Vaccine 26 (2008) 3389–3394

Contents lists available at ScienceDirect

Vaccine

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

Poly(propyleneimine) dendrimer and dendrosome mediated geneticimmunization against hepatitis B

Tathagata Duttaa,b,∗, Minakshi Garga, Narendra K. Jaina

a Department of Pharmaceutical Sciences, Dr. Hari Singh Gour University, Sagar-470003, MP, Indiab School of Pharmacy, University of Queensland, St Lucia, Brisbane, QLD 4067, Australia

t reseasmidn, wactionotoxiand Pomesmumwedy givial fo

3 wascompine le

ationzation

a r t i c l e i n f o

Article history:Received 12 September 2007Received in revised form 21 April 2008Accepted 24 April 2008Available online 12 May 2008

Keywords:DendrimerDendrosomeGenetic immunizationDNA vaccineHepatitis BGene delivery

a b s t r a c t

The purpose of the presention against hepatitis B. Plhepatitis B surface antigein different ratios. Transfetransfection. Results of cyt(p < 0.05) higher for PPI 75for preparation of dendroswas found to possess optiof HBsAg in CHO cells shostudies were carried out bdendrimeric or dendrosomIgG2a, IgG2b, biweekly. DFsubclasses under study asdeveloped very high cytokTh1 mediated. Our observpDNA for genetic immuni

1. Introduction

Genetic immunization is a promising approach to vaccination,induces potent cell-mediated immune response and in certaincases potent antibody response against the encoded antigen [1–5].In most of the cases naked plasmid DNA was administered by intra-muscular route, although other routes (e.g. intranasal, intravaginal,etc.) have been also investigated [2,3]. Following intramuscu-lar injection, the interstitial DNAase degrades most of the DNA,with the remaining material being taken up by the cells that aretransfected episomally [4]. These, however, lack co-stimulatorymolecules and are therefore unable to act as professional antigen-presenting cells (APC). However, there are reports suggesting thatthe APCs are transfected directly, which then get activated andmigrate to the draining lymph nodes for antigen presentation[6]. Various delivery systems were tested to date for gene deliv-ery and genetic immunization, which includes cationic liposomes[7], nanoparticles [8], and dendrimers [9]. Dendrimers are highly

∗ Corresponding author. Tel.: +91 7582 264712; fax: +91 7582 264712.E-mail addresses: duttatathababu@yahoo.com, jnarendr@yahoo.co.in (T. Dutta).

0264-410X/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.doi:10.1016/j.vaccine.2008.04.058

arch work is to explore the potential of dendrosomes in genetic immuniza-DNA encoding pRc/CMV-HBs[S] (5.6 kb), encoding the small region of the

s complexed with 5th generation poly(propyleneimine) dendrimer (PPI)of CHO cells revealed that a ratio of 1:50 for pDNA:PPI was optimum for

city studies showed that the toxicity of PPI–DNA complex was significantlyPI 100 as compared to the other PPI–DNA complexes. PPI 50 was employedby reverse phase evaporation method. The dendrosomal formulation DF3vesicle size, zeta potential and entrapment efficiency. In vitro production

that DF3 possess maximum transfection efficiency. In vivo immunizationng a single intramuscular injection of 10 �g of plasmid DNA (pDNA) or itsrmulation to female Balb/c mice, followed by estimation of total IgG, IgG1,found to elicit maximum immune response in terms of total IgG and itsared to PPI 50 and pDNA at all time points. Animals immunized with DF3vel. Higher level of IFN-� suggests that the immune response was strictly

s clearly prove the superiority of dendrosomes over PPI–DNA complex andagainst hepatitis B.

© 2008 Elsevier Ltd. All rights reserved.

branched and reactive three-dimensional polymers, with all bonds

emanating from a central core. The term ‘dendrimer’ originatedfrom the Greek word ‘dendron’ meaning tree and suffix ‘mer’from meros, denoting smallest repeating units. In recent yearsdendrimers have attracted enormous attention among researchersworking in the field of biomedical sciences. Attractive features likenanoscopic size, highly controllable molecular weight, large num-ber of readily accessible terminal functional groups and possibilityof encapsulating a guest molecule in internal cavities give den-drimers a distinct edge over other polymers for the delivery of drugs[10]. The nanoscopic size not only helps the dendrimers to evadethe reticuloendothelial system (RES) of the body but also makethem extremely important for intracellular drug delivery. Inher-ent cationic charge and spherical shape associated with the amineterminated dendrimers make them highly potent for the deliveryof genes and immunogens [11]. Poly(propyleneimine) dendrimers(PPI) are highly branched, globular dendrimers with primary aminogroups at the periphery. Due to the presence of a definite numberof primary amino groups, the PPI dendrimers possess a finite posi-tive charge, which is responsible for condensation of the DNA andsubsequent transfection [12]. They are well-known cationic genedelivery vectors and form a stable complex with DNA through elec-

