P

Es

ZEa

Ib

c

d

e

f

a

ARRAA

KDTELs

1

ho

TT

mST

t

0h

International Journal of Pharmaceutics 448 (2013) 231– 238

Contents lists available at SciVerse ScienceDirect

International Journal of Pharmaceutics

jo ur nal homep a ge: www.elsev ier .com/ locate / i jpharm

harmaceutical nanotechnology

valuation of cationic dendrimer and lipid as transfection reagents ofhort RNAs for stem cell modification

arrintaj Ziraksaza,b, Alireza Nomanic, Masoud Soleimanid,∗, Behnaz Bakhshandehe,hsan Arefianb, Ismaeil Haririanf, Majid Tabbakhiana,∗∗

Department of Pharmaceutics and Isfahan Pharmaceutical Sciences Research Center, School of Pharmacy and Pharmaceutical Sciences,sfahan University of Medical Sciences, Isfahan, IranDepartment of Molecular Biology and Genetic Engineering, Stem Cell Technology Research Center, Tehran, IranDepartment of Pharmaceutics, School of Pharmacy, Zanjan University of Medical Sciences, Zanjan, IranHematology Department, School of medical sciences, Tarbiat Modares University, Tehran, IranDepartment of Biotechnology, College of Science, University of Tehran, Tehran, IranDepartment of Pharmaceutics, School of Pharmacy, Tehran University of Medical Sciences, Tehran, Iran

r t i c l e i n f o

rticle history:eceived 28 November 2012eceived in revised form 7 March 2013ccepted 16 March 2013vailable online xxx

eywords:endrimerransfection efficiencymbryonic stem cellipidiRNA

a b s t r a c t

Nowadays a large number of clinical trials suffer from lacking an efficient method for drug delivery intotarget cells with minimal side effects. Due to the great significance of this issue in novel and effective ther-apies, more attempts are required in order to distinguish better conditions for biomedical drug delivery.Since embryonic stem cells (ESCs) are under scrutiny of many new studies, development of novel meth-ods for their genetical and functional modifications is of great value. On the other hand, the application ofshort nucleic acids in new therapeutic approaches is increasing. In this study the efficiency of small inter-fering RNA (siRNA) uptake with two transfection reagents generation five of polyamidoamine dendrimer(PAMAM G5) as a cationic dendrimer and N-[1-(2,3-dioleoyloxy)]-N,N,N-trimethylammonium propanemethyl-sulfate (DOTAP) as a cationic lipid and one commercially available reagent were evaluated inmouse ESCs using flow cytometry. Prior to the cellular investigations; atomic force microscopy; gel elec-trophoresis; siRNA binding and release assays; and size and zeta potential measurements were utilized tocharacterize the physicochemical properties of reagent-siRNA nano-complexes. The safety of the nano-complexes was subsequently assessed by MTT assay. Functional effects of siRNA (complementary strand

for OCT4 transcript) transfection in ESCs with the mentioned reagents were analyzed using a quantitativereal-time polymerase chain reaction (qPCR). Surprisingly DOTAP at higher molar ratios and PAMAM atlower molar ratios could successfully knock down the OCT4 transcription relatively better than commer-cial reagent. Our findings supported the appropriate efficiency of the mentioned transfection reagentsfor short nucleic acid transfection. From a clinical point of view, this research helps allocation of shortnucleic acids into stem cells therapies.

. Introduction

Gene/drug delivery is one of the most interesting topics thatelp to maximize the efficiency and minimize the adverse effectsf a treatment (Mellott et al., 2012; Pezzoli et al., 2012). Novel

∗ Corresponding author at: Hematology Department, Faculty of Medical Science,arbiat Modares University, P.O. Box: 14115-111, Tehran, Iran.el.: +98 21 88861065 7; fax: +98 21 8886 1065 7.∗∗ Corresponding author at: Department of Pharmaceutics and Isfahan Phar-

aceutical Sciences Research Center, School of Pharmacy and Pharmaceuticalciences, Isfahan University of Medical Sciences, P.O. Box: 8174673441, Isfahan, Iran.el: +98 311 7922585; fax: +98 311 6680011.

E-mail addresses: soleim m@modares.ac.ir (M. Soleimani),abbakhian@pharm.mui.ac.ir (M. Tabbakhian).

378-5173/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.ijpharm.2013.03.035

© 2013 Elsevier B.V. All rights reserved.

approaches in targeted gene/drug delivery are achieved throughthe design and development of appropriate methods which con-trol the delivery by means of external triggers such as patient’sphysiology or chemical composition of environment. The maindisadvantages of traditional gene delivery methods such as viralvectors are safety issues for instance immunogenicity, and car-cinogenicity which dissuade researchers to use them in vivo and,therefore, leading to delay in passing clinical trials phases to com-mercialization (Bakhshandeh et al., 2012a; Mellott et al., 2012;Pezzoli et al., 2012). In order to bypass these obstacles, manyresearchers have prompted to solve these problems by designing

different non-viral methods for both stable and transient expres-sion, including many commercial non-viral vectors (various lipids,polymers, dendrimers and peptides), electroporation (Yamanoet al., 2010), etc. Molecular constructs of drug/carriers should be

2 al of Pharmaceutics 448 (2013) 231– 238

fusioiothapis

iapourattS

morm(diwabe

eoc

mfcsfmatsdcdtsedafimaso

ie

Table 1Sequence information for Stealth OCT4 NM 013633 RNAi duplexes.

