adenovirus-mediated gene transfer in the midgestation fetal mouse

tawhpfltn

tprpele

tbirHalHnbl

wts

CI

Mf

JA

0CA

Adenovirus-Mediated Gene Transfer in the Midgestation Fetal Mouse

G. S. Lipshutz, M.D.,*,1 L. Flebbe-Rehwaldt, Ph.D.,† and K. M. L. Gaensler, M.D.†,2

ournal of Surgical Research 84, 150–156 (1999)rticle ID jsre.1999.5588, available online at http://www.idealibrary.com on

*Department of Surgery and †Department of Medicine, University of California at San Francisco, San Francisco, California 94143

de

gmg

btpoudtgtcypn

hAdasbbdcgvht

Presented at the Annual Meeting of the Association for Aca

Background. The development of strategies for generansfer in utero will make possible the amelioration,nd eventually the cure, of genetic diseases associatedith pre- and postnatal morbidity and mortality. Weave developed a murine model for in utero, intrahe-atic, adenovirus-mediated gene transfer in Day 15etuses and compared the level and distribution ofuciferase reporter gene expression in newborns withhose observed in adult animals injected intrave-ously.Material and methods. CD-1 fetuses underwent in-

rahepatic injection on Day 15 of gestation with 1 3 107

article-forming units (PFU) of an E1- and E3-deletedecombinant adenovirus containing the luciferase re-orter gene or with normal saline. At birth, pups wereuthanized, and the brain, heart, intestine, liver,ungs, and spleen harvested and analyzed for lucif-rase activity.Results. Two adenovirus-injected litters proceeded

o term and one female aborted. Tissues from 10 new-orn mice in the experimental group and 5 newbornsn the control group were analyzed; tissues from theemaining newborns were reserved for other studies.igh-level luciferase expression was detected in alldenovirus-injected newborn livers. Lower levels ofuciferase activity were detected in distant organs.epatic toxicity as determined by serum transami-ase elevations was observed in adult, but not in new-orn mice previously injected with the adeno-

uciferase virus.Conclusions. In utero intrahepatic gene deliveryith adenoviral vectors in the developing murine fe-

us is feasible and produces high-level gene expres-ion. These studies suggest that viral and nonviral

1 G.S.L. was supported in part by a scholarship from the American

wetw

ollege of Surgeons and K.M.L.G. was supported in part by Nationalnstitutes of Health Grant HL-53762.

2 To whom correspondence should be addressed at Department ofedicine, 3rd and Parnassus, U-436, Box 0793, University of Cali-

ornia, San Francisco, CA 94143.

150022-4804/99 $30.00opyright © 1999 by Academic Pressll rights of reproduction in any form reserved.

mic Surgery, Seattle, Washington, November 18–22, 1998

ene delivery vectors may be useful in the develop-ent of future approaches to prenatal treatment of

enetic disorders. © 1999 Academic Press

Key Words: in utero; gene therapy; adenovirus.

INTRODUCTION

Advances in molecular genetics have made it possi-le to define the genetic basis of many disorders andhen to screen for the presence of these mutationsrenatally. The development of approaches to replacer modify the expression of these abnormal genes intero will in the future ameliorate or even cure theseiseases. As the tissue sites of expression and regula-ory mechanisms that direct stage- and tissue-specificene expression are better defined, it will be possible toarget therapeutic gene expression to the appropriateell types and stage of development. In addition, anal-sis of gene expression following transfer in utero willrovide a valuable tool for understanding the mecha-isms of gene regulation during ontogeny.Replication-deficient recombinant adenoviral vectors

ave been widely used in gene transfer studies [1–8].denovirus is a nonenveloped virus with a 36-kbouble-stranded linear DNA genome that encompassesseries of early and late genes encoding regulatory and

tructural proteins, respectively. Advantages of recom-inant adenoviruses for gene transfer are that they cane grown and purified to high concentrations, trans-uce both dividing and terminally differentiated cells,arry large segments of cDNA, and produce high-levelene expression [1, 9]. In addition, these recombinantiruses have been extensively studied, their genomesave been characterized, and the viruses have beenested as gene delivery vectors in adult mammals
here they exhibit hepatic tissue tropism [2]. Genexpression following transduction with adenoviral vec-ors is transient and typically persists for severaleeks. The relatively short-lived expression, hepato-

tgvl

ataewigtaitopm

fiscffldb

EgasfpigsmCwpicI1Paw

s[eDomtd

sswt

iabtpSmijdWapmairu1w

adflwphbwt(4(r

aawveidp

wotacls

ipit

EN

oxicity of adenoviral proteins, as well as the immuno-enicity of adenoviruses, has limited the utility of theseectors for gene therapy purposes where long-term orifelong therapeutic gene expression is required [10].