ine 26

3390 T. Dutta et al. / Vacc

trostatic interactions. The dendrimer–DNA complex due to its netpositive charge penetrates the cell membrane and thereby releasesthe DNA leading to subsequent gene expression [13–15]. However,the dendrimers are reported to possess hemolytic toxicity and cyto-toxicity owing to their polycationic nature. Moreover, interactionwith the oppositely charged macromolecules in plasma results inpremature release of DNA within the blood [13,14]. Also, degrada-tion of the plasmid DNA by DNAase present in the plasma leads topoor gene expression in vivo [15,16].

Dendrosomes are novel, vesicular, spherical, supramolecularentities, containing entrapped dendrimer–DNA complex, possess-ing negligible hemolytic toxicity, higher transfection efficiencyand better in vivo acceptability. The transfection efficiency andtoxicity of dendrosomes composed of polyamidoamine (PAMAM)dendrimers were investigated in detail in our laboratory [17].

The purpose of the present research work is to explore the poten-tial of dendrosomes in genetic immunization against hepatitis B.The PPI dendrimer–DNA complex would be entrapped within thedendrosomes as opposed to being complexed externally with suchvesicles to form lipoplexes. Apart from protecting the plasmid DNAfrom nuclease degradation, the dendrosomes would be a suitablemeans of delivering their contents directly to the APC [18–21]. Theextended release of the contents of the dendrosomes would be anadded advantage.

2. Materials and methods

2.1. Materials

Fifth generation poly(propyleneimine) dendrimers (PPI) weresynthesized by divergent approach and extensively character-ized as reported earlier [22,23]. Phosphatidylcholine, cholesteroland 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bro-mide (MTT) were obtained from Sigma Chemicals (St. Louis. MO).Diethyl ether and dimethyl sulfoxide (DMSO) were purchasedfrom Rankem (Ranbaxy Labs Ltd., India). Dulbecco’s modifiedEagle’s minimum essential medium (DMEM) was obtained fromNissui Pharmaceuticals (Tokyo, Japan). Foetal calf serum (10%FCS) was obtained from Biowhittaker (Walkersville, Maryland).Horseradish peroxidase-conjugated goat antimouse immunoglob-ulin IgG1, IgG2a and IgG2b, and foetal calf serum were obtainedfrom Sera-Lab (Crawly Down, Sussex, UK). Recombinant hepatitisB surface antigen (HBsAg) (S region; ayw subtype) was supplied by

Genzyme diagnostics (Kingshill, Kent, UK). All other reagents wereof analytical grade.

Plasmid pRc/CMV-HBs[S] (5.6 kb) expressing the sequence cod-ing for the S (small) region of HBsAg (subtype ayw) was obtainedfrom Aldevron (Fargo, ND, USA). Chinese Hamster Ovary (CHO) cellswere benevolent gift of Dr. T. Velpandyan, All India Institute ofMedical Sciences, New Delhi.

The protocols for animal experiments were duly approved bythe Institutional Animal Ethics Committee, Dr. H.S. Gour University,Sagar, India (vide letter No. Animal Ethics Comm/DB/98 dated 28July 2006; Registration No. 379/01/ab/CPCSEA). Female Balb/c mice,6–8 weeks old, were obtained from Experimental Animal facility,All India Institute of Medical Sciences, New Delhi.

2.2. Preparation of dendrimer–DNA complex

The dendrimer–DNA complex was prepared [24] by addition ofDNA to varying proportion of PPI in phosphate buffer saline (PBS;pH 7.4) and incubated at 25 ◦C for 30 min to allow complex forma-tion. PPI–DNA complexes of different ratios of PPI and pDNA wereprepared as shown in Table 1.

(2008) 3389–3394

Table 1Zeta potential of various dendrimeric formulations

S. no. Formulation code pRc:CMV HBS: PPIdendimer ratio

Zeta potential (mV)

1 PPI 5 1:5 +2.1 ± 0.072 PPI 10 1:10 +4.6 ± 0.053 PPI 20 1:20 +8.0 ± 0.164 PPI 40 1:40 +16.9 ± 0.235 PPI 50 1:50 +21.3 ± 0.336 PPI 75 1:75 +31.7 ± 0.727 PPI 100 1:100 +43 ± 1.21

Values represent mean ± S.D. (n = 3).

