CiD

MHJa

4b

c

d

a

ARRAA

KJWCIN

1

teastgAsi

0d

Vaccine 27 (2009) 4370–4380

Contents lists available at ScienceDirect

Vaccine

journa l homepage: www.e lsev ier .com/ locate /vacc ine

oimmunization with an optimized IL15 plasmid adjuvant enhances humoralmmunity via stimulating B cells induced by genetically engineeredNA vaccines expressing consensus JEV and WNV E DIII

athura P. Ramanathand, Michele A. Kutzlerc, Yuan-Chia Kuoa, Jian Yana,arrison Liua, Vidhi Shaha, Amrit Bawaa, Bernard Sellingb, Niranjan Y. Sardesaid,

. Joseph Kimd, David B. Weinera,∗

Department of Pathology & Laboratory Medicine, University of Pennsylvania School of Medicine, 505 Stellar-Chance Building,22 Curie Blvd, Philadelphia, PA 19104, United StatesImpact Biologicals Inc., 520 Georgetown Road, Wallingford, PA 19086, United StatesDivision of Infectious Diseases & HIV Medicine, Drexel University College of Medicine, Philadelphia, PA 19104, United StatesVGX Pharmaceuticals, Inc., 450 Sentry Parkway, Blue Bell, PA 19422, United States

r t i c l e i n f o

rticle history:eceived 10 October 2008eceived in revised form 21 January 2009ccepted 29 January 2009vailable online 6 March 2009

eywords:EV

NVonsensus DNA vaccines

L15

a b s t r a c t

The Japanese encephalitis virus (JEV) and West Nile virus (WNV) are responsible for a large proportionof viral encephalitis in humans. Currently, there is no FDA approved specific treatment for either, thoughthere are attempts to develop vaccines against both viruses. In this study, we proposed novel geneticallyengineered DNA vaccines against these two neurotrophic flaviviruses. The structural domain III (DIII) ofE protein from these viruses is reported to carry dominant epitopes that induce neutralizing antibodies.Therefore we created consensus sequence of DIII domain across numerous strains of JEV and WNV. Basedon the consensus amino acid sequence, synthetic codon and RNA optimized DIII-expressing DNA vaccineconstructs with an efficient leader sequence were synthesized for immunization studies. In addition, wealso constructed a genetically engineered IL15 DNA vaccine molecular adjuvant for co-stimulating theimmune response against DIII clones. Vaccine constructs were delivered into BALB/C mice intramuscu-

®

eutralizing antibodies larly followed by electroporation using the CELLECTRA in vivo electroporator. We have observed thatthe combined delivery of both WNV DIII and IL15-ECRO DNA vaccine constructs resulted in not only thehighest level of antibody against DIII, but also enhanced cross reactivity with two other antigens tested.Also, coimmunization with IL15 plasmid further increased the immune response by four- to five-fold.Importantly, we have shown that IL15 coimmunization adjuvanted humoral responses against DIII anti-

l of alated

gens by elevating the levecontrol potential travel re

. Introduction

The flaviviruses, West Nile virus (WNV) and Japanese encephali-is virus (JEV) are responsible for a large proportion of viralncephalitis in humans [1,2]. WNV infects a wide range of aviannd mammalian species including humans. WNV has also beenhown to be transmitted through blood transfusion [3,4], organransplantation [4,5], and breast-feeding [6]. WNV covers a large

eographical area that includes the south of Europe, Africa, centralsia, and more recently North America. The virus often producesymptoms of meningoencephalitis in Israel, Egypt, India and Pak-stan. In Egypt, it is responsible for 3% of all meningitis and aseptic

∗ Corresponding author. Tel.: +1 215 662 2352; fax: +1 215 573 9436.E-mail address: dbweiner@mail.med.upenn.edu (D.B. Weiner).

264-410X/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2009.01.137

ntibody secreting B cells. Such a DNA vaccine approach may better help toinfectious agents such as JEV.

© 2009 Elsevier Ltd. All rights reserved.

encephalitis [7]. JEV is the single-most important cause of viralencephalitis in Asia, with case fatality rates averaging 30% [8,9].JEV is a major problem in South-East Asia, India, and China, wherethe virus is endemic. In recent years, JEV has spread to other geo-graphic areas such as Australia, and Pakistan, and has thus becomean important emerging virus infection in these areas. Japaneseencephalitis (JE) is an inflammatory disease of the central nervoussystem including the cerebrum, the cerebellum, and the spinal cord[10]. Several studies have suggested that a mouse-brain-derived,inactivated vaccine against JEV appeared to be effective [11–13].However, there are concerns about the immunogenicity and the

safety of this vaccine [14,15]. A live-attenuated and a primary ham-ster kidney (PHK) cell culture-based JEV vaccine has been licensedin China and shown to be effective. However, since PHK is not anapproved cell line for production of human vaccines and the useof products developed from mouse brain tissue is a matter of con-

Vaccin

cvrb

cvaaafesiEabtDepttsaia

asrirsscteeab[t[ha

oaio[Isvetoseo

cfrr

M.P. Ramanathan et al. /

ern, many countries including the US will not administer this JEVaccine. Therefore, a recombinant vaccine representing a smallestegion of the virus sufficient to induce focused neutralizing anti-odies, represents an attractive alternative [16].

The flaviviral E protein (53–55 kDa) is a typical membrane gly-oprotein, forming the outer structural protein component of theirus, and is responsible for a number of important processes suchs viral attachment, fusion, penetration, cell tropism, virulence andttenuation [17]. The 500-amino acid E protein is also the majorntigen responsible for eliciting neutralizing antibodies that con-er protection to the host. Structurally, the E protein exists as a flat,longated dimer on the surface of the virion, and consists of threetructural domains (DI, DII and DIII), of which DIII contains predom-nantly sub-complex and type-specific epitopes [18–20]. Hence, the

protein is the most important target for neutralizing antibodiesnd is therefore suitable for vaccine development. Several vaccinesased on DIII have been shown to be immunogenic and induce neu-ralizing antibodies under certain conditions [13,21–24]. However,III protein alone appears to be less immunogenic when deliv-red as a protein-immunogen, though the conformational status ofurified DIII regions resembles that of native DIII domain in E pro-ein [25]. Alka et al. optimized the conditions that could enhancehe immunity against DIII-based protein antigens [26]. Therefore, ifuitable modifications are added in the process of immunization ofDIII-based DNA vaccine to enhance its immunogenicity and abil-

ty to induce highly active neutralizing antibodies, this can becomesafe and effective vaccine against both JEV and WNV.

