Logout succeed
Logout succeed. See you again!

Non-Viral Vectors for Gene Therapy, Second Edition: Part 1 PDF
Preview Non-Viral Vectors for Gene Therapy, Second Edition: Part 1
1 Recent Advances in Non-viral Gene Delivery Christine C. Conwell and Leaf Huang CenterforPharmacogenetics, School ofPharmacy UniversityofPittsburgh Pittsburgh,Pennsylvania 15261 I. Introduction II. Cationic Lipids A. Ligand-mediated targeting B. Toxicity III. Cationic Polymers A. Novel polymer mixtures B. Toxicity IV. Triggered Release A. pH sensitivity B. Redox sensitivity V. Physical Delivery Methods for Naked DNA A. High pressure delivery methods B. Electroporation C. Laser beam gene transduction D. Ultrasound E. Magnetofection F. Photochemical internalization VI. Prospects Acknowledgment References AdvancesinGenetics,Vol.53 0065-2660/05$35.00 Copyright2005,ElsevierInc.Allrightsreserved. DOI:10.1016/S0065-2660(05)53001-3 4 Conwell and Huang ABSTRACT Genetherapyhasbeendeemedthemedicineofthefutureduetoitspotentialto treat many types of diseases. However, many obstacles remain before gene deliveryisoptimizedtospecifictargetcells.Overthelastseveraldecades,many approachestogenedeliveryhavebeencloselyexamined.Byunderstandingthe factors that determine the efficiency of gene uptake and expression as well as thosethatinfluencethetoxicityofthevector,wearebetterabletodevelopnew vectorsystems.Thischapterwillprovideabriefoverviewofrecentadvancesin gene delivery, specifically on the development of novel non-viral vectors. The following chapters will provide additional details regarding the evolution of non-viralgene delivery systems. (cid:1)2005,ElsevierInc. I. INTRODUCTION Genedeliveryholdsgreatpromiseasatherapeuticagentforavastarrayofmedical ailmentsincludingcancer,geneticdisordersandacquireddiseases.Theidealgene deliveryvehiclewouldexhibitcellspecificity,minimalimmuneresponse,efficient releaseofDNAintocells,andhavealargeDNAcapacity.Viralvectorssuchas attenuated viruses, adenoviruses and retroviruses, have thus far proven to have significantly more efficient gene expression than most non-viral vectors. The successofviralvectorshasbeenseverelylimitedduetothepotentialforaspecific immuneresponsetothevectorthatcouldhindergenedeliveryaswellaselicita severeinflammatoryreactionandcausenonspecificgeneintegrationintothehost genome.Non-viralvectorsarenotexpectedtoelicitaspecificimmuneresponseor randomly integrate DNA into the genomic DNA of the host and therefore are lookedtoasthefutureofgenedeliverysystems.Non-viralvectorsincludecationic polymer and lipid-based encapsulation of DNA as well as the delivery of naked DNAbyphysicalmechanisms.Althoughnon-viralvectorsareconsideredsuperior vehiclesforgenedeliveryduetotheirdecreasedimmunogenicity,theirsuccesshas been severely limited by inefficient cellular uptake and gene expression. Recent advances made in the field of non-viral gene delivery, specifically strategies to improvetargetspecificityandgeneuptakeandrelease,andtofurtherreducethe nonspecificimmuneresponse,areaddressedbrieflybelow. II. CATIONIC LIPIDS Felgneretal.reportedtheuseoflipid-basedvectorsinthelate 1980sandsince thenthesevesicleshavebeenconsideredoneofthemostpromisingmethodsfor non-viral gene delivery (Felgner et al., 1987). The cationic head groups make 1. RecentAdvancesinNonviralGeneDelivery 5 strong electrostatic associations with the DNA, eventually leading to the col- lapse of the anionic polymer. The length and degree of saturation of the lipid chainissignificantindeterminingthestabilityandtoxicityoftheliposome.To formastablecomplex,thecationiclipidisoftencombinedwithaneutrallipid and/or a helper lipid to form a liposome stable under physiological conditions. PlasmidDNAcontainingthegeneofinterestisincorporatedintotheliposome toform liposome/DNA complexes, orlipoplexes. A. Ligand-mediated targeting Significant efforts have been made to improve gene expression by modifying lipoplexestotarget-specificcelltypes.Byassuringdeliveryofthevectorstothe specific tissues, the probability of successful delivery is substantially increased. Significant advances have been made towards the targeting of tumor cells in vivo. Since cancer cells are known to over-express receptors (e.g., folate and transferrin), lipoplexes have been modified to contain ligands that are recog- nizedbytumorcells.Folatemoleculeshavebeentaggedtolipidsforincorpora- tion into the lipoplexes (Dauty et al., 2002; Zuber et al., 2003). In addition to target-specific molecules, short chain fragments of antibodies have also been used to target tumor cells. Specifically, fragments