The transfer of both reporter genes and biologicallyctive genes by adenoviral vector-mediated transfer inhe adult animal have been intensively investigatednd produce high-level gene expression. To assess thefficiency of in utero intrahepatic gene transfer in mice,e have used recombinant adenoviral vectors contain-

ng the luciferase reporter gene. In previous in uteroene transfer studies, adenovirus-mediated generansfer by injection into the amniotic fluid in ovinend rodent models has demonstrated expression lim-ted to the skin and proximal alimentary and respira-ory tracts of the fetus [11, 12]. Intratracheal deliveryf adenoviral vectors in ovine fetuses produced re-orter gene expression that was limited to the proxi-al airways [13].The fetal liver is an important potential target organ

or future gene therapy of a variety of genetic diseasesncluding thalassemia, Gaucher’s disease (and othertorage diseases), ornithine transcarbamylase defi-iency, and hemophilia. In these studies, we have per-ormed direct intrahepatic injection of midgestationetuses with an adenoviral vector containing the fireflyuciferase reporter gene and have examined the tissueistribution and level of luciferase expression in new-orn pups 5 days after in utero injection.

METHODS

Preparation of recombinant adenoviral vectors. AdCa18Luc, an1- and E3-deleted adenoviral vector containing the firefly luciferaseene [13] (provided by Dr. Frank Graham, McMaster University) isserotype 5, replication-deficient adenovirus. Preparation of viral

tocks was performed as described [14]. Briefly, 293 cells, a trans-ormed human embryonic kidney cell line that constitutively ex-resses adenovirus E1 proteins [15], was grown to ;80% confluencen 150-mm plates with Dulbecco’s modified Eagle’s medium (highlucose) containing 10% fetal calf serum and 1% penicillin/treptomycin. These cultures were infected with AdCa18Luc at aultiplicity of infection (m.o.i.) of 10 and incubated at 37°C in a 5%O2 atmosphere for 36–48 h. Cells and media were harvested,ashed with 20 mM Tris–HCl (pH 8.4), and chilled to 4°C prior toelleting and lysing of the cells by three freeze–thaw cycles on dryce. After centrifugation, supernatants were loaded onto a cesiumhloride gradient and placed in a Ti70.1 rotor (Beckman Instrumentsnc., Palo Alto, CA) followed by ultracentrifugation at 40,000 rpm for8 h at 4°C (L8-70M Ultracentrifuge, Beckman Instruments Inc.,alo Alto, CA). Bands were removed and desalted by dialysis at 4°Cgainst phosphate-buffered saline (PBS) with 10% glycerol. Samplesere aliquoted and frozen at 280°C.The concentration of virus preparations was determined both by

pectrophotometric absorbance at 260 nm [16] and by plaque assay14]. Plaque assays were performed in duplicate on 70–80% conflu-nt 293 cells in 35-mm plates. Viral dilutions were prepared in

LIPSHUTZ, FLEBBE-REHWALDT, AND GAENSLER: AD

MEM with 2% FBS and 1% penicillin/streptomycin and incubatedn cell monolayers for 2 h. Two percent agar at 42°C was overlaid ononolayers. Plates were incubated at 37°C with 5% CO2 and addi-

ional agar was added every 4 days. The final plaque count wasetermined on Day 14. Particle-to-plaque ratios were less than 100. a

