lentivirus delivery by adsorption to tissue engineering scaffolds

Lentivirus delivery by adsorption to tissueengineering scaffolds

Seungjin Shin,1 David M. Salvay,1 Lonnie D. Shea1,21Department of Chemical and Biological Engineering, Northwestern University, 2145 Sheridan Road E156,Evanston, Illinois 60208-31202The Robert H. Lurie Comprehensive Cancer Center of Northwestern University, Galter Pavilion,675 N. Saint Clair, 21st Floor, Chicago, Illinois 60611

Received 1 April 2009; revised 16 June 2009; accepted 9 July 2009Published online 13 October 2009 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.a.32619

Abstract: Biomaterial scaffolds capable of localized genedelivery are being investigated for numerous regenera-tive medicine applications and as model systems for fun-damental studies of tissue formation. In this manuscript,we investigate the delivery of lentivirus from a tissueengineering scaffold using a surface immobilization strat-egy. Poly(lactide-co-glycolide) (PLG) was employed asthe biomaterial for delivery, which has been widely usedfor a number of tissue engineering applications. The vi-rus was immobilized by freezing and subsequent lyophi-lization of the virus with the scaffold. The presence ofsucrose during freezing and lyophilization maintainedthe activity of the lentivirus, and was similar to an ade-novirus control. Collagen and fibronectin were investi-gated for their ability to enhance surface immobilization.Fibronectin modestly increased binding and transductionof the adenovirus, yet did not significantly impact the

lentivirus delivery. Most of the immobilized lentiviruswas released from the scaffold within 24 h. In vivo im-plantation of the scaffolds yielded transgene expressionthat persisted for at least 4 weeks. These findings indi-cate the potential for delivering lentivirus from tissueengineering scaffolds using a surface immobilizationstrategy. To our knowledge, this report is the first toinvestigate lentivirus delivery from porous tissue engi-neering scaffolds. Delivery of lentiviral vectors from PLGscaffolds could provide an efficient and versatile genedelivery system for use with in vitro and in vivo modelsof tissue formation, and ultimately for therapeutic appli-cations. � 2009 Wiley Periodicals, Inc. J Biomed MaterRes 93A: 1252–1259, 2010

Key words: PLG; lentivirus; tissue-engineering scaffold;surface immobilization; extracellular matrix

INTRODUCTION

Biomaterial scaffolds capable of localized genedelivery are being investigated for numerous regen-erative medicine applications and as model systemsfor fundamental studies of tissue formation. In manyinstances, vectors delivered from the scaffold in vivotarget cells within the host tissue, which can subse-quently serve as bioreactors for the localized produc-tion of tissue inductive factors. Vectors deliveredfrom the scaffold typically encode for secretedgrowth factors, such as angiogenic or neurotrophicfactors, or bone morphogenetic proteins (BMP).1–3

Since these factors are secreted into the local micro-

environment, a modest number of cells expressingthe transgene can stimulate numerous cells locally.However, an increasing number of vectors aretargeting factors that work intracellularly, such astranscription factors4,5 or shRNA.6 Vectors that targetintracellular processes will stimulate only those cellsexpressing the transgene, and will thus likely requiremore efficient delivery that can increase the numberof cells that express the transgene.

The mechanisms of gene delivery from scaffoldshave been described as a polymeric release orsurface immobilization. For the polymeric releasesystems, the vector is typically encapsulated in thematerial with the objective of obtaining a sustainedrelease. This approach requires that the vector bestable throughout the processing steps, and has beengenerally applied to nonviral approaches such asnaked plasmid, polyplexes, and lipoplexes.1,7–10 Thealternative approach involves surface immobilizationin which the scaffold is prefabricated and the vectoris subsequently immobilized. This approach avoids

Correspondence to: L. D. Shea; e-mail: [email protected] grant sponsor: NIH; contract grant numbers:

RO1 EB005678, R21 EB006520, RO1 EB003806

� 2009 Wiley Periodicals, Inc.

