Biodegradable Alginate Microspheres as aDelivery System for Naked DNA

Neeraj Aggarwal, Harm HogenEsch, Peixuan Guo, Adam North, Mark Suckow and Suresh K. Mittal

ABSTRACT

Sodium alginate is a naturallyoccurring polysaccharide that caneasily be polymerized into a solidmatrix to form microspheres. Thesebiodegradable microspheres wereused to encapsulate plasmid DNAcontaining the bacterial B-galactosi-dase (LacZ) gene under the controlof either the cytomegalovirus(CMV) immediate-early promoteror the Rous sarcoma virus (RSV)early promoter. Mice inoculatedorally with microspheres containingplasmid DNA expressed LacZ in theintestine, spleen and liver. Inocula-tion of mice with microspheres con-taining both the plasmid DNA andbovine adenovirus type 3 (BAd3)resulted in a significant increase inLacZ expression compared to thoseinoculated with microspheres con-taining only the plasmid DNA. Ourresults suggest that adenovirusesare capable of augumenting trans-gene expression by plasmid DNAboth in vitro and in vivo.

RESUME

L'alginate de sodium est unpolysaccharide naturel qui peutetre facilement polymerise en unematrice solide pour former desmicrospheres biodegradable. Desmicrospheres furent utilisees pourencapsuler de l'ADN plasmidiqueporteur du gene de la B-galactosi-dase bacterienne (LacZ) sous con-trole du promoteur du cytome-galovirus (CMV) ou du virus dusarcome de Rous (VSR). Le gene

LacZ fut exprime dans l'intestin, larate et le foie de souris inoculeespar voie orale avec des microspherescontenant de l'ADN plasmidique.L'inoculation de souris avec desmicrospheres contenant l'ADNplasmidique et l'adenovirus bovinde type 3 causa une augmentationsignificative du niveau d'expressiondu LacZ comparativement auniveau exprime chez des sourisinoculees uniquement avec l'ADNplasmidique. Les resultats obtenussuggerent que les adenovirus peu-vent faire augmenter l'expressionde transgene par l'ADN plasmi-dique autant in vitro que in vivo.

(Traduitpar le docteur Serge Messier)

INTRODUCTION

Immunization with plasmid DNAencoding for the protein of interesthas introduced a new concept in thearea of vaccine development and genetherapy (1-4). In addition to injectionof naked DNA into muscles (5,6), anumber of delivery systems such asbombardment with gold microparti-cles coated with DNA (7-9), incorpo-ration of DNA into liposomes andother polycationic lipids (10-12), bio-logical erodable polymers (13), andothers, have been evaluated for theirtransfection efficiency in culturedcells and animals. A number of het-erologous regulatory sequences havebeen tested to determine their role inenhancing foreign gene expression inexperimental animals inoculated withplasmid DNA. The cellular uptake ofthe naked DNA when injected into amuscle is significantly less, and fol-

lowing endocytosis a large portion ofthe DNA is degraded after fusion ofendosomes (carrying the DNA) withlysosomes (14,15).

Following binding to specific cell-surface receptors (16,17), adenovirusis internalized by receptor-mediated*endocytosis within a clathrin-coatedvesicle. The endosomal membranefuses with adenoviral capsid, trig-gered by exposure of hydrophobicresidues of the adenoviral capsid pro-teins, due to a change in the endoso-mal pH (18). This process results indisruption of endosomes before theyare fused with lysosomes. Conjuga-tion of naked DNA with adenovirususing transferrin or polylysine resultsin internalization of these conjugatesby receptor-mediated endocytosis anddisruption of endosomes (19,20). Itprevents lysis of DNA by lysosomalenzymes and thus provides higher lev-els of foreign gene expression.

Purified protein and viruses can beencapsulated in alginate microspheres(21,22). Oral immunization of mice,rabbits, and cattle with microspherescontaining antigenic proteins elicitedboth humoral (mucosal and systemic)and cell-mediated immune responses.Since alginate microspheres are easyto prepare in large quantities and puri-fied protein or live virus particles canbe encapsulated, we wanted to deter-mine whether this technology couldbe used for encapsulating plasmidDNA.