Target sequence 5′-CCCGGAAGAGAAAGCGAACTA-3′

Sense strand 5′-CGGAAGAGAAAGCGAACUATT-3′

32 Z. Ziraksaz et al. / International Journ

ormed preferably in the nanometeric size range to be efficientlyptaken by the cells. Precise control of the molecular structure byynthesizing nano-scaffolds and nano-macromolecular grafts canmprove access to the target site. In addition, uniformity in size andther characteristics such as hydrophilicity, and macromolecular-ty can overcome the risk of fast clearance from body’s excretoryrgans (Cheng et al., 2007; Jain and Asthana, 2007). Liposomes meethe majority of the requirements of an ideal carrier, however, theseighly dynamic vesicles depend not only on their monomer char-cteristics, but also on solution conditions such as concentration,H, and temperature. This may not be favorable to maintain the

ntactness of the carrier in the diverse environment of a biologicalystem (Croy and Kwon, 2006; Kwon, 2003).

In the last two decades, dendrimers have gained popularityn the field of drug and gene delivery due to their nanometricnd homogenous size and molecular weight range. They can dis-lay different changeable surface functional groups for medicalr biological application. The final architecture of dendrimers issually globular, with surface functional groups which are mainlyesponsible for interaction with their environment. This makes thection of these dendritic units more versatile, allowing for drugso be encapsulated or attached to the surface or peripheral groupshat modulate the solubility and toxicity (Liu and Frechet, 1999;venson and Tomalia, 2005).

This offers an advantage over many nanoscale systems andakes it appropriate for more widespread drug/gene delivery

bjectives such as controlling drug payload, solubilization, andelease kinetics in vitro and in vivo (Svenson, 2009). Some of theost preferred classes among dendrimers are polyamidoamine

PAMAM) dendrimers, poly-l-lysine dendrimers, melamine-basedendrimers and polypropylenimine (PPI) dendrimers. The efforts

n dendrimeric system offer an opportunity for the precise andell-controlled delivery of anticancer, antimalarial, antitubercular,

nti-inflammatory agents, and many other potent and promisingiologically active molecules such as therapeutic genes (Gajbhiyet al., 2009).

Besides the efficacy, the biocompatibility of any system is anssential issue when considering it as a drug carrier. The toxicityf dendrimers is directly proportional to their generation number,oncentration, and surface functional groups.

Regarding the fact that the net surface charge of both geneticaterials (DNA or RNA) and cell surfaces are negative, the success-

ul delivery of the polyanionic biomaterials depends on applyingationic carriers (Xiong et al., 2011). As seen for many other cationicystems, including liposomes and micelles, dendrimers with sur-ace groups bearing a cationic charge can readily damage cell

embranes, leading to cell lysis, especially at higher generations. In study carried out by Duncan et al., the effect of dendrimer genera-ion, charge, and concentration on tissue uptake and transport wastudied (Duncan and Izzo, 2005). Overall, the anionic PAMAM den-rimers, G2.5 and G3.5, displayed faster transfer rates through theell layers and lower tissue deposition than amine-terminated den-rimers. Attempts in reducing toxicity have paved the way to maskerminal amine groups or substituting them with other groups,uch as hydroxyl, polyethylene glycol (PEG) or fatty acids(Kolhatkart al., 2007). However, decreasing the charge ratios of the den-rimers as non-viral vectors using above-mentioned methods candversely affect the DNA and RNA compaction, protection, andnally the biological consequences in delivery of these geneticaterials. Moreover, to reduce the cytotoxicity of dendrimers, an

ppropriate approach is to choose a lower generation. For example,electing G5 instead of G6 confirmed that it can reduce the toxicity

f the particles significantly (Nomani et al., 2010).

Nowadays it is obvious that a variety of diseases involvembalanced or over-expressed form of particular genes and muchffort has been made to down-regulate the expression of the

Antisense strand 5′-UAGUUCGCUUUCUCUUCCGGG-3′

Type and modification DNA, Sense: 3′-Cy5

genes at RNA or protein levels (Bakhshandeh et al., 2012b). Oneapproach to change the levels of gene expression relies on thepost-transcriptional level through the use of RNA interference(RNAi)-based technologies. Small interfering RNA (siRNA) technol-ogy indicates that the use of double-stranded RNA can induce genesilencing by causing sequence-specific degradation of complemen-tary mRNA molecules. Since its discovery in Caenorhabditis Elegansin 1998, it has been applied with great success in many studiesboth in vitro and in vivo (Gavrilov and Saltzman, 2012; Mohr andPerrimon, 2012).

Embryonic stem cells (ESCs) can provide a very attractive supplyin medicine, drug discovery, and developmental biology. Becauseof their capacity for controllable multilineage differentiation, ESCshave been used extensively as a source of cells for tissue and cell-based therapies (Hashemi et al., 2011). Although the membranesof stem cells are similarly negative in charge, they are more sensi-tive to the harsh situation (for example when treated with routinecationic non-viral vectors) and unfortunately are trickier to trans-fect than differentiated cells. As a result, the transfection of someof ESCs requires safer and more efficacious carriers than normaltransfection conditions.

Octamer-binding transcription factor 4 (OCT4) is a pluripotencymarker which is expressed continuously in stem cells (Kellnerand Kikyo, 2010). Herein, we choose mouse embryonic stem cells(mESCs) as a ESCs model to evaluate the transfection efficiency andsafety of three selected non-viral vectors to deliver siRNA (againstOCT4 transcript). PAMAM G5 as a cationic dendrimer and N-[1-(2,3-dioleoyloxy)]-N,N,N-trimethylammonium propane methyl-sulfate(DOTAP) as a cationic lipid were evaluated as non-viral transfectionreagents.