2.3. Zeta potential and particle size distribution

The zeta potential of various formulations of PPI–DNA complexhaving different ratios, was determined using Zetasizer 3000 HS(Malvern Instruments, UK).

2.4. Optimization of dendrimer DNA ratio

The amount of antigen secreted were determined using Chi-nese Hamster Ovary (CHO) cells transfected with PPI–DNA complexhaving different ratios of PPI and pRc/CMV-HBs[S]. Cells weremaintained in a 37 ◦C humidified 10% CO2 incubator with Mini-mal Essential Media (Hyclone, Logan, UT), 10% fetal bovine serum(Clontech, Palo Alto, CA), supplemented with non-essential aminoacids (Gibco, Gathersburg, MD). Briefly, CHO cells were seededat a density of 7 × 105 ml in 35 mm culture dishes (Fisher Scien-tific, Pittsburgh, PA) for 24 h, followed by a 3 h transfection withPPI–DNA complex. Cells were subsequently washed with completemedium followed by 48 h incubation. CHO cell culture super-natants were recovered and concentrated approximately five foldusing Centricon-10 filter units (Amicon, Bedford, MA). The level ofsecreted surface antigen was measured using a quantitative ELISA.

2.5. Enzyme linked immunosorbent assay (ELISA)

In vitro secreted HBsAg was measured using a quantitative cap-ture ELISA. Briefly, Immulon 4 HBX (Costar, Cambridge, MA) 96-wellmicrotiter plates were coated with monoclonal anti-HBs antibody(Aldevron, Fargo, ND) at 10 �g/ml diluted in coating buffer (0.05 Msodium bicarbonate, pH 9.6). Plates were allowed to incubate for1 h at room temperature. Following coating, plates were blocked

(50 mM Tris, 0.15 M NaCl, 1% BSA, pH 8.0) for 1/2 h at room tempera-ture. Concentrated CHO cell supernatants were applied in duplicateand incubated for 1 h at room temperature, after which a poly-clonal goat–anti-HBs antibody (Cliniqa Corp., Fallbrook, CA) wasapplied. Rabbit–anti-goat IgG conjugated to alkaline phosphatase(Cappel, Aurora, OH) was added for 1 h at room temperature fol-lowed by p-nitrophenylphosphate substrate (Sigma, St. Louis, MO).Plates were read at 405 nm using a Bio-Tek EL-311 plate reader(Bio-Tek Instruments, Winooski, VT). The relative amount of HBsantibody in each sample was interpolated from a standard curveusing purified HBs antigen (Cliniqa Corp., Fallbrook, CA). All washsteps were performed using wash buffer (50 mM Tris, 0.14 M NaCl,0.05% Tween 20, pH 8.0).

Total IgG, IgG1 and IgG2a serum anti-HBs antibody responseswere measured using mouse IgG quantitative kits (Bethyl Labora-tories, Montgomery, TX). Briefly, 96-well Immulon 4 HBX plates(Costar, Cambridge, MA) were coated with anti-mouse IgG, IgG1or IgG2a (for standard curve) in addition to a polyclonal goat–anti-HBs antibody (Cliniqa Corp., Fallbrook, CA) in coating buffer (0.05 Msodium bicarbonate, pH 9.6) for 1 h at room temperature. Plateswere then blocked with blocking buffer (50 mM Tris, 0.15 M NaCl,

ine 26

T. Dutta et al. / Vacc

1% BSA, pH 8.0) for 1/2 h. Purified reference mouse serum wasdiluted in sample diluent (50 mM Tris, 0.15 M NaCl, 1% BSA, 0.05%Tween 20, pH 8.0) at varying concentrations to generate the stan-dard curve (10–200 ng/ml), which was included on each plate.Purified HBs antigen (Cliniqa, Fallbrook, CA), diluted in sample dilu-ent (10 �g/ml) was then applied and incubated for 1 h at roomtemperature. Following incubation, serial dilutions of mouse serumsamples were applied and incubated for 1 h such that at least onevalue would fall within the range of the standard curve. After wash-ing, goat–anti-mouse horseradish peroxidase labeled secondaryantibody was applied and incubated for 1 h. Antibody detection wascompleted with the addition of tetramethylbenzidine (TMB) sub-strate (KPL, Gaithersburg, MD) followed by a 50-min incubation.Product formation was terminated with the addition of 2N H2SO4stop solution [21]. The absorbance of each well was measured at450 nm using a Bio-Tek EL-311 plate reader (Bio-Tek Instruments,Winooski, VT). All wash steps were performed using wash buffer(50 mM Tris, 0.14 M NaCl, 0.05% Tween 20, pH 8.0).

