DA

SUa

b

a

ARRAA

KCLHHC

1

wvrppntsnrcctic

f

0d

Vaccine 27 (2009) 6582–6588

Contents lists available at ScienceDirect

Vaccine

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

evelopment of candidate combination vaccine for hepatitis E and hepatitis B:liposome encapsulation approach

hubham Shrivastavaa, Kavita S. Lolea, Anuradha S. Tripathya,mesh S. Shaligramb, Vidya A. Arankallea,∗

Hepatitis Department, National Institute of Virology, 130/1, Sus Road, Pashan, Pune 411021, Maharastra, IndiaSerum Institute of India Ltd., 212/2, Hadapsar, Off Soli Poonawala Road, Pune, India

r t i c l e i n f o

rticle history:eceived 9 June 2009eceived in revised form 27 July 2009ccepted 7 August 2009vailable online 9 September 2009

eywords:ombination vaccineiposomeepatitis Eepatitis B

a b s t r a c t

To reduce extra injections, cost and ensure better coverage, use of combination vaccines is preferable.An attempt was made to evaluate the encapsulation of hepatitis E virus neutralizing epitope (NE) regionand hepatitis B virus surface antigen (HBsAg) in liposomes as DNAs, proteins and DNA + protein. Micegroups were immunized with different liposome-encapsulated formulations and monitored for anti-HEV and anti-HBs titres, IgG subtypes, antigen-specific lymphocyte proliferation and cytokine levels. Theprotective levels of anti-HBs and in vitro virus-binding capacity of anti-HEV antibodies were assessed.Liposome-encapsulated DNA either singly or in combination did not elicit antibody response. Anti-HEVand anti-HBs IgG titres of individual component of protein alone (Lipo-E-P/Lipo-B-P) or DNA + protein for-mulations (Lipo-E-DP/Lipo-B-DP) were comparable to respective titres in combination vaccine of protein(Lipo-BE-P) and DNA + protein formulations (Lipo-BE-DP). IgG1 levels were significantly higher in Lipo-

o-delivery approach BE-P group whereas, equivalent levels of IgG1 and IgG2a were observed in Lipo-BE-DP group against bothcomponents of the vaccine. Combination vaccine group showed mixed Th1/Th2 cytokine profile. Lipo-

BsAgnd n

some entrapped NE and Hto both the components a

. Introduction

Viral hepatitis is an important cause of morbidity and mortalityorldwide, though the incidence with respect to different viruses

aries in different geographical areas. In addition to reducing theisk of exposure to these viruses, prophylactic immunization haslayed crucial role in controlling some of these infections [1,2]. Atresent, potent vaccines for hepatitis A and B as well as a combi-ation vaccine is commercially available [3,4]. A successful clinicalrial for hepatitis E vaccine has been reported [5,6]. Despite exten-ive efforts worldwide, availability of hepatitis C vaccine in theear future seems remote [7]. Being dependent on hepatitis B viruseplication for its multiplication, control of HBV would automati-ally eliminate hepatitis delta virus (HDV) infections. Thus, if wean develop a combination vaccine for hepatitis A, B and E, four ofhe five major hepatitis viruses can be controlled. This is especially

mportant for developing countries wherein, cost as well as vaccineoverage are major concerns.

Considering the unavailability of a suitable cell-culture systemor most hepatitis viruses, recombinant proteins/DNAs are usu-

∗ Corresponding author. Tel.: +91 20 25871194; fax: +91 20 25871895.E-mail address: varankalle@yahoo.com (V.A. Arankalle).

264-410X/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2009.08.033

in protein and DNA + protein formats induce excellent immune responseeed to be evaluated in higher animals.

© 2009 Elsevier Ltd. All rights reserved.

ally evaluated as vaccine candidates. However, such candidates aregenerally less immunogenic and for the generation of robust andsustained humoral/cell-mediated immune responses required forviral clearance, development/evaluation and selection of adjuvantsbecomes obligatory. Alum is commonly used as an adjuvant for sev-eral bacterial and viral vaccines [8]. Among several adjuvants testedin animal studies, aluminium salts, liposomes and MF59 have beenapproved for use in humans [9,10].

Liposome-mediated vaccine delivery increases the shelf lifeof encapsulated antigens by protecting DNAs from nucleaseattack. Liposome also acts as immunoadjuvant by facilitating slowrelease of antigens [11–13]. Several studies have demonstratedthe modulation of immune response of the target antigens by theco-entrapment of cytokine expressing plasmids or CpG motifs inthe liposome for enhanced humoral and cell-mediated immunity[14–17].

Earlier, we showed superiority of liposome-encapsulated NE(neutralization epitopes present within the open reading frame-2, ORF2 of hepatitis E virus) DNA and corresponding protein over

DNA-prime protein boost approach in rhesus monkeys [18]. Thisapproach was extended to develop a combination vaccine for hep-atitis E and B using NE and hepatitis B surface antigen (HBsAg) asproteins and/or DNAs in various combinations. This study providesevidence for such a possibility.

accine

2

2

2

fiBgatsg

2

S(Pv

2

NrPADawwsT

2

fSwwowa

TD

S. Shrivastava et al. / V

. Materials and methods

.1. Recombinant DNA and protein preparations

.1.1. Construction of recombinant DNA moleculesThe small envelope gene of hepatitis B virus (685 bp) was ampli-

ed with primers containing restriction sites and was cloned inamHI/EcoRI sites of the pVAX1 vector (Invitrogen, Life Technolo-ies, Carlsbad). Similarly, NE of hepatitis E virus (450 bp) was clonedt BamHI site of the pVAX1 vector through blunt-end ligation. Bothhe recombinant plasmids were amplified in Escherichia coli DH5�train and purified using QIAGEN Endofree Mega Plasmid kit (Qia-en, Germany).