In this study, we have tested three types of DIII sequencess putative vaccine targets including consensus JEV DIII, consen-us WNV DIII and a chimeric JEV sequence in which WNV DIIIesidues [19,20] that were shown to be determinants for thenduction of neutralizing antibodies against WNV, were incorpo-ated at the corresponding positions of JEV DIII domain. All theseequences are consensus across fifteen different strains. Recenttudies have suggested that “centralized” immunogens such asonsensus immunogens or ancestral immunogens may be usefulo minimize the degree of sequence dissimilarity among differ-nt virus strains [27]. In our previous work, we have built novelngineered HIV and Influenza consensus optimized immunogensnd the results suggested that these novel immunogens inducedroader cellular immune responses than native DNA immunogens28,29]. Hence, consensus sequences have become an importantool for inducing immune responses against specific infections28,31]. Based on the generated consensus sequence, syntheticuman codon optimized DIII-encoding sequences were generateds DNA immunization cassettes for further study.

Recently, several strategies aimed at improving the magnitudef the cellular immune responses induced by DNA vaccines, suchs codon optimization and RNA optimization, and the addition ofmmunoglobulin E (IgE) leader sequences that have weak RNA sec-ndary structure have been studied and applied successfully to HIV28,32], HPV [33] and Influenza [29,34] DNA vaccine development.n combination with these expression enhancing features, inclu-ion of co-stimulatory molecules and/or adjuvants and improved inivo delivery such as in vivo electroporation [35] may enhance thexpression of the target antigen, in the context of vaccination. Elec-roporation has classically been used in vitro to enhance the deliveryf plasmids inside the cells using cell culture techniques. Recenttudies, however, have shown its promise in enhancing the deliv-ry and expression of plasmid DNA in vivo, leading to the generationf more potent immune responses [33,34].

IL15 is one of the important DNA vaccine adjuvants, and is aytokine that is known to have significant impact on CD8+ T cellunction [36,37]. IL15 has been shown to improve the immuneesponse against tumors [38] in specific disease models. It has beeneported that IL15 plays an essential role in the proliferation of

e 27 (2009) 4370–4380 4371

memory CD8+ T cells [39,40]. The important role of IL15 in the con-trol of all phases of T cell differentiation into a memory phenotype,including proliferation of antigen-specific T cells, rescue of T cellsfrom death during the contraction phase of the immune responseand proliferation/maintenance of memory T cells [41] has been welldocumented. In an effort to recruit this molecule as a potential DNAadjuvant, Kutzler et al. optimized and identified features that areimportant for enhanced expression in the context of vaccination[39]. In that report, it was determined that immunization with opti-mized IL15 in combination with HIV-1 gag DNA constructs resultedin a significant enhancement in antigen-specific CD8+ T cell prolif-eration, IFN-gamma secretion, strong induction of long-lived CD8+T cell responses, and protected the mice against a lethal mucosalchallenge with influenza virus [39]. Moreover, these studies indi-cated that IL15 could be used as an adjuvant during multiple boostimmunizations, as no immune tolerance was observed as measuredby anti-IL15 antibodies (unpublished observations). In addition tothe important role that IL15 plays in regulation of T cell function, ithas also been determined that IL15/IL15Ra pathways are integral tothe maturation and function of antigen presenting cells [42]. There-fore, due to the regulation of T and antigen presenting cell functionand differentiation by IL15, we hypothesized that coimmunizationwith IL15 as an immune adjuvant may lead to enhanced antibodyproduction as a result of enhanced CD4 T cell help and APC function.Based on these studies, we created a synthetic codon optimizedIL15 gene expression vector and used it as a DNA vaccine adjuvantfor enhancing immune response against WNV and JEV DIII. In thisstudy, we observe an enhanced immune response through combi-nation of all these parameters with small quantity of WNV or JEVDNA vaccine. Furthermore, we present a direct evidence for induc-tion of improved B cell immune response via co-vaccination withIL15 DNA plasmid.

2. Materials and methods

2.1. Cell lines

HeLa and RD (rhabdomyosarcoma) cells were obtained fromthe American Type Culture Collection (Manassas, VA). They weremaintained in DMEM medium supplemented with 10% heat inacti-vated fetal bovine serum, penicillin G (100 U/ml) and streptomycin(100 �g/ml) at 37 ◦C in 5% CO2.

2.2. Consensus DNA vaccine constructs

To generate JEV and WNV E DIII consensus sequences,approximately fifteen different sequences representing differentgeographical regions were collected from the Gene Bank (NCBI) toavoid sampling bias and aligned using MegAlign (DNASTAR, Madi-son, WI). IgE-leader sequence was added to the amino terminusof resulting consensus sequence and two stop codons were addedat the end of the open reading frames. KpnI and PstI sites werefused to 5′ and 3′ ends of the coding sequence respectively. Thecomplete sequence was subjected to codon optimization and RNAoptimization by using GeneOptimizer (GENEART, Regensburg, Ger-many). The codon optimized synthetic sequences were then clonedinto pVAX1 expression vector (Invitrogen) using routine cloningstrategies.

2.3. Cloning of human IL15-ECRO into pVAX1 vector and IL15expression analysis

Native IL15 contains two alternative leader peptides that are notonly involved in the regulation of IL15 translation, but also directits intracellular trafficking. The classical long (48 aa) signal pep-tide is associated with all secreted forms of IL15, while IL15 that

4 Vaccin

csmaptvmusGdhi(tTrsaadoTfacgg(ciwfp

2

agf(s2waafStoPiP

2

pv(lawa

372 M.P. Ramanathan et al. /

ontains the short 21 aa signal peptide is not secreted but rathertored intracellularly [41]. Design for the plasmid form of opti-ized human IL15 (pIL15-ECRO) requires replacing the LSP with

n “optimized” IgE-leader designed by our laboratory for increasedrotein expression [28,39]. Moreover, the codon usage was adaptedo the codon bias of Homo sapiens genes resulting in a high CAIalue (non-optimized: 0.66; optimized: 0.98). Since human andouse genes share 73% homology, we utilized pIL15-ECRO in vivo

sing the mouse model. For design and synthesis, codons wereelected so that regions of very high (>80%) or very low (<30%)C content was avoided where possible. In this regard, it has beenetermined that the wild type IL15 gene uses rare codons with aigh frequency and the GC content was quite low (35%) which facil-

tates quick mRNA turnover. Therefore, GC content was increased57%) to prolong mRNA half-life. During the optimization process,he following cis-acting sequence motifs were avoided: internalATA-boxes, chi-sites and ribosomal entry sites, AT-rich or GC-ich sequence stretches, ARE, INS, CRS sequence elements, repeatequences and RNA secondary structures, (cryptic) splice donornd acceptor sites and branch points. Following analysis, 3 neg-tively cis-acting motifs were identified and removed. The finalesign of the gene contained 100% congruence with mature formf human IL15 with IgE-leader replacing the wild type LSP form.he synthetic codon optimized human IL15 gene was assembledrom synthetic oligonucleotides and cloned into pVAX1 using EcoRInd XhoI restriction sites by Geneart, Inc. (Germany). The finalonstruct was verified by sequencing and found to be 100% con-ruence. One day before transfection, 7.0 × 105 HeLa or RD cellsrown in D10 medium were seeded onto 60 mm culture dishesFalcon). The cells were then transfected with 5 �g of the IL15onstructs with DOTAP liposome transfection kit (Roche Biochem-cals, CA) according to the manufacturer’s protocol. Supernatants

ere collected at 24, 48, and 72 h post transfection and analyzedor protein levels via ELISA (R & D) as per the manufacturer’srotocols.

