral translation, replication, pathogenesis and possibly assembly [6-13]. Based on its position within the genome, sequence/structure similarity, as well as functional implications, the core element of mosquito-borne flavivirus can be divided into three domains: 1) the ‘promoter’ region that is indispensable for genome replication and includes the 3’ stem-loop structure (3’SL) and conserved sequence 1 (CS1); 2) an enhancer region that is nonessential for replication yet enhances RNA synthesis and/or translation efficiency, consisting of the CS2 (conserved sequence 2) containing dumb-bell like structure and its duplicate(s) in the case of some mosquito-borne flaviviruses; 3) the XRN-1 stalling structure that is required for subgenomic flavivirus (sf)RNA production. The sfRNAs were initially detected in mice infected with Murray Valley encephalitis virus (MVEV) [14] and were characterized in cell culture using Japanese encephalitis virus [15]. Recently, Pijlman et al. demonstrated that the production of these sfRNAs relies on the host enzyme XRN-1 and that their production is an important determinant for viral cytopathogenicity and pathogenesis [16]. Recent studies demonstrated that irrespective of their host species and route of transmission, all flaviviruses including the tentative members

107 Chapter 5 of the genus, produce one or two sfRNAs [14-20]. They vary from 0.3 kb to 0.5 kb, which correlates with the length of the core element of the viral 3’-UTR. Biochemical assays with purified yeast XRN-1 as well as specific knockdown experiments confirmed that sfRNAs are produced through incomplete degradation of the viral genomic RNA by the highly processive 5’→3’ exoribonuclease XRN-1, a cellular enzyme that plays a key role in P-body associated RNA degradation [16;19;20]. Based on the combined outcome of primer extension analysis on viral RNA from infected cells, comparative computer-aided RNA folding, RNA structure probing, site- directed mutagenesis of infectious cDNA clones and in vitro XRN-1 assays, it was predicted that the XRN-1 stalling site in the genome of mosquito-borne flaviviruses would fold into a unique RNA pseudoknot structure. Although this structure is reminiscent of an H-type RNA pseudoknot, it differs in that it is characterized by a structurally conserved side stem- loop (Fig. 1) and a relatively short linker sequence [16;18;19]. Apart from the generally well conserved RNA structural elements of the XRN-1 stalling site, Pijlman et al. also identified several nucleotides that appear to be conserved within the XRN-1 stalling sequence of

UA C G pk’ G C pk’ pk3’ UGA G C G A U G C U G U G A G U CU A G U G A U A G A A A G G C A G G A A C U C A C G U G G A C C G C C A C G CG A G C AC G C C G e2 U A G C A U C G PSK G C

PSK C G A U A U U U C C C GC C PSK e3 C G G C G U U C A C G A U U A G U U U UG C C A A AGGGC C G U G C G A G C A A G G G C A C A A GA C UA G C C C C G UA G A U A C A U G C C G G C G U C GC A U G C G C A U G CG e1 A U A U U A G C C G C G pk3 C G G C A U pk C G U G C G AU A U G A A G U U G G A C A G C C G A C C U U AGAA C A A C U U G A G U A A A U A C U U U G G C C A G U C A U U G U pk 10.531 nt linker insertion 10.629 nt 10.501 nt 10.581 nt 10.252 nt 10.343 nt

YFV WNV-NY99 MODV Fig. 1 Model for the YFV, WNV and MODV XRN-1 stalling signals Nucleotides 10,531-10,629 of the YFV-17D genomic sequence were predicted to form complex stem-loop structures that are responsible for the stalling of XRN-1, and thus the production of YFV sfRNA. The predicted secondary structure is represented here: the bottom stem is namedFig. 1e1, the upper stem e2, and the side hairpin e3. The loop sequence GCUGU (pk3’, highlighted in grey) within the stem-loop e2 was predicted to interact with downstream nucleotides ACAGC (pk3, highlighted in grey), shown with a connecting line, forming a pseudoknot structure (PSK). Nucleotides within this region that were found to be conserved among the mosquito-borne flaviviruses are highlighted in red. The two conserved nucleotides (A-10540 and C-10593), which were mutated as described in the text, are highlighted in yellow. The position where stretches of A and U were inserted to construct the linker insertion mutants are indicated with a green arrow. XRN-1 stalling signal from WNV-NY99 (nt10,501-10,581) and MODV (nt 10,252-10,343) that were used to replace that of YFV in the SLE chimeras are shown here, with sequences involved in pseudoknot interaction (PSK) highlighted in grey and labeled as pk’ and pk. Nucleotides within WNV-NY99 XRN-1 stalling signal that were found to be conserved among the mosquito-borne flaviviruses are highlighted in red, and the side stem-loop structure that was used to replace e3 of YFV in the FLYF-e3WNV mutant is highlighted in blue.

108 RNA structural features for sfRNA production mosquito-borne flaviviruses (Fig.1). For YFV-17D, it was shown that a sequence of ~80 nucleotides comprising the stem-loop structure SLE as well as the downstream nucleotides complementary to the bulge in SLE was sufficient to function as XRN-1 stalling signal in an heterologous expression system [19]. Similar RNA structures, although slightly different in some details, were also predicted to be XRN-1 stalling sites of NKV flaviviruses and insect- specific flaviviruses [20]. In this chapter, we report the results of a more detailed mutagenesis study of the YFV- 17D XRN-1 stalling site, which aimed to investigate the importance of the unique structural features and conserved nucleotides of this particular RNA pseudoknot in sfRNA production and its contribution to cytopathogenicity.

Materials and Methods

Cell culture

The origin and culture conditions of the BHK-21J and Vero E6 cells used in this study have been described before [21].

Recombinant DNA techniques and plasmid construction

Unless described in more detail, general standard nucleic acid methodologies were used throughout this study [22]. Competent Escherichia coli DH5α were used for cloning [23]. The nucleotide numbering of constructs containing YFV-derived inserts and the resulting full-length clones was based on the published YFV-17D sequence [24] (GenBank accession no. X03700).

The shuttle plasmid pBsk-YFV9845-10861 [19] that contained the complete YFV-17D 3’ UTR was used as a template for site-directed mutagenesis using the QuickChange strategy (Stratagene) for the introduction of mutations into the stem-loop E (SLE) region (nomenclature according to [25]). The nucleotide sequence of the SfiI - XbaI fragment encompassing the

SLE of the mutated pBsk-YFV9845-10861 derivatives was verified by sequencing and cloned into pACNR-FLYF17Da [21] to yield full-length YFV-17D cDNA clones with mutations of interest.

In vitro RNA transcription and the production of virus stocks

Plasmid DNA for in vitro RNA transcription was purified using the Nucleobond AX DNA isolation kit (Macherey-Nagel). All plasmids were linearized with AflII, and used for in vitro transcription, as described previously [26].

109 Chapter 5

Stocks of the cDNA-derived viruses were obtained by electroporating BHK-21J cells with full-length YFV transcripts [26]. Medium was harvested when cytopathic effect (CPE) became visible. Virus titers were determined by plaque assay on BHK-21J cells, as described previously [26;27].

Analysis of viral RNA synthesis and sfRNA production

Approximately 1 million BHK-21J cells were seeded in a 10-cm2 plate and were subsequently infected with either the parental YFV-17D or mutant viruses at m.o.i. 5. 3[H]-uridine labeling was performed in the presence of actinomycin D at 24-30 hpi as described previously [21]. Total 3[H]-uridine labeled RNA was isolated using Tripure (Roche), denatured with glyoxal and analyzed on 0.8% agarose gels [22]. YFV sfRNA production was analyzed using non-radioactive RNA that was obtained as described above and used for Northern blotting followed by hybridization as reported previously [19].

Immunostaining of viral infectious centers on Vero cells

Vero cells were seeded in 10-cm2 plates and were infected in duplicate with an amount of virus that was expected to yield a countable number of individual, non-overlapping foci using a similar procedure as for a standard plaque assay. After the regular 4-day incubation at 37°C with 5% CO2, the Avicel overlay was removed and the cells were fixed with 3.5% formaldehyde for 20 minutes. One of the duplicates was stained with crystal violet to monitor plaque formation, whereas the other plate was used for immunostaining of YFV plaques or infectious centers using a monoclonal antibody directed to NS5 [28] as described previously for Modoc virus [29].

110 RNA structural features for sfRNA production

Results

The YFV XRN-1 stalling signal can be replaced by XRN-1 stalling sequences of other flaviviruses without dramatically affecting sfRNA production and viral cytopathic effect on Vero cells. Given the observation that all members of the Flavivirus genus produce an sfRNA in infected cells, we were curious to find out whether the YFV XRN-1 stalling signal could be exchanged by the stalling sequence of another flavivirus without affecting sfRNA production and plaque phenotype of the chimeric YFV. Two recombinant viruses were constructed in which the YFV SLE region (nt 10,531-10,629) was replaced by a functionally comparable sequence of WNV (nt 10,501-10,581) [18] or MODV (nt 10,252-10,343) [30], respectively (Fig. 1). Virus stocks of these recombinant viruses were obtained by harvesting the media from BHK-21J cells that had been transfected with in vitro transcribed RNA derived from pACNR-YFV-SLE-WNV or pACNR-YFV-SLE-MODV. These viruses were subsequently used to infect cells to monitor viral RNA synthesis and sfRNA production. Northern blot hybridization analysis of the RNA from cells infected with these viruses, using oligonucleotide YFV1632 as a probe, revealed they all produced two sfRNAs (Fig. 2A and 2B). The size variation of these sfRNAs correlated with the size differences between either the WNV- or the MODV- derived insert that had been used to replace the YFV XRN-1 stalling signal. Primer extension was used to to determine the 5’ end of the sfRNAs produced by the recombinant viruses and revealed that XRN-1 was stalled at exactly the same nucleotide as in WNV and MODV, respectively [18;20] (data not shown). Plaque assays with these recombinant viruses on Vero cells were used to obtain an indication as to whether the sfRNAs produced by these viruses were serving a similar function in promoting cytopathogenicity as the authentic YFV-17D sfRNA. As illustrated in Fig. 2C, YFV-SLE-MODV was able to form plaques on Vero cells, whereas YFV-17D-ΔSLE, which does not produce any sfRNA [19], was not able to form plaques on Vero cells, despite the fact that this virus was forming infectious foci on Vero cells. Taken together, these results clearly demonstrated that the XRN-1 stalling signal from another flavivirus could substitute for that of YFV-17D without impairing virus replication and dissemination. Moreover, the chimeric viruses still maintained the ability to produce sfRNA and to induce cytopathic effect on Vero cells.

Duplex formation of the bottom stem e1 is not sufficient for sfRNA production.

Previous mutagenesis studies on the e2 region of the SLE [19] revealed that base-pairing of the top stem within the YFV XRN-1 stalling site was more important than the nucleotide sequence per se [19]. To gain a better understanding of the RNA structure requirements for XRN-1 stalling, the importance of the e1 stem in sfRNA production was addressed

111 Chapter 5 A) 1 2 3 B) 1 2 3

A) 1 2 3 B) 1 2 3

C) wtYFV YFV-SLE-MODV YFV-∆SLE

C) wtYFV YFV-SLE-MODV YFV-∆SLE

Fig. 2 Analysis of SLE-chimeric mutants A) Analysis of sfRNA production. BHK cells were infected with viruses (M.O.I. 5) derived from the wildtype pACNR- Fig. 2 FLYF (lane 1), pACNR-YFV-WNV-SLE (lane 2) or were mock infected (lane 3). Intracellular RNA was isolated from infected cells at 30 hpi and was analyzed by Northern blot hybridization as described in the Materials and Methods section. Fig. 2 B) Analysis of sfRNA production. BHK cells were infected with viruses (M.O.I. 5) derived from the wildtype pACNR- FLYF (lane 2), pACNR-YFV-MODV-SLE (lane 3) or were mock infected (lane 1). Intracellular RNA was isolated from infected cells at 30hpi and analyzed by Northern blot hybridization as described in the Materials and Methods section. C) Morphology of plaques (upper panel) and foci (lower panel) produced by viruses derived from the wildtype pACNR-FLYF (left panels), pACNR-YFV-SLE-MODV (middle panels) and pACNR-FLYF-ΔSLE (right panels), which were analyzed with standard plaque assay and immunostaining methods on Vero cells as described in the Materials and Methods section.

112 RNA structural features for sfRNA production in this study. Three YFV-17D mutants were created. In mutant YFV-DD, the bottom four nucleotides of the 5’ arm of e1 (5’-UCAG-3’) were replaced by a complementary sequence (5’-CUGG-3’), whereas in YFV-CC, 3’ arm of the stem (5’-CUGG-3’) was replaced by its complement (5’-UCAG-3’). In both mutants, the base-pairing of the e1 stem region of the SLE was predicted to be disrupted. In the third mutant, the 5’- (5’-UCAG-3’) and 3’- (5’- CUGG-3’) arms were swapped to create YFV-DC, in which the RNA secondary structure of the e1 region was expected to be restored. 3[H]-uridine labeling of BHK-21J cells infected with either wt YFV-17D or the CC, DD or DC mutants revealed that all three YFV-17D mutants replicated with similar efficiency as the parental virus (Fig. 3A). Northern blot analysis of YFV sfRNA production by these mutants revealed that, as hypothesized, neither YFV-CC nor YFV- DD produced detectable amounts of sfRNA (Fig. 3B). Surprisingly, mutant YFV-DC, in which the potential to form an SLE-e1-like stem was expected to be restored also failed to produce any sfRNA (Fig. 3B). Although the results presented above did not rule out the importance of base-pairing in the SLE-e1 stem, they clearly indicated that there are additional RNA sequence and/or structure requirements in this region for XRN-1 stalling.

Nucleotides in stem e1 conserved within mosquito-borne flaviviruses are important for sfRNA formation.

The above findings led to the hypothesis that apart from the conserved duplex formation of the e1 stem, the actual nucleotide sequence in the e1 region of the SLE also contributes to XRN-1 stalling. As highlighted in Fig. 1, several nucleotides that make up the SLE-e1 in mosquito-borne flaviviruses appear to be conserved, suggesting that they may play a role in efficient XRN-1 stalling. Point mutations including all three alternative nucleotides were therefore introduced at position A-10540 and C-10593. Based on the predicted structure for SLE-e1, an A-10540 to C or U mutation would introduce a bulge in the stem, whereas the A to G mutation would replace the original A-U base pair by a non-Watson-Crick G-U base pair. Likewise, C-10593 was mutated to an A, G or U, respectively. Except for the C to U mutation that would have the potential to maintain base-pairing, the other substitutions (C to A or G) were expected to disrupt the base-pairing in SLE-e1. Virus stocks of all these mutants were used to determine the replication efficiency by 3[H]-uridine RNA labeling and to analyze sfRNA production by Northern blot hybridization. As shown inFig. 4A, all A-10540 mutants replicated with similar efficiencies as the parental YFV-17D. Surprisingly, Northern blot analysis of these mutants revealed that both the A→C and A→G mutants were able to produce sfRNA, albeit with lower efficiency compared to the wildtype virus, irrespective of whether base-pairing at this particular position was maintained. The A→U mutant did not yield sfRNA, revealing a yet-to-be-defined requirement for this specific position that must be unrelated to base-pairing. The C to A, U, G substitutions at position 10593 had no effect

113 Chapter 5 A)A) 1 2 3 4 5 B) 1 2 3 4 5

A)A) 1 2 3 4 5 B) 1 2 3 4 5

C) wtFLYF SLEe1DD SLEe1CC SLEe1DC

C) wtFLYF SLEe1DD SLEe1CC SLEe1DC

Fig. 3 Analysis of SLE-e1 stem mutants A) Analysis of viral RNA synthesis. BHK-21J cells were Fig.infected 3 with viruses (M.O.I. 5) derived from pACNR-FLYF (lane 1), pACNR-FLYF-CC (lane 3), pACNR-FLYF-DD (lane 4), pACNR-FLYF-DC (lane 5), or were mock infected (lane 3 2). Intracellular RNA was isolated and analyzed by [H]-RNAFig. 3 labeling as described in the Materials and Methods section. B) Analysis of sfRNA production. BHK cells were infected with viruses (M.O.I. 5) derived from the wildtype pACNR- FLYF (lane 1), pACNR-FLYF-CC (lane 3), pACNR-FLYF-DD (lane 4), pACNR-FLYF-DC (lane 5), or were mock infected (lane 2). Intracellular RNA was isolated and was analyzed by Northern blot hybridization, as described in the Materials and Methods section. C) Morphology of plaques (upper panel) and foci (lower panel) produced by viruses derived from the wildtype pACNR-FLYF (column 1), pACNR-FLYF-DD (column 2), pACNR-FLYF-CC (column 3), and pACNR-FLYF-DC (column 4) were analyzed with standard plaque assay and immunostaining methods on Vero cells as described in the Materials and Methods section.