.1.2. Recombinant proteinsHBsAg protein (rHBsAg) expressed in yeast was obtained from

erum Institute of India, Pune, India. The recombinant NE proteinrNEp) was expressed in prokaryotic system and purified usingroBond purification resin (Invitrogen, Carlsbad) as described pre-iously [19].

.2. Liposome preparations

The DNA and corresponding proteins (either HBsAg and/orE) were co-entrapped into liposomes by dehydration and

ehydration method [20]. Phosphatidyl Choline (PC), Dioleoylhosphatidyl Ethanolamine (DOPE) and Dioleoyloxy Trimethylmmonium Propane (DOTAP) were used in the molar ratio of 4:2:1.NA + protein were mixed together and added to the small unil-mellar vesicle suspension in the mass ratio of 1:200. The mixtureas freeze-dried and rehydrated with PBS. Mice were inoculatedith 50 �l liposome suspension subcutaneously. Different lipo-

ome formulations used for mice immunization are depicted inable 1.

.3. Mice immunizations

This study was approved by the Institutional Ethical Committeeor animals and experiments were conducted as per the guidelines.ix- to eight-week-old female Swiss Albino mice (10 mice/group)

ere immunized subcutaneously at 0, 4, and 8 weeks intervalith various liposome formulations. Mice were bled from retro-

rbital plexus before immunization and at regular intervals of 2eeks after first dose. Sera were stored at −20 ◦C until further

nalysis.

able 1etails of liposome-encapsulated immunogens used for mice immunization.

Liposome formulations Dose/components

1. Lipo-pv 1 �g empty pVAX1 vector2. Lipo-E-D 1 �g NE DNA3. Lipo-E-P/Lipo-E-1P 1 �g NE protein4. Lipo-E-DP/Lipo-E-1D1P 1 �g NE DNA + 1 �g NE protein5. Lipo-B-D 1 �g HBV DNA6. Lipo-B-P/Lipo-B-1P 1 �g HBV protein (HBsAg)7. Lipo-B-DP/Lipo-B-1D1P 1 �g HBV DNA + 1 �g HBV protein8. Lipo-BE-D 1 �g HBV DNA + 1 �g NE DNA9. Lipo-BE-P/Lipo-BE-1P 1 �g HBV protein + 1 �g NE protein10. Lipo-BE-DP/Lipo-BE-1D1P 1 �g HBV DNA + 1 �g HBV

protein + 1 �g NE DNA + 1 �g NEprotein

11. Lipo-E-10D1P 10 �g NE DNA + 1 �g NE protein12. Lipo-B-10D1P 10 �g HBV DNA + 1 �g HBV protein13. Lipo-BE-10D1P 10 �g HBV DNA + 1 �g HBV

protein + 10 �g NE DNA + 1 �g NEprotein

14. Lipo-E10D-B1P 10 �g NE DNA + 1 �g HBV protein15. Lipo-B10D-E1P 10 �g HBV DNA + 1 �g NE protein

27 (2009) 6582–6588 6583

2.4. ELISA for antibody responses

For the detection and titration of anti-HEV antibodies,recombinant protein-based ELISAs were used according to theprotocol described earlier [19,21]. Coating antigens were eitherrORF2p or rNEp. Identical protocol was followed for anti-HBsELISA except that rHBsAg (5 �g/well) was used as the coatingantigen.

ELISAs for IgG subtype analysis were performed in the same wayas mentioned above except that after incubation with test serum,appropriate anti mouse isotype antibodies (Sigma Chemicals, St.Louis, MO) were added and incubated for 30 min at 37 ◦C. HRP-conjugated rabbit anti-goat IgG (Sigma Chemicals, St. Louis, MO)was used as the detector antibody.

AXSYM AUSAB assay (Abbott Laboratories) was used to quantifyanti-HBs levels in International Units according to the manufac-turer’s protocol. Antibody levels were expressed as mIU/ml.

2.5. In vitro virus-binding assay for HEV

In the absence of availability of an in vitro or in vivo neutraliza-tion assay for the detection of anti-HEV neutralizing antibodies, anELISA based virus-binding/neutralization assay was performed asdescribed earlier [19].

2.6. Lymphocyte proliferation assay (LPA)

Ten mice from each group (Table 1) were sacrificed at 2–3 weekspost-last dose and spleens were harvested. Single cell suspension ofsplenocytes was prepared in complete RPMI 1640 medium (Gibco,Invitrogen, USA) with 10% FBS (Gibco, Invitrogen, USA) contain-ing antibiotics. LPA was carried out as described previously [19].Results are expressed as stimulation indices (SI). Mice showing SIvalues ≥ 3 were considered responders.

2.7. Cytokine assay

Cytokine assay was carried out as described previously [22].For the detection of cytokines in culture supernatants, BD Cyto-metric Bead Array Mouse Th1/Th2 Cytokine and Inflammation kit(BD Biosciences, USA) were used according to the manufacturer’sguidelines.

2.8. Statistical analyses

To compare mean log titres between inter- and intra-groups,t-test was used. Mann–Whitney test was used to compare SI val-ues. Wilcoxon signed rank test and Mann–Whitney test were usedto compare cytokine levels between dependent and independentgroups, respectively. SPSS 9.0 software was used for statisticalanalyses. Non-responders in each group were included in theanalyses.

