GENE DELIVERY

Journal of Controlled Release 113 (2006) 261 – 270

www.elsevier.com/locate/jconrel

Prophylactic anti-tumor effects in a B cell lymphoma model

with DNA vaccines delivered on polyethylenimine (PEI)
functionalized PLGA microparticles
Sudhir Pai Kasturi a,1 , Hong Qin b,1 , Kassandra S. Thomson a , Shihab El-Bereir b ,
Soung-chul Cha b , Sattva Neelapu b , Larry W. Kwak b , Krishnendu Roy a,⁎
a
Department of Biomedical Engineering, The University of Texas at Austin, ENS 610, C0800, 1 University Station, Austin, Texas 78712, USA
b
Department of Lymphoma and Myeloma, Center for Cancer Immunology Research,
The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030, USA
Received 7 February 2006; accepted 19 April 2006
Available online 3 May 2006

Abstract

Idiotypic sequences, specific to the hypervariable regions of immunoglobulins expressed by malignant B cells offer a therapeutic target in B
cell lymphoma. Efficient approaches have been described to clone a single chain fragment of the tumor immunoglobulin (Ig) comprising of heavy
and light Ig chains (sFv) fused with proinflammatory chemokines. Tumor associated, poorly immunogenic self antigens encoded by plasmid DNA
(pDNA) have been rendered immunogenic by chemokine fusion, thereby targeting to antigen presenting cells (APCs) which differentially express
chemokine receptors. Here we present an injectable (parenteral) approach using synthetic polymer based cationic microparticle formulations for
enhancing the potency of such chemokine/self antigen expressing plasmid construct. Branched and linear polyethyleneimine (PEI) were
conjugated on poly (D, L lactide-co-glycolide) (PLGA) microparticles using carbodiimide chemistry followed by efficient loading of plasmid
DNA. In addition to imparting significant buffering ability to these cationic microparticles, flow cytometry studies indicate that these DNA loaded
microparticles significantly up regulate CD80 and MHC class II markers in phagocytic RAW264.7 cells, indicating intrinsic adjuvant effects.
Intradermal injections in Balb/c mice with these formulations induced significant protection upon tumor challenge with 2.5 times the minimal
lethal dose. Long term survival rates were significant (p b 0.05) in comparison with saline injected controls or blank microparticles. Further studies
indicated that intramuscular delivery might provide better protection compared to intradermal injections and perform similar to gene gun mediated
administration. We conclude, based on these promising in vivo results, that such surface-functionalized microparticles offer an attractive strategy to
improve the potency of self antigen-based cancer DNA vaccines.
© 2006 Elsevier B.V. All rights reserved.

Keywords: DNA vaccine; PLGA; PEI; Surface functionalization; Cationic microparticles; Lymphoma; Cancer vaccines

1. Introduction potential as safe, practical and efficient vaccine has not trans-
lated into human applications [2,5–7]. The major limitations
Genetic immunization has evolved to be an attractive alter- have been inefficient delivery of the pDNA to antigen presenting
native to live attenuated pathogens in terms of safety, ease of cells (APCs) and the lack of specific adjuvants that can effec-
mass production and cost [1–4]. Plasmid DNA (pDNA) vac- tively augment the immune response [7,8].
cines have an established record of efficacy in pre-clinical In cancer immunotherapy most antigens are self or host de-
models of allergies, infectious diseases and cancers [2,5]. Des- rived and therefore are weakly immunogenic [6]. Thus, sig-
pite significant success in a variety of animal models, their nificant challenges exist in realizing the potential of pDNA-based
cancer vaccines [5]. In recent years, polymer based delivery sys-
⁎ Corresponding author. Tel.: +1 512 232 3477. tems have gained increasing interest in APC-targeted vaccine
E-mail address: kroy@mail.utexas.edu (K. Roy). delivery [9–11]. The eventual success of these carriers lies in
1
These authors contributed equally to this work. developing systems which carry either intrinsic adjuvant ability or
0168-3659/$ - see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jconrel.2006.04.006
GENE DELIVERY