114 A) B) 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 RNA structural features for sfRNA production

A) B) 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8

g. 4B

g. 4B

C) wtFLYF A10540C A10540G A10540T C10593A C10593G C10593T

C) wtFLYF A10540C A10540G A10540T C10593A C10593G C10593T

Fig. 4 Analysis of SLE-e1 conserved nucleotide mutants A) Analysis of viral RNA synthesis. BHK-21J cells were infected with viruses (M.O.I. 5) derived from pACNR-FLYF (lane 1), pACNR-FLYF-A10540C (lane 3), pACNR-FLYF-A10540GFig. 4 (lane 4), pACNR-FLYF-A10540T (lane 5), pACNR- FLYF-C10593A (lane 6), pACNR-FLYF-C10593G (lane 7), pACNR-FLYF-C10593T (lane 8), or were mock infected (lane 2). Intracellular RNA was isolated and analyzed by 3[H]-RNA labeling as described in the Materials and Methods section. B) Analysis of sfRNA production. BHK cells were infectedFig. 4 with viruses (M.O.I. 5) derived from the wildtype pACNR- FLYF (lane 1), pACNR-FLYF-A10540C (lane 3), pACNR-FLYF-A10540G (lane 4), pACNR-FLYF-A10540T (lane 5), pACNR-FLYF-C10593A (lane 6), pACNR-FLYF-C10593G (lane 7), pACNR-FLYF-C10593T (lane 8), or were mock infected (lane 2), and the production of sfRNA was analyzed by Northern blot hybridization, as described in the Materials and Methods section. C) Morphology of plaques (upper panel) and foci (lower panel) produced by viruses derived from the wildtype pACNR-FLYF (column 1), pACNR-FLYF-A10540C (column 2), pACNR-FLYF-A10540G (column 3), pACNR-FLYF- A10540T (column 4), pACNR-FLYF-C10593A (column 5), pACNR-FLYF-C10593G (column 6), pACNR-FLYF- C10593T (column 7) were analyzed with standard plaque assay and immunostaining methods on Vero cells as described in the Materials and Methods section.

115 Chapter 5 on the YFV-17D replication efficiency either (Fig. 4A). However, no YFV sfRNA production was detected for the C→A or C→G mutants. On the contrary, sfRNA was produced in cells infected with the C→U mutant, in which the original G-C base pair had been replaced with a G-U base pair (Fig. 4B), suggesting that base-pairing of this particular residue is required for sfRNA production. Despite clearly detectable sfRNA production for mutants A→C, A→G, C→U, none of them was able to form plaques on Vero cells, and the size of their foci appeared to correlate with the amount of sfRNA produced (Fig. 4C). Mutant A10540G that produced most abundant sfRNA also yielded the biggest foci, whereas the A→U, C→A, C→G mutants that failed to produce sfRNA made the smallest foci on Vero cells (Fig. 4C).

The side stem-loop structure e3 rather than the nucleotide sequence is essential for sfRNA production.

Most if not all XRN-1 stalling sites that have been identified in flaviviruses are characterized by a small side hairpin that branches from the 5’ strand of the major hairpin [16;19;20] (Fig.1). This structure is referred to as SLE-e3 [25] in YFV and is quite variable in sequence among the different flaviviruses. Two YFV-17D mutants were created to study the role of SLE-e3 in stalling XRN-1. In YFV-∆SLE-e3 this side loop was deleted whereas in YFV-SLE- e3WNV the YFV sequence of e3 was replaced by the equivalent structure in SL-II [18] of WNV, which shares little sequence similarity with that of YFV (Fig. 1). BHK-21J cells were infected with parental YFV-17D in parallel with these two mutant viruses. Virus replication efficiency was studied by 3[H]-uridine labeling of viral RNA. As expected, both YFV-∆SLE-e3 and YFV-SLE-e3WNV were replication competent (Fig. 5A), although viral RNA synthesis was clearly less efficient in YFV-∆SLE-e3 compared to YFV-17D or YFV-SLE-e3WNV infected cells. The effect of deleting SLE-e3 or substituting it by the comparable WNV RNA structure on sfRNA production was analyzed by Northern blotting. As illustrated in Fig. 5B, no YFV sfRNA was detected in cells infected with YFV-ΔSLE-e3, whereas YFV-SLE-e3WNV produced quantities of sfRNA comparable with those generated by the wildtype virus. These results clearly demonstrated that this relatively small hairpin structure in the 5’ part of the XRN-1 stalling signal is absolutely required for sfRNA synthesis. YFV-SLE-e3WNV formed plaques with a similar size as those of YFV-17D on Vero cells upon crystal violet staining, whereas no plaques were observed for YFV-ΔSLE-e3 under the same conditions (Fig. 5C). The results of the infectious center assay on Vero cells for YFV-17D and these two mutant viruses revealed that YFV-SLE-e3WNV formed similar foci as the wt virus and the tiny foci produced by YFV- ΔSLE-e3 may be accounted for by a combined effect of the slower rate of RNA synthesis and a deficiency in sfRNA production of this particular mutant (Fig.5C).

116 A) 1 2 3 4

RNA structural features for sfRNA production

A) 1 2 3 4

B) 1 2 3 4 C) wtYFV FLYF-∆SLEe3 FLYF-SLEe3WNV

B) 1 2 3 4 C) wtYFV FLYF-∆SLEe3 FLYF-SLEe3WNV

Fig. 5 Analysis of SLE-e3 mutants A) Analysis of viral RNA synthesis. BHK-21J cells were infected with viruses (M.O.I. 5) derived from pACNR-FLYF (lane 2), pACNR-FLYF-ΔSLEe3 (lane 3), pACNR-FLYF- e3 WNV (lane 4), or were mock infected (lane 1), and viral RNA synthesis was analyzed by 3[H]-RNA labelingFig. as described 5 in the Materials and Methods section. B) Analysis of sfRNA production. BHK cells were infected with viruses (M.O.I. 5) derived from the wildtype pACNR- FLYF (lane 2), pACNR-FLYF-ΔSLEe3 (lane 3), pACNR-FLYF- e3 WNV (lane 4), or were mock infected (lane 1). Intracellular RNA was isolated and analyzed by Northern blot hybridization, as described in the Materials and Methods section. C) Morphology of plaques (upper panel) and foci (lower panel) produced by of viruses derived from the wildtype pACNR-FLYF (column 1), pACNR-FLYF-ΔSLEe3 (column 2), and pACNR-FLYF- e3 WNV (column 3), were analyzed with standard plaque assay and immunostaining methods on Vero cells as described in the Materials and Methods section. Fig. 5

117 Chapter 5

Extending the linker between S1 and L3 negatively affects sfRNA production.

Compared to the typical H-type pseudoknot, in which L1 spans S2 and crosses the major groove of the helix and L3 spans S1 and crosses the minor groove [31], the number of nucleotides between the YFV SLE equivalent in mosquito-borne flaviviruses andthe downstream nucleotides that were predicted to be involved in the RNA pseudoknot interaction is relatively short. In fact, in the case of YFV, it was predicted to comprise only one nucleotide (a “G” residue), whereas in WNV the linker comprises two nucleotides (Fig.1). Although theoretically possible, it was hard to envision how such a short stretch of nucleotide(s) could cross the minor groove and contribute to the stability of the overall structure of the XRN-1 stalling site in mosquito-borne flaviviruses. It is also noteworthy that the comparable L3 region for the predicted XRN-1 stalling site of the NKV- and insect flaviviruses is significantly longer. This resulted in the hypothesis that optimizing the length of L3 might increase the stability of the predicted XRN-1 stalling structure, and thus the production of more sfRNA. To verify this hypothesis, five YFV-17D mutants were created in which the L3 linker sequence was extended by one, two, three, four, or six nucleotides. The inserted nucleotides consisted of “A” and “U” residues to minimize the possibility that these insertion would induce the formation of an alternative RNA structure that could interfere with the XRN-1 stalling properties of SLE. Insertion of these nucleotides had no effect on the overall rate of viral RNA synthesis of these particular mutants (Fig. 6A). However, contrary to our expectations, increasing the length of L3 had a negative effect on the quantity of YFV sfRNA that was produced. The insertion of a single nucleotide already diminished sfRNA production and extending the length of the insertion quickly led to the abolishment of sfRNA synthesis (Fig. 6B) and yielded viruses that were no longer capable of forming plaques on Vero cells (Fig. 6C). From these experiments, it can be concluded that extension of the predicted L3 RNA element of the YFV XRN-1 stalling RNA pseudoknot negatively affected XRN-1 stalling of this RNA structure.

118 A) 1 2 3 4 5 6 7

RNA structural features for sfRNA production

A) 1 2 3 4 5 6 7

B) 1 2 3 4 5 6 7 C) wtYFV FLYF-A FLYF-AAT

B) 1 2 3 4 5 6 7 C) wtYFV FLYF-A FLYF-AAT

Fig. 6 Analysis of linker insertion mutants A) Analysis of viral RNA synthesis. BHK-21J cells were infected with viruses (M.O.I. 5) derived from pACNR-FLYF (lane 1), pACNR-FLYF-A (lane 3), pACNR-FLYF-AT (lane 4), pACNR-FLYF-AAT (lane 5), pACNR-FLYF-ATAT (lane 6), pACNR-FLYF-ATTAAT (lane 7), or were mock infected (lane 2). Intracellular RNA was isolated and analyzed by 3[H]-RNA labeling as described in the Materials andFig. Methods 6 section. B) Analysis of sfRNA production. BHK cells were infected with viruses (M.O.I. 5) derived from wildtype pACNR-FLYF (lane 1), pACNR-FLYF-A (lane 3), pACNR-FLYF-AT (lane 4), pACNR-FLYF-AAT (lane 5), pACNR-FLYF-ATAT (lane 6), pACNR-FLYF-ATTAAT (lane 7), or were mock infected (lane 2). Intracellular RNA was isolated and analyzed by Northern blot hybridization, as described in the Materials and Methods section. C) Morphology of plaques (upper panel) and foci (lower panel) produced by viruses derived from the wildtype Fig. 6 pACNR-FLYF (left panels), pACNR-FLYF-A (middle panels) and pACNR-FLYF-AAT (right panels) were analyzed with standard plaque assay and immunostaining on Vero cells as described in the Materials and Methods section.

119 Chapter 5

Discussion

Subgenomic RNA production is a defined feature for some +ssRNA families, such as coronaviruses and togaviruses, while for flaviviruses it has long been accepted that plus- strand genomic RNA, minus-strand RNA and replication-related replicative-form and intermediate RNAs are the only virus-derived RNA species present in infected cells. This notion was recently challenged by the identification of an additional non-coding subgenomic flavivirus RNA species (sfRNA) that is co-linear with the viral 3’-UTR [14-17;20]. Different from the corona- or togaviruses, for which subgenomic mRNA is generated during transcription by the viral RNA dependent RNA polymerase, flaviviruses appear to adopt a novel post- transcriptional mechanism for sfRNA production, i.e., by the use of the host enzyme XRN-1 [16;19]. XRN-1 is the main cellular exoribonuclease involved in 5’ → 3’ mRNA degradation in eukaryotes. It is a well conserved, highly processive enzyme that has been reported to be slowed down or stalled in vitro by some artificial high-ordered RNA structures, such as oligo(G) tracts and stable stem-loop structures near the 5’ end of an RNA [32]. A long stem structure with four Gs buried at the bottom was demonstrated to protect the 20S RNA Narnavirus genome from XRN-1 degradation [33]. For the production of sfRNA, XRN-1 was proposed to be stalled upstream of a stem-loop structure stabilized by a pseudoknot interaction that is formed at the 5’ border of the flavivirus 3’-UTR core element, designated as SLE in YFV [19;25] and SLII in WNV [16;18]. Sequence comparison and computer-aided structure prediction revealed the presence of comparable RNA pseudoknot structures in the 3’-UTR of all other members of the Flavivirus genus. In fact, these structures were shown to be able to stall XRN-1 both in vitro and in vivo resulting in the production of sfRNA in all flaviviruses [20;34]. A pseudoknot is an RNA structure minimally composed of two helical segments connected by single-stranded regions or loops [35]. The best characterized type of pseudoknot is the hairpin type (H-type) pseudoknot, which has two base-paired stem regions (designated as S1 and S2) and, depending on the number of loop bases that participate in the pseudoknotting interaction, two or three single-stranded loops (L1, L2 and L3). The base-paired stems (i.e., S1 and S2) stack coaxially to form a quasi-continuous helix, and L1 spans S2 and crosses the deep groove of the helix, whereas L3 spans S1 and crosses the shallow groove [31]. For the YFV XRN-1 stalling signal, the whole SLE could be considered as S1 that however is longer and different from a typical H-type pseudoknot S1, because it actually comprises two stems (designated as e1 and e2 [25], Fig. 1) that are separated by a side hairpin structure (e3). Compared to a typical H-type pseudoknot, the crossing sequence linking the two stems (L3) in the mosquito-borne flaviviruses is rather short, so it is hard

120 RNA structural features for sfRNA production to imagine how such a short L3 (only one nucleotide in the case of YFV) would be able to manage that crossover, and more generally how these stems (e1, e2, e3 and S2 formed by pseudoknot interaction) are oriented in space. Apart from this well conserved RNA pseudoknot structure, several apparently conserved nucleotides were identified in the bottom stem (e1) of the XRN-1 stalling site in mosquito-borne flaviviruses [16]. The special structural features of this type of pseudoknot probably determines its special function as a roadblock for the proceeding XRN-1, since no other pseudoknot structures tested thus far are able to fulfill the same function (unpublished data). The objective of thistudy s was to expand our knowledge on the critical RNA sequence/structure features that render SLE an efficient XRN-1 stalling site. Mutations were introduced to alter the characteristic RNA structures and several, apparently conserved nucleotides of the typical XRN-1 stalling site for mosquito-borne flaviviruses as represented by YFV-17D. The mutant viruses were compared to the parental virus in various aspects of viral life cycle, such as RNA synthesis efficiency, sfRNA production and cytopathogenicity. Similar to what was previously described for the top-stem e2 of SLE [19], the side hairpin structure e3 is required for XRN-1 stalling, albeit with no sequence specificity, as the equivalent side stem-loop of WNV XRN-1 stalling structure (SL-II) that bears little sequence similarity to e3 could replace e3 without affecting viral genomic RNA synthesis or sfRNA production. On the contrary, base-pairing at the bottom stem e1 does not appear to be the sole requirement for successful XRN-1 stalling. As expected, YFV mutants in which the base- pairing within stem e1 was disrupted were unable to produce sfRNA and had an attenuated phenotype. Surprisingly, the mutant virus in which base-pairing at e1 was predicted to be restored was unable to produce sfRNA, despite the fact that this mutant virus replicated with nearly the same efficiency as the parental virus. This finding suggests that in addition to the RNA structure, the nucleotide sequence of stem e1 is also of functional importance in stalling XRN-1. This is clearly different from what has been reported for YFV SLE stem e2 [19] and what was observed for the side stem-loop e3, for which base-pairing appears to be the sole determinant for XRN-1 stalling. Identical results were obtained when making comparable mutations in the equivalent region of MODV (Royo and Bredenbeek, unpublished results). To determine whether specific nucleotides within stem e1 play essential rolesin XRN-1 stalling, two nucleotides conserved in the XRN-1 stalling site of mosquito-borne flaviviruses were targeted for mutagenesis in the background of YFV-17D. Restoration of the base-pairing potential did seem to restore sfRNA formation, since both the A10540G and C10593U mutants were able to produce sfRNA, yet the A10540G mutant that was expected to introduce a small bulge in e1 was also able to generate sfRNA in infected cells. Strikingly, none of these single-nucleotide substitution mutants were able to form plaques on Vero cells, irrespective of the amount of sfRNA produced. An interesting finding was that the foci

121 Chapter 5 size on Vero cells appeared to be correlated with the amount of sfRNA produced. In fact, all sfRNA-null mutants made in this study were consistently less cytopathogenic on Vero cells, forming tiny foci and no plaques, and those capable of producing sfRNA yet with lower efficiency formed bigger foci. One possible explanation is that cytopathogenic phenotype is somehow related to sfRNA quantity; therefore, sfRNA produced by the substitution mutants may have been too low to induce CPE and thus plaque formation on Vero cells. Alternatively, the changed nucleotides could somehow alter sfRNA conformation, resulting in the loss of cytopathogenicity. Notably, this difference in plaque/foci formation was not observed when BHK-21J cells were infected. On BHK-21J cells, all of the sfRNA mutants produced plaques, with sizes similar to that produced by parental YFV-17D. This cell line specificity could not be explained by innate immune evasion, a function proposed for sfRNA [36], since neither of these cell lines could produce type I interferon, suggesting sfRNA is involved in other interferon-independent pathways that contribute to cell death. Although the mutagenesis study underlined the importance of both sequence and structure features present in YFV SLE region for sfRNA production and functionality, it does not provide us with sufficient information to speculate on the conformation adopted by this special pseudoknot structure that stalls the highly processive XRN-1. Hopefully, the 3-D structure of this RNA region will be resolved in the near future to direct further mutagenesis studies. We have to bear in mind, however, that crystallization with or without the presence of viral/host factors could dramatically change RNA conformation. The 3-D structure obtained with these methods might not represent the whole picture in infected cells, where RNA could dynamically change conformation and interact with other proteins or nucleic acid sequences. For example, the conserved nucleotides located in the bottom e1 stem could be interacting with other nucleotides, thus stabilizing the higher-order RNA structure; alternatively, they could bind proteins (XRN-1, for instance), which either locks it at this position or induces a switch in RNA 3-D conformation. In this sense, structure probing, especially the recently proposed in vivo probing techniques [37], could present us with RNA conformations in a physiologically relevant environment. Instead of exhaustive mutagenesis studies, specific mutagenesis studies directed by structural information coupled with chemical probing and biochemical analysis would hopefully help solve the unanswered questions surrounding the XRN-1 stalling signal and provide more insights into sfRNA function.

