0022-538X/93/063644-05$02.00/0
CopyrightX)1993, American Society forMicrobiology
An
Immortalized Human Fibroblast
Cell Line Is
Permissive
for Human
Cytomegalovirus Infection
TERESA COMPTON
Departmentof MedicalMicrobiologyandImmunology, Universityof
Wisconsin-MadisonMedical
School, Madison,
Wisconsin
53706-1532Received 21 December1992/Accepted21March1993
Human foreskin fibroblasts (HFF)wereimmortalized via retrovirus-mediatedgenetransferofthe E6andE7
genes ofhuman papillomavirus type16. An immortalized fibroblast (IF)cell linewhichwasmorphologically
akin to the parental cell line was isolated. The IF cell line was evaluated for permissiveness to human
cytomegalovirus (HCMV) infection after the IF cell linesurpassed the normalpassage limitation ofdiploid
fibroblasts. Western immunoblot analysisofrepresentative HCMV-encodedimmediate-early (72-kDa),early
(gB), andlate(gH)geneproductsdemonstrated that the IF cell lineproducedtheseproteinsanalogoustothose
produced bytheparental 11FFcells. SimilarquantitiesofinfectiousviruswereproducedintheIFand 11FFcell
lines as determined in one-step growth curve experiments. Compared with the HFFcells, morphologically
identical plaqueswereproducedinthe IFcellline inapproximately10to12days postinfection.Thesefindings indicate that fibroblast cell lines immortalized with transforming genes of human papillomavirus retain
complete permissivenesstoHCMVinfection and supportplaqueformation. TheIFcelllinewillbe usefulfor
future genetic analysisofHCMV.
Human cytomegalovirus (HCMV) is a significant human
pathogen responsible for severe clinical manifestations in
immunosuppressed individuals (1, 15, 17, 25, 29, 30).
Mo-lecular characterization of HCMV has proven challenging becauseof itslonglifecycleand limited hostrangeinvitro.
Secondary culturesof human diploid fibroblasts from
vari-ousoriginsarepredominatelyusedtopropagateHCMV,and these cell lines are uniquely capable ofproducing plaques.
Other celllines, such as humanbrain endothelial cells (24, 28),human umbilical vein endothelial cells(14, 31,33, 38),a
teratocarcinoma cell line (23), cultures of human
macro-phagesandmonocytic cell lines(16, 39), andhuman
astro-cytomacelllines(2, 21),arealso knowntobepermissivefor theproductionofHCMV; however,noneofthese cells lines
fullysupport plaque formation.Apermanently growingcell line which is permissive for productive HCMV infection
andsupportsplaqueformation wouldbeanextremely useful
and convenient tool to study HCMV infections in vitro. It
hasbeen demonstrated thattheE6 andE7genesofhuman
papillomavirus (HPV)type 16 have the capacity to immor-talize primary human keratinocytes, the permissive host
for HPV infections (13). In the present study, the E6 and
E7 genes of HPV type 16 along with the neomycin
resis-tance marker were stably introduced into the genome of
human foreskinfibroblasts (HFF)andanimmortalized
fibro-blast (IF) cell line was isolated. The resultant cell line is
thus far immortal in its growth properties and remains permissive to productive HCMV infection and plaque
for-mation.
HFF cells were purchased from Clonetics, San Diego,
Calif., and propagated as previously described (8, 27). An
amphotropicretrovirus-packaging cell line which produces the recombinant virus LXSN16E6E7 as described by
Hal-bert et al. (13) was provided by Denise Galloway (Fred
Hutchinson Cancer Research Center, Seattle, Wash.). The
genomeofthisvirus containsaneomycinresistance marker,
theE6 andE7genesoftype16 HPV flanked bythe Moloney murine leukemia virus long terminal repeats. The E6-E7
recombinant stock produced by thepackagingcell linewas
usedtoinfect 50%confluentmonolayersofpassage11HFF cells. Briefly, 1 ml of virus stock plus 3 ml of complete medium (Dulbecco modified Eagle medium plus 10% fetal bovine serum) containing 4 ,ug of Polybrene per ml was
addedtothe monolayersfor 2 h and further incubatedwith additionalPolybrene-containing medium for 5 h.Fresh
com-plete mediumwasadded overnight. On day2,thecellswere
tyrpsinized and plated in limiting dilutions. On day 3, the
mediumwas supplemented with 200 ,ug ofG418 per ml to
select neomycin-resistantcolonies. A kill curve analysis of
the HFF cells indicated that 200 ,ug of G418 per ml was
appropriatefortheHFF cell line(datanotshown). Approx-imately 2 weeks later, individual neomycin-resistant colo-nieswereevident and 12cloneswereselectedandexpanded.