This paper describes the encapsula-tion of plasmid DNA into alginatemicrospheres as a vehicle for genedelivery. The use of bovine adeno-virus type 3 (BAd3) to enhance thelevel of foreign gene expression

Department of Veterinary Pathobiology (Aggarwal, HogenEsch, Guo, North, Mittal); Laboratory Animal Practice (Suckow); School of VeterinaryMedicine, Purdue University, West Lafayette, Indiana 47907 USA.

Dr. Aggarval's present address is Institute for Animal Health, Pirbright Laboratory, Surrey, UK.Dr. Suckow's present address is Freimann Life Science Center, University of Notre Dame, Notre Dame, Indiana, USA.Address correspondence and reprint requests to: Dr. S.K. Mittal, 1290 Lynn Hall, Department of Veterinary Pathobiology, School of VeterinaryMedicine, Purdue University, West Lafayette, Indiana 47907 USA; telephone: 765-496-2894; fax: 765-494-9830; e-mail: skmittal@vet.purdue.edu.Received October 14, 1998.

Can J Vet Res 1999; 63: 148-152148

in vivo by encapsulating the plasmidDNA mixed with BAd3 is alsodiscussed.

MATERIALS AND METHODS

CELL CULTURE AND VIRUS

MDBK, PK-15, and 3T3 cells ofbovine, porcine and murine origin,respectively, were obtained fromAmerican Type Culture Collection(ATCC), and grown as monolayercultures using Eagle's minimumessential medium (MEM, Life Tech-nologies Inc.) supplemented with 10%fetalClone III (HyClone LaboratoriesInc.) and 50 ,g/mL gentamicin.BAd3, obtained from ATCC, wasgrown in MDBK cells and purified bycesium chloride density-gradient cen-trifugation (23). The titer of the puri-fied virus preparation was determinedby plaque assay on MDBK cells.

PLASMIDS

Plasmids pTK21CMVgalSV40(provided by Dr. Guo, Department ofVeterinary Pathobiology, Purdue Uni-versity, West Lafayette, Indiana) andpREP9gal (24) contain the bacterialB-galactosidase (LacZ) gene under thecontrol of the cytomegalovirus(CMV) immediate-early promoter andthe Rous sarcoma virus (RSV) earlypromoter, respectively. The LacZgene in both of these plasmid wasunder the control of the simian virus40 (SV40) polyadenylation signal.Plasmid DNA was purified by isopyc-nic centrifugation in cesium chloride-ethidium bromide gradients (25).

TRANSFECTION OF CELLS

One day before transfection, semi-confluent monolayers of PK- 15 and3T3 cells were harvested with trypsin,counted, and 1 to 2 X 105 cells wereadded into each well of 6-well platesand incubated at 37°C in 5% CO2. Thecell monolayers were washed twicewith OPTI-MEM I (Life TechnologiesInc.). Plasmid DNA (1 or 5 ,ug) wasmixed with 5 jig of Lipofectin (LifeTechnologies Inc.) and incubated atroom temperature following the man-ufacturer's protocol. The DNA-lipo-some mixture was added dropwise tothe cell monolayers, covered withOPTI-MEM I, and incubated for30 min at 370C. At this time point,a set of monolayers transfected

with either pTK21CMVgalSV40 orpREP9gal were infected with BAd3 ata multiplicity of infection (m.o.i.) of500 plaque forming units (pfu) percell. At 24 h post-transfection, themedium was replaced with MEM con-taining 5% fetalClone III. The cellswere harvested by scraping at 48 and72 h post-transfection and the cellpellet was assayed for LacZ activity.