3. Results

3.1. Assessment of humoral response

3.1.1. Anti-HEV antibody responses in mice immunized withLipo-E-D, Lipo-E-P, Lipo-E-DP, Lipo-BE-D, Lipo-BE-P andLipo-BE-DP formulations

Use of either rORF2p or rNEp as detecting antigens in ELISA gave

similar results. None of the mice immunized with Lipo-E-D devel-oped anti-HEV antibodies (Table 2). Mice immunized with Lipo-E-Pand Lipo-E-DP showed 90% and 80% seroconversions respectively4 weeks post-first dose. Anti-HEV antibody titres and percent sero-conversions increased further with subsequent two doses (Table 2).

6584 S. Shrivastava et al. / Vaccine 27 (2009) 6582–6588

Table 2Percent seroconversions and titres of anti-HEV and anti-HBs antibodies at different time points in mice immunized with three doses of different formulations.

Liposome formulations % seroconversions (mean log anti-HEV titres ± SE)

Post-first dosea Post-second doseb Post-third dosec,d

Lipo-E-D 0 0 0Lipo-E-P 90 (1.78 ± 0.25) 100 (2.56 ± 0.25) 100 (3.14 ± 0.21)Lipo-E-DP 80 (2.06 ± 0.42) 100 (3.57 ± 0.14) 100 (3.63 ± 0.08)Lipo-BE-D 0 0 0Lipo-BE-P 70 (1.68 ± 0.39) 100 (3.05 ± 0.27) 100 (3.41 ± 0.14)Lipo-BE-DP 90 (2.81 ± 0.42) 90 (3.58 ± 0.38) 100 (3.72 ± 0.21)

Liposome formulations % seroconversions (mean log anti-HBs titres ± SE)

Post-first dosea Post-second doseb Post-third dosec,d

Lipo-B-D 0 0 0Lipo-B-P 90 (1.92 ± 0.32) 100 (3.20 ± 0.15) 100 (3.39 ± 0.19)Lipo-B-DP 80 (1.77 ± 0.34) 100 (3.68 ± 0.14) 100 (3.93 ± 0.16)Lipo-BE-D 0 0 10 (0.37 ± 0.07)Lipo-BE-P 70 (1.26 ± 0.25) 100 (3.08 ± 0.19) 100 (3.41 ± 0.16)Lipo-BE-DP 60 (1.61 ± 0.40) 90 (3.27 ± 0.26) 100 (3.54 ± 0.17)

a Percent seroconversion at 4 weeks post-first dose.b Percent seroconversion at 4 weeks post-second dose.

ons w

Mh(D1

Fwl(dt

c Percent seroconversion at 2–3 weeks after the last dose.d Mean log titres comparing at post-third dose between liposome formulati

ice immunized with Lipo-E-1D1P and Lipo-E-10D1P developed

igher anti-HEV titres than the mice immunized with Lipo-E-1Pp = 0.046 and 0.011) (Fig. 1a). However, 10-fold increase in NENA did not enhance anti-HEV titres when Lipo-E1D1P and Lipo-E-0D1P were compared (p = 0.09). Encapsulation of NE protein with

ig. 1. (a) Serum anti-HEV IgG titres at 2 weeks post-third dose in mice immunizedith different liposome formulations at 0, 4 and 8 weeks. Each symbol represents

og titre of a single mouse. × shows mean ± SE for the group. *p-Value < 0.05 (t-test).b) Serum anti-HBs IgG titres at 2 weeks post-third dose in mice immunized withifferent liposome formulations at 0, 4 and 8 weeks. Each symbol represents logitre of a single mouse. × shows mean ± SE for the group. *p-Value < 0.05 (t-test).

ith p-value ≤ 0.05 considered to be significant.

non-specific DNA (Lipo-B10D-E1P) induced comparable antibodytitres to NE protein encapsulated with 1 �g or 10 �g of specific DNA(Lipo-E-1D1P and Lipo-E-10D1P, p = 1.00 and 0.32, respectively)(Fig. 1a).

Among mice groups receiving combination vaccines, none ofthe mice immunized with Lipo-BE-D formulation seroconvertedto anti-HEV (Table 2). Four weeks post-first dose, 70% and 90%mice seroconverted to anti-HEV in Lipo-BE-P and Lipo-BE-DPgroups, respectively. Antibody titres and percent seroconversionsincreased further with subsequent doses (Table 2). Anti-HEV titresin mice immunized with Lipo-BE-P and Lipo-BE-DP were compara-ble (p = 0.24) (Fig. 1a). Anti-HEV titres in Lipo-BE-P and Lipo-BE-DPgroups were comparable to titres in mice immunized with Lipo-E-P and Lipo-E-DP, respectively (p = 0.30 and 0.69). Anti-HEV titresin Lipo-BE-10D1P and Lipo-BE-1P groups were also comparable(p = 0.09) (Fig. 1a).

3.1.2. Anti-HBs antibody responses in mice immunized withLipo-B-D, Lipo-B-P, Lipo-B-DP, Lipo-BE-D, Lipo-BE-P andLipo-BE-DP formulations

None of the mice immunized with Lipo-B-D developed anti-HBsantibodies (Table 2). Mice immunized with Lipo-B-P and Lipo-B-DP showed 90% and 80% seroconversions, respectively, to anti-HBsantibodies 4 weeks post-first dose (Table 2). The anti-HBs titres inLipo-B-1D1P and Lipo-B-10D1P groups were significantly higherthan that of Lipo-B-1P group (p = 0.039 and 0.015) (Fig. 1b). Encap-sulation of non-specific DNA (Lipo-E10D-B1P) resulted in higherantibody titres as well when compared to the Lipo-B-1P group(p = 0.003). Comparable antibody titres were recorded in Lipo-E10D-B1P and Lipo-B-10D1P (p = 0.91).