262 S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270

designs where the antigen/soluble adjuvant is presented in a pre-clinical evaluation of surface-functionalized cationic micro-
format relevant to interact with toll-like receptor (TLR) molecules particles as efficient carriers of pDNA vaccines in a murine
crucial in innate immune cascades [10–13]. Although gene gun model of B cell lymphoma. Specifically, we have synthesized
mediated immunization has shown tremendous promise in larger surface-functionalized PLGA microparticles with branched and
animals and humans to effectively prime both humoral and linear PEI, compared and evaluated their ability to activate
cytotoxic responses, their safety, reproducibility and cost feasi- APCs in vitro and generate a protective anti-tumor immune
bility for widespread application remains to be developed [14,15]. response when administered via different routes in vivo.
Further, processes for large scale manufacturing of particle Second generation, pDNA-based Idiotypic vaccines stream-
mediated epidermal delivery (PMED) of DNA vaccines are yet to line the approach of realizing patient specific identification of
be validated and issues with long term stability, specific pro- the Idiotype target and quick turn around of cloning a fusion
cedures and specialized technologies remains to be addressed construct of the weakly immunogenic self antigen with a variety
[14], especially for mass immunization and for applications in of pro inflammatory chemokines [37]. Such fusion pDNA
developing countries. Combinatorial delivery systems which can antigen has not only proven to be beneficial in circumventing
incorporate vaccine antigens along with pH sensitive or transgene the requirement of patient specific Idiotype protein purification,
expression enhancing polymers, co-localize immunopotentiators but also proven to be a highly adaptable method to balance
(cytokines, CpG oligos, MPL-A etc) or incorporate chemokines humoral and cellular immunity [37,38]. Cationic microparticles
to attract and target immature dendritic cells and other antigen with adsorbed plasmid DNA have demonstrated long term
presenting cells, could provide, highly potent and attractive alter- expression and induction of potent CD4+ mediated humoral
natives to current delivery systems [13,16–20]. response in addition to induction of CTL responses in mice
Synthetic polymer based micro/nanoparticulate systems have models and non human primates [10,34,35]. Hence, we believe
shown considerable promise to significantly enhance the potency that microparticle mediated delivery of this targeted fusion
of protein and DNA vaccines at low doses [9–11]. Encapsulation construct could significantly enhance long term anti-tumor
of proteins and pDNA inside polymer microparticles has been immunity against B cell lymphoma.
extensively reported for parenteral or oral delivery in a variety of
pre-clinical and clinical models [9,21–25]. However significant 2. Materials and methods
limitations exist in encapsulation of antigens including high shear
during emulsification, elevated temperatures, aqueous/organic 2.1. Polymers and reagents
interface and low pH microclimate, which could lead to low levels
of loading and compromised bioactivity [26–29]. In addition the Acid end terminated (uncapped end groups) poly (lactide-co-
limited diffusional release of the high molecular weight pDNA glycolide) (PLGA 50:50, Resomer® RG502H) with inherent
from these particles leads to low effective dose of the antigen. viscosity (i.v) 0.16–0.2 (Mw ∼ 11,000 Da), and RG503H, (i.v)
Several strategies have been proposed to rationally increase 0.32–0.44, (Mw ∼ 29,000 Da) was purchased from Boehringer
the release rates of protein and DNA antigens in time scales Ingelheim (Virginia, USA). Branched Polyethyleneimine
relevant to life spans of APCs [30–32]. One approach using pH Mw ∼ 70,000 Da was purchased from Polysciences as a 30%
sensitive polymer microparticles have demonstrated enhanced aqueous solution and used as received. Linear PEI 25,000 Da was
pDNA delivery following uptake by APCs followed by bulk purchased from Polysciences as a dry powder. Branched PEI
dosage delivery at early time points [19,32,33]. An alternative 25,000 Da was purchased from Aldrich, Milwaukee, USA. Poly
approach has been to adsorb plasmid DNA on the surface of (vinyl alcohol) Mw ∼ 31,000 (88% hydrolyzed) was purchased
cationic microparticles [29,34,35]. The advantages of such an from Fluka (Sigma–Aldrich, St Louis, MO, USA). MES was
approach could include a) improved interaction of the pDNA purchased from Acros Organics (Fischer Scientific, USA).
antigen with the TLR receptors in the phagocytic pathways b) Plasmid DNA, MCP3-sFv20 encoding the B cell Idiotype antigen
improved loading levels compared to encapsulated designs fused with the monocyte chemotactic protein (MCP)-3 [38] was
especially during scale-up and c) options to encapsulate and co- amplified by Aldevron LLC for in vivo studies (endotoxin con-
deliver other biomolecules such cytokines or lipophilic adju- centrations b 100 EU/mg). Plasmids were ordered in TE buffer pH
vants such as MPL-A to improve the vaccine potency [16,20]. 8.0 for long term storage, ethanol precipitated and resuspended in
Depending on the surface modification strategies chosen to impart RNase and DNAse free water for subsequent loading on micro-
cationicity, protection to the adsorbed DNA from DNAse's has particles. All other marker plasmids were obtained from Aldev-
also been shown [29,12,35]. ron. FlTC-labeled antibodies (CD80, I-A, F4/80) were purchased
Recently, we demonstrated a novel approach in synthesis of from E-biosciences, CA, USA.
cationic microparticles by conjugating branched PEI to the
surface of pre-synthesized PLGA microparticles [36]. These 2.2. Cells and cell lines
formulations were sufficiently cationic to adsorb plasmid DNA,
carried inherent buffering ability due to the presence of PEI RAW264.7 macrophage cells were purchased from Amer-
suggesting a possible “proton sponge effect”, were non toxic at ican Type Culture Collection (ATCC, Manassas, VA) and cul-
increasing doses as shown by MTT cytotoxicity assays and tured in full DMEM media with 10% fetal bovine serum as
finally enhanced gene transfer in vitro suggesting ability to recommended by ATCC. All other cell culture reagents were
transfect APCs. Here, we report further characterization and purchased from Invitrogen, Carlsbad, CA.
 GENE DELIVERY
S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270 263

2.3. Animals Corporation, Holstville, NY). All sample microparticles were

suspended in 1 mM KCl. 10 readings per sample were recorded
Female Balb/C mice were obtained from NCI, Bethesda, and the mean and standard error of the distribution of charged
Maryland. Mice were housed at the M.D. Anderson Cancer particles were noted. BrookHaven recommended material with
Research Center animal housing facility. Mice were ∼ 8–10 zeta potential − 53 mV was used as a reference.
weeks old during the first immunization. All experiments were
approved by the Institutional animal care and use committee at 2.6.2. Microparticle sizing
the M. D. Anderson Cancer Center. Animal care was provided Particle size was analyzed by laser diffraction using a Mal-
in accordance to the procedure outlined in the Animal Care and vern Mastersizer (Malvern, MA, USA). All samples were sus-
Use Handbook (The University of Texas, M.D. Anderson Can- pended in ultrapure water and added drop wise to the stirred cell
cer Center). till a required level of obscuration was obtained. The volume
average distribution mode was chosen to note the size distri-
2.4. PLGA microparticle synthesis butions of the suspensions. A D (v, 0.9) volume distribution
reading indicate that 90% of the population is within a given
PLGA microparticles were synthesized using a double size range. All measurements were obtained for at least 3 inde-
emulsion, solvent evaporation method as reported earlier [36]. pendently synthesized batches and an average percentile distri-
Double emulsion was chosen as the method of synthesis in order bution is reported for each sample.
to ensure that, in the future, other therapeutic molecules (e.g.
adjuvants, chemokines, SiRNA etc.) could be co-delivered 2.6.3. Buffering ability of branched vs linear PEI conjugated
along with surface loaded pDNA. Briefly, 350 mg of RG502H PLGA microparticles
PLGA polymer was dissolved in 7 ml of dichloromethane Acid titration experiments were conducted as reported earlier
(Fischer Scientific, PA) to yield a 5% w/v polymer solution. [36]. Briefly, a 10 mg/ml suspension in 150 mM NaCl of the
300 μl of deionized water was added to the organic phase and respective cationic or unmodified microparticles was adjusted
emulsified at 10,000 rpm using a benchtop Silverson SL2T to ∼ pH 9.0. Varying volumes of 0.1 N HCl or 0.01 N HCl was
homogenizer for 2 min. The first emulsion was added to 50 ml of added drop wise and continued till the effective pH of the
a 1% PVA solution and emulsified for another minute. The water suspension reached less than 4.0. The trend in pH change versus
in oil in water emulsion was added to 50 ml of 1% PVA solution HCl added was plotted to assess the buffering ability of the
and magnetically stirred for 3.5 h for solvent evaporation. The various formulations.
resulting polymeric microparticles were washed thrice with
deionized water, lyophilized and stored at − 20 °C. 2.7. Plasmid DNA loading and quantification