Animal procedures. All procedures were approved by the Univer-ity of California, San Francisco (UCSF), Committee on Animal Re-earch. Nonpregnant and 13-day timed-pregnant CD-1 female miceere purchased from Charles River Breeding Laboratories (Wilming-

on, MA). Vaginal plug dates were recorded as Day 1 of gestation.On Day 15 of gestation, pregnant females were anesthetized by

ntraperitoneal injection of 600 ml of 2.5% tribomoethanol in tert-amyllcohol (Sigma–Aldrich Research, St. Louis, MO) diluted in phosphate-uffered saline. Midline laparotomy was performed and each horn ofhe gravid uterus was exposed. Borosilicate microinjection pipets wererepared with a heated-filament electrode puller (Sutter Instrument,an Rafael, CA), beveled, and attached to a microprocessor-controlledicroinjector (Harvard Instruments, South Natick, MA). Fetuses were

njected by a transuterine approach into the fetal liver. Following in-ection, the uterus was returned to the abdominal cavity and the ab-ominal wall closed in two layers with 4-O Dexon (Davis 1 Geck,ayne, NJ). Animals were kept under a warming lamp until awake

nd active, and then housed in an undisturbed environment untilarturition. Control animals were injected with normal saline or re-ained uninjected. Adenovirus stocks were diluted in normal saline

nd 1 3 107 plaque-forming units (PFU) of AdCa18Luc was injectednto the hepatic parenchyma in a volume of 5 ml; injected controlseceived 5 ml of normal saline by the same route. Adult CD-1 micenderwent intravenous lateral tail vein injection (200 ml) with either3 109 PFU of AdCa18Luc (n 5 5) or normal saline (n 5 5). All animalsere bled and sacrificed 5 days after injection.Assays for luciferase activity. Newborn animals were euthanizedfew hours after birth. Serum was obtained and animals were

issected and individual organs harvested. Tissues were placed inat-bottom 2.0-ml small-capped tubes (USA-Scientific, Ocala, FL)ith 200 ml of luciferase lysis buffer (Promega, Madison, WI) andlaced on ice. Adult animals were euthanized and tissues similarlyarvested; the brain, liver, and lungs were placed in 500 ml of lysisuffer and the intestine, heart, and spleen placed in 300 ml. Tissuesere homogenized with a hand-held homogenizer (Omni Interna-

ional, Marietta, GA). Specimens were centrifuged at 1400 rpmBrinkmann Instruments Inc., Westbury NY) for 5 min and placed at°C. Luciferase levels were determined in triplicate by luminometryAnalytical Luminescence Laboratory, Sparks, MD) and recorded aselative light units (RLU).Protein assay. Samples were centrifuged at 14,000 rpm for 5 min

nd dilutions prepared in 96-well plates. A Lowry-based proteinssay [17] was performed (Bio-Rad, Hercules, CA) and specimensere analyzed on a microplate reader (Molecular Dynamics, Sunny-ale, CA) at 750 nm. Results were pooled and mean and standardrror of the mean (SEM) calculated. Luciferase levels were normal-zed after the protein concentration of individual tissue extracts wasetermined and the luciferase activity plotted for each tissue as RLUer microgram of tissue protein.Determination of serum alanine aminotransferase (ALT). Bloodas drawn from the facial vein of newborn mice or from the retro-rbital plexus in adult mice. Samples were allowed to clot in sterileubes at room temperature. After 20 min, samples were centrifugedt 3000 rpm for 5 min. Serum was separated from packed red bloodells and analyzed on a Hitachi 747 Colorimetric Enzymatic Ana-yzer (IDEXX Veterinary Services, West Sacramento, CA). The re-ults were pooled and mean and SEM calculated.Histological examination. Right lobes of adult livers were placed

n 10% neutral-buffered formalin at room temperature. Tissues wererocessed in a series of increasing alcohol concentrations and placedn paraffin wax by standard techniques [14]. Ten-micrometer sec-ions were made and stained with hematoxylin and eosin.

151OVIRAL GENE EXPRESSION IN THE FETAL MOUSE

RESULTS

To optimize transuterine injection procedures andssess the accuracy of intrahepatic (IH) injections, day

1mcn

heF ;i lledm n.

1

FIG. 1. Technique for intrahepatic injection of fetal mice. (A) Tetuses at this age have a crown-to-rump length of 12 mm and weigh

njection needle is inserted into the hepatic parenchyma and is fiethylene blue dye is seen in fetal blood vessels, optic disc, and ski