Only 10% mice immunized with Lipo-BE-D developed anti-HBsantibodies at 4 weeks post-third dose (Table 2). In Lipo-BE-P andLipo-BE-DP groups, 4 weeks after the first dose, 70% and 60% miceseroconverted to anti-HBs antibodies. Antibody titres and per-cent seroconversions increased further with subsequent two doses(Table 2). Anti-HBs titres in mice immunized with Lipo-BE-P and

Lipo-BE-DP were comparable (p = 0.61). Anti-HBs titres of Lipo-BE-Pand Lipo-BE-DP groups were comparable to Lipo-B-P and Lipo-B-DPgroups, respectively (p = 0.91 and 0.11). Synergistic effect of co-delivery of DNA + protein was also evident in Lipo-BE-10D1P whencompared with that of Lipo-BE-1P (p = 0.0005) (Fig. 1b).

S. Shrivastava et al. / Vaccine 27 (2009) 6582–6588 6585

Fig. 2. (a) Serum anti-HEV IgG isotype titres at 2 weeks post-third dose in miceimmunized with different liposome formulations at 0, 4 and 8 weeks. Values arerIlt

3

3

wi(pintl(

3

tLgoo

3

3i

cwsl

Fig. 3. (a) Cytokines profile in splenocytes after in vitro stimulation with NE antigenat 2 weeks post-third dose in mice immunized with different liposome formulationsat 0, 4 and 8 weeks. Values are represented as mean (pg/ml) ± SE. *p-Value < 0.05

inclusion of DNA led to further increase in IL2 levels in Lipo-E-DPgroup as compared to Lipo-E-P (p = 0.002) (Table 3).

In Lipo-BE-D group, no significant increase in cytokines levelswas observed as compared to pVAX1 (Fig. 3a). In both Lipo-BE-P and Lipo-BE-DP groups, IL2 levels were significantly higher than

Table 3IL2 and IL4 cytokine levels in splenocytes at 2 weeks post-third dose of immuniza-tion. Values are represented as mean (pg/ml) ± SEa.

Liposome formulations Cytokine [mean (pg/ml) ± SE] valuesafter stimulation with NE

IL2 IL4

Lipo-pv 1.93 ± 0.90 0.97 ± 0.47Lipo-E-D 1.94 ± 0.50 0.60 ± 0.30Lipo-E-P 8.58 ± 4.01 1.35 ± 0.30Lipo-E-DP 50.65 ± 10.24 7.34 ± 5.92Lipo-BE-D 2.59 ± 0.45 1.38 ± 0.37Lipo-BE-P 7.66 ± 1.39 2.20 ± 0.19Lipo-BE-DP 9.00 ± 0.79 2.48 ± 0.44

Liposome formulations Cytokine [mean (pg/ml) ± SE] valuesafter stimulation with HBsAg

IL2 IL4

Lipo-pv 3.38 ± 0.58 1.80 ± 0.43Lipo-B-D 3.06 ± 0.62 1.21 ± 0.28

epresented as mean log 10 titre ± SE. *p-Value < 0.05 (t-test). (b) Serum anti-HBsgG isotype titres at 2 weeks post-third dose in mice immunized with differentiposome formulations at 0, 4 and 8 weeks. Values are represented as mean log 10itre ± SE. *p-Value < 0.05 (t-test).

.2. Antibody isotype profiling

.2.1. Hepatitis EIn mice immunized with Lipo-E-P group, IgG1 isotype titres

ere significantly higher than those of IgG2a (p = 0.003). Micemmunized with Lipo-E-DP also showed predominance of IgG1p = 0.002) with significantly higher levels of IgG2a as com-ared to Lipo-E-P group (p ≤ 0.001) (Fig. 2a). In Lipo-BE-P

mmunized mice group, predominance of IgG1 over IgG2a wasoted (p < 0.001). In Lipo-BE-DP immunized mice group, nei-her IgG1 nor IgG2a was predominant (p = 0.103) (Fig. 2a). IgG2aevels were significantly higher in Lipo-BE-DP than Lipo-BE-Pp = 0.006).

.2.2. Hepatitis BIn mice immunized with Lipo-B-P group, the predominant iso-

ype was IgG1 (p = 0.016) whereas IgG2a was predominant inipo-B-DP immunized mice group (p = 0.006) (Fig. 2b). In Lipo-BE-Proup, predominance of IgG1 over IgG2a was noted (p = 0.008) andn the other hand, in Lipo-BE-DP group, predominance of IgG2aver IgG1 was observed (p = 0.044) (Fig. 2b).

.3. Assessment of cellular immune response

.3.1. Cytokine profile in response to rNEp and rHBsAg antigensn mice immunized with different liposome formulations

In response to in vitro stimulation with both antigens, weompared Th1 and Th2 cytokines levels in mice immunizedith different liposome formulations. In the absence of antigen-

timulations, the levels of cytokines were below the detectionimits.

(Mann–Whitney test). (b) Cytokines profile in splenocytes after in vitro stimulationwith HBsAg at 2 weeks post-third dose in mice immunized with different liposomeformulations at 0, 4 and 8 weeks. Values are represented as mean (pg/ml) ± SE.*p-Value ≤ 0.05 (Mann–Whitney test).