In thestudy shown here, asingle neomycin-resistant clone
thatwasmorphologicallyverysimilartotheparentalcellline
waschosenfor furtheranalysis. The IF cell linewas
main-tained under selectivepressureand has beenpassagedupto
50 times, whereas normal HFF cells reach senesence
be-tweenpassages23 and28.
One-step growthcurveanalysiswasperformedtoevaluate
the IF cell line forpermissivenessto HCMV infection. The
AD169 strain of HCMV was used for these studies and
propagated in HFFcellsaspreviouslydescribed (8, 27).For
thesingle-cycle growth analysis, 10-cm2 culture dishes of IF (passage 40)orHFF(passage 15) cellswereinfected with104
PFU of HCMVperdish (multiplicity of infection [MOI] = 0.1). Following a 90-min adsorption period, the inoculum
wasremovedandmediumcontaining 2% fetal bovineserum
(maintenancemedium) wasaddedto eachplate. The plates
wereincubatedat37°C andharvestedondays 2, 5, 7, 9, and
13 postinfection. On the indicated day postinfection, one
dish from each cell line was collected for plaque titration
analysis. Plaque titrations were performed as described
previously (8). As shown in Fig. 1, virusproduction peaked
in both the HFF and IF cell lines approximately 7 days postinfection.Theyieldofinfectious viruswasgreaterthan
3644
on November 9, 2019 by guest
http://jvi.asm.org/
_~~
HFF SIF+HCMV Mo +HCMV Mo
Days Post-infection:
Immediate Early (72 kDa)
0 3 6 9 12 15
Days post-infection
FIG. 1. One-step growthcurveanalysis of HCMV production in HFF and IF cell lines. HFForIF cellswereinfected with HCMV (AD169) at an MOI of 0.1. After adsorption for 90 min at 37°C, monolayers werewashed with serum-free medium and incubated with maintenance medium. Viruswasharvestedasindicated inthe text,and titersweredetermined by plaque assays onHFF
mono-layers.
10-fold higher from the IF cells in the results of the experi-ment presented here. In repeat experiments, the yield of AD169wasgenerally higher from the IF cells by 10-fold than
that from the HFF cells. It isnotknownwhether this slight
difference is strainspecificorbiologically significant. Highly
similargrowthcurves were also obtained whenonly extra-cellularvirus was recovered, although the titers were
ap-proximately100-fold lower(datanot shown).
HCMVgeneexpressioninpermissive fibroblasts is char-acterizedby sequential expressionof the viralgenomethat is
temporally regulated anddesignated into three classes:
im-mediate-early(IE), early,and late(34, 36).Thesynthesisof the major IEprotein, a 72-kDa protein, andtwo glycopro-teins, gB and gH, was examined in HFF and IF cells at
various times postinfection byWestern immunoblot
analy-sis. The major IE gene product, a 72-kDa protein, was
detected witha polyclonal antiserum raised against
recom-binant-derived, purified72-kDaprotein (26).Thisantiserum,
which recognizes both the 72-kDa (IE1) and 86-kDa (IE2) proteins, was agenerous gift fromJay A. Nelson, Oregon
Health SciencesUniversity,Portland. HFForIFcellswere
cultured in 35-mm-diameter dishes and infected with HCMV
atanMOIof 3. On the indicatedday postinfection, the cells were harvested by scrapingwith a rubber policeman,
pel-leted by centrifugation, washed with phosphate-buffered
saline (PBS), and suspended in glass-distilled water. After
disruption by sonication, the cell extract was treated with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer and resolved in triplicate on