2. Materials and methods

2.1. RNA oligonucleotides

Stealth RNAi duplexes (Invitrogen, Qiagen Corporation,Carlsbad, CA, http://www.invitrogen.com) were designedto OCT4 (NM 013633) using the BLOCK-iT RNAi Designer athttp://www.invitrogen.com/rnaidesigner. Double stranded siRNAduplexes were prepared by annealing equimolar amounts of senseand antisense strands in a buffer containing 30 mM HEPES–KOH(pH 7.4), 100 mM potassium acetate, 2 mM magnesium acetate.The annealing reaction was performed by heating to 90 ◦C for1 min and incubating at 37 ◦C for 60 min. The sense strand of eachsiRNA duplex consisted of a 24 nt target sequence followed by aDNA overhang. The antisense strand was composed of nucleotidescomplementary to the target sequence and a DNA overhang. Allsingle-stranded and double-stranded siRNAs were resuspended inDEPC-treated water to a final concentration of 20 �M. Sequenceinformation for the Stealth RNAi duplexes and correspondingStealth controls used in this study is provided in Table 1.

2.2. Dendrimer synthesis

Amine-terminated generation five of polyamidoamine den-drimer (PAMAM) was synthesized and characterized according tothe method described previously (Nomani et al., 2010). Briefly, thesynthesis was performed via iterative reactions of Michael addition

al of P

olmImaiF

2

d2aouar

2

7Fouoflas

2

ttabaa

2

iaLtbos

2

p2daflDirc

Z. Ziraksaz et al. / International Journ

f methylacrylate to the ethylenediamine for half generations fol-owed by exhaustive amidation of half generations by high excess

ethanolic ethylenediamine in the next step for full generations.n each step of the synthesis the product was purified by various

ethods including azeotropic distillation, high vacuum overnightnd ultrafiltration with MWCO of 3000 Da followed by freeze dry-ng of the samples. The final product was characterized by 1H NMR,T-IR and gel permeation chromatography (GPC).

.3. Formation of PAMAM G5/siRNA complexes

Particles composed of PAMAM G5 and siRNA were formed inouble-distilled water (ddH2O) at N/P ratios of 0.5, 1, 2.5, 5, 10 and0. Briefly, siRNA stock solution was diluted in ddH2O to obtain

10.4 �g/ml concentration solution. Then the determined volumef dendrimer solution (100 �l) was added to an equal siRNA vol-me and incubated for 20 min at room temperature. Commerciallyvailable Arrest-InTM (Invitrogen Co., Carlsbad, USA) transfectingeagent was used according to manufacturer’s instruction.

.4. Preparation of DOTAP liposomes

Solutions of DOTAP in chloroform were evaporated at 60 ◦C from00–150 milibar under argon with gradual decrease in pressure.or complete drying, the samples were exposed to argon gas forne minute. Then the tubes were placed in vacuum desiccatornder high vacuum for a couple of hours. When the trace of therganic solvent was removed, the lipid film was purged with argonor 15 min. Multilamellar liposomes were formed when the driedipid film was resuspended in deionized water, heated for 45 mint 45 ◦C, then vortexed. The size of the vesicles was reduced byonication for at least 10 min.

.5. Formation of DOTAP/siRNA complexes

The DOTAP/siRNA complexes were prepared by adding the dilu-ions of aqueous liposome suspension to an appropriate amount ofhe siRNA stock solution (10.4 �g/ml in distilled water). Immedi-tely after mixing, the complex suspension was gently agitated by arief period of pippeting and left for 20 min at room temperature tollow complex formation. Complexes were analyzed immediatelyfter the incubation period.

.6. Particle size and zeta potential analysis

The size or zeta potential of the complexes was evaluatedmmediately after preparation using Malvern Nanosizer ZN seriesccording to the manufacturer’s instructions (Malvern Instrumentstd, Worcestershire, UK). The particle size analysis was based onhe intensity values. The disposable capillary cells (catalog num-er: DTS1061) were used for size and zeta potential measurementsf each sample at the same time. The pH of the medium used forize and zeta analysis was maintained at 7.

.7. Atomic force microscopy (AFM) analysis of the complexes

For AFM experiments, PAMAM G5/siRNA and DOTAP/siRNAarticles were formed at N/P ratios of 0.5, 1, 2.5, 5, 10 and0. The 5 �l of 100-fold diluted siRNA-containing particles wasropped on a freshly cleaved mica sheet, dried at room temper-ture and washed twice by ddH2O. After exposing to a gentle airow for 10 min, the analysis by AFM was performed using a DME

ualScope/RasterscopeTM SPM (DME Co., Denmark). The AFM stud-

es were performed at AC mode, spring constant of 2.8 N/m withesonance frequency of 100 kHz, speed of 30–40 �m/s, and forceonstant of 42 N. The processing of topographic or phase images

harmaceutics 448 (2013) 231– 238 233

was carried out using DME SPM software version 2.1.1.2 (DME Co.,Denmark).

2.8. Agarose gel electrophoresis

Gel electrophoresis experiments were performed using 2% (w/v)agarose gel in a 1 M Tris–acetate–EDTA (TAE) buffer solution.PAMAM/siRNA and DOTAP/siRNA complexes containing 0.1 �g ofsiRNA were mixed with 2 �l of loading dye and loaded onto gel. FreesiRNA was used as control. The samples were subjected to 110 V for20 min. Then, the gel was incubated in 0.5% ethidium bromide solu-tion for further 20 min and visualized under UV illumination by UVdocumentation device.