exposure of the vector to the biomaterials processingsteps, and may thus retain the vector activity. Thisapproach has been employed widely with nonviralvectors, including naked plasmid,11 polyplexes,12

lipoplexes,12–14 and viral vectors such as adenovi-rus.15–18 The immobilization strategies have involvednonspecific interactions between the vector andmaterial,9,12,15 layer by layer deposition,19 biotin-avi-din conjugation of the vectors to the material,16,20

and antibody binding of vectors.21 Lentiviral vectorsare emerging as a prominent research tool due tothe stable and long term expression, the hightransduction efficiency and the low immunogenic-ity,22 yet have not been employed for delivery frombiomaterials.

In this report, we investigate the delivery of lenti-viral vectors from biomaterial scaffolds composed ofpoly(lactide-co-glycolide) (PLG) using a surfaceimmobilization strategy. PLG scaffolds have beenwidely used for numerous tissue engineering appli-cations, such as bone and nerve regeneration andislet transplantation.11,23,24 We investigated immobili-zation through nonspecific binding of the virus toPLG or to extracellular matrix (ECM) coated PLG.To maximize surface association of the virus, vectorswere dried onto the scaffold following lyophilizationor freezing in the presence or cryoprotectants. Bind-ing and release of the virus were characterized onPLG disks in vitro, and the conditions that maxi-mized gene transfer were subsequently translated tomicroporous scaffolds for in vitro and in vivo testing.Delivery of lentiviral vectors from PLG scaffoldscould provide an efficient and versatile gene deliv-ery system for use with in vitro and in vivo modelsof tissue formation, and ultimately for therapeuticapplications.

METHODS

Virus production

Lentivirus and adenovirus were prepared for the studiesusing established techniques. Lentivirus was produced inHEK-293T cells grown in DMEM plus 10% FBS at 378C,and 5% CO2. The lentiviral packaging vectors (pMDL-Gag-Pol, pRSV-Rev, pIVS-VSV-G), previously described by Dullet al.25 were cotransfected along with plenti-CMV-GFP,plenti-CMV-bgal, or plenti-CMV-luciferase into 293T cellsusing Lipofectamine 2000 (Roche Biosciences, Palo Alto,CA) After 48 h of transfection, the supernatant was col-lected and filtered (0.45 micron). Viruseswere thenconcen-tratedusing PEG-it (System Biosciences, Mountain, CA),with the precipitated lentiviruses resuspended with PBS or1 M scurose PBS. The virus titer was determined by HIV-1p24 Antigen ELISA Kit (ZeptoMetrix Co., Buffalo).Recombinant adenovirus carrying GFP (rAd-GFP) wasconstructed using Transpos-AdTM method (Qbiogene,

Carlsbad, CA) according to the manufacturer’s protocol.Viruses were produced in HEK-293A cells and purified byViraBindTM Adenovirus Purification Kit (Cell Biolabs, SanDiego, CA). The adenoviral titer was determined byTCID50 assay on HEK-293A cells.

PLG disks for virus immobilization

Disks were fabricated for in vitro studies that investigatevirus activity. For disk fabrication, PLG (75:25 mole ratioof lactide to glycolide i.v. 5 0.6–0.8 dL/g) (BoehringerIngelheim Chemical, Petersburg, VA) pellets were heatedto 828C and pressed into a flat disk using a 5 kg weight.The temperature was slowly decreased from 828C to 378Cand the disks were placed at 378C overnight. Disks had aradius of �1.5 cm and were stored at room temperatureuntil use. For protein coating of PLG, disks were incubatedwith collagen (10 or 50 lg in 30 lL) or fibronectin (2 or10 lg in 30 lL) and dried overnight. PLG or ECM-coatedPLG disks were covered by 10 lL of virus solution (4 3109 LP/mL in 1 M sucrose PBS) for 1 h at 48C andlyophilized for 24 h.