122 RNA structural features for sfRNA production

Reference List

[1] Kuno G, Chang GJ, Tsuchiya KR, Karabatsos N, Cropp CB. Phylogeny of the genus Flavivirus. J Virol 1998 Jan;72(1):73-83.

[2] Stollar V, Thomas VL. An agent in the Aedes aegypti cell line (Peleg) which causes fusion of Aedes albopictus cells. Virology 1975 Apr;64(2):367-77.

[3] Crabtree MB, Sang RC, Stollar V, Dunster LM, Miller BR. Genetic and phenotypic characterization of the newly described insect flavivirus, Kamiti River virus. Arch Virol 2003 Jun;148(6):1095-118.

[4] Hoshino K, Isawa H, Tsuda Y, et al. Genetic characterization of a new insect flavivirus isolated from Culex pipiens mosquito in Japan. Virology 2007 Mar 15;359(2):405-14.

[5] Lindenbach BD, Rice CM. Molecular biology of flaviviruses. Adv Virus Res 2003;59:23-61.

[6] Mandl CW, Kunz C, Heinz FX. Presence of poly(A) in a flavivirus: significant differences between the 3’ noncoding regions of the genomic RNAs of tick-borne encephalitis virus strains. J Virol 1991 Aug;65(8):4070- 7.

[7] Wallner G, Mandl CW, Kunz C, Heinz FX. The flavivirus 3’-noncoding region: extensive size heterogeneity independent of evolutionary relationships among strains of tick-borne encephalitis virus. Virology 1995 Oct 20;213(1):169-78.

[8] Poidinger M, Hall RA, Mackenzie JS. Molecular characterization of the Japanese encephalitis serocomplex of the flavivirus genus. Virology 1996 Apr 15;218(2):417-21.

[9] Wang E, Weaver SC, Shope RE, Tesh RB, Watts DM, Barrett AD. Genetic variation in yellow fever virus: duplication in the 3’ noncoding region of strains from Africa. Virology 1996 Nov 15;225(2):274-81.

[10] Gritsun TS, Venugopal K, Zanotto PM, et al. Complete sequence of two tick-borne flaviviruses isolated from Siberia and the UK: analysis and significance of the 5’ and 3’-UTRs. Virus Res 1997 May;49(1):27-39.

[11] Proutski V, Gould EA, Holmes EC. Secondary structure of the 3’ untranslated region of flaviviruses: similarities and differences. Nucleic Acids Res 1997 Mar 15;25(6):1194-202.

[12] Rauscher S, Flamm C, Mandl CW, Heinz FX, Stadler PF. Secondary structure of the 3’-noncoding region of flavivirus genomes: comparative analysis of base pairing probabilities. RNA 1997 Jul;3(7):779-91.

[13] Mandl CW, Holzmann H, Meixner T, et al. Spontaneous and engineered deletions in the 3’ noncoding region of tick-borne encephalitis virus: construction of highly attenuated mutants of a flavivirus. J Virol 1998 Mar;72(3):2132-40.

[14] Urosevic N, van MM, Mansfield JP, Mackenzie JS, Shellam GR. Molecular characterization of virus-specific RNA produced in the brains of flavivirus-susceptible and -resistant mice after challenge with Murray Valley encephalitis virus. J Gen Virol 1997 Jan;78 ( Pt 1):23-9.

[15] Lin KC, Chang HL, Chang RY. Accumulation of a 3’-terminal genome fragment in Japanese encephalitis virus-infected mammalian and mosquito cells. J Virol 2004 May;78(10):5133-8.

[16] Pijlman GP, Funk A, Kondratieva N, et al. A highly structured, nuclease-resistant, noncoding RNA produced by flaviviruses is required for pathogenicity. Cell Host Microbe 2008 Dec 11;4(6):579-91.

[17] Scherbik SV, Paranjape JM, Stockman BM, Silverman RH, Brinton MA. RNase L plays a role in the antiviral response to West Nile virus. J Virol 2006 Mar;80(6):2987-99.

123 Chapter 5

[18] Funk A, Truong K, Nagasaki T, et al. RNA structures required for production of subgenomic flavivirus RNA. J Virol 2010 Nov;84(21):11407-17.

[19] Silva PA, Pereira CF, Dalebout TJ, Spaan WJ, Bredenbeek PJ. An RNA pseudoknot is required for production of yellow fever virus subgenomic RNA by the host nuclease XRN1. J Virol 2010 Nov;84(21):11395-406.

[20] Silva PA, Dalebout TJ, Gultyaev AP, Olsthoorn RC, Bredenbeek PJ. Characterization of the sfRNAs that are produced in cells infected with no known vector and cell fusing agent. 2013. Unpublished Work

[21] Bredenbeek PJ, Kooi EA, Lindenbach B, Huijkman N, Rice CM, Spaan WJ. A stable full-length yellow fever virus cDNA clone and the role of conserved RNA elements in flavivirus replication. J Gen Virol 2003 May;84(Pt 5):1261-8.

[22] Sambrook J, Fritsch T, Maniatis T. Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1989.

[23] Inoue H, Nojima H, Okayama H. High efficiency transformation of Escherichia coli with plasmids. Gene 1990 Nov 30;96(1):23-8.

[24] Rice CM, Lenches EM, Eddy SR, Shin SJ, Sheets RL, Strauss JH. Nucleotide sequence of yellow fever virus: implications for flavivirus gene expression and evolution. Science 1985 Aug 23;229(4715):726-33.

[25] Olsthoorn RC, Bol JF. Sequence comparison and secondary structure analysis of the 3’ noncoding region of flavivirus genomes reveals multiple pseudoknots. RNA 2001 Oct;7(10):1370-7.

[26] Silva PA, Molenkamp R, Dalebout TJ, et al. Conservation of the pentanucleotide motif at the top of the yellow fever virus 17D 3’ stem-loop structure is not required for replication. J Gen Virol 2007 Jun;88(Pt 6):1738-47.

[27] Matrosovich M, Matrosovich T, Garten W, Klenk HD. New low-viscosity overlay medium for viral plaque assays. Virol J 2006;3:63.

[28] Buckley A, Gaidamovich S, Turchinskaya A, Gould EA. Monoclonal antibodies identify the NS5 yellow fever virus non-structural protein in the nuclei of infected cells. J Gen Virol 1992 May;73 ( Pt 5):1125-30.

[29] Jiang X, Silva PA, Dalebout TJ, Bredenbeek P. An infectious Modoc Virus cDNA as a tool to study conserved 3’-UTR RNA elements in flaviviruses with no known vector. 2013.

[30] Leyssen P, Charlier N, Lemey P, et al. Complete genome sequence, taxonomic assignment, and comparative analysis of the untranslated regions of the Modoc virus, a flavivirus with no known vector. Virology 2002 Feb 1;293(1):125-40.

[31] Brierley I, Pennell S, Gilbert RJ. Viral RNA pseudoknots: versatile motifs in gene expression and replication. Nat Rev Microbiol 2007 Aug;5(8):598-610.

[32] Poole TL, Stevens A. Structural modifications of RNA influence the 5’ exoribonucleolytic hydrolysis by XRN1 and HKE1 of Saccharomyces cerevisiae. Biochem Biophys Res Commun 1997 Jun 27;235(3):799-805.

[33] Esteban R, Vega L, Fujimura T. 20S RNA narnavirus defies the antiviral activity of SKI1/XRN1 in Saccharomyces cerevisiae. J Biol Chem 2008 Sep 19;283(38):25812-20.

[34] Knaap R., Jiang X, Olsthoorn RC, Bredenbeek PJ. RNA structure requirements for XRN1 stalling in the tick- born encephalitis virus 3’untranslated region. 2013. Unpublished Work

[35] Staple DW, Butcher SE. Pseudoknots: RNA structures with diverse functions. PLoS Biol 2005 Jun;3(6):e213.

124 RNA structural features for sfRNA production

[36] Schuessler A, Funk A, Lazear HM, et al. West Nile virus non-coding subgenomic RNA contributes to viral evasion of type I interferon-mediated antiviral response. J Virol 2012 Feb 29.

[37] Spitale RC, Crisalli P, Flynn RA, Torre EA, Kool ET, Chang HY. RNA SHAPE analysis in living cells. Nat Chem Biol 2013 Jan;9(1):18-20.

125 126 Chapter 6

Trans-complementation of yellow fever virus cytopathogenicity by cells expressing subgenomic flavivirus (sf)RNA

Xiaohong Jiang, Robert C. M. Knaap, Tim J. Dalebout and Peter J. Bredenbeek

127 Chapter 6

Abstract

All flaviviruses produce a small, non-coding RNA derived from viral 3’ untranslated region (3’- UTR) during infection. This RNA is denoted subgenomic flavivirus (sf)RNA, and is produced from incomplete degradation of the viral genomic RNA by the host exoribonuclease XRN-1. XRN-1 processing was predicted to be stalled by a non-H-type RNA pseudoknot structure formed at the 5’ border of the viral 3’-UTR core region. Although the precise function of sfRNA remains to be elucidated, sfRNA production correlates with cytopathogenicity and pathogenesis during viral infection. In this study, expression plasmids for the production of the yellow fever virus (YFV) and West Nile virus (WNV) sfRNAs were constructed and were shown to produce sfRNAs that were identical to those derived from viral infection both in vitro and in vivo. With the expression plasmids, SW13 cell populations constitutively expressing WNV or YFV sfRNA were established and shown to enhance cytopathic effect upon infection by wildtype YFV-17D or a mutant virus YFV-17D-BB that was deficient in sfRNA production. These cell lines could be used to test other hypothesis regarding sfRNA functionality in the presence or absence of viral infection and provide more insight into the molecular determinants for cytopathogenicity of flaviviruses.

128 Trans-complementation of sfRNA function

Introduction

Flaviviruses are small, enveloped viruses with a positive-sense single-stranded RNA genome. It has long been held that only three species of virus derived RNA are present in cells infected with flavivirus, i.e., positive-sense genomic RNA, negative-sense RNA, and intermediates generated during RNA replication (replicative-form RNA). However, this notion has been challenged by the recent discovery of another small viral RNA that is co-linear with the distal part of the flavivirus 3’ untranslated region (3’-UTR) [1-5]. This RNA was denoted subgenomic flavivirus (sf)RNA and has been found to be produced in both mammalian and insect cell lines infected with arthropod-borne (arbo), no-known-vector (NKV) and insect- specific flaviviruses [1-4;6]. The production of sfRNA was proposed to be a hallmark for the Flavivirus genus, since this feature has been found to be conserved within members of the Flavivirus genus, whereas related viruses belonging to other genera of the Flaviviridae do not produce similar RNA species [2;5]. Unlike subgenomic plus- and/or minus-strand RNAs generated by other plus- strand RNA viruses like corona- and togaviruses, the sfRNA is not produced by the viral RNA-dependent RNA polymerase. It is, in fact, an incomplete degradation product of the viral genomic RNA by a host ribonuclease, XRN-1, which is stalled at the 5’ border of the 3’-UTR core element of the viral genome [2;3;5;7] that contains essential RNA elements for virus replication, translation and possibly encapsidation [8]. XRN-1 is well conserved among eukaryotes and is a highly processive 5’→3’ exoribonuclease associated with mRNA turnover. After shortening of the poly(A) tail and decapping by DCP1 and DCP2, mRNA with a 5’ monophosphate becomes susceptible to XRN-1 hydrolysis. All the enzymes involved in the 5’→3’ mRNA decay pathway have been shown to localize in mRNA-processing bodies (P-bodies) [9]. This rather unexpected involvement of XRN-1 in the flavivirus life cycle has been firmly established usingin vitro assays with commercially available purified yeast XRN- 1 as well as XRN-1 knockdown experiments in cell culture [2;10]. The actual stalling site for XRN-1 was mapped to a non-H-type RNA pseudoknot structure that is predicted to be present in the 3’-UTR of every member of the Flavivirus genus that yet has little to no sequence similarity, especially when viruses from different clusters are compared [5;11]. This RNA pseudoknot, or the production of sfRNA, is not required for viral RNA synthesis, polyprotein translation or virion formation, as sfRNA-null mutants in which either this RNA pseudoknot was deleted or mutated still replicated with similar efficiency as the wildtype viruses and demonstrated growth kinetics that were at most slightly delayed compared to the parental viruses [2;3;7;12]. However, flaviviruses that did not produce sfRNA showed a remarkable reduction of cytopathogenicity in cell culture [2;12] and pathogenesis in mice [2;7]. The molecular mechanisms behind these observations

129 Chapter 6 remain to be uncovered. Several possible functions associated with sfRNA were proposed. The sfRNA was proposed to be part of the viral immune evasion strategy, subverting type I interferon mediated innate immune response in mammalian hosts [13] and suppressing the major antiviral pathway, i.e., RNA interference in insect cells [14]. Besides, sfRNA was hypothesized to be a precursor for the production of a microRNA (miRNA) in insect cell lines [15]. These suggested sfRNA function, however, do not provide a unifying explanation as to why all flaviviruses, irrespective of their host, produce sfRNA. A recent study revealed the interaction between sfRNA and a more universal pathway, i.e., sfRNA sequestering XRN-1 of the infected cells and thus altering overall gene expression profile [10]. In this chapter, we describe the construction and characterization of cell lines that express either the YFV or WNV sfRNA and demonstrate that these cell lines can be used to trans-complement the lack of cytopathogenicity of a YFV mutant that does not produce sfRNA as well as enhance the cytopathogenicity of wildtype YFV on SW13 cells.

Materials and methods

Cell culture

BHK-21J [16] and SW13 cells were grown at 37°C in 5% CO2 in Dulbecco’s modified Eagle’s medium (DMEM; Lonza) supplemented with 8% fetal calf serum (Bodinco BV), 100 U/mL penicillin (Lonza), 100 U/mL streptomycin (Lonza) and 5% non-essential amino acids (GE Healthcare). HEK293T cells were grown under the same condition, except for the addition of L-glutamine (2 mM; Lonza) into culture media.

Recombinant DNA techniques and plasmid construction

In general, standard nucleic acid methodologies were used [17]. Competent E. coli DH5α was used for cloning [18]. The complete 3’-UTR sequence of YFV-17D and WNV were cloned as 3’-trailer downstream of enhanced green fluorescent protein (eGFP) coding sequence in the mammalian expression vector pcDNA3. The hepatitis delta virus ribozyme (HDVr) was derived from a West Nile virus replicon cDNA and was fused to the last nucleotide of the flaviviral 3’-UTR in the pcDNA3 constructs by PCR [19]. As a control, plasmid pcDNA3-eGFP- YFV3’-UTR-BB was constructed, in which the stem-loop structure in the XRN-1 stalling RNA pseudoknot within pcDNA3-eGFP-YFV3’-UTR was disrupted by Quickchange site-directed mutagenesis (Strategene). A previous study showed that this mutation incapacitated the stalling site and thus disabling the production of sfRNA [3]. The characteristics of the viral 3’-UTRs that were cloned in the pcDNA3-eGFP-HDVr vector are summarized in Fig. 1B.

