Ultrasound Exposure Of Lipoplex Loaded Microbubbles Facilitates Direct Cytoplasmic Entry Of The Lipoplexes

Transcript

  Ultrasound Exposure of Lipoplex Loaded MicrobubblesFacilitates Direct Cytoplasmic Entry of the Lipoplexes Ine Lentacker, †,‡ Nan Wang, † Roosmarijn E. Vandenbroucke, † Jo Demeester, † Stefaan C. De Smedt,* ,† and Niek N. Sanders §  Laboratory of General Biochemistry and Physical Pharmacy, Ghent Research Group on Nanomedicine, Faculty of Pharmaceutical Sciences, Ghent Uni V ersity, Harelbekestraat 72, B-9000 Ghent, Belgium, and Laboratory of Gene Therapy, Department of Nutrition,Genetics and Ethology, Faculty of Veterinary Medicine, Ghent Uni V ersity, Heidestraat 19, B-9820 Merelbeke, Belgium Received September 8, 2008; Revised Manuscript Received December 3, 2008; AcceptedDecember 30, 2008 Abstract:  Recently we reported that the transfection of cells by PEGylated lipoplexes becomessignificantly better by binding the PEGylated lipoplexes to the surface of microbubbles andapplying ultrasound. To further optimize this gene delivery system it is important to understandthe working mechanism. This paper elucidates the cellular entry path of these lipoplexes. Theresults clearly show that the PEGylated lipoplexes, released from the microbubbles upon applyingultrasound, are not taken up by endocytosis, the most common route for nanoparticles to entercells. Our data demonstrate that, upon implosion of the microbubbles, the PEGylated lipoplexesare released and are most probably able to passively diffuse through the cell membrane poresor become injected in the cytoplasm of the target cells. This is attractive as the  in vivo   use ofPEGylated nanoparticles remains currently limited due to a decreased cellular uptake andinefficient escape of the PEGylated nanoparticles from the endosomes. Keywords:  Ultrasound; microbubbles; intracellular pathway; PEGylated lipoplexes; microjets Introduction The development of safe and efficient gene deliverysystems is crucial for  in  V i V o  gene therapy. Viral deliverysystems are the most efficient gene delivery systems.However, their  in V i V o  use has been limited after the deathof several patients during clinical trials with both adenoviraland adenoassociated viral (AAV) vectors. 1 Furthermore, viraldelivery systems are expensive while, especially, AAVvectors cannot host (very) large transgenes. Also the risk of insertional mutagenesis and severe immune responses limittheir  in V i V o  use. In contrast, nonviral gene delivery systemshave several advantages: easy and cheap production and thepossibility of incorporating large plasmids. 2 Furthermore theycause a relatively lower immune response. 2 To improve theefficiency of nonviral delivery systems researchers haveupgraded them with different functionalities like targetingligands and fusiogenic peptides to enable their endosomalescape. The latter is important as almost all nonviral vectorsare taken up via endocytosis. Additionally, to avoid (a)aggregation in blood and (b) interaction with blood com-pounds like albumin, the surface of many types of nonviraldelivery systems has been covered with polymers like poly * Corresponding author. Mailing address: Laboratory of GeneralBiochemistry and Physical Pharmacy, Ghent Research Groupon Nanomedicine, Faculty of Pharmaceutical Sciences, GhentUniversity, Harelbekestraat 72, B-9000 Ghent, Belgium.E-mail: [email protected]. Tel:  + 32 (09) 264 8076. Fax:  + 32 (09) 264 81 89. † Laboratory of General Biochemistry and Physical Pharmacy,Ghent Research Group on Nanomedicine, Faculty of Phar-maceutical Sciences. ‡ E-mail: [email protected]. § Laboratory of Gene Therapy, Department of Nutrition, Geneticsand Ethology, Faculty of Veterinary Medicine. (1) Marshall, E. Gene Therapy Death Prompts Review of AdenovirusVector.  