Assay of active viruses

The activity of the virus was determined to investigatethe effect of virus processing. Active virus was measuredby counting tranduced cells. Samples were serially diluted(from a dilution of 1021 to 1026) in PBS and mixed withHEK-293T cells growing in a monolayer. At 3 days afterinfection, GFP positive cells were counted under the fluo-rescence microscope. Bioactivity of the surface-lyophilizedvirus on the PLG disk was assessed by measuring the GFPpositive cell area on the disk at 2 days after cell seeding.Captured image of GFP-positive area were analyzed byScion Image software (Scion, Frederick, MD). Color imageswere then converted to gray scale and thresholded tomark the pixel area. Pixel number was counted to indicatethe level of cell infection.

3D PLG scaffolds with immobilized virus

Polymer scaffolds were fabricated using a gas foaming,particulate leaching process. PLG was dissolved indichloromethane to 6% (w/w) solution, which was thenemulsified in 1% poly(vinyl alcohol) to create micro-spheres. The scaffold was constructed by mixing 1.5 mg ofmicrospheres with 50 mg of NaCl (250 lm < d < 425 lm),and then compression molding the mixture in a 5 mm KBrdie at 1500 w using a Carver press. The scaffold was thenequilibrated with high pressure CO2 gas (800 w) for 16 hin a custom-made pressure vessel. Afterwards, the pres-sure was released over a period of 25 min, which serves tofuse adjacent microspheres creating a continuous polymerstructure. To remove the salt, each scaffold was leached in5 mL of water for 4 h with fresh water replacement after2 h. Similar to the PLG disks, collagen (100 lg) or fibronec-tin (20 lg) were mixed with PLG scaffolds and dried over-night. PLG or ECM-coated PLG scaffold were mixed with

LENTIVIRUS DELIVERY BY ADSORPTION TO TISSUE ENGINEERING SCAFFOLDS 1253

Journal of Biomedical Materials Research Part A

10 lL of virus solution (1011 LP/mL in 1 M sucrose PBS)for 1 h at 48C and lyophilized for 24 h. Release studieswere performed by washing the scaffolds with PBS andsubsequent immersion in a 24 well plate, which contained500 lL of 10% FBS DMEM medium at 378C. The mediumwas sampled (10 lL) at 1, 4, 24 hr time points, seriallydiluted (from a dilution of 0 to 1024), and infected to theHEK293T cells. At 3 days after infection, GFP positive cellswere counted.

In vitro cell transduction on the 3D PLG scaffold was an-alyzed by seeding and culturing cells on PLG scaffoldswith immobilized virus. Lentivirus (10 lL) expressingbeta-galactosidase (lenti-ßgal, 1011 LP/mL in 1 M sucrosePBS) were mixed with ECM (50 lg of collagen or 10 lg offibronectin; PLG-collagen or PLG-fibronectin) and lyophi-lized with 3D PLG scaffold. Scaffolds were washed withPBS and 5 3 105 of 293 cells were seeded on the top of thescaffolds. The expression of beta-galactosidase was visual-ized at 3 days after transduction by staining with X-gal.26

Bioluminescence imaging

Lentivirus expressing luciferase (lenti-Luc, 3 3 108 LP)was lyophilized on the surface of PLG scaffolds for 24 hand implanted subcutaneously into male CD1 mice (20–22g). In vivo luciferase expression was monitored using anIVIS imaging system (Xenogen Corp., Alameda, CA),which includes a cooled CCD camera. For imaging, theanimals were injected i.p. with D-luciferin (Molecular Ther-apeutics, MI; 150 mg/kg body wt) using 28-gauge insulinsyringes. The animals were placed in a light-tight chamberand bioluminescence images were acquired (every 5 minfor a total of 20 min) until the peak light emission wasconfirmed. Gray-scale and bioluminescence images weresuperimposed using the Living Image software (XenogenCorp.). A constant size region of interest was drawn overthe implantation site. The signal intensity was reported asan integrated light flux (photons/s), which was deter-mined by IGOR software (WaveMetrics, OR). Backgroundphoton fluxes were obtained using the same proceduresbefore the injection of D-luciferin. Care and use of the labo-ratory animals followed the guidelines established by theNorthwestern University Institutional Animal Care andUse Committee (IACUC).