MICROENCAPSULATION OF PLASMIDDNA AND BAD3

The protocol for preparing alginatemicrospheres was modified from apreviously described procedure (21).To obtain a total of 10 mL of suspensionof microspheres for each preparation,phosphate-buffered saline, (PBS)pH 7.2, 5 X 109 pfu purified prepara-tion of BAd3, 1.5 mg pTK21CMV-galSV40, 1.5 mg pREPgal, 1.5 mgpTK21CMVgal SV40 + 5 X 109 pfuBAd3 or 1.5 mg pREPgal + 5 X 109pfu BAd3 were mixed with sodiumalginate solution and emulsified withoil to form microspheres which werestabilized by calcium chloride andzinc chloride. Thus, 1 mL suspensionof microspheres may contain a maxi-mum of either 5 x 108 pfu BAd3,150 jig plasmid DNA, or 5 X 108 pfuBAd3 and 150 jig plasmid DNA. Themajority of microspheres were within5-10 jim in diameter as measured byMicrotrak Particle Analyzer. Allwashings generated during the pro-cess of microencapsulation were col-lected and divided into 2 portions.One portion was extracted with phe-nol and chloroform, DNA was precip-itated with ethanol and resuspended ina minimum volume of TE buffer(10 mM Tris, 1 mM EDTA, pH 8.0).Uncut DNA was run on an agarose gelby electrophoresis, stained with ethid-ium bromide and visualized under UVlight. The second portion of all wash-ings was centrifuged at 27 000 rpmfor 2 h at 4°C in Beckman TiS0.1rotor. The pellets were resuspended inminimum volumes of PBS and titratedfor infectious virus particles byplaque assay on MDBK cells.

ANIMAL INOCULATION

A total of eighteen 6- to 8-week-oldBALB/c mice were randomly groupedinto 6 groups (3 animals per group)and were inoculated orally using agavage needle and a syringe at Day 1,Day 2 and Day 3 with 1 mL suspen-

sion of alginate microspheres contain-ing either PBS, BAd3, pTK21CMV-galSV40, pREP9gal, pTK21CMV-galSV40 + BAd3, or pREP9gal +BAd3. Animals were sacrificed atDay 5 by an overdose of sodium bar-biturate and the small intestine,spleen, and liver were collected anddivided into 2 portions. One portionof various tissues was weighed andused to assay for LacZ activity. Theother portion was embedded in Tis-sue-Tek OTC compound (Miles Sci-entific, Inc.) and stored at -70°Cuntil use. This portion was used forimmunohistochemical and histochem-ical analyses.

B-GALACTOSIDASE ASSAY

The protocol to measure LacZactivity was adapted from a previ-ously described procedure (25). Thecell pellets were resuspended in thecell extraction buffer (250 mM Tris-HCl (pH 7.8), 0.5% NP40, 1 mMphenylmethylsulfonylfluoride (PMSF)),vortexed, and the supernatants weresaved for LacZ assay. The mouse tis-sues were homogenized in the cellextraction buffer using a tissumizerand the supernatants were used toassay for LacZ activity. A 40-jiLsample of various dilutions of cell ortissue extracts was mixed with350 ,iL of the sodium phosphate solu-tion (100 mM sodium phosphate(pH 7.5), 10 mM KCI, 1 mM MgSO4,and 50 mM 2-mercaptoethanol). Afterincubation at 37°C for 5 min, 132 ,iLof the ONGP solution (0.4% o-nitro-phenol B-D-galactopyranoside (ONGP)in 100 mM sodium phosphate, pH 7.5)was added and the incubation wascontinued for 1 h. The enzyme reac-tion was stopped with 172 I,L of 1 MNa2CO3 and the intensity of the yel-low color that developed was mea-sured spectrophotometrically at 420nm. Various dilutions of purified bac-terial LacZ (Sigma Chemical Co., Inc.)were used as a standard for LacZ assay.

HISTOCHEMICAL STAINING

Frozen tissues sections were pre-pared using a cryomicrotome (LeicaCM1800), were fixed with acetoneand stored at -70°C. The histochemi-cal assay to detect LacZ activityin situ was modified from a previ-ously described protocol (26). Briefly,sections were cut from frozen tissuesand fixed with 0.5% glutaraldehyde in

149

B)

- 1+5

5+5

- +5+V

5+5+V

I

I"

z

ISO

160

140 -

120 -

100 -

t0 -

60 -

40

20

048 72

Time p.ut-lramskecdi. (h)

- 1+5

5+5

- I+S+V

5+S+V

48 72

Tim p.uI-Iramuf.ci.m (h)

C)

ISo

160

140

120 -

I 100 -

I o 5m I+s+v

60 S+S+V

40

48 72TMm pms-tramfecd.m (b)

