Epitope resurfacing on dengue virus-like particle vaccine preparation to induce broad neutralizing antibody

Dengue virus (DENV), a member of the family Flaviviridae, is a mosquito-borne pathogen with four distinct serotypes, DENV-1 to DENV-4 (Lindenbach and Rice, 2001). It has been estimated that DENV infects about 390 million individuals globally each year, resulting in 96 million clinically apparent infections ranging from mild fever to the life-threatening dengue hemorrhagic fever (DHF) or dengue shock syndrome (DSS) (Bhatt et al., 2013). Although the chimeric yellow fever 17D-derived, tetravalent dengue vaccine (CYD-TDV) has recently been approved by the governments of a few DENV-circulating countries, the unexpected low vaccine efficacy in dengue-naïve children or children younger than 6 years old has limited the use of this vaccine in the 9 – 45 age group living in endemic countries (Guy and Jackson, 2016). Developing a new strategy by either improving the efficacy of CYD-TDV (WHO, 2017; Aguiar et al., 2016; Halstead, 2017) or the second-generation vaccine candidates (Kirkpatrick et al., 2016) is essential to broaden the coverage for all vulnerable populations.

The envelope (E) protein of DENV on the surface of viral particles is the major target of neutralizing antibodies (NtAbs) (Roehrig, 2003; Pierson et al., 2008). The ectodomain of the E protein contains three distinct domains, EDI, EDII, and EDIII, which are connected by flexible hinges to allow rearrangement of domains during virus assembly, maturation and infection (Modis et al., 2003). However, the immune response from humans who have recovered from primary DENV infections is dominated by cross-reactive (CR), non-NtAbs which recognize mainly pre-membrane (prM) or fusion loop of the E (FLE) protein (Dejnirattisai et al., 2010). During virus replication, newly synthesized DENVs are assembled as immature particles in the endoplasmic reticulum at neutral pH, followed by translocation through the trans-Golgi network and low-pH secretory vesicles (Chambers et al., 1990). The pr portion of the prM protein, positioned to cover the fusion loop (FL) peptide at the distal end of each E protein, prevents premature fusion during the virion maturation process along the egress pathway from the infected cells (Li et al., 2008; Kostyuchenko et al., 2013). For DENV to become fully infectious, the pr molecule has to be removed by a cellular furin protease during the egress process (Rodenhuis-Zybert et al., 2010). However, furin cleavage in the low-pH secretory vesicles is thought to be inefficient; hence, DENV particles released from infected cells are heterogeneous populations with different degrees of cleavage and release of the pr portion of prM to form the mature membrane (M) protein (Pierson and Diamond, 2012). Immature DENV (imDENV) particles consist of uncleaved prM protein and partially immature particles (piDENV) containing both prM and M proteins, while fully mature DENV (mDENV) contains only M protein (Pierson and Diamond, 2012; Junjhon et al., 2010). Functional analyses have revealed that a completely immature flavivirus lacks the ability to infect cells unless in the presence of anti-prM antibodies through a mechanism called antibody-dependent enhancement (ADE) of infection (Dejnirattisai et al., 2010; Rodenhuis-Zybert et al., 2010; da Silva Voorham et al., 2012; Rodenhuis-Zybert et al., 2011). ADE plays an important role in dengue pathogenesis, and is potentially modulated by the antibody concentration and the degree of virion maturity (Nelson et al., 2008; Rodenhuis-Zybert et al., 2015).

Virus-like particles (VLPs) containing flavivirus prM/E proteins have been demonstrated to be a potential vaccine candidate, since their ordered E structures are similar to those on the virion surface and also undergo low-pH-induced rearrangements and membrane fusion similar to viral particles (Allison et al., 1995). Also, VLP vaccines present several advantages since they are highly immunogenic, non-infectious, and accessible to quality control as well as increased production capacity. However, we theorized that the process of VLP maturation might be similar to that of dengue viral particles, with heterogeneous pr-containing structures. In this study, we engineered and characterized the structure of mature DENV-2 VLPs through cryo-EM. The immunological properties of mature VLPs were further compared with its immature counterparts.

As we theorized, the process of VLP maturation are similar to that of dengue viral particles so that the wild-type particles (wtD2VLP) produced from the cells are partially immature containing both prM and M proteins (Figure 1—figure supplement 1). To overcome this problem, we engineered the DENV-2 VLPs as completely mature particles by manipulating the furin cleavage site at the junction of pr and M in a DENV-2 VLP-expressing plasmid (Chang et al., 2003). For comparison, we also generated immature VLP (imD2VLPs) by mutating the minimal furin cleavage motif REKR to REST (Li et al., 2008), which resulted in completely uncleaved prM as detected by western blot and ELISA (Figure 1). Generation of mature DENV-2 VLP (mD2VLP) was much more challenging. First, we performed multiple sequence alignments of the pr/M junction from several flaviviruses, including the distance-related cell fusion agent virus (CFAV), and analyzed their furin cleavage potential using the algorithm PiTou 2.0²⁴. DENV-3 pr/M junction had the lowest predicted Pitou score of 6.87, while WNV had the highest score of 15.4 (Figure 1A). Secondly, we determined the cleavage efficiency of D2VLP by replacing P1-8 at the pr/M junction site of DENV-2 with sequences from different flaviviruses. Although the Pitou prediction of the CFAV pr/M junction did not result in a high score for cleavage, replacement with P1-8 of CFAV resulted in the most efficient furin cleavage for D2VLP, with only 31% of prM remaining uncleaved as compared to the 100% uncleaved imD2VLP (Figure 1A). An additional mutation in the P3 residue from E to S of the CFAV P1-8 construct reduced the prM signaling to 13.7% of that of imD2VLP. We thus named this construct with the replacement of CFAV P1-8 plus the P3 E to S mutation as mD2VLP. The pr/M cleavage of this construct was increased to nearly 90% (Figure 1B). Similar to wtD2VLP purified from plasmids transfected culture supernatants, the mD2VLP and imD2VLP showed consistent conformational integrity as ascertained by similar equilibrium banding profiles in 5 – 25% sucrose density gradients, as well as comparable particle size distributions as determined by negative staining electron microscopy, and complex glycosylation on the E proteins (Figure 1—figure supplement 2).

Figure 1 with 2 supplements see all

Download asset Open asset

Previous study has shown molecular organization of recombinant VLPs from tick-borne encephalitis virus (TBEV) (Ferlenghi et al., 2001). However, the cryo-EM results showed low resolution due to particle size heterogeneity and no immunological data can be derived from such structure. To better understand if the structural organization of mD2VLP is similar to that of TBEV and its immunological characteristics, purified mD2VLPs were subjected to cryo-EM analysis. The results demonstrated that the mD2VLPs preparation had a variable size distribution (Figure 2—figure supplement 1). In order to perform the single particle 3D reconstruction, first of all, we used a dengue virus-specific Mab (MAb32-6) (Li et al., 2012) to identify the mD2VLPs, it turned out that the spherical particles with a diameter of ~31 nm (Figure 2—figure supplement 2), which were also the major population as shown in Figure 2—figure supplement 1B, were well bound by antibodies. Besides, the particles with diameter of ~31 nm had more solid features in 2D image analyses (Figure 2—figure supplement 1) and were able to obtain stable structure potentially than other groups. Therefore, we continued and focused on this population for further 3D reconstruction and image analyses. The results showed that the 31nm-diameter mD2VLPs at a resolution of 13.1 Å had a hollow structure and smooth surface with protrusions around the 5-fold positions (Figure 2A, left, and Figure 2—figure supplement 3). Fitting the atomic E and M surface proteins PDB: 3J27) into the cryo-EM density map showed that the immunogold labelled mD2VLPs had 60 copies of E packed as dimers in a T = 1 icosahedral surface lattice (Figure 2A, right). The apparent differences of structural features on the mD2VLPs compared to the mature native virion particles (Kostyuchenko et al., 2013; Zhang et al., 2013a) was a slimmer lipid bilayer and the rearrangements of the surface proteins. First, the central section through the viron and m2DVLP reconstructions showed that the bilayer of mD2VLP was relatively thinner than that of viron. The distance between the exterior leaflet and the interior leaflet of mD2VLP and virion was 12 Å and 14 Å, respectively (Figure 2B). A similar T = 1 icosahedral symmetry of the E-protein arrangement and thinner lipid bilayer were also found in TBEV VLPs (Ferlenghi et al., 2001). Second, interpretation of the map showed that E dimer subunits moved apart from each other causing less density in the intra-dimeric interphase (Figure 3C). Because of this loose interaction, a groove located within the E-protein dimer near the icosahedral 2-fold vertices on mD2VLP was noted (Figure 3A, right, and Figure 3—figure supplement 1). Third, the central section showed that there was a solid density in the H-layer, which is composed of H helices of E and M protein stem regions, while the density in the T-layer, which contains transmembrane helices of E and M proteins, was weaker in the VLP (Figure 2B). Each E protein consists of an E ecto-domain and a stem region that connects the ecto-domain to its transmembrane region. The stem contains two α-helices, one of which interacts with the viral lipid membrane and the other with the E ecto-domain. This looser interaction of E protein dimers was further stabilized by the E protein rearrangement and the closer interaction with one of the α-helices, which resulted in the solid density beneath the ecto-domain as shown in Figure 2B. It is worth noting that the stem region of this mD2VLP was replaced into Japanese Encephalitis viral (JEV) sequence; it was previously suggested that this replacement increased the hydrophobicity for the interaction with the lipid membrane and enhanced the secretion of dengue VLP (Purdy and Chang, 2005).