Statistical analysis

The statistical significance of the differences betweengroups was analyzed by one-way ANOVA followed byTukey’s post hoc test for multiple comparisons. A level ofp < 0.05 was accepted as significant.

RESULTS

Lyophilization and viral activity

Initial studies investigated conditions to lyophilizeviruses without loss of activity, which will be used

in subsequent steps to promote association of the vi-rus with the biomaterial surface. Sucrose was addedto the solution as a cryoprotectant to maintain activ-ity during the freezing and dehydration process.Viruses that express GFP (lenti-GFP, rAd-GFP) werefrozen or lyophilized in PBS or 1 M sucrose PBS andthe infectious activity were compared [Fig. 1(A,B)].The presence of sucrose during freezing relative toPBS alone significantly enhanced the virus activityfor both lentivirus and adenovirus (p < 0.01), thoughlentivirus retained �80% activity without sucrose[Fig. 1(B)]. Sucrose was essential during the lyophili-zation step, as less than 10% activity was maintainedfor lentivirus and adenovirus that was lyophilizedwithout sucrose. Lyophilization in the presence ofsucrose retained more than 90% of the activity forboth adenovirus and lentivirus. The concentration ofsucrose is critical to the virus stability [Fig. 1(C)],and the activity of adenovirus was more sensitive tothe sucrose concentration compared to the lentivirus.

Virus retention on PLG disks

We subsequently investigated the lyophilization ofvirus onto PLG disks, with and without adsorbedECM proteins, and determined the amount boundfrom the extent of GFP expression. Lyophilizationonto the disks was performed with 1 M sucrose PBSbuffer, followed by cell seeding. The activity of viruson the surface was between 6% and 7% for the lenti-virus and adenovirus (Fig. 2). PLG disks with colla-gen or fibronectin enhanced the amount of virusbound by �1% to 2% respectively; however, theseproteins did not significantly impact the binding oflentivirus.

The transduction efficiency from the surface im-mobilized virus was subsequently compared to thetransduction efficiency that can be obtained by thebolus addition of virus to the culture media, withthe latter approach being the traditional method oftransduction (Fig. 3). The transduced cell area ofsurface lyophilization was significantly increased by1.8 times with lentivirus and 2.2 times with adenovi-rus compared to that of suspended virus (p < 0.05).Different concentration and types of ECM coatinghas no significant effect on the transduction effi-ciency. These data suggest that lyophilized virus onPLG or ECM-coated PLG can enhance the transduc-tion on the surface of PLG and effectively decreasethe concentration of virus needed for transduction.The results with adenovirus are consistent withother reports for adenovirus immobilization to bio-materials.15 The transduction efficiency achievedwith lentivirus was approximately fivefold less thanthat obtained with adenovirus; however, the resultsindicate that substantial activity of the lentivirus is

1254 SHIN, SALVAY, AND SHEA


retained, which has not been previously reported.All subsequent studies focus only on lentivirusdelivery from PLG scaffolds.

In vitro release and transduction from 3Dmicroporous scaffold

Lentivirus immobilization and transduction from3D microporous scaffold was subsequent investi-gated. Lentivirus was lyophilized onto the scaffold,and the amount of virus released was characterized(Fig. 4). The percentage of virus release at 4 h was70.8%, 65.3%, and 61.3% for naked PLG, collagen-coated PLG, and fibronectin-coated PLG scaffoldrespectively. Within 24 h, more than 80% of viruswas released regardless of the scaffold condition.The rapid release from the scaffold was expected

considering the relatively low retention efficiency onthe 2D PLG disks.