2.5. Branched and linear PEI conjugated cationic PLGA MCP3-sFv20 plasmid was loaded on cationic microparticles
microparticle synthesis by modifying the technique described earlier [36]. This method
was specifically adopted to obtain suspendable and injectable
Branched PEI conjugation was carried out as described earlier formulations for in vivo studies. Briefly, ∼ 60 μg of plasmid
[36]. The process of conjugation was slightly modified for linear DNA (1.2 wt.% target loading) was diluted in 1 ml of PBS (pH
PEI conjugation. Briefly, 5 M excess of linear PEI was dissolved adjusted to 6.00). 5 mg of cationic microparticle suspension was
in 0.2 M MES buffer, pH 6.5. This solution was heated to 55 °C in suspended in 1 ml of PBS pH 6.00. 1 ml of the suspended
a water bath till the polymer was completely dissolved and microparticles was added drop wise to the pDNA solution and
allowed to cool back to room temperature. 40 mg of unmodified incubated at 4 °C on an end to end shaker. DNA loaded micro-
PLGA microparticles was weighed out in a 1.5 ml microcen- particles were recovered by centrifugation at 8500 rpm (Ep-
trifuge tube and suspended in 1 ml of 0.2 M MES buffer (pH 5.2). pendorf 5810R, Eppendorf, Germany). Post centrifugation, the
25 M excess of sulfo-NHS was added in an ice cold MES buffer. supernatant was analyzed using the ND-1000 spectrophotom-
40 M excess of EDC was added to the suspension under mild eter (Wilmington, DE). Absolute amounts of DNA loaded on
vortexing and then allowed to activate the carboxyl groups on the the microparticles were calculated with a mass balance app-
microparticle surface for 2 h. The activated formulation was roach by subtracting the amounts of DNA in the supernatant
added drop wise to the linear PEI solution under magnetic stirring. from the initial amount of pDNA used.
The conjugation process was allowed to proceed for 3.5 h. These
linear PEI conjugated particles were washed extensively 4× with 2.8. In vitro release of pDNA
1 M NaCl solution, passed through a 38 μm mesh, lyophilized and
stored at −20 °C till further use. The DNA loaded microparticles were resuspended in PBS
pH 7.2 and incubated at 37 °C on an end to end shaker. Each
2.6. Particle characterization independent experiment was done in triplicates for each for-
mulation and experiments were repeated at least thrice. Super-
2.6.1. Zeta potential evaluation natants were collected at various times up to 21 days. Picogreen
Zeta potential evaluation of the PEI conjugated microparti- double stranded DNA reagent based fluorometric estimation
cles was conducted using ZetaPlus (BrookHaven Instruments was used to quantify the amount of plasmid in the release
GENE DELIVERY

264 S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270

Table 1 (control) and the absorbance values from formulation treated

Characterization of surface charge, size distribution and pDNA loading cells were converted to relative cell viabilities. At least 2
efficiency of branched and linear PEI conjugated PLGA microparticles
independent experiments were conducted with each formu-
Type of formulation Zeta Percent size DNA loading lation in triplicates.
potential less than (wt/wt) (μg/mg)
in (mV) 10 μm
2.10. Flow cytometry analysis of cell surface markers in RAW
Unmodified PLGA − 10.413 +/ − 0.191 100% 0.268 +/ − 0.079
cells treated with pDNA loaded cationic microparticles
microparticles
Branched PEI70kDa 37.69 +/ − 0.107 85% 7.493 +/ − 0.30
Conjugated PLGA RAW264.7 macrophage cells were seeded in 6 well plates at
microparticles 8 × 105 cells/well and cultured overnight. 100 μg of either
Branched PEI25kDa 37.74 +/ − 0.68 NA 6.04 +/ − 0.816 unmodified PLGA microparticles or pDNA loaded branched
conjugated PLGA
and linear PEI conjugated microparticles were added to the wells
microparticles
Linear PEI25kDa 44.925 +/ − 0.262 100% 4.80 +/ − 1.19 with fresh medium. 100 ng of lipopolysaccaride/ml of culture
conjugated PLGA media was used as a positive control and untreated cells were
microparticles used as negative controls. Fc block (Pharmingen, San Diego,
Size and zeta potential values represent an average of 3 independent batches of USA) was used prior to staining to minimize non specific
cationic microparticles. DNA loading data represents a mean of 3 independent adsorption and FITC-Rat IgG2a isotype control (Pharmingen,
loading experiments. USA) was used for background and non specific fluorescence.
Cells were incubated with the respective formulations for 48 h
samples and the cumulative release was plotted as shown in before staining with FITC conjugated CD80, MHC Class II (I-
Fig. 2. A) and F4/80 antibodies for FACS analysis. Propidium iodide
gating was used to eliminate dead cells and 50,000 total events
2.9. Cytotoxicity evaluation using MTT assay were collected for analysis.

An established MTT assay [36] was used to evaluate the 2.11. In vivo immunization and A20 B cell lymphoma tumor
cytotoxicity of branched and linear PEI conjugated micro- challenge
particles upon interaction with phagocytic RAW 264.7
murine macrophage cells in vitro. Briefly, 2 × 104 RAW 2.11.1. Comparison of branched and linear PEI conjugated
cells were seeded in 96 well tissue culture plates and PLGA microparticle formulation
cultured overnight. Microparticle formulations were added to 10 Balb/C mice per group were immunized thrice in 2 week
the wells (total volume 0.2 ml) at a concentration of 0.5, intervals as described by Biragyn et al. [38]. PBS and/or irrelevant
0.75 and 1.0 mg/ml. These doses exceed the amounts of plasmid DNA injected animals were used as negative controls.
polymer needed to either stimulate cells in other adjuvant Branched PEI70 kDa or linear PEI25 kDa conjugated cationic
experiments conducted (Section 2.10) or as reported earlier PLGA microparticles with adsorbed MCP3-sFv20 pDNAvaccine
for transfection studies [36]. Microparticles were incubated were injected intradermally at dose equivalent of 25 μg pDNA.
for 24 h, rinsed with PBS and the MTT assay was Gene gun immunization was carried out as a positive control as
conducted following manufacturer's protocol (Sigma described earlier [38] with approximately 2 μg pDNA dose
Aldrich, St. Louis, MO). Absorbance (O.D) readings at equivalent. Mice were challenged intraperitoneally 2 weeks after
570 nm of untreated cells were used as 100% viable