Science  1999 , 2244–2245 . (2) Patil, S. D.; Rhodes, D. G.; Burgess, D. J. DNA-based therapeuticsand DNA delivery systems: a comprehensive review.  AAPS J. 2005 ,  7   (1), E61 - E77 . articles 10.1021/mp800154s CCC: $40.75  ©  2009 American Chemical Society  VOL. 6, NO. 2, 457–467  MOLECULAR PHARMACEUTICS  457Published on Web 01/30/2009  ethylene-glycol (PEG). 3 - 5 However, PEGylation drasticallylowers the transfection efficiency of nonviral vectors byhampering their cellular uptake and endosomal release.Therefore, PEG chains have been attached to lipids andpolymers via chemical bonds that become cleaved in theacidic environment of the late endosomes. However, thesynthesis of such so-called “bioresponsive” carriers remainsrather difficult. 6,7 Recently, the use of ultrasound and microbubbles hasgained more and more attention to deliver drugs, especiallynucleic acids. 8 Although microbubbles are currently used ascontrast agent in ultrasound imaging, 9 they can also provokeseveral cellular effects. At low ultrasound intensities, themicrobubbles oscillate linearly in the acoustic pressurewaves, a phenomenenon called stable cavitation. This resultsin microstreaming, which affects the cellular membrane whenthe microbubbles are located close enough to the cells. 10 Athigher ultrasound intensities, the expansions of the mi-crobubbles become larger followed by a violent collapse of the microbubbles that results in shock waves that cantemporarily perforate cell membranes (this phenomenon iscalled sonoporation). This collapse is due to the inertia of the inrushing fluid and is therefore called inertial cavitation. 11 Ultrasound assisted drug delivery has many advantages, andone of the most attractive properties is the potential for timeand space controlled delivery of drugs. On top, microbubblesand ultrasound energy are considered relatively safe as bothhave been applied in medical imaging for several years. 12 Ultrasound in combination with microbubbles has beenintensively evaluated to enhance the delivery of   naked   DNA(genes and antisense oligonucleotides) and siRNA. 13 - 21 However, to obtain a significantly higher biological effectlarge amounts of DNA and siRNA are required. This is dueto the fact that naked DNA and siRNA are sensitive todegradation by nucleases, which are widely distributed inthe body. Also, as the ultrasound induced pores in the cellmembranes are short lived, large amounts of DNA andsiRNA are required near the pores to ensure a sufficient influxinside the cells.To overcome the limitations of (a) naked DNA combinedwith ultrasound and microbubbles and (b) PEGylated li-poplexes (which suffer from an inefficient cellular uptakeand endosomal escape), we previously coupled PEGylatedlipoplexes onto ultrasound responsive microbubbles (Figure1), 22,23 the idea being that the lipoplexes will be releasedupon ultrasound treatment and that they will be more easilytransported inside the cell. Applying ultrasound to PEGylatedlipoplexes bound to microbubbles resulted in much highertransfection when compared to the transfection obtained withfree PEGylated lipoplexes with or without the use of microbubbles and ultrasound (Figure 2B). To further opti-mize this new delivery system it is necessary to understand (3) Eliyahu, H.; Servel, N.; Domb, A. J.; Barenholz, Y. Lipoplex-induced hemagglutination: potential involvement in intravenousgene delivery.  Gene Ther.  2002 ,  9  (13), 850–858 . (4) Sakurai, F.; Nishioka, T.; Yamashita, F.; Takakura, Y.; Hashida,M. Effects of erythrocytes and serum proteins on lung accumula-tion of lipoplexes containing cholesterol or DOPE as a helperlipid in the single-pass rat lung perfusion system.  Eur. J. Pharm. Biopharm.  2001 ,  52  (2), 165–172 . (5) Moghimi, S. M.; Szebeni, J. Stealth liposomes and long circulatingnanoparticles: critical issues in pharmacokinetics, opsonization andprotein-binding properties.  