.4. Indirect immunofluorescent assay

HeLa cells were used to perform indirect immunofluorescentssays routinely. The cells were seeded in a two-chamber slide andrown for overnight in D10 medium prior to their use in trans-ection. They were transfected with vaccine constructs or pVAX1 �g/well) using FuGENE 6 Transfection Reagent (Roche). Thirtyix hours post transfection, the cells were fixed with methanol for0 min at room temperature and washed gently with PBS. Theyere incubated with the sera from DNA vaccinated mice for 90 min

nd washed again. Subsequently, the samples were incubated withnti-mouse FITC-conjugated secondary antibody (Sigma–Aldrich)or 45 min. 40,6-Diamido-2-phenylindole hydrochloride (DAPI;igma–Aldrich) was added to the solution of secondary antibodyo counterstain nuclear contents in order to show the total numberf cells available in the given field. The images were acquired usinghase 3 Pro program for fluorescent microscopy (Media Cybernet-cs, Silver Spring, MD) as TIFF files and further analyzed using Adobehotoshop program.

.5. In vitro expression of vaccine constructs

The ability of constructs to generate their gene products withredicted molecular mass was confirmed prior to their use inaccination studies. TNT T7 in vitro transcription/translation kit

Promega, Madison, WI) was used to generate 35S-methionineabeled protein samples a T7-promoter in the pVAX1 backbones per the supplier’s protocols. The radiolabeled protein samplesere immunoprecipitated using anti-WNV E or anti-JEV E antibody

nd the immunoprecipitated complexes were electrophoresed on

e 27 (2009) 4370–4380

a 15% SDS-PAGE gel (BioRad). The gel was fixed using phospho-enhancer solution (Amersham) and dried using a vacuum drier(Biorad). Autoradiography was performed to detect the signals fromthe 35S-labeled gene products.

2.6. Bacterial expression E DIII proteins

The DIII-coding regions were released from their pVAX1 back-bone by digestion with KpnI/PstI. The released fragments weregel-purified, then ligated into an appropriately digested bacterialexpression vector, pQE30. The ligated products were transformedinto JM109 cells, which were plated onto LB agar plates containing50 �g/ml ampicillin. Colonies formed overnight at 37 ◦C. Glycerolstocks were prepared from individual colonies. For protein expres-sion, a 3 �l aliquot of glycerol stock was added to a 16 mm × 150 mmtube containing 6 ml of LB broth containing 50 �g/ml ampicillin.The resulting bacterial pellet was processed and the lysates werepassed through Ni-columns as per the standard protocols foraccomplishing purification of the proteins samples. The eluted pro-tein samples were further dialyzed. The resulting purified proteinsamples were used as coating antigens in ELISA as well as B cellELISPOT assays.

2.7. Mice studies

2.7.1. Mice and immunizationThe quadriceps muscles of six to eight week-old female BALB/c

mice (Charles River, Wilmington, MA) were injected three timesand electroporated, each with 10 �g of DNA at biweekly inter-vals. For DNA immunizations, the mice were separated into groupsof five mice each and immunized by electroporation with pIL15-ECRO and pVAX (control group) or pJEV DIII or pJE/WNV DIII orpWNV DIII, or pJEV DIII + pWNV DIII, respectively. Briefly, square-wave pulses were delivered through a triangular 3-electrodearray consisting of 26 gauge solid stainless steel electrodes. Twoconstant-current pulses of 0.1 A were delivered for 52 ms/pulseseparated by a 1 s delay using the CELLECTRA® adaptive constant-current device (VGX Pharmaceuticals Inc., The Woodlands, TX).The sequence of events for plasmid administration/elctroporationwas as follows: place a disposable electrode array in the recep-tacle of the handle, press initiation button on handle and enteranimal experimental group number, inject plasmid DNA usinginsulin syringe, immediately place the array into the area sur-rounding the injection site, press initiation button on handle, andafter 4 s countdown, pulses will be delivered. The EP conditionswere 0.1 A, 2 pulses, 52 ms/pulse, 1 s between pulses. All elec-trodes were completely inserted into the muscle during treatment.Mice were housed and treated at the University of Pennsylvania,and cared for under the guidelines of the NIH and the Univer-sity of Pennsylvania Institutional Animal Care and Use Committee(IACUC).

2.7.2. DIII antibody ELISA assayThe DIII protein suspensions were thawed, and the tubes were

vortexed to suspend to the fine protein particulates. Small aliquotsof the resuspended protein samples were dissolved in Tris–Ureabuffer (62 mM Tris–HCl/8 M urea, pH 8.0) to produce a 10 �g/mlsolution. Aliquots of 100 �l (1 �g) of diluted protein samples wereapplied to the wells of a HisGrab copper-coated high-binding capac-ity plates (Pierce) and incubated for overnight at 4 ◦C. The nextday, plates were washed with PBST (PBS, 0.05% Tween 20), blocked

for 1 h with 3% BSA in PBST, and incubated with serial dilutions ofserum from immunized and naïve mice for 1 h at 37 ◦C. Bound IgGwas detected using goat anti-mouse IgG-HRP (Research Diagnos-tics, NJ) at a dilution of 1:10,000. Bound enzyme was detected bythe addition of the chromogen substrate solution Tetramethylben-

M.P. Ramanathan et al. / Vaccine 27 (2009) 4370–4380 4373

Fig. 1. Generation of JEV and WNV vaccine constructs. (A) Organization of flaviviral polyprotein with cleavage sites catalyzed by viral and host cell proteases. A threedimensional conformation of flaviviral E shows three distinct domains, I, II and III. On the right side, plasmid map reveals various elements of vaccine constructs used in thisstudy. (B) The consensus amino acid sequence of DIII is given. IgE-leader sequence, fused at the N-terminus, for maximizing the secretion process is underlined. (C) Restrictiondigestion analysis of vaccine constructs encoding JEV DIII, WNV DIII and chimeric JE/WNV DIII fragments indicates the length of DIII-coding region (435 bp). On the left end ofthe lanes, the DNA ladders are shown. (D) In vitro expression of DNA vaccine constructs. 35S-labeled gene products generated from DNA vaccine constructs were resolved in aS y of ae e const by RT-a sfecte

zB

2

tcteawl(rR

DS gel, as described in Section 2. The protein products corresponded to the mobilitmpty vector indicates specificity of this reaction. E, expression analysis DIII-vaccinransfection, they were analyzed for the presence of DIII-specific mRNA transcriptsppropriate primers, about 435-bp products were observed from DIII construct tran

idine (TMB; R&D Systems), and read at 450 nm on a Biotek EL312eio-Kinetics reader. All serum samples were tested in duplicate.