Fig. 2. Release profiles of plasmid DNA loaded on branched and linear PEI
Fig. 1. Branched PEI conjugated PLGA microparticles offer significantly conjugated PLGA microparticles. DNA loaded microparticles batches were
increased buffering ability compared to linear PEI conjugated or unmodified incubated in PBS (pH 7.2) at 37 °C. Microparticles were centrifuged at regular
PLGA microparticles in acid titration experiments. intervals and the supernatant collected and analyzed using the picogreen assay.
 GENE DELIVERY
S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270 265

lPEI, yielded a highly positive surface charge of ∼ 45 mV as

indicated in Table 1. Control microparticles (particles treated
similarly with lPEI and NHS but without EDC) treated with high
salt washes showed a potential of ∼ 0–7 mV (data not shown)
indicating specificity of the conjugation process. These results are
consistent with our previously reported data for branched PEI
[36]. Particle sizing was performed using the Malvern Mastersizer
and the volume average size distributions are consistent for
unmodified and cationic polymer modified microparticles (Table
1). Most particles had an average diameter of 2 μm with majority
of the microparticles having a diameter less than 10 μm. Size
evaluation using scanning electron microscopy on pDNA loaded
microparticles did not indicate any significant changes in particle
size (data not shown.
Fig. 3. Branched PEI (bPEI) and linear PEI (lPEI) conjugated PLGA We have previously reported that bPEI conjugation imparts
microparticles are non toxic to phagocytic cells in culture. Cytotoxicity was significant buffering ability on PLGA microparticles [36], an
evaluated using as a mitochondrial enzyme activity (MTT) assay following
incubation with various cationic microparticle formulations for 24 h. Micro-
effect that might lead to efficient endosomal escape through the
particles synthesized with two different molecular weights of PLGA (RG503H proton sponge effect [39]. As shown in Fig. 1, comparison of
and RG502H, Boehringer Ingelheim) were used for cytocompatibility linear PEI and branched PEI conjugated microparticles indi-
evaluation. O.D. values of MTT assay for untreated cells were used as control cated that linear PEI does not impart significant buffering ability
(100% viability). O.D. values for various particle formulations are expressed in while high molecular weight (70 kDa) branched PEI imparts
comparison with untreated cells.
strong buffering ability on PLGA microparticles.

the last immunization with 5 × 105 cells (2.5 times the minimum 3.2. Plasmid DNA loading and in vitro release
lethal dose). Mice survival was followed up to 80 days. Non
parametric log rank test (Kaplan–Meier) was used to establish pDNA loading was achieved using a slight modification of the
significance between the formulation groups and the negative incubation technique described earlier [36]. Unlike the loading of
control with respect to the survival rates over the time of ob- pDNA on CTAB coated microparticles which are protonated at
servation. P value of b 0.05 between the negative control groups physiological pH [29], PEI is not fully protonated at physiological
and the formulation groups were considered significant. pH (only 20% of the nitrogen atoms are charged at physiological
pH) [40]. This was further analyzed by zeta potential titration
2.11.2. Comparison of route of delivery (intramuscular vs. curves where the zeta potential was less than 10 mV at pH 7.0
intradermal) (data not shown). Hence the pH of the DNA adsorption buffer
10 Balb/C mice per group were immunized using pDNA solution was adjusted to pH 6.0 to charge the PEI for efficient
loaded on to branched PEI-functionalized microparticles using DNA adsorption. The loading levels were reproducible with a
either intramuscular or intradermal injections. Intradermal mean loading of ∼ 7 ug/mg (N N 40) for branched PEI70 kDa
delivery was performed as described above. For intramuscular conjugated cationic microparticles as indicated in Table 1. Zeta
injections, microparticles carrying 25 μg of pDNA (analyzed potential measurements following pDNA loading indicated a
from DNA loading experiments) were resuspended in 100 μl of highly negative surface charge (∼ −59 mV) indicating a DNA
saline and injected bi-laterally (50 μl each) into the quadriceps saturated surface (data not shown).
muscles. In addition blank PEI-PLGA microparticles were used Recent reports have compared the efficacy of soluble
as a control. Tumor challenge was carried out as described before branched and linear PEI in terms of their ability to condense
and mice survival was followed up to 80 days. To increase the DNA to form nanocomplexes [41,42]. Branched PEI was found
number of animals per group groups common between the two to strongly condense pDNA to yield smaller complexes com-
studies were combined for statistical analysis. Statistical analysis pared to linear PEI [42]. The strength of condensation had a
was conducted as explained in Section 2.11.1 (log rank test). profound effect on early vs. late onset of gene expression as seen
in in vitro cell culture models. We observed similar differences in
3. Results the loading ability of bPEI conjugated microparticles in
comparison with the lPEI conjugated particles. Both branched
3.1. Microparticle synthesis and characterization PEI70 kDa and 25 kDa conjugated microparticles demonstrated
higher DNA adsorption ability (Table 1). Furthermore, pDNA
Linear PEI (lPEI) with a molecular weight of 25 kDa was used release rates were significantly faster for linear PEI-PLGA
to compare the efficacy of branched vs. linear PEI functionalized particles, releasing 60% of the surface-loaded DNA within
PLGA microparticles. lPEI conjugation was achieved using a 3 weeks while only 28% was released from the surface of 25 kDa
EDC/NHS chemistry similar to that developed for branched PEI branched PEI-PLGA particles and ∼ 20% from the 70 kDa
(bPEI) [36]. Zeta Potential analysis of the lPEI conjugated micro- branched PEI-PLGA conjugated microparticles (Fig. 2). Based
particles, following high salt wash to remove surface adsorbed on these in-vitro release data and transfection data reported
GENE DELIVERY

266 S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270

Fig. 4. Flow cytometry of RAW murine macrophage cells incubated with various particle formulations and controls. Up regulation of key activation/maturation
markers (MHC class II (I-A), CD80 and F4/80) were evaluated following treatment with. LPS, unmodified PLGA microparticles, pDNA loaded branched PEI
conjugated microparticles and pDNA loaded linear PEI conjugated microparticles. (A) Isotype control (open histogram) (B) Histograms for various activation markers
(open histograms). Solid histograms indicate untreated cells.

earlier [36] 70 kDa branched PEI conjugated microparticles and relevant for pDNA complexation, as typically used in transfection
25 kDa linear PEI conjugated PLGA microparticles were used in studies in vitro. This confirms that minimal amounts of PEI
vivo to compare for quantitative differences in generation of anti- (quantity analyzed by fluorescamine assay, b6 μg/mg) [36]
tumor protective effects in established animal models. imparts no significant toxicity to cells in culture.