3.3.1.1. Hepatitis E. No significant increase in cytokine levels wasnoted in Lipo-E-D as compared to pVAX1 group (Fig. 3a). In bothLipo-E-P and Lipo-E-DP groups, Th1 immune profile was evidentas levels of IL2 and IFN� were significantly higher than IL4 andIL10, respectively [p = 0.012 and 0.012] (Fig. 3a, Table 3). However,

Lipo-B-P 73.01 ± 23.92 1.38 ± 0.32Lipo-B-DP 256.21 ± 98.12 0.59 ± 0.29Lipo-BE-D 3.19 ± 0.54 1.53 ± 0.47Lipo-BE-P 33.69 ± 6.62 3.54 ± 0.35Lipo-BE-DP 82.23 ± 21.16 3.80 ± 0.18

a n = 8 mice per group.

6586 S. Shrivastava et al. / Vaccine 27 (2009) 6582–6588

Table 4Lymphocyte proliferative response in mice immunized with different liposome formulations at 0, 4 and 8 weeks. Stimulation index ≥ 3 were considered as responder to theparticular antigen.

Approach Recall antigen No. of mice tested Responders (%) Median Range

Lipo-pv 5 0 (0%) 0.80 0.38–1.19Lipo-E-D 8 5 (62.5%) 3.17 0.76–9.47Lipo-E-P 8 5 (62.5%) 4.04 0.63–24.99Lipo-E-DP 7 5 (71.4%) 4.19 0.39–62.68Lipo-BE-D NE 6 4 (66.6%) 4.23 0.92–37.28Lipo-BE-P 6 4 (66.6%) 4.12 1.72–10.75Lipo-BE-DP 7 6 (85.7%) 5.35 0.93–15.69

Lipo-pv 5 0 (0%) 0.58 0.37–1.27Lipo-B-D 8 1(12.5%) 1.01 0.65–4.71Lipo-B-P 8 5 (62.5%) 5.04 2.09–16.20Lipo-B-DP HBsAg 8 6 (75%) 7.29 0.95–23.81Lipo-BE-D 6 2 (33.3%) 2.38 0.63–21.90

IBtpsr

3oILgDcILt

(Lhb0iL

3m

3

a6dsLmpLa

3

Lag

Lipo-BE-P 6Lipo-BE-DP 7

L4 (p = 0.012). The levels of IFN� and IL10 were comparable in Lipo-E-P group (p = 0.07), whereas IFN� levels were significantly higherhan IL10 in Lipo-BE-DP group (p = 0.017) (Fig. 3a). Th2 immunerofile was also evident in Lipo-BE-P immunized mice group withignificant rise in IL4 and IL10 levels as compared to Lipo-E-P,espectively (p = 0.031 and 0.006).

.3.1.2. Hepatitis B. No significant increase in cytokine levels wasbserved in Lipo-B-D group when compared to pVAX1 (Fig. 3b).L2 levels were significantly higher than IL4 in both Lipo-B-P andipo-B-DP groups, respectively (p = 0.012 and 0.012). In Lipo-B-Proup, levels of IFN� and IL10 were comparable (p = 0.21) and afterNA inclusion (Lipo-B-DP), Th1 profile was evident with signifi-ant increase in IFN� levels compared to IL10 (p = 0.017) (Fig. 3b).ncreased shift towards Th1 immune profile was also evident inipo-B-DP group with significant increase in IL2 levels as comparedo Lipo-B-P (p = 0.046) (Table 3).

As compared to pVAX1, significant increase in IL10 levelsp = 0.014) was observed in Lipo-BE-D group (Fig. 3b). In bothipo-BE-P and Lipo-BE-DP groups, IL2 levels were significantlyigher than IL4 (p = 0.012). IFN� and IL10 levels were compara-le in Lipo-BE-P and Lipo-BE-DP groups, respectively (p = 0.57 and.26) (Fig. 3b). Th2 immune profile was also evident in Lipo-BE-P

mmunized mice with significant rise in IL4 levels as compared toipo-B-P group (p = 0.001) (Fig. 3b and Table 3).

.4. Lymphocyte proliferative responses to rNEp and rHBsAg inice immunized with different liposome formulations

.4.1. Hepatitis EMice immunized with pVAX1 vector did not respond to rNEp

nd rHBsAg. In Lipo-E-D group, T cell activation was observed in2.5% mice against rNEp, though anti-HEV antibodies were notetected in this group. In both Lipo-E-P and Lipo-E-DP groups,imilar T cell proliferation response was observed (Table 4). Inipo-BE-D group, again T cell activation was observed in 67% ofice in spite of absence of anti-HEV antibodies. Comparable T cell

roliferation was recorded in mice immunized with Lipo-BE-P andipo-BE-DP against rNEp. Thus, 62–85% mice responded to rNEp inll different formulations (Table 4).

.4.2. Hepatitis BT cell proliferation was weakly observed in both Lipo-B-D and

ipo-BE-D groups, similar to antibody responses (Tables 2 and 4)gainst rHBsAg. In Lipo-B-P, Lipo-B-DP, Lipo-BE-P and Lipo-BE-DProups, 62–85% mice responded to rHBsAg (Table 4).

4 (66.6%) 3.51 1.86–24.726 (85.7%) 7.59 1.95–46.02

3.5. Evaluation of protective immune response

3.5.1. Hepatitis ESerum samples showing >50% reduction in OD values in ELISA

after incubation with the hepatitis E virus was considered as anevidence for the presence of virus-binding/neutralizing antibodies.Inhibition of almost 100% was exhibited by mice sera immunizedwith different formulations, demonstrating ability of these anti-bodies to bind to the native virus. When virus-binding capacity ofserially diluted mice sera was compared with anti-ORF2 IgG titresin ELISA, similar titres were observed.

