An Inactivated Cell Culture Japanese Encephalitis Vaccine (JE-ADVAX) Formulated with Delta Inulin Adjuvant Provides Robust Heterologous Protection against West Nile Encephalitis via Cross-Protective Memory B Cells and Neutralizing Antibody

Abstract

West Nile virus (WNV), currently the cause of a serious U.S. epidemic, is a mosquito-borne flavivirus and member of the Japanese encephalitis (JE) serocomplex. There is currently no approved human WNV vaccine, and treatment options remain limited, resulting in significant mortality and morbidity from human infection. Given the availability of approved human JE vaccines, this study asked whether the JE-ADVAX vaccine, which contains an inactivated cell culture JE virus antigen formulated with Advax delta inulin adjuvant, could provide heterologous protection against WNV infection in wild-type and β2-microglobulin-deficient (β2m^−/−) murine models. Mice immunized twice or even once with JE-ADVAX were protected against lethal WNV challenge even when mice had low or absent serum cross-neutralizing WNV titers prior to challenge. Similarly, β2m^−/− mice immunized with JE-ADVAX were protected against lethal WNV challenge in the absence of CD8⁺ T cells and prechallenge WNV antibody titers. Protection against WNV could be adoptively transferred to naive mice by memory B cells from JE-ADVAX-immunized animals. Hence, in addition to increasing serum cross-neutralizing antibody titers, JE-ADVAX induced a memory B-cell population able to provide heterologous protection against WNV challenge. Heterologous protection was reduced when JE vaccine antigen was administered alone without Advax, confirming the importance of the adjuvant to induction of cross-protective immunity. In the absence of an approved human WNV vaccine, JE-ADVAX could provide an alternative approach for control of a major human WNV epidemic.

INTRODUCTION

West Nile virus (WNV) is a mosquito-borne flavivirus that is antigenically classified as a member of the Japanese encephalitis (JE) serocomplex, a group of neurotropic viruses that predominantly infects birds but can cause fatal encephalitis in humans and horses (1). The clinically most important virus belonging to the serocomplex is Japanese encephalitis virus (JEV), which is widely distributed in Asia and in recent decades has spread into India, Pakistan, and the Asia-Pacific region (2). WNV is present in Africa, Europe, the Middle East, Asia, Australia (subtype Kunjin), and the Americas. Clinical manifestations of WNV vary and may include fever, headache, severe muscle weakness, confusion, seizures, tremors, generalized paresis, hypertonia, and loss of coordination (3, 4). The virus first emerged in the United States in 1999, and it is estimated that from 1999 to 2010 over 3 million persons were infected with WNV in the United States; 25% of infections resulted in West Nile fever, and over 12,000 human cases of West Nile neuroinvasive disease were detected, with ∼10% of these resulting in fatality (5). The year 2012 saw a particularly severe WNV epidemic in the United States, with CDC reporting 5,674 total cases, including 2,873 with neuroinvasive disease and 286 deaths (www.cdc.gov/ncidod/dvbid/westnile/).

As demonstrated by animal studies, vaccination is an effective means for preventing WNV encephalitis, and the introduction of licensed veterinary vaccines has significantly reduced the incidence of equine disease (6), albeit at the cost of reducing the ability to use horses as sentinels of WNV spread (7). The first equine vaccine, introduced in 2002, comprised formalin-inactivated WNV adjuvanted with MetaStim adjuvant (West Nile Innovator; Pfizer) (8). Horses that received two doses and that were challenged 1 year postvaccination showed markedly reduced WNV viremia, affecting just 5% of immunized horses but 82% of controls (9). An alternative equine WNV vaccine is based on a live chimeric canary poxvirus vector carrying the WNV membrane (prM) and envelope (E) proteins (10–12). Yet another equine live chimeric WNV vaccine, made from insertion of prM and E genes into the yellow fever virus backbone (PreveNile/Intervet) (11, 13), was recalled in 2010 due to severe vaccine adverse events, including deaths (14), but was subsequently rereleased as an inactivated vaccine.

Unfortunately, there is still no approved WNV vaccine for humans, although several candidates are in early-stage clinical trials (reviewed in reference 6). This poses a significant problem when major human outbreaks of WNV occur, such as the recent 2012 U.S. epidemic (5), but also for WNV researchers, who desire protection against laboratory exposure. Research over the last 50 years has shown that infection with one JE serocomplex virus can provide protective immunity against heterologous viruses in the group, raising the possibility of cross-protective vaccination against antigenically related flaviviruses (15). Given the more advanced stage of development of human JEV vaccines, several of which are already approved, a key question is whether a JEV vaccine might confer cross-protection against WNV. The first approved human JEV vaccine (JE-VAX) was a mouse brain formalin-inactivated virus preparation developed in Japan in the 1960s (reviewed in reference 16). Prompted by the outbreak of WNV encephalitis in the United States, the cross-protective value of JE-VAX against WNV was tested in a small human vaccine trial but failed to induce cross-neutralizing WNV antibodies (17). However, it is still possible that newer-generation JEV vaccines under development, such as inactivated cell culture JEV vaccines containing novel adjuvants (18) or live attenuated vaccines (19, 20), could confer cross-protection against WNV (reviewed in reference 15).