2.9. siRNA condensation and relaxation assay

To confirm the efficient condensation of siRNA by PAMAM G5and DOTAP, the ethidium bromide expelling analysis (Ruponenet al., 1999) was carried out for the prepared complexes of each usedvector at N/P ratios from 0.03 to 40. Briefly, the different N/P ratiosof either PAMAM or DOTAP with a fixed amount of 40 pmol/50 �lof siRNA were prepared in each well of Corning® 96-well whitepolystyrene plate, and then incubated for 20 min followed by addi-tion of 10 �l of 2.5 �g/ml ethidium bromide. The fluorescenceintensity of each N/P ratio was recorded in triplicate at excitationand emission wavelengths of 530 and 610 nm, respectively usingTecan Infinite® 200 PRO microplate reader. The sample of ethid-ium bromide-added free siRNA was considered as 100% unboundsiRNA and the results were expressed as per cent to this control.The stability of the complexes was further evaluated in the pres-ence of heparin which is considered to be a competing polyanionfor the siRNA in the complexes (Ruponen et al., 1999). To this pur-pose, an amount of 10 �l of heparin (50 IU/ml) was added to eachprepared N/P ratio in a Corning® 96-well white polystyrene plateand after a gentle mixing the samples were incubated for 20 minat room temperature. The corresponding fluorescence intensitieswere measured in triplicate and expressed as the per cent to thefluorescence of 100% unbound siRNA sample.

2.10. Cell viability

The effect of the prepared Arrest-InTM, PAMAM, andDOTAP/siRNA complexes on the cell viability was tested by3-(4,5-dimethythiazol-2-yl)-2,5-diphenyl tetrazolium bromide(MTT) assay. One day before the transfection, mESCs were seededonto 96-well culture plates (10,000 cells/well). A volume of 100 �lof each complex was added to each well and incubated for 4 h.The cells were then washed twice by PBS and fresh medium addedto the wells followed by incubation of the cells for further 48 h.Finally, MTT assay was performed according to the previouslydescribed method (Ziraksaz et al., 2013).

2.11. Cell culture

Mouse emberyonic stem cells (mESCs) were maintained onmitomycin-C (Santacruz biotechnology, cat no. Sc 3514b) treatedmouse embryonic fibroblast (MEF) feeder layers. Then harvestedmESCs were cultured in gelatin-coated 25-cm2 flasks and main-tained in knockout F12 (Gibco), supplemented with 20% embryonic

serum (ES), and 106 U/ml leukemia inhibitory factor (LIF) (Mil-lipore). All media contained 100 �g/ml penicillin (Sigma) and100 �g/ml streptomycin (Gibco). mESCs were passaged twice perweek with daily media replacement.

234 Z. Ziraksaz et al. / International Journal of Pharmaceutics 448 (2013) 231– 238

Fig. 1. Physicochemical evaluations of siRNA complexed by PAMAM and DOTAP. Size and zeta potential analysis of DOTAP/siRNA (A and C) and PAMAM G5/siRNA (B andD) complexes were investigated under various N/P ratios in distilled water at pH 7. The results are shown as mean (n = 3) ± standard deviation. Condensation and relaxationa accorp Hepr arried

2

umeauAtfDfmcr

2

llcawl

2

mctmppmu4

nalysis of DOTAP/siRNA (E) and PAMAM G5/siRNA (F) complexes were carried outerformed in the presence of heparin and represented as each DOTAP or PAMAM ±etardation assay of DOTAP/siRNA (G) and PAMAM G5/siRNA (H) complexes were c

.12. Transfection

Plates (4-well) used for transfection were coated overnightsing 0.1% (v/v) gelatin in PBS. mESCs were plated in 500 �l ESCsedia per well at 24 h prior to the transfection. The transfection

xperiment was performed with labeled siRNA (Cy5-labeled siRNA)nd an optimized 80 pmol anti-oct4 siRNA per well to evaluate theptake of siRNA. Commercial lipo-polymer transfection reagent,rrest-InTM, was added according to the manufacturer’s instruc-

ion. Anti-oct4 siRNA was mixed with G5 dendrimer and DOTAPor at least 10 min in 200 �l serum-free and antibiotic-free F12MEM. The cell-containing wells were incubated by 500 �l of each

ormed complexes for 4 h. Then, the media were replaced with freshedia containing serum. After 48 h of mild shaking incubation, the

ells were analyzed by a quantitative real-time polymerase chaineaction (qPCR) and flow cytometry.

.13. Flow cytometry

Cellular uptakes of 500 �l lipoplexes (20 �M DOTAP/Cy5-abeled siRNA) and 500 �l dendriplexes (20 �M PAMAM G5/Cy5-abeled siRNA) after 6 h incubation were quantified by flowytometry. After washing twice by PBS, the cells were trypsinizednd then centrifuged for 4 min at 1200 rpm. The resuspended cellsere fixed by paraformaldehyde 4% and analyzed using a 2-beam

aser FACS Caliber and CellQuest software.

.14. qPCR analysis

PBS-washed cells were lysed by Trizol (qiagen) according to theanufacturer’s instruction. Total RNA was reverse transcribed into

DNA using Malonney murine leukemia virus (M-MuLV) reverseranscriptase (Fermentas), random hexamer primer 6-mer (Fer-

entas), and dNTPs (Fermentas). The qPCR experiments wereerformed using a standard SYBR Green PCR kit (Fermentas)

rotocol on a RotorGene 6000 instrument (Corbett). Data were nor-alized to an endogenous control gene (�-actin) and calibrated to

ntransfected cells. The sequences of the primers specific for OCT- were as follows: forward primer, GGA TGG CAT ACT GTG GAC;

ding to the method described in the manuscript. The relaxation experiments were inside the graph. The results are as mean (n = 3) ± standard deviation. Agarose gel

out at various N/P ratios. Lane siRNA corresponds to the naked siRNA sample.

reverse primer, CTT GGC AAA CTG TTC TAG C. The sequences for�-actin were as follows: forward primer, CTT CTT GGG TAT GGAATC CTG; reverse primer, GTG TTG GCA TAG AGG TCT TTA C. Allreactions were run in triplicate and the threshold cycle averagewas used for data analysis by Rotor-gene Q software (Corbett). Therelative mRNA expression levels were calculated based on the �Ctmethod.