3.5.2. Hepatitis BIn the absence of anti-HBV neutralizing antibody assay on

account of unavailability of a robust in vitro system or small ani-mal model for the virus, based on the efficacy trials of the vaccinein humans, it is universally accepted that 10 mIU/ml representsprotective titre and all studies adhere to this criteria [23]. Anti-HBs titres from all mice have therefore been expressed as mIU/mlreflecting extent of protective titres. Anti-HBs levels (mIU/ml)in different groups were: Lipo-B-P (983.17 ± 227.14), Lipo-B-DP(3638.00 ± 987.37), Lipo-BE-P (2752.87 ± 811.40) and Lipo-BE-DP (4612.00 ± 422.72). Anti-HBs levels were significantly higherin Lipo-B-DP group as compared to Lipo-B-P group (p = 0.026)whereas, comparable antibody levels were documented in Lipo-BE-P to that of Lipo-BE-DP (p = 0.062). No significant rise in antibodylevels was observed in Lipo-BE-P and Lipo-BE-DP to that of Lipo-B-P and Lipo-B-DP, respectively (p = 0.06 and 0.38). All mice showedmore than 200 mIU/ml anti-HBs antibodies, well above the protec-tive level of 10 mIU/ml.

4. Discussion

Following successful documentation of efficacy of encapsula-tion of NE DNA and corresponding protein of HEV in liposomesin rhesus monkeys [18], we continued our efforts in developing acombination vaccine for hepatitis E and B employing this approachwith preliminary evaluation in mice. In this study, we used NE DNAand NE protein of HEV (based on our results) and S gene of HBVand HBsAg (proved beyond doubt to be excellent immunogen foralmost three decades) as the components of the candidate combi-nation vaccine. The combination included DNAs or proteins or both

DNAs and proteins encapsulated together in liposomes.

The possibility of using DNA alone was negated by the absenceof seroconversion in mice immunized with 1 �g and 10 �g NEDNA or S DNA singly or in combination of both. Evaluation ofproteins yielded encouraging results. With 1 �g of recombinant

accine

pteDoDr(rnwa

orkstrewtbImwdftirIooewFaFat

B(niIiIawamC[

cotLHifTcs

S. Shrivastava et al. / V

roteins, 90% seroconversion was observed against NE and HBsAg,he combination resulted in 70% seroconversion with respect toach immunogen 4 weeks after first dose (Table 2). Addition ofNA did not alter the results. Anti-HEV and anti-HBs IgG titresf individual component of protein alone (Lipo-E-P/Lipo-B-P) orNA + protein formulations (Lipo-E-DP/Lipo-B-DP) were compa-

able to the respective titres in combination vaccine of proteinLipo-BE-P) and DNA + protein formulations (Lipo-BE-DP). Theseesults clearly show that it is possible to get rid of the DNA compo-ent (requiring dealing with lot of regulatory issues) of the vaccineithout compromising on the performance of the individual as well

s combination vaccines.We further evaluated the role of encapsulation of higher amount

f specific DNAs and non-specific DNAs in enhancing antibodyesponse. When specific DNA quantity was raised from 1 to 10 �geeping protein constant at 1 �g (Lipo-E10D1P/Lipo-B10D1P),uperiority of co-delivery approach was evident when comparedo the respective protein counterparts (Lipo-E-1P/Lipo-B-1P) aseported by Laing et al. [24] in case of influenza. However, Laingt al. [24] did not observe superiority of co-delivery approachhen similar liposome formulation was used for HBsAg. They fur-

her added mannosylated lipid in the formulation and observedetter immune response as compared to Engerix-B vaccine [24].n combination vaccine, superiority of co-delivery approach was

aintained in case of hepatitis B, whereas no such enhancementas observed with the E component (Fig. 1a, b). Concentrationependent enhancing effect of specific DNA on the antibody titresor both hepatitis E and B were observed in single vaccine formula-ions. However, in combination vaccine when higher DNA amounts used, this effect was observed only for B and not for E probablyeflecting intrinsic properties of the antigens under investigation.nterestingly, even with non-specific DNA, similar rise in titres wasbserved only for hepatitis B and not for E, suggesting adjuvant rolef any DNA in modulating immune response against HBsAg. Gurselt al. [25] documented similar adjuvanting effect of plasmid DNAith co-entrapped DNA + HBsAg protein liposome formulations.

or prophylaxis, combination protein vaccine producing excellentntibody titres against both the components would be preferable.ormulations with the addition of different concentrations of DNAre worth pursuing as therapeutic candidate vaccines needing fur-her in-depth analysis.

Liposome encapsulation of proteins (Lipo-E-P/Lipo-B-P/Lipo-E-P) favored Th2 profile as evidenced by IgG subtype analysisFig. 2a, b). On the other hand, skewing towards Th1 profile wasoted after the addition of HBV DNA to HBsAg (Fig. 2b) suggesting

nvolvement of DNA in inducing cell-mediated immune response.mportantly, though the addition of NE DNA to NE protein resultedn significant increase of IgG2a levels (Fig. 2a), predominance ofgG1 was maintained. In combination vaccine (Lipo-BE-DP), bal-nced IgG1/IgG2a levels (NE) and predominance of IgG2a (HBsAg)ere observed. Encapsulation of non-specific DNA in liposomes

lso led to shift towards Th1 type (data not shown). Enhance-ent of antibody production and switch towards Th1 pathway by

pG motifs was shown for liposome-encapsulated HCV NS3 protein26].