Figure 2 with 3 supplements see all

Download asset Open asset

Figure 3 with 3 supplements see all

Download asset Open asset

Next, we speculated that the structure of mD2VLP with a groove located within the E-protein dimer could affect the epitope presentation on the surface of the particles. First, we calculated the solvent accessibility of mD2VLP and compared the results with that of DENV-2 virion. This rearrangement of E protein under T = 1 symmetry on the VLP surface exposed more accessible epitopes (48.2%; 191 amino acid with >30% solvent accessibility among 396 amino acids of the entire E protein) compared to DENV-2 virions (43.4%; 172 amino acid), particularly located at fusion loop, aa 239, aa 251 – 262 of EDII which together formed the groove, and at A strand, cd loop and G strand of EDIII which surrounded a 5-fold axis (Figure 3A, left, Figure 3—figure supplement 2). In silico footprint analysis (Figure 3B) showed that those cryptic residues on the E protein, which were previously buried inside of the native DENV virion and only interacted with the broad NtAbs while the virion ‘breathed' (Lok, 2016), were better exposed on the m2DVLP. Specifically, the epitopes interacting with MAb 1A1D-2, which binds to the virus at 37 ^oC (Lok et al., 2008); MAb 2D22, which could block two-thirds of all dimers on the virus surface, depending on the strain (Fibriansah et al., 2015a); and the ‘E-dimer-dependent epitope’, which is recognized by broadly neutralizing MAbs EDE1 and 2 (Rouvinski et al., 2015), were all well-exposed in VLPs without steric hindrance. Second, a panel of murine MAb, recognizing different domains of E protein and used previously for antigenic mapping (Crill et al., 2012), was subjected to binding-ELISA. The results confirmed that most of the conformational-dependent Mabs preferentially binds to mD2VLP, in particularly those recognizing domain I/II (Figure 4).

Figure 4

Download asset Open asset

Figure 4—source data 1

source data for antibody mapping results in Figure 4.

https://doi.org/10.7554/eLife.38970.014

Download elife-38970-fig4-data1-v2.xls

To further determine the influence of prM cleavage on the immunogenicity of D2VLP, we immunized groups of 4-week-old BALB/c mice with the purified wtD2VLP, imD2VLP or mD2VLP (4 μg/mouse) twice at a 4 week interval. To better represent the neutralizing antibodies at the late convalescent phase, serum samples collected at 8 weeks after the boost were analyzed by antigen-specific ELISA for antibody response against homologous D2VLP antigens with different maturity profiles. In order to precisely quantify the amount of dengue-specific antibodies, we used the same quantity of purified VLPs in the antigen-capture ELISA (Figure 5A). WtD2VLP and imD2VLP induced similar ELISA titers of DENV-2-specific IgG antibody against three VLP antigens. However, stronger reactivity against the mature antigen was noted with mD2VLP immunization (1:15,347 vs. 1:6381 vs. 2529 for mD2VLP, imD2VLP and wtD2VLP, respectively) (Figure 5B). We also analyzed the neutralizing ability of these sera against the four serotypes of DENV using a 50% antigen focus-reduction micro neutralization test (FRμNT₅₀). The mD2VLP immunization group induced a higher and broader neutralizing antibody response against all 4 serotypes of DENV (FRμNT₅₀ for DENV-1 = 1:331, DENV-2 = 1:597, DENV-3 = 1:70 and DENV-4 = 1:141), as compared to the imD2VLP vaccinated group (FRμNT₅₀ for DENV-1 = 1:100, DENV-2 = 1:207 DENV-3 = 1:64 and DENV-4 = 1:76) (Figure 5C).

Figure 5 with 4 supplements see all

Download asset Open asset

Figure 5—source data 1

source data for antibody mapping results in Figure 5.

https://doi.org/10.7554/eLife.38970.020

Download elife-38970-fig5-data1-v2.xlsx

Since the amino acid sequence is identical except for the mutations at the furin cleavage site, the difference in antibody binding and neutralizing activity between mD2VLP and imD2VLP vaccinated mice sera could result from the difference in induction of CR anti-prM non-NtAbs or anti-E NtAbs recognizing structure-dependent epitopes. To address whether the higher neutralization activity induced by mD2VLP was partly due to the reduction of anti-prM antibodies that are known to be cross-reactive but have no neutralizing activity (Dejnirattisai et al., 2010), we performed epitope-blocking ELISAs by using an anti-prM-specific monoclonal antibody (MAb 2H2). MAb 2H2 blocked only 7.59% of the activity of anti-mD2VLP sera but blocked up to 35.26% of antibodies from the imD2VLP-vaccinated groups (Figure 5—figure supplement 1). To avoid steric hindrance due to Mab 2H2 binding, we performed site-directed mutagenesis on three amino acids of pr protein based on the following criteria: (1) the conservation of amino acids among all four serotypes; (2) residues interacting with MAb 2H2 (Wang et al., 2013); (3) residues not interfering with prM and E interaction (Li et al., 2008). As shown in Figure 5—figure supplement 2, K26P was a key residue, which significantly decreased the binding of both MAb 2H2 and 155–49 (another cross-reactive murine Mab recognizing pr protein (Huang et al., 2006). However, only the mutant Δ2H2-imD2VLP with F1A, K21D and K26P amino acid triple mutations completely prevent the binding of both MAb 2H2 and 155–49 but not that of DENV-2, E-specific Nt-MAb 3H5 (Figure 5—figure supplement 3). The sera from immunized mice were tested for their binding ability to the wild-type and mutant-Δ2H2 imD2VLP. The differences in binding of imD2VLP and Δ2H2 imD2VLP were greater for the anti-imD2VLP sera than the anti-mD2VLP sera (Figure 5D and E), suggesting that compared to mD2VLP, imD2VLP was more likely to induce 2H2-like anti-prM antibodies. Thus, mD2VLP has the advantage of eliciting lower levels of anti-prM CR antibody and inducing antibodies with greater neutralizing activity.

The preservation of neutralizing epitopes on the surface of VLPs is critical for efficient production of broadly neutralizing antibody responses. Recent studies have suggested that human monoclonal antibodies (MAbs) reactive with all dengue serotypes can neutralize DENV in the low picomolar range. These MAbs have a preference to bind at the envelope dimer epitopes preserved on virion particles with a high degree of maturity (Rouvinski et al., 2015; Dejnirattisai et al., 2015). We next investigated if a broad neutralization response to mD2VLP immunization was due to the induction of antibodies with a preference to bind at the E dimer epitopes preserved on the surface of mature VLP (Rouvinski et al., 2015; Dejnirattisai et al., 2015). The splenocytes from mD2VLP-vaccinated mice were used to perform fusion and generate hybridomas. Among 2836 hybridomas screened, two MAbs with high reactivities to VLP measured by ELISA and FRμNT₅₀ were shown in Figure 6. As shown in Figure 6A, DM8-6 is a serotype-specific MAb reactive only with DENV-2; while MAb DM25-3 recognized all four serotypes of DENV. Next, we tested the neutralizing activity of these two MAbs against four serotypes of DENV. DM8-6 showed good neutralizing activity against DENV-2 at a concentration of 0.037 μg/mL in FRμNT₅₀ but poorly neutralized the other three serotypes. Consistent with the ELISA results, DM25-3 neutralized all four serotypes in FRμNT₅₀ at 0.32, 0.38, 0.24 and 0.58 μg/mL for DENV-1 to DENV-4, respectively (Figure 6B).