Prog. Lipid Res.  2003 ,  42  (6), 463–478 . (6) Meyer, M.; Wagner, E. Recent developments in the applicationof plasmid DNA-based vectors and small interfering RNAtherapeutics for cancer.  Hum. Gene Ther.  2006 ,  17   (11), 1062–1076 . (7) Wong, J. B.; Grosse, S.; Tabor, A. B.; Hart, S. L.; Hailes, H. C.Acid cleavable PEG-lipids for applications in a ternary genedelivery vector.  Mol. Biosyst.  2008 ,  4  (6), 532–541 . (8) Hernot, S.; Klibanov, A. L. Microbubbles in ultrasound-triggereddrug and gene delivery.  Ad  V . Drug Deli V ery Re V .  2008 ,  60  (10),1153–1166 . (9) Riess, J. G. Fluorocarbon-based injectable gaseous microbubblesfor diagnosis and therapy.  Curr. Opin. Colloid Interface Sci.  2003 , 8  (3), 259–266 . (10) Ohl, C.-D.; Arora, M.; Ikink, R.; de Jong, N.; Versluis, M.; Delius,M.; Lohse, D. Sonoporation from jetting cavitation bubbles.  Biophys. J.  2006 ,  91  (11), 4285–4295 . (11) Newman, C. M. H.; Bettinger, T. Gene therapy progress andprospects: Ultrasound for gene transfer.  Gene Ther.  2007 ,  14  (6),465–475 . (12) Grayburn, P. A. Current and future contrast agents.  Echocardio-graphy  2002 ,  19  (3), 259–265 . (13) Bekeredjian, R.; Chen, S. Y.; Grayburn, P. A.; Shohet, R. V.Augmentation of cardiac protein delivery using ultrasound targetedmicrobubble destruction.  Ultrasound Med. Biol.  2005 ,  31  (5), 687–691 . (14) Duvshani-Eshet, M.; Machluf, M. Therapeutic ultrasound opti-mization for gene delivery: A key factor achieving nuclear DNAlocalization.  J. Controlled Release  2005 ,  108  (2 - 3), 513–528 . (15) Kinoshita, M.; Hynynen, K. A novel method for the intracellulardelivery of siRNA using microbubble-enhanced focused ultra-sound.  Biomed. Biophys. Res. Commun.  2005 ,  335  (2), 393–399 . (16) Kinoshita, M.; Hynynen, K. Intracellular delivery of Bak BH3peptide by microbubble-enhanced ultrasound.  Pharm. Res.  2005 , 22  (5), 716–720 . (17) Manome, Y.; Nakayama, N.; Nakayama, K.; Furuhata, H. In-sonation facilitates plasmid DNA transfection into the centralnervous system and microbubbles enhance the effect.  Ultrasound  Med. Biol.  2005 ,  31  (5), 693–702 . (18) Mehier-Humbert, S.; Bettinger, T.; Yan, F.; Guy, R. H. Plasmamembrane poration induced by ultrasound exposure: Implicationfor drug delivery.  J. Controlled Release  2005 ,  104  (1), 213–222 . (19) Newman, C. M.; Lawrie, A.; Brisken, A. F.; Cumberland, D. C.Ultrasound gene therapy: On the road from concept to reality.  Echocardiography  2001 ,  18  (4), 339–347 . (20) Pislaru, S. V.; Pislaru, C.; Kinnick, R. R.; Singh, R.; Gulati, R.;Greenleaf, J. F.; Simari, R. D. Optimization of ultrasound-mediated gene transfer: comparison of contrast agents andultrasound modalities.  Eur. Heart J.  2003 ,  24  (18), 1690–1698 . (21) Vannan, M.; McCreery, T.; Li, P.; Han, Z. G.; Unger, E.; Kuersten,B.; Nabel, E.; Rajagopalan, S. Ultrasound-mediated transfectionof canine myocardium by intravenous administration of cationicmicrobubble-linked plasmid DNA.  J. Am. Soc. Echocardiogr. 2002 ,  15  (3), 214–218 . articles  Lentacker et al. 458  MOLECULAR PHARMACEUTICS  VOL. 6, NO. 2  the mechanism which explains this higher transfectionefficiency. Therefore, the aim of this work was to elucidatethe cellular pathway by which PEGylated lipoplexes, uponrelease from the microbubbles by ultrasound, enter the cells. Materials and Methods Preparation and Characterization of Lipid MicrobubblesContaining DSPE-PEG-Biotin.  Liposomes containing DPPC(dipalmitoylphosphatidylcholine) and DSPE-PEG-biotin (1,2-distearoyl- sn -glycero-3-phosphoethanolamine-  N  -(biotinyl(poly-ethyleneglycol)2000)) in a 95:5 molar ratio were preparedas previously described. 