Figure 1. Expression of LacZ in cells initially transfected with plasmid DNA carrying theLacZ gene and subsequently infected with BAd3. PK-15, and 3T3 cells were transfected witheither pTK21CMVgalSV40 or pREP9gal and subsequently infected with BAd3 at a m.o.i. of500 pfu/cell. At 48 and 72 h post-transfection the cells were harvested, cell extracts were pre-pared and used to assay for LacZ activity. Mock and virus-infected cell extracts were used asnegative controls. Purified bacterial LacZ was used as a standard. Each bar representsthe mean of 2 independent samples. A) PK-15 cells transfected with pTK21CMVgalSV40,B) PK-15 cells transfected with pREP9gal, and C) 3T3 cells transfected with pTK21CMV-galSV40; 1 + 5, 1 ,ug DNA and 5 ,ug liposomes; 5 + 5, 5 ,ug DNA and 5 ,ug liposomes; 1 + 5 +V, 1 ,g DNA + 5 ,ug liposomes + BAd3; 5 + 5 + V, 5 ,ug DNA + 5 ,ug liposomes + BAd3.

PBS. After rinsing twice with PBScontaining 1 mM MgCl2, sectionswere overlayed with the X-gal solu-tion (1 mg/mL 5-bromo-4-chloro-3-indolyl-B-D-galactoside (X-gal),5.0 mM K3Fe(CN)6, 5.0 mM K4Fe-(CN)6, and 1.0 mM MgCl2 in PBS)and incubated overnight at room tem-perature. The slides were washed withPBS and counterstained with eosin.The blue deposits indicated LacZactivity.

RESULTS

B-GALACTOSIDASE EXPRESSION INCELLS TRANSFECTED WITH PLASMIDDNA AND BAD3

We tested the effciency of 2 het-erologous promoters, the cytomegalo-virus (CMV) immediate-early pro-moter and the Rous sarcoma virus(RSV) early promoter, to drive tran-sient expression of the LacZ genewhen cells initially transfected withplasmid DNA are subsequentlyinfected with BAd3. Since BAd3 doesnot replicate in either PK-15 cells or3T3 cells (data not shown), virusinfection of these cells will not causecell lysis. Expression of LacZ in PK-15 cells transfected with pTK21CMV-galSV40 or pREPgal and subse-

quently infected with BAd3 wasapproximately 1.7 to 6.4- or 0.7 to2.3-fold higher, respectively, com-pared to transfected cells withoutvirus infection (Fig. IA; 1B). Simi-larly, expression of LacZ in 3T3 cellstransfected with pTK21CMVgalSV40and subsequently infected with BAd3was approximately 4.3 to 8.6-foldhigher compared to transfected cellswithout virus infection (Fig. IC). Wedid not get detectable levels of LacZexpression with pREPgal in 3T3 cellsunder the conditions used forpTK21CMVgalSV40. The maximumlevel of LacZ expression with pREP-gal + BAd3 even in PK-15 cells wasapproximately half compared to thatof pTK21CMVgalSV40 + BAd3.Expression of LacZ in 3T3 cells withpTK21CMVgalSV40 + BAd3 wasapproximately half of that obtained inPK-15 cells. The choice of both pro-moter and cell line seems to beresponsible for the levels of reportergene expression in a transient expres-sion assay. Expression of LacZ with aplasmid carrying the LacZ gene underthe control of the mouse cyto-megalovirus (MCMV) immediate-early promoter was approximately3 times higher than that with thehuman CMV promoter (unpublishedresults). We did not find reporter gene

1.50

0aa

of

U0

300.0U

m plX-21

M pREP

M plT-21+V

3 pREP+V0.50

0.00_nuade Spleen Liver

Figure 2. Expression of LacZ in tissues ofmice inoculated with microspheres contain-ing plasmid DNA and BAd3. The 6- to8-week-old BALB/c mice were inoculatedorally at Day 1, Day 2 and Day 3 with algi-nate microspheres containing either PBS,BAd3, pTK21CMVgalSV40, pREP9gal,pTK21CMVgalSV40 + BAd3, or pREP9gal+ BAd3. At Day 5 post-inoculation, the smallintestine, spleen and liver were collected andused to assay for LacZ activity. Purified bac-trial LacZ was used as a standard. Each barrepresents the mean value from 3 mice. pTK-21, pTK21CMVgalSV40; pREP, pREP9gal;pTK-21+V, pTK21CMVgalSV40 + BAd3;pREP+V, pREP9gal + BAd3.

expression proportionate to theamount of plasmid DNA used. Thiswas mainly due to limiting amounts ofliposomes when higher amounts ofplasmid DNA were used, since byincreasing quantities of liposomeswith increasing amounts of DNA, thelevels of foreign gene expressionwere proportionate to the amount ofDNA used (data not shown). How-ever, this does not have a significanteffect on the interpretation of ourresults.

