The many faces of the flavivirus NS1 protein offer a multitude of options for inhibitor design

Antiviral Research 130 (2016) 7e18

Contents lists available at ScienceDirect

Antiviral Research journal homepage: www.elsevier.com/locate/antiviral

Review

The many faces of the flavivirus NS1 protein offer a multitude of options for inhibitor design Daniel Watterson a, Naphak Modhiran a, Paul R. Young a, b, * a

Australian Infectious Diseases Research Centre, School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, QLD 4072, Australia b Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD 4072, Australia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 14 December 2015 Received in revised form 23 February 2016 Accepted 28 February 2016 Available online 2 March 2016

The flavivirus non-structural protein, NS1, is an unusual viral gene product. Despite the recent unveiling of its atomic structure (Akey et al., 2014), and a growing list of host molecules with which it has been found associated, the primary function of NS1 remains elusive. It assumes many diverse roles including direct participation in the flaviviral replication complex and virion maturation. In its secreted form it is a hexameric lipoparticle that is involved in systemic immune and endothelial cell modulation. In this review we highlight recent advances in elucidating the molecular mechanisms underpinning NS1 function and present the current state of play and some future prospects for NS1 targeted antiviral strategies. This article forms part of a symposium on flavivirus drug discovery in Antiviral Research. © 2016 Elsevier B.V. All rights reserved.

Keywords: Flavivirus Inhibitors Dengue Antiviral Complement

Contents 1. 2. 3. 4.

5.

6.

7.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Early interventions e taking aim at NS1 expression, processing and post-translational modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 A sugary target eGlycosylation is a key NS1 component and proven antiviral target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 A fatty target e NS1 membrane interactions and lipidic cargo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 4.1. Structural insights into lipid association . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 4.2. Dimer contacts within the hexamer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Multifunctional sNS1 in vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 5.1. Complement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 5.2. Immune activation and endothelial dysfunction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 5.3. NS1 as a viral toxin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Functional assays for the development of NS1 antivirals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 6.1. Liposome remodeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 6.2. Immune activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

1. Introduction

* Corresponding author. Australian Infectious Diseases Research Centre, School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, QLD 4072, Australia E-mail address: [email protected] (P.R. Young). http://dx.doi.org/10.1016/j.antiviral.2016.02.014 0166-3542/© 2016 Elsevier B.V. All rights reserved.

First described as a soluble complement fixing (SCF) antigen more than four decades ago (Brandt et al., 1970b; Russell et al., 1970; Smith et al., 1970), the unusual properties of the flavivirus NS1 protein captivated the fledgling flavivirus research field. Early

8

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

biophysical investigations teased apart the flaviviral antigens on the basis of their distinct biochemical and morphological characteristics, leading to the recognition of SCF as a non-structural antigen (Brandt et al., 1970a; Cardiff et al., 1971). These seminal studies were subsequently supported at the genetic level with the sequencing of the first flaviviral genome, in which the SCF was designated NS1(Rice et al., 1985). A potential driving role for SCF (NS1) in disease pathogenesis was postulated early on, in light of the observed association between high levels of complement consumption and the more severe form of dengue disease, dengue shock syndrome (DSS) (Bokisch et al., 1973, 1973). Subsequent investigation into the interplay between dengue (DENV) and complement yielded further evidence that immune complexes, whether virion or NS1 derived, potentiate disease severity (Ruangjirachuporn et al., 1979; Sobel et al., 1975; Theofilopoulos et al., 1976). Indeed, the extent and interactions between NS1 and the complement pathway have since been expanded considerably (covered below), however a clear paradigm of the role it plays in disease progression is yet to emerge. In parallel with the putative roles proposed on the basis of clinical investigations, insight into the role of NS1 in the flavivirus lifecycle at the molecular level continues to evolve. First observed at the protein level as a 46 kDa species (Smith and Wright, 1985), it is now known that NS1 comes in a variety of flavours, with distinct populations incorporating alternate post-translational modifications and quaternary structures that traffic to different locations in and from the infected cell. The much-anticipated recent elucidation of the crystal structure of NS1 has shed new light on these alternate NS1 species and opened new avenues for antiviral development (Akey et al., 2014; Edeling et al., 2014). In addition, our recent demonstration that the DENV NS1 engages TLR4, with the consequent induction of cytokine production and vascular permeability, provides the first mechanistic explanation to link NS1 with disease severity (Beatty et al., 2015; Modhiran et al., 2015). Armed with a solid framework for known structure-function relationships, researchers are now poised to probe deep into the underlying molecular mechanisms underpinning the diverse array of functions that have been proposed for NS1. In this review we will present the current state of play in NS1 research from the perspective of the newly available structural information together with the known functional attributes of NS1. We will cover NS1 interactions in the viral lifecycle, from protein translation and trafficking through to the role the secreted form plays in systemic disease. 2. Early interventions e taking aim at NS1 expression, processing and post-translational modification Early events in the flavivirus lifecycle, including viral entry, nucleocapsid uncoating and translation have been well documented (Pierson and Kielian, 2013). In brief, following nucleocapsid disassembly, the viral RNA is released into the cytoplasm and relocates to the ER (Fig. 2, step 1). Once engaged with the membranebound ribosomal machinery, the single open reading frame encoded by the positive sense flaviviral genome serves as a template for the first round of translation into the viral polyprotein which is cotranslationally cleaved by viral and host proteases. During translation, NS1 is translocated into the lumen of the ER via a signal peptide encoded by the C terminal residues within the upstream flavivirus envelope (E) gene (Falgout et al., 1989). Within the lumen, the N-terminus of NS1 is released after processing by resident host signal peptidase (Nowak et al., 1989). Cleavage of the NS1/NS2A junction also occurs within the ER lumen, however the identity of the protease responsible is not yet known. This cleavage motif is unique within the flaviviral genome, and when considering NS1 biosynthesis chronologically, it represents the first potential target

for the development of NS1 targeting antiviral therapies. The putative recognition motif, an octapeptide representing the last 8 amino acids in NS1 (L/M-V-X-S-X-V-X-A), is the minimal requirement for cleavage and is conserved across the flaviviral family (Hori and Lai, 1990), suggesting a common target for broad spectrum anti-flaviviral therapy. This processing also requires the presence of the N-terminal portion of NS2A, and it has been speculated that NS2A itself may mediate this cleavage event (Falgout et al., 1989). However, truncation and mutagenesis studies do no support this hypothesis, and the anonymous protease, likely of host origin, will remain off the antiviral target list until its identity is revealed. The NS1 protomer contains 12 cysteines that pair to form 6 disulfide bonds that mediate intra-domain stabilization. Reflecting the importance of these residues, all 12 cysteines are conserved across the flaviviurses that infect mammalian and avian species (Wallis et al., 2004). Interestingly, only 10 of these cysteines are conserved for NS1 proteins encoded by flaviviruses that only infect insects, (own observations). The inter-residue linkages were first assessed via mass spectrometry (Wallis et al., 2004), and have been further clarified in the resolved crystal structures for West Nile virus (WNV) and dengue virus (DENV) NS1. The critical nature of these bonds has also been demonstrated in mutagenesis studies (Pryor and Wright, 1993), indicating that an antiviral strategy targeting protein disulfide isomerase (PDI, Fig. 2, step 1) may prove effective in preventing correct NS1 folding and function. The development of non-toxic inhibitors to block the activity of protein disulfide isomerases has been challenging, however recent advances in reversible PDI inhibitors with good pharmacokinetic (PK) characteristics indicates that this approach is realizable (Kaplan et al., 2015), but has yet to be applied within the flavivirus context. While still within the ER, the NS1 monomer dimerises and acquires a partially hydrophobic nature. Atomic structures of the DENV and WNV NS1 dimer have revealed a hydrophobic domain comprised of the N-terminal region of NS1 (Fig. 2). The first ~20 amino acids of each monomer within the dimer form an interlocking beta-roll domain, which contributes substantially to the dimeric interface (Fig. 1A). Together with a hydrophobic protrusion located approximately 150 amino acids downstream, these domains are proposed to insert into the lumenal membrane and account for the close association between NS1 and lipids (Fig. 1A and Fig. 2, step 2). In addition to this hydrophobic interface we had earlier identified the incorporation of a glycosylphosphatidylinositol (GPI) moiety within a sub-population of NS1 (Jacobs et al., 2000). The importance of the GPI addition remains unclear, and has been the subject of subsequent investigation (Noisakran et al., 2007). We observed a reduction in virus replication within GPI deficient cell lines, providing evidence of a putative avenue for antiviral intervention. Although untested in the flaviviral context, disruption of GPI-protein clustering using GPI analogues has been shown effective in the reduction of prion (PrPsc) formation (Bate et al., 2010). In this example, treating with GPI analogues prevented sequestration of the PrPsc into cholesterol rich lipid microdomains (rafts) and subsequent transport to the cell surface, which may have parallels for NS1 membrane targeting. Further work will be required to assess the validity of this approach within the context of flavivirus infection. 3. A sugary target eGlycosylation is a key NS1 component and proven antiviral target Concurrent with protomer folding and proteolytic processing within the lumen, N-linked glycan addition within NS1 is mediated by the oligosaccharyl transferase complex (OTC, Fig. 2 step 1) (Winkler et al., 1988). An atypical property for a non-structural viral

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

9

Fig. 1. NS1 protein structure. A) Domain arrangement within the NS1 dimer. Domains of only one NS1 subunit are coloured for clarity. Indicated domains include the wing domain (yellow), the central b-ladder (red) and the hydrophobic b-roll and connecting domain (blue and orange). N-linked glycans that were resolved in the crystal structure are indicated. A 90
rotation reveals the proposed membrane topology, with the hydrophobic undercarriage inserting into the cellular membrane. Insets show observed flexibility within the exposed “greasy loop”. B) NS1 hexameric organization. Left, the NS1 hexamer is shown side on with one dimer domain coloured as per (A) and the other two dimers displayed as molecular surfaces (grey). Rotation through 90
shows the central cavity which is proposed to house the lipidic cargo. Insets highlight the different inter-dimer contacts observed between DENV and WNV NS1. No molecular contacts were resolved within the WNV hexamer (form 2) in comparison to DENV where N terminal residues may play a critical role in hexamer stabilization.