Figure 6 with 3 supplements see all

Download asset Open asset

Figure 6—source data 1

source data for antibody mapping results in Figure 6.

https://doi.org/10.7554/eLife.38970.025

Download elife-38970-fig6-data1-v2.xlsx

To determine if MAb DM25-3 recognized quaternary structure-dependent epitopes presented only on mature virion particles (Fibriansah et al., 2015a; Rouvinski et al., 2015), DM25-3 was tested for its binding activity to both mD2VLP and imD2VLP using antigen-capture ELISA. MAb DM25-3 recognized mD2VLP very well, but reacted poorly with imD2VLP (Figure 6C). Next, we performed site-directed mutagenesis on the fusion loop peptide in domain II and A-strand in domain III, both of which are important binding regions for CR group/complex neutralizing antibodies (de Alwis et al., 2011; Tsai et al., 2013) (Figure 6—figure supplement 1). The results suggested that amino acid E-101 was likely important in the binding site of DM25-3. Residue 101 on the fusion loop of E protein is conserved among all four serotypes of DENV and can only be exposed when virion particles undergo low-pH induced conformational change or during the ‘breathing’ state (Zhang et al., 2015; Zhang et al., 2013b). Figure 3C shows that residue 101 is located at the touching point between EDII and EDIII of the dimeric molecule. Since residue 101 is a critical residue in stabilizing E-protein dimers on DENV virion, mutation on 101 would disrupt viral particle formation (Zhang et al., 2013a; Rouvinski et al., 2017). However, mD2VLP with a mutation of residue 101 can still form particles (Figure 6—figure supplement 2), which further support our cryoEM model that the E:E protein interactions on VLP are loose. When the E:E protein interaction was looser, residue 101 at the groove within the dimeric molecule was more exposed (64%) than on the native virion (23.8%, PDB:3J27) or on the native virion during the 37°C ‘breathing’ state (45.3%, PDB:3ZKO). We also generated a recombinant virus of DENV-2 whose EDIII was exchanged for a consensus EDIII (Chen et al., 2013), and the recombinant virus PL046cEDIII was recovered from the transfected cell culture supernatants for use in ELISA and FRμNT. Compared to the parental DENV-2 strain PL046, there was significant loss of binding of MAb DM25-3 and immune mouse sera to PL046cEDIII (Figure 6D), suggesting that DM25-3 could be an E-dimer inter-domain antibody whose binding footprint is sensitive to amino acid changes in the EDIII domain. By comparing FRμNT₅₀ tests using murine antisera on both PL046 and PL046cEDIII viruses, we found that anti-mD2VLP sera showed a greater difference in neutralizing activity against PL046 and PL046cEDIII than did anti-imD2VLP sera, indicating a greater conformational dependence for anti-mD2VLP-triggered neutralization (Figure 6E).

The major obstacle of dengue vaccine development is the lack of a suitable small animal model to evaluate vaccine efficacy. Therefore, we decided to test whether the monovalent mD2VLP immunogen could provide protection from a lethal dose challenge of heterologous dengue virus serotypes using suckling mice developed in a previous study (Chang et al., 2003; Galula et al., 2014; Hughes et al., 2012). The advantage of using sucking mice to test vaccine efficacy is that the protection from dengue virus challenge can only come from maternal antibodies generated by vaccinated female mice and passively transferred to the their babies. The immunization and challenge schedule is illustrated in Figure 7A. As shown in Figure 7B, suckling mice born from mothers vaccinated with mD2VLP were all protected from challenge with all four serotypes of dengue virus.

Figure 7

Download asset Open asset

Figure 7—source data 1

source data for antibody mapping results in Figure 7.

https://doi.org/10.7554/eLife.38970.027

Download elife-38970-fig7-data1-v2.xlsx

How VLP immunogens mimic their viral counterparts structurally and how the neutralizing epitopes are preserved on the VLP surface area of significant interest in the development of good vaccine candidates. Recent studies suggest that potent human neutralizing antibodies with broad reactivity across dengue serocomplex can be generated from dengue patients, particularly after secondary infection (Rouvinski et al., 2015; Dejnirattisai et al., 2015; Tsai et al., 2013). Our unique findings here suggested that mD2VLP, with diameter of 31 nm and a scaffolding of multiple E-protein dimers similar to that of virion, could be capable of inducing such CR antibodies with broad neutralizing activity. This monovalent mature-form dengue VLP with ‘epitope re-surfaced’ has the potential to be the ‘epitope-focused’ antigens when combined with other live attenuated dengue vaccine to induce higher and broader neutralizing antibodies (Rouvinski et al., 2017; Rey et al., 2018).

Very few flavivirus studies explored the immunogenicity of prM-reduced VLP antigens and its capability in inducing broad NtAbs (Rouvinski et al., 2017; Keelapang et al., 2013; Suphatrakul et al., 2015; Metz et al., 2017). In TBEV study, VLP preparation showed a range of size distribution with the majority fell into 31 nm (Ferlenghi et al., 2001). Greater than 90% of the particles are mature and immunogenic. On the contrary, WNV VLPs were secreted as large (40–50 nm) and small (20–30 nm) particle sizes, which showed different maturity and immunogenicity. The large VLPs with the size of virions (50 nm) are more mature and induced higher NtAb titers and more potent protection against WNV challenge than the small immature VLPs (Ohtaki et al., 2010). Similar design of prM-reduced D2VLP was used as immunogens but was produced from mosquito cells and such VLP required adjuvants to boost immunity (Suphatrakul et al., 2015). Other attempts using E-dimers without the co-expression of prM cannot only display the quaternary structure epitopes but also reduce the exposure of FLE epitopes (Rouvinski et al., 2017; Metz et al., 2017; Metz et al., 2018). However, the immunogenicity of such E-dimers with only two copies of E could be lower than VLP. The mD2VLP used in this study was produced from mammalian cells with the features similar to dengue virion, such as glycosylation pattern, particle distribution in gradient after high-speed centrifugation, epitopes preserved on the surface by the mapping of monoclonal antibodies, and mostly importantly, multiple copies of E structurally packed as dimers on the particle surface lattice (Figure 1—figure supplement 2 and Figure 5—figure supplement 1). Compared to imD2VLP (35%), only 7.6% of prM-recognizing antibodies were induced by mD2VLP through 2H2-blocking assay. Current dengue vaccine candidates in the market such as CYD-TDV or others in clinical trial are either prM-possessing forms or E-protein monomer which is lack of quaternary epitopes (Wichmann et al., 2017). Our study suggested that mD2VLP has the potential to be a safer immunogen as the second-generation dengue vaccine.

In addition, mD2VLP-vaccinated mice produce antisera predominantly composed of about 48% of DM25-3-like antibodies (Figure 5—figure supplement 4). The packing of E dimers on VLPs does not just preserve quaternary structure epitopes on the surface lattice as suggested previously (Metz et al., 2018; Crill et al., 2009), but also provides a unique surface-accessible structure with increased epitope accessibility. Usually these epitopes are cryptic in mature virions maintained at 28°C and in a neutral-pH environment (Lok, 2016). By superimposing the E-dimer-dependent quaternary epitopes (Lok et al., 2008; Fibriansah et al., 2015a; Rouvinski et al., 2015) onto the structure of mD2VLP, the binding footprints of these antibodies were highly overlapping and formed a ‘neutralization-sensitive hotspot’ on mD2VLP (Figure 3—figure supplement 3). In spite of its size heterogeneity, the underlying mechanisms of why mD2VLPs are able to stimulate an elevated and broader immune response could be based on the following: (1) epitope accessibility exposed at the grooves within E-protein dimers, which govern the generation of neutralizing antibodies; (2) removal of decoy epitopes presented on prM-containing structures found in imD2VLPs; and (3) inter-dimeric epitope accessibility due to the 5-fold and 3-fold openings of the E protein arrangement of T = 1. Other mechanisms cannot be excluded such as reducing the production of FLE-like antibodies (ie. MAb E53 or 4G2), which have a preference of binding to spikes on noninfectious, immature flavivirions by engaging the highly conserved fusion loop that has limited solvent exposure of the epitope on mature virions (Cherrier et al., 2009).