2.6. Preparation of dendrosomes

The optimized PPI–DNA complex was selected for the prepa-ration of dendrosomes. Dendrosomes were prepared as describedearlier [17]. Briefly, phosphatidylcholine (PC) and cholesterol (C)were dissolved in diethyl ether, in varying molar ratios, to whichwas added a solution of optimized dendrimer–DNA complex(100 mg/ml) in PBS (pH 7.4). The mixture was sonicated for 5 minwith a pause of 1 min over an ice bath. Resulting thick emulsionwas vortexed to remove traces of organic solvent. Dendrosome for-mation takes place by phase inversion. A laboratory homogenizer(EmulsiFlex-C5, Avestin, Ottawa, ON, Canada) was used to reducethe particle size of the dendrosomes. Multiple passes (5k–10k psi)were often needed to obtain the desired particle size (200 nm).Purification of dendrosomes was carried out by, passing themthrough sephadex G 50 mini-column. The column was eluted with0.5 ml PBS (pH 7.4).

2.7. Determination of entrapment efficiency

The entrapment efficiency (ratio of amount of PPI–DNA complexentrapped to the amount of complex added expressed in percent-age) was determined by lysing dendrosomes with n-propanol. Adispersion of dendrosomes in PBS (pH 7.4) was mixed with anequal volume of n-propanol and vortexed for 10 min. The result-

ing solution was centrifuged for 15 min at 2500 rpm. Supernatantwas withdrawn and analysed at 260 nm after appropriate dilutionsagainst a PPI dendrimer solution as reagent blank.

2.8. In vitro transfection of CHO cells

CHO cells were transfected with optimized dendrosomal for-mulation and the amount of HBsAg secreted in vitro, was measuredby ELISA as described earlier. Simultaneously, cytotoxicity studieswere carried out using standard procedure for MTT assay [25,26].

2.9. In vivo immunization

Female Balb/c mice, 6–8 weeks old were employed for thestudy. The protocol for animal experiments were duly approvedby the Institutional Animal Ethics Committee, Dr. H.S. Gour Uni-versity, Sagar, India (vide letter No. Animal Ethics Comm/DB/98dated 28 July 2006; Registration No. 379/01/ab/CPCSEA). Ani-mals were given intramuscular injection of 10 �g (in 0.1 ml PBS)of either naked DNA (pRc/CMV-HBs[S]), or its PPI DNA com-plex or dendrosomal formulation containing equivalent amount of

(2008) 3389–3394 3391

pRc/CMV-HBs[S](pDNA). Sera samples collected at different timeintervals were tested for anti HBSsAg IgG1, IgG2a, IgG2b by ELISAas described earlier. Endogenous levels of interferon-g (IFN-g) andinterleukin-4 (IL-4) in whole spleens were determined by themethod of Nakane et al. [27] as modified by de Souza et al. [28,21].Individual spleens from mice injected intramuscularly twice with10 �g naked or liposome-entrapped pDNA and with 1 �g HBsAg (in0.2 ml of 0.9% NaCl) intravenously 24 h before death, were weighed,homogenized in ice-cold RPMI containing 1% 3-(cholamidopropyl-o-dimethylammonio)-1-propanesulphonate (CHAPS, Sigma) in aDounce tissue homogenizer and 10% (w/v) homogenates were pre-pared. Homogenates were left on ice for 1 h and insoluble debriswas removed by centrifugation at 2000 × g for 20 min. Standardcapture ELISA s were used to determine IFN-� and IL-4 levels.

2.10. Statistical analysis

Statistical analysis was performed with Graph Pad Instat Soft-ware (Version 3.00, Graph Pad Software, San Diego, CA, USA) usingone-way ANOVA followed by Tukey–Kramer multiple comparisontest. Difference with p > 0.05 was considered statistically insignifi-cant, whereas p < 0.001 was considered a very significant difference.

3. Results and discussion

3.1. Preparation and in vitro evaluation of dendrimer–DNAcomplex

Dendrimer–DNA complexes were prepared by varying the ratioof dendrimer and plasmid DNA and the zeta potential of the result-ing complexes were estimated. Fifth generation PPI dendrimer wasselected as it contains 64 primary amino groups at its periphery,providing the positive charge necessary for condensing the DNAas well as transfection, but avoid the possible structural defectsassociated with generations higher than that. With the increase indendrimer to DNA ratio, the zeta potential was found to increase(Table 1). Since transfection efficiency as well as cytotoxicity is afunction of the charge present on the surface of the carrier sys-tem, in vitro transfection of CHO cells as well as cytotoxicity studieswere carried out. The results of cytotoxicity studies clearly sug-gest that with the increase in PPI:DNA ratio and the zeta potential,cytotoxicity of the complex increases. Charge dependent toxicityof dendrimers is well documented [22,25,26]. It has been reported

earlier that with the increase in positive charge, cytotoxicity of thecarrier system increases. With the modification of terminal aminogroups, the positive charge and hence the cytotoxicity of den-drimers decreases, which clearly signifies that the positive chargeis the sole factor responsible for the cytotoxicity of dendrimer aswell as PPI–DNA complex [17,22].