A)

Igo

160

I

140

120

100

so

60

40

0

150

MICROENCAPSULATION OF PLASMIDDNA AND BAD3

In order to determine whether algi-nate microspheres could be used forencapsulating plasmid DNA for genedelivery and effect of BAd3 on levelof foreign gene expression by plasmidDNA in vivo, we generated micro-spheres containing either plasmidDNA, virus or both. None of thewashings showed DNA in detectableamounts (data not shown) indicatingthat the efficiency of microencapsula-tion of plasmid DNA was high. Infec-tious virus titers from various wash-ings were insignificant (data notshown). Since the process of microen-capsulation did not involve any treat-ment that obviously could decreasevirus infectivity, we assumed that themicroencapsulation of BAd3 was alsoefficient.

EXPRESSION OF B-GALACTOSIDASE INTISSUES OF MICE INOCULATED WITHMICROSPHERES

Mice were orally inoculated withmicrospheres containing either plas-mid DNA, virus or both and the intes-tine, spleen and liver were collectedand analyzed for LacZ expression. Inmice inoculated with microspherescontaining only pTK21CMVgalSV40,reporter gene expression wasobserved in intestine, whereas in micereceiving microspheres containingpTK21CMVgalSV40 and BAd3,LacZ expression was observed inintestine, spleen and liver (Fig. 2). Inmice inoculated with microspherescontaining either pREPgal or pREP-gal + BAd3, LacZ expression wasobserved in intestine, spleen and liver(Fig. 2). Expression of reporter genewas comparatively less in liver witheither plasmid. In mice inoculatedwith plasmid DNA and virus, expres-sion of LacZ in intestine, spleen andliver was approximately 2-fold higherthan the expression obtained with theplasmid DNA only.The intestine, spleen and liver sec-

tions from mice inoculated withmicrospheres containing plasmidDNA and BAd3 were analyzed forLacZ expression by cytochemicalstaining. The reporter gene expressionwas evident from blue color develop-ment in some cells in all 3 tissuestested (Fig. 3 D, E, F). None of thecontrol tissues had color developmentabove background (Fig. 3 A, B, C).

Figure 3. Histochemical analysis for LacZ in tissues of mice inoculated with microspherescontaining plasmid DNA and BAd3. The 6- to 8-week-old BALB/c mice were inoculated orallyat Day 1, Day 2 and Day 3 with alginate microspheres containing either BAd3 orpTK21CMVgalSV40 + BAd3. At Day 5 post-inoculation, the small intestine, spleen and liverwere collected and embedded in OTC compound. Frozen sections were analyzed for LacZexpression using X-gal. A) liver, B) intestine, and C) spleen sections from the animal inocu-lated with BAd3. D) liver, E) intestine, and F) spleen sections from the animal inoculated withpTK21CMVgalSV40 + BAd3.

Similarly, the intestine, spleen andliver sections from mice inoculatedwith microspheres containing plasmidDNA and virus were analyzed forexpression of LacZ by immunohisto-chemical staining using a monoclonalantibody against LacZ. The reportergene expression was evident frombrown color deposits in some cells inall 3 tissues tested (data not shown).As expected, control tissues did notdevelop color above background.Since cytochemical and immunohisto-chemical examinations were notquantitative, mouse tissues were notanalyzed from all groups.

