Envelope-modified tetravalent dengue virus-like particle vaccine: implication for flavivirus vaccine design

16 Dengue viruses (DENV) infect 50-100 million people each year. The spread of DENV17 associated infections is one of the most serious public health problems worldwide, as there is no 18 widely available vaccine or specific therapeutic for DENV infections. To address this, we 19 developed a novel tetravalent dengue vaccine utilizing virus-like particle (VLP) technology. We 20 created recombinant DENV1-4 VLPs by co-expressing precursor membrane (prM) and envelope 21 (E) proteins, with a F108A mutation in the fusion loop structure of E to increase the production 22 of VLPs in mammalian cells. Immunization with DENV1-4 VLPs as individual, monovalent 23 vaccines elicited strong neutralization activity against each DENV serotype in mice. When 24 immunized as a tetravalent vaccine, DENV1-4 VLPs elicited high levels of neutralization 25 activity against all four serotypes simultaneously. The neutralization antibody response induced 26 by the VLPs was significantly higher than DNA or recombinant E proteins immunization. 27 Moreover, antibody-dependent enhancement (ADE) was not observed against any serotype at 28 1:10 serum dilution. We also demonstrated that Zika virus (ZIKV) VLP production level was 29 enhanced by introducing the same F108A mutation in ZIKV envelope protein. Taken together, 30 these results suggest that our strategy for DENV VLP production is applicable to other flavivirus 31 VLP vaccine development, due to the similarity in their viral structures and describes the 32 promising development of an effective tetravalent vaccine against the prevalent flavivirus. 33 34 Importance: The dengue virus poses one of the most serious public health problems worldwide, 35 and the incidence of diseases caused by the virus has increased dramatically. Despite decades of 36 effort, there is no effective treatment against dengue. A safe and potent vaccine against dengue is 37 still needed. We have developed a novel tetravalent dengue vaccine using virus-like particle 38 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom (VLP) technology, which is non-infectious as it lacks viral genome. Previous attempts by other 39 groups with dengue virus VLPs resulted in generally poor yields. We found that a critical amino 40 acid mutation in the envelope protein enhances the production of VLP. Our tetravalent vaccine 41 elicited potent neutralizing antibody responses against all four serotypes. Our finding can also be 42 applied to vaccine development against other flaviviruses, such as Zika virus or West Nile virus. 43 44 Introduction 45 Dengue is a mosquito-borne disease caused by dengue virus (DENV), and has been recognized 46 as a serious public health problem worldwide. DENV is a positive-strand RNA virus that belongs 47 to the Flavivirus genus of the Flaviviridae family. There are four DENV serotypes co-circulating 48 in endemic areas, which share 60-75% identity at the amino acid level but are clinically 49 indistinguishable (1). Infection by any of the four serotypes of DENV causes dengue fever, 50 which is a flu-like febrile illness, and occasionally progresses to life-threatening dengue 51 hemorrhagic fever or dengue shock syndrome (2). About 50% of the world’s population is 52 currently at risk of DENV infection (3). There remains no effective dengue-specific antiviral 53 treatment or therapy, and vector control efforts to prevent the spread of DENV have been 54 ineffective (4). Therefore, an effective vaccine is viewed as one of the most desired method to 55 control this disease. 56 A major challenge in dengue vaccine development is the existence of closely related four 57 DENV serotypes. After an initial infection with one DENV serotype, individuals who are 58 subsequently exposed to any of the other serotypes are more likely to develop a more severe case 59 of the disease, due to a phenomenon known as antibody-dependent enhancement (ADE): it has 60 been reported that non-neutralizing levels of anti-DENV antibody can enhance viral entry into 61 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom host cells by forming a DENV-antibody complex (5-7). There is the concern that an incomplete 62 immune response upon a first immunization may cause ADE-mediated severe dengue disease 63 during the period between the first and the last immunization. Hence, there is a need for a safe 64 and highly efficacious dengue vaccine that provides long-lasting immunity against all four 65 serotypes simultaneously with a short immunization schedule. 66 Currently, CYD-TDV (Dengvaxia®) is the only licensed dengue vaccine in the world. 