Cell transduction on the 3D PLG scaffold was inves-tigated in vitro using lentiviruses expressing ß-galacto-sidase. At 3 days after cell seeding, ß-galactosidaseexpression was visualized to identify the distributionof transduced cells within the scaffold (Fig. 5). In PLGand ECM-coated PLG scaffold, cell transduction wasobserved throughout the PLG scaffold (Fig. 5). Thesedata demonstrate that surface-lyophilized lentivirusprovide effective and localized transgene expressionon the surface of tissue engineering PLG scaffold.

In vivo cell transduction on 3D PLG scaffold

Lentivirus-lyophilized PLG scaffolds were thenimplanted to mice subcutaneously to investigate the

Figure 1. Virus activity following lyophilization: (A) Images of infected cells expressing GFP (lenti-GFP, rAd-GFP) forviruses that were frozen or lyophilized in 1 M sucrose. Bar represents 100 lm. The quantity of active virus expressed as a per-centage of the titer of 1 M Sucrose PBS-Frozen control (B, C). The symbol * indicates statistical significance relative to PBS-Ly-ophilized group at p < 0.01. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]



ability to promote long term and localized expres-sion in vivo. Bioluminescence imaging was employedto quantify luciferase expression following delivery

of a lentivirus encoding for luciferase. Transgeneexpression was localized to the implantation site forall time points, indicating that expression at off-tar-get sites was minimal (Fig. 6). Additionally, trans-gene expression persisted for at least 4 weeks in vivoand was consistent for all animals implanted. Takentogether, these results indicate that lentivirus immo-bilized to microporous scaffolds may be a valuabletool to promote gene transfer in vivo.

DISCUSSION

In this manuscript, we investigate the delivery oflentivirus and adenovirus from a tissue engineeringscaffold using a surface immobilization strategy.Drying of viral and nonviral fectors onto biomaterialsurfaces has been employed to maximize surfaceimmobilization.9,15,27 For lyophilization onto the sur-face, the freezing and dehydration processes can sig-nificantly decrease the activity of the vectors, andthus cryoprotectants are used. Sucrose is a com-monly used stabilizer that is able to maintain the ac-tivity of proteins, and viral and nonviral vectors dur-ing freezing and dehydration. Lyophilization hasbeen examined as an alternative pathogen inactiva-tion procedure,28 but under the sucrose formulation,it has been used to preserve the viral activity ofadenovirus and adeno-associated virus for a longperiod.29 For both adenovirus and lentivirus, concen-trations of 0.5 M retain �80% of the virus activity,with 1 M retaining more than 95% of the activity.Sucrose concentrations on the order of 1 M havebeen previously used with adenovirus.15

Figure 3. Transduction efficiency for surface and bolus delivery: Transduction efficiency for PLG disks (0.32 cm2) coatedwith collagen (10 or 50 lg) or fibronectin (2 or 10 lg), with cells infected by bolus delivery (control) or following virusimmobilization. Transduction efficiency was calculated by the pixels/mm2. The symbol * and ** indicate statistically signif-icant differences relative to bolus viral delivery group at p < 0.05 and p < 0.01, respectively.

Figure 2. Virus retention 2D PLG disks: (A) Image ofinfected cells expressing GFP for immobilization of Lenti-GFP (4 3 107 LP) and rAd-GFP (4 3 104 TCID50) in 1 Msucrose PBS buffer on PLG, with and without ECM modi-fication. Bar represents 100 lm. (B) The amount of theretained virus was calculated by the GFP positive area.Values are mean 6 SD. [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]



Surface immobilization can be performed withretention of activity; however, the amount of immo-bilized vector is relatively low on PLG. The lowamount of immobilized lentivirus indicates arelatively low affinity for the material, which is alsoconsistent with near complete release from the scaf-fold within 24 h. Adenovirus binding to naked PLGwas similar to the lentivirus. Previous studies usingadenovirus immobilization to hydroxyapatite disks(HA) indicated that greater than 30% of the virusremained on the material for up to 16 h, indicatingthat the PLG surface is less efficient at binding ade-novirus than HA. We investigated the inclusion ofECM proteins to increase binding of the virus, assome viruses have previously been shown to associ-ate with specific ECM proteins.30 Additionally, ECMproteins have enhanced the binding and delivery ofseveral nonviral vectors through nonspecific interac-