In vitro transfection of CHO cells with naked plasmid DNA(pDNA) as well as its dendrimeric formulations exhibited pro-duction of HBsAg, estimated by ELISA. All the dendrimer-basedformulations exhibited significantly (p < 0.001) higher productionof HBsAg as compared to that pDNA. The in vitro expression ofpDNA was found to increase with dendrimer DNA ratio upto a cer-tain limit beyond which it again starts declining. The in vitro HBsAgproduction was found to be maximum at a ratio of DNA:dendrimer,1:50 (PPI 50), which is 61 times higher than pDNA and 2.6 timeshigher than PPI 5 (Fig. 1). Having established that the PPI 50 pos-sesses the highest transfection efficiency and in vitro productionof HBsAg, which is found to be extremely significant (p < 0.001),PPI 50 was selected for the preparation of dendrosomes as well asfurther studies. The significant decrease in transfection efficiencyin case of PPI 75 and PPI 100 can be explained on the basis of the

ine 26

antibody level of animals immunized with either 10 �g of pDNA, orPPI 50 or DF3 formulation containing equivalent amount pDNA ispresented in Fig. 4. Results clearly suggest that DF3 produces sig-

3392 T. Dutta et al. / Vacc

Fig. 1. In vitro transfection of CHO cells with various dendrimeric formulations ofpDNA, based on PPI–DNA ratio. Values are expressed as mean ± S.D. (n = 5).

higher zeta potential and the cytotoxicity associated with them.The cytotoxicity of PPI 75 and PPI 100 was found to be significantlyhigher (p < 0.001) as compared to other formulations, which mightbe the reason behind poor gene expression and in vitro productionof HBsAg (Fig. 2). These observations are in accordance with theresults of earlier researchers [17,24].

3.2. Preparation and characterization of dendrosomes

The entrapment efficiency, zeta potential and vesicle size of den-drosomal formulations of different compositions are presented inTable 2. It was observed that an optimum entrapment efficiency of46.79 ± 1.33% was attained with a 7:3 molar ratio of PC:C, in dendro-somal formulation DF3. Decrease in entrapment efficiency above amolar ratio of 7:3 of PC:C might be due to the increased rigidity ofthe vesicles. DF3 was also found to possess optimum vesicle sizeand zeta potential necessary for transfection. All the dendrosomalformulations were found to be non-toxic to the CHO cells as evidentfrom the results of MTT assay (Fig. 2).

Results of in vitro transfection studies suggest that DF3 possesshighest transfection efficiency, which was found to be extremelysignificant (p < 0.001). While the in vitro production of HBsAgby cells transfected with DF3 was found to be 12.3 ± 0.98-foldhigher than that of PPI 50, in case of DF1 and DF6 it was foundto be 4.5 ± 0.74 and 6.8 ± 0.46-fold higher respectively (Fig. 3).The greater transfection efficiency, as evident from the higherproduction of HBsAg, of DF3 is attributed to its higher entrap-ment efficiency as well as positive zeta potential. Based on theseobservations, formulation DF3 was selected for further in vivoimmunization studies.

Fig. 2. Cytotoxicity of various dendrimeric and dendrosomal formulations. PPIdenotes 5th generation poly(propyleneimine) dendrimer, while pDNA denotespRc/CMV-HBs[S]. All values are expressed as mean ± S.D. (n = 5).

(2008) 3389–3394

Fig. 3. In vitro transfection of CHO cells with various dendrosomal formulations.Values = mean ± S.D. (n = 5).

