楊 陽 郭 霞

(1四川大學華西醫(yī)院生物治療國家重點實驗室,成都610041;2揚州大學化學化工學院,江蘇揚州225002)

1 Introduction

Cationic liposomes and polymers are widely used as nonviral(or synthetic)vectors bothin vitroandin vivo.1-3Both plasmid DNA and siRNA bind to cationic liposomes and polymersviaelectrostatic interaction between the anionic phosphodiester backbones and the positively charged groups in cationic delivery agents and hence,form a complex protecting nucleic acid from degradation by nuclease.

Polyethylenimine(PEI)has been one of the most widely studied synthetic cationic gene vectors due to its superior transfection efficiency.4-8The high transfection efficiency of PEI/DNA complexes is attributed to the unique capacity of PEI tobuffer endosomes according to the proton sponge hypothesis.9,10Large PEI macromolecules(＞5 kDa),HMW-PEI,are generally non-degradable and cannot be cleared from the body,while low molecular weight PEI(LMW-PEI),by contrast,has been demonstrated to have a relatively low cytotoxicity.11The complexes formed by LMW-PEI and nucleic acid have significantly large diameters and show a tendency towards aggregation,while those formed by HMW-PEI and nucleic acid are stable,with a satisfactory size for gene delivery.12PEI with molecular weight of 25 kDa(PEI 25 kDa)has thus been usually used.11,12However,its high cytotoxicity partly caused by its high molecular weight and charge density seriously hampered its therapeutic use.In the past decade,many researchers have focused on reducing the nonspecific cytotoxicity of HMW-PEI while maintaining its superior gene delivery efficiency through structure modification.For example,PEI co-polyethyleneglycol(PEG)was synthesized for efficient gene delivery and gene silencing.13The incorporation of PEG segments was not only helpful to solve the cytotoxicity problem,but also improved the efficiency of gene release.A hyperbranched PEI modified with poly(γ-benzyl-L-glutamate)was found to reduce the cytotoxicity significantly and exhibit 4 times higher gene silencing efficiency than PEI 25 kDa.14Enhanced chemotherapy efficacy was also observed using PEI-grafted graphene oxide.15

At the same time,since LMW-PEI has been demonstrated to have a relatively low cytotoxicity,extensive efforts have been focused on developing LMW-PEI based vectors for gene delivery.By using gel permeation chromatography,the PEI with molecular weight being ca 4000-10000 Da was obtained from PEI 25 kDa,symbolized as PEI F25-LMW.PEI F25-LMW exhibited a much higher gene delivery efficacy.16Biodegradable cross-linked poly(amino alcohol esters)based on LMW-PEI and(coixan polysaccharide)-graft-LMW-PEI folate were also developed for gene delivery.11,17Recently,It has been found that incorporation of hydrophobic components into the cationic polymer backbone can enhance both the stability of the complex with nucleic acid and the gene delivery efficiency.18,19

In spite of these extensive efforts,one major problem still exists;these cationic delivery vehicles have strong interaction with blood components.1-3,20So far,very limited information for an effective gene therapy in the clinic has been obtained.3Therefore,development of negative-charged vectors should be a strategy for this drawback.19In this paper,we use the branched PEI with a molecular weight of 2 kDa(PEI 2 kDa),oleic acid,and oligonucleotide to construct a stable and negative-charged lipopolyplex.

2 Experimental

2.1 Materials

PEI 2 kDa(Aldrich),haloperidol(powder,Sigma),oleic acid(OA,≥99%,Sigma),3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium bromide(MTT,≥97.5%,Sigma),and trypan blue(Bioreagent powder,Aldrich).1,2-Distearoyl-sn-glycero-3-phosphoethanol-amine-N-[methoxy(polyethyleneglycol-2000)]ammonium salt(DSPE-PEG,Fig.S1 in Supporting Information)was purchased from Avanti Polar Lipids,Inc.(Alabaster,AL).The phosphorothioate oligodeoxynucleotide(ODN)with antisense sequence 5?-CCCAGCCTTCCAGTCCCTTG-3?and sense sequence 5?-CAAGGGACTGGAAGGCTGGG-3?,and 5?-FAM(fluorescein)labeled antisense sequence were synthesized by Dharmacon(Lafayette,Co.)in deprotected,desalted,and annealed form.

NCI-H-460 cells(human lung cancer cells)were obtained from American Type Culture Collection and were stably transduced with GL3 firefly luciferase gene using a retroviral vector produced in Dr.Pilar Blancafort′s lab at the University of North Carolina at Chapel Hill.The cells were maintained in RPMI 1640 cell culture medium with 10%fetal bovine serum(Invitrogen,Carlsbad,CA),100 U·mL-1penicillin,and 100 μg·mL-1streptomycin(Invitrogen,Carlsbad,CA).