protein, the presence of glycans was one of the first features of interest identified within NS1, drawing the attention of early researchers (Westaway, 1975). In the wake of these original studies, there now exists a wealth of evidence that not only highlights glycosylation as a critical NS1 component but also identifies glycosylation as a validated NS1-directed antiviral target. Examining the multiple roles glycosylation plays within the NS1 context provides focus on the potential for antiviral intervention. Sequence analysis first revealed 2e3 putative N-linked glycosylation sites for the flavivirus NS1, depending on the virus analysed (Rice et al., 1985; Trent et al., 1987). It is likely these differences represent variations on an immune evasion strategy, however they may also play a role in receptor binding and cellular activation (discussed below). After the addition of high-mannose carbohydrate moieties within the OTC, NS1 separates into three

distinct populations; a significant fraction is incorporated in the viral replication complex and associated vesicle packets (VPs) (Fig. 2, step 3) (Khromykh et al., 1999; Lindenbach and Rice, 1997, 1999; Mackenzie et al., 1996), a second minor population is trafficked to the plasma membrane (mNS1, Fig. 2, step 5) (Schlesinger et al., 1990; Winkler et al., 1989) and the third is secreted into the extracellular milieu (sNS1, Fig. 2, step 6) (Crooks et al., 1990, 1994; Flamand et al., 1999). The secreted form is first trafficked via the Golgi where exposed carbohydrate moieties are trimmed and processed to more complex forms by resident glycotransferases and glycosidases (Fig. 2, step 4). Interestingly, there may be a requirement of high mannose processing for efficient secretion as little or no secreted NS1 is observed in infected insect cells, which lack the glycosylation machinery to process NS1 to the complex carbohydrate form (Brooks, 2006). Therefore blocking either initial

10

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

Fig. 2. NS1 within the infected cell. 1) During translation NS1 is translocated into the ER lumen where an unidentified protease cleaves at the NS1/2A junction. Protomer folding occurs concomitantly with N-linked glycan addition in the oligosaccharyl transferase complex (OTC) and disulfide bond formation mediated by protein disulfide isomerase (PDI). 2) NS1 monomer association into dimers results in the exposure of a hydrophobic undercarriage that associates with the lumenal membrane. 3) A dimeric NS1 species plays a critical role in the viral replication complex through interactions with NS4A and NS4B. NS1 has also been recently proposed to interact with prM/E, linking replicative complexes within the vesicle packets (VPs) together with virion formation. 4) NS1 glycan processing within the Golgi is mediated by glycosidases and transferases. 5) A smaller subset of NS1 is modified with a GPI anchor, a modification that is observed within the cell surface mNS1 population. 6) Within the Golgi, NS1 dimers associate on cholesterol lipid microdomains, subsequently pinching off to form a soluble hexameric lipoparticle. The ratio of lipid components (phospholipid, PL; Cholesterol, Chol; Di- and Tri-glycerides, DG and TG) to a single NS1 molecule are as indicated.

glycan attachment at the ER translocon or subsequent trimming and processing steps within the Golgi could suppress NS1 secretion and downstream activity. However, our group and others have successfully produced secreted recombinant NS1 through both baculovirus and transient plasmid based expression systems within a variety of insect cells (Gutsche et al., 2011; Modhiran et al., 2015; Muller et al., 2012). Thus it may be that, in the context of an infected cell, interactions with other viral proteins and the lumenal membrane diverts NS1

towards the replication complex while glycan processing acts to prevent these interactions tipping the equilibrium towards hexamer formation and secretion. Transient expression systems driving co-expression of NS1 with other non-structural proteins that have been shown to induce membrane proliferation independently may provide insight into these questions (Miller et al., 2007; Roosendaal et al., 2006). It is of note that all known flaviviruses, including insect only viruses, incorporate a glycosylation site at Asn 207, while there are

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

additional sites at Asn 130 for DENV1-4, yellow fever virus (YFV), Japanese encephalitis virus (JEV); Asn 130 and Asn 175 for WNV, Saint Louis encephalitis virus (SLEV) and Murray Valley encephalitis virus (MVEV); and Asn 85 and Asn 223 for the tick-borne viruses Tick-borne encephalitis virus (TBEV) and Louping ill virus (LIV). These differences correlate with processing status, as Asn207 remains unprocessed in mammalian cells (Post et al., 1991) and although solvent exposed in the resolved crystal structures (Fig. 2A), the carbohydrate moiety points downward towards the proposed position of the lumenal membrane (or inwards towards the interface of the hexamer) potentially excluding the activity of the glycoprocessing machinery. Indeed, glycosylation abolishing mutations at this site dramatically affect NS1 oligomer assembly and secretion (Crabtree et al., 2005; Pryor and Wright, 1994; Somnuke et al., 2011b), highlighting a critical structural role for this modification. Conversely, sites at Asn 130 and Asn 175 are processed to a complex carbohydrate form within mammalian cells in both natural infection and recombinant expression settings (Dalgarno et al., 1986). Such details help inform glycan targeting antiviral strategies, as compounds that target glycosylation directly, such as tunicamycin, will impact on all sites including Asn207, while blockade of glycosidase activity will impact processing of sites at Asn 130 and 175 specifically. The first examination of glycan directed antivirals against flaviviruses focused on the virion, where treatment of cells with inhibitors of alpha-glycosidases, Castanospermine (CST) and deoxynojirimycin (DNJ), was shown to decrease infectious virus production (Courageot et al., 2000). NS1 was soon to attract attention though, and in a 2002 study, Wu et al. demonstrated significant reduction of NS1 secretion in addition to an effect on virion production in the presence of the iminosugar derivative Nnonyl-deoxynojirimycin (NN-DNJ) (Wu et al., 2002). Furthermore, the researchers noted that treatment reduced NS1 association with calnexin, suggesting a direct effect on protein folding, which may have a secondary impact through diminishing NS1 levels within the replication complex in addition to inhibition of secretion. Notably, NN-DNJ proved effective in treating JEV in a murine model, although no direct effect on NS1 activity was determined in this system. A following study utilizing replicon based systems suggested the efficacy of CST could be rationalized by the effect on virion secretion and infectivity, however no attempt was made to quantify secreted NS1 levels in either in vitro or in vivo models (Whitby et al., 2005). In fact, the pro-drug form of CST (6-O-butanoyl Castanospermine or Celgosivir) has since been shown to directly reduce sNS1 levels in vivo (Schul et al., 2007) and subsequent work revealed accumulation of NS1 in the ER along with upregulation of the unfolded protein response (UPR) after Celgosivir treatment (Rathore et al., 2011). After establishment of the Celgosivir pharmacokinetic (PK) profile, dose and schedule efficacy within a lethal mouse model (Watanabe et al., 2012), Celgosivir was translated to clinical phase trials. However, no viral load reduction or fever ablation was observed in the phase 1b proof of concept trial (Low et al., 2014). Interestingly, time to clearance of serum NS1 was significantly reduced in the treated group up to day 7 post treatment (log-rank p ¼ 0.026). Given the direct role proposed for NS1 in severe disease pathogenesis it is tempting to speculate that Celgosivir may reduce the more severe manifestations of dengue, such as vascular permeability syndrome, however a larger trial will be required to assess this possibility (Halstead, 2014). In addition to Celgosivir, the flaviviral community continues to develop iminosugar derivatives with anti-flaviviral activity (Chang et al., 2011, 2009, 2013; Gu et al., 2007; Perry et al., 2013; Warfield et al., 2015; Yu et al., 2012). These compounds, typically in the form of a DNJ based head group with various N-alkyl side chains attached, have been developed in an effort to enhance glycosidase

11

inhibition, increase cellular uptake and improve PK profile. While these studies continue to overlook the potential NS1 directed activity of these compounds, a recent directed evolution study using the DNJ analogue UV-4B observed positive selection within the NS1 gene in response to treatment (Plummer et al., 2015). However, the generation of escape mutants also highlights the ability of flaviviruses to respond to glycan targeted antivirals, which may need to be considered for use in combination with other antiviral strategies. Nevertheless, it seems likely that many of these derivatives will be revisited in light of our recent observations that NS1 likely plays a leading role in dengue pathology through direct activation of the innate immune response and perturbation of vascular permeability. 4. A fatty target e NS1 membrane interactions and lipidic cargo Concomitant with glycan processing and entry into the secretory pathway, NS1 transitions from a membrane associated dimer into a hexameric lipo-particle (Fig. 2, step 6). The exact mechanics involved in the transition between dimer to hexamer are still unknown, however the recent NS1 crystal structures contain some tantalizing clues. An atomic structure for NS1 has been a longsought goal of the flaviviral research community, however attempts to crystalize NS1 were met with significant difficulty. Progress towards a structural model for NS1 was made with the release of two low-resolution, single-particle based reconstructions of hexameric NS1 (Gutsche et al., 2011; Muller et al., 2012). Both studies revealed an open barrel like structure with three asymmetrically aligned densities likely to represent the NS1 dimer (Fig. 1A). Analysis of the secreted particle using mass spectrometry and thin layer chromatography revealed the particle contains a significant lipid component, with triglycerides, mono- and diacylglycerols, cholesterol esters and phospholipids among the various lipid species identified (Fig. 2, step 6) (Gutsche et al., 2011). This lipid cargo is thought to fill the interior of the barrel like structure, where tight packing would allow the estimated ~70 individual lipid molecules to fill the interior volume of the hexamer. Similar lipidic profiles were obtained for NS1 recovered from DENV1 infected vero cells and recombinant DENV2 NS1 produced in Drosophila Schneider 2 cells (S2). Taken together with the apparent stoichiometric quantities of lipid species within the NS1 hexamer it appears the NS1 cargo manifest is tightly regulated, and possibly mediated by structural interactions. The discrete nature of the cargo also suggests avenues for antiviral intervention, as alteration in the lipids available to NS1 can be expected to alter both hexamer formation and downstream functions. It is probable that the lipid component presented the major obstacle to the generation of high-quality NS1 crystals for X-ray diffraction studies. In their recent publication of the atomic structure of NS1, Akey and colleagues were able to overcome these difficulties by obtaining lipid-free NS1 using detergent solubilization followed by reformation of soluble hexamers using sequential size exclusion chromatographic steps (Akey et al., 2014). Utilizing this method for recombinant NS1 derived from both WNV and DENV2 based constructs, crystal structures were resolved that revealed a striking similarity in fold between the dimeric form of the two viral proteins, a finding in marked contrast to their significant sequence divergence (ca. RMSD 0.845 Å, 54.7% sequence identity). Intriguingly, two forms of WNV NS1 were resolved, assuming either perfect hexameric symmetry with a central cavity open on both ends (closed form), or with C3 symmetry and splayed open to reveal a single, much wider opening at one end of the hexamer (open form). These forms may be analogous to the cryoEM (D3 symmetry) and negative stain reconstructions (C3