67 CYD-TDV is a live attenuated tetravalent dengue vaccine developed by Sanofi Pasteur, and 68 requires three injections over one extended year (0/6/12-month immunization schedule) to 69 induce a well-balanced antibody response against all four serotypes (6, 8). The overall pooled 70 vaccine efficacy for symptomatic dengue during the first 25 months post-dose 1 was 60.3% for 71 all participants (9). However, efficacy in children under 9 years was lower at 44.6% with 70.1% 72 efficacy in seropositive and 14.4% efficacy in seronegative children (9). The vaccine was 73 licensed only for persons 9 to 45 years of age in the dengue endemic countries. Furthermore, 74 interim results from long-term safety follow up studies demonstrated an increased risk for 75 hospitalization of vaccine-sensitized individuals (10), suggesting that the ADE-related concerns 76 are relevant. 77 Virus-like particle (VLP) vaccine is an alternative feasible approach to live attenuated 78 vaccines. VLPs are self-assembled particles consisting of viral structural proteins, which mimic 79 the conformation of the authentic native virus but lacks its genomic DNA or RNA (11, 12), and 80 are the basis of a number of safe, marketed vaccines against hepatitis B and human 81 papillomavirus (13). VLPs are highly immunogenic due to the resemblance of their morphology 82 to that of authentic viruses, and safe because they are non-infectious and do not pose a risk of 83 reversion to virulence. Another key advantage for using VLPs to develop a dengue vaccine 84 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom include a short vaccination schedule, which will reduce the risk of ADE-mediated severe dengue 85 cases. Furthermore, a multivalent VLP-based dengue vaccine is expected to elicit balanced 86 antibody responses against all four DENV serotypes as there is no concern for replication 87 interference, which is a key issue for a live attenuated dengue vaccine (14). 88 It has been reported that the co-expression of flavivirus precursor membrane (prM) and 89 envelope (E) produces VLPs, with several reports by various groups on DENV VLP production 90 in insect cell, yeast and mammalian cell, however, DENV VLP production level was relatively 91 low (15-21). For our study, we created novel DENV VLPs for all four serotypes and produced 92 them at high yields by introducing mutations into the E protein. Tetravalent vaccination with 93 DENV VLPs elicited high titers of neutralizing antibody (NAb) against all four serotypes 94 simultaneously in mice. Our tetravalent DENV VLP thus bears the potential of serving as a next 95 generation dengue vaccine, and as a template for other flavivirus vaccine development strategies. 96 97 Results 98 Amino acid mutation in E produces DENV1 VLP at high yield. 99 The DENV genome encodes three structural proteins (capsid; C, prM and E) that form the virus 100 coat, and seven non-structural proteins that take part in virus replication within host cells. The 101 complex of prM and E plays important roles in virus assembly and fusion to the host cell 102 membrane. The E protein lies on the surface of the dengue virion and plays a direct role in cell 103 entry, thus recognized as a target for dengue vaccine development (22). We created an 104 expression plasmid DENV1 prM-E, containing the hydrophobic signal sequence (ss, located 105 between C and prM), prM, and E genes of DENV1 (Fig. 1A). When transfected into 293F cells, 106 DENV1 prM-E VLP was secreted into the culture supernatant. However, the amount of VLP in 107 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom the supernatant was very low as measured by western blotting using anti-DENV1-4 E 108 monoclonal antibody (Fig. 1B, middle lane). 109 To overcome the low yield, we focused on modifying DENV E for the following reasons: 110 DENV E protein undergoes structural conformational changes when the dengue virion fuses with 111 the host cell membrane. In a previously reported work on alphavirus VLP vaccine development, 112 it was shown that mutating E2 protein in the region of the conformational change greatly 113 increased VLP yield (23). Since flaviviruses and alphaviruses both use mechanisms of class II 114 fusion proteins for viral entry (24), we hypothesized that a modification to the amino acid(s) 115 which impacts E protein conformational change may also increase DENV VLP yield. We 116 selected 27 amino acids that could potentially impact the conformational change of the DENV1 117 E protein, introduced a single amino acid mutation for each residue, and assessed VLP 118 production levels in the culture supernatant (Table 1). As a result, we found that the substitution 119 of phenylalanine 108 to alanine (F108A) greatly increased the DENV1 VLP production. The 120 production of DENV1 prM-EF108A VLP (hereafter called DENV1 VLP) was about 16-fold higher 121 than wild-type DENV1 prM-E (Fig. 1B). 