tions.12,31 Adenovirus binding was enhanced by 50%in the presence of ECM proteins, though bindingremained relatively low (12%), and lentivirus bind-ing to the ECM modified surfaces was similar to theunmodified surfaces. Alternative strategies may facil-itate virus immobilization to surfaces, such as bioti-nylation of the virus for binding to avidin modifiedsurfaces32 or the modification of the materials withantibodies to the virus.17 These approaches have thepotential to enhance surface binding and subsequentdelivery, yet require more sophisticated modifica-tions to either the vector or the material.

Although the quantity of immobilized vector islow, the high efficiency of the vector enables sub-stantial transgene expression that persists for at least4 weeks in vivo. The expression level and duration isconsistent with previous reports of adenovirus deliv-ery from biomaterial scaffolds.15 Lentivirus is anintegrating virus, thus expression is expected to per-sist for long-times; however, long term expressionhas been observed at this site without integration.Subcutaneous delivery of naked plasmid from thescaffolds has produced transgene expression thatpersisted for 105 days, which does not likely occurby integration.1 However, the duration of expressionobserved with naked plasmid delivery at the subcu-taneous site has not extended to other sites, such asthe intraperitoneal fat or spinal cord.10,24 Interest-ingly, the expression levels with lentivirus deliverywere between 2 and 10-fold higher than thatobtained by delivering hundreds of micrograms ofnaked plasmid at the same site. Thus, the expressionlevels are similar to or greater with the lentivirusdelivery using substantially less vector.

For implantation of porous scaffolds, we havepreviously demonstrated that transgene expression

Figure 4. Virus release from 3D PLG scaffold: Lentivirusencoding GFP(Lenti-GFP, 3 3 108 LP in 1 M sucrose PBS)were deposited alone, or following mixture with ECM(PLG-Col and PLG-Fib). Released virus was measured byGFP expression after incubation with 293 cells.

Figure 5. In vitro cell transduction on 3D PLG scaffold: Transduction of 293 cells on scaffolds with deposited lentivirusenoding beta galactosdiase (Lenti-bgal, 3 3 108 LP in 1 M sucrose PBS). Lentivirus was deposited on (A) unmodified scaf-fold, and (B) collagen and (C) fibronectin modified scaffolds. X-gal staining was performed 3 days after cell seeding.[Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]



occurs immediately adjacent to the scaffold andwithin the scaffold.10 Consistent with these results,the in vivo imaging studies indicate that expressionis localized to the implant site. The cell types typi-cally targeted by the gene delivery are thoseinvolved in the response to the injury created by im-plantation and the foreign body (i.e., the scaffold).Our previous studies with nonviral vectors indicatedthat macrophages were the primary cell type thatexpressed the transgene,10 whereas others deliveringnonviral vectors have demonstrated that endothelialcells are transfected.33

In conclusion, we demonstrate the ability todeliver lentivirus from porous PLG scaffolds in vitroand in vivo using a surface immobilization strategy.The binding efficiency of the lentivirus is relativelylow and unaffected by the presence of either colla-gen or fibronectin. Though the quantity immobilizedis low, the activity remains high and is able to trans-fect large number of cells in vitro and to produce sig-

nificant and long-term expression in vivo. To ourknowledge, this report is the first to investigate lenti-virus delivery from a tissue engineering scaffold.Lentivirus is widely used as a research tool due totheir efficiency and broad tropism. Additionally,libraries of genes and shRNA sequences in lentiviralconstructs are commercially available, which in com-bination with the broad tropism and efficiency of thelentivirus, can facilitate studies of tissue formationthat may ultimately provide therapeutic strategies.

References

1. Jang JH, Rives CB, Shea LD. Plasmid delivery in vivo fromporous tissue-engineering scaffolds: Transgene expressionand cellular transfection. Mol Ther 2005;12:475–483.