12

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

symmetry) previously reported (Gutsche et al., 2011; Muller et al., 2012). The two crystal conformations observed for WNV NS1 and the two distinct EM single-particle reconstructions, both suggest dynamic movement between the dimers, with the open form possibly representing a structural intermediate that is formed during the process of hexamer formation. The equimolar distribution of triglycerides to NS1 monomers within the secreted particle lead Gutsche et al. to suggest a model where NS1 dimers are able to specifically “pinch off” a subset of lipids from neutral lipid rich proto-lipid droplets (LDs) within the ER (Gutsche et al., 2011). Thus, 3 dimers may rearrange into an organized array and interact with specific lipid sub-populations before lifting up and closing to encase their lipid cargo in the form of the mature hexameric lipoparticle (Fig. 2, step 6). The elucidation of bound detergent moieties within the open crystal form of WNV NS1 at the inter-dimeric boundary also supports this model. This process would be dependent on both the immediate NS1 concentration and the local lipidic environment. Accordingly, both direct NS1 targeted and host lipid directed strategies are promising avenues for antiviral development. Lipid directed antiviral intervention has been a focus of ongoing flaviviral research, with clear established targets including the enveloped virion and the lipid rich microenvironment of the replication complex. Cholesterol has been identified as a central factor in flaviviral replication and modulation of cholesterol synthesis has been realized as an antiviral strategy in vitro (Mackenzie et al., 2007). Foreshadowing the intimate lipid association revealed from structural studies, the identification of a GPI anchor modification of a sub-population of NS1 suggested a tight relationship between NS1 and host lipids (Jacobs et al., 2000). Subsequent work using both immunofluorescence and membrane floatation approaches demonstrated that the GPI linked NS1 co-localizes with the established lipid raft markers CD55 and GM1 on the cell surface (Noisakran et al., 2008). Furthermore, after revealing the specific lipidic components of the NS1 lipo-particle, Gutsche et al. went on to substantiate the importance of these interactions in sNS1 formation through the use of lipidic pathway inhibitors. These experiments revealed a significant reduction in recombinant sNS1 secretion upon treatment with Niacin, an inhibitor of triglyceride synthesis acting through diacylglycerol acyltrasferase 2 (Kamanna and Kashyap, 2008), as well as Methyl-b-cyclodextrin, a cholesterol sequestering compound (Zidovetzki and Levitan, 2007). sNS1 production from infected cells was also impeded in the presence of D-Threo-1-phenyl-2-palmitoylamino-3-pyrrolidino-1-propanol (P4), an inhibitor of glucosylceramide synthase that abrogates glycosphingolipid incorporation into lipid rafts. These preliminary findings were encouraging and have formed the basis of a patent covering lipidic inhibitors targeting NS1 production (Ermonval et al.). Further work has confirmed the antiviral potential of cholesterol targeting compounds in treating flavivirus infection (Poh et al., 2012). However, as with the glycosidase inhibitors, research attention has yet to shift to anti-NS1 related activity. Instead, studies so far have focused on the role of lipid droplets (LDs) during flavivirus infection, and a range of reports have linked LDs and the host enzyme fatty acid synthetase (FASN) to flaviviral replication. These enquiries have identified a role for non-structural protein 3 (NS3) and capsid (C) in the modulation of host cell lipid metabolism through recruitment of FASN and remodeling of LDs to favour viral replication and particle formation. It seems a likely prospect that NS1 function, both within the replication complex and/or through the secretory pathway, will be perturbed in the presence of compounds targeting host cell lipid metabolism such as those used in these studies. However, careful experimental design will be critical

in teasing out the detail as isolating specific effects on any particular viral protein represents a major challenge in light of the interconnected nature of the host lipid metabolome. On the other hand, host-targeting inhibitors that interrupt multiple viral functions will pose a formidable barrier to selective resistance. So while promising advances have been made thus far, further clinical development together with progress in understanding the underlying mechanisms of lipid targeting anti-flaviviral approaches should be a priority. 4.1. Structural insights into lipid association Within the quaternary organization of the hexamer, the atomic details revealed by the crystal structures identified motifs within the NS1 dimer that are likely to play key structural roles in lipid association and hexamer formation. NS1 is arranged into three distinct domains; a hydrophobic b-roll (aa 1e29) which, together with a downstream loop, appears as a hydrophobic protrusion from the cross-like structure of the dimer which is formed by a central bladder domain and a globular a/b domain with a RIG-I like fold (Fig. 1A). Within the dimer, the two a/b domains project laterally from either sides of the elongated central b-ladder and have been dubbed the “wing” domain. This organization imparts an amphipathic like overall structure to the dimer, with the inner hydrophobic face comprised of the b-roll and associated loop structure proposed to intercalate with the lumenal facing leaflet of the ER membrane. These insights provide the blueprints for novel NS1 targeted antivirals that could be designed to bind to and subvert the lipid binding domains, or mimetic compounds based on the critical lipid associated motifs identified in the structure. Interestingly, NS1 in its lipid free state was observed to bind to and remodel cholesterol rich liposomes, in a process that might mimic the native insertion into the lumenal membrane. This property provides a potential avenue for investigating NS1-lipid interactions, and was explored by Akey et al. in the first structure-based investigation of NS1 function. Using site-directed mutagenesis the researchers probed these potential interactions through the insertion of charged amino acids in place of conserved hydrophobic residues within the exposed loop (Fig. 1A, insets). Remarkably, no differences in liposome remodeling relative to wild-type protein were observed. Conversely, a reduction in viral replication efficiency was observed when these same mutations were incorporated into an infectious clone, leading the researchers to suggest that this loop may be involved in other roles within the viral life cycle such as interaction with the hydrophobic viral cofactors of the replication complex. However, in a subsequent commentary (Akey et al., 2015) it was noted that these same mutations resulted in enhanced secretion of NS1, which could impact on replication efficiency through skewing the ratio of intracellular vs. extracellular NS1 levels. Of particular note, the enhanced secreted NS1 derived from these mutant constructs was dimeric in nature, demonstrating that hexamerization is not a strict requirement for secretion, at least in a recombinant insect cell system. It will therefore be important to assess the oligomeric nature as well as the levels of secreted NS1 in the presence of inhibitors targeting lipid association. This result also reiterates the critical nature of this exposed loop, which although conserved, can adopt different conformations as was observed between the two crystal forms of WNV NS1 (Fig. 1A, insets). These subtle differences in orientation may underlie the lipid loading of the NS1 hexamer, where the exposed hydrophobic residues potentiate incorporation of the lipidic cargo and repositioning of the loop facilitates the dimer to hexamer transitions. Although preliminary in nature, these findings suggest that this region is of particular interest for the design of inhibitors that could subvert

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

hexamer formation and lipid association. Missing from the crystal structure is the exposed loop between the b5 and b6 strands within the wing domain, which was unresolved in both DENV and WNV structures. This loop is a major antigenic site and is well conserved between the mammalian and avian flaviviruses. Recent work has suggested this region may play an important role in mediating NS1 interaction with the envelope protein, thus tethering the replication complex with virion formation (Scaturro et al., 2015). It will be of interest to determine whether this novel insight can be translated into alternative NS1 directed antiviral interventions.

4.2. Dimer contacts within the hexamer The closed hexameric crystallographic form of DENV and WNV NS1 appears to match well with the reported cryo-EM reconstruction of native NS1, suggesting it is an accurate representation of the quaternary organization within the mature hexameric lipoparticle secreted from infected cells. This is a surprising finding given the lack of lipid in the crystal structure form. Indeed, there is limited to no inter-dimer contact within the crystallographic hexamer and it is expected that the lipid cargo contributes substantially to the structural integrity of the hexamer form. Despite the very similar dimeric architecture between DENV and WNV NS1, close comparison of the hexameric structures reveals significant differences in the inter-dimer contacts (Fig. 1B, insets). In fact, there are no residues within bonding distance within the WNV hexamer (form 2 or closed) and only one residue (K11) within the resolved DENV hexameric structure that is within bond forming distance. However, the two residues at the tip of the DENV extended loop (F162 and G163) were not resolved and may contribute to interdimer contact, potentially through piepi interactions with the DENV specific W9 and F162. Although remarkable, these findings sit well with a previous study from Youn et al. that suggested a role for an N-terminal motif in NS1 secretion (Youn et al., 2010). Here the researchers demonstrated that the exchange of residues between WNV and DENV at this site (RQ10NK) resulted in an enhanced secretion within a WNV infectious clone, reminiscent of the native DENV phenotype. While additional structure-function studies are now required to expand on these limited interpretations, the strength of both structural and biological data supporting a key role for this motif suggests it could be another attractive target for the design of peptidomimetic compounds for disruption of NS1 secretion and hexamer formation.