2.15. Statistics

Student’s t-test or one-way ANOVA was used for comparisonof two or more than two groups, respectively. P value < 0.05 wasconsidered as significantly different in all cases. All experimentswere performed in triplicate, unless otherwise stated, and the datawere shown as mean ± standard deviation (SD).

3. Results

3.1. Evaluation of size, surface charge and morphology ofreagent-siRNA complexes

The results of size and zeta potential analysis for complexes ofDOTAP and PAMAM G5 with siRNA are presented in Fig. 1A–D.Generally, at higher N/P ratios DOTAP/siRNA complexes showedsmaller size and higher zeta potential. Here, the N/P ratio of 0.5showed the biggest size among the other N/P ratios revealing thatthe amount of cationic lipid was not enough to neutralize thecharges and compact the siRNA efficiently. The results demon-strated that the differences among the sizes of N/P ratios of 5, 10and 20 were not statistically significant. Despite the slight changesof size over the N/P ratio higher than one, evaluation of the surfacecharges of the complexes showed the significant increases withincreasing N/P ratios. The highest charge was observed at N/P ratioof 20 which was around 35–40 mV.

Contrary to DOTAP, the size of PAMAM G5 complexes did not

show any reasonable change over the increase in N/P ratios of thecomplexes (Fig. 1B). Zeta potential of PAMAM/siRNA complexes areshown to be negative at N/Ps of lower than 2.5, then reaching asteady state charge of around 20–25 mV at N/Ps of 5 and higher.

Z. Ziraksaz et al. / International Journal of Pharmaceutics 448 (2013) 231– 238 235

F and Pf .5 (f),

cdpa9ceNr

3

itacbsfwtpaprt

3

tgwauha

4

e

by flow cytometry (Fig. 5). The uptake efficiency of naked siRNA wasonly 1.9%. Meanwhile, the efficiency of Arrest-In/siRNA complexesuptake was shown to be on average of 39.15%. For PAMAM G5, theuptake efficiencies were 25.6% and 35% for N/P ratios of 10 and 20,

ig. 2. Atomic force microscopy phase or topographic images of DOTAP/siRNA (a–e)or DOTAP were as 0.5 (a), 2.5 (b), 5 (c), 10 (d) and 20 (e), and for PAMAM were as 0

AFM evaluation showed almost spherical morphologies for theomplexes (Fig. 2). Generally, the projection diameter of complexesecreased along with increase in N/P ratios. For DOTAP/siRNA com-lexes, higher N/P ratios resulted in smaller particles and theirverage diameters at N/P ratios of 0.5, 2.5, 5, 10 and 20 were 883,92, 471, 486 and 348 nm, respectively. The PAMAM G5/siRNAomplexes showed a better homogeneity in their sizes than DOTAP,specially at two N/P ratios of 5 and 10. Their average diameters at/P ratios of 0.5, 2.5, 5 and 10 were 163.5, 191, 57, and 51 nm,

espectively.

.2. Condensation and relaxation of siRNA

Fig. 1E and F shows the siRNA condensation and relaxation stud-es by PAMAM (Fig. 1F) and DOTAP (Fig. 1E). This figure confirmshat both PAMAM and DOTAP are able to form tight complexesnd block the access of ethidium bromide to their internal siRNAargos. However, at maximum compaction still 20% of siRNA coulde detected by the dye. This is supposed to be due to the part ofiRNA strands which remained unprotected on the particle sur-aces and were exposed to dye intercalation. It seemed that PAMAMas more efficient in compaction than DOTAP. While N/Ps higher

han 10 were necessary for DOTAP to reach its steady state of com-action, PAMAM could protect the maximum siRNA at N/P ratio ofround one. At the same time, incubation of PAMAM/siRNA com-lexes with heparin resulted in 20–40% relaxation of siRNA at N/Patios of 0.5–20 but DOTAP showed no significant relaxation duringhe similar incubation time and heparin concentration.

.3. Agarose gel electrophoresis analysis

siRNA stability analysis by agarose gel electrophoresis showedhat DOTAP was able to stabilize and retard siRNA efficiently in theel at all tested N/P ratios (Fig. 1G). No distinct band related to siRNAas detected on the gel in case of DOTAP complexes. For PAMAM

t lower N/P ratios (0.5, 1 and 2.5) most of the siRNAs remainednbound and were detectable as smear on the gel. However, atigher N/P ratios (10 and 20) the dendrimer completely stabilizednd compacted the siRNA during the experiment (Fig. 1H).

. Cytotoxicity analysis

mESCs were subjected to the used vectors to find their cytotoxicffects. First the mESCs were isolated from their cultured colonies

AMAM/siRNA (f–i) complexes formed at different N/P ratios. The selected N/P ratios2.5 (g), 5 (h), and 10 (i).

on MEFs (Fig. 3). MTT assay for PAMAM G5, DOTAP/siRNA at dif-ferent N/P ratios (called with siRNA samples in Fig. 4) and also thesame related concentration and composition for each N/P but with-out the addition of corresponding amounts of siRNA (called here aswithout siRNA treatments) are presented in Fig. 4. The F12 mediumgroup (untreated cells) was considered as control. In the pure siRNAgroup, more than 80% of cells were alive. For both dendrimer andDOTAP, all vectors alone and in the complexed forms showed aregular pattern: decreasing in cell viability along with increasingin charge ratio. While 80% of cells were viable at both N/P ratios 0.5and 1, at N/P ratios of 10 and 20 about 40% of cells died. The resultsalso showed that the complexes of PAMAM/siRNA were slightlymore toxic than DOTAP at higher N/P ratios.