Advax adjuvant is based on immunologically active delta inulin microparticles (21) and has proved successful in both animal studies and human vaccine trials (22). In preclinical models, Advax was shown to enhance the immunogenicity of a broad range of viral vaccines, including pandemic influenza virus (23), HIV (24), hepatitis B virus (25), African horse sickness virus (26), and JEV (18, 27). JE-ADVAX is a recently developed JEV vaccine based on the combination of Advax adjuvant with a Vero cell-derived, formalin-inactivated JEV antigen (28). JE-ADVAX was shown in a previously published study (18) to provide immunogenicity against JEV superior to that provided by either the older mouse brain JE-VAX vaccine, an inactivated cell culture JEV vaccine adjuvanted with aluminum hydroxide (JESPECT; Novartis) (29), or a live attenuated chimeric yellow fever virus-JEV vaccine recently licensed in Australia and Thailand (Chimerivax-JE; Sanofi-Pasteur). When administered to mice or horses, JE-ADVAX induced high titers of not only neutralizing antibody against JEV but also cross-neutralizing antibody against WNV and the related flavivirus Murray Valley encephalitis virus (MVEV) (18). In this study, we asked whether protection against heterologous WNV challenge could be achieved by a dual- or even single-dose JE-ADVAX immunization regimen. We secondarily sought to identify the immunological mechanisms underlying the heterologous protection. The results confirm that JE-ADVAX is able to provide heterologous protection against WNV, which it does via induction of a cross-protective memory B-cell population together with enhanced production of serum cross-neutralizing antibodies.

MATERIALS AND METHODS

Animals.

C57BL/6 (B6) and congenic β2-microglobulin-deficient (β2m^−/−) mice (30) were bred under specific-pathogen-free conditions and supplied by the Animal Breeding Facility at the John Curtin School of Medical Research, The Australian National University (ANU), Canberra, Australia. Alternatively, B6 mice were sourced from the Charles River Laboratories. Female mice >7 weeks of age were used in all experiments, unless indicated otherwise in the figure legends, and animals were randomized to treatment groups. All animal experiments in Australia and the United States were approved by and conducted in accordance with the guidelines of the ANU or The University of Queensland Animal Ethics Committees and by the Institutional Animal Care and Use Committee of Utah State University, respectively.

Virus.

WNV strain NY99 (31) was propagated in MA-104 or C6/36 cells, and MVEV prototype strain 1-51 was propagated in suckling mouse brain (32). Virus stocks were stored in single-use aliquots at −80°C and were titrated by plaque assay on Vero cells (33). Mice were challenged by subcutaneous (s.c.) injection into the footpad or the inguinal area, as indicated in the figure legends, of a defined dose of virus diluted in minimal essential medium (MEM).

Vaccines and adjuvants.

Inactivated Vero cell-grown JE vaccine (ccJE) (28) was a gift of the Kitasato Institute, Japan, and was supplied at a concentration of 40 μg/ml and further diluted in saline to obtain a 0.5- or 5-μg dose per mouse. Advax adjuvant was provided by Vaxine Pty. Ltd., Adelaide, Australia, as two formulations (Advax-1 and -2), both at a delta inulin concentration of 50 mg/ml, and diluted in saline for a final dose of 1 mg/mouse. Advax-2 differed from Advax-1 only in that it also contained a small amount of oligonucleotide (CpG7909, 10 μg/dose). Delivery of JE-ADVAX was by the intramuscular (i.m.) route in a 0.1-ml injection. Control equine WNV vaccine (West Nile Innovator) was purchased from Pfizer. The vaccine was supplied at the usual horse dose in 1 ml and was administered as 2 doses of 50 μl/mouse injected 3 weeks apart.

Infectious cell culture assay.

Virus titers in serum and tissues were assayed using the virus yield assay, where a specific volume of tissue homogenate was added to the first tube of a series of dilution tubes. Serial dilutions were made and added to Vero cells. Six days later the viral cytopathic effect was used to identify the endpoint of infection. Four replicates were used to calculate the infectious doses per ml or g of sample (34).

WNV neutralizing antibody assay.