.7.3. Splenocyte purificationMice were sacrificed one week after the third immunization and

he spleens from each mouse were harvested and pooled in a 15 mlonical tube containing RPMI 1640 (one tube for each experimen-al group). In a sterile tissue culture hood, the pooled spleens fromach experimental group were crushed in a sterile bag using Stom-cher blender (Brinkmann Instruments, Inc). The splenocytes were

ashed and pelleted (1200 rpm); the pellet was treated with ACK

ysing buffer (Biosource) for 5–10 min, and then washed, pelleted1200 rpm) and put through a 70 �m cell strainer to remove anyemaining spleen organ stroma. The splenocytes were washed inPMI 1640 twice, resuspended in R10 medium (RPMI 1640 plus

pproximately 16.0 kDa in mass. Lack of corresponding protein product from pVAX1tructs in RD cells. Cells were transfected with vaccine constructs and two days postPCR. Mock-transfected cells failed to indicate appropriate amplified product. Usingd cells confirming the presence of corresponding DIII-encoding transcripts.

10% FBS), and counted (cell viability is determined using trypanblue stain) using a hemocytometer for their use in ELISPOT assays.

2.7.4. B Cell ELISPOT assayThe ELISPOT 96-well plates (Millipore) were coated with 100 �l

(2 �g/ml) of purified proteins and incubated for overnight at 4 ◦C.Separate plates were used for each coating antigen. The follow-ing day, plates were washed and blocked for 2 h with 1% BSA.Two hundred and fifty thousand splenocytes from DNA vaccinatedmice were added to each well and incubated for 5–6 h at 37 ◦C in

5% CO2 in the presence of RPMI 1640 (negative control). Follow-ing incubation, the cells were washed and incubated for overnightat 4 ◦C with biotinylated anti-mouse antibody (R&D Systems). Theplates were washed, and streptavidin–alkaline phosphatase (R&DSystems) was added to each well and incubated for 2 h at room tem-

4374 M.P. Ramanathan et al. / Vaccine 27 (2009) 4370–4380

Fig. 2. (A) Localization of three different IL15 isoforms. HeLa cells were transfected with plasmids expressing IL15-SSP or IL15-LSP or IL15-ECRO isoforms and were analyzedby immunofluorescence assay after 36 h of transfection. The cells were fixed, incubated with monoclonal anti-IL15 antibody, followed by incubation with anti-mouse FITC-conjugated secondary antibody. DAPI was used to counter stain the nuclear contents of the cells. Empty vector was used as negative control and the mock-transfected cellsdid not yield any specific staining (a–c). The IL15-SSP isoform is not secreted, but rather stored intracellularly appearing in nuclear components (d–f), whereas the IL15-LSPisoform (g–i) and human codon optimized IL15 (j-l) are associated with secreted IL15 as clearly revealed in the cytoplasmic regions. (B) Secretion level of IL15 isoforms. RDc ernatc f secrea

p3cedEa

2

oew

ells were transfected with IL15 expression constructs and after two days, the suponstruct yielded highest level of IL15 secretion. (C) In order to compare the level os fold increase over the -SSP form.

erature. The plates were again washed, and 5-bromo-4-chloro-′-indolylphosphate p-toluidine salt and nitro blue tetrazoliumhloride (chromogen color reagent; R&D Systems) were added toach well. The plates were then rinsed with distilled water andried at room temperature. Spots were counted by an automatedLISPOT reader (Cellular Technology Ltd.). Results are expressed asnumber of antigen-specific antibody secreting cells (ASCs).

.8. Statistical analysis

All of the values are expressed as the mean ± standard errorf the mean (SEM) calculated from triplicate samples from eachxperimental group. Where appropriate, the statistical differenceas assessed by using a two-tailed, paired Student’s t Test and

ants were analyzed for the presence of secreted IL15 protein by ELISA. IL15-ECROtion, the amount of secreted IL15 from IL15-LSP and -ECRO constructs were shown

yielded a specific p value for each experimental group. Routinelythe data shown were representative of at least three independentexperiments done in duplicate or triplicate.

3. Results

3.1. Generation of consensus E DIII immunogens

Fig. 1A reveals schematically the organization of flaviviral

polyprotein and the conformational features of E protein. In aneffort to develop an immunogen with the ability to induce highlycross-reactive cellular immune responses, amino acid sequencesJEV and WNV E DIII domain were downloaded from publicdatabases. The sequences were chosen from diverse geographical

Vaccine 27 (2009) 4370–4380 4375

lstgssnco(gbsscma(

3

taWtTmfpcpacttTss

3

wiAatcaahcsast

3

cEfta

Fig. 3. Bacterial production and purification of DIII protein fragments. DIII-encodingregions were cloned into pQE30-expression vector and protein samples were puri-fied from bacterial lysates as described in Section 2. Coomassie staining of PAGEindicates the samples collected from different stages of purification steps. Afterbinding to the Ni-chelated column, elusion buffer with 20 mM-imidazole failed to

M.P. Ramanathan et al. /

ocations and were non-recombinant. MegAlign (DNASTAR, Madi-on, WI) program was used to align the amino acid sequences andhe most common amino acid at each position was chosen. Weenerated three constructs, a consensus WNV DIII construct, con-ensus JEV DIII construct and a chimeric JEV construct in whicheveral residues from WNV DIII that were shown to be determi-ants of inducing neutralizing antibodies were incorporated at theirorresponding positions of JEV DIII, without altering the lengthf DIII fragment; this antigen is referred as JE/WNV DIII domainFig. 1B). The rationale for this chimeric clone was to have a sin-le vaccine construct that could provide enhanced immunity tooth WNV and JEV together, instead of two individual vaccine con-tructs. After generation of the consensus sequence, an IgE-leaderequence was fused to the amino terminus of the DIII domainoding sequence. Based on the amino acid sequence, human opti-ized synthetic gene sequences were constructed was created for

ll three vaccine constructs, and cloned into the pVAX1 vectorFig. 1C).

.2. Validation of consensus antigens

DIII-vaccine constructs were validated for their ability to expressheir gene products with appropriate molecular mass. By usingn in vitro translation/transcription system (Promega, Madison,I), 35S-labeled radioactive protein products were generated from

hese constructs, immunoprecipitated, and resolved by SDS-PAGE.hese plasmids generated gene products of about 16.0 kDa inass corresponding to the predicted length of the open reading

rame (Fig. 1D). In order to confirm the presence of appro-riate mRNA transcripts by RT-PCR during their expression inells, RD cells were transfected with either empty vector orVAX1-borne DIII domain constructs. The presence of cDNA ofbout 420 bps from transfected cells expressing individual DIIIonstructs was observed only from the cells transfected withhe DIII constructs. Lack of corresponding signals from mock-ransfected cells confirms the specificity of this assay (Fig. 1E).hese results together validated JEV, WNV and JE/WNV DIII con-tructs and hence they were further subjected to immunizationtudies.