of short chain antibodies for transferrin have been covalently conjugated to the liposome to form immuno- lipoplexes (Xu et al., 2002). Both systems showed an increased affinity for binding tumor tissues in vitro and the immunolipoplex binding was also increased in vivo (Dauty et al., 2002; Xu et al., 2002; Zuber et al., 2003). To further enhance the efficiency of these targeted vectors, polyethylene glycol (PEG) was attached to the vectors prior to delivery (i.e., PEGylation). This modification was expected to increase lipoplex solubility as well as improve circulation of the vector (Ogris et al., 1999). The PEG-modified vectors were found to have enhanced binding of lipoplexes to the target cancer cells as compared tounmodified lipoplexes(Yu et al.,2004;Zuber et al., 2003). B. Toxicity The toxicity of lipoplexes has been a major limitation for their use as in vivo gene delivery systems. As mentioned above, lipid-based vectors do not trigger a cellular immune response (i.e., specific recognition), however these vectors may be recognized as foreign and initiate the production of cytokines such as tumor necrosis factor-(cid:1) (TNF-(cid:1)), interferon-(cid:2) (INF-(cid:2)), interleukin-6 (IL-6), and IL-12. The toxicity of the lipoplexes may be largely attributed to the composition of the liposomes. Optimization of lipid-based vectors (e.g., incor- poration of various concentrations of helper lipids) has been relatively exhausted in recent years, as has the development of novel cationic lipids. 6 Conwell and Huang Therefore, it has been necessary to address the toxicity of lipoplexes by other methods. Tan et al. have shown that the sequential injection of liposome and plasmid DNA can significantly reduce the inflammatory response induced by systemic gene delivery (Tan et al., 2001). By first injecting liposome, then waiting a short time and injecting the plasmid DNA, the authors were able to decrease the levels of the cytokines, TNF-(cid:1) and IL-12 by greater than 80% as compared to lipoplex delivery (Tan et al., 2001). More recently, Liu et al. described the preparation of a nonimmunostimulatory lipid-based vector. In contrast to the standard lipoplex, this vector contains lipid, DNA and an inflammatory suppressor molecule that specifically inhibits the production of the cytokine, NF-(cid:3) B (Liu et al., 2004b). These “safeplexes” successfully delivered DNA to a number of tissues, all of which displayed a significant decrease in TNF-(cid:1)ascomparedtolipoplexdelivery.Additionally,deliveryofsafeplexesdid nottriggerasignificantincreaseinthelevelsofIL-12andINF-(cid:2),whichalsoact asindicatorsoftheseverityoftheinducedimmuneresponse.Thus,theaddition of an immunosuppressor molecule within the lipoplex can significantly reduce the toxicity associated with lipid-based non-viral vectors. These advances amongothers will beaddressed indetail inChapter8. III. CATIONIC POLYMERS Cationic polymers condense DNA into compact structures by neutralizing the anionic charge on the DNA. The resulting cationic polymer/DNA complexes, or polyplexes, encapsulate the DNA into small particles for gene delivery. Commonpolycationsincludepolylysine,polyaminessuchaspolyethylenimine, histoneproteins,polyarginine-containingproteins(i.e.,protamine,HIV-TAT), andcationicdendrimers.Aswithlipids,notallcationicpolymersareoptimalfor gene delivery and issues such as efficiency and toxicity must be considered. Many cationic polymers that condense DNA can not withstand the stringent conditionsofdelivery,suchashighionicstrength,thereforeitisessentialtouse polymers thatbindtightly, butreversibly, to the DNA. A. Novel polymer mixtures Polyethylenimines (PEI) have been shown tohave oneof the highest transfec- tion efficiencies of all cationic polymers and have become a favorite non-viral genedeliveryvectorinthelastdecade(Boussifetal.,1995).PEIisappealingasa deliveryvectorbecauseithasahighchargedensityforoptimalDNAcondensa- tion as well as the ability to act as a “proton sponge”, which promotes release fromendosomalcompartments(Kircheisetal.,2001).Toxicityissues,however, havelimitedtheoverallsuccessofthiscationicpolymerforinvivogenedelivery. 