The current obstacle in developing the second generation dengue vaccine is that dengue vaccine-induced neutralizing antibodies failed to correlate with or predict vaccine mediated protection (Yang et al., 2018; Moodie et al., 2018). The discrepancy between neutralization titer and protection against all four serotypes of DENV can also be found in our study (Figure 5C and Figure 7B). Previous studies have shown that neutralizing activity can be modulated by the maturity of virion in flavivirus (Pierson et al., 2008; Nelson et al., 2008). Several murine cross-reactive weak neutralizing antibodies, such as 4G2, can have neutralizing activity against all four serotypes of DENV at high concentrations, although they can enhance the virus replication while at lower concentrations (Zellweger et al., 2010). To confirm this argument, we mutated amino acid 101 on fusion peptide of imD2VLP and compared the loss of binding of antibodies from the immuned mice sera with the wild-type imD2VLP. Mutation of amino acid 101 on imD2VLP disrupted the binding of murine Mab 4G2. Meanwhile, the results showed that the proportion of 4G2-like antibodies among the sera receiving imD2VLP immunization was greater than that of sera either receiving mD2VLP or wtD2VLP (Figure 6—figure supplement 3). Recent studies suggested that the most potent neutralizing antibodies came from those recognizing quaternary epitopes on the smooth surface of dengue virions (Rouvinski et al., 2015; Fibriansah et al., 2015b; Crowe, 2017). Our study suggested that different types of D2VLP could raise different proportions of NtAbs, which are highly structure-dependent and could sometimes contribute to similar level of FRNT50, such as against DENV-3 (Figure 5C). Despite the size heterogeneity, we have found that a mature form of monovalent VLP from dengue virus serotype two with ‘epitope re-surfaced’ is efficient in inducing elevated and broadly NtAbs targeting quaternary epitopes. We cannot exclude the possibility of that the protection of mD2VLP immunization against all four serotypes of DENV challenge could also be due to non-neutralizing mechanism such as antibody-dependent cell-mediated cytotoxicity (ADCC). Therefore, it is crucial to understand the type of neutralizing antibodies induced by vaccination and establish the association between the level of such B-cell mediated immunological response and protection in the future study (Katzelnick et al., 2017; Flipse and Smit, 2015; Henein et al., 2017).

The major limitation of the current study is the unknown structure of imD2VLP or wtD2VLP to provide the proper explanation for the antibody response (Figure 5) or murine Mab mapping results (Figure 4). Our in silico footprint analysis showed the epitopes of Mab 1A1D-2 on DIII were well exposed on mD2VLP and was compared to DENV-2 virion (Figure 3), instead of imD2VLP. By comparing the structure of mD2VLP and D2 virion, it might be due to the groove within the E-protein dimer exposed several amino acids, including those located in both DI/II and DIII, previously suggested to be cryptic on virion (Figure 3A). However, our Mab mapping experiment was performed to compare between mD2VLP and imD2VLP. The results did confirm that several conformational-dependent Mabs, particularly those recognizing DI/II, were affected more by the structure of VLP with different maturity (Figure 4). The mapping results of Mab 1A1D-2 showing similar binding activity to both mD2VLP and imD2VLP (Figure 4D) might indicate DIII was less affected by VLP maturity. This is consistent with the previous study in TBEV VLPs (Lorenz et al., 2003). Our current follow-up study is to solve the structure of imD2VLP, which might give a clear and direct picture of structural differences from Mab mapping.

Despite of size heterogeneity, our unique findings in this study showed that a mature, monovalent DENV-2 VLP with diameter of 31 nm possessing grooves within the E protein dimeric molecules on the surface of particles has the potential to induce highly protective, NtAbs against heterologous dengue viruses from all four serotypes. When combined delivery with other live attenuated vaccine, such as CYD-TDV or dengue tetravalent vaccine, there is potential to provide not just T-cell immunity, but also higher broad NtAb response through ‘epitope-focusing’ while exploring the next generation dengue vaccine in humans (Crill et al., 2012; Rey et al., 2018; Flipse and Smit, 2015). The strategy here may also provide a new direction for the development of other flavivirus vaccines, including Zika virus.

The cryo-EM density map of mD2VLP has been deposited to Electron Microscopy Data Bank under accession number EMD-6926. All data generated or analyzed during this study are included in the manuscript and supporting files.

1. 1. Shang-Rung Wu
   2. Day-Yu Chao

   (2018) Electron Microscopy Data Bank

   ID EMD-6926. CryoEM structure of mature dengue virus-like particle.

   http://emsearch.rutgers.edu/atlas/6926_summary.html

1. 1. Aguiar M
   2. Stollenwerk N
   3. Halstead SB

   (2016) The Impact of the Newly Licensed Dengue Vaccine in Endemic Countries

   PLoS Neglected Tropical Diseases 10:e0005179.

   https://doi.org/10.1371/journal.pntd.0005179

   • PubMed
   • Google Scholar
2. 1. Allison SL
   2. Stadler K
   3. Mandl CW
   4. Kunz C
   5. Heinz FX

   (1995)

   Synthesis and secretion of recombinant tick-borne encephalitis virus protein E in soluble and particulate form

   Journal of Virology 69:5816–5820.

   • PubMed
   • Google Scholar
3. 1. Bhatt S
   2. Gething PW
   3. Brady OJ
   4. Messina JP
   5. Farlow AW
   6. Moyes CL
   7. Drake JM
   8. Brownstein JS
   9. Hoen AG
   10. Sankoh O
   11. Myers MF
   12. George DB
   13. Jaenisch T
   14. Wint GR
   15. Simmons CP
   16. Scott TW
   17. Farrar JJ
   18. Hay SI

   (2013) The global distribution and burden of dengue

   Nature 496:504–507.

   https://doi.org/10.1038/nature12060

   • PubMed
   • Google Scholar
4. 1. Cavallo L
   2. Kleinjung J
   3. Fraternali F

   (2003) POPS: a fast algorithm for solvent accessible surface areas at atomic and residue level

   Nucleic Acids Research 31:3364–3366.

   https://doi.org/10.1093/nar/gkg601

   • PubMed
   • Google Scholar
5. 1. Chambers TJ
   2. Hahn CS
   3. Galler R
   4. Rice CM

   (1990) Flavivirus genome organization, expression, and replication

   Annual Review of Microbiology 44:649–688.

   https://doi.org/10.1146/annurev.mi.44.100190.003245

   • PubMed
   • Google Scholar
6. 1. Chang GJ
   2. Hunt AR
   3. Davis B

   (2000) A single intramuscular injection of recombinant plasmid DNA induces protective immunity and prevents Japanese encephalitis in mice

   Journal of Virology 74:4244–4252.

   https://doi.org/10.1128/JVI.74.9.4244-4252.2000

   • PubMed
   • Google Scholar
7. 1. Chang GJ
   2. Hunt AR
   3. Holmes DA
   4. Springfield T
   5. Chiueh TS
   6. Roehrig JT
   7. Gubler DJ

   (2003) Enhancing biosynthesis and secretion of premembrane and envelope proteins by the chimeric plasmid of dengue virus type 2 and Japanese encephalitis virus

   Virology 306:170–180.

   https://doi.org/10.1016/S0042-6822(02)00028-4

   • PubMed
   • Google Scholar
8. 1. Chen YC
   2. Huang HN
   3. Lin CT
   4. Chen YF
   5. King CC
   6. Wu HC

   (2007) Generation and characterization of monoclonal antibodies against dengue virus type 1 for epitope mapping and serological detection by epitope-based peptide antigens

   Clinical and Vaccine Immunology 14:404–411.