5. Multifunctional sNS1 in vivo Perhaps the most notable aspect of NS1 is the high level of secreted protein found circulating in the blood of flavivirus infected individuals. It appears that the levels of secreted NS1 differs significantly between flaviviruses, although there is limited reliable in vivo data to allow accurate sNS1 quantification outside of DENV infection. The levels of sNS1 in DENV are substantial, with levels in excess of 10 mg/ml reported (Alcon et al., 2002). The high level of NS1 in the bloodstream, which is present early during infection, has made NS1 a primary biomarker in disease diagnosis. The timing and level of sNS1 within DENV has also lead many researchers to propose a direct role for sNS1 in disease pathology. A wide variety of extracellular functions have now been attributed to sNS1. In the following we will highlight some of the central paradigms that have been established with a focus on potential targets for antiviral therapies.

13

5.1. Complement The first potential functional role for sNS1 as a soluble complement fixing antigen or SCF, was identified even prior to its designation as NS1. Since these landmark studies, the complex nature and sheer magnitude and variety of interactions between NS1 and components of the complement pathway is slowly becoming apparent. Although far from complete, the current data suggest that NS1 is able to mediate complement activation through immune complex formation as well as a direct interaction with complement pathway participants. As the consumption of complement and elevated levels of C3a, C5a and sC5b-9 have been linked to disease severity in the case of DENV (Avirutnan et al., 2006; Bokisch et al., 1973; Russell et al., 1969), a mechanistic understanding of the sNS1 complement nexus should provide therapeutic opportunities to alleviate disease pathology. Following the observation that NS1 can drive complement activation in the presence of antiserum, a raft of papers reported on the role complement and immune complexes play in potentiating DENV disease severity. However, after this early focus, the global dengue research community turned its attention to the role of lymphocytes and inflammatory cytokine production (discussed below) and complement focused research essentially went into hibernation. It took the development of quantitative NS1 capture assays (Alcon et al., 2002; Young et al., 2000) and their use in defining a correlation between NS1 levels and disease severity (Avirutnan et al., 2006; Libraty et al., 2002) before the NS1 and complement story returned to researcher attention. Both cell surface mNS1 and secreted sNS1 have been observed to drive complement activation in the presence of anti-serum (Fig. 3, inset 2). It is difficult to compare the relative potency of mNS1 to sNS1 with respect to complement activation, however it is worth noting that sNS1 can activate the complement pathway in the absence of anti-NS1 antibody. Nevertheless, elevated complement consumption is observed in the presence of anti-NS1 antibodies or convalescent human serum. Activation in the presence of physiologically relevant levels of NS1 and anti-serum is observed to go to completion, with the production of the soluble terminal complement complex (SC5b-9) produced even in the absence of anti-serum (Avirutnan et al., 2006). The high levels of circulating NS1 at the same time as the anamnestic rise of anti-NS1 antibodies in secondary infection suggest that immune complexes are a critical mediator of complement activation. Such activation would be expected to play some, if not a key role in the enhanced disease pathology of secondary DENV infection. Conversely, examples of protective, complement fixing anti-NS1 antibodies have been reported (Schlesinger et al., 1993), and may confer protection via antibodydependent cell-mediated cytotoxicity (ADCC), although both Fcgamma receptor -dependent and -independent mechanisms have been observed (Chung et al., 2006b). The oligomeric nature of the hexameric NS1 lipoparticle may also play a potentiating role in the initiation of the activation cascade as the recruitment and activation of C1 has been found to be dependent on the formation of antigen bound IgG hexamers (Diebolder et al., 2014). Understanding these interactions will be key to vaccine design and any attempt to intervene in NS1 activity using therapeutic antibodies. In addition to immune complex driven complement activation, a direct interaction between NS1 and a variety of complement pathway members has been reported (Fig. 3, inset 1). Among a growing list of NS1 interacting partners within the complement pathway are the complement regulation protein factor H (fH), complement inhibitory factor clusterin, complement proteins C4, C4b, C1q and proC1s/C1s (Avirutnan et al., 2010, 2011; Chung et al., 2006a; Kurosu et al., 2007; Schlesinger, 2006; Silva et al., 2013).

14

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

Fig. 3. The many functions for extracellular sNS1. Many of these functions and/or associations are limited to select flaviviruses. 1) sNS1 has been shown to interact directly with various components of the complement pathway and can exhibit both stimulating and inhibitory regulation. 2) sNS1 and mNS1 antibody complexes have been shown to stimulate complement activation. 3) Auto-antibodies generated against DENV NS1 have been shown to bind directly to host proteins (HP) present on endothelial cells leading to activation, cellular dysfunction and leak. DENV anti-NS1 antibodies have also been shown to bind platelets, blocking activation. 4) DENV, WNV and YFV sNS1 have been shown to interact with the cell surface receptor TLR4 on the surface of CD14þ monocytes, inducing cellular activation and the release of inflammatory cytokines. 5) DENV sNS1 has been shown to activate endothelial cells directly leading to a loss of endothelial barrier integrity through a TLR4 dependent pathway. Inflammatory cytokines released from sNS1 stimulated monocytes may also lead to endothelial damage and leak.

These include interactions that could result in both inhibitory and activating roles for NS1 within the complement cascade, and further investigation of these interactions within the context of the recent structural and functional revelations are now required. It is important to point out that not all of these interactions have been confirmed across all Flaviviruses tested and so there is a degree of evolutionary adaption that may be unique to particular virus-host engagement. Detailed structural probing of the various complement interactions is yet to come, however previous work has demonstrated a requirement for glycosylation at N130 for efficient recognition of C1s and C4 by DENV NS1 (Somnuke et al., 2011a). It is not possible to confirm the specific requirement of a carbohydrate moiety with respect to these interactions as this site is a determinant of hexamer formation and secretion and may reduce binding indirectly by alteration of the quaternary structure of the NS1 lipoparticle. Also of note is the structural homology observed between the central bladder domain of NS1 and that of complement binding partners from other pathogens (Akey et al., 2015). A common feature of these binding partners is an anti-parallel b-sheet fold that is recognized by a conserved complement control protein (CCP)

domain within the complement partners. Interestingly the CCP domain is found in a number of complement proteins proposed to associate with NS1, including fH, C1s, C4. Thus the exposed position and structural homology of the central NS1 b-ladder domain makes a compelling candidate for complement protein binding. Mutagenesis and binding studies should quickly reveal the potential in vivo role for these interactions and identify regions within NS1 that mediate complement pathway modulation. However, while the identification of such sites will inform novel antiviral design, successful intervention will also require a comprehensive understanding of the nuanced relationship between NS1 and the complement system, which has the potential to both enhance and alleviate pathogenesis. 5.2. Immune activation and endothelial dysfunction Disruption of endothelial cell layer integrity is a fundamental component of flaviviral pathology in humans. In the case of DENV and YFV, the endothelial cell barrier that makes up the capillary wall is damaged, resulting in fluid loss and haemorrhage (Basu and Chaturvedi, 2008; Monath and Barrett, 2003). While in the

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

encephalitic flaviviruses, including JEV and WNV, the dysregulation of the blood brain barrier is a critical factor in the development of central nervous system (CNS) pathology (Neal, 2014). These effects are tightly associated with the levels of inflammatory factors produced during infection and NS1 has long been a primary suspect in immune activation and downstream endothelial damage and associated disease pathology. As discussed, NS1 and NS1 immune complexes are able to activate the complement pathway resulting in the production of the terminal complement factor SC5b-9 and the membrane attack complex C5b-9. Furthermore, this finding is mirrored in vivo where sC5b-9 levels are correlated with NS1 and are associated with disease severity. In addition to complement, DENV NS1 is also reported to induce auto-antibodies which are proposed to cross react with proteins present on platelets and endothelial cells (Fig. 3, inset 3). A wide range of host proteins have been identified as cross-reactive targets (Cheng et al., 2009; Chuang et al., 2011, 2014; Liu et al., 2011; Sun et al., 2007, 2015; Yin et al., 2013). Interestingly, although most NS1 directed antibodies have been shown to provide a level of passive protection in vivo, a number of anti-NS1 preparations have been shown to increase morbidity and induce apoptosis in endothelial cells in vitro (Lin et al., 2002). DENV NS1 MAb binding to endothelial cells can also lead to the activation of NF-kB resulting in the production of the inflammatory cytokines IL-6, IL-8 and MCP-1 (Lin et al., 2005). A key understanding of the epitopes within NS1 that can elicit auto-antibodies is critical for the production of NS1 based vaccines or NS1 targeting antibodies. Some efforts into identifying potential problematic epitopes has been established using homologous peptide sequence alignment of a number of host proteins that were shown to cross-react with NS1 sera including ATP synthase b chain, PDI, vimentin and heat shock protein 60 (Cheng et al., 2009). This study revealed a motif within aa311-330 in DENV NS1 that is homologous to a number of the identified host proteins, and a synthetic peptide corresponding to this sequence was observed to react with antibodies against some of the identified proteins. These findings can now be revisited in light of the published structure with the aim of developing rational vaccine approaches without the risk of auto-antibody induction. 5.3. NS1 as a viral toxin In addition to the role of NS1 immune complexes and autoantibodies, a direct role for NS1 in immune activation and endothelial cell dysfunction has now been established. In two complementary studies from our laboratory and that of Eva Harris, NS1 treatment alone was demonstrated to elicit the pathological hallmarks of dengue disease, endothelial leak and inflammatory cytokine activation (Beatty et al., 2015; Modhiran et al., 2015). Using an animal model of disease, the work of Beatty et al. observed increased disease symptoms with the co-administration of recombinant NS1 together with virus. Going further, they revealed that the intravenous delivery of NS1 to naive animals produced observable morbidity, and was accompanied by cytokine induction and endothelial barrier breakdown. Maximal effects were recorded 3 days post administration mirroring the lag in disease severity and viral replication kinetics that is observed in humans. Complementing these findings, work in our laboratory revealed the molecular mechanisms underpinning the activity of sNS1. Using knock out cell lines and reconstituted in vitro receptor systems we established that sNS1 is recognized by the pathogen-associated molecular pattern (PAMP) receptor TLR4. TLR4 is required for both monocyte activation (Fig. 3, panel 4) and direct NS1-endothelial interactions (Fig. 3, panel 5). This activity mimics that of the endotoxin LPS and when examined from this perspective there are striking similarities between the manifestations of severe dengue