4.1. Investigation of transfection efficiencies of the complexes

The efficiencies of Cy5-conjugated-siRNA uptake by PAMAM G5,DOTAP, and Arrest-InTM (as a commercial control) were evaluated

Fig. 3. A representative phase contrast microscopy image of mouse embryonic stemcell (mESC) colonies on mouse embryonic fibroblasts (MEFs).

236 Z. Ziraksaz et al. / International Journal of Pharmaceutics 448 (2013) 231– 238

Fig. 4. MTT assay graphs of PAMAM/siRNA (A) and DOTAP/siRNA (B) complexes. Multiple PAMAM/siRNA and DOTAP/siRNA complexes at various N/P ratios (gray columnsw AM ana weigm stan

rar

4O

(tP2

r(

Fc

hich are called with siRNA samples) were compared with their corresponding PAMre called without siRNA samples), Arrest-InTM at its manufacturer’s recommendedeans of six different measurements relative to the negative control (DMEM-F12) ±

espectively. Using DOTAP as transfection reagent resulted in 59.8%nd 56.5% uptake efficiencies of siRNAs for N/P ratios of 10 and 20,espectively.

.2. Functional analysis of siRNA transfection by evaluation ofCT4 transcription

In order to evaluate the functional transfection of siRNA80 picomol, complement of OCT4 transcript), we investigatedhe transcription of OCT4 quantitatively in transfected mESCs byAMAM G5 and DOTAP/siRNA at N/P ratios of 0.5, 1, 2.5, 5, 10 and

0. Untreated cells were applied as negative controls.

qPCR showed that the OCT4 transcriptions were significantlyeduced in the mentioned groups compared to untreated cellsFig. 6). Remarkably, the transcription of OCT4 in the PAMAM group

ig. 5. The flow cytometry evaluation of Cy5-labaled siRNA uptake in mESCs by PAMAommercial transfection reagent and negative control of untreated mESCs.

d DOTAP concentration alone and without addition of siRNA (black columns whichht ratio, siRNA alone and negative control of DMEM-F12. All represented data aredard error of mean (SEM).

was significantly lower than that in the DOTAP group at lower N/Pratios such as 0.5, 1 and 2.5. In contrast, the DOTAP groups weremore successful to knock down OCT4 transcription compared toPAMAM groups at higher N/P ratios such as 5, 10 and 20.

5. Discussion

To facilitate genetic modifications of stem cells, the efficienciesof two transfection reagents (PAMAM G5 dendrimers; as cationicdendrimer and DOTAP; as cationic lipid) were compared in trans-ferring siRNA into the mESCs.

The dominant mechanism directing the interaction betweenthe negative nucleic acids and cationic carriers is often the elec-trostatic forces which will condense the DNA or RNA strands intonanocomplexes (Bloomfield, 1997; Kong et al., 2012). In this study,

M and DOTAP at different N/P ratios of 10 and 20 compared with Arrest-InTM as

Z. Ziraksaz et al. / International Journal of Pharmaceutics 448 (2013) 231– 238 237

F at difT

eptHpto2is

sbtscNaTN2

tr

iPhmgpigsttcrpra

woloit

ig. 6. Quantitative evaluation of OCT4 transcription (siRNA knock-down) in mESCshe data are mean of three independent measurements ± standard deviation.

valuations of the size and morphology of reagent-siRNA com-lexes showed that in general along with increase in N/P ratios,he diameter of these nanocomplexes decreased (Figs. 1 and 2).owever, unlike DOTAP, the size of the PAMAM complexes did noterfectly show reasonable trend over increase in the N/P ratios ofhe complexes, which might imply the heterogeneity of complexesf small oligos when compacted by PAMAM G5 (Nomani et al.,010). In contrast to the size results, zeta potentials analysis was

n line with the N/P increments. The higher N/P ratios accordinglyhowed the higher zeta potential.

The results of the size measured by AFM and Nanosizer showome significant discrepancy among the complexes prepared byoth carriers. The most probable explanation would be related tohe basics of each technique and sample preparation. For AFM theamples are dried on mika sheets and the projection diameter of theomplexes will finally be measured while the size data in case ofanosizer are gathered for the complexes by laser light scatterings a rough estimation when they are dispersed in distilled water.he detailed explanation of the discrepancy between the AFM andanosizer data can be found in various literatures (de Assis et al.,008; Singh et al., 2011).

Tracking the prepared complexes on the gel following the elec-rophoresis revealed the complete binding of siRNAs and theiretardation in the gel by the used reagents.

While the results of condensation assay was interestinglyn favor of PAMAM, the relaxation of prepared complexes ofAMAM/siRNAs was easier than DOTAP/siRNA in the presence ofeparin. This is in line with the gel electrophoresis results andeans that the protection and stability of DOTAP complexes are

enerally higher than PAMAM G5. Ruponen et al. (1999) com-ared complexes of different non-viral carriers with plasmid DNA

n view of their stability in the presence of polyanions such aslycosaminoglycans and heparin sulphate. According to their clas-ification, DOTAP and PAMAM G6 dendrimer can be categorized ashe moderately sensitive vectors to the polyanions. They concludedhat molecules which have stable positive charges such as DOTAPan be even more stable in the presence of polyanions. Our findingelated to the relaxation behavior of DATAP can be described by itsermanent positive charges of quaternary amines which results inelatively higher resistance to the relaxation by polyanionic hep-rin.