52 JOURNAL OF SURGICAL RESEAR

5 murine fetuses were initially injected with 0.25%ethylene blue. Fetuses were viewed under magnifi-

ation (Fig. 1A) and injected through the left abdomi-al wall below the diaphragm (Fig. 1B). After injection

fvol

fetal mouse is transilluminated allowing the fetus to be oriented.300 mg. (B) Transillumination allows visualization of the liver. The

with 0.25% methylene blue. (C) Uninjected control; (D) injected

: VOL. 84, NO. 2, JUNE 15, 1999
CH
etuses were immediately sacrificed and dissected toerify the site of injection. All fetuses (12 of 12) dem-nstrated a single puncture site in the left superficialobe of the liver and hepatic staining. Occasionally,

ii1

ljcoien

I

ia[enahtsTRpiR

A

it1lbatfo

ftgAhAmgAs1rpemim

H

tsnapvRi(RtgtB

va

uc.a .

nadvertent intravenous injection occurred, resultingn the appearance of dye in vessels systemically (Figs.C and 1D).The efficiency and reproducibility of gene transfer fol-

owing IH injection of fetuses were then assessed by in-ection of AdCa18Luc, a recombinant adenoviral vectorontaining the luciferase gene. This vector has deletionsf the E1a, E1b, and E3 genes and insertion of a cassettencluding the human cytomegalovirus (CMV) immediatearly gene promoter, the luciferase cDNA, and polyade-ylation sequences from SV40 (Fig. 2).

ntravenous Injection in Adult Mice

To assess the activity of AdCa18Luc, intravenousnjection of the virus was performed in adult animalst doses successfully administered in other studies18]. Adult animals (26.5 6 3.37 g) were injected withither 1 3 109 PFU of AdCa18Luc (3.7 3 107 PFU/g) orormal saline. All animals (n 5 10) survived to 5 daysfter injection for analysis. Luciferase expression wasighest in the liver (660,050 6 429,105 RLU/mg pro-ein), lung (8313.7 6 2055.9 RLU/mg protein), andpleen (4063.4 6 1273 RLU/mg protein) (see Fig. 5B).he heart (299.9 RLU/mg protein), brain (230.2LU/mg protein), and intestine (84.5 6 41.1 RLU/mgrotein) demonstrated lower levels. Background levelsn normal saline-injected controls were 0.4 6 0.03LU/mg protein.

nimal Survival and Reproducibility of AdenoviralExpression

The tissue distribution and expression of luciferasen newborns were then determined following IH injec-ion at day 15 of gestation. Five microliters containing3 107 PFU of AdCa18Luc was injected into the fetal

iver (3.3 3 107 PFU/g tissue), a dose that is compara-le to the dose administered intravenously in adult

FIG. 2. Linear representation of adenoviral genome in AdCa18Llength of 10 kilobases (HCMV, human cytomegalovirus promoter)


nimals. Four pregnant CD-1 females were anesthe-ized at day 15 of gestation; the fetuses of three of theemales underwent transuterine intrahepatic injectionf 1 3 107 PFU of AdCa18Luc vector. The fetuses of the

l8R7

ourth female were injected with normal saline. All ofhe pregnant females recovered uneventfully from sur-ery; however, one of the females injected withdCa18Luc delivered prematurely and cannibalizeder pups. The two remaining females injected withdCa18Luc, as well as the saline-injected control fe-ale, proceeded to term and delivered on day 20 of

estation. All of the 22 pups injected in utero withdCa18Luc and 14 of the 15 pups injected with normalaline (93%) were liveborn. In other studies with day5 transuterine intrahepatic fetal injection of adenovi-us (n 5 25 pregnant females) or normal saline (n 5 19regnant females), we have not observed any differ-nce in the rate of preterm delivery, fetal survival, oraternal cannabilization of neonates between fetuses

njected with viral vectors and those injected with nor-al saline (data not shown).

epatic Luciferase Levels after in Utero Injection

Reproducibility of adenoviral delivery and transduc-ion of tissue following intrahepatic injection were as-essed by quantitating luciferase expression in theewborn liver at day of life 1. High-level luciferasectivity expressed as RLU per microgram of tissuerotein was detected in all 10 livers analyzed. Thealues obtained ranged from 14,156.07 to 48,113.16LU/mg protein (Fig. 3). The level of luciferase activity

n newborn animals (n 5 10) was highest in the liverFig. 4A). The mean level was 16,305 6 4826 (SEM)LU/mg protein. However, high levels of luciferase ac-

ivity were also detected in other intraabdominal or-ans such as the spleen (1355.3 6 405.1 RLU/mg pro-ein) and intestine (78.2 6 33.69 RLU/mg protein).ackground tissue levels averaged 0.1 RLU/mg protein.Interestingly, organs distant from the site of adeno-

iral injection also demonstrated luciferase activity inll of the animals injected in utero, albeit at lower