With the inclusion of DNA, activation of Th1 pathway as indi-ated by the significant increase in IL2 and IFN� levels werebserved in Lipo-E-DP and Lipo-B-DP groups when compared tohe respective protein counterparts. Though both Lipo-B-P andipo-BE-P induced balanced Th1/Th2 cytokines with respect toBsAg, a shift from Th1 (Lipo-E-P) to a balanced Th1/Th2 cytokines

n combination protein vaccine may prove useful in protectionrom hepatitis E. In Lipo-BE-DP immunized mice, production ofh1 cytokines (NE) and balanced Th1/Th2 cytokines (HBsAg) indi-ates influence of HBsAg protein and/or DNA on the cytokinehift.

27 (2009) 6582–6588 6587

Discrepancy in the isotype analysis and cytokine profiles wasobserved in Lipo-E-P and Lipo-E-DP groups while all other com-binations showed good correlation. These results indicate that NEprotein has an intrinsic property of eliciting Th1 cytokines whereasHBsAg induces balanced Th1/Th2 cytokine levels. Ramakrishnaet al. [27] documented similar discrepancy in case of Japaneseencephalitis viral antigen immunized mice. In case of NE protein,probably early cytokines may be playing crucial role in determin-ing polarization of T cells which become resistant to any furtherchange in prevailing cytokine environment [27].

Activation of Th1 cytokines (IL2, IFN�) has been shown tobe a pre-requisite for protection against intracellular pathogenswhereas, Th2 cytokines play important role in antibody produc-tion [28]. We have documented elevation of IL2 and IFN� levelsin mice immunized with proteins (NE, HBsAg or combination) orDNA + protein (proteins and corresponding DNAs, either singly orin combination) formulations. Antiviral role of IFN� by loweringHBV mRNA levels in the livers of transgenic mice has been demon-strated. This observation supported the role of Th1 cytokines in therecovery from infection [29]. T cell proliferation was also observedin protein and DNA + protein formulations against both NE andHBsAg suggesting involvement of CD4 + T helper cells [30]. Inter-estingly, despite the absence of anti-HEV antibodies in Lipo-E-Dand Lipo-BE-D groups, T cell proliferative responses were observedin about 60% of mice. Thus, activation of cytokines and lymphocyteproliferation provides evidence for the generation of cell-mediatedimmunity by all the candidate vaccines.

An important issue with vaccine development is the availabilityof correlates of protection for the disease under consideration. Forboth hepatitis E and B, a direct correlation of antibody titres andprotection has been documented [6,23,31]. Except DNA, all otherformulations generated excellent antibody titres with respect toboth the antigens deserving evaluation in higher animals.

In conclusion, our data shows that a combination vaccine forhepatitis E and B is possible using liposome entrapment of NEand HBsAg as proteins or in combination with specific/non-specificDNAs. Considering the comparable immune responses generatedby protein alone components, this approach seems worth pursingwithout addressing several regulatory issues with use of DNA. Weneed to include and evaluate the hepatitis A component as well tohave a combination vaccine for hepatitis A, B and E. Vaccine can-didate with DNA component inducing robust antibody responsealong with enhanced cell-mediated immunity may be tried fortherapeutic applications.

Acknowledgements

The authors are grateful to Dr. A.C. Mishra, Director, NationalInstitute of Virology for all the encouragement. We also thank Dr.S.V. Kapre for his kind help in liposome formulations. Ms. ShubhamShrivastava would like to thank Indian Council of Medical Researchfor providing Senior Research Fellowship. Thanks are also due toMs. Hemangini for assistance in flow cytometry. We appreciate thetechnical assistance of Mr. Bipin Tilekar and Mr. Shirish Vaidya. Wewish to thank Mr. A.M. Walimbe for assistance in statistical analysis.

References

[1] Vogt TM, Wise ME, Bell BP, Finelli L. Declining hepatitis A mortality in the UnitedStates during the era of hepatitis A vaccination. J Infect Dis 2008;197:1282–8.

[2] Chang MH, Chen CJ, Lai MS, Hsu HM, Wu TC, Kong MS, et al. Universal Hepati-

tis B vaccination in Taiwan and the incidence of hepatocellular carcinoma inchildren. N Engl J Med 1997;336:1855–9.

[3] Nothdurft HD. Hepatitis A vaccines. Expert Rev Vaccines 2008;7:535–45.[4] Van Damme P, Van Herck K. A review of the efficacy, immunogenicity and

tolerability of a combined hepatitis A and B vaccine. Expert Rev Vaccines2004;3:249–67.

6 accine

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

588 S. Shrivastava et al. / V

[5] Krawczynski K. Recombinant hepatitis E virus vaccine: a successful Phase II trialin a disease-endemic region. Expert Rev Gastroenterol Hepatol 2007;1:239–42.

[6] Shrestha MP, Scott RM, Joshi DM, Mammen Jr MP, Thapa GB, Thapa N,et al. Safety and efficacy of a hepatitis E virus vaccine. New Engl J Med2007;356:895–903.

[7] Stoll-Keller F, Barth H, Fafi-Kremer S, Zeisel MB, Baumert TF. Developmentof hepatitis C virus vaccines: challenges and progress. Expert Rev Vaccines2009;8:333–45.

[8] Gupta RK. Aluminium compounds as vaccine adjuvants. Adv Drug Deliv Rev1998;32:155–72.

[9] Kovarik J, Siegrist. Optimization of vaccine responses in early life: the role ofdelivery systems and immunomodulators. Immunol Cell Biol 1998;76:222–36.