3.3. Cell toxicity of branched and linear PEI conjugated 3.4. Up regulation of cell surface markers in phagocytic cells
cationic PLGA microparticles in vitro

MTT assay was used to examine the cytotoxicity effects of RAW 264.7 cells have been used extensively for the analysis of
these cationic formulations in a mouse phagocytic RAW264.7 cell activation/maturation of antigen presenting cells in vitro [43,44].
line as reported earlier [10,36]. Untreated cells were used as Incubation of cells with pDNA loaded PEI-functionalized micro-
negative control. Cells were incubated for 24 h at increasing particle formulations for 48 h revealed significant up regulation of
doses. PLGA microparticle formulations, surface-functionalized the co-stimulation marker CD80 and antigen processing marker
with either branched or linear PEI, did not show any significant MHC class II (I-A) (Fig. 4). This up regulation was lower than
difference in cytotoxicity compared to unmodified PLGA LPS and soluble CpG stimulated cells, but consistently higher
particles (Fig. 3). Untreated cells were used as controls and the than control PLGA microparticles. F4/80, a macrophage marker
OD value at 570 nm for untreated cells were considered as 100% was also used to stain the cells to confirm the macrophage phe-
viability. As shown, cells treated with the particle-PEI formula- notype of the cells used. No change was seen in the F4/80 marker
tions showed, on average, 90–100% viability. Earlier observa- for the formulation treated batches but modest up regulation was
tions have confirmed the toxicity of soluble PEI used at doses seen on treatment with LPS only.
 GENE DELIVERY
S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270 267

tentially better protection against tumor challenge for intradermal

route of immunization.
Intramuscular administration of pDNA vaccines are known
to provide improved Th1 type response compared to gene gun
mediated delivery [3,10,15]. In order to evaluate whether intra-
muscular administration could provide significantly enhanced
protective anti-tumor immunity, branched PEI conjugated mi-
croparticles were administered using both intradermal and in-
tramuscular injections. Survival analysis clearly indicated an
improved long term survival rate of mice immunized intramus-
cularly with bPEI-PLGA microparticles (Fig. 5B). The
protective response was similar to that obtained through gene
gun immunization. Log rank tests yielded a statistically
significant enhanced survival rates for both intramuscular
injections of pDNA loaded PEI-PLGA microparticles and
with gene gun mediated routes of immunization (p b 0.01)
compared to the saline or irrelevant plasmid injected controls.

4. Discussion

Several strategies have been reported to enhance the potency

of pDNA vaccines in preclinical animal models and humans,
including electroporation, co-administration of transgene en-
hancing polymers, use of synthetic micro/nanoparticles sys-
tems, biolistic administration of DNA using gene gun or co-
administration of recombinant cytokines, chemokines or other
soluble adjuvants [3,10,15,17,20]. Although naked pDNA ad-
ministration has shown strong immune responses in mouse
Fig. 5. In vivo tumor challenge study. Charts represent the survival rates of Balb/C models, success in primate models and humans has been limited
mice challenged with 2.5 times the minimal lethal dose of tumor cells (A20 B cell [7]. It has been shown that intramuscular naked DNA injections
lymphoma) following three immunizations using pDNA loaded branched and could lead to highly efficient transfection of muscle cells lead-
linear PEI conjugated microparticles. Gene gun mediated immunization was used ing to an immune response against the transgene [45]. However,
as a positive control. (A) Direct comparison of branched versus linear PEI
no clear evidence has been found towards effective transfection
functionalized particles following intradermal immunization (B) Comparison of
intradermal versus intramuscular immunization using branched PEI conjugated of APCs using naked pDNA alone [46]. Additional strategies to
microparticles. Statistical analysis was performed using a log rank test for survival improve the transfection of APCs are required in order to further
data. P values indicated are between experimental group and control group. improve the immunogenicity and clinical applicability of pDNA
vaccines [10,47].
3.5. Tumor challenge and enhanced survival of formulation Recent studies using cationic microparticles with adsorbed
treated mice pDNA have demonstrated expression of the transgene in APCs
in the draining lymph nodes [34,46,47]. A key observation has
In vivo experiments were conducted to compare the per- been that the activation/maturation state of APCs taking up the
formance of linear vs. branched PEI conjugated cationic micro- particle-delivered antigens is crucial in mediating tolerance vs.
particles in inducing prophylactic anti-tumor effects on pre- an immunogenic response [11]. Specifically, for tumor asso-
immunized Balb/c mice against A20 B cell lymphoma tumor ciated self antigens, which are weakly immunogenic, it might be
challenge. Branched PEI70 kDa conjugated PLGA particles necessary to co-stimulate APCs processing the gene-encoded
were chosen based on the in vitro results and the significant antigen. We hypothesized that covalently functionalizing the
buffering ability it had demonstrated over the low molecular surface of biodegradable PLGA microparticles with a transfec-
weight branched PEI batches. Gene gun mediated immunization tion enhancing polycation, PEI, would provide a cationic mi-
was used as a positive control as it has been an established route croparticle system that would a) efficiently adsorb pDNA for
of delivery of these chemokine fused plasmids reported so far surface presentation b) demonstrate increased DNA loading
[38]. Intradermal injections of branched and linear PEI conju- capacity compared to encapsulated designs, c) facilitate endo-
gated microparticles were performed three times at 2 week in- somal escape and d) mediate efficient immune response against
tervals. Survival rates of mice were followed post tumor challenge weakly immunogenic antigens.
for more than 2 months. Statistical analysis (log rank test) of the We have synthesized branched and linear PEI conjugated
survival data revealed a clear advantage of branched PEI-PLGA cationic PLGA microparticles with majority of the microparti-
microparticles over linear PEI-PLGA particles (Fig. 5A). How- cles within 10 μm as analyzed using light scattering measure-
ever, gene gun mediated administration of pDNA showed po- ments and scanning electron micrograph analysis. Zeta potential
GENE DELIVERY