130 Trans-complementation of sfRNA function

Plasmid transfection and generation of eGFP-flavivirus3’-UTR expressing cell populations

Transfection of DNA into SW13 cells was mediated by polyethylenimine (PEI; Polysciences Inc.). Briefly, 13 μg of plasmid DNA was diluted in 150 mM NaCl to a total volume of 430μl. Subsequently, 50 μl PEI (1 mg/ml) was added to 380μl of 150 mM NaCl and thoroughly mixed with the DNA containing solution. The mixture was incubated for 10 minutes at room temperature and was added to ~80% confluent SW13 cells in a 75 cm2 flask. After a 24 h incubation of cells with DNA-PEI complexes, culture medium was refreshed. Two days post transfection, cells were put under selective pressure by adding 250μg/mL G418 (Geneticin; GE Healthcare) to the medium of SW13 cells. After two weeks of selection, cells were analyzed on a FACSAriaIII machine (BD Biosciences) and the top 10% eGFP expressing cells were collected. These cells were maintained in selective media that contained 250 μg/ mL G418. For transient expression experiments, HEK293T cells were used given their favorable response to DNA transfection. The cells were grown to 70 - 80% confluency in 3.5 cm2 dishes and 1.5 μg of the expression plasmids were transfected using a downscaled version of the protocol described above (total volume added for transfection was 110 μl).

In vitro XRN-1 assay

In vitro XRN-1 assay was performed as described previously [3]. Briefly, pcDNA3-eGFP-YFV3’- UTR and pcDNA3-eGFP-YFV3’-UTR-BB plasmids were linearized at the unique XhoI site downstream of the HDVr and purified by phenol-chloroform extraction. Approximately 1-2 μg of the linearized DNA was used as template for in vitro RNA transcription using T7-Scribe™ RNA IVT Kit (Cell Script). After DNase treatment, RNA was purified by LiCl precipitation. 8 μg of RNA was incubated with 1.32 units of tobacco acid pyrophosphatase (TAP; Epicentre) at 37°C for 2 h to obtain RNA with a 5’ monophosphate. After phenol-chloroform extraction and ethanol precipitation, 4 μg of TAP-treated RNA was incubated with 1 Unit of XRN-1 (Terminator 5’-phosphate-dependent exonuclease; Epicentre) at 30°C for 1.5 h.

Northern blot analysis

Total RNA was isolated from infected or transfected cells using Tripure (Roche). For Northern blotting, 10μg of total RNA isolated from either infected or transfected cells or 0.2μg of XRN-1 treated in vitro transcribed RNA was denatured with formaldehyde, separated on a formaldehyde containing 1.5% agarose gel and blotted onto Hybond-N+ membrane (GE- Healthcare). RNA was immobilized by baking the membrane at 80°C for 2 - 3 h. The blots were hybridized to 32P-labeled oligonucleotides as described previously [17;20].

131 Chapter 6

Complementation assay

SW13 cells expressing the YFV-17D or WNV sfRNA were seeded in 6-well clusters. The cells were washed once with PBS and inoculated with 100-200 pfu of YFV-17D or YFV-17D-BB (titers determined on BHK cells). After 1 h of adsorption at 37°C, the inocula were removed and the cells were overlaid with complete DMEM medium containing 1.2% Avicel (FMC BioPolymer). Plaques were visualized by crystal violet staining at 4 days post infection.

Results

Construction of expression plasmids for the production of authentic WNV or YFV-17D sfRNA

The basic vector that is used for this study was pcDNA3-eGFP, in which eGFP was inserted into pcDNA3.1 (Invitrogen) by using the BamHI and EcoRI sites. The complete WNV 3’-UTR and the HDVr sequence were obtained by PCR from the WNV replicon [19] and cloned into pcDNA3-eGFP with EcoRI and XhoI sites , resulting in the construction of pcDNA3-eGFP- WNV3’-UTR. This plasmid was further used to construct similar plasmids containing the 3’- UTR of other flaviviruses like YFV-17D. Generally, complete flavivirus 3’-UTR fragments were produced by PCR using a forward primer that contained an EcoRI adaptor sequence and a reverse primer that is complementary to the last ~35 nt of the flavivirus fused to the most 5’ 35 nt of the HDVr sequence, within which a unique RsrII site was included that was used for cloning. The general feature of the expression cassette cloned in pcDNA3.1 vector was depicted in Fig. 1A-1. Transfection of these plasmids into mammalian cells was expected to result in the transcription of mRNA that is terminated at the transcription termination site present in pcDNA3.1 and get polyadenylated (Fig. 1A-2). This “precursor” RNA was expected to be cleaved in cis by the HDVr (Fig. 1A-2), resulting in the eGFP mRNA containing a complete flavivirus 3’-UTR sequence (Fig. 1A-3). If this RNA would enter the 5’→3’ RNA decay pathway, it would be decapped and then hydrolyzed by XRN-1 (Fig. 1A-3) until the enzyme reaches the XRN-1 stalling site that is present in the viral 3’-UTR, resulting in the production of authentic flavivirus sfRNA (Fig. 1A-4).

In vitro and in vivo analysis of YFV sfRNA production from pcDNA3-eGFP- YFV3’-UTR

To verify that pcDNA3-eGFP-YFV3’-UTR can be used as a template for the production of YFV sfRNA, the plasmid was linearized with XhoI and used as the template for in vitro

132 A) Trans-complementation of sfRNA function

A)

B)

Construct NCBI number Nucleotides Length (nts) B) pcDNA3-eGFP-YFV3’-UTR X03700 10351-10862 512

pcDNA3-eGFP-WNV3’-UTR M12294 10377-11029 653 Figure 1. Characteristics Construct of the expression plasmidsNCBI number for the production Nucleotides of flavivirus Length sfRNA (nts) A) SchematicpcDNA3-eGFP-YFV3’-UTR representation offRNA s production X03700 from pcDNA3-eGFP-flavivirus3’-UTR 10351-10862 plasmids 512

General outlinepcDNA3-eGFP-WNV3’-UTR of the cassettes that were M12294cloned in the pcDNA3 10377-11029 vector downstream 653 of the CMV promoter (black) that drives RNA polymerase II transcription: eGFP coding sequence (green), YFV-17D or WNV3’-

UTR sequence (yellow), the hepatitis delta virus ribozyme sequence (HDVr; grey) and the RNA polymerase II transcription terminator and polyadenylation sequence (blue). Unique DNA restriction enzyme recognition sites used to construct the plasmids are also depicted.Fig. 1 Transcription of eGFP mRNA by cellular RNA polymerase II and o-transcriptionalc cleavage by HDVr Ribozyme cleaved, decapped eGFP RNA becomes susceptible to XRN-1 hydrolysis. Stalling of XRN-1 at the non-H-type pseudoknotFig. present 1 in the flavivirus 3’-UTR results in the production of flavivirus sfRNA. B) Information of viral 3’-UTR sequences that were cloned as 3’-trailer downstream of eGFP coding sequence in the pcDNA3-eGFP-flavivirus3’UTR plasmids. transcription of the eGFP-YFV3’-UTR RNA. Plasmid pcDNA3-eGFP-YFV3’-UTR-BB was used as a negative control. RNA transcribed from this plasmid was not expected to produce sfRNA due to the disruption of the RNA structure that was required to stall XRN-1. After treatment

133 Chapter 6 with TAP, the in vitro transcripts with 5’ monophosphate were incubated with XRN-1 and subsequently analyzed by Northern blotting and hybridization using oligonucleotide 1632 that was complementary to nt 10,690 to 10,708 in the YFV 3’-UTR as a probe. As shown in Fig. 2A, two bands were observed in all four lanes of the Northern blot that represented the eGFP gene containing RNAs that were not digested by XRN-1 and that either still contained the HDVr sequence (eGFP RNA+HDVr) or from which the HDVr sequence and poly(A) tail were removed due to the autocatalytic cleavage of the HDVr sequence (eGFP RNA). Upon treatment with XRN-1, two faster migrating RNA species representing the YFV sfRNA with or without the HDVr and the poly(A) sequence were detected for the transcript containing the wildtype YFV-17D 3’-UTR sequence (lane1). In contrast, eGFP-YFV3’UTR-BB transcript from pcDNA3-eGFP-YFV3’UTR-BB that was treated in the same manner did not yield detectable amounts of sfRNA, despite the production of mRNA (lane 3). Neither sfRNA-related bands were detected in samples that were not subjected to XRN-1 treatment (lanes 2 & 4). To determine whether transfection of mammalian cells with these expression plasmids would result in the production of authentic sfRNA, HEK293T cells were transfected with pcDNA3-eGFP-YFV3’-UTR or pcDNA3-eGFP-YFV3’-UTR-BB plasmids. At 48 h post transfection, total RNA was isolated from the cells and analyzed for sfRNA production by Northern blot hybridization, using oligonucleotide 1632 as a probe. Intracellular RNA isolated from HEK293T cells transfected with pcDNA3-eGFP-YFV3’-UTR contained two sfRNAs (Fig. 2B, lane 2) that co-migrated with the two sfRNAs produced in BHK-21J cells infected with YFV (Fig. 2B, lane 1). No sfRNA was detected in HEK293T cells transfected with the mutant plasmid pcDNA3-eGFP-YFV3’-UTR-BB, despite the efficient transcription of the mRNA, as indicated by two arrows that represented eGFP mRNA with or without HDVr and the poly(A) sequence. Two other less prominent RNA species were consistently observed in intracellular RNA isolated from HEK293T cells transfected with the plasmids. The faster migrating RNA (labeled as A) could represent the YFV sfRNA to which the HDVr and poly-A sequence were still attached due to incomplete cleavage of the precursor RNA bythe ribozyme. Alternatively, this RNA A could originate from an alternative weak XRN-1 stalling site that is upstream of the XRN-1 stalling site in the viral 3’-UTR. A similar explanation could be valid for the production of RNA B. Taken together, these in vitro and in vivo data clearly demonstrated that these pcDNA3-eGFP-flaviviral3’-UTR plasmids could be used for the production of authentic flavivirus sfRNAs.

134 Trans-complementation of sfRNA function

A) B) YFV YFV YFV YFV (AB) YFV (BB) (AB) (AB) (BB) XRN1 + - + - BHK 293T

YFV Genome

eGFP RNA+HDVr eGFP RNA+HDVr eGFP RNA eGFP RNA

B

A sfRNA+HDVr sfRNA 1 sfRNA 1 sfRNA 2

1 2 3 4 1 2 3 Figure 2. In vitro and in vivo production of sfRNA from pcDNA3-eGFP-YFV3’-UTR and pcDNA3-eGFP-YFV3’-UTR- BB A) Northern blot using a YFV 3’-UTR specific probe (oligonucleotideFig. 2 YFV1632) on in vitro transcripts of pcDNA3- eGFP-YFV3’-UTR (lanes 1 and 2) and pcDNA3-eGFP-YFV3’-UTR-BB (lanes 3 and 4). RNA in lanes 2 and 4 was treated with tobacco acid pyrophosphatase (TAP), whereas RNA in lanes 1 and 3 was treated with TAP and subsequently with purified yeast XRN-1. Bands corresponding to the eGFP RNA with or without HDVr and sfRNA with or without HDVr are indicated by arrows. B) Northern blot using oligonucleotide YFV 1632 as a probe on intracellular RNA isolated from HEK293T cells that were transfected with pcDNA3-eGFP-YFV3’-UTR (lane 2) or pcDNA3-eGFP-YFV3’-UTR-BB (lane 3). RNA was isolated at 48 h post transfection. Intracellular RNA isolated from BHK-21J cells infected with YFV was analyzed in parallel and served as sfRNA size marker (lane 1). Bands corresponding to YFV genome, eGFP RNA with or without HDVr and sfRNA 1 and 2 are indicated by arrows. Two less prominent bands labeled as A and B are discussed in the text.

135 Chapter 6

Establishment of SW13 cell populations constitutively expressing YFV-17D or WNV sfRNA

After establishing that transcripts derived from pcDNA3-eGFP-YFV3’-UTR were able to produce authentic sfRNA both in vitro and in vivo, populations of SW13 cells constitutively expressing YFV-17D or WNV sfRNA were established. To achieve this objective, SW13 cells were transfected with pcDNA3-eGFP-YFV3’-UTR or pcDNA3-eGFP-WNV3’-UTR, respectively. From 48 h post transfection onwards, the cells were grown in selective medium containing 250 mg/ml G418. After an additional 14 days of culturing under selective pressure, these cell cultures contained mixed populations of eGFP expressing and non-eGFP expressing G418-resistant SW13 cells, as was revealed by FACS analysis (Fig. 3A, middle panels) or fluorescent imaging (Fig. 3B, upper panel). The top 10% eGFP expressing SW13 cells in both cell populations were isolated by FACS to obtain a more homogeneous population of cells expressing relatively high levels of eGFP, under the assumption that these cells produced

A)

Before transfection

B) eGFP-WNV3’-UTR eGFP-YFV3’-UTR eGFP-WNV3’-UTR

Before cell sorting

After cell sorting

Figure 3. eGFP expression in G418 resistant SW13 cells transfected with pcDNA3-eGFP-YFV3’-UTR and pcDNA3- eGFP-WNV3’-UTR plasmids Fig. 3 A) FACS analysis of the G418 resistant SW13 cells transfected with pcDNA3-eGFP-YFV3’-UTR (left panels) or pcDNA3-eGFP-WNV3’-UTR (right panels) before and after eGFP expression-based cell sorting. The top panel shows fluorescence in non-transfected cells. The percentage gated represents the percentage of cells with eGFP expression level higher than that of non-transfected SW13 cells. B) G418 resistant SW13 cells obtained after transfection with the pcDNA3-eGFP-WNV3’-UTR plasmid were seeded on coverslips before (upper panel) and after cell sorting (lower panel). After fixation with 3.7% formaldehyde, cells were stained with Hoechst (blue) and visualized using a fluorescence microscope.

136 Trans-complementation of sfRNA function the largest quantities of sfRNA. As shown in the bottom panels of Fig. 3A, the cell sorting resulted in a significant increase both in the percentage of eGFP expressing cells as well as in the level of eGFP expression in the sorted cells, which correlated with fluorescent imaging results (Fig. 3B).

Trans-complementation of a YFV-17D sfRNA-null mutant using SW13 cell populations constitutively expressing YFV-17D or WNV sfRNA

The established cell populations were subsequently used to trans-complement the YFV-17D- BB mutant [3] that could not produce YFV sfRNA and was unable to form plaques on SW13 cells. After expansion of the sorted SW13 cell populations expressing the eGFP-YFV3’-UTR or eGFP-WNV3’-UTR, these cells as well as the parental, non-transfected SW13 cells were seeded in 10 cm2 plates and infected with approximately 100 to 200 pfu (titers determined on BHK cells) of the parental YFV-17D or YFV-17D-BB. After 1 h of adsorption, the inocula were removed and Avicel containing overlay was added. Four days later, these plaque assays were fixed and stained with crystal violet to visualize plaque formation. As shown in Fig. 4, YFV-17D-BB mutant was unable to form visible plaques on the non-transfected

SW13 SW13-eGFP-YFV3’-UTR SW13-eGFP-WNV3’-UTR

YFV17D

YFV17D-BB

Figure 4. Plaque morphology of YFV-17D and YFV-17D-BB on SW13 cells that do or do not express flavivirus sfRNA SW13 cells and G418 resistant SW13 cells expressing YFV Fig.or WNV 4 sfRNA were infected with either parental YFV- 17D (upper panels) or YFV-17D-BB that is deficient in sfRNA production (lower panels), and plaque morphology was analyzed as described in the Materials and Methods section.

137 Chapter 6

SW13 cells, however on cells expressing eGFP-YFV3’-UTR mRNA and thus YFV sfRNA, this sfRNA-null mutant virus was able to form small but clearly identifiable plaques, suggesting that the cytopathic characteristics of YFV-17D on SW13 cells were restored by the trans- complementation of the YFV sfRNA produced in these cells (Fig. 4). More surprisingly, SW13 cells expressing the eGFP-WNV3’-UTR mRNA and thus the authentic WNV sfRNA were also capable of trans-complementing the lack of cytopathogenicity of the YFV-17D-BB mutant, as judged from the obvious plaque formation in the lower right panel of Fig. 4. The validity of the observed trans-complementation of the YFV-17D-BB mutant in cells producing either YFV or WNV sfRNA was reinforced by the observation that infection of these cells with parental YFV-17D (capable of producing sfRNA) resulted in the formation of plaques that were much clearer and larger than those formed on the parental SW13 cells, where YFV- 17D infection resulted in not well-defined and turbid plaques (Fig. 4, upper left panel). The clearer and larger plaques observed for YFV-17D on these particular cell populations most

eGFP-YFV3’-UTR eGFP-WNV3’-UTR

After cell sorting

After passaging

Figure 5. Kinetics of eGFP expression of G418 resistant SW13 cells transfected with pcDNA3-eGFP-YFV3’-UTR and pcDNA3-eGFP-WNV3’-UTR plasmids after continuous passaging SW13 cells producing YFV or WNV sfRNA were analyzed by flow cytometry directly after cell sorting (upper panels) and after two weeks in culture (lower panels). The y-axis representing cell counts is plotted against eGFP expression level (x-axis). Fig. 5

138 Trans-complementation of sfRNA function likely originated from higher concentrations of sfRNA that were produced in these eGFP- flavivirus3’-UTR mRNA expressing cells.