3.3. In vivo immunization

After establishing that the in vitro production of HBsAg was sig-nificantly higher in case of cells transfected with the dendrosomalformulation DF3 as compared to the PPI–DNA complex or pDNA,in vivo immunization studies were carried out with the optimizedformulations viz. DF3 and PPI 50 and pDNA. Total IgG anti HBsAg

nificantly (p < 0.01) higher IgG response as compared to that of PPI50. pDNA on the other hand displayed a negligible antibody titer,which is found to be statistically insignificant (p > 0.05). Fig. 5(a)shows IgG1 response in mice 2, 4, 6, 8, 10 and 12 weeks afterinjection of the formulations. Results reveal that at all time pointsunder study, animals immunized with DF3 elicited higher immuneresponse in terms of IgG1, which is found to be extremely signif-icant (p < 0.0001). A similar trend is observed in case of IgG2a andIgG2b, where the dendrosomal formulation was found to elicit max-imum immune response as compared to that of PPI 50 and pDNAwithin the period of study (Fig. 5(B) and (c)). Animals immunizedwith various formulations of pDNA displayed a sharp increase inimmune response, which attained a peak value within 6–8 weeksafter immunization. While in case of pDNA and PPI 50, a sharpdecline in antibody titer was observed after 6 weeks, DF3 displayedalmost a steady level of IgG1, IgG2a, IgG2b after 6 weeks provingits efficacy in maintaining the antibody titer and rendering pro-tection against hepatitis B. Animals immunized with dendrosomal

Fig. 4. Immunization with dendrimeric and dendrosomal formulations of pRc/CMV-HBs[S]. Balb/c mice in groups of five were injected intramuscularly on 0th day,and total IgG level was estimated biweekly upto 12 weeks. Values are expressedas mean ± S.D. (n = 5) of log10 of the reciprocal end-point serial twofold serum dilu-tions required for OD readings to reach a value of about 0.200. Sera from untreatedmice gave log10 values less than 2.0.

T. Dutta et al. / Vaccine 26

Table 2Entrapment efficiency, zeta potential and vesicle size of various dendrosomal formulation

S. no. Formulation code Molar ratio (PC:C) Entrapm

1 DF1 9:1 21.83 ±2 DF2 8:2 34.17 ±4 DF3 7:3 46.79 ±5 DF4 6:4 39.59 ±6 DF5 5:5 34.26 ±7 DF6 4:5 31.45 ±Values represent mean ± S.D. (n = 3).

formulation displayed a slightly Th1 polarized immune responseas evident from higher IgG2a/IgG1 ratios. To further elucidate theeffect of antigen expression levels on the Th1/Th2 bias, IFN-� andIL-4 secretion was measured in splenic tissue culture supernatants

Fig. 5. Immunization with dendrimeric and dendrosomal formulations of pRc/CMV-HBs[S]. Balb/c mice in groups of five were injected intramuscularly on 0th day, andsera samples were collected biweekly and tested by ELISA for (A) IgG1, (B) IgG2a,and (C) IgG2b, upto 12 weeks. Values are expressed as mean ± S.D. (n = 5) of log10 ofthe reciprocal end-point serial twofold serum dilutions required for OD readings toreach a value of about 0.200. Sera from untreated mice gave log10 values less than2.0.

(2008) 3389–3394 3393

s

ent efficiency (%) Vesicle size (nm) Zeta potential (mV)

0.72 247 ± 2.3 18.34 ± 0.781.29 183 ± 1.7 23.41 ± 0.471.33 121 ± 2.9 29.33 ± 0.210.58 136 ± 3.2 26.17 ± 1.321.32 154 ± 1.4 25.75 ± 1.431.62 178 ± 1.6 21.86 ± 0.67

following in vitro stimulation with purified HBsAg. Cell-mediatedimmunity was measured in terms of endogenous IFN-� contentof the spleens of mice immunized with PPI50, pDNA and DF3,and injected intravenously with 1 mg encoded antigen 24 h beforedeath. Results (Fig. 6) indicate much greater levels of the cytokinein the spleen of mice immunized with the dendrosomal formula-tion of the plasmid (DF3). Animals immunized with the PPI 50 andDF3 produced significantly (p < 0.001) higher level of IFN-� sug-gesting that the immune response was strictly Th1 mediated. Itis well known that intramuscular injection of DNA vaccines elic-its predominately a Th1 biased immune response as evidenced bya predominance of IgG2a production against HBsAg [1,21,29]. Ourobservations are in accordance with that of the earlier researchers.

It is an well known fact that efficient transfection with cationicpolymers like dendrimers relies on a slight excess of positivecharge, which allows binding of the dendrimer–DNA complexeswith the anionic cell membrane and its subsequent internaliza-tion [13,17,24]. However, the cationic charge on the surface of thecomplex is masked by the plasma proteins, which impose a net neg-

Fig. 6. (A) Interleukin IL-4 and (B) interferon-� levels in the spleens of mice immu-nized with naked pDNA, PPI 50, DF3 formulation. Balb/c mice in groups of five wereinjected on 0th day with 10 �g pRc/CMV-HBs[S] either in the naked form or asits dendrimeric or dendrosomal formulation. Control represents cytokine levels innormal, non-immunized mice. Forty-one days after injection, mice were injectedintravenously with 1 �g HBsAg, killed 24 h later and their spleens subjected tocytokine analysis. Each bar represents the mean ± S.D. of a group of five mice.