.3. Expression analysis of IL15-ECRO construct

In order to confirm the localization pattern of IL15, HeLa cellsere transfected with each of these IL15 constructs and subjected to

mmunofluorescent analysis using monoclonal anti-IL15 antibody.s anticipated, IL15-SSP has been retained within the nuclear regionnd both IL15-LSP and IL15-ECRO displayed cytoplasmic localiza-ion patterns (Fig. 2A). Next, a transfection assay in RD cells wasarried out to compare the efficiency of IL15-constructs in theirbility to express and secrete functional IL15. The data showed thatsubstantial increase in protein production and secretion (87-foldigher than the native IL15, and 5.7-fold higher than the IL15-LSPonstruct) were observed from the human codon optimized con-truct, IL15-ECRO, as measured by specific ELISA analysis (Fig. 2Bnd C). These data clearly demonstrate that the pIL15-ECRO con-truct is highly efficient in expressing and secreting IL15 than all ofhe IL15 expressing plasmids that we have reported earlier [39].

.4. Production of JEV, WNV and JE/WNV DIII proteins

We were interested in generating purified protein samples of

onsensus antigens for their application in subsequent ELISA andLISPOT studies. For this purpose, DIII-encoding fragments wereurther subcloned into a bacterial expression vector, pQE30 andhe protein fragments were purified using Ni-chelated columnss described under Section 2. Fig. 3 describes the purity level of

release DIII fragments and imidazole at 250 mM was effective in selectively elutingpolyhistidine-tagged DIII fragments. The prominent bands of purified DIII fragmentswith appropriate molecular mass (16.0 kDa) are marked.

DIII fragments from different purification steps. The lanes contain-ing crude lysates from DIII-expressing bacterial cells indicated thateven without IPTG-induction, the level of the expression of DIII frag-ment was very prominent. In comparison to the ladder of samplespresent from total crude sample lanes, a distinct single band thatcorresponded to the mass of 16 kDa was observed from the elutedfractions containing 250 mM imidazole indicating a high level ofpurity of the expressed DIII fragments. Imidazole at 20 mM wasnot sufficient to release His-tagged DIII protein fragments fromNi-chelated matrix. Hence the choice of imidazole in the elusionbuffer was very specific to the elusion of the His-tagged protein. Thepurified fractions were dialyzed and stored in aliquots upon deter-mination of their concentration for their subsequent applicationsas coating antigens in this study.

3.5. Titration of anti-DIII antibodies from DNA vaccinated sera

In vivo electroporation (IVE) appears to be a promising new tech-nology for the delivery of DNA vaccines. In order to determine, ifcombined with our consensus antigens, IVE could induce a potentimmune response against synthetic DNA antigens, we immunizedand electroporated BALB/c mice with each of the individual vac-cine candidate in the presence or absence of pIL15-ECRO (Fig. 4A).Serum samples from DNA vaccinated mice obtained one week afterthird booster were assayed for the presence of respective antibodyagainst JEV and WNV by ELISA. All of the vaccinated mice had pro-duced antibodies against DIII domains. Sera from the entire DNAvaccinated mice were evaluated individually against all three anti-gens: JEV DIII, WNV DIII and JE/WNV DIII. As shown in Fig. 4B, ELISAdata has yielded interesting results. First, this assay clearly indicated

that the coimmunization of DIII constructs with IL15 plasmid hassignificantly enhanced antibody production against both JEV andWNV DIII antigens. Comparatively, the addition of IL15 has inducedhighest level of response against WNV DIII. Secondly, in the animal

4376 M.P. Ramanathan et al. / Vaccine 27 (2009) 4370–4380

Fig. 4. DIII antibody ELISA. (A) The schematic plan for vaccination and immunization is indicated. BALB/c mice were immunized two weeks apart with 20 �g vaccine constructso ter. (BD ith 1w s. The

gaaactcesaaabta

r pVAX1 empty vector via CELLECTRA® electroporator and sacrificed one week laNA vaccinated mice was incubated for 1 h at 37 ◦C on 96-well Ni-chelated plates was detected using anti-mouse IgG-HRP with routine color development procedure

roup that was not coimmunized with IL15, the combination of JEVnd WNV DIII expression plasmids induced higher levels of anti-DIIIntibody than elicited by these two vaccine constructs individu-lly. Thirdly, in contrary to the rationale that was set forth for theonstruction of chimeric JEV/WNV construct, the antibody produc-ion was found to be significantly lower with this chimeric vaccineonstruct. Although coimmunization with IL15 showed immunenhancement to the chimeric construct, DIII-specific antibody mea-ured in the sera from this group were the least immunogenics compared to the other constructs (Fig. 4). Thus, this chimeric

ntigen failed to serve as a potential antigen for inducing a strongntibody immune response, not only against JEV DIII and WNV DIII,ut also against the chimeric protein itself. Overall, the antibodyiter from the animals that received the combination of WNV DIIInd IL15 constructs was higher than all other experimental groups;

) Anti-DIII serum specifically reacts with purified DIII fragments. Serum from DIII�g/well of purified DIII protein samples. The production of appropriate antibodiesplates were read at 450 nm. Values represent the mean (±S.D) of triplicate wells.

most importantly serum from this group was found to cross reactwith all three antigens demonstrating the broad reactivity inducedby this combination.

3.6. Specificity of anti-DIII sera to DIII-expressing transfected cells

Sera from DNA vaccinated mice were also examined to verifywhether it can bind to the DIII antigens expressed in cells thatwere transfected with DIII-vaccine constructs. For this purpose,HeLa cells were transiently transfected with all three DIII-encoding

constructs and 36 h post transfection, the cells were fixed forimmunofluorescence analysis. The cells were first incubated withappropriate serum samples followed by FITC-conjugated anti-mouse secondary antibody. The nuclear content of the cells werecounterstained with DAPI. Anti-DIII sera from DNA vaccinated mice

M.P. Ramanathan et al. / Vaccine 27 (2009) 4370–4380 4377

F ells wT micew nuclem

salcAu

ig. 5. Staining of DIII-transfected cells with sera from DNA vaccinated mice. HeLa cwo days post transfection, they were fixed and incubated with serum from BALB/cith FITC-conjugated anti-mouse secondary antibody and DAPI which counter stainsice is presented below each set.

tained only the transfected cells that expressed WNV (Fig. 5A)

nd JEV DIII (Fig. 5B). The staining pattern resembled that of fullength flaviviral E [43]. Mostly importantly, the DIII-expressingells revealed a predominantly cytoplasmic localization pattern.nti-DIII-serum did not show any significant staining with eitherntransfected or pVAX1-transfected HeLa cells.