268 S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270

analysis indicated significant charge reversal following PEI PEI and (c) reduction in the number of PEI charged groups due
conjugation. In particular the linear PEI-PLGA particles had to covalent conjugation to the particle surface.
higher cationic charge than the branched PEI conjugated One advantage of delivering DNA vaccines through surface-
microparticles. However, as indicated in Table 1, they exhibited modified microparticles could be the direct interaction of the
less efficient pDNA loading compared to the branched PEI- CpG sequences in pDNA with the TLR-9 receptors in the
PLGA particles. This is consistent with earlier reports that linear phagocytic pathways of antigen processing cells (APCs). This
PEI is 90% protonated at physiological pH [41] yet, less would manifest in efficient activation/maturation of APCs upon
effective in condensing pDNA compared to branched PEI [42]. incubation with pDNA-loaded PEI-PLGA microparticles. To
Since linear PEI only has secondary amines compared to test this hypothesis, we conducted experiments with RAW
primary, secondary and tertiary amines in branch PEI (1:2:1, murine macrophage cells in vitro. RAW cells have been used as
primary :secondary: tertiary amines) it is conceivable that the a substitute to dendritic cells in evaluating adjuvant effects
charge density is much lower resulting in less effective because of the ease and convenience of culture in comparison
condensation. The in vitro release rates of surface adsorbed with isolating primary DC's [21,43]. As early as 48 h post
pDNA also indicate that the DNA adsorption by the 70 kDa addition and incubation of microparticles, clear morphological
branched PEI-PLGA microparticles was much stronger com- changes were seen in the cultured cells compared to the
pared to the low molecular weight branched PEI and the linear untreated controls (data not shown) similar to that reported by
PEI conjugated particles. We chose to investigate the linear PEI Saxena et al. [44]. Branched PEI70 kDa which demonstrated the
(fastest releasing) versus the 70 kda branched PEI (slowest best in vivo protection result, clearly performed the linear
release) functionalized microparticles for tumor challenge PEI25 kDa conjugated cationic formulation in up regulating
studies. Several recent designs have postulated that pDNA MHC Class II and CD80, indicators of APC maturation and
vaccines delivered using pH sensitive polymer systems generate activation. F4/80, a macrophage marker was only up regulated
stronger response partly due to their faster intracellular release upon exposure to LPS but remained unaltered for all other
kinetics, in time scales relevant to life span of most APCs formulations, suggesting a TLR-9 independent mechanism of
[32,33]. However, as demonstrated by our in vivo results com- regulation.
paring branched PEI or linear PEI conjugated particles, lower in Gene gun mediated biolistic administration of gold microbe-
vitro release rates did not correlate with tumor protection ads has been reported to be the most efficient delivery system for
efficacy in vivo. Thus, there appears to be a different mechanism this self antigen/chemokine fusion pDNA construct [38]. In fact,
through which the branched PEI particles mediate a stronger CD8+ T cell depletion studies had clearly demonstrated the need
protective immune response. It is possible that a longer antigen for a cell mediated immunity for long term, enhanced anti-tumor
expression (through slow release of the pDNA) is more relevant effects in the A20 tumor challenge model. Microparticle mediated
for a weakly immunogenic self-antigen, such as the Idiotype injectable systems are known to passively target phagocytic cells
antigen described here. However, persistence of formulations in and thus supports a direct priming mechanism hypothesis [47]. In
the secondary lymphoid organs and a possible transfer from any addition, long term release of pDNA from either encapsulated or
dying APCs to the resident APCs could not be ruled out. Future surface adsorbed formulations could strengthen the immune
mechanistic studies, involving gene expression analysis at the responses generated. Intradermal injections for the first round of
site of injection and the draining lymph nodes, are necessary to experiments were selected for a direct comparison of delivery to
further evaluate how these particles mediate a stronger immune the skin using different mechanisms (injection vs. biolistic).
response. Clearly gene gun, using a significantly lower pDNA dose
Clinical application of PEI as a non viral gene delivery (optimized in previous in vivo studies [38]) showed considerable
vehicle has been severely impeded by its inherent cytotoxicity advantages over the intradermal injection, although statistically
in vivo [41,48]. As shown in Fig. 3 PEI-PLGA particles at significant anti-tumor effects were seen with the branched
relevant doses did not show significant cytotoxicity in vitro. PEI70 kDa conjugated microparticles. We further compared an
Several strategies have been reported to minimize this cyto- intramuscular route of delivery to the intradermal route because of
xicity including using low molecular weight PEI cross linked to reported advantages of improved Th1 response [47]. The results
form larger molecular weight polymer which ensures degrada- demonstrate significant enhancement of long term anti-tumor
tion and elimination of the low molecular weight components, effect compared to intradermal administration. The survival rate
conjugation of low molecular weight PEI to gold nanoparticles was comparable to gene gun mediated delivery, albeit at a higher
to retain transfection properties or conjugating PEG to shield pDNA dose. Future studies, comparing various dose ranges,
some of the positive charges [41,49,50]. Most designs, com- could provide better insight on the efficacy of parenteral versus
bining PEI and PLGA [35,51] have focused on blending biolistic administration in this animal model.
approaches where varying amounts of PEI were blended with Although gene gun mediated immunization has shown
PLGA during particle synthesis. We hypothesized that covalent tremendous promise in larger animals and humans to effectively
conjugation would provide a more reproducible formulation prime both humoral and cytotoxic responses, their safety, re-
and ensure that PEI does not leach out following administration. producibility and cost feasibility for widespread application
This could also provide consistent scale up and eliminate PEI- remains to be developed [14,15]. Further, processes for large
related systemic and local cytotoxicity due to (a) minimal scale manufacturing of particle mediated epidermal delivery
quantity of polymer on the particle surface (b) absence of free (PMED) of DNA vaccines are yet to be validated and issues
 GENE DELIVERY
S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270 269