122 123 Development of DENV2, DENV3, DENV4, and ZIKV VLPs. 124 F108 is located in the fusion loop structure of DENV E protein, which consists of hydrophobic 125 amino acids and is responsible for the insertion of DENV E protein into the host cell membrane 126 (22). As this sequence is highly conserved among flaviviruses, we expected that the F108A 127 mutation would also increase the production of other serotypes of DENV VLPs. Hence, we first 128 constructed expression plasmids encoding DENV2 prM-E with or without the F108A mutation 129 (Fig. 2A, DENV2 prM-EF108A and prM-E). However, when transfected into 293F cells, DENV2 130 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom prM-E or prM-EF108A did not secrete VLPs into the culture supernatant as measured by western 131 blotting using anti-DENV2 E polyclonal antibody (Fig. 2B). DENV2 E protein is 495 amino 132 acids (aa) in length and has structurally distinct domains: envelope domain (ED) I, EDII, EDIII, 133 and stem and transmembrane anchor (ST/TM) region (22). It has been thought that ST/TM 134 contain a membrane retention signal which might contribute to the inefficient secretion of prM-E 135 VLP (25). To facilitate DENV2 VLP secretion, we next created a chimeric DENV2 construct by 136 replacing the C-terminal region of DENV2 E protein with the corresponding region of DENV1 E 137 protein. We prepared several chimeric DENV2 E proteins of different length C-terminal regions, 138 and assessed their VLP production in 293F cells. We found that replacing aa 297-495 region of 139 DENV2 E protein, which corresponds whole EDIII and ST/TM region with DENV1 E protein aa 140 297-495 region enabled efficient DENV2 VLP production (Fig. 2C, prM-EF108A_2/1). The 141 amount of secreted VLP with F108A mutation (prM-EF108A_2/1) was about 15-fold higher than 142 that of VLP without F108 mutation (prM-E_2/1) (Fig. 2B). Next, we used the same approach to 143 produce DENV3 and DENV4 VLPs. Similar to DENV2, the expression plasmids encoding 144 DENV3 and DENV4 prM-E or prM-EF108A did not secrete VLPs in the culture supernatants (Figs. 145 3A and 3B). In DENV3, VLP production was greatly increased when the EDIII and ST/TM of 146 DENV3 E protein was replaced with aa 297-495 of DENV1 E protein (prM-E_3/1), and it was 147 further enhanced by introducing the F108A mutation (Fig. 3A, prM-EF108A_3/1). Similar to 148 DENV2 VLP result, the production of DENV4 VLP with F108A mutation and EDIII and ST/TM 149 replacement (prM-EF108A_4/1) was about 14-fold higher than that of the VLP without F108 150 mutation (prM-E 4/1) (Fig. 3B). 151 We also tested this strategy to develop Zika virus (ZIKV) VLP. ZIKV is another 152 mosquito-borne flavivirus responsible for the recent outbreaks in South and Central America 153 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom (26), and its structure is similar to that of other flaviviruses including DENV (27). While the 154 expression of wild type ZIKV prM-E did not produce VLP well, prM-E with the F108A 155 mutation in E protein enabled ZIKV VLP production (Fig. 3C, prM-E and prM-EF108A). Unlike 156 DENV2-4, replacement of ZIKV EDIII and ST/TM with that of DENV1 did not produce VLPs 157 even with the F108A mutation (data not shown). We found that replacement of the ZIKV ST/TM 158 (aa 404-504) with the Japanese encephalitis virus (JEV) ST/TM in combination with F108A 159 mutation further increased ZIKV VLP production levels (Fig. 3C, prM-EF108A_ZIKV/JEV). 160 These data demonstrate that distinct combinations of manipulations to C-terminal region of the E 161 protein (replacement of ST/TM plus EDIII or ST/TM only) and the introduction of F108A 162 mutation may be applied to increase flavivirus VLP production. 163 164 Production and characterization of DENV1-4 VLPs. 165 The production levels of DENV2 and 4 VLP were relatively lower than that of DENV3 VLP. 166 The original DENV2 expression constructs tested in Fig. 2 was derived from the S1 vaccine 167 strain (28). To improve the yield of DENV2 VLP, we prepared another prM-EF108A_2/1 168 expression construct using a different DENV2 strain (D2/TO/UH04/1974) from American 169 genotype (29), and compared the VLP production level to the S1 strain. We ultimately selected 170 the American strain for the following studies based on its higher VLP yield compared to the S1 171 strain (1.8-fold, Fig. 4A). Next, to improve the yield of DENV4 VLP, we introduced an amino 172 acid mutation in the furin recognition site in prM. As DENV3 VLP constructs showed better 173 VLP production than the DENV2 VLP construct, we were prompted to match the DENV4 furin 174 recognition sequence to that of DENV3. As the substitution of the glutamic acid 90 by aspartic 175 acid (E90D) slightly improved DENV4 VLP expression (1.2-fold, Fig. 4B), we selected this 176 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom construct for the following studies. A summary of the DENV1-4 VLP constructs used for the 177 characterization and immunization studies are shown in Table 2. 