2.2 Preparations of PEI/ODN polyplex and OA/PEI/ODN lipopolyplex

OA micelles were prepared by suspending OA in water containing 0.2 mmol·L-1of NaOH(final pH＞10);the final concentration of OA is 80 mmol·L-1.This solution was then vortexed briefly and agitated overnight in a shaker under argon to ensure complete dissolution.PEI/ODN polyplex was prepared by adding appropriate ODN and PEI stock solutions in H2O(1 μg of ODN is equivalent to 3 nmol of phosphate,and 0.9 μg of PEI contains 10 nmol of amine nitrogen).16The mixture was incubated for 30 min at room temperature.Lipopolyplexes were prepared by adding an appropriate amount of OA micellar solution into the polyplexes,followed by incubation at room temperature for 30 min.The final concentration of OA is 2.5 mmol·L-1and the final content of ODN is 70 ng·μL-1.The ratio of the number of N atoms in PEI to that of P atoms in ODN,i.e.N/P ratio,is lower than 36,otherwise,precipitates appear.

PEGylated lipopolyplexes were prepared by post-insertion method.Briefly,an amount of DSPE-PEG stock solution(20 mg·mL-1)was added to the preformed lipopolyplexes;the final concentration of DSPE-PEG was 10%-20%.Then the mixture was incubated at 50-60°C for 30 min.The resulting formulations were allowed to cool to room temperature before use.

The particle size was measured by using dynamic light scattering on a Malvern Zetasizer 3000(Malvern Instruments,Malvern,UK).The image of the particles was determined with a transmission electronic microscope(TEM,TECNAI 12,Philip Apparatus Co.,Netherlands)by negative staining technique.Uranyl acetate solution(1%)was used as the staining agent.

2.3 Cellular growth inhibition study

NCI-H-460 cells(5000 cells/200 μL/well)were seeded into 96-well plates(Corning Inc.,Corning,NY).After 20 h incubation,cells were treated with different formulations in serum containing medium at 37°C for 48 h.Cell viability was then detected by MTT assay.Briefly,20 μL of MTT solution(5 mg·mL-1in PBS)was added to each well,and the cells were incu-bated for 4 h at 37°C.The media were removed,and the formazan crystals formed in cells were dissolved in dimethyl sulfoxide.Absorbance(A)at 570 nm was measured by a plate reader(Bioscan Inc.,Washington DC).The percent of viable cells compared to the untreated control cells is calculated according to the following equation.

2.4 In vitro cellular uptake study

Cellular uptake was determined by fluorescence spectrometer and flow cytometry.

NCI-H-460 cells(1×105cells/0.5 mL/well)were seeded in a 24-well plate(Corning Inc.,Corning,NY)20 h before experiments.Cells were then treated with different lipopolyplexes containing 250 nmol·L-1FAM labeled ODN in RPMI 1640 medium containing 10%fatal bovine serum at 37°C for 4 h.Cells were washed three times with phosphate buffer solution(PBS)followed by incubation with 300 μL lysis buffer(0.1%Triton X-100 in PBS)at 37°C for 1 h.Fluorescence intensity of cell lysate was determined by a Perkin Elmer LS 55 spectrometer;the excitation wavelength(λex)is 494 nm and the emission wavelength(λem)is 522 nm.For free ligand competition study,cells were co-incubated with 50 μmol·L-1haloperidol and different formulations.

Flow cytometry was used to evaluate the internalization efficiency of the lipopolyplex.After incubation with lipopolyplexes,cells were rinsed three times with PBS and collected by trypsinization.The collected cells were centrifuged at 100 g for 10 min and resuspended in PBS.Internalization efficiency was evaluated as the percentage of cells with lipopolyplex(or FAM labeled ODN)inside using a FACS Calibur system from a BD FACSCanto instrument(BD Biosciences,Franklin Lakes,NJ).