SDS-7.5% PAGEgelsfollowedbytransfer to nitrocellulose
membranes essentially asdescribed elsewhere (6).The effi-ciencyof transferwasmonitoredbyPonceau Sstaining.The
nitrocellulose membraneswere blocked in 10 mM Tris(pH
7.4)-150
mM NaCl plus 5% dried milk, 1% bovine serumalbumin, and 0.1% Tween 20 for 60 min at 37°C and then
subjected to three successive washes in 10 mM Tris (pH
7.4)-150mM NaCl-1%dried milk-0.1% Tween 20(WB) or
WB supplemented with 0.1% SDS (for polyclonal
antibod-ies).To detect theIEprotein,theanti-IEserum wasdiluted
Early (gB)
Late
(gH)
[image:2.612.65.304.69.251.2]1 3 5 7 9 9 1 3 5 7 9 9
FIG. 2. Synthesis of the major IE protein, gB, and gH as a function of time in HFForIFcells. HFForIFcellswereinfected withHCMV (AD169)atanMOI of 3. After adsorption for 90minat
37°C, the inoculum was removed and maintenance medium was added.Ontheindicatedday postinfection, the cellswereharvested, lysed, treated with SDS-PAGE sample buffer which contained 2-1-mercaptoethanol, resolved in triplicate by SDS-PAGE (7.5% acrylamide), and transferred to nitrocellulose asdescribed in the
text. The 72-kDa IE protein, gB, and gH were detected with monospecificantibodies followedby horseradish perosidase-conju-gated anti-species second antibody and detected with chemilumines-cence asdescribed in thetext.
inWBplus 0.1% SDS and incubated with the membrane for
60 min at room temperature. After the membrane was
washed, the primary antibodywasdetected with horseradish
perioxidase-conjugated goat anti-rabbit serum (Pierce).
Chemiluminesence (ECL; Amersham) was used to detect the antibody conjugates as describedby the manufacturer,
and the membraneswereexposedtoX-OMAT film(Kodak), generally for 10-30 s. As shown in Fig. 2, the major IE
proteinwasevident at 1 day postinfection inboth the HFF and IF cells. Thesteady-statelevels of the 72-kDaIE protein
in HFF cells declined after 3 days postinfection, afindingin
agreementwith that ofStenbergetal. (35).Incontrast, the levels of the IE protein remained fairly constant inthe IF
cells until 9days postinfection.
GlycoproteinBwaschosen as ageneral representativeof
anearlygeneproduct sincesynthesisofgBis detectable in
the presence of inhibitors ofDNAreplication (5). The gB glycoprotein is contained within thegC-I envelope
glycop-roteincomplexand is also referredtoasgpl30/55 (3, 9, 11,
12, 20). A monoclonal antibody (27-78) specific for glyco-proteinBwaskindly provided byWilliam J.Britt, University
ofAlabama, Birmingham. Purified 27-78immunoglobulinG
109
108
i07 106
105
104
103
102
*1141*
-72 kDa
Igpl30/116
-gp55
-gp86
HFF EF
on November 9, 2019 by guest
http://jvi.asm.org/
[image:2.612.326.566.74.373.2]wasprepared from ascitic fluidsbyprotein A chromatogra-phy. As described above, lysates of HFF or IF cellswere
transferredto nitrocellulose membranes and thegB protein
was detected with 27-78 immunoglobulin G (0.25 ,g/ml)
diluted in WB followed by horseradish peroxidase-conju-gated goat anti-mouse serum and with chemiluminesence.
Synthesis of the gB protein was detectable on day 3 and
remained evident in both cell lines forthe duration of the
experiment (Fig. 2). Inaddition, processing andproteolytic cleavage of the gB biosynthesis products appeared to be
identicalin the IFcells and HFF cells.