178 The optimized DENV1-4 VLPs were expressed in 293F cells and purified from culture 179 supernatants by utilizing a combination of an anion exchange column and a ceramic 180 hydroxyapatite column chromatography. The purified DENV1-4 VLP samples contained E (53 181 kDa) and prM (around 20 kDa) as major proteins when they were assessed by SDS-PAGE and 182 Coomassie blue staining (Fig. 5A). The VLP yield after the purification steps determined by 183 Bradford protein assay was up to 3 mg/L for DENV3 VLP, and up to 1.5 mg/L for DENV1, 2, 184 and 4 VLPs. Transmission electron microscopy (EM) images confirmed that DENV1-4 VLPs 185 exhibited electron-dense 35-50 nm spherical particles (Fig. 5B), which are similar to the 186 previously described DENV VLPs in size (16, 20, 30), but smaller than the dengue virions that 187 have a diameter of around 50 nm (31). 188 189 Immunogenicity and neutralizing antibody responses in mice induced by monovalent VLP 190 vaccines. 191 To evaluate immune responses of DENV VLPs, BALB/c mice were immunized with 20 μg of 192 monovalent DENV VLPs of each serotype or 20 μg of recombinant E proteins (rE, aa 46-413) 193 with aluminum hydroxide (Alum) adjuvant intramuscularly (i.m.) three times at 3-week intervals. 194 Two weeks after the last immunization, serum anti-flavivirus IgG titers were measured by 195 ELISA using purified JEV antigen (32). In comparison to the PBS control group, mice 196 immunized with rEs elicited anti-flavivirus antibodies. VLPs induced significantly higher anti197 flavivirus IgG titers than rEs (Fig. 6A). 198 NAb titers against DENV1-4 were determined by 50% focus reduction neutralization test 199 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom (FRNT50) (33). NAb titers induced by DENV rEs were below the detection limit except in the 200 case of mice immunized with DENV2 rE. On the other hand, DENV VLPs induced high NAb 201 responses against their homologous DENV serotype (Figs. 6B-E). The NAbs induced by DENV 202 VLPs were serotype cross-reactive, but no cross-reactivity to JEV was detected. Geomean 203 FRNT50 titers against homologous DENV in DENV VLP-immunized mice were 3,965 (DENV1), 204 2,071 (DENV2), 4,267 (DENV3), and 1,693 (DENV4). 205 206 Immunogenicity and neutralizing antibody responses in mice induced by tetravalent VLP 207 vaccines. 208 We next evaluated the immunogenicity of DENV VLPs in a tetravalent formulation in mice. 209 BALB/c mice were immunized with PBS or 80 μg of tetravalent DENV1-4 VLPs (20 μg per 210 serotype) with Alum adjuvant three times at three-week intervals. We also compared the 211 immunogenicity of DNA vaccine to VLP vaccine by immunizing mice with plasmid DNA 212 encoding wild type DENV prM-E (DNA_wild type) or encoding our DENV VLPs, which 213 contain the F108A mutation and/or chimeric E (DNA_F108A). The 80 μg of tetravalent DNA 214 plasmids (20 μg per serotype) were administered by i.m. injection with electroporation as 215 described previously (34). The tetravalent DNA vaccine encoding DENV VLPs (DNA_F108A) 216 induced higher immune responses compared to the DNA vaccine encoding wild type prM-E 217 (DNA_wild type). The mice immunized with tetravalent VLPs demonstrated the highest anti218 flavivirus IgG titers among the three groups (Fig. 7A). 219 Moreover, tetravalent VLPs elicited highly potent NAb response (Fig. 7B). Geomean 220 FRNT50 titers against the four DENV serotypes of tetravalent VLP-immunized mice were 7,479 221 (DENV1), 2,131 (DENV2), 3,202 (DENV3), and 869 (DENV4). NAb response was not detected 222 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom in DNA_wild type group (FRNT50 titers were below the detection limit of <80). DNA_F108A 223 group induced substantial NAb responses, however, FRNT50 titer was significantly lower, less 224 than 11% of those elicited in the VLP-immunized group. Geomean FRNT50 titers against the 225 four DENV serotypes in DNA_F108A immunized mice were 656 (DENV1), 227 (DENV2), 107 226 (DENV3), and below the detection limit of <80 (DENV4). These data suggest that DENV VLP 227 vaccine is a more effective vaccine form than the DNA vaccine format in the induction of potent 228 NAbs. 229 Next, we tested the infection-enhancing capacity of antibodies elicited by DENV VLP 230 immunization. Pre-existing, non-neutralizing anti-DENV antibody can enhance viral entry into 231 host cells by forming DENV-antibody complex (35). This phenomenon is known as ADE. To 232 assess ADE activity, we utilized the Fc gamma receptor (FcγR)-expressing BHK cell system (36). 233 Sera from mice immunized with DENV DNA vaccine, both in DNA_wild type and 234 DNA_F108A, enhanced DENV infection by 2to 3-fold at 1:10 serum dilution, demonstrating 235 that the induced antibodies possess ADE-activity to DENV (Fig. 8). In contrast, we did not 236 observe infection-enhancement for all four serotypes in VLP-immunized mice sera at 1:10 serum 237 dilution. When the sera from VLP-immunized mice were diluted more than at 1:100, infection238 enhancement was observed as expected, although no enhancement was observed against DENV1 239 in the dilutions we tested. 