1. RecentAdvancesinNonviralGeneDelivery 7 In an effort to circumvent toxicity issues while maintaining the transfection efficiency, modified PEI molecules have been designed. These modifications include investigating various molecular weights of branched and linear PEI, conjugation of PEI with PEG (Hong et al., 2004), methylation to a charged quaternary ammonium derivative ( Brownlie et al., 2004 ), and cholesterol–PEI conjugates (Furgeson et al., 2002, 2003 ), among others (Kichler, 2004). A recent study by Brownlie et al. used modular modifications of PEI to create several new variations of the molecule. The presence of PEG decreased toxicity of the PEI vector nearly 10-fold, whereas the quaternary ammonium derivate decreased the toxicity up to 4-fold (Brownlie et al., 2004). By modifying the PEI to create more biocompatible complexes, the toxicity of this vector may be reduced, providing a superior polymer for gene delivery. These alternative molecules will be addressed further in Chapter 9. Combining lipids and polymers to form new vectors (i.e., lipopoly- plexes) has provided additional options for more efficient cationic polymer-based vector delivery. Specifically, complexes containing lipid-protamine-DNA (LPD) as a modified gene delivery system were investigated for toxicity and efficiency as compared to standard lipid and polymer vectors. LPD was shown to have increased efficiency of gene delivery as compared to cationic liposomes ( Li and Huang, 1997; Li et al., 1998). A recent study by Arangoa et al. described LPD complexed with asialofetuin (AF), a target ligand-specific for the receptor found on hepatocytes in large numbers ( Arangoa et al ., 2003 ). The LPD-AF vectors had enhanced gene expression by approximately 3-fold over lipoplex-AF vectors and a 10-fold increase over the unmodified lipoplex, indicating that the presence of protamine significantly promoted gene delivery under these conditions. Fur- thermore, modified LPD vectors have been explored for treatment of tumors (Dileo et al., 2003a; Whitmore et al., 2001). Complexes containing CpG oligo- nucleotides wereshowntoelicit astrongimmuneresponse andincreasedcyto- kine production, leading to the reduction in tumor activity in vivo (Whitmore et al., 2001). The increased efficiency of LPD vectors as compared with other non-viralvectors,alongwithrecentadvancesintargetingofthecomplexes,has providedapromisingsystemfornon-viralgenedelivery. Block co-polymers have recently been developed as an adaptation of polymer-based gene delivery systems. Many of the modified vectors contain a frequentlyusedcationicpolymer(e.g.,PEI,poly-L-lysine(PLL),poly-histidine) that condenses DNA in combination with a stabilizing polymer such as PEG (Ahn et al., 2004; Miyata et al., 2004; Putnam et al., 2003). The production of copolymers allows for combinations of valuable characteristics in individual monomers to be combined to form a novel polymer. Recently, Li and Huang investigatedthecopolymer,poly(D,L-lactide-co-4-hydroxyl-L-proline)(PHLP). The polymer hydroxyproline is a component of naturally occurring substances such as gelatin and collagen, and therefore, should minimize toxicity of the 8 Conwell and Huang vector.Thisbiocompatiblecopolymerwasfoundtohaveincreasedgeneexpres- sionoverlongerperiodsoftimeandlowertoxicityascomparedtoPEIandPLL, making it an excellent candidate for further investigations (Li and Huang, 2004). Additionally, by creating biodegradable linkages between the polymers (i.e.,incorporationofestersorthiols),vectorsmaybedesignedtobeincreasingly biocompatible and have a higher propensity to release DNA under specific intracellular conditions (discussed below) (Miyata et al., 2004). Block copoly- mershavealsobeendesignedusingmoleculesthatarenotcationicinnature.A prominentexampleisthePluronic1polymers,whichconsistofvariousratiosof ethyleneoxideandpropyleneoxide(Kabanovetal.,2002).Thesepolymershave been shown to have increased gene expression as compared to naked DNA in skeletal muscle tissue. Additionally, efficient delivery of DNA was found with low concentrations of the copolymer, which further reduces toxicity issues (Lemieux et al., 2000). The use of copolymers as non-viral vectors allows for the adjustment of specific characteristics of the vector by manipulating the individual molecules, ratios and linkages incorporated, providing many oppor- tunitiestooptimizeencapsulationanddelivery. The use ofnatural molecules (e.g., albumin, chitosan, and gelatin) for theencapsulationofDNAforgenedeliveryisalsoapromisingoptionfornon- viral gene delivery. One advantage to using natural molecules is the reduced toxicity of the vector, partially due to the biodegradability of the polymer. Chitosan, a natural polysaccharide, has been successfully used to deliver DNA bothinvitroandinvivo(Chellatetal.,2005;Koping-Hoggardetal.,2004;Kumar et al., 2002; Mansouri et al., 2004). Nanospheres made from human serum albumin have also been shown to