Synthesis of glycoprotein H (gH) is not evident in the absence of viral DNAsynthesisand is thereforea
represen-tative of lategeneproducts (5).To detect gH,apolyclonal
antibody (no. 6824)wasgeneratedin arabbit after
immuni-zation withtwopeptidesderived from thegHsequence(10). Briefly, two peptides consisting of C-SYRSFQQLKAQD SLGQQPTT and C-SDLYTPCSSSSGRRDHSLERLTR
(amino acids 135 to 155 and 517 to 538, respectively)were
synthesized (kindly providedbyJohnson andJohnson,San
Diego, Calif.)andcoupledtoasinglekeyhole limpet
hemo-cyanin carrier by using a heterobifunctional cross-linking
reagent (sulfo-SMCC; Pierce). The peptide conjugate (500 ,ug)was injectedinto aNew Zealand Whiterabbit, and the animalwasboosted every4weeks. Ahigh-titer polyclonal
antiserumwasobtained. Theno.6824 antiserumrecognizes
the gp86 component of the gC-III complex (12) which containsthegHhomologof HCMV(10).To confirmthat this
reagentwasspecific for gH,amonoclonalantibody specific
forgH(14-4b)wasalsoprovided byWilliam J. Britt and used
to immunoprecipitate gH from infected cell extracts. The
immunoprecipitate was then transferred to nitrocellulose andpositivelydetected withtheno.6824antibody (datanot shown). The anti-gH serum was diluted in WB plus 0.1% SDS and detected as described above. In both cell lines, glycoprotein H was not readily detectable until 5 days postinfection and remained evident through 9 days
postin-fection(Fig. 2). Ingeneral, the kinetics ofexpressionofIE, early, and lategene products wasvery similar in both cell lines.
While the kinetics of detection of the three proteins
studied here were similar, levels of all three HCMVgene
productsappearedslightly higherin the IFcells. In particu-lar,the IEproteinwasdetectable until 9days postinfection,
while levels ofIE proteindeclined inthe HFFcellline after 3days postinfection.Todeterminewhether the IFcellswere more efficiently infected by HCMV, immunofluorescence
analysis was performed to examine the number of
IE-positivecellsas afunction oftime. Inthisexperiment,HFF
or IFcells cultured on glass coverslipswere infected with
HCMV (MOI = 3) for various periods of time. At the
indicated time, the cellswere fixed with 3%
paraformalde-hyde inPBS for20 min and then permeabilized with 0.1%
TritonX-100for 10min. The fixed monolayerswerewashed
withPBS plus 0.2% gelatin, and the cellswerereactedwith
the anti-IE antibody described above for 60 min at room temperature. Thecoverslipswere then washed three times
for 5 min each time with PBS-gelatin, and the primary antibodywasdetectedwithrhodamine-conjugatedgoat anti-rabbitserum(Kirkegaard&PerryLaboratories, Inc.,
Gaith-ersburg, Md.) for 60 min at room temperature. The wash steps were repeated, and the coverslips were mounted in
Gelvatol as described previously (7). The percentage of IE-positive cellswas determined, andthe data are summa-rized in Table 1. As shown in Table 1, the number of
IE-positivecellswasnotsignificantlygreaterintheIF cells
TABLE 1. Number ofIE-positiveHFForIFcells
as afunction of time
% of cellspositivefor IEproteinattheindicated
Cell time postinfectiona
line
18 h 36 h 72 h
HFF 66.1 95 33
IF 70.3 100 47
a Cells were fixed and stained at the indicated hour postinfection. A
minimum of 100IE-positive nucleiwerecounted in each sample.
than in theHFF cells. This result indicates that the IF cell line is not more efficiently infected than the parental fibro-blasts. These data and the results of the Western blot analysis (Fig. 2) suggest that synthesis of the major IE protein is sustainedlonger in the IF cells than in HFF cells, asituation which may be due to the presence of the HPV E6 andE7proteins.
To determine whether the IF cell line could supportplaque formation, HFF or IF cells cultured in 35-mm-diameter dishes were infected with 100 PFU of HCMV and incubated asdescribed above. Onday 12postinfection,themonolayers
werefixed with25% formalin and stained with0.05% meth-ylene blue. As shown inFig. 3, uninfectedmonolayers of IF and HFF cellswerevery similar in appearance. In infected
monolayers of IF and HFF cells, morphologically similar plaques were evident in the IF cells. Although the virus stock was grown in HFF cells,-comparable numbers of plaques were obtained in the IF cell line. Infectious virus
wasrecovered fromasingle plaque producedinthe IFcells,
as evidenced bysuccessfulreinfection of either HFForIF monolayers (data not shown). This finding suggests that
substantial adaptationto the IF cell line isnot requiredfor furthergrowth andpropagation.