Boxes within the genome indicate areas of deletion. Bar designates

EN
evels. Luciferase expression in the lungs was 360.85 63.82 RLU/mg protein, in the heart was 211.35 6 63.54LU/mg protein, and in brain extracts was 32.18 6.45 RLU/mg protein. Since the diaphragm is readily

veicoapsotlHdb

dregcltperc

A

liiaai3best

(b(moliemo1j5econc

ftviataepirtttel

iilm

1 CH

isible by fetal transillumination, expression of lucif-rase in the heart and lung is probably not due tonadvertent injection of adenovirus into the thoracicavity. Rather, we have demonstrated with injectionsf methylene blue that intrahepatic injection is associ-ted with leakage of the injected material into theeritoneal cavity, as well as direct introduction into theystemic circulation (Fig. 1D). Thus, systemic deliveryf some fraction of the AdCa18Luc vector is likely to behe source of luciferase in tissues distant from theiver, especially in the brain (Fig. 4A, lanes 1 and 6).owever, the level of luciferase expression in theseistant tissues was 10 to 20-fold lower (spleen andrain, respectively) than in the liver.Adenoviral transduction of the liver after in utero IH

elivery of the vector results primarily in hepatic pa-enchymal gene expression. Using a recombinant ad-noviral vector containing the b-galactosidase reporterene, hepatic parenchymal cells are the predominantell type transduced after intrahepatic delivery; se-ected sections have demonstrated staining of ;40% ofhe hepatocytes. Staining of occasional hematopoietic

FIG. 3. Expression of luciferase in newborn livers, 5 days follow-ng intrahepatic injection of AdCa18Luc. Tissue extracts from 10ndividual fetuses are depicted (lanes 1–10). y axis indicates relativeight units/mg tissue protein; x axis represents individual newborn

ice.


recursors was also observed (data not shown). How-ver, a systematic documentation of areas of adenovi-al uptake and reporter gene expression has not beenompleted.

tltg

LT Levels and Evidence of Hepatic Toxicity

We determined whether intrahepatic adenoviral de-ivery in neonates was associated with the hepatotox-city that has been observed with adenoviral injectionsn adult animals. To compare the relative toxicity ofdenovirus in fetal mice with that in adult animals, 5dult female CD-1 mice (6 weeks of age, 26.5 g) werenjected with 1 3 109 PFU of AdCa18Luc (average.7 3 107 PFU/g) by lateral tail vein. This is compara-le to the dose of adenovirus (3.3 3 107 PFU/g) deliv-red intrahepatically to fetuses in utero. ALT levels inerum were analyzed 5 days after in utero or lateralail vein injection.

The serum ALT levels in uninjected newborn micen 5 5) were 28 6 0.94 IU/liter on day of life 1. New-orn mice previously injected with AdCa18Luc in uteron 5 22) had ALT levels of 26.5 6 5.0 IU/liter. Newborn

ice that received 5 ml of normal saline in utero dem-nstrated an ALT of 17 6 1.2 IU/liter on the first day ofife 5 days after injection. Thus, there was no elevationn ALT levels observed 5 days after in utero injection ofither AdCa18Luc or normal saline. In contrast, adultice (n 5 5) injected intravenously with 1 3 109 PFU

f AdCa18Luc demonstrated an average ALT of18.4 6 1.3 IU/liter, 5 days after injection. Mice in-ected with normal saline had serum ALT levels of1.2 6 8.5 IU/liter, 5 days after injection. These twofoldlevations in adult ALT levels were associated with aellular infiltrate observed on histological examinationf formalin-fixed tissues (Figs. 5A and B). Immunophe-otypic characterization of these infiltrating mononu-lear cells is presently underway.