Serial dilutions of heat-inactivated plasma (56°C for 30 min) were incubated with an equal volume of WNV overnight at 4°C. On the next day, 200 μl of virus per plasma sample, including a virus control, positive-control plasma, and a cell culture medium control, was added to monolayers of Vero cells in 12-well tissue culture plates. Cultures were incubated for 1 to 1.5 h at 37°C while rocking gently every 15 min. After aspiration of the medium, 1.5 ml of 1.7% methylcellulose overlay medium was added. The culture was incubated at 37°C for 5 days, after which the medium was carefully removed and 750 μl of crystal violet stain with formaldehyde was added for 15 to 20 min. Stain was aspirated, and the wells were washed with water. Plaques were microscopically counted, and the reciprocal of the plasma dilution yielding a 50% reduction in virus control plaque numbers is reported as the 50% plaque reduction neutralization test (PRNT₅₀) titer. The PRNT₅₀ assay detection limit was a titer of 10.

Adoptive transfer of immune B or CD4⁺ T cells.

B6 donor mice were sacrificed at 3 weeks after completion of a two-dose immunization schedule with ccJE (0.5 μg) plus Advax-1 delivered 2 weeks apart. Spleens were aseptically removed, and single-cell suspensions were prepared by pressing the tissue gently through a fine-metal-mesh tissue sieve. Erythrocytes were lysed by suspending the splenocyte pellet in 4.5 ml distilled water, followed immediately by the addition of 0.5 ml of 10× phosphate-buffered saline (PBS). Lysed cells were discarded after centrifugation at 400 × g for 5 min. B-cell and CD4⁺ T-cell purification was performed as previously described (35). For B-cell enrichment, isolated splenocytes were incubated with 1:3 dilutions of anti-CD4 (RL172) plus anti-CD8 (31 M) supernatants in MEM plus 5% fetal bovine serum (FBS) for 30 min at 4°C, followed by incubation with rabbit serum complement (Cedarlane Laboratories) for 30 min at 37°C. Cells were washed twice with PBS before transfer into recipient mice. The efficiency of depletion of CD4⁺ and CD8⁺ cells was >95%, as assessed by flow cytometry (35). For CD4⁺ T-cell enrichment, B-cell depletion was first performed by incubating splenocytes with anti-CD19 magnetic beads (Miltenyi Biotec), followed by loading of cells onto magnetic columns according to the supplier's instructions. Effluent from the columns was collected, and cells were pelleted by centrifugation at 400 × g for 5 min. The efficiency of B-cell depletion was >99%, as assessed by flow cytometry (35). For further enrichment of CD4⁺ T cells, B-cell-depleted splenocytes were incubated with a 1:3 dilution of anti-CD8 (31 M) supernatant in MEM plus 5% FBS for 30 min at 4°C, followed by incubation with rabbit serum complement (Cedarlane Laboratories) for 30 min at 37°C. Cells were washed twice with PBS before transfer into recipient mice. The efficiency of depletion of CD8⁺ cells was more than 95%. Purified B or CD4⁺ T cells (10⁷ cells suspended in 100 μl PBS/mouse) were adoptively transferred by injection into the lateral tail vein of naive B6 recipient mice. B-cell transfer into 4-week-old mice was by the intraperitoneal (i.p.) route; B cells injected i.p. have been shown to distribute evenly in the recipient's lymphoid organs and with similar percentages as intravenously injected cells (36). Recipient mice were challenged a day later with 1 × 10⁴ PFU WNV or 1 × 10⁵ PFU MVEV via footpad injection.

Statistical analyses.

Mortality data were plotted into Kaplan-Meier curves and assessed for significance by the log-rank test. The Student t test was applied to assess differences between data gathered from two experimental groups. A P value of ≤0.05 was considered significant.

RESULTS

JE-ADVAX induces robust cross-protective immunity against virulent WNV.

Control mice receiving adjuvant alone or saline injection had a 70% mortality rate (14/20 mice) after challenge with WNV (Fig. 1A). In contrast, immunization of mice twice, 3 weeks apart, with ccJE (0.5 μg) plus either Advax-1 or Advax-2 provided complete cross-protection (mortality, 0/20 [0%]; P < 0.001), with two doses of ccJE without adjuvant providing only partial protection (mortality, 6/20 [30%]; P < 0.01; Fig. 1A). The amount of JEV antigen (0.5 μg) used for these two-dose regimens corresponded to ∼1/10 of the antigen content used in human two- or three-dose immunization schedules, which typically use 5 to 8 μg antigen per dose (29). Notably, the level of heterologous protection (100% survival) achieved with two doses of adjuvanted JE-ADVAX vaccine (Fig. 1A) equaled the protection (mortality, 0/20 [0%]) achieved with two doses of the homologous positive control, an equine inactivated adjuvanted WNV vaccine (West Nile Innovator; Pfizer).

Fig 1.

JE-ADVAX protection against heterologous WNV challenge. Groups of B6 mice (n = 20) were immunized with 2 doses of ccJE (0.5 μg) with or without Advax adjuvant or West Nile Innovator vaccine 3 weeks apart (A, C) or with a single dose of ccJE (5 μg) with or without Advax-1 or -2 (B, D). A negative-control group (n = 20) was immunized with saline. At 6 weeks postimmunization, mice were challenged with WNV (10⁵ PFU s.c.). Mice were monitored for morbidity and mortality and for body weight loss for 21 days. Asterisks denote significant differences relative to the saline group determined by the log-rank test (**, P < 0.01; ***, P < 0.001).