15

disease and those of septic shock. Unlike sepsis however, the aforementioned lag time between disease onset and severe symptom development provides a window of opportunity for the application of NS1/TLR4 targeted therapeutics. Although principally observational, both studies provide a blue print for future NS1 targeted interventions. The work of Beatty et al. illustrated the potential for rational vaccine design, using NS1 antiserum and monoclonal antibodies to inhibit NS1 function in vitro and in vivo. Although there is precedent within the literature for NS1 use in vaccine design (Chung et al., 2006b; Schlesinger et al., 1993), these findings, when combined with the recent NS1 structural data, provide the framework for novel vaccine approaches which could potentially block NS1 function without the off-target effects observed in the case of NS1 auto-antibodies or the complications arising from virion specific antibody dependent enhancement (ADE) of infection. Likewise, targeting the TLR4 receptor complex is a promising therapeutic strategy. Our work validated this approach using both TLR4 blocking antibodies and antagonists to inhibit NS1 activity within in vitro experimental models of monocyte activation and endothelial leak. Encouragingly, we were able to translate these findings to an in vivo model, where a reduction in the levels of leak in dengue infected mice was observed following treatment with LPS-RS, a naturally occurring antagonist of TLR4 signaling. Due to the pressing need for anti-sepsis strategies, there exists a number of potential treatments that could be re-purposed for the dengue setting. These include compounds that have entered or completed phase III trials such as the Lipid A analogue E5564 and TAK-242, which block TLR4 association with TIRAP-1 and TRAM mediated activation (Sha et al., 2007; Shirey et al., 2013). Other candidates that have progressed to phase II trials include another LPS analogue, E5531 (Kobayashi et al., 1998), and AV411, a phosphodiesterase inhibitor (Fort et al., 2005). Beyond these established examples there are also a range of inhibitors from various classes in preclinical trials which may prove useful in dengue treatment (Akashi et al., 2000; Daubeuf et al., 2007; Dunn-Siegrist et al., 2007; Kuronuma et al., 2009; Mizuno et al., 2004; Numata et al., 2010, 2012, 2013; Spiller et al., 2008; Wang et al., 2003). Successful translation of these potential antivirals will require a robust development pathway, using in vitro assessment of potential NS1 activity to facilitate QSAR before moving to animal models of disease and eventually human trials. While some of the approaches outlined above have the potential to move directly to clinical trials, a range of new functional approaches are available to researchers in the search for novel NS1 targeting antivirals and will help establish the function of NS1 in the context of other flaviviral diseases. 6. Functional assays for the development of NS1 antivirals Assays to monitor NS1 function have been largely absent despite the potential NS1 holds as an antiviral target. This is not all that surprising given the ambiguity surrounding NS1 form and function, prior to recent developments. Until recently the main avenue for investigating NS1 function has been through mutagenesis within a replicon or infectious clone setting. While these techniques have yielded some success, they cannot be translated to medium or highthroughput screening approaches as they are inherently time consuming and so lack scalability. Given the new functional and structural data at hand we will highlight a number of new potential in vitro approaches for the development of novel NS1 targeting antivirals. 6.1. Liposome remodeling In their landmark study describing the atomic structure of NS1,

16

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

Akey et al. also demonstrated that lipid free NS1 is able to intercalate and remodel liposomes. This process may be homologous to the native protein insertion into the lumenal membrane. Alternatively, it could reflect the ability of sNS1 to bind to the plasma membrane. Although these details remain to be elucidated, agents that prevent lipid remodeling by NS1 could be anticipated to corrupt NS1 function within the cellular context. Lipid remodeling could be monitored in real-time either through dynamic light scattering or through leakage assays using dye-loaded liposomes. 6.2. Immune activation The recent identification of TLR4 as the cellular receptor responsible for NS1 mediated cell activation opens a number of potential assay options to investigate NS1 targeted compounds that could inhibit NS1 directed cellular activation. There is a wealth of established assay systems to assess endotoxin activity within cellular systems (Chow et al., 1999; Hawkins et al., 2004; Huang et al., 2009; Shuto et al., 2005) and many could be translated to investigate NS1 activity. In the most simple case, the production of inflammatory cytokines could be measured either at the protein or mRNA level. Alternative approaches utilizing a reporter system would allow higher throughput screening of potential antiviral candidates, an approach that has been used with some success in the case of TLR4 antagonists (Wang et al., 2013; Yang et al., 2005). 7. Conclusions The interplay between virus, individual viral products and host cell factors in the kinetics and severity of flavivirus disease progression is complex, and NS1 is only one of a number of risk factors determining disease outcome and severity. However, the case for a significant role for NS1 in disease development is building with discoveries that include direct engagement with complement components, the induction of anti-NS1 antibodies that cross react with host components and the direct activation of both monocyte and endothelial cell populations. These findings have been complemented by recent structural details revealed by crystallographic and electron microscopy approaches. Despite this progress, there are yet many avenues left unexplored. Much of the detail has come from studying the role of DENV NS1, with little information available in the context of other flaviviruses. The high structural homology shared between WNV and DENV suggests a shared function, although this remains to be examined in light of the new functional findings for DENV NS1. It also remains to be seen what function cellular activation serves within the virus life cycle. Understanding the sometimes subtle but always far-reaching interplay between flaviviruses and the immune system will be critical to the development of new antiviral strategies and to progress those outlined in this review. References Akashi, S., Ogata, H., Kirikae, F., Kirikae, T., Kawasaki, K., Nishijima, M., Shimazu, R., Nagai, Y., Fukudome, K., Kimoto, M., Miyake, K., 2000. Regulatory roles for CD14 and phosphatidylinositol in the signaling via toll-like receptor 4-MD-2. Biochem. Biophys. Res. Commun. 268, 172e177. Akey, D.L., Brown, W.C., Dutta, S., Konwerski, J., Jose, J., Jurkiw, T.J., DelProposto, J., Ogata, C.M., Skiniotis, G., Kuhn, R.J., Smith, J.L., 2014. Flavivirus NS1 structures reveal surfaces for associations with membranes and the immune system. Science 343, 881e885. Akey, D.L., Brown, W.C., Jose, J., Kuhn, R.J., Smith, J.L., 2015. Structure-guided insights on the role of NS1 in flavivirus infection. Bioessays 37, 489e494. Alcon, S., Talarmin, A., Debruyne, M., Falconar, A., Deubel, V., Flamand, M., 2002. Enzyme-linked immunosorbent assay specific to Dengue virus type 1 nonstructural protein NS1 reveals circulation of the antigen in the blood during the acute phase of disease in patients experiencing primary or secondary infections. J. Clin. Microbiol. 40, 376e381.