240 241 NAb responses to DENV3 and DENV4 VLP in outbred mice. 242 Lastly, to evaluate the immunogenicity of DENV VLPs in a representative, genetically diverse 243 population, we immunized NIH Swiss outbred mice with DENV3 and DENV4 VLPs. We also 244 tested different doses to assess for a dose-response relationship between NAb titer and VLP dose. 245 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom Mice were immunized with monovalent DENV3 or DENV4 VLP at different doses as indicated 246 in the Fig.9 with Alum three times at 3-week intervals. FRNT50 assay using serum collected two 247 weeks post last immunization demonstrated that both VLPs elicited potent NAb response in the 248 outbred mice (Fig. 9). FRNT50 elicited by 20 μg of DENV3 and DENV4 VLP in NIH Swiss 249 mice were 5,106 and 751, respectively, which was similar to the FRNT50 in BALB/c mice 250 immunized with 20 μg of VLP. Potent NAb responses were observed at all doses tested. These 251 results suggest that our VLP vaccine could be effective at doses as low as 1.25 μg per each 252 serotype (Geomean FRNT50, 4,262 (DENV3), 1,541 (DENV4)). Taken together, these results 253 demonstrated that our DENV1-4 VLPs were highly immunogenic and our tetravalent DENV 254 VLP vaccine elicited potent NAb responses against all four DENV serotypes simultaneously. 255 256 Discussion 257 The global spread and persistence of DENV has made dengue vaccine development a pressing 258 matter. In this study, we found an effective strategy to produce DENV VLPs and developed a 259 novel tetravalent dengue VLP vaccine which induces strong neutralizing responses against all 260 four DENV serotypes. 261 The introduction of an F108A mutation in E protein increased VLP production in all four 262 serotypes of DENV as well as ZIKV. Amino acid residue F108 is located in the fusion loop of E 263 which is highly conserved among flaviviruses (37). During viral infection, the E protein 264 undergoes conformational changes that exposes the hydrophobic fusion loop on the viral surface. 265 The fusion loop is inserted into the host cell membrane and induces fusion of the viral and host 266 cell membranes that results in the release of the viral genome into the cytosol (22). It has been 267 reported that flavivirus VLPs composed of prM-E also undergo a similar process to fuse with the 268 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom cell membranes (38). This process may affect the viability of the cells producing VLPs and the 269 yield of VLP production. F108A mutation may prevent VLPs from fusing to 293F cells during 270 the production process, thereby leading to the high-level VLP production observed. 271 For DENV2-4, in addition to the F108A mutation, the replacement of C-terminal region 272 with that of the corresponding region of DENV1 was necessary to achieve efficient VLP 273 production (Figs. 2 and 3). ST/TM region is known to contain a membrane retention sequence, 274 which might prevent VLP budding from the plasma membrane (25). Replacing aa 297-495; 275 EDIII and ST/TM from DENV2-4 with that of DENV1 might impact the membrane retention 276 sequence in a manner that enhances the secretion of these VLPs. Importantly, based on the 277 dengue E dimer structure, F108 is in close contact with EDIII (39). The contact of F108A and 278 DENV1 EDIII in the DNEV2-4 VLP might also affect the structural stability in the context of 279 dengue E dimer structure form, which could increase the yield of VLPs. Further studies are 280 needed to support these hypotheses. 281 Although the EDIII of our DENV2-4 VLPs were replaced with DENV1 EDIII, they 282 induced serotype-specific NAbs in mice (Fig. 6), suggesting that there are neutralizing epitopes 283 located in EDI, EDII or EDI/EDII hinge region. These regions have been recognized as potent 284 targets of NAbs based on several mapping studies using mouse monoclonal antibodies that 285 showed potent neutralizing activities (40). Interestingly, recent studies showed that EDIII286 specific antibodies alone are unlikely to account for the strong NAb responses observed in 287 people naturally infected with DENV (41-43). In addition, potent DENV NAbs that bind to 288 epitopes around EDI/EDII were discovered in DENV-immune humans (44, 45). Our results 289 support the observations that EDI/EDII-targeting antibodies contribute to the serotype-specific 290 NAb responses. 291 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom When mice were immunized with the plasmid DNAs encoding our DENV VLPs (prM-E 292 with F108A mutation and C-terminal region replacement), higher NAb responses were observed 293 compared to the mice immunized with DNAs encoding wild-type prM-E (Fig. 7B). This could be 294 reflective of the different VLP production levels in vivo. Immunization with DNA induces host 295 cells to produce the encoding proteins, resulting in specific immune activation against them (46). 296 Indeed, when mice were immunized with DENV DNA vaccines, there was a significant 297 correlation between the level of in vitro VLP secretion by the DNA vaccine construct and the 298 elicited immune responses (29). Compared to the DNA vaccines or recombinant E proteins with 299 our VLP vaccine, we observed that mice immunized with VLP vaccine induced much higher 300 NAb responses (Figs 6B-D and 7B). VLP-based vaccines have been successful in protecting 301 humans from various viruses including Hepatitis B virus and Human Papillomavirus (13). 