ere transfected with pVAX or pWNV DIII or pJEV DIII-expressing vaccine constructs.immunized with these DNA vaccine constructs. Subsequently they were incubatedar contents of the cells. Lack of appropriate staining by sera from pVAX1-vaccinated

3.7. IL15 enhanced B cell activation against JEV and WNV antigens

Lastly, we wanted to determine whether the DIII-specific anti-body titers correspond to the presence of appropriate level ofDIII-antigen-specific B cells from the spleens of DNA vaccinatedmice. We employed B cell ELISPOT to address this question. The

4378 M.P. Ramanathan et al. / Vaccin

Fig. 6. DIII-specific B cell ELISPOT. BALB/c mice were immunized three times, eachtwo weeks apart, with 20 �g pVAX1 vector or DIII expression constructs and sac-rificed one week later. Splenocytes were harvested and cultured overnight in thepresence of R10 (negative control) or 2 �g/ml of one of the three purified DIII pro-tat

BsFsditscvcDlWogtDaacvti

expressed in insect cells. The major drawback of these sub-unit vac-

ein samples. Spot forming units were quantified by an automated ELISPOT reader,nd the raw values were normalized to SFU per million splenocytes. Values representhe mean (±S.D.) of triplicate wells.

cell ELISPOT technology allows the quantification of antigen-pecific antibody secreting cells within each immunization group.or this investigation, mouse spleens were prepared to mea-ure the frequency of B cells capable of producing antibodiesirected against DIII proteins, as described in Section 2. Interest-

ngly ELISPOT assay results concur with DIII antibody ELISA datahat were described above. Higher frequencies of mouse antibodyecreting cells specific for DIII were observed in mice that wereo-stimulated with IL15 than the groups that received DIII DNAaccine alone (Fig. 6). This assay also suggested that IL15 haslearly boosted antibody secreting cell responses against WNV DIIINA vaccine, as evidenced that this group exhibited the highest

evel of antibody secreting cells. Most importantly B cells fromNV DIII + IL15 group could cross react with all three antigens, as

bserved with anti-DIII ELISA data from this group. Next to thisroup, JEV DIII + IL15 displayed enhanced level of B cell responsehan the mice received JEV DIII alone. B cells from chimeric JE/WNVIII vaccinated mice exhibited a modest level of antigen-specificntibody secretions. Similar to the ELISA results, the chimericntigen is not as effective as the other constructs at eliciting B

ell response. Altogether, the ELISPOT results indicated the acti-ation of high-affinity DIII-specific B cells that corresponded tohe titers of DIII-specific antibodies as determined by ELISA stud-es.

e 27 (2009) 4370–4380

4. Discussion

JEV is an important pathogen because of its associated acute epi-demic viral encephalitis [44,45]. Further complicating matters forvaccine development, the RNA genome of flaviviruses is prone tohigh rates of mutation, resulting in genetically diverse JE virusescirculating in human populations in Asia and Australia [46,47].It is however generally considered that all strains in circulationbelong to a single serotype defined by neutralization parameters[48]. A recent molecular analysis of JEV strains from Asia classifiedthem into four distinct genotypic groups [48]. JEV vaccines that arecurrently available are based on only one strain of JEV; this highlevel of sequence diversity has led to questions and concerns aboutthe cross-protective effect of current JE vaccines against circulat-ing strains of this neurotrophic virus [14,15]. In order to developan antigen with an ability to induce highly reactive neutralizingantibodies, we employed the approach of generating consensusconstructs against this flavivirus.

JEV infections are regarded as one of the most serious viralcauses of encephalitis, with a mortality of up to 30–50% andhigh percentage of neurological sequelae in survivors [9]. Thus,mass immunization programs against Japanese encephalitis (JE)are generally recommended for the populations residing in theendemic areas by regional and international public-health author-ities, including WHO. In developed and non-endemic countries,JE is regarded as a rare and exotic disease. But in recent decades,case reports of infections in tourists and other travelers from non-endemic regions have been reported almost every year. However,vaccine coverage in the population of international travelers at riskis very low; due not only to a lack of awareness of the disease on thepart of travelers and their travel health advisers, but also because offear of the potential adverse reactions associated with the currentlylicensed mouse-brain-derived JEV vaccine [44]. Unfortunately therequirements for multiple-dose regimen and problems with reacto-genicity have complicated its use. An affordable vaccine that elicitsdurable immunity without the need for frequent boosters is neededfor controlling JE in developing countries.

WNV is the most widespread arbovirus in the world. Currentlythere is no specific treatment for WNV disease in humans, althoughseveral antiviral compounds and therapies are being tested in clini-cal trials. Human beings and Equidae are sensitive to WNV infection,although most cases are asymptomatic. In humans, approximately20% of WNV-infected individuals develop a mild febrile illness dur-ing 3–6 days. Less than 1% of infected patients progress to a nervousform of the diseases. In the nine years since its introduction into thewestern Hemisphere, WNV has established enzoonotic transmis-sion cycles in nearly every part of the United States and southernCanada. The fact that transmission has also been documented inthe Carribean and in Central and South Americas makes it likelythat WNV has wider distribution in that region than has beenreported. WNV transmission will be very difficult to prevent andcontrol because of the poor public-health infrastructure to dealwith vector borne and zoonotic diseases. Vaccines can be used toprevent WNV infection and antiviral treatments can be used to treatsevere disease. There are currently four licensed WNV vaccines forhorses and one licensed vaccine for domestic geese. All recombi-nant vaccines are constructed from the preM/E gene sequences ofthe highly pathogenic NY1999 strain. No vaccine against WNV hasbeen approved for use in humans.

Against WNV, there are few sub-unit vaccine candidates basedon recombinant E protein [49,50] or VLP (virus like particle)

cines is the high number of doses required for protective immunity.Vaccination may have been, at least partially responsible for thedecrease of incidence in equine cases in the USA since 2001, whileat the same time the number of human cases still continued to grow.

Vaccin

TwpalvcaDu

JfdaiaDilflhaTpiflrbtccpn

tttJcIeacalrJftddDrpsi

A

DaV

[

[

[

[

[

[

[

[

[

[

[

M.P. Ramanathan et al. /

he level of protection against European strain of vaccines preparedith American strain is unknown, but the need for vaccines is pre-onderant in the USA where highly pathogenic strains circulatend where most potential hosts are non-immune. The next chal-enge in vaccination will be the development of safe and effectiveaccine (s) in humans in the USA, mostly for elderly and immuno-ompromised individuals. Induction of high titers of neutralizingntibodies by expression of a shorter domain from WNV E, such asIII domain alone, in a DNA vector will represent a safe vaccine forse in humans.