24 Briefly, the lipids were put in around-bottomed flask, dissolved in chloroform. Subsequently,the solvent was removed via evaporation followed byflushing with nitrogen. The obtained lipid film was hydratedin HEPES buffer (20 mM HEPES, pH 7.4) at a final lipidconcentration of 5 mg/mL and incubated overnight at 4  ° Cto allow the formation of liposomes. The resulting liposomeswere first extruded through a polycarbonate membrane (poresize of 0.2  µ m) using a miniextruder (Avanti Polar Lipids,Alabaster, AL). Subsequently, the extruded liposomes weresonicated with a 20 kHz probe (Branson 250 sonifier,Branson Ultrasonics Corp., Danbury, CT) in the presenceof perfluorobutane gas (C 4 F 10 , MW 238 g/mol, F2 chemicals,Preston, Lancashire, U.K.). After sonication the microbubbleswere washed (to remove the excess of lipids) with 3 mL of fresh HEPES buffer and finally resuspended in 5 mL. Toenable the attachment of biotinylated lipoplexes, the bioti-nylated microbubbles were incubated with 500  µ L of avidin(10 mg/mL) and incubated for 10 min at room temperature.Subsequently, the microbubbbles were centrifuged andwashed again with 3 mL of fresh HEPES buffer. Finally themicrobubbles were resuspended in 5 mL of HEPES buffer.The concentration of the avidinylated microbubbles in thedispersions was determined with the aid of a Burker chamberand a light microscope and equaled 4 × 10 8 microbubbles/ mL. Preparation and Characterization of PEGylated Cat-ionic Liposomes and Lipoplexes.  The cationic lipid DOTAP(  N  -(1-(2,3-dioleoyloxy)propyl)-  N  ,  N  ,  N  -trimethylammonium (22) Lentacker, I.; De Geest, B. G.; Vandenbroucke, R. E.; Peeters,L.; Demeester, J.; De Smedt, S. C.; Sanders, N. N. Ultrasound-responsive polymer-coated microbubbles that bind and protectDNA.  Langmuir   2006 ,  22  (17), 7273–7278 . (23) Vandenbroucke, R. E.; Lentacker, I.; Demeester, J.; De Smedt,S. C.; Sanders, N. N. Ultrasound assisted siRNA delivery usingPEG-siPlex loaded microbubbles.  J. Controlled Release  2008 ,  126  (3), 265–273 . (24) Sanders, N. N.; Van Rompaey, E.; De Smedt, S. C.; Demeester,J. Structural alterations of gene complexes by cystic fibrosissputum.  Am. J. Respir. Crit. Care Med.  2001 ,  164  (3), 486–493 . Figure 1.  Schematic presentation of lipoplex loadedmicrobubbles. Figure 2.  (A) Schematic representation of theexperimental setup. Cells were grown on one side of anOpticell unit. For ultrasound exposure, Opticell plateswere turned upside down. In this way, microbubbleswere able to rise against the cell layer. (B) Thetransfection efficiency of lipoplex loaded microbubblesin the presence of ultrasound (black bars) compared tothe transfection efficiency of naked DNA and free 15mol % PEGylated lipoplexes in the absence (light graybars) and presence of microbubbles and ultrasound(dark gray bars). The background luciferase signal inuntreated cells is also shown (white bars). The trans-fection results, i.e., the extent of luciferase expression, areexpressed as RLU (RLU: relative light units) per mgprotein. * p  < 0.05. Ultrasound Exposure of Lipoplex Loaded Microbubbles  articles VOL. 6, NO. 2  MOLECULAR PHARMACEUTICS  459  chloride), the phospholipid DOPE (dioleoyl phosphatidylethanolamine), DSPE-PEG, DSPE-PEG-biotin and choles-teryl Bodipy FLC12 (cholesteryl 4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza- s -indacene-3-dodecanoate) were pur-chased from Avanti Polar Lipids. Cationic liposomescontaining DOTAP and DOPE in a 1:1 molar ratio with 0to 15 mol % DSPE-PEG or DSPE-PEG-biotin were preparedas described above.For the preparation of lipoplexes we used plasmid DNA(pDNA; pGL3, Promega, Leiden, The Netherlands) contain-ing the luciferase gene from  Photinus pyralis  as reporter (i.e.,firefly luciferase). The pDNA was amplified in  Escherichiacoli  and purified as described elsewhere. 