ine 26

[

[

[

[

[

[

[

[

[

[

[

[24] Qin L, Pahud DR, Ding Y, Bielinska AV, Kukowska-Lataloo JF, Baker Jr JR, et

3394 T. Dutta et al. / Vacc

ative charge on the surface of the delivery system [30]. Moreover,the DNA in dendrimer–DNA complex is not encapsulated and hencesusceptible to degradation by serum nuclease. Observations [31,32]that anionic molecules can act competitively to release DNA fromcomplexes with cationic vesicles or molecules raises the interestingpossibility that, in vivo, negatively charged serum components (e.g.proteins) may interfere with transfection by competing with DNAfor binding sites, thus bringing about its premature release [33].Since the PPI–DNA complex is entrapped within the dendrosomalvesicles, the anionic molecules not only fail to compete for bind-ing sites but the DNA is also protected from degradation by serumnuclease. This is reason behind greater transfection efficiency ofdendrosomal formulation as compared to that of PPI–DNA complexand higher immune response.

Therefore, it can be concluded that in contrast to naked DNAvaccination by intramuscular route, dendrosomal formulation ofplasmid DNA given by the same route has a different fate. Thereis considerable degradation of naked DNA in plasma with someof the surviving materials transfecting the myocytes and APC thusrequiring a large dose of the DNA for eliciting a substantial immuneresponse. Due to the high surface positive charge and endoso-molytic activity of PPI, though the PPI–DNA complex exhibits highertransfection efficiency and immune response as compared to thatof naked DNA, still it lags far behind dendrosomes. In dendro-somes, the vesicular layers largely protect the PPI–DNA complex.Moreover, it can be hypothesized that the lipid layers of the den-drosomes controls the release of the PPI–DNA complex, while someof the larger vesicles remain at the site of injection following theirdegradation by tissue phospholipases, the smaller ones deliveringand transfecting efficiently the APC in the draining lymph nodes.Our observations clearly prove the superiority of dendrosomes overPPI–DNA complex and pDNA for genetic immunization against hep-atitis B.

Acknowledgements

Tathagata Dutta would like to thank Indian Council for MedicalResearch (ICMR), New Delhi, for providing financial assistance inthe form of Senior Research Fellowship. The project was carriedunder ICMR-SRF grant 45/30/2003/BMS/Immunol.

References

[1] Ulmer JB, Donnelly JJ, Parker SE, Rhodes GH, Felgner PL, Darki VJ, et al. Heterol-ogous protection against influenza by injection of DNA encoding a viral protein.Science 1993;259:1745–9.

[2] Lewis PJ, Babiuk LA. DNA vaccines: a review. Adv Virus Res 1999;54:129–88.

[3] Gregoriadis G. Genetic vaccines: strategies for optimisation. Pharm Res1998;15:661–70.

[4] Davis HL, Whalen RG, Demeneix B. Direct gene transfer in skeletal musclein vivo: factors influencing efficiency of transfer and stability of expression.Human Gene Ther 1993;4:151–6.

[5] Tang DC, DeVit M, Johnston SA. Genetic immunization is a simple method foreliciting an immune response. Nature 1992;356:152–4.

[6] Chattergon MA, Robinson TM, Boyer JD, Weiner DB. Specific immune inductionfollowing DNA-based immunization through in vivo transfection and activationof macrophages: antigen presenting cells. J Immunol 1998;160:5707–18.

[7] Ishii N, Fukushima J, Kaneko T, Okada E, Tani K, Tanaka S, et al. Cationic lipo-somes are a strong adjuvant for a DNA vaccine of human immunodeficiencyvirus type 1. Aids Res Human Retrovirus 1997;13:1421–8.

[

[

[

[

[

[

[

[

[

(2008) 3389–3394

[8] Maruyama T, Ishihara JS, Kim SW, Kim T. Nanoparticle DNA carrier with poly(l-lysine) grafted polysaccharide co-polymer and poly(d,l-lactic acid). BioconjugChem 1997;8:735–9.

[9] Tomalia DA. Birth of a new macromolecular architecture: dendrimers as quan-tized building blocks for nanoscale synthetic polymer chemistry. Prog PolymSci 2005;30:294–324.

10] Liu MJ, Frechet JMJ. Designing dendrimers for drug delivery. Pharm Sci TechnolToday 1999;2:393–401.

[11] Tripathi PK, Khopade AJ, Dutta T, Umamaheshewari RB, Jain S, Saraf DK, etal. Dendrimer polycations for delivery of genes. Indian J Pharm Sci 2003;65:15.