DIII domain of E is highly conserved among several WNV andEV strains and this region functions as a receptor binding domain,orming a continuous polypeptide segment that can fold indepen-ently. Certain mutations within DIII have shown to affect virulencend tropism of flaviviruses [18–20]. Hence this domain plays crit-cal roles in the flaviviral life cycle. rDIII is quite a stable proteinnd hence can become an attractive antigen. Recombinant flaviviralIII fragments have been shown to be immunogenic and protective

n mice challenged with the respective virulent viruses, under-ining the suitability of DIII-based vaccine formulations againstaviviruses. However, earlier attempts made it clear that a relativelyigh concentration of rDIII was needed for induction of neutralizingntibody responses, indicating that rDIII is poorly immunogenic.he results from our study agree with this possibility that DIIIrotein itself may not be a better vaccine candidate. One possible

nterpretation is that the percentage of solubilized proteins thatolded correctly might be low and the purification procedures fol-owed by different investigators also may affect immunogenicity ofDIII. Without IL15 support, we could not find a satisfactory anti-ody response from the monomeric DIII-expressing constructs. Inhis study, we showed enhanced level of DIII antibody-producingells as well as antibody production by using only 20 �g of vac-ine DNA per animal. Hence DNA vaccine in combination with IL15lasmid could represent a better, more effective alternate for immu-ization against JEV and WNV.

Collectively, the results obtained in the present study indicatehat DIII domain is an interesting well-defined vaccine candidatehat in combination with a good adjuvant should be studied inhe context of generating protective immunity against WNV andEV. Four- to five-fold increases in the level of antibody-producingells due to the co-stimulation of IL15 provided direct evidence thatL15 can induce the production of neutralizing antibody throughnhancing antibody-producing memory B cells against WNV as wells JEV. In contrary to the rationale that we set forth for makinghimeric DIII protein, the resulting chimeric protein did not serves a better antigen as noted from failed antibody response and theeast number of antibody-producing memory B cells. One possibleeason could be that introduction of epitopes from WNV DIII intoEV DIII might have affected the conformation status that is neededor proper presentation for eliciting a strong immune response orhat the protein was less stable due to folding controls. This reportemonstrates a method by which an optimized cytokine plasmid-elivered IL15 could enhance memory B cells specific to flaviviralIII which is known for its role in inducing neutralizing antibody

esponse. Taken together, these data suggest a role for IL15 as aossible candidate adjuvant for WNV/JEV DNA vaccines. Additionaltudy of this approach in non-human primate models of infections warranted.

cknowledgements

This work was supported in part by NIH grants awarded to Prof.avid Weiner. The laboratory notes possible commercial conflictsssociated with this work, which may include: Wyeth, VGX, BMS,irxsys, Ichor, Merck, Althea, and Aldeveron and possibly others.

[

[

e 27 (2009) 4370–4380 4379

References

[1] Solomon T. Exotic and emerging viral encephalitides. Curr Opin Neurol2003;16(3):411–8.

[2] Solomon T. Recent advances in Japanese encephalitis. J Neurovirol2003;9(2):274–83.

[3] Harrington T, Kuehnert MJ, Kamel H, Lanciotti RS, Hand S, Currier M, etal. West Nile virus infection transmitted by blood transfusion. Transfusion2003;43(8):1018–22.

[4] Armstrong WS, Bashour CA, Smedira NG, Heupler FA, Hoeltge GA, MawhorterSD, et al. A case of fatal West Nile virus meningoencephalitis associatedwith receipt of blood transfusions after open heart surgery. Ann Thorac Surg2003;76(2):605–7.

[5] Shepherd JC, Subramanian A, Montgomery RA, Samaniego MD, Gong G,Bergmann A, et al. West Nile virus encephalitis in a kidney transplant recipient.Am J Transplant 2004;4(5):830–3.

[6] Pealer LN, Marfin AA, Petersen LR, Lanciotti RS, Page PL, Stramer SL, et al. Trans-mission of West Nile virus through blood transfusion in the United States in2002. N Engl J Med 2003;349(13):1236–45.

[7] Brinton MA. The molecular biology of West Nile Virus: a new invader of thewestern hemisphere. Ann Rev Microbial 2002;56:371–402.

[8] Kumar R, Tripathi P, Singh S, Bannerji G. Clinical features in children hospitalizedduring the 2005 epidemic of Japanese encephalitis in Uttar Pradesh, India. ClinInfect Dis 2006;43(2):123–31.

[9] Murgod UA, Muthane UB, Ravi V, Radhesh S, Desai A. Persistent movementdisorders following Japanese encephalitis. Neurology 2001;57(12):2313–5.

[10] Halstead SB, Jacobson J. Japanese encephalitis. Adv Virus Res 2003;61:103–38.[11] Schioler KL, Samuel M, Wai KL. Vaccines for preventing Japanese encephalitis.

Cochrane Database Syst Rev 2007;3. CD004263.12] Appaiahgari MB, Saini M, Rauthan M, Jyoti, Vrati S. Immunization with recombi-

nant adenovirus synthesizing the secretory form of Japanese encephalitis virusenvelope protein protects adenovirus-exposed mice against lethal encephalitis.Microbes Infect 2006;8(1):92–104.

13] Bharati K, Vrati S. Japanese encephalitis: development of new candidate vac-cines. Expert Rev Anti Infect Ther 2006;4(2):313–24.

[14] Hoke CH, Nisalak A, Sangawhipa N, Jatanasen S, Laorakapongse T, Innis BL, et al.Protection against Japanese encephalitis by inactivated vaccines. N Engl J Med1988;319(10):608–14.

[15] Takahashi H, Pool V, Tsai TF, Chen RT, The VAERS Working Group. Adverse eventsafter Japanese encephalitis vaccination: review of post-marketing surveillancedata from Japan and the United States. Vaccine 2000;18(26):2963–9.

[16] Diagana M, Preux PM, Dumas M. Japanese encephalitis revisited. J Neurol Sci2007;262(1–2):165–70.

[17] Lindenbach BD, Rice CM. Molecular biology of flaviviruses. Adv Virus Res2003;59:23–61.

[18] Heinz FX, Mandl CW, Holzmann H, Kunz C, Harris BA, Rey F, et al. The flavivirusenvelope protein E: isolation of a soluble form from tick-borne encephalitisvirus and its crystallization. J Virol 1991;65(10):5579–83.

[19] Nybakken GE, Oliphant T, Johnson S, Burke S, Diamond MS, Fremont DH. Struc-tural basis of West Nile virus neutralization by a therapeutic antibody. Nature2005;437(7059):764–9.

20] Oliphant T, Engle M, Nybakken GE, Doane C, Johnson S, Huang L, et al. Develop-ment of a humanized monoclonal antibody with therapeutic potential againstWest Nile virus. Nat Med 2005;11(5):522–30.

21] Martina BE, Koraka P, van den Doel P, Rimmelzwaan GF, Haagmans BL, OsterhausAD. DC-SIGN enhances infection of cells with glycosylated West Nile virusinvitro and virus replication in human dendritic cells induces production of IFN-alpha and TNF-alpha. Virus Res 2008;135(1):64–71.