   https://doi.org/10.1128/CVI.00249-06

   • PubMed
   • Google Scholar
9. 1. Chen HW
   2. Liu SJ
   3. Li YS
   4. Liu HH
   5. Tsai JP
   6. Chiang CY
   7. Chen MY
   8. Hwang CS
   9. Huang CC
   10. Hu HM
   11. Chung HH
   12. Wu SH
   13. Chong P
   14. Leng CH
   15. Pan CH

   (2013) A consensus envelope protein domain III can induce neutralizing antibody responses against serotype 2 of dengue virus in non-human primates

   Archives of Virology 158:1523–1531.

   https://doi.org/10.1007/s00705-013-1639-1

   • PubMed
   • Google Scholar
10. 1. Cherrier MV
    2. Kaufmann B
    3. Nybakken GE
    4. Lok SM
    5. Warren JT
    6. Chen BR
    7. Nelson CA
    8. Kostyuchenko VA
    9. Holdaway HA
    10. Chipman PR
    11. Kuhn RJ
    12. Diamond MS
    13. Rossmann MG
    14. Fremont DH

    (2009) Structural basis for the preferential recognition of immature flaviviruses by a fusion-loop antibody

    The EMBO Journal 28:3269–3276.

    https://doi.org/10.1038/emboj.2009.245

    • PubMed
    • Google Scholar
11. 1. Crill WD
    2. Hughes HR
    3. Delorey MJ
    4. Chang GJ

    (2009) Humoral immune responses of dengue fever patients using epitope-specific serotype-2 virus-like particle antigens

    PLoS One 4:e4991.

    https://doi.org/10.1371/journal.pone.0004991

    • PubMed
    • Google Scholar
12. 1. Crill WD
    2. Hughes HR
    3. Trainor NB
    4. Davis BS
    5. Whitney MT
    6. Chang GJ

    (2012) Sculpting humoral immunity through dengue vaccination to enhance protective immunity

    Frontiers in Immunology 3:334.

    https://doi.org/10.3389/fimmu.2012.00334

    • PubMed
    • Google Scholar
13. 1. Crowe JE

    (2017) Principles of broad and potent antiviral human antibodies: insights for vaccine design

    Cell Host & Microbe 22:193–206.

    https://doi.org/10.1016/j.chom.2017.07.013

    • PubMed
    • Google Scholar
14. 1. da Silva Voorham JM
    2. Rodenhuis-Zybert IA
    3. Ayala Nuñez NV
    4. Colpitts TM
    5. van der Ende-Metselaar H
    6. Fikrig E
    7. Diamond MS
    8. Wilschut J
    9. Smit JM

    (2012) Antibodies against the envelope glycoprotein promote infectivity of immature dengue virus serotype 2

    PLoS One 7:e29957.

    https://doi.org/10.1371/journal.pone.0029957

    • PubMed
    • Google Scholar
15. 1. de Alwis R
    2. Beltramello M
    3. Messer WB
    4. Sukupolvi-Petty S
    5. Wahala WM
    6. Kraus A
    7. Olivarez NP
    8. Pham Q
    9. Brien JD
    10. Brian J
    11. Tsai WY
    12. Wang WK
    13. Halstead S
    14. Kliks S
    15. Diamond MS
    16. Baric R
    17. Lanzavecchia A
    18. Sallusto F
    19. de Silva AM

    (2011) In-depth analysis of the antibody response of individuals exposed to primary dengue virus infection

    PLoS Neglected Tropical Diseases 5:e1188.

    https://doi.org/10.1371/journal.pntd.0001188

    • PubMed
    • Google Scholar
16. 1. de la Rosa-Trevín JM
    2. Otón J
    3. Marabini R
    4. Zaldívar A
    5. Vargas J
    6. Carazo JM
    7. Sorzano CO

    (2013) Xmipp 3.0: an improved software suite for image processing in electron microscopy

    Journal of Structural Biology 184:321–328.

    https://doi.org/10.1016/j.jsb.2013.09.015

    • PubMed
    • Google Scholar
17. 1. Dejnirattisai W
    2. Jumnainsong A
    3. Onsirisakul N
    4. Fitton P
    5. Vasanawathana S
    6. Limpitikul W
    7. Puttikhunt C
    8. Edwards C
    9. Duangchinda T
    10. Supasa S
    11. Chawansuntati K
    12. Malasit P
    13. Mongkolsapaya J
    14. Screaton G

    (2010) Cross-reacting antibodies enhance dengue virus infection in humans

    Science 328:745–748.

    https://doi.org/10.1126/science.1185181

    • PubMed
    • Google Scholar
18. 1. Dejnirattisai W
    2. Wongwiwat W
    3. Supasa S
    4. Zhang X
    5. Dai X
    6. Rouvinski A
    7. Jumnainsong A
    8. Edwards C
    9. Quyen NTH
    10. Duangchinda T
    11. Grimes JM
    12. Tsai WY
    13. Lai CY
    14. Wang WK
    15. Malasit P
    16. Farrar J
    17. Simmons CP
    18. Zhou ZH
    19. Rey FA
    20. Mongkolsapaya J
    21. Screaton GR

    (2015) A new class of highly potent, broadly neutralizing antibodies isolated from viremic patients infected with dengue virus

    Nature Immunology 16:170–177.

    https://doi.org/10.1038/ni.3058

    • PubMed
    • Google Scholar
19. 1. Eswar N
    2. Webb B
    3. Marti-Renom MA
    4. Madhusudhan MS
    5. Eramian D
    6. Shen MY
    7. Pieper U
    8. Sali A

    (2007) Comparative protein structure modeling using MODELLER

    Current Protocols in Protein Science 50:2.9.1–2.9.2.

    https://doi.org/10.1002/0471140864.ps0209s50

    • Google Scholar
20. 1. Ferlenghi I
    2. Clarke M
    3. Ruttan T
    4. Allison SL
    5. Schalich J
    6. Heinz FX
    7. Harrison SC
    8. Rey FA
    9. Fuller SD

    (2001) Molecular organization of a recombinant subviral particle from tick-borne encephalitis virus

    Molecular Cell 7:593–602.

    https://doi.org/10.1016/S1097-2765(01)00206-4

    • PubMed
    • Google Scholar
21. 1. Fibriansah G
    2. Ibarra KD
    3. Ng TS
    4. Smith SA
    5. Tan JL
    6. Lim XN
    7. Ooi JS
    8. Kostyuchenko VA
    9. Wang J
    10. de Silva AM
    11. Harris E
    12. Crowe JE
    13. Lok SM

    (2015a) DENGUE VIRUS. Cryo-EM structure of an antibody that neutralizes dengue virus type 2 by locking E protein dimers

    Science 349:88–91.

    https://doi.org/10.1126/science.aaa8651

    • PubMed
    • Google Scholar
22. 1. Fibriansah G
    2. Tan JL
    3. Smith SA
    4. de Alwis R
    5. Ng TS
    6. Kostyuchenko VA
    7. Jadi RS
    8. Kukkaro P
    9. de Silva AM
    10. Crowe JE
    11. Lok SM

    (2015b) A highly potent human antibody neutralizes dengue virus serotype 3 by binding across three surface proteins

    Nature Communications 6:6341.

    https://doi.org/10.1038/ncomms7341

    • PubMed
    • Google Scholar
23. 1. Flipse J
    2. Smit JM

    (2015) The Complexity of a Dengue Vaccine: A Review of the Human Antibody Response

    PLoS Neglected Tropical Diseases 9:e0003749.

    https://doi.org/10.1371/journal.pntd.0003749

    • PubMed
    • Google Scholar
24. 1. Fornili A
    2. Autore F
    3. Chakroun N
    4. Martinez P
    5. Fraternali F

    (2012) Protein-water interactions in MD simulations: pops/POPSCOMP solvent accessibility analysis, solvation forces and hydration sites

    Methods in Molecular Biology 819:375–392.

    https://doi.org/10.1007/978-1-61779-465-0_23

    • PubMed
    • Google Scholar
25. 1. Galula JU
    2. Shen WF
    3. Chuang ST
    4. Chang GJ
    5. Chao DY

    (2014) Virus-like particle secretion and genotype-dependent immunogenicity of dengue virus serotype 2 DNA vaccine

    Journal of Virology 88:10813–10830.