Avirutnan, P., Fuchs, A., Hauhart, R.E., Somnuke, P., Youn, S., Diamond, M.S., Atkinson, J.P., 2010. Antagonism of the complement component C4 by flavivirus nonstructural protein NS1. J. Exp. Med. 207, 793e806. Avirutnan, P., Hauhart, R.E., Somnuke, P., Blom, A.M., Diamond, M.S., Atkinson, J.P., 2011. Binding of flavivirus nonstructural protein NS1 to C4b binding protein modulates complement activation. J. Immunol. 187, 424e433. Avirutnan, P., Punyadee, N., Noisakran, S., Komoltri, C., Thiemmeca, S., Auethavornanan, K., Jairungsri, A., Kanlaya, R., Tangthawornchaikul, N., Puttikhunt, C., Pattanakitsakul, S.N., Yenchitsomanus, P.T., Mongkolsapaya, J., Kasinrerk, W., Sittisombut, N., Husmann, M., Blettner, M., Vasanawathana, S., Bhakdi, S., Malasit, P., 2006. Vascular leakage in severe dengue virus infections: a potential role for the nonstructural viral protein NS1 and complement. J. Infect. Dis. 193, 1078e1088. Basu, A., Chaturvedi, U.C., 2008. Vascular endothelium: the battlefield of dengue viruses. FEMS Immunol. Med. Microbiol. 53, 287e299. Bate, C., Tayebi, M., Williams, A., 2010. Glycosylphosphatidylinositol anchor analogues sequester cholesterol and reduce prion formation. J. Biol. Chem. 285, 22017e22026. Beatty, P.R., Puerta-Guardo, H., Killingbeck, S.S., Glasner, D.R., Hopkins, K., Harris, E., 2015. Dengue virus NS1 triggers endothelial permeability and vascular leak that is prevented by NS1 vaccination. Sci. Transl. Med. 7, 304ra141. Bokisch, V.A., Top Jr., F.H., Russell, P.K., Dixon, F.J., Muller-Eberhard, H.J., 1973. The potential pathogenic role of complement in dengue hemorrhagic shock syndrome. N. Engl. J. Med. 289, 996e1000. Brandt, W.E., Cardiff, R.D., Russell, P.K., 1970a. Dengue virions and antigens in brain and serum of infected mice. J. Virol. 6, 500e506. Brandt, W.E., Chiewslip, D., Harris, D.L., Russell, P.K., 1970b. Partial purification and characterization of a dengue virus soluble complement-fixing antigen. J. Immunol. 105, 1565e1568. Brooks, S.A., 2006. Protein glycosylation in diverse cell systems: implications for modification and analysis of recombinant proteins. Expert Rev. Proteomics 3, 345e359. Cardiff, R.D., Brandt, W.E., McCloud, T.G., Shapiro, D., Russell, P.K., 1971. Immunological and biophysical separation of Dengue-2 antigens. J. Virol. 7, 15e23. Chang, J., Schul, W., Yip, A., Xu, X., Guo, J.-T., Block, T.M., 2011. Competitive inhibitor of cellular alpha-glucosidases protects mice from lethal dengue virus infection. Antivir. Res. 92, 369e371. Chang, J., Wang, L., Ma, D., Qu, X., Guo, H., Xu, X., Mason, P.M., Bourne, N., Moriarty, R., Gu, B., Guo, J.-T., Block, T.M., 2009. Novel imino sugar derivatives demonstrate potent antiviral activity against flaviviruses. Antimicrob. Agents Chem. 53, 1501e1508. Chang, J., Warren, T.K., Zhao, X., Gill, T., Guo, F., Wang, L., Comunale, M.A., Du, Y., Alonzi, D.S., Yu, W., Ye, H., Liu, F., Guo, J.-T., Mehta, A., Cuconati, A., Butters, T.D., Bavari, S., Xu, X., Block, T.M., 2013. Small molecule inhibitors of ER a-glucosidases are active against multiple hemorrhagic fever viruses. Antivir. Res. 98, 432e440. Cheng, H.J., Lin, C.F., Lei, H.Y., Liu, H.S., Yeh, T.M., Luo, Y.H., Lin, Y.S., 2009. Proteomic analysis of endothelial cell autoantigens recognized by anti-dengue virus nonstructural protein 1 antibodies. Exp. Biol. Med. (Maywood) 234, 63e73. Chow, J.C., Young, D.W., Golenbock, D.T., Christ, W.J., Gusovsky, F., 1999. Toll-like receptor-4 mediates lipopolysaccharide-induced signal transduction. J. Biol. Chem. 274, 10689e10692. Chuang, Y.C., Lei, H.Y., Lin, Y.S., Liu, H.S., Wu, H.L., Yeh, T.M., 2011. Dengue virusinduced autoantibodies bind to plasminogen and enhance its activation. J. Immunol. 187, 6483e6490. Chuang, Y.C., Lin, Y.S., Liu, H.S., Yeh, T.M., 2014. Molecular mimicry between dengue virus and coagulation factors induces antibodies to inhibit thrombin activity and enhance fibrinolysis. J. Virol. 88, 13759e13768. Chung, K.M., Liszewski, M.K., Nybakken, G., Davis, A.E., Townsend, R.R., Fremont, D.H., Atkinson, J.P., Diamond, M.S., 2006a. West Nile virus nonstructural protein NS1 inhibits complement activation by binding the regulatory protein factor H. Proc. Natl. Acad. Sci. U. S. A. 103, 19111e19116. Chung, K.M., Nybakken, G.E., Thompson, B.S., Engle, M.J., Marri, A., Fremont, D.H., Diamond, M.S., 2006b. Antibodies against West Nile Virus nonstructural protein NS1 prevent lethal infection through Fc gamma receptor-dependent and -independent mechanisms. J. Virol. 80, 1340e1351. s, P., 2000. Courageot, M.-P., Frenkiel, M.-P., Duarte Dos Santos, C., Deubel, V., Despre a-Glucosidase inhibitors reduce dengue virus production by affecting the initial steps of virion morphogenesis in the endoplasmic reticulum. J. Virol. 74, 564e572. Crabtree, M.B., Kinney, R.M., Miller, B.R., 2005. Deglycosylation of the NS1 protein of dengue 2 virus, strain 16681: construction and characterization of mutant viruses. Arch. Virol. 150, 771e786. Crooks, A.J., Lee, J.M., Dowsett, A.B., Stephenson, J.R., 1990. Purification and analysis of infectious virions and native non-structural antigens from cells infected with tick-borne encephalitis virus. J. Chromatogr. 502, 59e68. Crooks, A.J., Lee, J.M., Easterbrook, L.M., Timofeev, A.V., Stephenson, J.R., 1994. The NS1 protein of tick-borne encephalitis virus forms multimeric species upon secretion from the host cell. J. Gen. Virol. 75 (Pt 12), 3453e3460. Dalgarno, L., Trent, D.W., Strauss, J.H., Rice, C.M., 1986. Partial nucleotide sequence of the Murray Valley encephalitis virus genome. Comparison of the encoded polypeptides with yellow fever virus structural and non-structural proteins. J. Mol. Biol. 187, 309e323. Daubeuf, B., Mathison, J., Spiller, S., Hugues, S., Herren, S., Ferlin, W., KoscoVilbois, M., Wagner, H., Kirschning, C.J., Ulevitch, R., Elson, G., 2007. TLR4/MD-2

D. Watterson et al. / Antiviral Research 130 (2016) 7e18 monoclonal antibody therapy affords protection in experimental models of septic shock. J. Immunol. 179, 6107e6114. Diebolder, C.A., Beurskens, F.J., de Jong, R.N., Koning, R.I., Strumane, K., Lindorfer, M.A., Voorhorst, M., Ugurlar, D., Rosati, S., Heck, A.J.R., van de Winkel, J.G.J., Wilson, I.A., Koster, A.J., Taylor, R.P., Saphire, E.O., Burton, D.R., Schuurman, J., Gros, P., Parren, P.W.H.I., 2014. Complement is activated by IgG hexamers assembled at the cell surface. Sci. (New York, N. Y. 343, 1260e1263. Dunn-Siegrist, I., Leger, O., Daubeuf, B., Poitevin, Y., Depis, F., Herren, S., KoscoVilbois, M., Dean, Y., Pugin, J., Elson, G., 2007. Pivotal involvement of Fcgamma receptor IIA in the neutralization of lipopolysaccharide signaling via a potent novel anti-TLR4 monoclonal antibody 15C1. J. Biol. Chem. 282, 34817e34827. Edeling, M.A., Diamond, M.S., Fremont, D.H., 2014. Structural basis of Flavivirus NS1 assembly and antibody recognition. Proc. Natl. Acad. Sci. U. S. A. 111, 4285e4290. Ermonval, M., Flamand, M., Gutsche, I., Kayal, S., Rey, F., Salmon, J., Use of a molecule able to interfere with the activity of non-structural 1 (NS1) protein produced by a Flavivirus, for preparing a medicament for inhibiting a Flavivirus infection in a susceptible host. WO2009106986eA2; US2009275499-A1; WO2009106986A3, 49. Falgout, B., Chanock, R., Lai, C.J., 1989. Proper processing of dengue virus nonstructural glycoprotein NS1 requires the N-terminal hydrophobic signal sequence and the downstream nonstructural protein NS2a. J. Virol. 63, 1852e1860. Flamand, M., Megret, F., Mathieu, M., Lepault, J., Rey, F.A., Deubel, V., 1999. Dengue virus type 1 nonstructural glycoprotein NS1 is secreted from mammalian cells as a soluble hexamer in a glycosylation-dependent fashion. J. Virol. 73, 6104e6110. Fort, M.M., Mozaffarian, A., Stover, A.G., Correia Jda, S., Johnson, D.A., Crane, R.T., Ulevitch, R.J., Persing, D.H., Bielefeldt-Ohmann, H., Probst, P., Jeffery, E., Fling, S.P., Hershberg, R.M., 2005. A synthetic TLR4 antagonist has antiinflammatory effects in two murine models of inflammatory bowel disease. J. Immunol. 174, 6416e6423. Gu, B., Mason, P., Wang, L., Norton, P., Bourne, N., Moriarty, R., Mehta, A., Despande, M., Shah, R., Block, T., 2007. Antiviral profiles of novel iminocyclitol compounds against bovine viral diarrhea virus, West Nile virus, dengue virus and hepatitis B virus. Antivir. Chem. Chemother. 18, 49e59. Gutsche, I., Coulibaly, F., Voss, J.E., Salmon, J., d'Alayer, J., Ermonval, M., Larquet, E., Charneau, P., Krey, T., Megret, F., Guittet, E., Rey, F.A., Flamand, M., 2011. Secreted dengue virus nonstructural protein NS1 is an atypical barrel-shaped highdensity lipoprotein. Proc. Natl. Acad. Sci. U. S. A. 108, 8003e8008. Halstead, S.B., 2014. Stumbles on the path to dengue control. Lancet Infect. Dis. 14, 661e662. Hawkins, L.D., Christ, W.J., Rossignol, D.P., 2004. Inhibition of endotoxin response by synthetic TLR4 antagonists. Curr. Top. Med. Chem. 4, 1147e1171. Hori, H., Lai, C.J., 1990. Cleavage of dengue virus NS1-NS2A requires an octapeptide sequence at the C terminus of NS1. J. Virol. 64, 4573e4577. Huang, L.-Y., DuMontelle, J.L., Zolodz, M., Deora, A., Mozier, N.M., Golding, B., 2009. Use of toll-like receptor assays to detect and identify microbial contaminants in biological products. J. Clin. Microbiol. 47, 3427e3434. Jacobs, M.G., Robinson, P.J., Bletchly, C., Mackenzie, J.M., Young, P.R., 2000. Dengue virus nonstructural protein 1 is expressed in a glycosyl-phosphatidylinositollinked form that is capable of signal transduction. FASEB J. 14, 1603e1610. Kamanna, V.S., Kashyap, M.L., 2008. Mechanism of action of niacin. Am. J. Cardiol. 101, 20Be26B. Kaplan, A., Gaschler, M.M., Dunn, D.E., Colligan, R., Brown, L.M., Palmer, A.G., Lo, D.C., Stockwell, B.R., 2015. Small molecule-induced oxidation of protein disulfide isomerase is neuroprotective. Proc. Natl. Acad. Sci. 112, E2245eE2252. Khromykh, A.A., Sedlak, P.L., Guyatt, K.J., Hall, R.A., Westaway, E.G., 1999. Efficient trans-complementation of the flavivirus kunjin NS5 protein but not of the NS1 protein requires its coexpression with other components of the viral replicase. J. Virol. 73, 10272e10280. Kobayashi, S., Kawata, T., Kimura, A., Miyamoto, K., Katayama, K., Yamatsu, I., Rossignol, D.P., Christ, W.J., Kishi, Y., 1998. Suppression of murine endotoxin response by E5531, a novel synthetic lipid A antagonist. Antimicrob. Agents Chemother. 42, 2824e2829. Kuronuma, K., Mitsuzawa, H., Takeda, K., Nishitani, C., Chan, E.D., Kuroki, Y., Nakamura, M., Voelker, D.R., 2009. Anionic pulmonary surfactant phospholipids inhibit inflammatory responses from alveolar macrophages and U937 cells by binding the lipopolysaccharide-interacting proteins CD14 and MD-2. J. Biol. Chem. 284, 25488e25500. Kurosu, T., Chaichana, P., Yamate, M., Anantapreecha, S., Ikuta, K., 2007. Secreted complement regulatory protein clusterin interacts with dengue virus nonstructural protein 1. Biochem. Biophys. Res. Commun. 362, 1051e1056. Libraty, D.H., Young, P.R., Pickering, D., Endy, T.P., Kalayanarooj, S., Green, S., Vaughn, D.W., Nisalak, A., Ennis, F.A., Rothman, A.L., 2002. High circulating levels of the dengue virus nonstructural protein NS1 early in dengue illness correlate with the development of dengue hemorrhagic fever. J. Infect. Dis. 186, 1165e1168. Lin, C.F., Chiu, S.C., Hsiao, Y.L., Wan, S.W., Lei, H.Y., Shiau, A.L., Liu, H.S., Yeh, T.M., Chen, S.H., Liu, C.C., Lin, Y.S., 2005. Expression of cytokine, chemokine, and adhesion molecules during endothelial cell activation induced by antibodies against dengue virus nonstructural protein 1. J. Immunol. 174, 395e403. Lin, C.F., Lei, H.Y., Shiau, A.L., Liu, H.S., Yeh, T.M., Chen, S.H., Liu, C.C., Chiu, S.C., Lin, Y.S., 2002. Endothelial cell apoptosis induced by antibodies against dengue virus nonstructural protein 1 via production of nitric oxide. J. Immunol. 169,