Regarding the cytotoxicity assessments, the higher charge ratiosere clearly related to the higher toxicity of mESCs. This effect

f higher cationic charges has been mentioned for various cell

ines in culture (Hunter, 2006). In our study, similar sensitivity wasbserved in case of mESCs (Fig. 4). The mechanism of cytotoxic-ty would be found in literatures (for example, Hunter, 2006, andherein references). On the other hand, the cytotoxicity of PAMAM

ferent N/P ratios of DOTAP and PAMAM/siRNA complexes using an optimized qPCR.

at higher N/P ratios seems to be higher than DOTAP. Apart fromtheir different molecular structures, higher cytotoxcicty of PAMAMwould also be related to its higher free cationic molecules at theseN/Ps. As Fig. 1E and F shows, at N/P around one PAMAM can com-pact siRNAs completely where the addition of more PAMAM willcause increment in the free cations which may correspond to itsslightly higher cytotoxicities.

For more cellular investigations, we applied Cy5-conjugated-siRNAs to evaluate the transfection efficiencies of these reagentsin mESCs. It is noteworthy to mention that without mild shak-ing incubation, the transfection efficiencies were too low (datanot shown). Flow cytometry analyses demonstrated that the deliv-ery efficiency of DOTAP was relatively higher than commercialArrest-InTM reagent while the yield of PAMAM was comparable toArrest-InTM efficiency (Fig. 5).

As the last analysis, we checked the transfection of siRNA(complement of OCT4 transcript, Table 1) by qPCR. ESCs expressOCT-4 as one of the most specific transcription factors besidesnonog and SOX2 which are pluripotency markers. Quantitativeevaluation of OCT4 transcription in transfected mESCs showed asignificant down-regulation. Surprisingly the knock-downing func-tions of both aforementioned reagents were relatively better thanthe positive control Arrest-InTM.

As illustrated in Fig. 6, the dendrimer complexes were generallysuccessful in silencing OCT4 at lower charge ratios (i.e. 0.5, 1 and2.5) while DOTAP complexes interfered better in OCT4 transcriptat higher charge ratios (i.e. 5, 10 and 20). Some studies referredto the excess amount of cationic charges of polycationic polymersas the interfering and harmful agents for the cellular physiologyand functions of biomacromolecules such as enzymes and proteins(Ruponen et al., 1999). These excess polycations might disrupt thefunction of siRNA-RISC complexes, too. RISC complexes are nec-essary in siRNA activity, therefore, PAMAM/siRNA efficiency couldbe expected to plummet with increasing the N/P ratio over 5. Inthe case of DOTAP complexes, besides its lower interference withessential anionic biomacromolecules, N/P ratios higher than 5 werecapable of condensing siRNA and forming stable and protectedcomplexes (Fig. 1E). These properties resulted in better activity ofits siRNA content in their target inside the cells at higher N/P ratios.

However, altogether, qPCR results (Fig. 6) showed that atoptimum N/P ratios, PAMAM was more than 100-fold better trans-fecting reagent than DOTAP for mESCs. Considering the flowcytometry results (Fig. 5), they might indicate that DOTAP is moreefficient in cell uptake (2-fold higher uptake). Contrasting to that,

the endosomal escape would be a trickier barrier for its complexesthan PAMAM complexes. This effect might impede the function oflots of administered siRNA complexes in case of DOTAP and lead toa lower activity compared to PAMAM as observed in qPCR results.

2 al of P

Hr

6

rcesofaa

rami(maeefebpc

C

A

Rali

R

B

B

B

38 Z. Ziraksaz et al. / International Journ

owever, the exact mechanism of each vector and the effect of N/Patio on siRNA activity need further detailed investigations.

. Conclusion

In this study the siRNA uptake efficiencies of two transfectioneagents (PAMAM G5 as dendrimer and DOTAP as lipid) and oneommercially available reagent were evaluated in mESCs by differ-nt physicochemical and biological assays. Since ESCs are under thecrutiny of many new researches in biomedical therapies, devel-pment of the novel and efficient methods for their genetic andunctional modifications is of great value. On the other hand, thepplication of short nucleic acids such as siRNAs in new therapeuticpproaches is increasing.

Prior to the cellular investigations, characterization of theeagent/siRNA nanocomplexes by gel electrophoresis, complex-tion and heparin relaxation, AFM, and size and zeta potentialeasurements were performed. In cellular assessment, the tox-

city of the reagents was considered. Functional effects of siRNAcomplementary strand for OCT4 transcript) transfection using the

entioned reagents in mESCs were analyzed by flow cytometrynd qPCR. OCT4 is a pluripotency marker gene that is expressedxtensively in stem cells. Our findings supported the appropriatefficiency of the mentioned transfection reagents for siRNA trans-ections. Due to the great significance of this issue in the novel andffective therapies, more attempts are required in order to designetter conditions for delivery of genetic materials. From a clinicaloint of view, this research helps allocation of siRNA into the stemells therapies.

onflicts of interest

The authors declare no conflicts of interest.

cknowledgements

This work was supported financially by Stem Cell Technologyesearch Center. The authors would like to thank Dr. Nikougoftarnd her colleagues for their collaboration in this study. We wouldike to thank the valuable work of Mr. J. Mohammadi in languagemprovement of this manuscript.

eferences

akhshandeh, B., Soleimani, M., Hafizi, M., Ghaemi, N., 2012a. A comparative studyon nonviral genetic modifications in cord blood and bone marrow mesenchymalstem cells. Cytotechnology 64, 523–540.

akhshandeh, B., Soleimani, M., Hafizi, M., Paylakhi, S.H., Ghaemi, N., 2012b.MicroRNA signature associated with osteogenic lineage commitment. Mol. Biol.Rep. 39, 7569–7581.

loomfield, V.A., 1997. DNA condensation by multivalent cations. Biopolymers 44,269–282.

harmaceutics 448 (2013) 231– 238

Cheng, Y., Gao, Y., Rao, T., Li, Y., Xu, T., 2007. Dendrimer-based prodrugs: design,synthesis, screening and biological evaluation. Comb. Chem. High ThroughputScreen 10, 336–349.