22] Martina BE, Koraka P, van den Doel P, van Amerongen G, Rimmelzwaan GF,Osterhaus AD. Immunization with West Nile virus envelope domain III protectsmice against lethal infection with homologous and heterologous virus. Vaccine2008;26(2):153–7.

23] Chu JH, Chiang CC, Ng ML. Immunization of flavivirus West Nile recombinantenvelope domain III protein induced specific immune response and protectionagainst West Nile virus infection. J Immunol 2007;178(5):2699–705.

24] Mota J, Acosta M, Argotte R, Figueroa R, Mendez A, Ramos C. Induction of pro-tective antibodies against dengue virus by tetravalent DNA immunization ofmice with domain III of the envelope protein. Vaccine 2005;23(26):3469–76.

25] Mason PW, Dalrymple JM, Gentry MK, McCown JM, Hoke CH, Burke DS, et al.Molecular characterization of a neutralizing domain of the Japanese encephali-tis virus structural glycoprotein. J Gen Virol 1989;70(Pt 8):2037–49.

26] Alka, Bharati K, Malik YP, Vrati S. Immunogenicity and protective efficacy of theE. coli-expressed domain III of Japanese encephalitis virus envelope protein inmice. Med Microbiol Immunol 2007;196(4):227–31.

27] Gaschen B, Taylor J, Yusim K, Foley B, Gao F, Lang D, et al. Diversity considerationsin HIV-1 vaccine selection. Science (New York, NY) 2002;296(5577):2354–60.

28] Yan J, Yoon H, Kumar S, Ramanathan MP, Corbitt N, Kutzler M, et al.Enhanced cellular immune responses elicited by an engineered HIV-1 subtypeB consensus-based envelope DNA vaccine. Mol Ther 2007;15(2):411–21.

29] Laddy DJ, Yan J, Khan AS, Andersen H, Cohn A, Greenhouse J, et al.

Electroporation of synthetic DNA antigens offers protection in nonhumanprimates challenged with highly pathogenic avian influenza virus. J Virol2009;83:4624–30.

31] Muthumani K, Lankaraman KM, Laddy DJ, Sundaram SG, Chung CW, Sako E, et al.Immunogenicity of novel consensus-based DNA vaccines against Chikungunyavirus. Vaccine 2008;26(40):5128–34.

4 Vaccin

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[cacy and durability of a recombinant subunit West Nile vaccine candidate in

380 M.P. Ramanathan et al. /

32] Kumar S, Yan J, Muthumani K, Ramanathan MP, Yoon H, Pavlakis GN, et al.Immunogenicity testing of a novel engineered HIV-1 envelope gp140 DNA vac-cine construct. DNA Cell Biol 2006;25(7):383–92.

33] Yan J, Harris K, Khan AS, Draghia-Akli R, Sewell D, Weiner DB. Cellular immunityinduced by a novel HPV18 DNA vaccine encoding an E6/E7 fusion consensusprotein in mice and rhesus macaques. Vaccine 2008;26(40):5210–5.

34] Laddy DJ, Yan J, Kutzler M, Kobasa D, Kobinger GP, Khan AS, et al. Hetero-subtypic protection against pathogenic human and avian influenza virusesvia in vivo electroporation of synthetic consensus DNA antigens. PLoS ONE2008;3(6):e2517.

35] Draghia-Akli R, Khan AS, Brown PA, Pope MA, Wu L, Hirao L, et al. Parametersfor DNA vaccination using adaptive constant-current electroporation in mouseand pig models. Vaccine 2008;26(40):5230–7.

36] Nishimura H, Fujimoto A, Tamura N, Yajima T, Wajjwalku W, Yoshikai Y. Anovel autoregulatory mechanism for transcriptional activation of the IL-15 geneby a nonsecretable isoform of IL-15 generated by alternative splicing. Faseb J2005;19(1):19–28.

37] Yajima T, Nishimura H, Sad S, Shen H, Kuwano H, Yoshikai Y. A novel role ofIL-15 in early activation of memory CD8+ CTL after reinfection. J Immunol2005;174(6):3590–7.

38] Ugen KE, Kutzler MA, Marrero B, Westover J, Coppola D, Weiner DB, et al. Regres-sion of subcutaneous B16 melanoma tumors after intratumoral delivery of anIL-15-expressing plasmid followed by in vivo electroporation. Cancer Gene Ther2006;13(10):969–74.

39] Kutzler MA, Robinson TM, Chattergoon MA, Choo DK, Choo AY, Choe PY, et al.Coimmunization with an optimized IL-15 plasmid results in enhanced functionand longevity of CD8 T cells that are partially independent of CD4 T cell help. JImmunol 2005;175(1):112–23.

40] Schluns KS, Lefrancois L. Cytokine control of memory T-cell development andsurvival. Nat Rev 2003;3(4):269–79.

[

e 27 (2009) 4370–4380

[41] Waldmann TA. The biology of interleukin-2 and interleukin-15: implicationsfor cancer therapy and vaccine design. Nat Rev 2006;6(8):595–601.

42] Ohteki T, Suzue K, Maki C, Ota T, Koyasu S. Critical role of IL-15-IL-15R forantigen-presenting cell functions in the innate immune response. Nat Immunol2001;2(12):1138–43.

43] Ramanathan MP, Chambers JA, Taylor J, Korber BT, Lee MD, Nalca A, et al. Expres-sion and evolutionary analysis of West Nile virus (Merion strain). J Neurovirol2005;11:544–56.

44] Oya A, Kurane I. Japanese encephalitis for a reference to international travelers.J Travel Med 2007;14(4):259–68.

45] Hemmerter S, Slapeta J, van den Hurk AF, Cooper RD, Whelan PI, Russell RC, etal. A curious coincidence: mosquito biodiversity and the limits of the Japaneseencephalitis virus in Australasia. BMC Evol Biol 2007;7:100.

46] Solomon T, Dung NM, Wills B, Kneen R, Gainsborough M, Diet TV, etal. Interferon alfa-2a in Japanese encephalitis: a randomised double-blindplacebo-controlled trial. Lancet 2003;361(9360):821–6.

[47] Solomon T, Ni H, Beasley DW, Ekkelenkamp M, Cardosa MJ, Barrett AD. Ori-gin and evolution of Japanese encephalitis virus in southeast Asia. J Virol2003;77(5):3091–8.

48] Tsarev SA, Sanders ML, Vaughn DW, Innis BL. Phylogenetic analysis suggestsonly one serotype of Japanese encephalitis virus. Vaccine 2000;18(Suppl.2):36–43.

49] Watts DM, Tesh RB, Siirin M, Rosa AT, Newman PC, Clements DE, et al. Effi-

protecting hamsters from West Nile encephalitis. Vaccine 2007;25(15):2913–8.

50] Lieberman MM, Clements DE, Ogata S, Wang G, Corpuz G, Wong T, et al. Prepara-tion and immunogenic properties of a recombinant West Nile subunit vaccine.Vaccine 2007;25(3):414–23.