    https://doi.org/10.1128/JVI.00810-14

    • PubMed
    • Google Scholar
26. 1. Guy B
    2. Jackson N

    (2016) Dengue vaccine: hypotheses to understand CYD-TDV-induced protection

    Nature Reviews Microbiology 14:45–54.

    https://doi.org/10.1038/nrmicro.2015.2

    • PubMed
    • Google Scholar
27. 1. Halstead SB

    (2017) Dengvaxia sensitizes seronegatives to vaccine enhanced disease regardless of age

    Vaccine 35:6355–6358.

    https://doi.org/10.1016/j.vaccine.2017.09.089

    • PubMed
    • Google Scholar
28. 1. Henein S
    2. Swanstrom J
    3. Byers AM
    4. Moser JM
    5. Shaik SF
    6. Bonaparte M
    7. Jackson N
    8. Guy B
    9. Baric R
    10. de Silva AM

    (2017) Dissecting antibodies induced by a chimeric yellow Fever-Dengue, Live-Attenuated, tetravalent dengue vaccine (CYD-TDV) in naive and Dengue-Exposed individuals

    The Journal of Infectious Diseases 215:351–358.

    https://doi.org/10.1093/infdis/jiw576

    • PubMed
    • Google Scholar
29. 1. Huang KJ
    2. Yang YC
    3. Lin YS
    4. Huang JH
    5. Liu HS
    6. Yeh TM
    7. Chen SH
    8. Liu CC
    9. Lei HY

    (2006) The dual-specific binding of dengue virus and target cells for the antibody-dependent enhancement of dengue virus infection

    The Journal of Immunology 176:2825–2832.

    https://doi.org/10.4049/jimmunol.176.5.2825

    • PubMed
    • Google Scholar
30. 1. Hughes HR
    2. Crill WD
    3. Chang GJ

    (2012) Manipulation of immunodominant dengue virus E protein epitopes reduces potential antibody-dependent enhancement

    Virology Journal 9:115.

    https://doi.org/10.1186/1743-422X-9-115

    • PubMed
    • Google Scholar
31. 1. Junjhon J
    2. Edwards TJ
    3. Utaipat U
    4. Bowman VD
    5. Holdaway HA
    6. Zhang W
    7. Keelapang P
    8. Puttikhunt C
    9. Perera R
    10. Chipman PR
    11. Kasinrerk W
    12. Malasit P
    13. Kuhn RJ
    14. Sittisombut N

    (2010) Influence of pr-M cleavage on the heterogeneity of extracellular dengue virus particles

    Journal of Virology 84:8353–8358.

    https://doi.org/10.1128/JVI.00696-10

    • PubMed
    • Google Scholar
32. 1. Katzelnick LC
    2. Coloma J
    3. Harris E

    (2017) Dengue: knowledge gaps, unmet needs, and research priorities

    The Lancet Infectious Diseases 17:e88–e100.

    https://doi.org/10.1016/S1473-3099(16)30473-X

    • PubMed
    • Google Scholar
33. 1. Keelapang P
    2. Nitatpattana N
    3. Suphatrakul A
    4. Punyahathaikul S
    5. Sriburi R
    6. Pulmanausahakul R
    7. Pichyangkul S
    8. Malasit P
    9. Yoksan S
    10. Sittisombut N

    (2013) Generation and preclinical evaluation of a DENV-1/2 prM+E chimeric live attenuated vaccine candidate with enhanced prM cleavage

    Vaccine 31:5134–5140.

    https://doi.org/10.1016/j.vaccine.2013.08.027

    • PubMed
    • Google Scholar
34. 1. Kirkpatrick BD
    2. Whitehead SS
    3. Pierce KK
    4. Tibery CM
    5. Grier PL
    6. Hynes NA
    7. Larsson CJ
    8. Sabundayo BP
    9. Talaat KR
    10. Janiak A
    11. Carmolli MP
    12. Luke CJ
    13. Diehl SA
    14. Durbin AP

    (2016) The live attenuated dengue vaccine TV003 elicits complete protection against dengue in a human challenge model

    Science Translational Medicine 8:330ra36.

    https://doi.org/10.1126/scitranslmed.aaf1517

    • PubMed
    • Google Scholar
35. 1. Köhler G
    2. Milstein C

    (1975) Continuous cultures of fused cells secreting antibody of predefined specificity

    Nature 256:495–497.

    https://doi.org/10.1038/256495a0

    • PubMed
    • Google Scholar
36. 1. Kostyuchenko VA
    2. Zhang Q
    3. Tan JL
    4. Ng TS
    5. Lok SM

    (2013) Immature and mature dengue serotype 1 virus structures provide insight into the maturation process

    Journal of Virology 87:7700–7707.

    https://doi.org/10.1128/JVI.00197-13

    • PubMed
    • Google Scholar
37. 1. Li L
    2. Lok SM
    3. Yu IM
    4. Zhang Y
    5. Kuhn RJ
    6. Chen J
    7. Rossmann MG

    (2008) The flavivirus precursor membrane-envelope protein complex: structure and maturation

    Science 319:1830–1834.

    https://doi.org/10.1126/science.1153263

    • PubMed
    • Google Scholar
38. 1. Li PC
    2. Liao MY
    3. Cheng PC
    4. Liang JJ
    5. Liu IJ
    6. Chiu CY
    7. Lin YL
    8. Chang GJ
    9. Wu HC

    (2012) Development of a humanized antibody with high therapeutic potential against dengue virus type 2

    PLoS Neglected Tropical Diseases 6:e1636.

    https://doi.org/10.1371/journal.pntd.0001636

    • PubMed
    • Google Scholar
39. 1. Liang JJ
    2. Liao CL
    3. Liao JT
    4. Lee YL
    5. Lin YL

    (2009) A Japanese encephalitis virus vaccine candidate strain is attenuated by decreasing its interferon antagonistic ability

    Vaccine 27:2746–2754.

    https://doi.org/10.1016/j.vaccine.2009.03.007

    • PubMed
    • Google Scholar
40. Book

    1. Lindenbach BD
    2. Rice CM

    (2001)

    "Flaviviridae: The Viruses and Their Replication." in Fields' Virology, 4

    Knipe D. M, Howley P. M, editors.  Lippincott Williams & Wilkins, Philadelphia.

    • Google Scholar
41. 1. Lok SM
    2. Kostyuchenko V
    3. Nybakken GE
    4. Holdaway HA
    5. Battisti AJ
    6. Sukupolvi-Petty S
    7. Sedlak D
    8. Fremont DH
    9. Chipman PR
    10. Roehrig JT
    11. Diamond MS
    12. Kuhn RJ
    13. Rossmann MG

    (2008) Binding of a neutralizing antibody to dengue virus alters the arrangement of surface glycoproteins

    Nature Structural & Molecular Biology 15:312–317.

    https://doi.org/10.1038/nsmb.1382

    • PubMed
    • Google Scholar
42. 1. Lok SM

    (2016) The interplay of dengue virus morphological diversity and human antibodies

    Trends in Microbiology 24:284–293.

    https://doi.org/10.1016/j.tim.2015.12.004

    • PubMed
    • Google Scholar
43. 1. Lorenz IC
    2. Kartenbeck J
    3. Mezzacasa A
    4. Allison SL
    5. Heinz FX
    6. Helenius A

    (2003) Intracellular assembly and secretion of recombinant subviral particles from tick-borne encephalitis virus

    Journal of Virology 77:4370–4382.

    https://doi.org/10.1128/JVI.77.7.4370-4382.2003

    • PubMed
    • Google Scholar
44. 1. Metz SW
    2. Gallichotte EN
    3. Brackbill A
    4. Premkumar L
    5. Miley MJ
    6. Baric R
    7. de Silva AM

    (2017) In vitro assembly and stabilization of dengue and zika virus envelope protein Homo-Dimers

    Scientific Reports 7:4524.

    https://doi.org/10.1038/s41598-017-04767-6

    • PubMed
    • Google Scholar
45. 1. Metz SW
    2. Thomas A
    3. White L
    4. Stoops M
    5. Corten M
    6. Hannemann H
    7. de Silva AM

    (2018) Dengue virus-like particles mimic the antigenic properties of the infectious dengue virus envelope

    Virology Journal 15:60.