DISCUSSION

In these studies we have developed a murine modelor transuterine fetal hepatic gene delivery. Followingransuterine fetal injections, there was 93–100% sur-ival and high-level hepatic reporter gene expressionn all liveborn animals. Pups appeared normal at birthnd showed no significant elevation of ALT levels, of-en associated with injection of adenoviral vectors indult animals. Despite the technical difficulties inher-nt in our in utero injection approach, high-level he-atic luciferase expression was reproducibly achievedn all newborn animals. A fraction of the AdCa18Lucecombinant adenovirus was also delivered to the sys-emic circulation, resulting in the transduction of dis-ant tissues and low-level expression of luciferase inhe brain, heart, and intestine. We have not as yetstablished whether expression in distant tissues fol-owing IH delivery of AdCa18Luc in utero is due to

: VOL. 84, NO. 2, JUNE 15, 1999

ransduction and expression of luciferase in the vascu-ar endothelium in these organs, or whether there isransduction of the tissue/parenchyma of distant or-ans by circulating recombinant adenovirus. If inad-

vtftva

iradgbmnt

(isiilt

aupsap

e al alb ht)

l

ertent systemic delivery and expression of a therapeu-ic gene were associated with pathological effects,uture strategies for in utero delivery could incorporateissue-specific promoter and enhancer elements intoector constructs and target gene expression to theppropriate tissues or cell types.Strategies for efficient gene delivery and expression

n the fetal liver and other tissues are important pre-equisites for successful in utero gene therapy-basedpproaches. These will be particularly important ineveloping approaches for ameliorating or correctingenetic disorders with prenatal and/or neonatal mor-idity or mortality. The immaturity of the fetal im-une system may allow prolonged expression fromonintegrating viral and nonviral gene therapy vec-ors. However, expression persisting to 2 weeks of life

FIG. 4. (A) Distribution of luciferase expression in newborn micuciferase expression at day 5 after intravenous injection in adult animackground levels in mock-injected control tissues (shown at the rig


FIG. 5. (A) Photomicrograph of adult mouse liver 5 days after intraveiver 5 days after intravenous injection of 1 3 109 pfu of AdCa18Luc (h

19 days after in utero injection) following in utero IHnjection of the Ad-b-galactosidase is similar to thateen when adult animals have received an intravenousnjection of the same vector. This decline in expressions most likely secondary to dilution of the episomallyocated viral genome as the animal grows, rather thano an immune response.

The transient high-level gene expression thatdenovirus-mediated gene delivery produces may beseful in providing expression of a protein that is tem-orally specific and abrogates the development of atructural abnormality. In addition, transient in uterond/or perinatal expression of a protein may delay theathological sequelae of a genetic disorder until otherherapeutic approaches may be implemented afterirth.

t day 1 of life after adenovirus delivery at E15. (B) Distribution ofs. Lanes A through F represent individual tissues and lane G depicts. Error bars represent SEM.

EN
tb

nous injection of normal saline. (B) Photomicrograph of adult mouseematoxylin and eosin, 2003 magnification).

boeevadhfvateoctpm

1

1

1

1

1

1

1

1

1

1 CH

Presently existing limitations for the use of recom-inant adenoviral vectors may be abrogated by the usef more recently constructed vectors in which all ad-noviral coding sequences have been deleted [19]. How-ver, first-generation adenoviral vectors produce toxiciral proteins that elicit a cellular immune responsend that ultimately results in the destruction of trans-uced cells. Furthermore, all human adenoviruses thatave undergone testing are able to oncogenically trans-orm rodent cells in vitro. Alternatively, other viralectors including adeno-associated virus, retroviruses,nd lentiviral constructs or nonviral DNA delivery sys-ems may provide more long-term therapeutic genexpression. Finally, in development of animal modelsf in utero gene therapy, the target disease must bearefully selected and issues regarding germ-line mu-ation and maternal–fetal risk must be thoroughly ex-lored prior to applying these approaches to the treat-ent of human disease.

REFERENCES

1. Brody, S. L., and Crystal, R. G. Adenovirus-mediated in vivogene transfer. Ann. NY Acad. Sci. 716: 90, 1994.

2. Herz, J., and Gerard, R. D. Adenovirus-mediated transfer of lowdensity lipoprotein receptor gene acutely accelerates cholesterolclearance in normal mice. Proc. Natl. Acad. Sci. USA 90: 2812,1992.

3. Kay, M. A., Landen, C. N., Rothenberg, S. R., Taylor, L. A.,Leland, F., Wiehle, S., Fang, B., Bellinger, D., Finegold, M.,Thompson, A. R., et al. In vivo hepatic gene therapy: Completealbeit transient correction of factor IX deficiency in hemophiliaB dogs. Proc. Natl. Acad. Sci. USA 91: 2353, 1994.