DISCUSSION

The efficiency of naked DNA entryinto cells and its survival in the cyto-plasm before it reaches the nucleusare 2 important factors that directlyaffect levels of transgene expressionin addition to the choice of heterolo-gous regulatory sequences. Receptor-mediated endocytosis could enhancethe efficiency of plasmid DNA entryinto cells (19,20). Various mecha-nisms that protect DNA from degra-dation within the cytoplasm willensure transport of the maximum

amount of DNA into the nucleus fortranscription. In order to extend therange of delivery systems for thenaked DNA, we encapsulated theplasmid DNA into alginate micro-spheres. Our results have 2 importantfindings: i) oral inoculation of micewith alginate microspheres containingplasmid DNA led to transgene expres-sion in the intestine, spleen and liver,and ii) microspheres containing boththe naked DNA and BAd3 causedenhanced reporter gene expressionin vivo compared to that of micro-spheres containing only plasmid DNA.

Since adenovirus can efficientlyenter a number of both replicating andquiescent cells, it may be used as amediator for macromolecular trans-port into cells. Exposure of culturedcells to human adenovirus andunmodified plasmid DNA led to coin-ternalization of the plasmid DNA andthe virus by receptor-mediated endo-cytosis (27). This resulted in augmen-tation of the transgene gene expres-sion compared to the cells exposedwith the plasmid DNA alone. SinceBAd3 does not replicate in culturedcells of murine or porcine origin, itwas necessary to determine whetherBAd3 entry into these cells will takeplace, resulting in enhancement in

151

gene expression of the naked DNA.There was an increase in LacZ expres-sion with BAd3 in cells transfectedwith either pTK21CMVgalSV40 orpREPgal.We did not attempt to determine the

exact mechanism involved in the pro-cess of BAd3-mediated augmentationof transgene expression by the nakedDNA, but it is probably due to inhibi-tion of DNA lysis by lysosomes. LacZexpression with the RSV promoterwas undetectable in 3T3 cells but inmouse tissues it was better than thatobtained with the CMV promoter.Therefore, it appears that there couldbe significant differences betweenin vitro and in vivo gene expressionlevels with the same gene construct.For many applications the use of onlyplasmid DNA encasulated into algi-nate microspheres may be sufficient,however, plasmid DNA could bemixed with an adenovirus if a higherlevel of transgene gene expression isrequired. Currently experiments are inprogress to determine whether encap-sulation of adenovirus with plasmidDNA affect levels and type of immuneresponse elicited against the transgeneproduct. This novel gene delivery sys-tem has potential for immunizationand gene therapy applications.

ACKNOWLEDGMENTS

We thank Drs. T.L. Bowersock andL. Woodyard for helpful discussionsand D. Kong for excellent technicalhelp. This work was supported bygrants from NIH (GM55168-01),Showalter Trust and Purdue ResearchFoundation to SKM.

REFERENCES

1. ULMER JB, DONNELLY JJ, PARKERSE, et al. Heterologous protection againstinfluenza by injection of DNA encoding aviral protein. Science 1993; 259:1745-1749.

2. ULMER JB. Elegantly presented DNAvaccines. Nat Biotechnol 1997; 15:842-843.

3. DONNELLY JJ, ULMER JB, LIU A.DNA vaccines. Life Sci 1997; 60:163-172.

4. FELGNER PL. Nonviral strategies forgene therapy. Sci Am 1997; 276: 102-106.

5. WOLFF JA, MALONE RW, WIL-LIAMS P, et al. Direct gene transfer intomouse muscle in vivo. Science 1990; 247:1465-1468.

6. DAVIS HL, WHALEN RG, DEMENEIXBA. Direct gene transfer into skeletal mus-cle in vivo: factors affecting efficiency oftransfer and stability of expression. HumGene Ther 1993; 4: 151-159.

7. YANG NS, BURKHOLDER J, ROBERTSB, MARTINELL B, MCCABE D. In vivoand in vitro gene transfer to mammaliansomatic cells by particle bombardment.Proc Natl Acad Sci USA 1990; 87:9568-9572.

8. WILLIAMS RS, JOHNSTON SA,RIEDY M, DEVIT MJ, MCELLIGOTTSG, SANFORD JC. Introduction of for-eign genes into tissues of living mice byDNA-coated microprojectiles. Proc NatlAcad Sci USA 1991; 88: 2726-2730.

9. TANG D, DEVIT M, JOHNSTON SA.Genetic immunization is a simple methodfor eliciting an immune response. Nature1992; 356: 152-154.