have improved transfection efficiency over naked DNA. Additionally, these vectors have minimized interactions with other intercellular components due to the presence of the albumin protein (Brzoska et al., 2004; Simoes et al., 2004). The advantages to investigating naturallyoccurringpolymersfornanospheresforgenedeliverywillbediscussed further inChapter12. B. Toxicity Thetoxicityofcationicpolymersisfrequentlyaresultofthequantityofpolymer required to achieve the optimal þ/(cid:1) charge ratio for the polyplex. In many instances, such as with PEI, the charge ratio that provides optimal efficiency is very near that which induces severe toxicity (Chollet et al., 2004). Polymer length/molecular weight has also been found to influence the toxicity of the polyplexes (Ahn et al., 2004; Fischer et al., 1999; Kramer et al., 2004; Wadhwa et al., 1997). Recent investigations have focused on creating new biodegrad- able polymers such as poly[(cid:1)-(4-aminobutyl)-L-glycolic acid] (PAGA) and a network of poly(amino ester) (n-PAE), which are expected to have reduced 1. RecentAdvancesinNonviralGeneDelivery 9 immunogenicity since they can be easily degraded within the host (Anderson et al., 2003; Lim et al., 2000, 2002). n-PAE has been found to elicit a reduced immuneresponsebutequivalenttransfectionefficiencyascomparedtoPEI(Lim et al., 2002). The development of new biodegradable polymers holds great promise for reducingthe toxicity ofcationic polymer-based delivery vectors. IV. TRIGGERED RELEASE A recurring issue with both lipid and cationic polymer-based non-viral vectors isthereleaseofDNAoncetheparticleistakenintothecell.Manyvectorsare able to efficiently bind to the target cells; however, the gene expression was lower than expected. A primary example of this is the cationic lipid containing the folate ligand. Particles were small (i.e., less than 50nm) and associationwithtargetcellswasobserved,yetgeneexpressionwasnotefficient (Dauty et al., 2002; Zuber et al., 2003). Some polymers bind DNA very tightly, whichprotectsitfromnucleasedegradationintheserum,buttheassociationis not easily reversible once the vector is inside the cell matrix. In order to enhance the escape of DNA from the vector, new polymers and lipids have been designed that are sensitive to intracellular conditions, such as decreasing pHanddenaturingconditions(AsokanandCho,2002;GuoandSzoka,2003). When the engineered vectors are exposed to specific conditions, the particles become unstable and develop defects that result in DNA leaking out of the complex. A. pH sensitivity VariationsinpHlevelshavebeenobservedinseveralcellularpathwaysaswell as within specific cellular compartments (i.e., endosomes) (Drummond et al., 2000;GuoandSzoka,2003).Forovertwodecades,thepHdecreaseobservedin the endosome has been exploited to enhance DNA release from non-viral vectors (Asokan and Cho, 2002; Yatvin et al., 1980). Destabilization of the vectors largely stems from protonation of neutral or negative components of the complex, which affects the overall structure and molecular interactions of the vector (Guo and Szoka, 2003; Thomas and Tirrell, 2000). Lipids have also been designed to contain an acid-sensitive linker region that is hydro- lyzed upon exposure to acidic conditions, creating discontinuity throughout the particle andpermitting the leakage ofDNA from the complex (Gerasimov etal.,1999;Thompsonetal.,1996).DevelopmentofpH-sensitivemoleculeshas beenshowntoincreasetheefficiencyofgeneexpressionbytheenhancedrelease of DNA into the endosomal compartment. Several pH-sensitive molecules are showninTable 1.1. 10 Conwell and Huang Table1.1. ExamplesofpH-SensitiveMoleculesUsedinNonviralGeneDelivery Modifiedfrom(AsokanandCho,2002;GuoandSzoka,2003). 1. RecentAdvancesinNonviralGeneDelivery 11 B. Redox sensitivity Thecytoplasmofthecellhasareducingenvironmentthatisusefulindenatur- ing disulfide bonds. Naturally occurring cationic polymers (i.e., protamines) active in DNA compaction in vivo have been shown to be stable at high ionic strengthsduetothepresenceofdisulfidelinkages.Inanenvironmentwithfree sulfhydryl groups, the linkages are reduced and DNA release was observed (Vilfan et al., 2004). Novel lipids and cationic polymers have cysteine residues incorporatedintotheircore,andfrequentlythelinkageofligandstothelipidsor polymersexistsviaareduciblemoiety(Dautyetal.,2001;GuoandSzoka,2001, 2003; Kwok et al., 2003). These molecules rely on basic redox chemistry to establish stable disulfide linkages intended to aid in vector efficiency (i.e., attaching target-specific ligand) and