To test whether a single dose of JE-ADVAX vaccine was still able to provide heterologous protection against WNV, mice were given a single immunization with a 10-fold higher dose of ccJE (5 μg) alone or with Advax adjuvant and challenged 6 weeks later. The dose used was in accordance with past experience that a higher antigen dose is required when attempting to achieve protective immunity with a single-dose regimen. A single dose of 5 μg ccJE in saline did not significantly improve protection against WNV (mortality, 11/20 [55%]; P = 0.3) over mock immunization (mortality, 15/20 [75%]; Fig. 1B). In contrast, a single dose of adjuvanted JE-ADVAX with Advax-1 (JE-ADVAX-1) (mortality, 1/20 [5%]; P < 0.0001) or JE-ADVAX with Advax-2 (JE-ADVAX-2) (mortality, 4/20 [20%]; P = 0.001; Fig. 1B) conferred significantly improved survival compared to mock immunization.

Group survival rates were closely reflected by the severity of clinical illness, as represented by body weight loss and disease signs (ruffled fur, hunched back, paralysis). Consistent with the rates of survival in each group, animals immunized with two 0.5-μg doses of adjuvanted JE-ADVAX vaccine or the control West Nile Innovator vaccine showed essentially no weight loss postchallenge (Fig. 1C), those immunized with a single 5-μg dose of JE-ADVAX-1 or JE-ADVAX-2 showed minimal weight loss (<2.5%) (Fig. 1D), those immunized with a single 5-μg dose of ccJE in saline lost ∼10% of their body weight, and mock-immunized mice lost 20% or more of their body weight (Fig. 1C).

Reduced viral burden in mice immunized with JE-ADVAX.

Plasma WNV titers in the mouse WNV model typically peak at about 3 days postinfection (37). Plasma was therefore obtained at day 3 postchallenge for measurement of WNV titers by cell culture endpoint assay. Whereas sham-immunized mice had high plasma WNV titers in the range of 10^5.3 to 10^6.5 infectious units (IU)/ml, mice that received two doses of ccJE (0.5 μg) with Advax-1 or -2 or the control West Nile Innovator vaccine showed significantly lower or undetectable plasma WNV titers (Fig. 2A). There was a nonsignificant trend toward reduced plasma WNV titers in mice that received a single dose of ccJE with Advax-2 but not Advax-1. These data correlated with the magnitude of the preexposure cross-neutralizing antibody titers against WNV in the different groups (Table 1), consistent with inhibition of virus replication being mediated by vaccine-induced serum cross-neutralizing antibody.

Fig 2.

Viral burden in vaccinated mice challenged with WNV. Groups (n = 5) of B6 mice were immunized twice with ccJE (0.5 μg) with or without Advax adjuvant or with West Nile Innovator vaccine 3 weeks apart or with a single dose of ccJE (5 μg) with or without Advax adjuvant. A negative-control group was immunized with saline. Six weeks after completion of the vaccination schedule, mice were challenged with WNV (10⁵ PFU s.c.). (A) At 3 days postchallenge, mice were bled and the WNV content in plasma was determined. (B) Animals were euthanized at 6 days postchallenge to determine the virus content in brain. Each symbol represents an individual mouse, and horizontal lines indicate geometric mean titers. The lower detection limit of the assay is indicated by the dashed line. Asterisks represent significant differences relative to saline-treated negative-control mice determined by the Wilcoxon-Mann-Whitney rank-sum test (*, P < 0.05; **, P < 0.01). TCID₅₀, 50% tissue culture infective dose.

Table 1.

Prechallenge neutralizing antibody titers against WNV

Vaccinea	No. of vaccine doses	Pooled WNV PRNT50 titer
Saline	1	<10
ccJE (5 μg)	1	10
ccJE (5 μg) + Advax-1	1	20
ccJE (5 μg) + Advax-2	1	40
ccJE (0.5 μg)	2	20
ccJE (0.5 μg) + Advax-1	2	320
ccJE (0.5 μg) + Advax-2	2	320
West Nile Innovator	2	1,250

^aGroups of B6 mice were immunized with a single dose of ccJE (5 μg) with or without Advax adjuvant, two doses of ccJE (0.5 μg) with or without Advax adjuvant, or West Nile Innovator vaccine 3 weeks apart. A negative-control group was immunized with saline. Prechallenge sera taken 6 weeks postimmunization and 1 day prior to challenge were pooled (20 mice/group), and the WNV-neutralizing antibody titer was measured by PRNT₅₀ assay.