17

657e664. Lindenbach, B.D., Rice, C.M., 1997. trans-Complementation of yellow fever virus NS1 reveals a role in early RNA replication. J. Virol. 71, 9608e9617. Lindenbach, B.D., Rice, C.M., 1999. Genetic interaction of flavivirus nonstructural proteins NS1 and NS4A as a determinant of replicase function. J. Virol. 73, 4611e4621. Liu, I.J., Chiu, C.Y., Chen, Y.C., Wu, H.C., 2011. Molecular mimicry of human endothelial cell antigen by autoantibodies to nonstructural protein 1 of dengue virus. J. Biol. Chem. 286, 9726e9736. Low, J.G., Sung, C., Wijaya, L., Wei, Y., Rathore, A.P., Watanabe, S., Tan, B.H., Toh, L., Chua, L.T., Hou, Y., Chow, A., Howe, S., Chan, W.K., Tan, K.H., Chung, J.S., Cherng, B.P., Lye, D.C., Tambayah, P.A., Ng, L.C., Connolly, J., Hibberd, M.L., Leo, Y.S., Cheung, Y.B., Ooi, E.E., Vasudevan, S.G., 2014. Efficacy and safety of celgosivir in patients with dengue fever (CELADEN): a phase 1b, randomised, double-blind, placebo-controlled, proof-of-concept trial. Lancet Infect. Dis. 14, 706e715. Mackenzie, J.M., Jones, M.K., Young, P.R., 1996. Immunolocalization of the dengue virus nonstructural glycoprotein NS1 suggests a role in viral RNA replication. Virology 220, 232e240. Mackenzie, J.M., Khromykh, A.A., Parton, R.G., 2007. Cholesterol manipulation by west nile virus perturbs the cellular immune response. Cell host microbe 2, 229e239. Miller, S., Kastner, S., Krijnse-Locker, J., Buhler, S., Bartenschlager, R., 2007. The nonstructural protein 4A of dengue virus is an integral membrane protein inducing membrane alterations in a 2K-regulated manner. J. Biol. Chem. 282, 8873e8882. Mizuno, T., Kurotani, T., Komatsu, Y., Kawanokuchi, J., Kato, H., Mitsuma, N., Suzumura, A., 2004. Neuroprotective role of phosphodiesterase inhibitor ibudilast on neuronal cell death induced by activated microglia. Neuropharmacology 46, 404e411. Modhiran, N., Watterson, D., Muller, D.A., Panetta, A.K., Sester, D.P., Liu, L., Hume, D.A., Stacey, K.J., Young, P.R., 2015. Dengue virus NS1 protein activates cells via Toll-like receptor 4 and disrupts endothelial cell monolayer integrity. Sci. Transl. Med. 7, 304ra142. Monath, T.P., Barrett, A.D., 2003. Pathogenesis and pathophysiology of yellow fever. Adv. Virus Res. 60, 343e395. Muller, D.A., Landsberg, M.J., Bletchly, C., Rothnagel, R., Waddington, L., Hankamer, B., Young, P.R., 2012. Structure of the dengue virus glycoprotein nonstructural protein 1 by electron microscopy and single-particle analysis. J. Gen. Virol. 93, 771e779. Neal, J.W., 2014. Flaviviruses are neurotropic, but how do they invade the CNS? J. Infect. 69, 203e215. Noisakran, S., Dechtawewat, T., Avirutnan, P., Kinoshita, T., Siripanyaphinyo, U., Puttikhunt, C., Kasinrerk, W., Malasit, P., Sittisombut, N., 2008. Association of dengue virus NS1 protein with lipid rafts. J. Gen. Virol. 89, 2492e2500. Noisakran, S., Dechtawewat, T., Rinkaewkan, P., Puttikhunt, C., Kanjanahaluethai, A., Kasinrerk, W., Sittisombut, N., Malasit, P., 2007. Characterization of dengue virus NS1 stably expressed in 293T cell lines. J. Virol. Methods 142, 67e80. Nowak, T., Farber, P.M., Wengler, G., Wengler, G., 1989. Analyses of the terminal sequences of West Nile virus structural proteins and of the in vitro translation of these proteins allow the proposal of a complete scheme of the proteolytic cleavages involved in their synthesis. Virology 169, 365e376. Numata, M., Chu, H.W., Dakhama, A., Voelker, D.R., 2010. Pulmonary surfactant phosphatidylglycerol inhibits respiratory syncytial virus-induced inflammation and infection. Proc. Natl. Acad. Sci. U. S. A. 107, 320e325. Numata, M., Kandasamy, P., Nagashima, Y., Posey, J., Hartshorn, K., Woodland, D., Voelker, D.R., 2012. Phosphatidylglycerol suppresses influenza A virus infection. Am. J. Respir. Cell Mol. Biol. 46, 479e487. Numata, M., Nagashima, Y., Moore, M.L., Berry, K.Z., Chan, M., Kandasamy, P., Peebles Jr., R.S., Murphy, R.C., Voelker, D.R., 2013. Phosphatidylglycerol provides short-term prophylaxis against respiratory syncytial virus infection. J. Lipid Res. 54, 2133e2143. Perry, S.T., Buck, M.D., Plummer, E.M., Penmasta, R.A., Batra, H., Stavale, E.J., Warfield, K.L., Dwek, R.A., Butters, T.D., Alonzi, D.S., Lada, S.M., King, K., Klose, B., Ramstedt, U., Shresta, S., 2013. An iminosugar with potent inhibition of dengue virus infection in vivo. Antivir. Res. 98, 35e43. Pierson, T.C., Kielian, M., 2013. Flaviviruses: braking the entering. Curr. Opin. Virol. 3, 3e12. Plummer, E., Buck, M.D., Sanchez, M., Greenbaum, J.A., Turner, J., Grewal, R., Klose, B., Sampath, A., Warfield, K.L., Peters, B., Ramstedt, U., Shresta, S., 2015. Dengue virus evolution under a host-targeted antiviral. J. Virol. 89, 5592e5601. Poh, M.K., Shui, G., Xie, X., Shi, P.Y., Wenk, M.R., Gu, F., 2012. U18666A, an intracellular cholesterol transport inhibitor, inhibits dengue virus entry and replication. Antivir. Res. 93, 191e198. Post, P.R., Carvalho, R., Galler, R., 1991. Glycosylation and secretion of yellow fever virus nonstructural protein NS1. Virus Res. 18, 291e302. Pryor, M.J., Wright, P.J., 1993. The effects of site-directed mutagenesis on the dimerization and secretion of the NS1 protein specified by dengue virus. Virology 194, 769e780. Pryor, M.J., Wright, P.J., 1994. Glycosylation mutants of dengue virus NS1 protein. J. Gen. Virol. 75 (Pt 5), 1183e1187. Rathore, A.P., Paradkar, P.N., Watanabe, S., Tan, K.H., Sung, C., Connolly, J.E., Low, J., Ooi, E.E., Vasudevan, S.G., 2011. Celgosivir treatment misfolds dengue virus NS1 protein, induces cellular pro-survival genes and protects against lethal challenge mouse model. Antivir. Res. 92, 453e460. Rice, C.M., Lenches, E.M., Eddy, S.R., Shin, S.J., Sheets, R.L., Strauss, J.H., 1985.