302 Because of their good safety profiles and strong immunogenicity, many other VLP-based 303 vaccine candidates such as those against CHIKV, influenza virus or norovirus are in clinical 304 trials or under preclinical studies (47, 48). Although there is no commercialized VLP-based 305 vaccine against flavivirus to date, previous studies have demonstrated that co-expression of prM 306 and E proteins could produce VLPs of West Nile virus, St. Louis encephalitis virus and JEV (49, 307 50). Recently, several reports showed that DENV VLPs could induce effective NAb responses in 308 mice when immunized as a monovalent vaccine (51, 52). However, tetravalent VLP vaccine 309 development has not been very successful. It has been reported that a tetravalent DENV VLP 310 vaccine candidate produced in Pichia pastoris was immunogenic in mice, but the NAb titers 311 were somewhat low (1:16 to 1:32 by Plaque reduction neutralization test (PRNT50), when 312 immunized with 25 μg per serotype) (53). In our study, tetravalent DENV VLP induced 1:869 to 313 1:7,479 by FRNT50, suggesting our tetravalent VLP vaccine can achieve more potent NAb 314 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom responses compared to existing candidates. 315 While high levels of neutralizing antibodies were detected in the VLP immunized group 316 (Fig. 7), ADE activity was absent at the lowest feasible dilution (1:10, Fig. 8). Strong 317 neutralizing activity in a conventional plaque reduction neutralization assay is associated with 318 absence of ADE when tested at the lowest feasible dilution (54). It has also been reported that 319 infection-enhancement activity is absent at low serum dilutions in convalescent patient samples 320 that possess high levels of neutralizing activity (55). Since immune enhanced viral replication 321 did not occur at the low dilution (1:10) of sera from the VLP immunized group but only occurred 322 at higher dilutions, it is unlikely that immune enhancement would occur in an in vivo condition, 323 e.g. with undiluted serum, although further precise preclinical and clinical studies regarding 324 ADE effect are needed. 325 Viral interference is a phenomenon observed in the tetravalent live attenuated dengue 326 vaccines. One or some replication dominant serotypes interfere with others, resulting in 327 unbalanced immune response (14). Tetravalent dengue VLP vaccine is highly unlikely to show 328 viral interference, because they do not replicate. While the live attenuated dengue vaccine, CYD329 TDV requires a one-year immunization schedule, our DENV VLP vaccine immunizations can be 330 completed within about two months, which should greatly reduce the risk of ADE-mediated 331 severe dengue cases. VLPs can be manufactured efficiently under current regulatory standards, 332 and the sequence can be matched easily to circulating virus strains or emerging new variants if 333 needed. In addition to these advantages, the high level of NAb response combined with high 334 levels of VLP production make our tetravalent VLP a promising next-generation dengue vaccine 335 candidate. Furthermore, we demonstrated that the F108A mutation and replacement of the C336 terminal region of E similarly enabled ZIKV VLP production in 293F cells. This strategy will 337 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom not only facilitate vaccine development against DENV but also against other flaviviruses such as 338 ZIKV or West Nile virus, due to the similarity across flavivirus conserved structures. 339 340 Materials and Methods 341 Vector construction 342 Plasmids encoding structure proteins of DENV, ZIKV and JEV were generated by gene 343 synthesis (GeneArt, Thermo Fisher Scientific). The signal sequence, prM, and E genes were 344 cloned into pUC119 based expression vector. The strains used are; DENV1 (Western Pacific 345 strain, Genbank# U88535.1), DENV2 (S1 vaccine strain, Genebank# M19197.1 and American 346 strain, Genebank# AY744147), DENV3 (Singapore 8120/95 strain, Genebank# AY766104.1), 347 DENV4 (ThD4_0476_97 strain, Genebank# AY618988.1), ZIKV (Genebank# KU312312) and 348 JEV (nakayama strain, Genebank# EF571853). Mutagenesis was performed by QuikChange 349 Lightning Site-Directed Mutagenesis Kit following the manufacturer’s instructions (Agilent 350 technologies). Chimeric constructs of DENV2, DENV3, DENV4 E with DENV1 E were 351 produced by standard overwrap extension PCR method. Briefly, ss, prM, and aa 1-296 region of 352 E from DENV2, DENV3 or DENV4 were amplified with a universal CMV-forward primer and 353 reverse primers that have connective sequences with DENV1 E at their 3’ ends. The aa 297-495 354 region of DENV1 E was amplified using forward primer that have same connective sequences at 355 5’ end and a reverse primer that recognizes poly-A region. These PCR products were mixed, 356 amplified, and cloned into NotI and BglII sites of pUC59 vector. For chimeric construct of ZIKV 357 and JEV, ZIKV ss, prM, E aa 1-403 and JEV E aa 400-500 were connected as same method as 358 above. 