with long term stability, specific procedures and specialized [16] C.S. Chong, M. Cao, W.W. Wong, K.P. Fischer, W.R. Addison, G.S.
technologies remains to be addressed [14], especially for mass Kwon, et al., Enhancement of T helper type 1 immune responses against
hepatitis B virus core antigen by PLGA nanoparticle vaccine delivery,
immunization and for applications in developing countries. J. Control. Release 102 (1) (2005) 85–99.
Another critical point is that the surface-functionalized micro- [17] A.V. Kabanov, E.V. Batrakova, S. Sriadibhatla, Z. Yang, D.L. Kelly, V.Y.
particles can potentially deliver multiple therapeutic agents Alakov, Polymer genomics: shifting the gene and drug delivery paradigms,
within a single carrier, pDNA and ODNs on the surface along J. Control. Release 101 (1–3) (2005) 259–271.
[18] X. Zhao, S. Jain, H. Benjamin Larman, S. Gonzalez, D.J. Irvine, Directed cell
with proteins, peptides, chemokines or cytokines carried inside
migration via chemoattractants released from degradable microspheres,
the same particle. We have demonstrated feasibility of this dual Biomaterials 26 (24) (2005) 5048–5063.
delivery before [36]. We are currently conducting immunization [19] S.R. Little, D.M. Lynn, S.V. Puram, R. Langer, Formulation and characteriza-
experiments with adjuvants encapsulated particles co-adminis- tion of poly (beta amino ester) microparticles for genetic vaccine delivery,
tered along with pDNA loaded formulations. Such a combina- J. Control. Release 107 (3) (2005) 449–462.
torial delivery approach could provide significantly improved [20] J. Kazzaz, M. Singh, M. Ugozzoli, J. Chesko, E. Soenawan, D.T. O'Hagan,
Encapsulation of the immune potentiators MPL and RC529 in PLG
immune response and therapeutic benefits compared to gene microparticles enhances their potency, J. Control. Release (2005).
gun mediated delivery. [21] A. Luzardo-Alvarez, N. Blarer, K. Peter, J.F. Romero, C. Reymond, G.
In conclusion, we have shown that branched PEI conjugated Corradin, et al., Biodegradable microspheres alone do not stimulate murine
PLGA particles are highly reproducible, show enhanced buf- macrophages in vitro, but prolong antigen presentation by macrophages in
fering, are non cytotoxic, demonstrate enhanced adjuvant effects vitro and stimulate a solid immune response in mice, J. Control. Release
109 (1–3) (2005) 62–76.
in phagocytic cells and provide significant protective anti-tumor [22] E.E. Sheets, R.G. Urban, C.P. Crum, M.L. Hedley, J.A. Politch, M.A. Gold,
effect in an animal model of B cell lymphoma. Such surface- et al., Immunotherapy of human cervical high-grade cervical intraepithelial
functionalized cationic particle formulations could provide an neoplasia with microparticle-delivered human papillomavirus 16 E7 plasmid
efficient platform for improving the potency of DNA vaccines, DNA, Am. J. Obstet. Gynecol. 188 (4) (2003) 916–926.
especially for weakly immunogenic self-antigens. [23] U. McKeever, S. Barman, T. Hao, P. Chambers, S. Song, L. Lunsford, et
al., Protective immune responses elicited in mice by immunization with
formulations of poly(lactide-co-glycolide) microparticles, Vaccine 20 (11–
Acknowledgements 12) (2002) 1524–1531.
[24] D.H. Jones, S. Corris, S. McDonald, J.C. Clegg, G.H. Farrar, Poly(DL-
The authors would like to thank Dr. Angela Bardo with the lactide-co-glycolide)-encapsulated plasmid DNA elicits systemic and
ICMB core facility at The University of Texas at Austin for her mucosal antibody responses to encoded protein after oral administration,
Vaccine 15 (8) (1997) 814–817.
assistance with the FACS. [25] F. Garcia, K.U. Petry, L. Muderspach, M.A. Gold, P. Braly, C.P. Crum, et
al., ZYC101a for treatment of high-grade cervical intraepithelial neoplasia:
References a randomized controlled trial, Obstet. Gynecol. 103 (2) (2004) 317–326.
[26] L. Li, S.P. Schwendeman, Mapping neutral microclimate pH in PLGA
[1] J.J. Donnelly, J.B. Ulmer, J.W. Shiver, M.A. Liu, DNA vaccines, Annu. microspheres, J. Control. Release 101 (1–3) (2005) 163–173.
Rev. Immunol. 15 (1997) 617–648. [27] K. Fu, D.W. Pack, A.M. Klibanov, R. Langer, Visual evidence of acidic
[2] S. Gurunathan, D.M. Klinman, R.A. Seder, DNA vaccines: immunology, environment within degrading poly(lactic-co-glycolic acid) (PLGA) micro-
application, and optimization⁎, Annu. Rev. Immunol. 18 (2000) 927–974. spheres, Pharm. Res. 17 (1) (2000) 100–106.
[3] J.J. Donnelly, B. Wahren, M.A. Liu, DNA vaccines: progress and challenges, [28] E. Walter, K. Moelling, J. Pavlovic, H.P. Merkle, Microencapsulation of DNA
J. Immunol. 175 (2) (2005) 633–639. using poly(DL-lactide-co-glycolide): stability issues and release character-
[4] M. Bivas-Benita, T.H. Ottenhoff, H.E. Junginger, G. Borchard, Pulmonary istics, J. Control. Release 61 (3) (1999) 361–374.
DNAvaccination: concepts, possibilities and perspectives, J. Control. Release [29] M. Singh, M. Briones, G. Ott, D. O'Hagan, Cationic microparticles: a
107 (1) (2005) 1–29. potent delivery system for DNA vaccines, Proc. Natl. Acad. Sci. U. S. A.
[5] M. Yu, O.J. Finn, DNAvaccines for cancer too, Cancer Immunol. Immunother. 97 (2) (2000) 811–816.
55 (2) (2006) 119–130. [30] Y.J. Kwon, S.M. Standley, S.L. Goh, J.M. Frechet, Enhanced antigen
[6] O.J. Finn, Cancer vaccines: between the idea and the reality, Nat. Rev., presentation and immunostimulation of dendritic cells using acid-degradable
Immunol. 3 (8) (2003) 630–641. cationic nanoparticles, J. Control. Release 105 (3) (2005) 199–212.
[7] M.A. Liu, J.B. Ulmer, Human clinical trials of plasmid DNA vaccines, [31] N. Murthy, M. Xu, S. Schuck, J. Kunisawa, N. Shastri, J.M. Frechet, A
Adv. Genet. 55 (2005) 25–40. macromolecular delivery vehicle for protein-based vaccines: acid-degradable
[8] Z. Cui, DNA vaccine, Adv. Genet. 54 (2005) 257–289. protein-loaded microgels, Proc. Natl. Acad. Sci. U. S. A. 100 (9) (2003)
[9] R. Langer, J.L. Cleland, J. Hanes, New advances in microsphere-based 4995–5000.
single-dose vaccines, Adv. Drug Deliv. Rev. 28 (1) (1997) 97–119. [32] S.R. Little, D.M. Lynn, Q. Ge, D.G. Anderson, S.V. Puram, J. Chen, et al., Poly-
[10] D.T. O'Hagan, M. Singh, J.B. Ulmer, Microparticles for the delivery of beta amino ester-containing microparticles enhance the activity of nonviral
DNA vaccines, Immunol. Rev. 199 (2004) 191–200. genetic vaccines, Proc. Natl. Acad. Sci. U. S. A. 101 (26) (2004) 9534–9539.
[11] V.W. Bramwell, Y. Perrie, Particulate delivery systems for vaccines, Crit. [33] C. Wang, Q. Ge, D. Ting, D. Nguyen, H.R. Shen, J. Chen, et al., Molecularly
Rev. Ther. Drug Carr. Syst. 22 (2) (2005) 151–214. engineered poly(ortho ester) microspheres for enhanced delivery of DNA
[12] D.T. O'Hagan, N.M. Valiante, Recent advances in the discovery and delivery vaccines, Nat. Matters 3 (3) (2004) 190–196.
of vaccine adjuvants, Nat. Rev. Drug Discov. 2 (9) (2003) 727–735. [34] X. He, L. Jiang, F. Wang, Z. Xiao, J. Li, L.S. Liu, et al., Augmented
[13] B. Pulendran, Modulating vaccine responses with dendritic cells and Toll- humoral and cellular immune responses to hepatitis B DNA vaccine
like receptors, Immunol. Rev. 199 (2004) 227–250. adsorbed onto cationic microparticles, J. Control. Release 107 (2) (2005)
[14] S. Mitragotri, Immunization without needles, Nat. Rev., Immunol. 5 (12) 357–372.
(2005) 905–916. [35] C.G. Oster, N. Kim, L. Grode, L. Barbu-Tudoran, A.K. Schaper, S.H.
[15] H.J. Dean, J. Haynes, C. Schmaljohn, The role of particle-mediated DNA Kaufmann, et al., Cationic microparticles consisting of poly(lactide-co-
vaccines in biodefense preparedness, Adv. Drug Deliv. Rev. 57 (9) (2005) glycolide) and polyethylenimine as carriers systems for parental DNA
1315–1342. vaccination, J. Control. Release 104 (2) (2005) 359–377.
GENE DELIVERY