The results of this study indicate thatdiploidcultures of HFFimmortalized with thetransforminggenesof HPV type 16retain theabilitytosupportproductive HCMV infection, including plaque formation. These data also demonstrate
thatthe E6 and E7 genes have thecapacitytoimmortalizea
celltype-fibroblasts-otherthan thepermissivecell
type-epithelial cells-for HPV infection. One-step growthcurve
experimentsanddetection of HCMV-encodedproteins
dem-onstrate that there were no significant differences in the general growth properties of HCMV in the IF cells. The
synthesisofthemajor IEprotein appeared tobe sustained longer in the IF cells, which likely explains the slightly
highertiters of virusproduced in these cells.
Herpesviruses,including HCMV,areamenable to genetic
manipulationviahomologous recombination (32). Strainsof viruscontaining whole-gene deletions or specific mutations
areextremely useful for functional analysis of the proteins encoded bythese genes. There are now several reports of the construction and production of deletion strains of
HCMV; however,in eachinstance, nonessential regions of the HCMV genome have been removed (4, 18, 19, 22, 37). The lack ofapermanently growing cell line which supports plaque formation has prevented the production of stains of virus with deletions in essential gene products, since a stably
transfected, complementing cell line which provides the gene product to be removed in trans is required. The
developmentofanimmortalized cell line which is permissive for HCMV infection and plaque formation represents a
major technical advancement that will be useful for future
genetic analysisofHCMV.
on November 9, 2019 by guest
http://jvi.asm.org/
MOCK
+HCMV
HFF
pll
IF
p41
FIG. 3. Morphological comparison of uninfectedandinfectedHFF and IFcell monolayers. HFF (passage 11; pll) or IF cells (passage 41;
p4l)wereeither mock infectedorinfectedwith100PFUofHCMV.Afteradsorption for 90 min at 37°C, the monolayers were overlaid with
agarose-containing medium and furtherincubated at37°Cfor12days. After fixationwith25%buffered formalin,the cell monolayers were
visualizedbystaining with0.05%methylene blue, observed, and photographed on a Televol inverted microscope. Magnification, x200.
I acknowledge Neil Cooper and Bonnie Bradt of The Scripps ResearchInstitute,LaJolla, Calif.,forassistancein theproduction
of theanti-gHantiserum.Sherry Colemanprovided experttechnical assistance. I thank my colleagues Curtis Brandt and Miroslav
Malkovskyin theDepartmentof MedicalMicrobiology and
Immu-nology, University of Wisconsin, forcriticalreadingof the
manu-script.
These studieswereaidedby grants IRG-35-341 from the Ameri-can Cancer Society and U01A131494 from the Midwest STD Re-searchCenter.
REFERENCES
1. Alford, C. A., and W. J. Britt. 1985. Cytomegalovirus, p.
629-660.In B.N. Fields, D. M. Knipe, R. M.Chanock,J. L.
Melnick,B. Roizman,and R. E. Shope (ed.),Virology. Raven Press, New York.
2. Boeckh,M.,C. A.Gleaves,R.Bindra,andJ. D.Meyers. 1991.
Comparison of MRC-5 and U-373MG astrocytoma cells for
detection of cytomegalovirus in shell vial centrifugation cul-tures. Eur. J.Clin.Microbiol. Infect. Dis. 10:569-572.
3. Britt,W.J., and L. G.Vugler. 1989.Processingof thegp55-116 envelope glycoproteincomplex (gB) of humancytomegalovirus.
J. Virol. 63:403-410.
4. Browne, H.,M.Churcher,and T. Minson. 1992. Construction
andcharacterization ofahuman cytomegalovirusmutantwith theUL18(classIhomolog)gene deleted. J.Virol. 66:6784-6787. 5. Compton,T. Unpublishedresults.
6. Compton, T.,and R.J.Courtney.1984.Virus-specific
glycopro-teins associated with the nuclear fraction of herpes simplex virus type 1-infected cells. J. Virol. 49:594-597.
7. Compton, T.,I. E.Ivanov,T.Gottlieb,M.Rindler,M.Adesnik,
andD. D. Sabatini. 1989. Asorting signal for the basolateral
deliveryof the vesicular stomatitis virus(VSV)Gproteinlies in its luminal domain:analysisof thetargetingofVSVG-influenza
-N
-f--.