359 360 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom Production and purification of VLPs 361 The VLPs were produced in Freestyle 293F cells (Thermo Fisher Scientific). The cells were 362 cultured in suspension in serum-free FreeStyle 293 Expression Medium (Thermo Fisher 363 Scientific) and transfected with VLP-expressing plasmids by using GeneX Plus transfection 364 reagent (ATCC) according to the manufacturer’s instructions. Four days after transfection, the 365 cell culture supernatant was harvested, then clarified by centrifugation and filtration with 0.45 366 μm polyethersulfone (PES) membrane. The culture supernatant was concentrated, and buffer367 exchanged to 2.5mM Sodium-Phosphate Buffer (Teknova) by KrosFlo Research II TFF Systems 368 with mPES MidiKros Filter Modules (Spectrum Laboratories). Further purification was 369 performed by NGC Quest 10 Chromatography Systems (Bio-Rad) with tandemly connected 370 HiTrap Q XL (GE Healthcare Life Sciences) and Foresight CHT Type II (Bio-Rad) columns. 371 The VLPs in the flow-through from Q XL column were captured by CHT Type II column and 372 eluted with 2.5 to 400 mM sodium phosphate gradient. The eluates containing VLPs were 373 concentrated by Amicon Ultra-15 centrifugal filter units (EMD Millipore) and filtered with 0.20 374 μm PES membrane. Total protein concentration of purified DENV VLP was measured by Quick 375 Start Bradford Protein Assay (Bio-Rad). Purity of the DENV VLP was assessed by SDS-PAGE 376 followed by Coomassie dye-based staining using InstantBlue Stain reagent (Expedeon). 377 378 Western blotting 379 VLP-containing culture supernatant was separated by Any kD Mini-PROTEAN TGX Precast 380 Protein Gels (Bio-Rad) electrophoresis, and the proteins were transferred onto nitrocellulose 381 membranes using Trans-Blot Turbo Transfer System (Bio-Rad). Membranes were blocked with 382 5% skim milk (Labscientific, Inc.) in tris-buffered saline containing 0.05% Tween 20 (TBS-T) 383 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom and treated with mouse anti-DENV1-4 E monoclonal antibody (clone 9F10, Santa Cruz 384 Biotechnology, 1:500), goat anti-DENV2 E polyclonal antibody (Santa Cruz Biotechnology, 385 1:1000), mouse anti-DENV4 E monoclonal antibody (clone 1H10-6, ATCC, 20 μg /mL), or anti386 Zika rabbit polyclonal antibody (IBT Bioservices, 0.5 μg/mL). Membranes were washed in TBS387 T, incubated with peroxidase (HRP)-conjugated antimouse, goat, or rabbit secondary antibody 388 (Santa Cruz Biotechnology, 1:5000), and developed using Clarity ECL Western Blot Substrate 389 (Bio-Rad). Imaging and data analysis were done by Image Lab software (ver. 5.2.1, Bio-Rad) 390 391 Electron microscopy analysis 392 The morphologies of DENV VLPs were analyzed at the Johns Hopkins School of Medicine 393 Microscope Facility. Briefly, the purified DENV VLPs were fixed in 4% formaldehyde and 394 placed on glow-discharged carbon-coated 200 Mesh copper grids. The grids were then stained 395 with 1% phosphotungstic acid and visualized by Philips CM120 Transmission Electron 396 Microscopy at 80 kV with AMT XR80 8-megapixel camera. 397 398 Mouse experiment 399 Immunization and serum sample preparation were conducted at Bioqual, Inc (Rockville, MD). 400 All animal experiments were conducted under Institutional Animal Care and Use Committee401 approved, and Office of Laboratory Animal Welfare-assured conditions. Female BALB/c mice 402 were purchased from Harlan (Frederick, MD). PBS, rE (Prospec, den-021, den-022, den-023, 403 and den-024) and VLP were administered intramuscularly with Alum at weeks 0, 3 and 6. DNAs 404 were immunized by intramuscularly injection followed by electroporation with a BTX 2 needle 405 array and a BTX model 830 electroporation generator (Harvard Instruments) using the following 406 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom parameters: Six pulses of 100 V with 50 ms durations and 200 ms between pulses. Blood 407 samples were taken at weeks 0, 5 and 8. Serum was prepared and heat-inactivated by incubating 408 at 60°C for 30 min. 409 410 Anti-Flavivirus IgG ELISA 411 Serum anti-flavivirus IgG titer was measured by enzyme-linked immunosorbent assay (ELISA). 412 A 96-well ELISA plate was coated with purified Japanese encephalitis virus (JEV, strain: 413 JaOArS982) at 250 ng/well at 4 °C overnight (32). The plate was blocked with Block Ace (DS 414 phama Biomedical) for 1 h at room temperature and washed with PBS-0.05% Tween-20 (PBS-T). 415 The test sera and standard control were diluted at 1:1,000 and were added to wells, incubated for 416 1 h at 37°C, and washed. 