2. De Laporte L, Shea LD. Matrices and scaffolds for DNAdelivery in tissue engineering. Adv Drug Deliv Rev 2007;59:292–307.

3. Huang YC, Simmons C, Kaigler D, Rice KG, Mooney DJ.Bone regeneration in a rat cranial defect with delivery of PEI-condensed plasmid DNA encoding for bone morphogeneticprotein-4 (BMP-4). Gene Ther 2005;12:418–416.

4. Trentin D, Hall H, Wechsler S, Hubbell JA. Peptide-matrix-mediated gene transfer of an oxygen-insensitive hypoxia-in-ducible factor-1alpha variant for local induction of angiogene-sis. Proc Natl Acad Sci USA 2006;103:2506–2511.

5. Phillips JE, Burns KL, Le Doux JM, Guldberg RE, Garcia AJ.Engineering graded tissue interfaces. Proc Natl Acad Sci USA2008;105:12170–12175.

6. Cheema SK, Chen E, Shea LD, Mathur AB. Regulation andguidance of cell behavior for tissue regeneration via thesiRNA mechanism. Wound Repair Regen 2007;15:286–295.

7. Whittlesey KJ, Shea LD. Nerve growth factor expression byPLG-mediated lipofection. Biomaterials 2006;27:2477–2486.

8. Huang YC, Riddle K, Rice KG, Mooney DJ. Long-term invivo gene expression via delivery of PEI-DNA condensatesfrom porous polymer scaffolds. Hum Gene Ther 2005;16:609–617.

9. Jang JH, Bengali Z, Houchin TL, Shea LD. Surface adsorptionof DNA to tissue engineering scaffolds for efficient genedelivery. J Biomed Mater Res A 2006;77:50–58.

10. Rives CB, des Rieux A, Zelivyanskaya M, Stock SR, LoweWL, Jr., Shea LD. Layered PLG scaffolds for in vivo plasmiddelivery. Biomaterials 2009;30:394–401.

11. De Laporte L, Cruz Rea J, Shea LD. Design of modular non-viral gene therapy vectors. Biomaterials 2006;27:947–954.

12. Bengali Z, Pannier AK, Segura T, Anderson BC, Jang JH,Mustoe TA, Shea LD. Gene delivery through cell culture sub-strate adsorbed DNA complexes. Biotechnol Bioeng 2005;90:290–302.

13. Bengali Z, Rea JC, Gibly RF, Shea LD. Efficacy of immobi-lized polyplexes and lipoplexes for substrate-mediated genedelivery. Biotechnol Bioeng 2009;30:2361–2368.

14. Pannier AK, Anderson BC, Shea LD. Substrate-mediateddelivery from self-assembled monolayers: Effect of surfaceionization, hydrophilicity, and patterning. Acta Biomater2005;1:511–522.

15. Hu WW, Wang Z, Hollister SJ, Krebsbach PH. Localized viralvector delivery to enhance in situ regenerative gene therapy.Gene Ther 2007;14:891–901.

16. Hu WW, Lang MW, Krebsbach PH. Development of adenovi-rus immobilization strategies for in situ gene therapy. J GeneMed 2008;10:1102–1112.

Figure 6. In vivo tranduction by lentivirus-lyophilizedPLG scaffold. (A) Bioluminescence imaging and quantifica-tion of firefly luciferase expression for 4 weeks followingsubcutaneous implantation of lentivirus immobilized PLGscaffolds. Lentivirus expressing luciferase (Lenti-luc, 3 3108 LP) was lyophilized onto the unmodified PLG scaffold.(B) Integrated light flux (photons/sec) as measured usingconstant-size regions of interest over the implant site (n 54 for day 3, n 5 7 for day 7, n 5 3 for day 14 and n 5 3for day 28 for experimental and background data). Scaf-folds lyophilized with lentivirus expressing luciferase (n),background, (l), and sham operation (^, n 5 1). Valuesare mean 6 S.E.M. [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]



17. Levy RJ, Song C, Tallapragada S, DeFelice S, Hinson JT, Vya-vahare N, Connolly J, Ryan K, Li Q. Localized adenovirusgene delivery using antiviral IgG complexation. Gene Ther2001;8:659–667.