    https://doi.org/10.1186/s12985-018-0970-2

    • PubMed
    • Google Scholar
46. 1. Modis Y
    2. Ogata S
    3. Clements D
    4. Harrison SC

    (2003) A ligand-binding pocket in the dengue virus envelope glycoprotein

    PNAS 100:6986–6991.

    https://doi.org/10.1073/pnas.0832193100

    • PubMed
    • Google Scholar
47. 1. Moodie Z
    2. Juraska M
    3. Huang Y
    4. Zhuang Y
    5. Fong Y
    6. Carpp LN
    7. Self SG
    8. Chambonneau L
    9. Small R
    10. Jackson N
    11. Noriega F
    12. Gilbert PB

    (2018) Neutralizing antibody correlates analysis of tetravalent dengue vaccine efficacy trials in Asia and Latin America

    The Journal of Infectious Diseases 217:742–753.

    https://doi.org/10.1093/infdis/jix609

    • PubMed
    • Google Scholar
48. 1. Nelson S
    2. Jost CA
    3. Xu Q
    4. Ess J
    5. Martin JE
    6. Oliphant T
    7. Whitehead SS
    8. Durbin AP
    9. Graham BS
    10. Diamond MS
    11. Pierson TC

    (2008) Maturation of West Nile virus modulates sensitivity to antibody-mediated neutralization

    PLoS Pathogens 4:e1000060.

    https://doi.org/10.1371/journal.ppat.1000060

    • PubMed
    • Google Scholar
49. 1. Ohtaki N
    2. Takahashi H
    3. Kaneko K
    4. Gomi Y
    5. Ishikawa T
    6. Higashi Y
    7. Kurata T
    8. Sata T
    9. Kojima A

    (2010) Immunogenicity and efficacy of two types of west nile virus-like particles different in size and maturation as a second-generation vaccine candidate

    Vaccine 28:6588–6596.

    https://doi.org/10.1016/j.vaccine.2010.07.055

    • PubMed
    • Google Scholar
50. 1. Pierson TC
    2. Fremont DH
    3. Kuhn RJ
    4. Diamond MS

    (2008) Structural insights into the mechanisms of antibody-mediated neutralization of flavivirus infection: implications for vaccine development

    Cell Host & Microbe 4:229–238.

    https://doi.org/10.1016/j.chom.2008.08.004

    • PubMed
    • Google Scholar
51. 1. Pierson TC
    2. Diamond MS

    (2012) Degrees of maturity: the complex structure and biology of flaviviruses

    Current Opinion in Virology 2:168–175.

    https://doi.org/10.1016/j.coviro.2012.02.011

    • PubMed
    • Google Scholar
52. 1. Purdy DE
    2. Chang GJ

    (2005) Secretion of noninfectious dengue virus-like particles and identification of amino acids in the stem region involved in intracellular retention of envelope protein

    Virology 333:239–250.

    https://doi.org/10.1016/j.virol.2004.12.036

    • PubMed
    • Google Scholar
53. 1. Rey FA
    2. Stiasny K
    3. Vaney MC
    4. Dellarole M
    5. Heinz FX

    (2018) The bright and the dark side of human antibody responses to flaviviruses: lessons for vaccine design

    EMBO Reports 19:206–224.

    https://doi.org/10.15252/embr.201745302

    • PubMed
    • Google Scholar
54. 1. Rodenhuis-Zybert IA
    2. van der Schaar HM
    3. da Silva Voorham JM
    4. van der Ende-Metselaar H
    5. Lei HY
    6. Wilschut J
    7. Smit JM

    (2010) Immature dengue virus: a veiled pathogen?

    PLoS Pathogens 6:e1000718.

    https://doi.org/10.1371/journal.ppat.1000718

    • PubMed
    • Google Scholar
55. 1. Rodenhuis-Zybert IA
    2. Moesker B
    3. da Silva Voorham JM
    4. van der Ende-Metselaar H
    5. Diamond MS
    6. Wilschut J
    7. Smit JM

    (2011) A fusion-loop antibody enhances the infectious properties of immature Flavivirus particles

    Journal of Virology 85:11800–11808.

    https://doi.org/10.1128/JVI.05237-11

    • PubMed
    • Google Scholar
56. 1. Rodenhuis-Zybert IA
    2. da Silva Voorham JM
    3. Torres S
    4. van de Pol D
    5. Smit JM

    (2015) Antibodies against immature virions are not a discriminating factor for dengue disease severity

    PLoS Neglected Tropical Diseases 9:e0003564.

    https://doi.org/10.1371/journal.pntd.0003564

    • PubMed
    • Google Scholar
57. 1. Roehrig JT

    (2003) Antigenic structure of flavivirus proteins

    Advances in Virus Research 59:141–175.

    https://doi.org/10.1016/S0065-3527(03)59005-4

    • PubMed
    • Google Scholar
58. 1. Rouvinski A
    2. Guardado-Calvo P
    3. Barba-Spaeth G
    4. Duquerroy S
    5. Vaney MC
    6. Kikuti CM
    7. Navarro Sanchez ME
    8. Dejnirattisai W
    9. Wongwiwat W
    10. Haouz A
    11. Girard-Blanc C
    12. Petres S
    13. Shepard WE
    14. Desprès P
    15. Arenzana-Seisdedos F
    16. Dussart P
    17. Mongkolsapaya J
    18. Screaton GR
    19. Rey FA

    (2015) Recognition determinants of broadly neutralizing human antibodies against dengue viruses

    Nature 520:109–113.

    https://doi.org/10.1038/nature14130

    • PubMed
    • Google Scholar
59. 1. Rouvinski A
    2. Dejnirattisai W
    3. Guardado-Calvo P
    4. Vaney MC
    5. Sharma A
    6. Duquerroy S
    7. Supasa P
    8. Wongwiwat W
    9. Haouz A
    10. Barba-Spaeth G
    11. Mongkolsapaya J
    12. Rey FA
    13. Screaton GR

    (2017) Covalently linked dengue virus envelope glycoprotein dimers reduce exposure of the immunodominant fusion loop epitope

    Nature Communications 8:15411.

    https://doi.org/10.1038/ncomms15411

    • PubMed
    • Google Scholar
60. 1. Schägger H

    (2006) Tricine-SDS-PAGE

    Nature Protocols 1:16–22.

    https://doi.org/10.1038/nprot.2006.4

    • PubMed
    • Google Scholar
61. 1. Sorzano CO
    2. Marabini R
    3. Velázquez-Muriel J
    4. Bilbao-Castro JR
    5. Scheres SH
    6. Carazo JM
    7. Pascual-Montano A

    (2004) XMIPP: a new generation of an open-source image processing package for electron microscopy

    Journal of Structural Biology 148:194–204.

    https://doi.org/10.1016/j.jsb.2004.06.006

    • PubMed
    • Google Scholar
62. 1. Suphatrakul A
    2. Yasanga T
    3. Keelapang P
    4. Sriburi R
    5. Roytrakul T
    6. Pulmanausahakul R
    7. Utaipat U
    8. Kawilapan Y
    9. Puttikhunt C
    10. Kasinrerk W
    11. Yoksan S
    12. Auewarakul P
    13. Malasit P
    14. Charoensri N
    15. Sittisombut N

    (2015) Generation and preclinical immunogenicity study of dengue type 2 virus-like particles derived from stably transfected mosquito cells

    Vaccine 33:5613–5622.

    https://doi.org/10.1016/j.vaccine.2015.08.090

    • PubMed
    • Google Scholar
63. 1. Tang G
    2. Peng L
    3. Baldwin PR
    4. Mann DS
    5. Jiang W
    6. Rees I
    7. Ludtke SJ

    (2007) EMAN2: an extensible image processing suite for electron microscopy

    Journal of Structural Biology 157:38–46.

    https://doi.org/10.1016/j.jsb.2006.05.009

    • PubMed
    • Google Scholar
64. 1. Tian S
    2. Huajun W
    3. Wu J

    (2012) Computational prediction of furin cleavage sites by a hybrid method and understanding mechanism underlying diseases

    Scientific Reports 2:261.

    https://doi.org/10.1038/srep00261

    • PubMed
    • Google Scholar
65. 1. Tsai WY
    2. Lai CY
    3. Wu YC
    4. Lin HE
    5. Edwards C
    6. Jumnainsong A
    7. Kliks S
    8. Halstead S
    9. Mongkolsapaya J
    10. Screaton GR
    11. Wang WK