270 S. Pai Kasturi et al. / Journal of Controlled Release 113 (2006) 261–270

[36] S.P. Kasturi, K. Sachaphibulkij, K. Roy, Covalent conjugation of polyethyle- [45] J.A. Wolff, V. Budker, The mechanism of naked DNA uptake and expression,
neimine on biodegradable microparticles for delivery of plasmid DNAvaccines, Adv. Genet. 54 (2005) 3–20.
Biomaterials 26 (32) (2005) 6375–6385. [46] M. Dupuis, K. Denis-Mize, C. Woo, C. Goldbeck, M.J. Selby, M. Chen, et
[37] M.M. Dar, L.W. Kwak, Vaccination strategies for lymphomas, Curr. al., Distribution of DNA vaccines determines their immunogenicity after
Oncol. Rep. 5 (5) (2003) 380–386. intramuscular injection in mice, J. Immunol. 165 (5) (2000) 2850–2858.
[38] A. Biragyn, K. Tani, M.C. Grimm, S. Weeks, L.W. Kwak, Genetic fusion [47] K.S. Denis-Mize, M. Dupuis, M. Singh, C. Woo, M. Ugozzoli, D.T.
of chemokines to a self tumor antigen induces protective, T-cell dependent O'Hagan, et al., Mechanisms of increased immunogenicity for DNA-based
antitumor immunity, Nat. Biotechnol. 17 (3) (1999) 253–258. vaccines adsorbed onto cationic microparticles, Cell. Immunol. 225 (1)
[39] M.X. Tang, F.C. Szoka, The influence of polymer structure on the interactions (2003) 12–20.
of cationic polymers with DNA and morphology of the resulting complexes, [48] P. Chollet, M.C. Favrot, A. Hurbin, J.L. Coll, Side-effects of a systemic
Gene Ther. 4 (8) (1997) 823–832. injection of linear polyethylenimine-DNA complexes, J. Gene Med. 4 (1)
[40] R. Kircheis, L. Wightman, E. Wagner, Design and gene delivery activity of (2002) 84–91.
modified polyethylenimines, Adv. Drug Deliv. Rev. 53 (3) (2001) 341–358. [49] Y.H. Kim, J.H. Park, M. Lee, Y.H. Kim, T.G. Park, S.W. Kim, Poly-
[41] U. Lungwitz, M. Breunig, T. Blunk, A. Gopferich, Polyethylenimine- ethylenimine with acid-labile linkages as a biodegradable gene carrier,
based non-viral gene delivery systems, Eur. J. Pharm. Biopharm. 60 (2) J. Control. Release 103 (1) (2005) 209–219.
(2005) 247–266. [50] M. Thomas, A.M. Klibanov, Conjugation to gold nanoparticles enhances
[42] K. Itaka, A. Harada, Y. Yamasaki, K. Nakamura, H. Kawaguchi, K. Kataoka, polyethylenimine's transfer of plasmid DNA into mammalian cells, Proc.
In situ single cell observation by fluorescence resonance energy transfer Natl. Acad. Sci. U. S. A. 100 (16) (2003) 9138–9143.
reveals fast intra-cytoplasmic delivery and easy release of plasmid DNA [51] M. Bivas-Benita, S. Romeijn, H.E. Junginger, G. Borchard, PLGA-PEI
complexed with linear polyethylenimine, J. Gene Med. 6 (1) (2004) 76–84. nanoparticles for gene delivery to pulmonary epithelium, Eur. J. Pharm.
[43] A.M. Krieg, CpG motifs in bacterial DNA and their immune effects, Annu. Biopharm. 58 (1) (2004) 1–6.
Rev. Immunol. 20 (2002) 709–760.
[44] R.K. Saxena, V. Vallyathan, D.M. Lewis, Evidence for lipopolysaccharide-
induced differentiation of RAW264.7 murine macrophage cell line into
dendritic like cells, J. Biosci. 28 (1) (2003) 129–134.