The major clinical disease manifestations of WNV infection are caused by virus spread into the central nervous system and the consequent encephalitis. Five animals per group were sacrificed on day 6 postchallenge for quantification of the WNV burden in the brain and for brain histology to be performed. Animals were not perfused prior to removal of the brain, and hence, a fraction of the virus measured in the brain homogenates on day 6 may reflect intravascular virus, although plasma WNV titers in this model typically peak at day 3 and have fallen by day 6 postinfection (37). Hence, the majority of virus measured at this time point could be anticipated to be within the brain parenchyma. Closely reflecting what was observed in the plasma, brain WNV titers were significantly reduced or absent in mice that received two doses of JE-ADVAX (0.5 μg) containing Advax-1 or -2 or that received the control West Nile Innovator vaccine (Fig. 2B). These data show that the JE-ADVAX 2-dose regimens significantly suppressed virus replication in the brain and extraneural tissues, thereby explaining the reduced severity of encephalitis and the absence of mortality in these groups. While there was only a nonsignificant trend in the reduction of brain WNV titers in mice that received a single dose of JE-ADVAX (5 μg) containing Advax-1 or -2 (Fig. 2B), both of these groups still showed marked reductions in mortality.

Histology, as summarized in Table 2, was performed on the brain stems of representative animals. Consistent with the pattern of clinical disease observed, no significant brain lesions were observed in animals that had received two doses of ccJE (0.5 μg) alone (Fig. 3C) or with Advax adjuvant (Fig. 3E), mild to moderate lymphocytic infiltration was found in some animals that received a single ccJE immunization alone (Fig. 3B) or ccJE with Advax adjuvant (Fig. 3D), and severe lymphocyte and neutrophil accumulation and multifocal neutrophilic meningitis were seen in two out of three infected mock-immunized control animals (Fig. 3A).

Table 2.

Histopathology of brain stem of selected groups 6 days after WNV challenge

Treatment group	Histological appearancea
Saline	Severe lymphocyte and neutrophil accumulation around multiple small vessels in the midbrain; moderate lymphocytic and multifocal neutrophilic meningitis in 2/3 animals; no abnormal lesions in 1/3 animals
ccJE (single dose of 5 μg antigen)	No abnormal lesions in 2/3 animals; moderate lymphocyte and neutrophil accumulation around scattered small vessels in the midbrain with localized meningitis in 1/3 animals
ccJE (two doses of 0.5 μg antigen)	No abnormal lesions in 3/3 animals
ccJE + Advax-1 (single dose of 5 μg antigen)	No abnormal lesions in 2/3 animals; moderate lymphocyte accumulation around a group of small vessels in the brain stem in 1/3 animals
ccJE + Advax-1 (two doses of 0.5 μg antigen)	No abnormal lesions in 3/3 animals
Naive controls, mock infected	No abnormal lesions in 3/3 animals

^aAt day 6 postchallenge, 5 mice per group were sacrificed and the skull was cut open to expose the brain. The brain was sagitally divided into two pieces, and one piece was placed in formaldehyde for subsequent serial section hematoxylin-eosin staining and histology, as described in the text, with representative images shown in Fig. 3. The other brain piece was used for WNV infectious titer assay (Fig. 2B).

Fig 3.

JE-ADVAX immunization reduces histological lesions in midbrain of mice challenged with WNV. Representative hematoxylin-eosin-stained photomicrographs obtained from the midbrain of groups of three mice that were necropsied 6 days after WNV challenge are shown. (A) Severe lymphocyte and moderate neutrophil accumulation (arrows) around multiple small vessels were seen in the midbrain of placebo (saline)-treated mice. (B, D) Mice that received a single dose of ccJE (B) or ccJE plus Advax-1 (D) showed moderate lymphocyte and neutrophil (arrow) accumulation around the small vessels. (C, E) Mice that received two doses of ccJE alone (C) or ccJE plus Advax-1 (E) showed no accumulation of lymphocytes around small vessels. Photomicrographs in panels a to f are higher-magnification images of the images in panels A to F, respectively. Bars = 100 μm.

Efficient JE-ADVAX cross-protection against WNV does not require CD8⁺ T cells.

We have recently shown that the memory B-cell response elicited by immunization with JE-ADVAX provided durable homologous protection against JEV in the absence of both detectable preexposure serum neutralizing antibody and CD8⁺ T cells (27). This protective function of memory B cells was demonstrated in mice lacking intact β2-microglobulin (β2m^−/− mice). These mice display a deficiency in CD8⁺ T cells (38) as well as markedly reduced serum IgG levels (39–41) as a result of the impaired ability, in the absence of β2m, of neonatal Fc receptor recycling of IgG (42). An additional advantage of the use of β2m^−/− mice for studies of the immunological mechanism of WNV protection is their high susceptibility to viral challenge (27). Consistent with the previously observed high susceptibility to JEV infection, high mortality was seen in naive β2m^−/− mice infected with a small s.c. dose (10³ PFU) of virulent WNV, with all animals succumbing to lethal encephalitis by day 13 postinfection (Fig. 4). In contrast, after immunization of β2m^−/− mice with two doses of JE-ADVAX, the majority were protected against lethal WNV challenge (survival, 11/13 [85%]; P < 0.001), notwithstanding undetectable serum JEV-neutralizing antibody prechallenge.