24 The pDNA wasdissolved in HEPES buffer, and the concentration was setat 1.0 mg/mL, taking into account that the absorption at 260nm of a 50  µ g/mL DNA solution equals 1. The pDNAshowed a high purity as the ratio of the absorption atrespectively 260 and 280 nm was between 1.8 and 2.0.Lipoplexes were prepared at a charge ratio of 4. The chargeratio is defined as the ratio of the number of the positivecharges (srcinating from DOTAP) to the number of thenegative charges (srcinating from the pDNA). pDNA wasfirst diluted in HEPES buffer to a concentration of 0.41 mg/ mL. Subsequently, the diluted pDNA was added to an equalvolume of cationic liposomes (5 mM DOTAP) resulting ina final  (  charge ratio of 4. Immediately after the additionof pDNA to the cationic liposomes, HEPES buffer was addeduntil the final concentration of pDNA in the system was0.126 mg/mL. This mixture was then vortexed and incubatedat room temperature for 30 min. To fluorescently label theliposomes 0.5 mol % cholesteryl Bodipy FLC12 (cholesteryl4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza- s -indacene-3-dodecanoate) was used (Molecular probes, Eugene OR). Melanoma Cells Stably Expressing Renilla Luciferase(rLuc).  BLM_rLuc cells stably expressing renilla luciferasewere generated by transfecting BLM-cells (melanoma cells) 25 with the pGL4.76_CMV plasmid. The pGL4.76_CMVplasmid was generated by ligating the PCR amplified(forward primer AATAGTCGACTAGTTATTAATAG-TAATCAA and reversed primer AATAGGATCCGATCT-GACGGTTCACTAAAC) and  Sal I/   Bam HI double digestedCMV promotor into the  Xho I/   Bgl II double digested pGL4.76plasmid (Promega, Leiden, The Netherlands). The resultingpGL4.76_CMV plasmid was linearized with the  Bam HIrestriction enzyme and complexed with linear polyethylen-imine (PEI; 22 kDa) to transfect the BLM cells. Transfectedcells were incubated in fresh medium for 48 h and thenselected with 250  µ g/mL hygromycin. After two weeks,clones were isolated and expanded. Subsequently, the gener-ated clones were analyzed and a renilla luciferase positiveclone was selected. Transfection Experiments.  BLM-cells were cultured inDulbecco’s modified Eagle’s medium (DMEM) with thegrowth factor F12 and phenol red containing 2 mMglutamine, 10% heat deactivated fetal bovine serum (FBS),1% penicillin-streptomycin (Gibco, Merelbeke, Belgium) andHEPES buffer (100 mM, pH 7.4). Cells were grown to 90%confluency in OptiCell units (Biocrystal, Westerville, OH)in a humidified incubator at 37  ° C and 5% CO 2 . Subse-quently, cells were washed with 10 mL of phosphate bufferedsaline (PBS, Gibco) and the transfection medium was added.A first transfection medium was prepared by mixing 130  µ L of PEGylated lipoplexes with 1 mL of the microbubblesuspension (containing 4 × 10 8 microbubbles). After 5 minof incubation at room temperature, OPTIMEM (Gibco,Merelbeke, Belgium) was added to a final volume of 10 mL.A second transfection medium was prepared in a similar wayexcept that the 130  µ L of PEGylated lipoplexes was replacedby an equal volume of HEPES buffer containing 16.5  µ g of pDNA, the same amount as present in the lipoplexes.The 10 mL of transfection medium was completely addedto the OptiCell units (surface 50 cm 2 ). Subsequently, the cellswere placed in a water bath at 37  ° C with an absorbing rubberat the bottom and immediately subjected to ultrasoundradiation. The ultrasound radiation was performed for 10 swith a sonitron 2000 (RichMar, Inola, OK) equipped with a22 mm probe. A schematic representation of the experimentalsetup used is displayed in Figure 2A. In all ultrasoundexperiments the following ultrasound settings were applied:1 MHz, 10% duty cycle and an ultrasound intensity of 2W/cm 2 . The areas treated with ultrasound were marked, andafter radiation, the OptiCells were incubated for an additional2 h at 37  ° C. At the end of this incubation period, thetransfection medium was removed and the cells were washedtwo times with PBS, before adding fresh