Poe
I "I
1, ,i
.4, I.: Jf"* -.,.
"N.S4.1
1,xv.A-qw,VF -tAt
19as
7.0
... L'A'. -A
!.
105:7
140.n-
4on November 9, 2019 by guest
http://jvi.asm.org/
[image:4.612.62.569.75.510.2]hemagglutinin chimeras. Proc. Natl. Acad.Sci. USA 86:4112-4116.
8. Compton, T., R. R. Nepomuceno, and D. M. Nowlin. 1992.
Human cytomegalovirus penetrates host cells by pH-indepen-dent fusionatthecellsurface. Virology 191:387-395.
9. Cranage,M.P., T.Kouzarides, A. T. Bankier, S.Satchwell,K. Weston, P. Tomlinson, B. Barrell, H. Hart, S. E. Bell, A. C. Minson, and G. L. Smith. 1986. Identification of the human cytomegalovirus glycoproteinBgeneandinductionof
neutral-izing antibodies via its expression in recombinant vaccinia
virus.EMBO J. 5:3057-3063.
10. Cranage, M. P., G. L. Smith, S. E. Bell, H. Hart, C. Brown,
A.T.Bankier, P. Tomlinson, B. G.Barrell, and T.C.Minson.
1988. Identification andexpression ofahumancytomegalovirus glycoprotein with homology totheEpstein-Barr virusBXLF2
product, varicella-zoster virus gpIII,andherpessimplex virus
type 1glycoprotein H. J. Virol. 62:1416-1422.
11. Gretch,D.R.,R. C.Gehrz, and M. F.Stinski. 1988. Character-izationofahumancytomegalovirusglycoprotein complex(gcI).
J. Gen. Virol.69:1205-1210.
12. Gretch, D. R., B. Kari, L. Rasmussen, R. C.Gehrz, and M. F. Stinski. 1988. Identification and characterizationof three
dis-tinct families ofglycoprotein complexes in the envelopes of
humancytomegalovirus. J. Virol.62:875-881.
13. Halbert, C. L., G. W. Demars, and D. A. Galloway. 1991. The E7 gene of human papillomavirus type 16 is sufficient for
immortalization of human epithelial cells.J.Virol. 65:473-478. 14. Ho, D.D., T. R. Rota,C.A. Andrews, and M. S. Hirsch. 1984.
Replication ofhuman cytomegalovirus in endothelial cells. J.
Infect. Dis. 150:956-957.
15. Hutter, J. A., J. Scott, T. Wreghitt, T. Higenbottam, and J.
Wallwork. 1989. Theimportance of cytomegalovirus disease in heartlungtransplant recipients. Chest 95:627-631.
16. Ibanez, C. E., R. Schrier, P.Ghazal, C. Wiley, and J. A. Nelson.
1991. Human cytomegalovirus productively infects primary differentiatedmacrophages.J. Virol.65:6581-6588.
17. Jacobson, M. A., and J. Mills. 1988. Serious cytomegalovirus disease in the acquired immunodeficiency syndrome (AIDS).
Ann. Intern. Med.108:585-594.
18. Jones, T. R., and V. P. Muzithras.1992.Acluster ofdispensable
geneswithin thehumancytomegalovirusgenomeshort compo-nent:IRS1, US1 through US5, and the US6 family. J. Virol. 66:2541-2546.
19. Jones, T. R., V. P. Muzithras, and Y. Gluzman. 1991. Replace-mentmutagenesis of thehuman cytomegalovirusgenome:US10
and US11 gene products are nonessential. J. Virol.
65:5860-5872.
20. Kari, B., Y. N. Liu, R. Goertz, N. Lussenhop, M. F. Stinski, and R.Gehrz. 1990.Structureandcompositionof afamilyofhuman
cytomegalovirus glycoprotein complexes designatedgC-I(gB).
J. Gen. Virol. 71:2673-2680.
21. Kari, B., R. Radeke, and R. Gehrz. 1992. Processing of human
cytomegalovirus envelope glycoproteins in and egress of
cyto-megalovirus from human astrocytoma cells. J. Gen. Virol. 73:253-260.