    (2013) High-avidity and potently neutralizing cross-reactive human monoclonal antibodies derived from secondary dengue virus infection

    Journal of Virology 87:12562–12575.

    https://doi.org/10.1128/JVI.00871-13

    • PubMed
    • Google Scholar
66. 1. Wang Z
    2. Li L
    3. Pennington JG
    4. Sheng J
    5. Yap ML
    6. Plevka P
    7. Meng G
    8. Sun L
    9. Jiang W
    10. Rossmann MG

    (2013) Obstruction of dengue virus maturation by Fab fragments of the 2H2 antibody

    Journal of Virology 87:8909–8915.

    https://doi.org/10.1128/JVI.00472-13

    • PubMed
    • Google Scholar
67. 1. WHO

    (2017) Dengue vaccine: who position paper, July 2016 - recommendations

    Vaccine 35:1200–1201.

    https://doi.org/10.1016/j.vaccine.2016.10.070

    • PubMed
    • Google Scholar
68. 1. Wichmann O
    2. Vannice K
    3. Asturias EJ
    4. de Albuquerque Luna EJ
    5. Longini I
    6. Lopez AL
    7. Smith PG
    8. Tissera H
    9. Yoon IK
    10. Hombach J

    (2017) Live-attenuated tetravalent dengue vaccines: The needs and challenges of post-licensure evaluation of vaccine safety and effectiveness

    Vaccine 35:5535–5542.

    https://doi.org/10.1016/j.vaccine.2017.08.066

    • PubMed
    • Google Scholar
69. 1. Yang Y
    2. Meng Y
    3. Halloran ME
    4. Longini IM

    (2018) Dependency of vaccine efficacy on preexposure and age: a closer look at a tetravalent dengue vaccine

    Clinical Infectious Diseases 66:178–184.

    https://doi.org/10.1093/cid/cix766

    • PubMed
    • Google Scholar
70. 1. Zellweger RM
    2. Prestwood TR
    3. Shresta S

    (2010) Enhanced infection of liver sinusoidal endothelial cells in a mouse model of antibody-induced severe dengue disease

    Cell Host & Microbe 7:128–139.

    https://doi.org/10.1016/j.chom.2010.01.004

    • PubMed
    • Google Scholar
71. 1. Zhang SL
    2. Tan HC
    3. Hanson BJ
    4. Ooi EE

    (2010) A simple method for Alexa Fluor dye labelling of dengue virus

    Journal of Virological Methods 167:172–177.

    https://doi.org/10.1016/j.jviromet.2010.04.001

    • PubMed
    • Google Scholar
72. 1. Zhang X
    2. Ge P
    3. Yu X
    4. Brannan JM
    5. Bi G
    6. Zhang Q
    7. Schein S
    8. Zhou ZH

    (2013a) Cryo-EM structure of the mature dengue virus at 3.5-Å resolution

    Nature Structural & Molecular Biology 20:105–110.

    https://doi.org/10.1038/nsmb.2463

    • PubMed
    • Google Scholar
73. 1. Zhang X
    2. Sheng J
    3. Plevka P
    4. Kuhn RJ
    5. Diamond MS
    6. Rossmann MG

    (2013b) Dengue structure differs at the temperatures of its human and mosquito hosts

    PNAS 110:6795–6799.

    https://doi.org/10.1073/pnas.1304300110

    • PubMed
    • Google Scholar
74. 1. Zhang X
    2. Sheng J
    3. Austin SK
    4. Hoornweg TE
    5. Smit JM
    6. Kuhn RJ
    7. Diamond MS
    8. Rossmann MG

    (2015) Structure of acidic pH dengue virus showing the fusogenic glycoprotein trimers

    Journal of Virology 89:743–750.

    https://doi.org/10.1128/JVI.02411-14

    • PubMed
    • Google Scholar

Author details

1. Wen-Fan Shen

   Microbial Genomics Ph.D. Program, National Chung Hsing University and Academia Sinica, Taichung City, Taiwan

   Contribution

   Data curation, Investigation, Visualization

   Contributed equally with

   Jedhan Ucat Galula, Jyung-Hurng Liu, Mei-Ying Liao and Day-Yu Chao

   Competing interests

   No competing interests declared

2. Jedhan Ucat Galula

   Graduate Institute of Microbiology and Public Health, College of Veterinary Medicine, National Chung-Hsing University, Taichung City, Taiwan

   Contribution

   Data curation, Investigation, Visualization

   Contributed equally with

   Wen-Fan Shen, Jyung-Hurng Liu, Mei-Ying Liao and Day-Yu Chao

   Competing interests

   No competing interests declared

3. Jyung-Hurng Liu

   Institute of Genomics and Bioinformatics, College of Life Science, National Chung-Hsing University, Taichung City, Taiwan

   Contribution

   Software, Formal analysis, Visualization

   Contributed equally with

   Wen-Fan Shen, Jedhan Ucat Galula, Mei-Ying Liao and Day-Yu Chao

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7173-0372

4. Mei-Ying Liao

   Graduate Institute of Microbiology and Public Health, College of Veterinary Medicine, National Chung-Hsing University, Taichung City, Taiwan

   Contribution

   Data curation, Formal analysis, Investigation

   Contributed equally with

   Wen-Fan Shen, Jedhan Ucat Galula, Jyung-Hurng Liu and Day-Yu Chao

   Competing interests

   No competing interests declared

5. Cheng-Hao Huang

   Graduate Institute of Microbiology and Public Health, College of Veterinary Medicine, National Chung-Hsing University, Taichung City, Taiwan

   Contribution

   Data curation, Investigation

   Competing interests

   No competing interests declared

6. Yu-Chun Wang

   Graduate Institute of Microbiology and Public Health, College of Veterinary Medicine, National Chung-Hsing University, Taichung City, Taiwan

   Contribution

   Data curation, Investigation

   Competing interests

   No competing interests declared

7. Han-Chung Wu

   Institute of Cellular and Organismic Biology, Academia Sinica, Taipei, Taiwan

   Contribution

   Resources, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-5185-1169

8. Jian-Jong Liang

   Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan

   Contribution

   Resources, Investigation

   Competing interests

   No competing interests declared

9. Yi-Ling Lin

   Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan

   Contribution

   Resources

   Competing interests

   No competing interests declared

10. Matthew T Whitney

    Division of Vector-Borne Diseases, Centers for Disease Control and Prevention, Fort Collins, Colorado, United States

    Contribution

    Data curation, Investigation

    Competing interests

    No competing interests declared

11. Gwong-Jen J Chang

    Division of Vector-Borne Diseases, Centers for Disease Control and Prevention, Fort Collins, Colorado, United States

    Contribution

    Resources, Supervision, Writing—review and editing

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0001-9959-6585

12. Sheng-Ren Chen

    Institute of Oral Medicine, College of Medicine, National Cheng Kung University, Tainan, Taiwan

    Contribution

    Data curation, Formal analysis, Investigation

    Competing interests

    No competing interests declared

13. Shang-Rung Wu

    1. Institute of Oral Medicine, College of Medicine, National Cheng Kung University, Tainan, Taiwan
    2. Institute of Basic Medical Sciences, College of Medicine, National Cheng Kung University, Tainan, Taiwan

    Contribution

    Conceptualization, Data curation, Software, Supervision, Funding acquisition, Validation, Investigation, Visualization, Writing—original draft, Project administration, Writing—review and editing

    For correspondence

    shangrungwu@gmail.com

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0003-1940-7612

14. Day-Yu Chao

    Graduate Institute of Microbiology and Public Health, College of Veterinary Medicine, National Chung-Hsing University, Taichung City, Taiwan

    Contribution

    Conceptualization, Resources, Data curation, Supervision, Funding acquisition, Validation, Investigation, Methodology, Writing—original draft, Project administration, Writing—review and editing

    Contributed equally with

    Wen-Fan Shen, Jedhan Ucat Galula, Jyung-Hurng Liu and Mei-Ying Liao

    For correspondence

    dychao@nchu.edu.tw

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0001-7139-026X

• 4,229

  views

• 712

  downloads

• 25

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.