culture medium.Each transfection was performed in 3-fold. Twenty-fourhours after transfection the firefly luciferase expression bythe cells was analyzed. Therefore, the culture medium wasremoved and cells were washed with PBS. The areas exposedto ultrasound (20 mm diameter) were cut from the opticellmembrane and brought into a 24-well plate. Eighty micro-liters of CCLR (Cell Culture Lyse Reagent, Promega, Leiden,The Netherlands) buffer was added to each well andincubated at room temperature for at least 20 min to allowcell lysis. Twenty microliteres of the cell lysate wastransferred to a 96-well plate and the luciferase activity wasmeasured using the Glomax 96 Microplate Luminometer(Promega) as described elsewhere. 26 An aliquot (20  µ L) of each cell lysate was also analyzed for protein concentrationusing the BCA protein Assay (Pierce, Rockford, IL).Transfection results are expressed as relative light units(RLU) per mg protein.In the transfection experiment with methyl-   -cyclodextrin(M   CD), the BLM cell line stably expressing renilla lu- (25) Quax, P. H. A.; Vanmuijen, G. N. P.; Weeningverhoeff, E. J. D.;Lund, L. R.; Dano, K.; Ruiter, D. J.; Verheijen, J. H. MetastaticBehavior of Human-Melanoma Cell-Lines in Nude-Mice Cor-relates with Urokinase-Type Plasminogen-Activator, Its Type-1Inhibitor, and Urokinase-Mediated Matrix Degradation.  J. Cell Biol.  1991 ,  115  (1), 191–199 . (26) von Gersdorff, K.; Sanders, N. N.; Vandenbroucke, R.; De Smedt,S. C.; Wagner, E.; Ogris, M. The internalization route resultingin successful gene expression depends on polyethylenimine bothcell line and polyplex type.  Mol. Ther.  2006 ,  14  (5), 745–753 . articles  Lentacker et al. 460  MOLECULAR PHARMACEUTICS  VOL. 6, NO. 2  ciferase (BLM_rLuc) was used. The cells were preincubatedfor two hours with M   CD before the addition of thelipoplexes or lipoplex loaded microbubbles. Ultrasoundradiation was carried out as described above. Both, fireflyand renilla luciferase were measured with the Dual-Lu-ciferase Reporter Assay System (Promega, Leiden, TheNetherlands) in the Glomax 96 Microplate Luminometer(Promega). Results were expressed as firefly RLU/renillaRLU. Photochemical Internalization Experiments.  The pho-tosensitizer (PS) TPPS2a, meso-tetraphenylporphine with twosulfonate groups on adjacent phenyl rings, was kindlyprovided by Dr. Anders Høgset (PCI Biotech AS, Oslo,Norway). The PS was light protected and stored at 4  ° C untiluse. Cells were exposed to blue light from the LumiSource,a bank of four light tubes emitting light in the region of 375 - 450 nm, with 13 mW/cm 2 irradiance (PCI Biotech AS,Oslo, Norway). Cells were incubated overnight with 0.8  µ g/ mL PS. The day after, the PS was removed and cells wereincubated with lipoplexes or transfected with lipoplex loadedmicrobubbles and exposed to ultrasound. After 2 h incubationtime, the transfection medium was removed and cells wereincubated at 37  ° C for an additional 2 h with culture medium.Subsequenly the cells were exposed to the Lumisource for40s and again placed at 37  ° C. Cells were analyzed 24 hlater. Confocal Experiments.  BLM-cells were seeded intoculture dishes or opticell plates one day before the confocalexperiment. The cell membrane was labeled with concanava-lin A-Alexa647 (Molecular Probes). The concanavalin Astock solution was diluted 10-fold in OPTIMEM and addedto the cells immediately before visualization. Cells wereincubated with the free PEGylated lipoplexes for respectively30 and 150 min at 37  ° C and 5% CO 2 . Ultrasound exposureof BLM-cells was performed immediately after addition of the lipoplex loaded microbubbles and cells were incubatedthen at 37  ° C for respectively 30 and 150 min. After theincubation time, cells were visualized using a Nikon EZC1-si confocal laser scanning microscope (Nikon, Brussels,Belgium) equipped with a 40 × objective. The 491 nm lineof this microscope was used to excite the Bodipy label. The639 nm line was used to excite concanavalin A-Alexa647. Propidium Iodide (PI) Uptake.  