22. Kaye, J., H. Browne, M. Stoffel, and T. Minson. 1992. The UL16 geneofhumancytomegalovirusencodes aglycoproteinthat is dispensableforgrowth in vitro. J. Virol.66:6609-6615. 23. LaFemina, R., and G. S. Hayward. 1986. Constitutive and
retinoicacid-inducible expression ofcytomegalovirus
immedi-ate-early genes in human teratocarcinoma cells. J. Virol. 58: 434-440.
24. Lathey, J. L.,C. A. Wiley, M. A.Verity,andJ. A. Nelson. 1990. Cultured human braincapillaryendothelial cells arepermissive
for infectionbyhumancytomegalovirus.Virology 176:266-273.
25. Meyers, J. D., N. Flournoy, and E. D. Thomas. 1986. Risk
factors forcytomegalovirusinfection afterhuman marrow
trans-plantation.J.Infect. Dis. 153:478-488. 26. Nelson, J. A.Unpublishedresults.
27. Nowlin, D. M., N. R.Cooper,andT.Compton. 1991. Expression
of ahumancytomegalovirus receptorcorrelates with
infectibil-ityofcells. J. Virol.65:3114-3121.
28. Poland, S. D., P. Costello,G. A.Dekaban, and G. P. Rice. 1990.
Cytomegalovirus in the brain: in vitro infection of human
brain-derivedcells. J.Infect. Dis. 162:1252-1262.
29. Rubin,R.H.,and R. B.Colvin. 1986.Cytomegalovirusinfection in renal transplantation: clinical importance and control, p.
283-304. In G. M.Williams,J. F.Burdick,and K. Solez(ed.), Kidney transplant rejection: diagnosis and treatment. Dekker
Press,New York.
30. Singh, N., J. S.Dummer, M.Ho, etal. 1988. Infectionswith
cytomegalovirus and otherherpesviruses in 121 liver transplant recipients: transmission by donated organ and the effects of
OKT3 antibodies. J. Infect. Dis. 158:124-131.
31. Smiley,M.L., E.C.Mar, and E.S.Huang.1988. Cytomegalo-virus infection and viral induced transformation of human endothelial cells. J. Med. Virol. 25:213-226.
32. Spaete,R.R., and E.S. Mocarski.1987. Insertion anddeletion
mutagenesis of the humancytomegalovirusgenome.Proc.Natl.
Acad. Sci.USA 84:7213-7217.
33. Span,A. H., W.Mullers, A. M.Miltenburg,andC.A.
Brugge-man. 1991. Cytomegalovirus inducedPMN adherencein
rela-tion to anELAM-1antigenpresentoninfected endothelialcell
monolayers. Immunology 72:355-360.
34. Spector, D. H.,K. M.Klucher, D. K. Rabert,and D. A.Wright. 1990. Human cytomegalovirus early gene expression. Curr. Top. Microbiol. Immunol. 154:21-45.
35. Stenberg,R.M.,A.S.Depto,J.Fortney,andJ. A. Nelson. 1989.
Regulated expression ofearly and late RNAs and proteins from the human cytomegalovirus immediate-early gene region. J.
Virol.63:2699-2708.
36. Stinski, M. F. 1990. Cytomegalovirus and its replication, p.
1959-1980.InB. N.Fields, D.M.Knipe, R. M.Chanock,J. L.
Melnick, B.Roizman,and R. E.Shope (ed.), Virology. Raven
Press,New York.
37. Takekoshi, M.,F. Maeda-Takekoshi, S. Ihara, S.Sakuma, and Y. Watanabe. 1991.Site-specific stable insertion into the human cytomegalovirusgenomeofaforeigngeneunder control ofthe SV40promoter.Gene 101:209-213.
38. Waldman, W. J., W. H. Roberts, D. H. Davis, M. V. Williams, D. D.Sedmak,and R. E.Stephens.1991.Preservationof natural endothelial cytopathogenicityofcytomegalovirus by propaga-tion in endothelial cells. Arch. Virol. 117:143-164.
39. Weinshenker,B.G.,S.Wilton,andG. P. A. Rice. 1988.Phorbol ester-induced differentiation permits productive human
cyto-megalovirus infection in a monocytic cell line. J. Immunol.
140:1625-1631.