10] O’Hagan DT. MF59 is a safe and potent vaccine adjuvant that enhances protec-tion against influenza virus infection. Expert Rev Vaccines 2007;6:699–710.

11] Gregoriadis G, Saffie R, de Souza EJ. Liposome mediated DNA vaccination. FEBSLett 1997;402:107–10.

12] Perrie Y, Frederik PM, Gregoriadis G. Liposome mediated DNA vaccination: theeffect of vesicle composition. Vaccine 2001;19:3301–10.

13] Gregoriadis G, Eacon A, Wilson CW, McCormack E. A role for liposomes ingenetic vaccination. Vaccine 2002;(Suppl):B1–9.

14] Gursel M, Gregoriadis G. The immunological co-adjuvant action of liposomalinterleukin-2: the role of mode of localisation of the cytokine and antigen inthe vesicles. J Drug Target 1997;5:93–8.

15] Okada E, Sasaki S, Ishii N, Aoki I, Yasuda T, Nishioka K, et al. Intranasalimmunization of a DNA vaccine with 1L-12 and Granulocyte-MacrophageColony-Stimulating Factor (GM-CSF)-expressing plasmids in liposomesinduces strong mucosal and cell-mediated immune responses against HIV-1antigens. J Immunol 1997;159:3638–47.

16] Babai I, Samira S, Barenholz Y, Zakay-Rones Z, Kedar E. A novel influenza subunitvaccine composed of liposome-encapsulated haemagglutinin/neuraminidaseand IL-2 or GM-CSF. I. Vaccine characterization and efficacy studies in mice.Vaccine 1999;17:1223–38.

17] Joseph A, Louria-Hayon I, Plis-Finarov A, Zeira E, Zakay-Rones Z, Raz E, etal. Liposomal immunostimulatory DNA sequence (ISS-ODN) an efficient par-

enteral and mucosal adjuvant for influenza and hepatitis B vaccines. Vaccine2002;20:3342–54.

18] Arankalle VA, Lole KS, Deshmukh TM, Srivastava S, Shaligram US. Challengestudies in Rhesus monkeys immunized with candidate hepatitis E vaccines:DNA, DNA-prime-protein-boost and DNA-protein encapsulated in liposomes.Vaccine 2009;27:1032–9.

[

27 (2009) 6582–6588

19] Deshmukh TM, Lole KS, Tripathy AS, Arankalle VA. Immunogenicity of candi-date hepatitis E virus DNA vaccine expressing complete and truncated ORF2 inmice. Vaccine 2007;25:4350–60.

20] Gregoriadis G, McCormack B, Obrenovic M, Saffie R, Zadi B, Perrie Y. Vaccineentrapment in liposomes. Methods 1999;19:156–62.

21] Arankalle VA, Lole KS, Deshmukh TM, Chobe LP, Gandhe SS. Evaluation ofhuman (genotype 1) and swine (genotype 4)-ORF-2 based ELISAs for anti-HEVIgM and IgG detection in an endemic country and search for type 4 human HEVinfections. J Viral Hepat 2007;14:435–45.

22] Saravanabalaji S, Tripathy AS, Dhoot RR, Chadha MS, Kakrani AL, ArankalleVA. Viral load, antibody titers and recombinant open reading frame 2 protein-induced Th1/Th2 cytokines and cellular immune responses in self-limiting andfulminant hepatitis E. Intervirology 2009;52:78–85.

23] Jack AD, Hall AJ, Maine N, Mendy M, Whittle HC. What level of hepatitis Bantibody is protective? J Infect Dis 1999;179:489–92.

24] Laing P, Eacon A, McCormack E, Gregoriadis G, Frisch E, Schuber F. The co-delivery approach to liposomal vaccines: application to the development ofinfluenza A and hepatitis B vaccine candidates. J Liposome Res 2006;16:229–35.

25] Gursel M, Tunca S, Ozkan M, Ozcengiz G, Alaeddinoglu G. Immunoadjuvantaction of plasmid DNA in liposomes. Vaccine 1999;17:1376–83.

26] Jiao X, Wang Richard Y-H, Qiu Q, Alter Harvey J, Shih Wai-Kuo J. Enhancedhepatitis C virus NS3 specific Th1 immune responses induced by co-deliveryof protein antigen and CpG with cationic liposomes. J Gen Virol 2004;85:1545–53.

27] Ramakrishna C, Ravi V, Desai A, Subbakrishna DK, Shankar SK, Chandramuki A. Thelper responses to Japanese encephalitis virus infection are dependent on theroute of inoculation and the strain of mouse used. J Gen Virol 2003;84:1559–67.

28] Seder RA, Paul WE. Acquisition of lymphokine producing phenotype by CD4+T cells. Annu Rev Immunol 1994;12:635–73.

29] Mancini M, Hadchouel M, Tiollais P, Michel ML. Regulation of hepatitis Bvirus mRNA expression in a hepatitis B surface antigen transgenic mousemodel by IFNg secreting T cells after DNA based immunization. J Immunol1998;161:5564–70.

30] Chow YH, Chiang BL, Lee YL, Chi WK, Lin WC, Chen YT, et al. Development of

Th1 and Th2 populations and the nature of immune responses to hepatitis Bvirus DNA vaccines can be modulated by codelivery of various cytokine genes.J Immunol 1998;160:1320–9.

31] Centers for Disease Control. Protection against viral hepatitis: recommen-dations of the Immunization Practises Advisory Committee (ACIP). MMWRRecomm Rep 1990;39:1–26.