Decrease of eGFP expression in SW13 cell populations over cultivation

Despite the promising results presented above, the described cell populations were apparently under genetic pressure to minimize the expression of the eGFP-flavivirus3’- UTR RNA and/or the production of sfRNA. As shown in Fig. 5, only a few passages under standard culture condition (in the presence of G418) already resulted in the accumulation of cells expressing lower levels of eGFP in comparison to the cells directly after sorting, as monitored by flow cytometry. These more heterogeneous cell populations in terms of eGFP expression and therefore most likely also in the amount of sfRNA produced were no longer able to complement the lack of cytopathogenicity of the YFV-BB-mutant (data not shown).

Discussion

It is now generally accepted that all flaviviruses produce subgenomic noncoding RNA (sfRNA) that results from the stalling of host 5’→3’ exoribonuclease XRN-1 by an RNA pseudoknot structure formed at the 5’ border of the distal, more conserved core element of the viral 3’-UTR [1-3;5;7;11]. The production of sfRNA in cells infected with mosquito-borne and no-known-vector (NKV) flaviviruses is associated with viral cytopathogenicity in cell culture [2;5;12] and pathogenesis in mice [2], as was shown with WNV, YFV-17D and Modoc virus (MODV) mutants that were deficient in sfRNA production, although the underlying mechanism is currently not well understood. Several other potential functions of sfRNA in the virus life cycle have also been suggested. The sfRNA was reported to be involved in viral evasion of host immune responses in WNV infected mammalian cells [13] and antagonizing RNA interference in both mammalian and insect cells infected with WNV and DENV [14]. It has also been hinted that WNV sfRNA might actually function as a precursor for a virus-derived miRNA in WNV infected mosquito cells [15] and as an inhibitor of Japanese encephalitis virus minus strand RNA synthesis and genome translation [21]. In view of the conservation of sfRNA production in the whole Flavivirus genus, it is rather surprising that sfRNA appears to play various roles. Perhaps there is a major benefit for flavivirus fitness and survival to select in favor of this feature during evolution and host adaptation, and these seemingly disparate pathways are somehow intertwined in a yet to be uncovered manner. As new tools for further unveiling sfRNA function, this report describes the establishment of SW13 cell lines that produce authentic YFV-17D or WNV sfRNA. In order to produce sfRNAs identical to those generated during infection, the complete WNVor

139 Chapter 6

YFV-17D 3’-UTR sequence was cloned as a 3’-trailer downstream of the recombinant eGFP coding sequence that is under the transcriptional control of a cytomegalovirus (CMV) early promoter. The autocatalytic hepatitis delta virus ribozyme (HDVr) was introduced to achieve authentic, non-polyadenylated 3’ end for YFV or WNV sfRNA. In vitro assays with commercially available purified XRN-1 and transfection of HEK293 and SW13 cells with the expression plasmids revealed that the XRN-1 stalling site was fully functional and that sfRNA was produced from the transcribed recombinant eGFP mRNA. A trans-complementation assay was used to evaluate whether these produced sfRNAs were biologically active. The ability of a YFV sfRNA-null mutant to form plaques on these cell lines was used as a readout for cytopathogenicity. Not only the YFV sfRNA, but surprisingly also the WNV sfRNA were both able to complement the lack of cytopathogenicity of the YFV-17D-BB mutant [3] on SW13 cells, as judged by the clearly visible plaques formed on the sfRNA expressing cells. The involvement of sfRNA in enhancing cytopathogenic effect (CPE) was further corroborated by the observation that the parental YFV-17D producing sfRNA [2;3] formed larger and clearer plaques on the SW13 cells expressing YFV or WNV sfRNA as compared to those formed on original SW13 cells. This enhanced cytopathogenicity of YFV-17D in SW13 cells expressing YFV and WNV sfRNA was most likely due to the higher concentrations of sfRNA produced in these cells in comparison to the original SW13 cells infected with YFV-17D. The fact that WNV sfRNA can trans-complement the cytopathogenicity of YFV-17D mutants like YFV-17D-BB seems to indicate that the enhancement of cytopathogenicity by sfRNA could be associated with conserved RNA sequences and/or structures, such as conserved sequence 1 and 2 (CS1, CS2), or pseudoknot 1 and 2 that are present in the 3’- UTR core region of the mosquito-borne flaviviruses [22-24], or to a broader extent, RNA elements conserved among the whole Flavivirus genus, for example, the 3’ stem-loop (3’ SL) [22;23]. It is therefore tempting to construct cell lines harboring sfRNA from other clusters of flaviviruses, and to evaluate their ability of cross-cluster complementation. If cross-cluster complementation is not possible, RNA elements that are featured in the mosquito-borne flaviviruses can be further pinpointed. It was previously reported that dengue virus (DENV) and YFV-17D mutants that lacked the CS2 sequence showed similar replication efficiency as the wildtype viruses, but were significantly less cytopathogenic [20;25]. To test whether it is indeed the CS2 region or other conserved RNA elements that are essential for the enhancement of cytopathogenicity, cell lines expressing sfRNA that lack certain sequences and/or structures or sfRNA derived from chimeric 3’-UTR that combines conserved RNA elements from different clusters will be informative. On the other hand, if cross-complementation is possible between viruses belonging to different clusters of flaviviruses, RNA elements conserved among the whole genus might be involved in the induction of cytopathic effect by mosquito-borne flavivirus in mammalian hosts. Alternatively, the stalling of XRN-1 at the 5’ end of the sfRNA, irrespective of the origin

140 Trans-complementation of sfRNA function of sfRNA, may actually sequester this ribonuclease. Given the high concentration of sfRNA especially during the later stages of infection [1-3;5;21], it is likely that the squelch of XRN-1 by sfRNA will impose a significant impact on the host 5’→3’ mRNA decay pathway and thus disrupt the homeostasis in the cells and, especially in combination with an ongoing infection, induce CPE. In fact, a recent publication by Moonet al. demonstrated that XRN-1 was indeed sequestered by WNV and DENV sfRNA in mammalian cells, resulting in significant changes in the mRNA transcription profile due to the alteration of mRNAs stability [10]. To further corroborate the causal relationship between XRN-1 suppression and cytopathogenicity, cells expressing XRN-1 stalling sites from other flaviviruses or even strong XRN-1 stalling sites from other origins [26;27] would be of great value. The affected mRNA turnover pathway, and therefore the disruption of cell homeostasis also provides an explanation as to why the SW13 cells that continuously expressed the eGFP mRNA and thus the XRN-1 suppressing sfRNA were relatively unstable and showed a reduced level of eGFP expression and sfRNA production upon repeated passaging. To counteract the negative effect from altered expression profile, it is likely that cells, especially those that produced high concentrations of sfRNA, were actively down-regulating the expression of the flaviviral 3’-UTR containing eGFP mRNA via genetic (for instance by deleting part of the viral 3’-UTR while keeping the G418 resistance gene) or epigenetic (such as by methylation of the CMV promoter) means; or the high expressors were gradually lost from the population due to apoptosis. Irrespective of the cause, the accumulation of lower expressors unfortunately minimizes the experimental window for these cells and restricts the use of this sfRNA producing cell line approach. Ideally, inducible cell lines are better suited for this type of study, in that transcription would be shut off during routine passage, while GFP and thus sfRNA expression could be turned on for the study of sfRNA functionality. In summary, cells expressing authentic sfRNA could serve as powerful tools for a better understanding of sfRNA function. Combined with biochemical analysis and other approaches, it will hopefully help elucidate how this RNA species contributes to viral fitness, and point out alternative strategies for vaccine and/or medication development for the medically important flaviviruses.

141 Chapter 6

Reference List

[1] Lin KC, Chang HL, Chang RY. Accumulation of a 3’-terminal genome fragment in Japanese encephalitis virus-infected mammalian and mosquito cells. J Virol 2004 May;78(10):5133-8.

[2] Pijlman GP, Funk A, Kondratieva N, et al. A highly structured, nuclease-resistant, noncoding RNA produced by flaviviruses is required for pathogenicity. Cell Host Microbe 2008 Dec 11;4(6):579-91.

[3] Silva PA, Pereira CF, Dalebout TJ, Spaan WJ, Bredenbeek PJ. An RNA pseudoknot is required for production of yellow fever virus subgenomic RNA by the host nuclease XRN1. J Virol 2010 Nov;84(21):11395-406.

[4] Urosevic N, van MM, Mansfield JP, Mackenzie JS, Shellam GR. Molecular characterization of virus-specific RNA produced in the brains of flavivirus-susceptible and -resistant mice after challenge with Murray Valley encephalitis virus. J Gen Virol 1997 Jan;78 ( Pt 1):23-9.

[5] Silva PA, Dalebout TJ, Gultyaev AP, Olsthoorn RC, Bredenbeek PJ. Characterization of the sfRNAs that are produced in cells infected with no known vector and cell fusing agent. 2013. Unpublished Work

[6] Scherbik SV, Paranjape JM, Stockman BM, Silverman RH, Brinton MA. RNase L plays a role in the antiviral response to West Nile virus. J Virol 2006 Mar;80(6):2987-99.

[7] Funk A, Truong K, Nagasaki T, et al. RNA structures required for production of subgenomic flavivirus RNA. J Virol 2010 Nov;84(21):11407-17.

[8] Lindenbach B, Rice CM. Flaviviridae: The Viruses and Their Replication. In: Knipe D.M, Howley P.M., editors. Fields virology. 5 ed. Philadelphia, Lippincott Williams and Wilkins, 2007: p. 991-1042.

[9] Eulalio A, Behm-Ansmant I, Izaurralde E. P bodies: at the crossroads of post-transcriptional pathways. Nat Rev Mol Cell Biol 2007 Jan;8(1):9-22.

[10] Moon SL, Anderson JR, Kumagai Y, et al. A noncoding RNA produced by arthropod-borne flaviviruses inhibits the cellular exoribonuclease XRN1 and alters host mRNA stability. RNA 2012 Nov;18(11):2029-40.

[11] Knaap R., Jiang X, Olsthoorn RC, Bredenbeek PJ. RNA structure requirements for XRN1 stalling in the tick- born encephalitis virus 3’untranslated region. 2013. Unpublished Work

[12] Jiang X, Silva PA, Dalebout TJ, Bredenbeek P. An infectious Modoc Virus cDNA as a tool to study conserved 3’-UTR RNA elements in flaviviruses with no known vector. 2013.

[13] Schuessler A, Funk A, Lazear HM, et al. West Nile virus non-coding subgenomic RNA contributes to viral evasion of type I interferon-mediated antiviral response. J Virol 2012 Feb 29.

[14] Schnettler E, Sterken MG, Leung JY, et al. Noncoding flavivirus RNA displays RNA interference suppressor activity in insect and Mammalian cells. J Virol 2012 Dec;86(24):13486-500.

[15] Hussain M, Torres S, Schnettler E, et al. West Nile virus encodes a microRNA-like small RNA in the 3’ untranslated region which up-regulates GATA4 mRNA and facilitates virus replication in mosquito cells. Nucleic Acids Res 2012 Mar;40(5):2210-23.

[16] Lindenbach BD, Rice CM. trans-Complementation of yellow fever virus NS1 reveals a role in early RNA replication. J Virol 1997 Dec;71(12):9608-17.

[17] Sambrook J, Fritsch T, Maniatis T. Molecular cloning: a Laboratory Manual. 1998. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

142 Trans-complementation of sfRNA function

[18] Inoue H, Nojima H, Okayama H. High efficiency transformation of Escherichia coli with plasmids. Gene 1990 Nov 30;96(1):23-8.

[19] Tilgner M, Shi PY. Structure and function of the 3’ terminal six nucleotides of the west nile virus genome in viral replication. J Virol 2004 Aug;78(15):8159-71.

[20] Bredenbeek PJ, Kooi EA, Lindenbach B, Huijkman N, Rice CM, Spaan WJ. A stable full-length yellow fever virus cDNA clone and the role of conserved RNA elements in flavivirus replication. J Gen Virol 2003 May;84(Pt 5):1261-8.

[21] Fan YH, Nadar M, Chen CC, Weng CC, Lin YT, Chang RY. Small noncoding RNA modulates japanese encephalitis virus replication and translation in trans. Virol J 2011;8:492.

[22] Hahn CS, Hahn YS, Rice CM, et al. Conserved elements in the 3’ untranslated region of flavivirus RNAs and potential cyclization sequences. J Mol Biol 1987 Nov 5;198(1):33-41.

[23] Markoff L. 5’- and 3’-noncoding regions in flavivirus RNA. Adv Viruses R 2003;59:177-228.

[24] Olsthoorn RC, Bol JF. Sequence comparison and secondary structure analysis of the 3’ noncoding region of flavivirus genomes reveals multiple pseudoknots. RNA 2001 Oct;7(10):1370-7.

[25] Men R, Bray M, Clark D, Chanock RM, Lai CJ. Dengue type 4 virus mutants containing deletions in the 3’ noncoding region of the RNA genome: analysis of growth restriction in cell culture and altered viremia pattern and immunogenicity in rhesus monkeys. J Virol 1996 Jun;70(6):3930-7.

[26] Esteban R, Vega L, Fujimura T. 20S RNA narnavirus defies the antiviral activity of SKI1/XRN1 in Saccharomyces cerevisiae. J Biol Chem 2008 Sep 19;283(38):25812-20.

[27] Poole TL, Stevens A. Structural modifications of RNA influence the 5’ exoribonucleolytic hydrolysis by XRN1 and HKE1 of Saccharomyces cerevisiae. Biochem Biophys Res Commun 1997 Jun 27;235(3):799-805.

143 144 Chapter 7

Epilogue

145 Chapter 7

In spite of the increasing incidence of non-communicable diseases, infectious diseases remain among the leading causes of human mortality and morbidity, accounting for more than 20% of the 57 million deaths worldwide in 2008 [1]. Contrary to predictions in the early 1960’s [2], our battle with infectious agents is far from reaching an end. With the old foes and newly emerging nuisances, it is just more realistic to accept the ever-continuing struggle with pathogens as a perpetual war. Among the emerging and reemerging infectious diseases, those caused by arthropod-borne viruses (arboviruses) display unique features, making them difficult to tackle. These viruses establish persistent infection in invertebrate vectors (e.g., mosquitoes, ticks, midges, sandflies), and can be transmitted to vertebrate hosts (humans, other mammals and birds) under favorable conditions [3;4]. Except for African swine fever virus, all arboviruses that infect vertebrates are RNA viruses [5], which is probably explained by the greater genetic plasticity and higher mutation rates exhibited by RNA viruses, which likely facilitate alternating replication in vertebrate and invertebrate hosts. Notorious arboviruses include members of the Bunyaviridae such as Rift Valley fever virus and Crimean-Congo hemorrhagic fever virus, as well as West Nile virus (WNV), Japanese encephalitis virus (JEV) and Dengue virus (DENV), which all belong to theFlavivirus genus within the Flaviviridae. In order to curb the spread of diseases caused by arboviruses, a comprehensive understanding of the molecular characteristics of their replication, virus-host interactions and host adaptation is of great value. Compared to our understanding of DNA viruses, however, our knowledge of the molecular biology and biochemistry of RNA viruses was lagging behind in the early days, since most of the molecular biology techniques aimed at modifying nucleic acids available were directed at DNA substrates. The development of the so called “infectious cDNA clone technology”, however, greatly spurred RNA virus research, especially that of plus-strand RNA viruses. For these viruses, the genomic RNA not only serves as the genetic material, it is also directly translated into viral proteins. Therefore, the introduction of a plus-strand RNA virus genome into a cell can initiate the complete viral life cycle and result in the production of progeny virions that can start the next round of infection. Based on this principle, the infectious clone technology was developed and became widely used for RNA viruses like picornaviruses [6-15] and alphaviruses [16-19], yet this approach turned out to be more challenging when applied to flaviviruses, largely due to the genetic instability of full-length flavivirus cDNAs during propagation in E. coli [20-22]. Initially the instability problem was circumvented by using an in vitro ligation approach in which full-length templates for run-off RNA transcription were constructed by in vitro ligation of appropriate restriction fragments [20;21]. Alternative strategies that were subsequently developed to overcome the instability included: 1) the use of different bacterial strains for cDNA transformation and propagation [22]; 2) insertion of short introns to prevent the unwanted expression of ORFs present in the viral genome that are potentially