PI was added to the cellsin a concentration of 25  µ g/mL in OPTIMEM. The PI wasadded either before the exposure of the cells to lipoplexloaded microbubbles and ultrasound or afterward. The uptakeof PI resulted in the appearance of red fluorescent nuclei,that were visualized with the 639 nm laser of the NikonEZC1-si confocal microscope. Statistical Analysis.  All the data in this report areexpressed as mean  (  standard deviation (SD). For thetransfection results, Student’s  t   test was used to determinewhether data groups differed significantly from each other.A  p -value lower than 0.05 was considered statisticallysignificant. Results and Discussion Design and Transfection Efficiency of Lipoplex LoadedMicrobubbles.  As schematically presented in Figure 1, weearlier succeeded in coupling highly PEGylated lipoplexesonto lipid microbubbles with the aid of an avidin - biotinlink. 22 Biotinylated lipid microbubbles consisting of DPPCand DSPE-PEG-biotin were prepared next to DOTAP/DOPEbased lipoplexes with 15 mol % DSPE-PEG-biotin. Subse-quently, avidin was bound to the biotinylated microbubblesand mixed with the biotinylated lipoplexes. As publishedpreviously, 22 exposure of these lipoplex loaded microbubblesto ultrasound caused a massive release of intact lipoplexesand drastically increased the transfection efficiency of thePEGylated lipoplexes. Figure 2B shows the transfection of cells by respectively free PEGylated lipoplexes (light graybars), PEGylated lipoplexes physically mixed with mi-crobubbles (dark gray bars) and lipoplex loaded microbubblesafter exposure to ultrasound (black bars). Only the PEGylatedlipoplexes that were coupled to the microbubbles were ableto transfect the cells after ultrasound radiation (black bars).It has been postulated that sonication of free lipoplexes couldincrease their transfection efficiency, but this was notconfirmed in our experiments 27 - 29 in agreement with theobservations by Mehier-Humbert and colleagues. 18 Consider-ing the giant increase in transfection efficiency after couplingthe PEGylated lipoplexes to the microbubbles, we wantedto elucidate the differences in cellular uptake betweenrespectively “free” PEGylated lipoplexes and PEGylatedlipoplexes released from the lipoplex loaded microbubblesby ultrasound. Which Mechanism Do Lipoplexes Use To Enter Cellsafter Exposure of Lipoplex Loaded Microbubbles toUltrasound? (a) Influence of the Endocytic InhibitorMethyl-   -cyclodextrin (M   CD).  The intracellular uptakeof lipoplexes has been intensively studied and has beenascribed to endocytosis. 30 It is believed that the main reasonsfor the low transfection efficiency of highly PEGylatedlipoplexes are their limited endocytic uptake (as the PEGchains prevent association of the lipoplexes with the cellularmembrane) and, especially, their difficulties to escape from (27) Unger, E. C.; McCreery, T. P.; Sweitzer, R. H. Ultrasoundenhances gene expression of liposomal transfection.  In V est. Radiol. 1997 ,  32  (12), 723–727 . (28) Koch, S.; Pohl, P.; Cobet, U.; Rainov, N. G. Ultrasoundenhancement of liposome-mediated cell transfection is caused bycavitation effects.  Ultrasound Med. Biol.  2000 ,  26   (5), 897–903 . (29) Lawrie, A.; Brisken, A. F.; Francis, S. E.; Cumberland, D. C.;Crossman, D. C.; Newman, C. M. Microbubble-enhanced ultra-sound for vascular gene delivery.  Gene Ther.  2000 ,  7   (23), 2023–2027 . (30) Wasungu, L.; Hoekstra, D. Cationic lipids, lipoplexes andintracellular delivery of genes.  J. Controlled Release  2006 ,  116  (2), 255–264 . Ultrasound Exposure of Lipoplex Loaded Microbubbles  articles VOL. 6, NO. 2  MOLECULAR PHARMACEUTICS  461