Internalization efficiency=(cell numbers with FAM

3 Results and discussion

3.1 OA/PEI/ODN lipopolyplex formation

When the molecular weight of PEI is lower than or around 5000,the polyplexes of PEI with nucleic acid are unstable and show a tendency towards aggregation.12Therefore,it is not strange that the size of the polyplex formed between PEI 2 kDa and ODN is very unstable,around or bigger than 1000 nm 30 min after the preparation(exemplified by Fig.1A,where the N/P ratio is 27),and changes with time.The value of PDI(polydispersity index)is 0.475 or bigger,also time dependent.What is interesting is that the addition of OA micelles shows an obvious effect on the polyplex properties.When N/P ratio is around 10,the particles are still not stable,with a very big PDI.When N/P ratio is between 10 and 36,the addition of OA micelles can increase the stability of the polyplex and decrease the size.The particles were placed at room temperature up to 1 week and no aggregation was observed;the distribution of the particles remained a major peak at ca 133-212 nm(PDI:around 0.250)with aζpotential between-59 and-41 mV(exemplified by Fig.1(B,C),where the N/P ratio is 27).When N/P ratio is bigger than 36,precipitates appear.Thus,the N/P ratio between 10 and 36 is essential to a stable OA/PEI/ODN system.Table S1 in Supporting Information lists the values of size andζpotential.

Fig.1 Size distribution by intensity for(A)PEI/ODN/H2O and(B)OA/PEI/ODN/H2O systems and(C)the ζ potential for OA/PEI/ODN/H2O system

OA micelles can transform to vesicles when pH value is between 7 and 9.21-23The ratio of the non-ionized form to the ionized form is critical for the vesicle stability.As the pH reduces to values around 9,unilamellar liposomes begin to appear and then constitute the dominating aggregate structure until the pH value reduces to values around 8.With further acidification,the particles show an increasing tendency to aggregate.21-23

To confirm that OA,PEI,and ODN together formed the lipopolyplex,we determined the size andζpotential of the particles formed by adding OA micelles into H2O and PEI aqueous solution,respectively.When OA micelles were added into pure water,the system showed a pH value of ca.7,and hence,the particles formed were polysized(PDI:0.617,Fig.S2A)with aζpotential of-75 mV(Fig.S2B).When OA micelles were added into PEI aqueous solution,the size distribution of the particles had a single peak at ca 100 nm(exemplified by Fig.S2C,where N/P ratio is 27)with a much lower PDI and a higherζpotential(-49 mV,Fig.S2D).The pH value of OA/PEI/H2O system was around 7.5,only a little higher than that of OA/H2O system and still lower than 8.The more uniformed particles formed in the OA/PEI/H2O system should be mainly due to the N atoms in PEI,which exhibit a high affinity to the protons in the non-ionized form of OA and hence,increase the content of the ionized form.Theζpotential for the PEI/OA/ODN/H2O system is more negative than that of PEI/OA/H2Osystem(i.e.-55vs-49 mV),suggesting that OA,PEI,and ODN together form the lipopolyplex.The inset in Fig.1B shows the image of the lipopolyplex.The encapsulation of ODN in the particles(which will be discussed in the next section)also confirms the lipopolyplex formation.

Fig.2 (A)Size distribution by intensity and(B)ζ potential for OA/PEI/ODN/DSPE-PEG/H2O system

Although negatively charged vectors have weak interaction with blood component,the high negative charges on the vectors could reduce their binding affinity with cells.In practice,theζpotential should be less negative than-30 mV.DSPEPEG has been widely used to shield the positive charges on positive-charged liposomes.1-3In the present study,we found that with N/P ratio between 18 and 36,DSPE-PEG can also shield the negative charges on OA/PEI/ODN nanoparticles effectively;ζpotential was reduced to between-25 and-30 mV while the size just changed a little(Fig.2 and Table S1).Thus,the N/P ratio in the following experiments is between 18 and 36.

3.2 Encapsulation efficiency of ODN in and cytotoxicity of lipopolyplex

Fig.3A shows the UV-Vis spectra of ODN aqueous solution(curve 1),the filtrates of ODN aqueous solution(curve 2),OA/ODN aqueous solution(curve 3),and OA/PEI/ODN/H2O system(curve 4)obtained by using 100000 centrifugal filter devices(Amicon Ultra-0.5,Millipore Co.).The membrane of the centrifugal filter absorbs only a few ODN molecules(panel A,curves 1vs2).Fig.3B shows that the filter can block the OA aggregates and OA/PEI aggregates efficiently(curves 3 and 4vscurves 1 and 2).Thus,the encapsulation efficiency of ODN in OA/ODN and OA/PEI/ODN systems can be evaluated by comparing curve 2,panel A with curves 3 and 4,panel A.OA vesicles can only bind 15%of ODN(curves 3vs2,panel A),consisting with well-known information,i.e.very few nucleic acid molecules can be encapsulated in negatively charged vesicles.However,more than 80%of ODN are encapsulated in the lipopolyplex(curves 4vs2,panel A),confirming again that OA,PEI,and ODN together form the lipopolyplex.Scheme S1 illustrates the formation of the lipopolyplex.Moreover,inserting DSPE-PEG does not decrease the encapsulation of ODN(data not shown).