Fig 4.

JE-ADVAX provides cross-protection against WNV in mice lacking CD8⁺ T cells. Groups of 8-week-old β2m^−/− mice were immunized with two doses of JE-ADVAX (0.5 μg) delivered 2 weeks apart or sham vaccinated with saline. Six weeks after completion of the vaccination schedule, mice were inoculated s.c. with 10³ PFU of WNV. Morbidity and mortality were recorded daily, and surviving mice were monitored for 21 days. The data shown were constructed from two independent experiments. Asterisks denote statistical significance (***, P < 0.001).

Given our previous finding of the pivotal role of JE-ADVAX-induced memory B cells in homologous JEV protection, we asked whether a similar mechanism might contribute to the heterologous WNV protection provided by JE-ADVAX. To address this question purified B or CD4⁺ T cells were isolated from B6 donor mice 3 weeks after JE-ADVAX (0.5 μg) immunization and adoptively transferred into naive B6 recipients that were challenged s.c. on the following day with 10⁴ PFU of WNV. B cells from JE-ADVAX-immunized donor mice significantly improved survival in the recipient group after challenge with WNV (survival, 13/15 [87%]; P < 0.05) compared to that in naive mice (survival, 13/26 [50%]; Fig. 5A). Naive mice were used as controls for this experiment, as we have previously shown comparable susceptibility to virus challenge in unmanipulated naive mice and mice that received adoptive transfers of naive splenocytes (35, 43). Further subsets of mice from each group were tested for WNV viremia. By day 3 postchallenge, all naive mice had detectable viremia (10³ to 10⁴ IU/ml; n = 5), whereas all recipients of B cells from JE-ADVAX-immunized mice had undetectable viremia (<10² IU/ml; n = 5), consistent with an ability of adoptively transferred memory B cells to control viral replication. Recipients of CD4⁺ T cells from JE-ADVAX-immunized mice (Fig. 5A) displayed a slight nonsignificant trend toward improved survival (survival, 6/9 [66%]) relative to that for the naive control group (survival, 13/26 [50%]).

Fig 5.

JE-ADVAX induces B cells able to mediate cross-protection against related flaviviruses. Eight-week-old B6 donor mice were immunized with two doses of JE-ADVAX (0.5 μg) delivered 2 weeks apart. Purified B and CD4⁺ T cells were isolated from spleens at 3 weeks postimmunization and adoptively transferred into 6-week-old B6 recipient mice. Naive B6 mice that did not receive any cells served as a control group. (A) One day after splenocyte transfer, all mice were challenged s.c. with 10⁴ PFU of WNV and mortality was monitored in each group_. (B) Purified B cells from JE-ADVAX-immunized or naive donor mice were adoptively transferred into 4-week-old B6 recipient mice, which were then challenged 1 day later s.c. with 10⁵ PFU of MVEV and monitored twice daily for mortality. Asterisks denote statistical significance (*, P < 0.05). Combined statistical analysis of the survival data for groups of immune B-cell recipients challenged with either WNV or MVEV relative to those for the control groups showed that the difference was highly significant (P < 0.001, two-tailed Fisher's exact test).

The cross-protection offered by JE-ADVAX-induced memory B cells against challenge with MVEV, another JE serocomplex flavivirus, was similarly assessed by adoptive transfer of memory B cells from JE-ADVAX-immunized donor mice and compared to that achieved after transfer of naive B cells from nonimmunized mice (Fig. 5B). Young animals were used for this study, given the resistance of adult mice to extraneural infection with MVEV (33). Transfer of immune B cells from JE-ADVAX-immunized donors significantly improved survival after MVEV infection (survival, 6/10 [60%]; P < 0.05) compared to that after transfer of naive B cells (survival, 5/20 [25%]), further supporting a role of JE-ADVAX-induced memory B cells in mediating flavivirus cross-protection.