10. FELGNER PL, RINGOLD GM. Cationicliposome-mediated transfection. Nature1989; 337: 387-388.

11. BARTHEL F, REMY JS, LOEFFLERJP, BEHR JP. Gene transfer optimizationwith lipospermine-coated DNA. DNA CellBiol 1993; 12: 553-560.

12. DEMENEIX BA, ABDEL-TAWEB H,BENOIST C, SEUGNET I, BEHR JP.Temporal and spatial expression oflipospermine-compacted genes transferredinto chick embryos in vivo. Biotechniques1994; 16: 496-501.

13. MATHIOWITZ E, JACOB JS, JONGYS, et al. Biologically erodable micro-spheres as potential oral drug delivery sys-tems. Nature 1997; 386: 410-414.

14. ZENKE M, STEINLEIN P, WAGNERE, COTTEN M, BEUG H, BIRNSTIELML. Receptor-mediated endocytosis oftransferrin-polycation conjugates: an effi-cient way to introduce DNA intohematopoietic cells. Proc Natl Acad SciUSA 1990; 87: 3655-3659.

15. COTTEN M, WAGNER E, ZATLOU-KAL K, PHILLIPS S, CURIEL DT,BIRNSTIEL ML. High-efficiency recep-tor-mediated delivery of small and large(48 kilobase) gene constructs using theendosome disruption activity of defectiveor chemically inactivated adenovirus parti-

cles. Proc Natl Acad Sci USA 1992; 89:6094-6098.

16. BERGELSON JM, CUNNINGHAM JA,DROGUETT G, et al. Isolation of a com-mon receptor for Coxsackie B viruses andadenoviruses 2 and 5. Science 1997; 275:1320-1323.

17. HONG SS, KARAYAN L, TOURNIERJ, CURIEL DT, BOULANGER PA. Ade-novirus type 5 fiber knob binds to MHCclass I alpha2 domain at the surface ofhuman epithelial and B lymphoblastoidcells. EMBO J 1997; 16: 2294-2306.

18. SETH P. Mechanism of adenovirus-medi-ated endosome lysis: role of the intact ade-novirus capsid structure. Biochem BiophyRes Comm 1994; 205: 1318-1324.

19. CURIEL DT, AGARWAL S, WAGNERE, COTTEN M. Adenovirus enhancementof transferrin-polylysine-mediated genedelivery. Proc Natl Acad Sci USA 1991;88: 8850-8854.

20. COTTEN M, WAGNER E, BIRNSTIELML. Receptor-mediated transport of DNAinto eukaryotic cells. Methods Enzymol1993; 217: 618-644.

21. BOWERSOCK TL, HOGENESCH H,SUCKOW M, et al. Oral vaccination withalginate microsphere systems. J Cont Rel1996; 39: 209-220.

22. BOWERSOCK TL, HOGENESCH H.Oral immunization using microparticles.In: Park K, ed. Controlled Drug DeliveryChallenges and Strategies. Washington:American Chemical Society, 1997:269-288.

23. GRAHAM FL, PREVEC L. Manipula-tion of adenovirus vectors. In: Murray EJ,ed. Methods in Molecular Biology: GeneTransfer and Expression Protocols, Vol. 7.Clifton: The Humana Press, 1991:109-128.

24. SCHOLZ E, ZHANG C, GUO P. Trans-activation of the early SV40 promoter byavian infectious laryngotracheitis virus inavian cells. J Virol Methods 1993; 45:291-301.

25. SAMBROOK J, FRITSCH EF, MANI-ATIS T. Molecular Cloning: A LaboratoryManual. Cold Spring Harbor: Cold SpringHarbor Laboratory Press, 1989.

26. ENGELHARDT JF, YE X, DORANZ B,WILSON JM. Ablation of E2A in recom-binant adenoviruses improves transgenepersistence and decreases inflammatoryresponse in mouse liver. Proc Natl AcadSci USA 1994; 91: 6196-6200.

27. YOSHIMURA K, ROSENFELD MA,SETH P, CRYSTAL RG. Adenovirus-mediated augmentation of cell transfectionwith unmodified plasmid vectors. J BiolChem 1993; 268: 2300-2303.

152