1:1,000 diluted HRP conjugated anti-mouse IgG antibody (American 417 Qualex) was added and incubated for 1 h at 37°C. Wells were washed and developed with o418 phenylenediamine dihydrochloride (OPD, Sigma) with 0.05M citrate phosphate buffer and 419 0.03% H2O2 solution by incubating for 30 min in the dark. Reaction was stopped by adding 1N 420 HCl and read OD at 492 nm. Titer of sample serum IgG titer was calculated from standard curve 421 (33). A sample titer ≥ 3,000 were interpreted as anti-flavivirus IgG positive. 422 423 Fifty Percent Focus Reduction Neutralization test (FRNT50) 424 Vaccine-immunized mice sera were serially diluted and mixed with 40 to 50 focus-forming units 425 of virus (DENV1 99st12A strain -genotype IV, DENV2 oost22A strain -Asian 2, DENV3426 SLMC50 strain -genotype I, DENV4-SLMC318 straingenotype I and JEV S-982 strain427 genotype III) and incubated at 37°C for 1 h. Serum-virus mixture was inoculated to Vero cell 428 monolayer in 96-well plate and incubated at 37°C for 1 h, then 1.25% methylcellulose 4000 in 429 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom 2% FCS MEM was added to wells and incubated at 37°C for 3 days for DENV and 36 h for JEV. 430 The plates were washed with phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde 431 solution for 30 min and washed. Then the cells were permeabilized with 1% NP-40 solution for 432 30 min and washed. The plates were blocked with BlockAce for 30 min and treated with 1:1,500 433 diluted pooled human sera having high anti-flavivirus IgG titers (33) for 1 h at 37°C. 434 Subsequently, HRP-conjugated goat anti-human IgG (American Qualex, 1:1,500) was added and 435 incubated at 37°C for 1 h. 0.5 mg/ml of 3,3'-diaminobenzidine tetrahydrochloride (Wako) with 436 0.03% of H2O2 solution was added and incubated for 10 min for staining. After washing and air 437 drying, the number of foci per well were counted using a biological microscope. The reciprocal 438 of the endpoint serum dilution that provided 50% or greater reduction in the mean number of foci 439 relative to the control wells that contained no serum was considered to be the FRNT50 titer. 440 441 Antibody-dependent dengue virus infection enhancement assay 442 FcγR-expressing BHK cell lines (36) were seeded in to a 96 well plate and cultured in EMEM 443 supplemented with 10 % heat-inactivated FBS and G418. Vaccine immunized mouse serum or 444 mouse anti-Flavivirus E monoclonal antibody 4G2 were serially diluted from 1:10 to 1:10,000 445 with 10% FBS /EMEM, and mixed with 30-50 focus-forming units of virus (same DENV strains 446 as those used in FRNT assay) and incubated at 37°C for 1 h. Virus-immune complex were added 447 to each well of the cells, and cultured for 48 h. The cells were washed with PBS once, dried by 448 air, and fixed with ice-cold methanol-acetone (1:1). After the plates are air-dried, the wells were 449 blocked with 1% horse serum in PBS for 10 min, and washed with PBS for three times. The plate 450 was then treated with anti-flavivirus antibody (mAb 4G2 at a dilution of 1:1000) at 37°C for 30 451 min, biotin conjugated anti-mouse IgG antibody (Vector Laboratories, 1:500) at room 452 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom temperature for 30 min, and ABC (Vector Laboratories) solution at room temperature for 30 min. 453 The plate was washed with PBS three times after each incubation step. VIP solution (Vector 454 Laboratories) was added and incubated at room temperature for few minutes until color was 455 developed. The plate was washed with PBS once, infected cells were quantitated (Keyence BZ456 X710 microscope). Fold-enhancement values were calculated using the following formula: 457 (mean infected cells count using FcγR-expressing BHK cells with the addition of mouse serum 458 sample)/ (mean plaque count using FcγR-expressing BHK cells in the absence of test sample). 459 Infection enhancement (measured as ADE activity) was tested using serum samples that was 460 diluted from 1:10 to 1:10,000. The fold enhancement values were determined as follows: (the 461 mean number of DENV infected-cells in wells treated with serum samples)/ (the mean number of 462 plaques in wells with monolayers incubated in the absence of test samples). The mean value of at 463 least 3 negative control wells plus three times the standard deviation (SD) value was used as the 464 cut-off value to determine which samples had ADE activity (56). 465 466 Acknowledgements 467 The authors thank K. Tolliver, M. Nakata, J. Sastri, E. Cho-Fertikh and G. Moonsammy (VLP 468 Therapeutics) for facilitation of collaborations, manuscript preparation and helpful discussions, 469 H. Anderson, S. Kar (Bioqual) and S. Cherukuri (Noble Life Sciences) for managing animal 470 experiments, B. Smith (Johns Hopkins University) for EM analysis. A.U., M.I., M.M.N.T., 471 M.L.M, A.S., K.M., and W.A. performed the research; A.U., M.M.N.T., M.L.M., A.S., M.I., 472 K.M. S.K., R.U., K.M, and W.A. analyzed data; A.U., M.M.N.T., M.L.M., A.S., K.M. and W.A. 473 wrote the paper and all authors participated in manuscript revisions. R.U., and W.A conceived, 474 directed, and supervised the studies. 475 on O cber 6, 2017 by gest http/jvi.asm .rg/ D ow nladed fom