DISCUSSION

The current absence of an approved human WNV vaccine poses a significant problem when major human outbreaks of WNV, such as the recent 2012 U.S. epidemic, occur. While WNV vaccines are in various stages of clinical development, it is unclear when one or more of these might be approved for human use. Approaches include inactivated WNV vaccines (44), E glycoprotein conjugated to virus-like particles (45), live attenuated WNV vaccines (46, 47), a single-cycle replicon vaccine (48), a plasmid DNA vaccine (49), and chimeric viral vectors expressing the key WNV prM and E proteins on the backbone of either a canary pox virus (50), dengue virus (51), influenza virus (52), or yellow fever virus (53). Among the most advanced candidates in human development is Chimerivax-WN02, a live attenuated chimeric vaccine based on the yellow fever virus 17D clone. An early version of this vaccine caused high levels of viremia in a phase 1 study (53), but after further plaque purification of the chimeric virus, a phase 2 study achieved ∼96% seroconversion with variable viremia for up to 14 days postimmunization (54). Another yellow fever virus-WNV chimeric vaccine (PreveNile; Intervet) was withdrawn from the veterinary market in 2010 due to serious adverse effects, including deaths (14), and was subsequently rereleased as an inactivated vaccine. Although Chimerivax-WN02 differs from the withdrawn equine yellow fever virus chimeric vaccine in having two extra mutations in the E protein, the potential impact of the equine safety issues on future FDA approval of a human live chimeric WNV vaccine is not clear.

Given the availability of approved human JEV vaccines and the close antigenic relatedness of the JE serocomplex family of flaviviruses, this study sought to address whether an appropriately adjuvanted JEV vaccine could provide sufficient heterologous cross-protection against WNV infection to potentially allow the JE-ADVAX vaccine to be used to protect against human WNV outbreaks. Support for this hypothesis came from previous immunogenicity studies showing that a two-dose regimen of JE-ADVAX generated cross-neutralizing antibodies in mice or horses with titers generally considered protective against WNV (strain Kunjin) and MVEV (18). The current results confirm that JE-ADVAX, particularly when administered as a two-dose prime-boost regimen, provided robust protection against lethal WNV challenge, with prevention of weight loss, reduced brain lesions, and a reduction in viral load in the plasma and brain of challenged animals. This protection obtained with the two-dose regimen of heterologous JE-ADVAX vaccine equaled the protection obtained with the homologous West Nile Innovator equine vaccine control. This supports the feasibility of using JE-ADVAX to control human WNV outbreaks in the absence of an approved human WNV homologous vaccine.

While a single dose of JE-ADVAX also provided significant protection against WNV, this did not equal the protection provided by the two-dose regimen, despite the former using a 5-fold higher antigen dose. The 2-dose regimen with JE-ADVAX elicited ∼10-fold higher cross-neutralizing anti-WNV antibody responses than the 1-dose regimen, indicating that prime-boost regimens remain the most antigen efficient to induce high homologous and heterologous titers. Nevertheless, single-dose regimens remain attractive from an epidemic control perspective, and it will be interesting to test whether higher-dose, single immunizations of JE-ADVAX could achieve WNV protection equivalent to that achieved with the 2-dose regimen.

The importance of humoral immunity in vaccine-mediated protection against flavivirus infection is well documented (reviewed in reference 55). For instance, it was shown that passive immunization with neutralizing monoclonal antibody provided protection against WNV challenge (56, 57), and the ability of a two-dose regimen of formalin-inactivated WNV vaccine to protect against lethal WNV challenge was equally efficient in CD8⁺ T-cell-depleted or CD8⁺-sufficient animals (58), supporting the idea that humoral immunity alone is sufficient to protect against WNV without the need for CD8 T-cell immunity. However, the relative contributions of preexisting serum antibody versus those of memory B cells and those of CD8⁺ T cells in vaccine-induced cross-protective immunity among JE serocomplex viruses deserve further study. While passive transfer of immune serum from vaccinated to naive mice was shown to be able to mediate cross-protection among JE serocomplex viruses (20, 59), a recent study demonstrated the broader antigen specificity of memory B cells than serum antibody derived from long-lived plasma cells in mice following primary WNV infection or vaccination (60). Importantly, this broader antigen specificity of memory B cells accounted for the enhanced neutralization of E-protein antigenic variants of WNV (60) and supports the data that we show here that JE-ADVAX-induced memory B cells, when transferred to naive hosts, were able to provide heterologous protection against either WNV or MVEV challenge in the absence of prechallenge neutralizing antibody. This extends our previous finding that JE-ADVAX-induced memory B cells are able to mediate homologous JEV protection (27) and suggests that memory B cells can play a role at least equal to that of cross-reactive serum neutralizing antibody in offering protection against homologous and heterologous flaviviruses. Indeed, when challenge occurs in the absence of preformed cross-reactive serum neutralizing antibody, such as in β2m^−/− mice, memory B cells almost certainly assume a dominant role in cross-protection. This raises the important question of whether it is an identical or different population of JE-ADVAX-induced memory B cells that mediates homologous and heterologous flavivirus protection, a question that we plan to address in future studies.