Fig.3 (A)UV-Vis spectra of ODN aqueous solution(curve 1)and the filtrates of ODN aqueous solution(curve 2),OA/ODN/H2O(curve 3),and OA/PEI/ODN/H2O systems(curve 4)by using 100000 centrifugal filter devices;(B)UV-Vis spectra of OA/PEI/H2O system(curve 1),OAaqueous solution(curve 2),and their filtrates(curves 3 and 4)by using 100000 centrifugal filter devices

Fig.4 Cell viability upon incubating with unmodified(i.e.naked)and modified lipopolyplexes for 48 h as demonstrated by MTT assay

Fig.4 shows the cell viability of NCI-H-460 cells following incubation for 48 h with lipopolyplexes by the MTT assay.Dif-ferent from the PEI-based vectors,11,12all the lipopolyplexes in this study show very little cytotoxicity.This should be attributed to the low molecular weight of PEI.

3.3 In vitro cellular uptake of lipopolyplex

To investigate the delivery efficiency of the lipopolyplexes,we performed the cellular uptake study using FAM labeled ODN(FAM-ODN).As shown in Fig.5,the fluorescence intensity for the cell lysate from cells treated with free FAM-ODN was very low,same as the background fluorescence,while those from the cells treated with lipopolyplexes were very high.Moreover,the addition of haloperidol significantly reduced the fluorescence intensity from the cells treated with DSPE-PEG modified lipopolyplexes,but showed very little effect on that from the cells treated with unmodified(or naked)lipopolyplexes.It is generally thought that haloperidol can bind to the surface of cancer cells with high affinity.24However,haloperidol cannot inhibit the fluorescence intensity completely.Combining with the fact that haloperidol shows very little effect on the fluorescence intensity from cells treated with naked lipopolyplexes,there should have another binding form between the lipopolyplex and cells.

Since the fluorescence intensity depends on the tightness of complex or time course of release of free nucleic acid into the cytosol,25the cellular uptake efficiency cannot be quantified from Fig.5.Flow cytometry was then used to determine the internalization efficiency.As seen in Fig.6,the percentage of the cells emitting fluorescence was higher than 90%for DSPEPEG modified lipopolyplex and 40%for naked lipopolyplex,respectively.The addition of trypan blue into the samples can quench the fluorescence from the cells treated with naked lipopolyplex markedly,while leave that from the cells treated with DSPE-PEG lipopolyplex unchanged.Trypan blue cannot penetrate into the cellular membrane.It can quench the fluorescence of the chromophore absorbed on the cellular surface,but has no effect on the fluorescence of the chromophore entering cells.Thus,trypan blue is used to differentiate internalization from absorption.Therefore,the internalized efficiency of DSPE-PEG modified lipopolyplex is higher than 90%while that of naked lipopolyplex is only～20%.

Fig.5 Fluorescence intensities of cell lysates from untreated cells and from cells treated with free FAM labeled ODN and lipopolyplexes containing FAM labeled ODN

Fig.6 Cellular uptake by flow cytometry

4 Conclusions

This study developed a negatively charged lipopolyplex based on LMW-PEI(2 kDa)and OA as a gene vector.The usage of PEI 2 kDa may expand the application of PEI in gene delivery due to the high toxicity of HMW-PEI.Although the encapsulation efficiency of nucleic acids within anionic vectors is usually very low,the negatively charged lipopolyplex discussed in the present study can encapsulate high content of nucleic acid.By modification with DSPE-PEG,this lipopolyplex can exhibit a high efficient cellular internalization.Moreover,this strategy is very simple;neither sonication nor extrusion is needed(which has been utilized widely for liposome-basedvectors to get homogeneous particles),which may imply useful information for the development of efficient vectorsin vivo.

Supporting Information:available free of chargeviathe internet at http://www.whxb.pku.edu.cn.

(1) Xu,L.;Anchordoquy,T.J.Pharmaceutical Sci.2011,100,38.doi:10.1002/jps.22243

(2) Malam,Y.;Lim,E.J.;Seifalian,A.M.Current Medicinal Chemistry2011,18,1067.doi:10.2174/092986711794940860

(3) Guo,X.;Huang,L.Accounts of Chemical Research2012,45,971.doi:10.1021/ar200151m

(4)Boussif,O.;Lezoualc′h,F.;Zanta,M.A.;Mergny,M.D.;Scherman,D.;Demeneix,B.;Behr,J.P.Proc.Natl.Acad.Sci.U.S.A.1995,92,7297.doi:10.1073/pnas.92.16.7297

(5)Godbey,W.T.;Barry,M.A.;Saggau,P.;Wu,K.K.;Mikos,A.G.J.Biomed.Mater.Res.2000,51,321.