18. Fishbein I, Alferiev I, Bakay M, Stachelek SJ, Sobolewski P,Lai M, Choi H, Chen IW, Levy RJ. Local delivery of gene vec-tors from bare-metal stents by use of a biodegradable syn-thetic complex inhibits in-stent restenosis in rat carotidarteries. Circulation 2008;117:2096–2103.

19. Jewell CM, Lynn DM. Multilayered polyelectrolyte assem-blies as platforms for the delivery of DNA and other nucleicacid-based therapeutics. Adv Drug Deliv Rev 2008;60:979–999.

20. Segura T, Anderson BC, Chung PH, Webber RE, Shull KR,Shea LD. Crosslinked hyaluronic acid hydrogels: A strategyto functionalize and pattern. Biomaterials 2005;26: 359–371.

21. Jin X, Mei L, Song C, Liu L, Leng X, Sun H, Kong D, LevyRJ. Immobilization of plasmid DNA on an anti-DNA anti-body modified coronary stent for intravascular site-specificgene therapy. J Gene Med 2008;10:421–429.

22. Bennett J. Immune response following intraocular delivery ofrecombinant viral vectors. Gene Ther 2003;10:977–982.

23. Salvay DM, Rives CB, Zhang X, Chen F, Kaufman DB, LoweWL, Jr., Shea LD. Extracellular matrix protein-coated scaf-folds promote the reversal of diabetes after extrahepatic islettransplantation. Transplantation 2008;85:1456–1464.

24. Laporte LD, Yang Y, Zelivyanskaya ML, Cummings BJ,Anderson AJ, Shea LD. Plasmid releasing multiple channelbridges for transgene expression after spinal cord injury. MolTher 2008.

25. Dull T, Zufferey R, Kelly M, Mandel RJ, Nguyen M, TronoD, Naldini L. A third-generation lentivirus vector with aconditional packaging system. J Virol 1998;72:8463–8471.

26. Bondi A, Chieregatti G, Eusebi V, Fulcheri E, Bussolati G.The use of beta-galactosidase as a tracer in immunocyto-chemistry. Histochemistry 1982;76:153–158.

27. De Laporte L, Lei Yan A, Shea LD. Local gene delivery fromECM-coated poly(lactide-co-glycolide) multiple channel bridgesafter spinal cord injury. Biomaterials 2009;30:2361–2368.

28. Uhlenhaut C, Dorner T, Pauli G, Pruss A. Effects of lyophili-zation on the infectivity of enveloped and non-envelopedviruses in bone tissue. Biomaterials 2005;26:6558–6564.

29. Croyle MA, Cheng X, Wilson JM. Development of formula-tions that enhance physical stability of viral vectors for genetherapy. Gene Ther 2001;8:1281–1290.

30. Bajaj B, Lei P, Andreadis ST. High efficiencies of gene trans-fer with immobilized recombinant retrovirus: Kinetics andoptimization. Biotechnol Prog 2001;17:587–596.

31. Bengali Z, Rea JC, Shea LD. Gene expression and internaliza-tion following vector adsorption to immobilized proteins:Dependence on protein identity and density. J Gene Med2007;9:668–678.

32. Parrott MB, Adams KE, Mercier GT, Mok H, Campos SK,Barry MA. Metabolically biotinylated adenovirus for cell tar-geting, ligand screening, and vector purification. Mol Ther2003;8:688–700.

33. Riddle KW, Kong HJ, Leach JK, Fischbach C, Cheung C,Anseth KS, Mooney DJ. Modifying the proliferative state oftarget cells to control DNA expression and identifying celltypes transfected in vivo. Mol Ther 2007;15:361–368.



lentivirus delivery by adsorption to tissue engineering scaffolds

Documents