4. Haddada, H., Cordier, L., and Perricaudet, M. Gene therapyusing adenovirus vectors. Curr. Top. Microbiol. Immunol.199(Pt. 3): 297, 1995.

5. Zsengeller, Z. K., Wert, S. E., Hull, W. M., Hu, X., Yei, S.,Trapnell, B. C., and Whitsett, J. A. Persistence of replication-deficient adenovirus-mediated gene transfer in lungs ofimmune-deficient (nu/nu) mice. Hum. Gene Ther. 64: 457, 1995.

6. Trapnell, B. C., and Gorziglia, M. Gene therapy using adenovi-ral vectors. Curr. Opin. Biotechnol. 5: 617, 1994.


7. Stratford-Perricaudet, L. D., Levrero, M., Chasse, J.-F., Perri-caudet, M., and Briand, P. Evaluation of the transfer and ex-pression of an enzyme-encoding gene using a human adenovi-rus vector. Hum. Gene Ther. 1: 241, 1990.

1

8. McCray, P. B., Jr., Armstrong, K., Zabner, J., Miller, D. W.,Koretzky, G. A., Couture, L., Robillard, J. E., Smith, A. E., andWelsh, M. J. Adenoviral-mediated gene transfer to fetal pulmo-nary epithelia in vitro and in vivo. J. Clin. Invest. 95: 2620,1995.

9. Vincent, M. C., Trapnell, B. C., Baughman, R. P., Wert, S. E.,Whitsett, J. A., and Iwamoto, H. S. Adenovirus-mediated genetransfer to the respiratory tract of fetal sheep in utero. Hum.Gene Ther. 6: 1019, 1995.

0. Yang, Y., Nunes, F. A., Berencsi, K., Furth, E. E., Gonczol, E.,and Wilson, J. M. Cellular immunity to viral antigens limitsE1-deleted adenoviruses for gene therapy. Proc. Natl. Acad.Sci. USA 91: 4407, 1994.

1. Holzinger, A., Trapnell, B. C., Weaver, T. E., Whitsett, J. A.,and Iwamoto, H. S. Intra-amniotic administration of an adeno-viral vector for gene transfer to fetal sheep and mouse tissues.Pediatr. Res. 38: 844, 1995.

2. Douar, A.-M., Adebakin, S., Themis, M., Pavirani, A., Cook, T.,and Coutelle, C. Foetal gene delivery in mice by intra-amnioticadministration of retroviral producer cells and adenovirus.Gene Ther. 4: 883, 1997.

3. Addison, C. L., Hitt, M., Kunsken, D., and Graham, F. L.Comparison of the human versus murine cytomegalovirus im-mediate early gene promoters for transgene expression by ad-enoviral vectors. J. Gen. Virol. 78: 1653, 1997.

4. Hitt, M., Bett, A. J., Prevec, L., and Graham, F. L. Part 2:Viruses: Construction and Propagation of human adenovirusvectors. In J. E. Celis (Ed.), Cell Biology: A Laboratory Manual.San Diego: Academic Press, 1994. Pp. 479–490.

5. Graham, F. L., Smiley, J., Russell, W. C., and Nairn, R. Char-acteristics of a human cell line transformed by DNA from hu-man adenovirus type 5. J. Gen. Virol. 36: 59, 1977.

6. Chardonnet, Y., and Dales, S. Early events in the interaction ofadenovirus with HeLa cells: I. Penetration of type 5 and intra-cellular release of the DNA genome. Virology 40: 462, 1970.

7. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.Protein measurement with the Folin phenol reagent. J. Biol.Chem. 193: 265, 1951.

8. Wang, L., Zoppe, M., Hackeng, T. M., Griffin, J. H., Lee, K.-F.,and Verma, I. M. A factor IX-deficient mouse model for hemo-philia B gene therapy. Proc. Natl. Acad. Sci. USA 94: 11563,1997.

: VOL. 84, NO. 2, JUNE 15, 1999

9. Morsy, M. A., Gu, M., Motzel, S., Zhao, J., Lin, J., Su, Q., Allen,H., Franlin, L., Parks, R. J., Graham, F. L., et al. An adenoviralvector deleted for all viral coding sequences results in enhancedsafety and extended expression of a leptin transgene. Proc.Natl. Acad. Sci. USA 95: 7866, 1998.

adenovirus-mediated gene transfer in the midgestation fetal mouse

Documents