Croy, S.R., Kwon, G.S., 2006. Polymeric micelles for drug delivery. Curr. Pharm. Des.12, 4669–4684.

de Assis, D.N., Mosqueira, V.C.F., Vilela, J.M.C., Andrade, M.S., Cardoso, V.N., 2008.Release profiles and morphological characterization by atomic force microscopyand photon correlation spectroscopy of 99mTechnetium-fluconazole nanocap-sules. Int. J. Pharm. 349, 152–160.

Duncan, R., Izzo, L., 2005. Dendrimer biocompatibility and toxicity. Adv. Drug Deliv.Rev. 57, 2215–2237.

Gajbhiye, V., Palanirajan, V.K., Tekade, R.K., Jain, N.K., 2009. Dendrimers as thera-peutic agents: a systematic review. J. Pharm. Pharmacol. 61, 989–1003.

Gavrilov, K., Saltzman, W.M., 2012. Therapeutic siRNA: principles, challenges, andstrategies. Yale J. Biol. Med. 85, 187–200.

Hashemi, S.M., Soudi, S., Shabani, I., Naderi, M., Soleimani, M., 2011. The promotionof stemness and pluripotency following feeder-free culture of embryonic stemcells on collagen-grafted 3-dimensional nanofibrous scaffold. Biomaterials 32,7363–7374.

Hunter, A.C., 2006. Molecular hurdles in polyfectin design and mechanistic back-ground to polycation induced cytotoxicity. Adv. Drug Deliv. Rev. 58, 1523–1531.

Jain, N.K., Asthana, A., 2007. Dendritic systems in drug delivery applications. ExpertOpin. Drug Deliv. 4, 495–512.

Kellner, S., Kikyo, N., 2010. Transcriptional regulation of the Oct4 gene, a mastergene for pluripotency. Histol. Histopathol. 25, 405–412.

Kolhatkar, R.B., Kitchens, K.M., Swaan, P.W., Ghandehari, H., 2007. Surface acety-lation of polyamidoamine (PAMAM) dendrimers decreases cytotoxicity whilemaintaining membrane permeability. Bioconjug. Chem. 18, 2054–2060.

Kong, W.H., Bae, K.H., Jo, S.D., Kim, J.S., Park, T.G., 2012. Cationic lipid-coated goldnanoparticles as efficient and non-cytotoxic intracellular siRNA delivery vehi-cles. Pharm. Res. 29, 362–374.

Kwon, G.S., 2003. Polymeric micelles for delivery of poorly water-soluble com-pounds. Crit. Rev. Ther. Drug Carrier Syst. 20, 357–403.

Liu, M., Frechet, J.M., 1999. Designing dendrimers for drug delivery. Pharm. Sci.Technol. Today 2, 393–401.

Mellott, A.J., Forrest, M.L., Detamore, M.S., 2012. Physical non-viral gene deliverymethods for tissue engineering. Ann. Biomed. Eng..

Mohr, S.E., Perrimon, N., 2012. RNAi screening: new approaches, understandings,and organisms. Wiley Interdiscip. Rev. RNA 3, 145–158.

Nomani, A., Haririan, I., Rahimnia, R., Fouladdel, S., Gazori, T., Dinarvand, R., Omidi,Y., Azizi, E., 2010. Physicochemical and biological properties of selfassembledantisense/poly(amidoamine) dendrimer nanoparticles: the effect of dendrimergeneration and charge ratio. Int. J. Nanomed. 5, 359–369.

Pezzoli, D., Chiesa, R., De Nardo, L., Candiani, G., 2012. We still have a long way togo to effectively deliver genes! J. Appl. Biomater. Funct. Mater. 10, e82–e91.

Ruponen, M., Yla-Herttuala, S., Urtti, A., 1999. Interactions of polymeric andliposomal gene delivery systems with extracellular glycosaminoglycans: physi-cochemical and transfection studies. Biochim. Biophys. Acta – Biomembr. 1415,331–341.

Singh, S.K., Prasad Verma, P.R., Razdan, B., 2011. Atomic force microscopy,transmission electron microscopy, and photon correlation spectroscopy: threetechniques for rapid characterization of optimized self-nanoemulsiying drugdelivery system of glibenclamide, carvedilol, and lovastatin. J. Disper. Sci. Tech-nol. 32, 538–545.

Svenson, S., 2009. Dendrimers as versatile platform in drug delivery applications.Eur. J. Pharm. Biopharm. 71, 445–462.

Svenson, S., Tomalia, D.A., 2005. Dendrimers in biomedical applications – reflectionson the field. Adv. Drug Deliv. Rev. 57, 2106–2129.

Xiong, F., Mi, Z., Gu, N., 2011. Cationic liposomes as gene delivery system: transfec-tion efficiency and new application. Pharmazie 66, 158–164.

Yamano, S., Dai, J., Moursi, A.M., 2010. Comparison of transfection efficiency of

nonviral gene transfer reagents. Mol. Biotechnol. 46, 287–300.

Ziraksaz, Z., Nomani, A., Ruponen, M., Soleimani, M., Tabbakhian, M., Haririan, I.,2013. Cell-surface glycosaminoglycans inhibit intranuclear uptake but promotepost-nuclear processes of polyamidoamine dendrimer-pDNA transfection. Eur.J. Pharm. Sci. 48, 55–63.