12] Zeng F, Zimmerman SC. Dendrimers in supramolecular chemistry; from molec-ular recognition to self-assembly. Chem Rev 1997;97:1681–756.

13] Bielinska AV, Lataloo JFK, Jhonson J, Tomalia DA, Baker Jr JR. Regulation of invitro gene expression using antisense oligonucleotides or antisense expres-sion plasmids transfected using starburst PAMAM dendrimers. Nucleic AcidRes 1996;24:2176–84.

14] Bielinska AV, Latalto JFK, Baker Jr JR. The interaction of plasmid DNA withpolyamidoamine dendrimers; mechanism of complex formation and analysisof alternations induced in nuclear sensitivity and transcriptional activity of thecomplexed DNA. Biochem Biophys Acta 1997;1353:180–6.

15] Lechardeur D, Sohn KJ, Haardt M, Joshi PB, Monek M, Grahm RW, et al. Metabolicstability of plasmid DNA in the cytosol: a potential barrier to gene transfer. GeneTher 1999;6:482–9.

16] Nishikawa M, Huang L. Non-viral vectors in the new millennium: delivery bar-riers in gene transfer. Hum Gene Ther 2001;12:861–4.

[17] Dutta T, Agashe HB, Vijayarajkumar P, Joshi M, Jain NK. Dendrosome based genedelivery. J Exp Nanosci 2006;1(2):235–48.

[18] Gregoriadis G, Saffie R, Hart SL. High yield incorporation of plasmid DNAwithin liposomes: effect on DNA intergrity and transfection. J Drug Target1996;3:469–75.

19] Velinova M, Read N, Kirby C, Gregoriadis G. Morphological observations on thefate of liposomes in the regional lymph nodes after footpad injection into rats.Biochim Biophys Acta 1996;1299:207–15.

20] Tumer A, Kirby C, Senior J, Gregoriadis G. Fate of cholesterol rich unilamellarliposomes containing 111-I-labelled bleomycin after subcutaneous injectioninto rats. Biochim Biophys Acta 1983;760:119–25.

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

22] Dutta T, Agashe HB, Garg M, Balasubramanium P, Kabra M, Jain NK.Poly(propyleneimine) dendrimer based nanocontainers for targeting ofefavirenz to human monocytes/macrophages in vitro. J Drug Target2007;15(1):89–98.

23] De Brabender Van Den Berg EMM, Meijer EW. Poly(propylene) imine den-drimers: large scale synthesis by heterogeneously catalyzed hydrogenation.Angew Chem Int Ed Eng 1993;32(a):1308–11.

al. Efficient transfer of genes into murine cardiac graft by starburst PAMAMdendrimers. Hum Gene Ther 1998;9:553–8.

25] Malik N, Wiwattanapatapee R, Klopsch R, Lovenz K, Frey H, Weener JW, et al.Dendrimers relationship between structure and biocompatibility in vitro andpreliminary studies on biodistribution of 125I labeled polyamidoamine den-drimers in vivo. J Control Rel 2000;65:133–48.

26] Jevprasesphant R, Penny J, Jalal R, Attwood D, Mc Keown NB, D’Emanuel A. Theinfluence of surface modification on the cytotoxicity of PAMAM dendrimers.Int J Pharm 2003;252(1/2):263–6.

27] Nakane A, Numata A, Minagawa T. Endogenous tumor necrosis factor,interleukin-6, and gamma interferon levels during Listeria monocytogenesinfection in mice. Infect Immunol 1992;60:523–8.

28] de Souza JB, Ling IT, Ogun SA, Holder AA, Play fair JHT. Cytokines and antibodysubclass associated with protective immunity against blood-stage malaria inmice vaccinated with the C terminus of merozoite surface protein 1 plus a noveladjuvant. Infect Immunol 1996;64:3532–6.

29] Brazolot Millan CL, Weeratna R, Siegrist C-A, Davis HL. CpG DNA can inducestrong Th1 humoral and cell-mediated immune responses against hepatitis Bsurface antigen in young mice. Proc Natl Acad Sci USA 1998;95:1553–8.

30] Black CDV, Gregoriadis G. Interaction of liposomes with blood plasma proteins.Biochem Soc Trans 1976;4:253–6.

31] Perrie Y, Gregoriadis G. Liposome-entrapped plasmid DNA: characterisationstudies. Biochim Biophys Acta 2000;1475:125–32.

32] Zelphati O, Szoka Jr FC. Cationic liposomes as an oligonucleotide carrier: mech-anism of action. J Liposome Res 1996;7:31–49.

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