146 Epilogue toxic to E. coli [23]; 3) exploiting low-copy number vectors for the assembly of full-length flavivirus cDNA clones [24;25]. A stable full-length clone for yellow fever virus strain 17D (YFV-17D) was successfully constructed in a low-copy number vector [25], which not only helped to deepen our insights into the molecular biology of this prototypic flavivirus [25], but also facilitated the design and construction of YFV-17D based recombinant vaccines [26-41]. Ever since its development in the 1930s [42], YFV-17D has stood as a paradigm for a successful live attenuated viral vaccine, with a great record of safety and efficacy, rendering it an appealing vector for genetically engineered recombinant vaccines against other diseases. In Chapter 2,YFV- 17D was exploited to express the glycoprotein (GPC) sequence of Lassa virus (LASV). The recombinant virus was proposed as a bivalent vaccine candidate against both YFV and LASV that circulate in the same region in West Africa. While characterizing the recombinant YFV- 17D expressing LASV GPC, genetic instability was encountered, as illustrated by the gradual and partial loss of the insert. Considering the duplication of the YFV NS1 signal peptide coding sequence flanking the GPC insert, we tried to stabilize the clone by changing the codons of the downstream signal peptide so as to minimize sequence similarity, which was postulated to induce homologous recombination and thus the loss of the insert. This strategy, however, turned out to be unsuccessful. Detailed analysis of recovered recombinant viruses strongly suggested that the large size of the insert was the major problem, which has been encountered and reported in other studies aimed at constructing recombinant flaviviruses [30;37]. Despite the lack of direct evidence for insert size limitations, most likely due to size restrictions of the icosahedral capsid, our success in stabilizing the viruses by using smaller inserts corresponding to the individual subunits GP1 and GP2 of the LASV GPC, provided support for this hypothesis. Therefore, it is advisable to take the length of the heterologous insert into consideration while constructing recombinant flaviviruses or RNA viruses with an icosahedral capsid structure in general. Another approach for the construction of infectious clones is the so-called infectious DNA (iDNA) or DNA/RNA layered approach, by which +ssRNA viruses are recovered in vivo after transfection of plasmid DNA carrying the viral genomic cDNA under the transcriptional control of a eukaryotic promoter. Using iDNA, the procedure for recovering infectious virus particles can be simplified, since there is no need forin vitro transcription; moreover, DNA is much more stable at room temperature in comparison to the RNA genome or RNA viruses. The first infectious cDNA clone for an animal RNA virus was reported for poliovirus [43], and was later described for viruses belonging to other groups [44;45]. Attempts to make infectious YFV DNA were again hampered by the genetic instability of the plasmidin E. coli. The addition of the cytomegalovirus (CMV) promoter sequence at the 5’ end of the YFV-17D cDNA appeared to exacerbate the instability problem. Spurious transcription from eukaryotic promoters in bacteria were proposed to be the cause, and several strategies

147 Chapter 7 to reduce the expression of toxic products were proposed, both at the transcriptional level or at the translational level [46]. In Chapter 3, we managed to stabilize the YFV-17D infectious DNA by replacing the low-copy number vector pACNR with a bacterial artificial chromosome (BAC), pBeloBAC11, which further reduced the copy number in E. coli. The iDNA was characterized in cell culture and was proven to be able to produce infectious progeny viruses. As proof of concept, the immunogenicity of the YFV-17D iDNA was studied in mice. Unfortunately, mice are not ideal animal models for YFV, because they rapidly resolve the infection after subcutaneous or intravenous administration or develop neurotropic symptoms upon intracerebral inoculation, whereas human infection leads to viscerotropic diseases, instead of encephalitis or other neurological disorders. The mechanism behind the murine resistance to YFV, be it the vaccine strain 17D or the virulent YFV-Asibi strain [47], is not fully understood and was postulated to be related to the murine innate immune response towards YFV infection [47;48]. We did not detect any neutralizing antibodies in the sera of HLA-A2.1 transgenic C57BL/6 mice vaccinated with the infectious YFV-17D DNA, despite readily detectable YFV specific CD8+ T cell response (Chapter 3). It was not clear to what extent the genetic background of C57BL/6 mice could affect B cell activation by YFV antigens, yet the lack of viremia in the sera seemed to suggest quick and efficient clearance of the virus. During the course of this study, Meier et al. reported a study with interferon α/β receptor deficient mice (A129 mice), and proposed to use these mice as better small animal models for the study of YFV-Asibi infection and YFV-17D vaccination [47]. Therefore, we tested immunogenicity of the YFV-17D iDNA in both the interferon receptor knockout A129 mice and the congenic 129 mice. We were indeed able to detect humoral immune responses in A129 mice vaccinated with YFV17D iDNA, but not in the immuno-competent 129 mice, confirming the role of the murine innate immune system in the rapid clearance of YFV infection in mice. Due to time and funding constraints, however, our experiments were relatively limited, so larger scale vaccination-challenge experiments in A129 mice as well as in primates are needed to further evaluate the potential of the YFV-17D iDNA as a vaccine candidate, and as a platform for the development of DNA based recombinant YFV vaccines. To spare vaccine dose and to facilitate the application in human beings, more acceptable delivery routes (microneedle patch, for instance [49]) could be explored. Despite the success of the YFV-17D vaccine [50-53] and the promise of the YFV-17D based chimeric vaccine platform [33;34;54-61], arthropod-borne (arbo-) flaviviruses still pose a great threat to human beings. A better understanding of viral-host interactions will provide valuable guidelines for the development of better vaccines and/or treatments. The host environment is by nature hostile towards any parasites, with no exception to viruses. The biological instinct is to either keep the pathogens at bay or to distinguish the non-self from self and to destroy the ‘invader’ upon their intrusion. During their evolution, mammals

148 Epilogue have developed an armamentarium involving a network of pathways. The first line of defense is the innate immune response, and it comes as no surprise that almost all animal viruses have evolved strategies to evade or delay the host’s innate immune responses. In fact, flavivirus-encoded proteins have been shown to inhibit both the interferon induction pathway and the interferon signaling pathways (reviewed in [62;63]). Interestingly, the recently identified subgenomic flavivirus RNA (sfRNA) has also been implicated as a viral innate immune evasion strategy [64]. Whereas subgenomic RNA production is a well known characteristic for certain +ssRNA families (e.g., coronaviruses and alphaviruses), the flavivirus subgenomic RNA that is colinear with the distal part of the viral 3’-UTR was not discovered in cells or animals infected with flavivirus until recently [65-68]. This unique small RNA species is clearly different from subgenomic (m)RNAs produced by other plus-strand RNA viruses in that it does not encode for any viral protein. Even more remarkably, sfRNA is not directly produced by transcription involving the viral RdRP. Instead, sfRNAs were confirmed to be incomplete 5’→3’ degradation products of the viral genomes by the host exoribonuclease XRN-1 [68;69]. Though uncommon to RNA viruses that in general possess a compact genome, virus-derived noncoding RNAs are, on the other hand, transcribed by many DNA viruses, which have been found to interact with various cellular pathways and to serve distinct functions. For example, Marek’s disease virus noncoding RNA was found to be a precursor in the production of miRNA [70], whereas Herpesvirus saimiri noncoding RNA serves as an antagonist for cellular miRNA [71]. A 2.7kb CMV-encoded RNA was reported to regulate mitochondria-induced cell death [72]. In fact, WNV sfRNA was suggested to be further processed to yield an miRNA-like small RNA that was identified in infected mosquito cells [73], although it cannot be ruled out that the WNV genome RNA actually serves as the precursor for this small RNA, given the sequence identity between the sfRNA and viral 3’-UTR. However, it is more likely that viral genomic RNA, intertwined with viral as well as host proteins and wrapped in translation and replication complexes, is inaccessible to the miRNA machinery, whereas sfRNA has been found to localize to the P-bodies, where shRNA that is produced by Drosha in the nucleus undergoes further processing and is eventually assembled into the RISC complex [68]. The sfRNA has also been shown to interfere with RNAi in both mammalian and insect cells [74], and was suggested to be correlated with viral cytopathogenicity and pathogenesis [68;69]. Nevertheless, contrary to the CMV-encoded 2.7kb RNA that prevents cells from entering apoptosis, flavivirus subgenomic RNA appeared to enhance cytopathic effect (CPE) during viral infection, since mutants incapable of producing sfRNA were unable to induce plaque formation in several mammalian cell lines, yielded relatively small infectious centers, and were less pathogenic in small animal models compared to sfRNA-producing wildtype viruses [68;69] (Chapters 4 and 5). It is paradoxical that killing instead of sustaining the host could benefit the virus, or does the accumulation of sfRNA serve as a temporal switch from

149 Chapter 7 replication to the release of progeny viruses that can potentially be facilitated by the onset of CPE? In order to complete their replicative cycle and perpetuate, arboviruses also need to survive and thrive in arthropod vectors, where RNAi plays a major role in the host responses that combat infection. Just as plant and insect viruses, arboviruses have also evolved strategies to counteract this host response by expressing RNA interference suppressors (RSS). Strikingly, no flavivirus-encoded RSS proteins have been identified thus far, yet the noncoding sfRNA was proposed to (at least in part) fulfill the role of RNAi suppressor in insect cells [74]. Given the differentexperimental settings and approaches used for the study of sfRNA function, it is rather difficult to reach a consensus on sfRNA functionality. It is also hard to envision how these apparently different functions in immune modulation, RNAi suppression, miRNA production and inhibition of viral minus strand RNA synthesis and translation [64;73-75] are somehow connected to contribute to virus fitness. On the one hand, the conservation of sfRNA production throughout theFlavivirus genus seems to suggest a crucial function that has been preserved through evolution. Alternatively, sfRNA may interact with distinctive host factors to fulfill disparate roles in viruses that belong to different clusters and exhibit a unique host range. It is beneficial for arbo-flaviviruses to suppress the RNAi pathway in insects so as to reach high titers that will maximize their chances of transmission to a mammalian host. RNAi suppression would also benefit replication of insect-specific flaviviruses in insects. If this is the sole function of sfRNA, however, what would be the rationale of sfRNA production in NKV flaviviruses that do not require an arthropod vector for viral transmission. On the other hand, an immune evasion strategy in mammalian hosts will be beneficial for both the arbo- and NKV- flaviviruses, but does not provide an explanation for sfRNA production by insect flaviviruses. In this sense, there must have been either functional divergence during evolution or the use of a certain conserved pathway, such as RNAi, apoptosis, or RNA degradation pathways, which somehow contributes to viral fitness in disparate hosts. In order to test the proposed sfRNA functions in an experimental setting independent from viral replication or viral protein expression, yet still pertaining to the intracellular physiological environment, mammalian and insect cell lines expressing sfRNA of flaviviruses belonging to various clusters could be very helpful. We have established mammalian cell lines harboring a gene cassette containing the complete 3’-UTR sequence of YFV-17D or WNV downstream of a reporter gene (eGFP) under the transcriptional control of a CMV promoter. These cell lines constitutively express authentic YFV-17D or WNV sfRNA. Our data showed that not only the YFV-17D sfRNA, but also the WNV sfRNA could complement the lack of cytopathogenic effect of an sfRNA-null YFV mutant (Chapter 6). This finding can be explained if conserved RNA elements within the 3’-UTR of mosquito-borne flaviviruses

150 Epilogue interact with either host factors or viral components and promote cytopathogenicity. Alternatively, the inhibition of XRN-1 activity [76] per se might enhance cytopathology upon virus infection. We can test the latter hypothesis by using mammalian cell lines expressing XRN-1 stalling signals from other flaviviruses (e.g., NKV flaviviruses) with the trans- complementation approach. In addition, insect cell lines expressing sfRNA from insect- specific or arbo-flaviviruses can be established to study whether sfRNA by itself antagonizes cellular RNA interference pathways, and whether perturbation of this pathway contributes to the fitness of flaviviruses belonging to other clusters, or unrelated insect or arboviruses in general. Although no consensus has been reached regarding sfRNA functionality, the production of sfRNA has been unambiguously confirmed to be driven by a host enzyme, XRN-1 [68;69;77;78]. XRN-1 is a well conserved and highly processive cellular exoribonuclease that functions as the key enzyme in the 5’→3’ mRNA degradation pathway in eukaryotic cells. So far, only a few non-viral RNA structures like oligo(G) stretches and large thermodynamically favored stem-loop structures have been reported to be able to stall XRN-1, although the stalling efficiency of the stem-loop structures was much lower compared to that achieved by oligo(G) tracts or flavivirus stalling signals [79]. Some other structures of viral origin have also been reported to defy 5’→3’ degradation, such as the long stem structure with four Gs buried at the bottom of the 20S RNA of Narnavirus [80], and a 58-nucleotide sequence containing stem-loop structures responsible for the accumulation of small viral non-coding RNA in plants and protoplasts infected with red clover necrotic mosaic virus (RCNMV) [81]. For the production of sfRNA, XRN-1 was proposed to be stalled upstream of acomplex RNA structure involving a stem-loop structure that is stabilized by a pseudoknot interaction [69;77;78;82]. With their relatively small genome size and limited coding capacity, it is not uncommon for RNA viruses to usurp specific RNA motifs or versatile RNA structures to fulfill important roles in their replication. Pseudoknots, RNA tertiary structures involving at least two interacting stem-loop structures, are especially prevalent in RNA viruses [83] and have been identified as essential regulators in transcription and replication, inducers of ribosomal frameshifting, or structures that help to maintain RNA conformation for enzymatic function, as in ribozymes or telomerases (reviewed in [83]). A frameshift-inducing pseudoknot was recently identified in the NS1 coding sequence of flaviviruses belonging to the JEV serogroup. This pseudoknot is required for a frameshift that results in the production of a larger version of the viral NS1 protein [84]. Additional RNA pseudoknots were predicted to be formed in the 3’-UTR of the mosquito-borne flaviviruses, namely PSK1, PSK2 and PSK3 [85]. PSK1 and PSK2 were shown to be important for viral replication and/or translation [86] (Bredenbeek et al., unpublished data), whereas the more upstream pseudoknot (PSK3) was found to be nonessential for virus replication [69]. This XRN-1 stalling RNA pseudoknot structure exhibits

151 Chapter 7 special features that distinguish it from classical H-type RNA pseudoknots. It does bear some structural resemblance to the RNase L antagonizing enterovirus RNA, which competitively inhibits the antiviral endoribonuclease domain of RNase L, thus protecting the viral genome from endoribonuclease cleavage [87]. However, in vitro studies did not provide convincing evidence for inhibition of RNase L activity by sfRNA [64]. Interestingly, the non-H-type RNA pseudoknot structure in sfRNA was found to inhibit the enzymatic activity of another cellular ribonuclease, i.e., XRN-1, such that cellular mRNA turnover is perturbed, resulting in a different gene expression profile in cells that are infected with wildtype WNV producing sfRNA versus cells infected with a mutant WNV that is unable to stall XRN-1 or uninfected cells [76]. The accumulation of uncapped mRNA could potentially affect other cellular pathways, such as apoptosis and the stress response, possibly explaining the observation that sfRNA-deficient WNV and YFV-17D mutants are less cytopathogenic in certain cells of mammalian origin. Irrespective of their eraction int partner, the function that RNA pseudoknot structures fulfill largely depends on the RNA conformation as well as its thermostability. It can be envisioned, therefore, that the change of a single nucleotide within the XRN- 1 stalling structure may incapacitate XRN-1 stalling (Chapter 5). Despite our extensive mutagenesis studies of the XRN-1 stalling site of YFV-17D and MODV (Chapter 4 and Chapter 5), representing the mosquito-borne and NKV flaviviruses, a solid conclusion regarding the sequence/structure requirements for XRN-1 stalling is far from being reached. It is hard to envision how the loops and stems are oriented in space, or how XRN-1 binding might change the 3-D conformation, and we are not yet clear whether there is any specific nucleotide that is critical in the RNA-protein interaction. A better understanding of this structure probably requires additional data from in vivo biochemical probing analysis [88] and mutagenesis studies and would ideally involve structural biology (e.g., X-ray crystallography or N.M.R.). Thanks to the versatile tools and novel techniques developed in molecular biology in the last two decades of the previous century, we have been able to construct infectious clones of flaviviruses, with which extensive knowledge has been gained on virus biology and which have been used to develop vaccines against pathogenic flaviviruses such as JEV [39;40]. Nonetheless, question marks still remain regarding the molecular determinants of host restriction, sfRNA function and the sequence/structure requirements for sfRNA production, etc. Studies aiming to address these fundamental questions will help fill our knowledge gap in the understanding of virus-host and/or virus-vector interaction, which will in turn point out directions for the design of better vaccines and antivirals to combat the (re) emerging flaviviruses and to limit the impact of these pathogens on society.