18

D. Watterson et al. / Antiviral Research 130 (2016) 7e18

Nucleotide sequence of yellow fever virus: implications for flavivirus gene expression and evolution. Science 229, 726e733. Roosendaal, J., Westaway, E.G., Khromykh, A., Mackenzie, J.M., 2006. Regulated cleavages at the West Nile virus NS4A-2K-NS4B junctions play a major role in rearranging cytoplasmic membranes and Golgi trafficking of the NS4A protein. J. Virol. 80, 4623e4632. Ruangjirachuporn, W., Boonpucknavig, S., Nimmanitya, S., 1979. Circulating immune complexes in serum from patients with dengue haemorrhagic fever. Clin. Exp. Immunol. 36, 46e53. Russell, P.K., Chiewsilp, D., Brandt, W.E., 1970. Immunoprecipitation analysis of soluble complement-fixing antigens of dengue viruses. J. Immunol. 105, 838e845. Russell, P.K., Intavivat, A., Kanchanapilant, S., 1969. Anti-dengue immunoglobulins and serum beta 1 c-a globulin levels in dengue shock syndrome. J. Immunol. 102, 412e420. Scaturro, P., Cortese, M., Chatel-Chaix, L., Fischl, W., Bartenschlager, R., 2015. Dengue virus non-structural protein 1 modulates infectious particle production via interaction with the structural proteins. PLoS Pathog. 11, e1005277. Schlesinger, J.J., 2006. Flavivirus nonstructural protein NS1: complementary surprises. Proc. Natl. Acad. Sci. U. S. A. 103, 18879e18880. Schlesinger, J.J., Brandriss, M.W., Putnak, J.R., Walsh, E.E., 1990. Cell surface expression of yellow fever virus non-structural glycoprotein NS1: consequences of interaction with antibody. J. Gen. Virol. 71 (Pt 3), 593e599. Schlesinger, J.J., Foltzer, M., Chapman, S., 1993. The Fc portion of antibody to yellow fever virus NS1 is a determinant of protection against YF encephalitis in mice. Virology 192, 132e141. Schul, W., Liu, W., Xu, H.Y., Flamand, M., Vasudevan, S.G., 2007. A dengue fever viremia model in mice shows reduction in viral replication and suppression of the inflammatory response after treatment with antiviral drugs. J. Infect. Dis. 195, 665e674. Sha, T., Sunamoto, M., Kitazaki, T., Sato, J., Ii, M., Iizawa, Y., 2007. Therapeutic effects of TAK-242, a novel selective Toll-like receptor 4 signal transduction inhibitor, in mouse endotoxin shock model. Eur. J. Pharmacol. 571, 231e239. Shirey, K.A., Lai, W., Scott, A.J., Lipsky, M., Mistry, P., Pletneva, L.M., Karp, C.L., McAlees, J., Gioannini, T.L., Weiss, J., Chen, W.H., Ernst, R.K., Rossignol, D.P., Gusovsky, F., Blanco, J.C., Vogel, S.N., 2013. The TLR4 antagonist Eritoran protects mice from lethal influenza infection. Nature 497, 498e502. Shuto, T., Kato, K., Mori, Y., Viriyakosol, S., Oba, M., Furuta, T., Okiyoneda, T., Arima, H., Suico, M.A., Kai, H., 2005. Membrane-anchored CD14 is required for LPS-induced TLR4 endocytosis in TLR4/MD-2/CD14 overexpressing CHO cells. Biochem. Biophys. Res. Commun. 338, 1402e1409. Silva, E.M., Conde, J.N., Allonso, D., Nogueira, M.L., Mohana-Borges, R., 2013. Mapping the interactions of dengue virus NS1 protein with human liver proteins using a yeast two-hybrid system: identification of C1q as an interacting partner. PLoS One 8, e57514. Smith, G.W., Wright, P.J., 1985. Synthesis of proteins and glycoproteins in dengue type 2 virus-infected vero and Aedes albopictus cells. J. Gen. Virol. 66 (Pt 3), 559e571. Smith, T.J., Brandt, W.E., Swanson, J.L., McCown, J.M., Buescher, E.L., 1970. Physical and biological properties of dengue-2 virus and associated antigens. J. Virol. 5, 524e532. Sobel, A.T., Bokisch, V.A., Müller-Eberhard, H.J., 1975. C1q deviation test for the detection of immune complexes, aggregates of IgG, and bacterial products in human serum. J. Exp. Med. 142, 139e150. Somnuke, P., Hauhart, R.E., Atkinson, J.P., Diamond, M.S., Avirutnan, P., 2011a. Nlinked glycosylation of dengue virus NS1 protein modulates secretion, cellsurface expression, hexamer stability, and interactions with human complement. Virology 413, 253e264. Somnuke, P., Hauhart, R.E., Atkinson, J.P., Diamond, M.S., Avirutnan, P., 2011b. Nlinked glycosylation of Dengue virus NS1 protein modulates secretion, cellsurface expression, hexamer stability, and interactions with human complement. Virology 413, 253e264. Spiller, S., Elson, G., Ferstl, R., Dreher, S., Mueller, T., Freudenberg, M., Daubeuf, B., Wagner, H., Kirschning, C.J., 2008. TLR4-induced IFN-gamma production increases TLR2 sensitivity and drives Gram-negative sepsis in mice. J. Exp. Med. 205, 1747e1754.

Sun, D.S., Chang, Y.C., Lien, T.S., King, C.C., Shih, Y.L., Huang, H.S., Wang, T.Y., Li, C.R., Lee, C.C., Hsu, P.N., Chang, H.H., 2015. Endothelial cell sensitization by death receptor fractions of an anti-dengue nonstructural protein 1 antibody induced plasma leakage, coagulopathy, and mortality in mice. J. Immunol. 195, 2743e2753. Sun, D.S., King, C.C., Huang, H.S., Shih, Y.L., Lee, C.C., Tsai, W.J., Yu, C.C., Chang, H.H., 2007. Antiplatelet autoantibodies elicited by dengue virus non-structural protein 1 cause thrombocytopenia and mortality in mice. J. Thromb. Haemost. 5, 2291e2299. Theofilopoulos, A.N., Wilson, C.B., Dixon, F.J., 1976. The Raji cell radioimmune assay for detecting immune complexes in human sera. J. Clin. Invest 57, 169e182. Trent, D.W., Kinney, R.M., Johnson, B.J., Vorndam, A.V., Grant, J.A., Deubel, V., Rice, C.M., Hahn, C., 1987. Partial nucleotide sequence of St. Louis encephalitis virus RNA: structural proteins, NS1, ns2a, and ns2b. Virology 156, 293e304. Wallis, T.P., Huang, C.Y., Nimkar, S.B., Young, P.R., Gorman, J.J., 2004. Determination of the disulfide bond arrangement of dengue virus NS1 protein. J. Biol. Chem. 279, 20729e20741. Wang, R., Stephens, J., Lacy, M.J., 2003. Characterization of monoclonal antibody HTA125 with specificity for human TLR4. Hybrid. Hybridomics 22, 357e365. Wang, X., Smith, C., Yin, H., 2013. Targeting toll-like receptors with small molecule agents. Chem. Soc. Rev. 42, 4859e4866. Warfield, K.L., Plummer, E., Alonzi, D.S., Wolfe, G.W., Sampath, A., Nguyen, T., Butters, T.D., Enterlein, S.G., Stavale, E.J., Shresta, S., Ramstedt, U., 2015. A novel iminosugar UV-12 with activity against the diverse viruses influenza and dengue (novel iminosugar antiviral for influenza and dengue). Viruses 7, 2404e2427. Watanabe, S., Rathore, A.P., Sung, C., Lu, F., Khoo, Y.M., Connolly, J., Low, J., Ooi, E.E., Lee, H.S., Vasudevan, S.G., 2012. Dose- and schedule-dependent protective efficacy of celgosivir in a lethal mouse model for dengue virus infection informs dosing regimen for a proof of concept clinical trial. Antivir. Res. 96, 32e35. Westaway, E.G., 1975. The proteins of Murray Valley encephalitis virus. J. Gen. Virol. 27, 293e292. Whitby, K., Pierson, T.C., Geiss, B., Lane, K., Engle, M., Zhou, Y., Doms, R.W., Diamond, M.S., 2005. Castanospermine, a potent inhibitor of dengue virus infection in vitro and. Vivo. J. Virol. 79, 8698e8706. Winkler, G., Maxwell, S.E., Ruemmler, C., Stollar, V., 1989. Newly synthesized dengue-2 virus nonstructural protein NS1 is a soluble protein but becomes partially hydrophobic and membrane-associated after dimerization. Virology 171, 302e305. Winkler, G., Randolph, V.B., Cleaves, G.R., Ryan, T.E., Stollar, V., 1988. Evidence that the mature form of the flavivirus nonstructural protein NS1 is a dimer. Virology 162, 187e196. Wu, S.-F., Lee, C.-J., Liao, C.-L., Dwek, R.A., Zitzmann, N., Lin, Y.-L., 2002. Antiviral effects of an iminosugar derivative on flavivirus infections. J. Virol. 76, 3596e3604. Yang, Q.W., Mou, L., Lv, F.L., Zhu, P.F., Wang, Z.G., Jiang, J.X., Wang, J.Z., 2005. Novel TLR4-antagonizing peptides inhibit LPS-induced release of inflammatory mediators by monocytes. Biochem. Biophys. Res. Commun. 329, 846e854. Yin, Y., Jiang, L., Fang, D., Jiang, L., Zhou, J., 2013. Differentially expressed genes of human microvascular endothelial cells in response to anti-dengue virus NS1 antibodies by suppression subtractive hybridization. Viral Immunol. 26, 185e191. Youn, S., Cho, H., Fremont, D.H., Diamond, M.S., 2010. A short N-terminal peptide motif on flavivirus nonstructural protein NS1 modulates cellular targeting and immune recognition. J. Virol. 84, 9516e9532. Young, P.R., Hilditch, P.A., Bletchly, C., Halloran, W., 2000. An antigen capture enzyme-linked immunosorbent assay reveals high levels of the dengue virus protein NS1 in the sera of infected patients. J. Clin. Microbiol. 38, 1053e1057. Yu, W., Gill, T., Wang, L., Du, Y., Ye, H., Qu, X., Guo, J.-T., Cuconati, A., Zhao, K., Block, T.M., Xu, X., Chang, J., 2012. Design, synthesis, and biological evaluation of n-alkylated deoxynojirimycin (DNJ) derivatives for the treatment of dengue virus infection. J. Med. Chem. 55, 6061e6075. Zidovetzki, R., Levitan, I., 2007. Use of cyclodextrins to manipulate plasma membrane cholesterol content: evidence, misconceptions and control strategies. Biochim. Biophys. Acta 1768, 1311e1324.