stability while also providing a method to increase DNA release from the complexonce inside the cell. V. PHYSICAL DELIVERY METHODS FOR NAKED DNA Delivery of naked DNA to cells elicits minimal immune response as compared toDNAencapsulatedinlipidsorcationicpolymers.Thelackofimmunogenici- ty of naked DNA makes it a good prospect for gene therapy. The limitations with this approach arise in that naked DNA is unprotected against nuclease degradation and the DNA does not have target specificity. Thus, the actual physicaldeliveryofnakedDNAmustbedirectedtowardsthetissuesofinterest sincenotargetligandsareattachedtotheDNA.Advanceshavebeenmadein thedevelopmentofmethodologiestoimprovetargeteddeliveryofnakedDNA. Many of the methods described briefly below will be examined in detail in Part IV.NakedDNA, Oligonucleotides andPhysical Methods. A. High pressure delivery methods Manymethodsforthe delivery ofnakedDNA involvethe useofhigh-pressure methods to force the DNA into the desired tissues. Particle bombardment methodology(i.e.,genegun)providesapromisingmechanismforgenedelivery due to the small quantities of DNA required and the minimal toxicity. To deliverDNAbygenegun,theDNAmustfirstbecoatedontogoldortungsten microparticles and then delivered to the cytoplasm of target cells by a pressur- ized blast from the gene gun (Yang et al., 1990). The success of this method has been severely limited by the small area of tissue to which vector is delivered.Anewhigh-pressuregenegunhasrecentlybeendevelopedthatuses helium gas to initiate a high-pressure blast, which delivers gold particles sus- pended in ethanol into target tissue. Delivery with the improved gene gun 12 Conwell and Huang increased gene delivery to muscle tissue by several orders of magnitude as compared to the conventional gene gun (Dileo et al., 2003b). Penetration of subdermaltissues(i.e.,muscle)morethandoubledthelengthofgeneexpression previously observed. HydroporationmethodsdeliverDNAtothetargetbyinjectingalarge volume of solution with significant force (i.e., a burst of DNA-containing solution). Recent studies by Zhang et al. have shown that by hydrodynamic injection, delivery vectors were able to reach the target tissue (i.e., liver hepa- tocytes) via tail vein injection by traveling through the hepatic vein (Zhang et al., 2004). The increased pressure from the injection increases membrane permeability by creating defects in the membrane of the hepatocytes, thereby increasing vector uptake. Hydroporation has also recently been explored for gene delivery tothe kidneys as wellas tomuscletissue(Maruyamaet al.,2002; Zhangetal.,2001).Althoughincreasedgenedeliveryandexpressionhavebeen observedbythisdeliverymethod,itremainsaveryinvasivetechniquethathas beenfoundtoincreasebloodpressureanddecreaseheartrateduetothevolume introduced into the system. Jet injection, like hydroporation, delivers a DNA-containing solution totargettissueviaahigh-pressuremechanism.Thedevelopmentofanovellow volume‘highspeedjetinjector’allowsfortheefficientdeliveryofsmallvolumes of DNA and is not affected by the length of DNA used (Walther et al., 2001). The modified jet injector can penetrate tissues up to 10 mm deep and has displayed transfection efficiencies similar to those observed with particle bom- bardmenttechniques.Genedeliverybyjetinjectionisanappealingmechanism becauseDNAremainedintactthroughoutthedeliveryprocess(i.e.,notsheared bythe pressure) with no tissuedamage orimmuneresponse observed. Another alternative to hydrodynamic injection involves manually massaging liver tissue after intravenous injection of naked DNA (Liu and Huang, 2002a). Mechanical massage of the liver (MML) by applying pressure totheabdomenofthemouseinfourshortintervals,increasedgeneuptakeinto theliverascomparedtodeliveryofnakedDNAalone;however,itwasapproxi- mately an order of magnitude lower than that observed by hydrodynamic injection. Further investigations of the mechanism by which gene expression was enhanced suggest that permeability of liver tissue is increased for several minutes after mechanical massage (Liu et al., 2004a). Delivery of DNA by MML did not have increased toxicity as compared to the control. In vivo studies of miceinhepaticfailurerevealedthatthosemicetreatedwithhepatocytegrowth factor(HGF)byintravenousinjectionfollowedbyMMLsurvived,whereasmice treatedwithemptyplasmidorleftuntreatedexpiredwithinapproximately20hr of the diagnosis (Liu and Huang, 2002a). Thus, this less invasive technique may beusedtoeffectivelydeliverDNAtotheliverwithreducedtoxicitycompared tohydrodynamic injection.