152 Epilogue

Reference List

[1] WHO health report, 2008. 2008.

[2] Cockburn A. The Evolution and Eradication of Infectious Diseases. 1963. Baltimore, Johns Hopkins University Press.

[3] Ciota AT, Kramer LD. Insights into arbovirus evolution and adaptation from experimental studies. Viruses 2010 Dec;2(12):2594-617.

[4] Gould EA, Higgs S. Impact of climate change and other factors on emerging arbovirus diseases. Trans R Soc Trop Med Hyg 2009 Feb;103(2):109-21.

[5] Weaver SC, Reisen WK. Present and future arboviral threats. Antiviral Res 2010 Feb;85(2):328-45.

[6] Cohen JI, Ticehurst JR, Feinstone SM, Rosenblum B, Purcell RH. Hepatitis A virus cDNA and its RNA transcripts are infectious in cell culture. J Virol 1987 Oct;61(10):3035-9.

[7] Duechler M, Skern T, Blaas D, Berger B, Sommergruber W, Kuechler E. Human rhinovirus serotype 2: in vitro synthesis of an infectious RNA. Virology 1989 Jan;168(1):159-61.

[8] Duke GM, Palmenberg AC. Cloning and synthesis of infectious cardiovirus RNAs containing short, discrete poly(C) tracts. J Virol 1989 Apr;63(4):1822-6.

[9] Klump WM, Bergmann I, Muller BC, Ameis D, Kandolf R. Complete nucleotide sequence of infectious Coxsackievirus B3 cDNA: two initial 5’ uridine residues are regained during plus-strand RNA synthesis. J Virol 1990 Apr;64(4):1573-83.

[10] Mizutani S, Colonno RJ. In vitro synthesis of an infectious RNA from cDNA clones of human rhinovirus type 14. J Virol 1985 Nov;56(2):628-32.

[11] Roos RP, Stein S, Ohara Y, Fu JL, Semler BL. Infectious cDNA clones of the DA strain of Theiler’s murine encephalomyelitis virus. J Virol 1989 Dec;63(12):5492-6.

[12] Sarnow P. Role of 3’-end sequences in infectivity of poliovirus transcripts made in vitro. J Virol 1989 Jan;63(1):467-70.

[13] Tangy F, McAllister A, Brahic M. Molecular cloning of the complete genome of strain GDVII of Theiler’s virus and production of infectious transcripts. J Virol 1989 Mar;63(3):1101-6.

[14] van der Werf S, Bradley J, Wimmer E, Studier FW, Dunn JJ. Synthesis of infectious poliovirus RNA by purified T7 RNA polymerase. Proc Natl Acad Sci U S A 1986 Apr;83(8):2330-4.

[15] Zibert A, Maass G, Strebel K, Falk MM, Beck E. Infectious foot-and-mouth disease virus derived from a cloned full-length cDNA. J Virol 1990 Jun;64(6):2467-73.

[16] Davis NL, Willis LV, Smith JF, Johnston RE. In vitro synthesis of infectious venezuelan equine encephalitis virus RNA from a cDNA clone: analysis of a viable deletion mutant. Virology 1989 Jul;171(1):189-204.

[17] Kuhn RJ, Niesters HG, Hong Z, Strauss JH. Infectious RNA transcripts from Ross River virus cDNA clones and the construction and characterization of defined chimeras with Sindbis virus. Virology 1991 Jun;182(2):430- 41.

153 Chapter 7

[18] Liljestrom P, Lusa S, Huylebroeck D, Garoff H. In vitro mutagenesis of a full-length cDNA clone of Semliki Forest virus: the small 6,000-molecular-weight membrane protein modulates virus release. J Virol 1991 Aug;65(8):4107-13.

[19] Rice CM, Levis R, Strauss JH, Huang HV. Production of infectious RNA transcripts from Sindbis virus cDNA clones: mapping of lethal mutations, rescue of a temperature-sensitive marker, and in vitro mutagenesis to generate defined mutants. J Virol 1987 Dec;61(12):3809-19.

[20] Sumiyoshi H, Hoke CH, Trent DW. Infectious Japanese encephalitis virus RNA can be synthesized from in vitro-ligated cDNA templates. J Virol 1992 Sep;66(9):5425-31.

[21] Rice CM, Grakoui A, Galler R, Chambers TJ. Transcription of infectious yellow fever RNA from full-length cDNA templates produced by in vitro ligation. New Biol 1989 Dec;1(3):285-96.

[22] Lai CJ, Zhao BT, Hori H, Bray M. Infectious RNA transcribed from stably cloned full-length cDNA of dengue type 4 virus. Proc Natl Acad Sci U S A 1991 Jun 15;88(12):5139-43.

[23] Yamshchikov V, Mishin V, Cominelli F. A new strategy in design of +RNA virus infectious clones enabling their stable propagation in E. coli. Virology 2001 Mar 15;281(2):272-80.

[24] Ruggli N, Tratschin JD, Mittelholzer C, Hofmann MA. Nucleotide sequence of classical swine fever virus strain Alfort/187 and transcription of infectious RNA from stably cloned full-length cDNA. J Virol 1996 Jun;70(6):3478-87.

[25] Bredenbeek PJ, Kooi EA, Lindenbach B, Huijkman N, Rice CM, Spaan WJ. A stable full-length yellow fever virus cDNA clone and the role of conserved RNA elements in flavivirus replication. J Gen Virol 2003 May;84(Pt 5):1261-8.

[26] Bonaldo MC, Caufour PS, Freire MS, Galler R. The yellow fever 17D vaccine virus as a vector for the expression of foreign proteins: development of new live flavivirus vaccines. Mem Inst Oswaldo Cruz 2000;95 Suppl 1:215-23.

[27] Bonaldo MC, Garratt C,R Caufour PS, et al. Surface expression of an immunodominant malaria protein B cell epitope by yellow fever virus. J Mol Biol 2002 Jan 25;315(4):873-85.

[28] Bonaldo MC, Garratt C, R Marchevsky RS, et al. Attenuation of recombinant yellow fever 17D viruses expressing foreign protein epitopes at the surface. J Virol 2005 Jul;79(13):8602-13.

[29] Bonaldo MC, Garratt C, R Freire MS, Galler R. Expression of foreign protein epitopes at the surface of recombinant yellow fever 17D viruses based on three-dimensional modeling of its envelope protein. Cell Biochem Biophys 2006;44(3):313-24.

[30] Bonaldo MC, Mello SM, Trindade GF, et al. Construction and characterization of recombinant flaviviruses bearing insertions between E and NS1 genes. Virol J 2007;4:115.

[31] Bonaldo MC, Martins MA, Rudersdorf R, et al. Recombinant yellow fever vaccine virus 17D expressing simian immunodeficiency virus SIVmac239 gag induces SIV-specific CD8+ T-cell responses in rhesus macaques. J Virol 2010 Apr;84(7):3699-706.

[32] Bredenbeek PJ, Molenkamp R, Spaan WJ, et al. A recombinant Yellow Fever 17D vaccine expressing Lassa virus glycoproteins. Virology 2006 Feb 20;345(2):299-304.

[33] Guy B, Saville M, Lang J. Development of Sanofi Pasteur tetravalent dengue vaccine. Hum Vaccin 2010 Sep 16;6(9).

154 Epilogue

[34] Guy B, Guirakhoo F, Barban V, Higgs S, Monath TP, Lang J. Preclinical and clinical development of YFV 17D-based chimeric vaccines against dengue, West Nile and Japanese encephalitis viruses. Vaccine 2010 Jan 8;28(3):632-49.

[35] Jiang X, Dalebout TJ, Bredenbeek PJ, et al. Yellow fever 17D-vectored vaccines expressing Lassa virus GP1 and GP2 glycoproteins provide protection against fatal disease in guinea pigs. Vaccine 2011 Feb 1;29(6):1248-57.

[36] McAllister A, Arbetman AE, Mandl S, Pena-Rossi C, Andino R. Recombinant yellow fever viruses are effective therapeutic vaccines for treatment of murine experimental solid tumors and pulmonary metastases. J Virol 2000 Oct;74(19):9197-205.

[37] Stoyanov CT, Boscardin SB, Deroubaix S, et al. Immunogenicity and protective efficacy of a recombinant yellow fever vaccine against the murine malarial parasite Plasmodium yoelii. Vaccine 2010 Jun 23;28(29):4644-52.

[38] Tao D, Barba-Spaeth G, Rai U, Nussenzweig V, Rice CM, Nussenzweig RS. Yellow fever 17D as a vaccine vector for microbial CTL epitopes: protection in a rodent malaria model. J Exp Med 2005 Jan 17;201(2):201- 9.

[39] Appaiahgari MB, Vrati S. IMOJEV((R)): a Yellow fever virus-based novel Japanese encephalitis vaccine. Expert Rev Vaccines 2010 Dec;9(12):1371-84.

[40] Appaiahgari MB, Vrati S. Clinical development of IMOJEV (R)--a recombinant Japanese encephalitis chimeric vaccine (JE-CV). Expert Opin Biol Ther 2012 Sep;12(9):1251-63.

[41] Franco D, Li W, Qing F, et al. Evaluation of yellow fever virus 17D strain as a new vector for HIV-1 vaccine development. Vaccine 2010 Aug 9;28(35):5676-85.

[42] Theiler M, Smith HH. The use of yellow fever virus modified byin vitro cultivation for human immunization. J. Exp. Med. 65, 787-800 (1937). Rev Med Virol 2000 Jan;10(1):6-16.

[43] Racaniello VR, Baltimore D. Cloned poliovirus complementary DNA is infectious in mammalian cells. Science 1981 Nov 20;214(4523):916-9.

[44] Herweijer H, Latendresse JS, Williams P, et al. A plasmid-based self-amplifying Sindbis virus vector. Hum Gene Ther 1995 Sep;6(9):1161-7.

[45] Lee C, Calvert JG, Welch SK, Yoo D. A DNA-launched reverse genetics system for porcine reproductive and respiratory syndrome virus reveals that homodimerization of the nucleocapsid protein is essential for virus infectivity. Virology 2005 Jan 5;331(1):47-62.

[46] Mishin VP, Cominelli F, Yamshchikov VF. A ‘minimal’ approach in design of flavivirus infectious DNA. Virus Res 2001 Dec 4;81(1-2):113-23.

[47] Meier KC, Gardner CL, Khoretonenko MV, Klimstra WB, Ryman KD. A mouse model for studying viscerotropic disease caused by yellow fever virus infection. PLoS Pathog 2009 Oct;5(10):e1000614.

[48] Sangster MY, Mackenzie JS, Shellam GR. Genetically determined resistance to flavivirus infection in wild Mus musculus domesticus and other taxonomic groups in the genus Mus. Arch Virol 1998;143(4):697-715.

[49] Prausnitz MR, Mikszta JA, Cormier M, Andrianov AK. Microneedle-based vaccines. Curr Top Microbiol Immunol 2009;333:369-93.

[50] Monath TP. Yellow fever vaccine. Expert Rev Vaccines 2005 Aug;4(4):553-74.

155 Chapter 7

[51] WHO. The yellow fever initiative, 2007. www.who.int/csr/disease/yellowfev/introduction/en/index/html

[52] Barrett AD. Yellow fever vaccines. Biologicals 1997 Mar;25(1):17-25.

[53] Pulendran B. Learning immunology from the yellow fever vaccine: innate immunity to systems vaccinology. Nat Rev Immunol 2009 Oct;9(10):741-7.

[54] Arroyo J, Miller C, Catalan J, et al. ChimeriVax-West Nile virus live-attenuated vaccine: preclinical evaluation of safety, immunogenicity, and efficacy. J Virol 2004 Nov;78(22):12497-507.

[55] Chambers TJ, Nestorowicz A, Mason PW, Rice CM. Yellow fever/Japanese encephalitis chimeric viruses: construction and biological properties. J Virol 1999 Apr;73(4):3095-101.

[56] Galler R, Marchevsky RS, Caride E, et al. Attenuation and immunogenicity of recombinant yellow fever 17D-dengue type 2 virus for rhesus monkeys. Braz J Med Biol Res 2005 Dec;38(12):1835-46.

[57] Guirakhoo F, Arroyo J, Pugachev KV, et al. Construction, safety, and immunogenicity in nonhuman primates of a chimeric yellow fever-dengue virus tetravalent vaccine. J Virol 2001 Aug;75(16):7290-304.

[58] Guirakhoo F, Kitchener S, Morrison D, et al. Live attenuated chimeric yellow fever dengue type 2 (ChimeriVax-DEN2) vaccine: Phase I clinical trial for safety and immunogenicity: effect of yellow fever pre- immunity in induction of cross neutralizing antibody responses to all 4 dengue serotypes. Hum Vaccin 2006 Mar;2(2):60-7.

[59] Halstead SB, Thomas SJ. New Japanese encephalitis vaccines: alternatives to production in mouse brain. Expert Rev Vaccines 2011 Mar;10(3):355-64.

[60] Monath TP, Soike K, Levenbook I, et al. Recombinant, chimaeric live, attenuated vaccine (ChimeriVax) incorporating the envelope genes of Japanese encephalitis (SA14-14-2) virus and the capsid and nonstructural genes of yellow fever (17D) virus is safe, immunogenic and protective in non-human primates. Vaccine 1999 Apr 9;17(15-16):1869-82.

[61] Monath TP, Levenbook I, Soike K, et al. Chimeric yellow fever virus 17D-Japanese encephalitis virus vaccine: dose-response effectiveness and extended safety testing in rhesus monkeys. J Virol 2000 Feb;74(4):1742- 51.

[62] Munoz-Jordan JL, Sanchez-Burgos GG, Laurent-Rolle M, Garcia-Sastre A. Inhibition of interferon signaling by dengue virus. Proc Natl Acad Sci U S A 2003 Nov 25;100(24):14333-8.

[63] Munoz-Jordan JL, Fredericksen BL. How flaviviruses activate and suppress the interferon response. Viruses 2010 Feb;2(2):676-91.

[64] Schuessler A, Funk A, Lazear HM, et al. West Nile virus non-coding subgenomic RNA contributes to viral evasion of type I interferon-mediated antiviral response. J Virol 2012 Feb 29.

[65] Urosevic N, van MM, Mansfield JP, Mackenzie JS, Shellam GR. Molecular characterization of virus-specific RNA produced in the brains of flavivirus-susceptible and -resistant mice after challenge with Murray Valley encephalitis virus. J Gen Virol 1997 Jan;78 ( Pt 1):23-9.

[66] Lin KC, Chang HL, Chang RY. Accumulation of a 3’-terminal genome fragment in Japanese encephalitis virus-infected mammalian and mosquito cells. J Virol 2004 May;78(10):5133-8.

[67] Scherbik SV, Paranjape JM, Stockman BM, Silverman RH, Brinton MA. RNase L plays a role in the antiviral response to West Nile virus. J Virol 2006 Mar;80(6):2987-99.

156 Epilogue

[68] Pijlman GP, Funk A, Kondratieva N, et al. A highly structured, nuclease-resistant, noncoding RNA produced by flaviviruses is required for pathogenicity. Cell Host Microbe 2008 Dec 11;4(6):579-91.

[69] Silva PA, Pereira CF, Dalebout TJ, Spaan WJ, Bredenbeek PJ. An RNA pseudoknot is required for production of yellow fever virus subgenomic RNA by the host nuclease XRN1. J Virol 2010 Nov;84(21):11395-406.

[70] Burnside J, Bernberg E, Anderson A, et al. Marek’s disease virus encodes MicroRNAs that map to meq and the latency-associated transcript. J Virol 2006 Sep;80(17):8778-86.

[71] Cazalla D, Yario T, Steitz JA. Down-regulation of a host microRNA by a Herpesvirus saimiri noncoding RNA. Science 2010 Jun 18;328(5985):1563-6.

[72] Reeves MB, Davies AA, McSharry BP, Wilkinson GW, Sinclair JH. Complex I binding by a virally encoded RNA regulates mitochondria-induced cell death. Science 2007 Jun 1;316(5829):1345-8.

[73] Hussain M, Torres S, Schnettler E, et al. West Nile virus encodes a microRNA-like small RNA in the 3’ untranslated region which up-regulates GATA4 mRNA and facilitates virus replication in mosquito cells. Nucleic Acids Res 2012 Mar;40(5):2210-23.