(6)Bonner,D.K.;Zhao,X.;Buss,H.;Langer,R.;Hammond,P.T.J.Control.Release2013,167,101.doi:10.1016/j.jconrel.2012.09.004

(7)Mahato,M.;Kumar,P.;Sharma,A.K.Molecular Biosystems2013,9,780.doi:10.1039/c3mb25444e

(8) Patnaik,S.;Gupta,K.C.Expert Opinion on Drug Delivery2013,10,215.doi:10.1517/17425247.2013.744964

(9)Ogris,M.;Walker,G.;Blessing,T.;Kircheis,R.;Wolschek,M.;Wagner,E.J.Control.Release2003,91,173.doi:10.1016/S0168-3659(03)00230-X

(10) Urban-Klein,B.;Werth,S.;Abuharbeid,S.;Czubayko,F.;Aigner,A.Gene Therapy2005,12,461.doi:10.1038/sj.gt.3302425

(11)Li,S.;Wang,Y.;Zhang,J.;Yang,W.H.;Dai,Z.H.;Zhu,W.;Yu,X.Q.Mol.Biosyst.2011,7,1254.doi:10.1039/c0mb00339e

(12) Kunath,K.;Harpe,A.V.;Fischer,D.;Petersen,H.;Bickel,U.;Voigt,K.;Kissel,T.J.Control.Release2003,89,113.doi:10.1016/S0168-3659(03)00076-2

(13)Tsai,L.R.;Chen,M.H.;Chien,C.T.;Chen,M.K.;Lin,F.S.;Lin,K.M.C.;Hwu,Y.K.;Yang,C.S.;Lin,S.Y.Biomaterials2011,32,3647.doi:10.1016/j.biomaterials.2011.01.059

(14) Chen,J.;Tian,H.;Guo,Z.;Xia,J.;Kano,A.;Maruyama,A.;Jing,X.;Chen,X.Macromol.Biosci.2009,9,1247.doi:10.1002/mabi.v9:12

(15) Zhang,L.;Lu,Z.;Zhao,Q.;Huang,J.;Shen,H.;Zhang,Z.Small2011,7,460.doi:10.1002/smll.201001522

(16) Werth,S.;Urban-Klein,B.;Dai,L.;H?bel,S.;Grzelinski,M.;Bakowsky,U.;Czubayko,F.;Aigner,A.J.Control.Release2006,112,257.doi:10.1016/j.jconrel.2006.02.009

(17) Jiang,Q.;Shi,P.;Li,C.;Wang,Q.;Xu,F.;Yang,W.;Tang,G.Macromol.Biosci.2011,11,435.doi:10.1002/mabi.v11.3

(18)Roesler,S.;Koch,F.P.V.;Schmehl,T.;Weissmann,N.;Seeger,W.;Gessler,T.;Kissel,T.J.Gene Med.2011,13,123.doi:10.1002/jgm.v13.2

(19)Alshamsan,A.;Haddadi,A.;Incani,V.;Samuel,J.;Lavasanifar,A.;Uludag,H.Molecular Pharmaceutics2009,6,121.doi:10.1021/mp8000815

(20) Boura,J.S.;Santos,F.D.;Gimble,J.M.;Cardoso,C.M.P.;Madeira,C.;Cabral,J.M.S.;Silva,C.L.D.Human Gene Therapy Methods2013,24,38.doi:10.1089/hgtb.2012.185

(21) Morigaki,K.;Walde,P.Current Opinion in Colloid&Interface Science2007,12,75.doi:10.1016/j.cocis.2007.05.005

(22) Ferreira,D.A.;Bentley,M.V.L.B.;Karlsson,G.;Edwards,K.Int.J.Pharm.2006,310,203.doi:10.1016/j.ijpharm.2005.11.028

(23) Edwards,K.;Silvander,M.;Karlsson,G.Langmuir1995,11,2429.doi:10.1021/la00007a020

(24) Wang,B.;Rouzier,R.;Albarracin,C.T.;Sahin,A.;Wagner,P.;Yang,Y.;Smith,T.L.;Meric-Bernstam,F.;Marcelo,A.C.;Hortobagyi,G.N.;Pusztai,L.Breast Cancer Research and Treatment2004,87,205.doi:10.1007/s10549-004-6590-0

(25) Vader,P.;Aa,L.J.v.d.;Engbersen,J.F.J.;Storm,G.;Schiffelers,R.M.J.Control.Release2010,148,106.doi:10.1016/j.jconrel.2010.06.019