β2m^−/− mice have multiple immune defects, including a lack of CD8⁺ T cells (38) and compromised production of gamma interferon (IFN-γ) (61), a cytokine shown to be critical in recovery from primary infection with WNV (62, 63). They also lack neonatal Fc receptor-mediated protection of IgG catabolism (reviewed in references 64 and 65) and, as we have shown here, are extremely susceptible to WNV challenge compared to the susceptibility of wild-type mice. Given its multiple immune deficiencies and heightened susceptibility to mortality from flavivirus infection, the β2m^−/− mouse model provides a robust test of WNV vaccine efficacy. Despite the heightened disease susceptibility of β2m^−/− mice, immunization with JE-ADVAX provided robust heterologous protection against WNV, as previously shown for homologous protection of β2m^−/− mice against JEV challenge (27). This was despite the JE-ADVAX-immunized β2m^−/− mice having undetectable JEV- and WNV-neutralizing titers prior to WNV challenge. The protection obtained in the β2m^−/− mouse model confirms that the heterologous protection conferred by JE-ADVAX does not depend on memory CD8⁺ T cells, high levels of IFN-γ production, or prechallenge serum neutralizing antibody. However, the result does not exclude the possibility not investigated here that memory CD8⁺ T cells play a supportive role in JE-ADVAX-mediated protection, given previous demonstrations that CD8 T cells can contribute to single-dose, although not two-dose, WNV vaccine protection (58), adoptive transfer of WNV-specific T cells was able to protect against WNV infection (66), and induction of WNV-specific T cells by DNA immunization with single-chain HLA-A2 major histocompatibility complex trimer molecules incorporating an immune-dominant WNV peptide derived from the E protein was also able to provide protection (67).

Concern has been raised regarding the risk of vaccine enhancement of related flavivirus infections due to waning or subprotective antibody titers, such as those seen in antibody-mediated disease enhancement after a primary dengue virus infection (68). Reassuringly, a recent trial of a quadrivalent attenuated dengue vaccine showed no evidence of dengue disease enhancement, even though the vaccine failed to induce protective antibody titers against several serotypes (69). Furthermore, even if JE-ADVAX-induced serum WNV titers were to wane over time, this does not necessarily mean that WNV protection would be lost or disease enhancement seen. JE-ADVAX-immunized animals remained protected against lethal JEV challenge even in the absence of measurable serum antibody, likely by virtue of a long-lived population of protective memory B cells (27). Furthermore, in the current study, single-dose-immunized animals that had low or undetectable prechallenge WNV antibody titers showed no evidence of disease enhancement. The durability of the heterologous WNV protection mediated by JE-ADVAX is currently being tested in an ongoing study, which will also look for any evidence of disease enhancement as WNV antibody titers wane over time.

These results extend the findings on the ability of the Advax adjuvant to enhance the protection provided by viral vaccines, as previously demonstrated for JEV (18), seasonal (70) and pandemic (23) influenza virus, and hepatitis B virus (25), among others. Why the Advax adjuvant is so effective in enhancing homologous and heterologous protection of viral vaccines is still not known, although the explanation may lie in its broad enhancement of all arms of the adaptive immune response, including antibody-secreting cells, memory B cells, serum immunoglobulin levels, and memory T cells belonging to each of the Th1, Th2, and Th17 subsets (25, 70). This broad-based immune arsenal recruited by Advax should be ideal for homologous and heterologous virus protection. This may help explain the ability of the JE-ADVAX vaccine to induce cross-protective antibodies and heterologous protection against WNV, in contrast to a previous human trial where the unadjuvanted JE-VAX vaccine failed to induce cross-protective WNV antibodies (17). Another contributor to the unique success of JE-ADVAX in inducing cross-protective WNV antibodies is likely the greater antigenicity of the ccJE antigen compared to that of the earlier mouse brain-derived JE-VAX antigen (28). This likely relates to the less harsh formalin inactivation step used for production of the ccJE antigen (28), which may prevent damage to critical epitopes on the JEV E protein needed for induction of cross-protective neutralizing antibodies. In support of this possibility, two doses of ccJE antigen even without adjuvant induced low titers of cross-neutralizing antibodies and provided partial protection against WNV.

Current data suggest that the Advax adjuvant works by recruiting and priming antigen-presenting cells, e.g., dendritic cells, that then have an enhanced ability to take up antigen and activate antigen-specific T and B cells (25). Unlike alum adjuvant, Advax does not require a temporal association with the antigen for its adjuvant effect, having an adjuvant effect even when injected 24 h prior to injection of the antigen (25). While additional aspects of the mechanism of action of this unique polysaccharide adjuvant will be important to decipher, the critical attributes for human vaccine adjuvants are efficacy, safety, and tolerability (71, 72). Reassuringly, there was no evidence of any local or systemic toxicity from JE-ADVAX immunization, which supports the safety and tolerability data seen in recent human trials (22). A clinical trial is currently being planned to test JE-ADVAX in humans for its ability to induce long-term neutralization of both JEV and WNV. If successful, this could lead to development of a single-antigen vaccine protective against multiple JE serocomplex flaviviruses, including JEV and WNV.