[74] Schnettler E, Sterken MG, Leung JY, et al. Noncoding flavivirus RNA displays RNA interference suppressor activity in insect and Mammalian cells. J Virol 2012 Dec;86(24):13486-500.

[75] Fan YH, Nadar M, Chen CC, Weng CC, Lin YT, Chang RY. Small noncoding RNA modulates japanese encephalitis virus replication and translation in trans. Virol J 2011;8:492.

[76] Moon SL, Anderson JR, Kumagai Y, et al. A noncoding RNA produced by arthropod-borne flaviviruses inhibits the cellular exoribonuclease XRN1 and alters host mRNA stability. RNA 2012 Nov;18(11):2029-40.

[77] Knaap R., Jiang X, Olsthoorn RC, Bredenbeek PJ. RNA structure requirements for XRN1 stalling in the tick- born encephalitis virus 3’untranslated region. unpublished

[78] Silva PA, Dalebout TJ, Gultyaev AP, Olsthoorn RC, Bredenbeek PJ. Characterization of the sfRNAs that are produced in cells infected with no known vector and cell fusing agent. unpublished

[79] Poole TL, Stevens A. Structural modifications of RNA influence the 5’ exoribonucleolytic hydrolysis by XRN1 and HKE1 of Saccharomyces cerevisiae. Biochem Biophys Res Commun 1997 Jun 27;235(3):799-805.

[80] Esteban R, Vega L, Fujimura T. 20S RNA narnavirus defies the antiviral activity of SKI1/XRN1 in Saccharomyces cerevisiae. J Biol Chem 2008 Sep 19;283(38):25812-20.

[81] Iwakawa HO, Mizumoto H, Nagano H, et al. A viral noncoding RNA generated by cis-element-mediated protection against 5’->3’ RNA decay represses both cap-independent and cap-dependent translation. J Virol 2008 Oct;82(20):10162-74.

[82] Funk A, Truong K, Nagasaki T, et al. RNA structures required for production of subgenomic flavivirus RNA. J Virol 2010 Nov;84(21):11407-17.

[83] Brierley I, Pennell S, Gilbert RJ. Viral RNA pseudoknots: versatile motifs in gene expression and replication. Nat Rev Microbiol 2007 Aug;5(8):598-610.

[84] Melian EB, Hinzman E, Nagasaki T, et al. NS1’ of flaviviruses in the Japanese encephalitis virus serogroup is a product of ribosomal frameshifting and plays a role in viral neuroinvasiveness. J Virol 2010 Feb;84(3):1641- 7.

[85] Olsthoorn RC, Bol JF. Sequence comparison and secondary structure analysis of the 3’ noncoding region of flavivirus genomes reveals multiple pseudoknots. RNA 2001 Oct;7(10):1370-7.

157 Chapter 7

[86] Manzano M, Reichert ED, Polo S, et al. Identification of cis-acting elements in the 3’-untranslated region of the dengue virus type 2 RNA that modulate translation and replication. J Biol Chem 2011 Jun 24;286(25):22521-34.

[87] Townsend HL, Jha BK, Han JQ, Maluf NK, Silverman RH, Barton DJ. A viral RNA competitively inhibits the antiviral endoribonuclease domain of RNase L. RNA 2008 Jun;14(6):1026-36.

[88] Spitale RC, Crisalli P, Flynn RA, Torre EA, Kool ET, Chang HY. RNA SHAPE analysis in living cells. Nat Chem Biol 2013 Jan;9(1):18-20.

158 Summary

159 Summary

Although mankind has claimed victories over many pathogens in recent history, infectious diseases still remain among the leading causes of human morbidity and mortality. In contrast to the (nearly) eradicated smallpox virus and poliovirus that have human beings as their sole host, arthropod-borne viruses (arboviruses) are more difficult to eliminate, due to their use of invertebrate vectors for replication and transmission to humans and other mammalian hosts. Among the emerging and reemerging infectious agents, viruses belonging to the Flavivirus genus pose great threats to humans all over the world, causing febrile diseases, hemorrhagic fever, encephalitis or even death. A general introduction to flaviviruses is presented in Chapter 1 that could help better understand and appreciate the topics covered by this thesis. In order to conquer these medically important viruses, a comprehensive understanding of their molecular biology, as well as their interactions with the host and the vector is required. The successful development of reverse genetic systems for arbo- flaviviruses has not only facilitated the research into their biology but also openedup new possibilities to attenuate these viruses for the development of vaccine candidates to protect mankind and livestock from the diseases caused by mosquito- and tick-borne flaviviruses. The experimental chapters in this thesis mainly focus on the construction and characterization of flavivirus infectious clones and the application of these useful tools to address fundamental questions on virus biology or to explore novel vaccine platforms based on the yellow fever virus vaccine strain (YFV-17D). With a great record of safety and efficacy and the availability of a full-length cDNA clone, YFV-17D has been explored as a vector for the development of chimeric or recombinant vaccines. Chapter 2 describes the construction and characterization of recombinant YFV- 17D viruses expressing Lassa virus (LASV) glycoprotein coding sequences as bivalent vaccine candidates against wildtype YFV and LASV that circulate in the same region in West Africa. Despite successful mosquito control and vaccination campaigns in the past, resurgence of yellow fever has been observed in recent years, which is mainly due to a build-up of unprotected populations in high-risk areas. Chapter 3 of this thesis describes the construction and characterization of an alternative YFV-17D vaccine candidate based on a DNA launched YFV-17D infectious clone. With the incorporation of eukaryotic transcription promoter and terminator sequences and a hepatitis delta virus ribozyme sequence, YFV-17D infection could be initiated by the introduction of this YFV-17D infectious DNA (iDNA) into cells or organisms. The immunological characteristics of the YFV-17D iDNA were evaluated in murine models. This DNA/LAV (live attenuated vaccine) layered approach could potentially reduce the costs of YFV-17D vaccine production and facilitate vaccine transportation and distribution. In contrast to the pathogenic arthropod-borne flaviviruses, no known vector (NKV) flaviviruses that have been exclusively isolated from mammals are less-well studied, which

160 Summary is in part due to the lack of infectious clones. Chapter 4 describes the construction and characterization of the first NKV flavivirus infectious clone, which was further utilized to study the functional significance of RNA elements present in Modoc virus (MODV) 3’-UTR (untranslated region), as a representative of the NKV flaviviruses. The differences/similarities between NKV flavivirus 3’-UTR and that of vector-borne flaviviruses are discussed in this chapter. A recently discovered common feature for all flaviviruses is the production of a small flavivirus RNA (sfRNA) in infected cells or animals. The production of this sfRNA results from the incomplete degradation of the viral genome by the host exoribonuclease, XRN-1. XRN-1 was predicted to be stalled by a non-H-type RNA pseudoknot structure that is present is the 3’-UTR of every flavivirus studied thus far. Chapter 5 presents a detailed mutagenesis study of many of the unique sequence and structural characteristics of the XRN-1 stalling site in mosquito-borne flaviviruses using YFV-17D as a model. The impact of these mutations on virus replication, sfRNA production and viral cytopathogenicity were analyzed in detail. Chapter 6 describes the establishment of mammalian cell lines that constitutively expressed sfRNA from YFV-17D and WNV, which were further analyzed for their ability to trans-complement the cytopathogenicity of a YFV-17D mutant that was deficient in sfRNA production. Surprisingly, trans-complementation of the sfRNA-null YFV-17D mutant was not only observed in cell lines expressing the YFV-17D sfRNA but also in cell lines expressing the sfRNA of WNV. This cell line approach could be valuable to further our understanding of sfRNA function within flavivirus life cycle. Chapter 7 of this thesis is a short epilogue, in which the findings presented in the experimental chapters are summarized and discussed in the context of recently published findings and future research directions are suggested to improve the YFV-17D based vaccine platforms or to address the yet-to-be-solved questions about sfRNA production and function.

161 162 Samenvatting

163 Samenvatting

Alhoewel de mensheid in de laatste eeuw grote successen heeft geboekt in het bestrijden van parasieten, bacteriën en virussen, zijn infectieziekten nog steeds een belangrijke oorzaak van morbiditeit en mortaliteit bij mensen. In tegenstelling tot het reeds geëlimineerde pokkenvirus (variola) en het bijna uitgeroeide poliovirus, welke beide uitsluitend de mens als gastheer gebruik(t)en, is het vrijwel onmogelijk om de pathogene flavivirussen uit te roeien. Dit is vooral een gevolg van het feit dat deze virussen naast de mens een geleedpotige gastheer, zoals een teek of mug, gebruiken als een vector voor de verspreiding en als gastheer voor de vermeerdering van het virus. Verschillende van deze zogenaamde “arthropod-borne” (arbo) flavivirussen, zoals het West Nile virus (WNV) in Noord-Amerika, het dengue virus (DENV) in Zuidoost-Azië en het gele koortsvirus (YFV) in Afrika en Zuid-Amerika behoren tot de belangrijkste zogenaamde “emerging” en “re-emerging” virussen. Ieder jaar infecteren deze virussen grote aantallen mensen, hetgeen kan resulteren in een levensbedreigende encefalitis (WNV) dan wel zeer ernstige hemorragische koorts (DENV en YFV). Naast fundamenteel onderzoek aan de moleculaire biologie van de replicatie van flavivirussen worden in dit proefschrift, aan de hand van de vaccinstam van het gele koorts virus, ook alternatieve mogelijkheden gepresenteerd voor de vervaardiging en het gebruik van geattenueerde flavivirussen als recombinant vaccin. Hoofdstuk 1 geefteen overzicht van de biologie van de flavivirussen. De moleculaire biologie, alsmede de onderlinge verwantschap en de transmissieroutes van de verschillende flavivirussen worden in dit hoofdstuk beschreven. Deze uit fundamenteel- en veldonderzoek verkregen kennis is cruciaal voor het begrijpen van de complexe virus – gastheer interacties en dient als basis voor de ontwikkeling van strategieën om de invloed van flavivirus infecties op de maatschappij en economie terug te dringen. De overige hoofdstukken van dit proefschrift beschrijven de vervaardiging van zogenaamde “infectieuze cDNAs” van flavivirussen en het gebruik daarvan voor zowel de toepassing als alternatieve recombinante vaccins tegen flavivirussen, alsmede voor fundamenteel onderzoek in de moleculaire biologie van deze virussen. Bescherming tegen infectie van het eerder genoemde gele koorts virus is mogelijk door vaccinatie met de van het wildtype virus afgeleide, maar geattenueerde YFV-17D stam. Het YFV-17D vaccin is uitermate doeltreffend en veilig. Tientallen miljoenen mensen zijn inmiddels door vaccinatie met dit virus beschermd tegen gele koorts. In hoofdstuk 2 wordt beschreven hoe deze gunstige eigenschappen van het YFV-17D virus gebruikt kunnen worden om met moleculair biologische technieken een recombinant virus te construeren dat in principe niet alleen bescherming geeft tegen gele koorts, maar ook tegen een ander pathogeen, in dit geval het Lassa virus. De in dit hoofdstuk beschreven in vivo experimenten met de cavia als diermodel tonen aan dat deze benadering in principe succesvol kan zijn.

164 Samenvatting

Ondanks de beschikbaarheid van een goed vaccin is er de laatste decennia sprake van een hernieuwde verspreiding van het gele koorts virus en een toename van het aantal gele koorts- infecties van mensen in met name Afrika. Dit is onder andere het gevolg van het teruglopende aantal gevaccineerden in gebieden waar het virus endemisch is. In hoofdstuk 3 van dit proefschrift wordt een alternatief YFV-17D vaccin beschreven dat is gebaseerd op infectieus YFV-17D RNA dat zowel in celkweek als in muizen tot expressie komtna immunisatie met DNA in plaats van het virus. Alhoewel verder onderzoek zeker noodzakelijk is, blijkt uit de analyse van immunologische karakteristieken van de met het YFV DNA gevaccineerde muizen, dat deze strategie in principe werkt. Het gebruik van infectieus DNA als vaccin zou in de toekomst de productie van het YFV-17D vaccin kunnen vereenvoudigen, terwijl de betere thermostabiliteit van DNA, in vergelijking tot het huidige YFV-17D vaccin de distributie zal vergemakkelijken. Naast de voor mensen pathogene arbo-flavivirussen is er een groep minder goed bestudeerde flavivirussen waarvoor geen vector bekend is. Deze “no known vector” (NKV) flavivirussen zijn uitsluitend in zoogdieren aangetroffen. Ook mensen kunnen door sommige van deze NKV flavivirussen worden geïnfecteerd, maar infectie is tot op heden niet geassocieerd met ziekte. In hoofdstuk 4 van dit proefschrift wordt de constructie en karakterisatie van de eerste infectieuze cDNA kloon voor een NKV flavivirus beschreven. Deze Modoc virus kloon is vervolgens gebruikt om de structuur en functie van geconserveerde RNA elementen in het 3’ uiteinde van het virale genoom in de replicatie van het virus te bestuderen. Uit dit onderzoek blijkt dat het 3’ uiteinde van Modoc virus naast RNA sequenties welke qua structuur en functie geconserveerd zijn tussen de NKV- en arbo-flavivirussen, RNA elementen bevat welke uniek zijn voor de NKV flavivirussen en essentieel zijn voor virus replicatie. Tot voor kort werd algemeen aangenomen dat in tegenstelling tot diverse groepen andere RNA virussen, zoals de Nido- en Alphavirussen, flavivirussen geen subgenome RNAs zouden produceren. Uit recent onderzoek is echter gebleken dat naast het genoom RNA, alle flavivirussen een niet-coderend RNA molecuul met een lengte van 250 tot 500 nucleotiden produceren. Dit zogenaamde “small flavivirus” (sf) RNA heeft exact dezelfde nucleotidensequentie als het 3’ uiteinde van het virale genoom. Productie van dit sfRNA is echter niet het resultaat van virale transcriptie, maar het gevolg van de onvolledige afbraak van het virale genoom door het gastheer endo-ribonuclease XRN1. Een unieke tertiaire RNA structuur in het 3’ uiteinde van het genoom stopt de van 5’ naar 3’ verlopende hydrolyse van het virale RNA door XRN1, hetgeen resulteert in de accumulatie van sfRNA. In hoofdstuk 5 worden de resultaten van een uitgebreide mutagenese van deze XRN1-stoppende RNA pseudoknot structuur in YFV-17D beschreven. Uit de verkregen resultaten blijkt dat voor het succesvol stoppen van XRN1 en dus sfRNA productie, niet alleen de structuur van de RNA pseudoknot, maar ook specifieke nucleotiden op bepaalde posities in deze structuur

165 Samenvatting van belang zijn. De aangebrachte mutaties blijken echter geen significante invloed op de genoom RNA synthese en virusproductie te hebben. Uit eerder onderzoek is gebleken dat de productie van sfRNA bij WNV en YFV correleert met de pathogeniciteit van het virus. Het moleculair-biologische principe dat aan deze waarneming ten grondslag lag was echter onbekend. In hoofdstuk 6 wordt de constructie van cellijnen beschreven, die constitutief het sfRNA van YFV-17D ofWNV produceren. Uit het beschreven werk blijkt dat het mogelijk is om de cytopathogeniciteit van een sfRNA- negatieve mutant van YFV-17D te herstellen door trans-complementatie met het sfRNA dat in deze cellijnen wordt geproduceerd. Het maakt hierbij geen verschil of het aangeboden sfRNA van YFV of WNV afkomstig is. Blijkbaar is de functie van het sfRNA niet virusspecifiek en goed geconserveerd bij de arbo-flavivirussen. Het laatste hoofdstuk van dit proefschrift is een korte epiloog waarin de experimentele resultaten uit hoofdstuk 2 t/m 5 worden samengevat en bediscussieerd in de context van recent gepubliceerde data en waarin suggesties worden gedaan met betrekking tot toekomstig onderzoek.

166 CURRICULUM VITAE

167 Curriculum vitae

Xiaohong Jiang was born on April 10th. 1983 in Shanghai, China. In 2001, she finished secondary education in Shanghai High School. In the same year, she was enrolled into the School of Life Science in Fudan University, where she studied until 2005 and graduated with a B.Sc. diploma. She joined the Department of Medical Microbiology in Leiden University Medical Center (LUMC) in April 2006 and was trained as an intern for one year under the supervision of Dr. P.J. Bredenbeek, Dr. M. Kikkert and Prof. Dr. E. J. Snijder. She continued as a Ph.D. student (Onderzoeker in Opleiding) in the same department to carry out the research as described in this thesis, under the supervision of Dr. P.J. Bredenbeek and Prof. Dr. W. J. M. Spaan. Since July 2012, Xiaohong has been working as a research scientist in the Asian Veterinary Research&Development Center at Boehringer Ingelheim in Shanghai to continue her career as a virologist.

168